Where Democrats Disagree: Citizens’ Normative Conceptions of Democracy

715398 research-article2017 PSX0010.1177/0032321717715398Political StudiesLandwehr and Steiner Article Political Studies Where Democrats Disagree: 1­–19 © The Author(s) 2017 Reprints and permissions: Citizens’ Normative sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0032321717715398 https://doi.org/10.1177/0032321717715398 Conceptions of Democracy journals.sagepub.com/home/psx Claudia Landwehr and Nils D Steiner Abstract While support for the essential norms of liberal electoral democracy is high in almost all developed democracies, there is arguably also a gap between democratic aspirations and democratic practice, leading to dissatisfaction among citizens. We argue that citizens may hold very different normative conceptions of democracy which are equally compatible with support for liberal democracy, but lead to different expectations where institutional design and democratic practice are concerned. Satisfaction with democracy may thus depend on congruence between such normative conceptions and institutionally entrenched norms. Drawing on survey data from Germany with a comprehensive item battery on attitudes towards democratic decision-making, we identify four distinct factors leading to disagreements over democratic decision-making. We explore how these are related to personality, styles of cognition and political attitudes, and show that different expectations arise from them, such that regime support is affected by the normative conception(s) of democratic decision-making individuals subscribe to. Keywords political support, process preferences, attitudes to democracy, citizens, Germany Accepted: 29 March 2017 Introduction Shrinking turnout rates, political disenchantment and the success of populist parties in many European countries give rise to concerns about the state and future of democracy. While support for the idea of liberal electoral democracy is still very high in established democracies, there exists what Pippa Norris terms a ‘democratic deficit’ between demo- cratic ideals and democratic practice (Norris, 2011). Citizens’ everyday experiences and perceptions of politics often seem to be characterised by disappointments, resulting in a lack of trust in professional politicians and the institutions of representative democracy. Department of Political Science, Johannes Gutenberg Universität Mainz, Mainz, Germany Corresponding author: Claudia Landwehr, Department of Political Science, Johannes Gutenberg Universität Mainz, Jakob Welder Weg 12, Mainz 55099, Germany. Email: [email protected] 2 Political Studies 00(0) The origins of this disappointment remain controversial in the literature: Norris has argued that increasingly critical citizens demand more and more effective opportunities for politi- cal participation (Norris, 1999). Dalton’s ‘new politics’-hypothesis points in a similar direction: for a younger generation with post-materialist attitudes, representative democ- racy may simply not be enough democracy (Dalton, 1999). Hibbing and Theiss-Morse have countered these arguments with a diagnosis of ‘stealth democracy’, according to which many citizens have a deep dislike of political conflict and, having no interest in participation themselves, would prefer political decisions to be taken by unelected experts and professional business people (Hibbing and Theiss-Morse, 2002). More recently, numerous authors have pointed out how socio-economically disadvantaged groups have increasingly withdrawn from political participation, thus reducing incentives for politi- cians to respond to their interests and preferences (e.g. Bartels, 2009; Schäfer, 2013). According to this view, the ‘bottom third’ of society that is increasingly left behind is not reached by offers to participate on the input side of democracy and is understandably frus- trated with policy outputs that do anything but alleviate their exclusion. This group may be particularly vulnerable to populist mobilisation, which, given the recent electoral success of right-wing populists in many countries, may be the most significant contemporary con- cern. Populist parties and politicians also appeal to many middle-class citizens, though, and several researchers have identified a specific set of attitudes motivating support for them (Akkerman et al., 2013; Spruyt et al., 2016). However, these diagnoses are not mutually exclusive. Critical citizens may well exist besides stealth democrats, populists and a bottom third of society that is excluded both socio-economically and politically. If many citizens are dissatisfied with democracy’s performance, this may be for entirely different reasons: Citizens are likely to expect dif- ferent things from democracy and detect different shortcomings. In this article, we are interested in this very sphere of reasonable disagreement between democrats, that is, between people who in principle agree on the principles of polyarchy and the rule of law – a sphere of disagreement that in our eyes remains underexplored in the existing literature. We thus assume that citizens even and particularly in pluralist and consolidated democ- racies have differing normative conceptions of democracy from which different demo- cratic aspirations follow. While there is apparently a strong consensus where a conceptual core of liberal electoral democracy is concerned (Ferrin and Kriesi, 2016), a second layer of more specific normative orientations towards democratic decision-making is clasped around this core. This second layer of normative conceptions of democracy is more con- troversial than the core and may differ between societal groups and individuals. It consti- tutes a sphere of democratic disagreement, a sphere in which people who support the fundamental norms of liberal electoral democracy have different normative ideas about democracy’s input side. It is one of the aspects of collective life on which democrats disagree. To back up this claim, we conduct a series of analyses with data from a survey that was fielded in the German GESIS Panel (GESIS, 2017) in 2015. Our questionnaire uses a longish list of controversial statements about democracy with which good democrats may just as well agree or disagree to differing extents. The motivation behind our question- naire is to study citizens’ normative orientations towards democratic decision-making procedures more comprehensively, in order to inductively identify latent factors of disa- greement and study their determinants and consequences for political support. We discuss our motivation against the background of existing research on citizens’ orientations Landwehr and Steiner 3 towards democracy in more detail in section ‘Where Democrats Agree and Disagree’. In section ‘Four Conceptions of Democratic Decision-Making’, we present the methodol- ogy and the results of the dimensional analysis. By way of explorative factor analyses, we derive four distinct latent factors that structure attitudes towards democratic decision- making, thus pointing towards differing normative conceptions of democracy. We label these factors (1) trustee conception of democracy, (2) anti-pluralist scepticism, (3) delib- erative proceduralism and (4) populist majoritarianism. We go on to identify determinants of these four factors in section ‘Determinants of Citizens’ Normative Conceptions of Democracy’, considering pre-political characteristics such as individual personality and styles of cognition as well as feelings of political efficacy as relevant determinants. Finally, and most importantly, we explore implications of normative conceptions of democracy for preferences over alternative democratic decision-making procedures and political support (section ‘Implications of Citizens’ Normative Conceptions of Democracy’). We conclude that satisfaction with democracy may significantly depend on coherence between the own conception of democracy and the specific institutionalisation of democracy in one’s country. Given the different conceptions that prevail in pluralist societies, such coherence can never be fully achieved for all. However, democratic and inclusive meta-deliberation, that is, deliberation about the virtues and challenges of dem- ocratic decision-making procedures, could help to maintain and renew the procedural consensus upon which our democracy rests. Where Democrats Agree and Disagree Support for democracy relies upon a procedural consensus. Where substantial conflicts cannot be resolved, this procedural consensus allows societies to deal with conflict peace- fully and constructively. Where a procedural consensus prevails, losers in votes and elec- tions nonetheless accept decisions and remain loyal to the system because they can hope to win support for their positions at a later time. In secular, pluralist societies, procedures produce legitimacy and constitute the way to manage dissent (see Luhmann, 1983). In a consolidated democracy like Germany, support for democratic government is almost unanimous, meaning that to a significant degree, the procedural consensus pre- vails. Data from the European Social Survey (round 6, 2012) show little variation in sup- port for items that specify the rule of law and the minimal requirements for electoral democracy that characterise a Dahl-style polyarchy (Dahl, 1971), with mean values rang- ing between 9 and 10 on a 10-point scale (see Ferrin and Kriesi, 2016). Norris uses similar items that specify a system of rule of law plus Dahlsian polyarchy to describe the evolu- tion of attitudes to democracy as a learning process (Norris, 2011: ch. 8): in pre-demo- cratic societies and young democracies, citizens may still hold ‘authoritarian’ or ‘instrumental’ conceptions of democracy, which in the process of democratic consolida- tion are replaced with ‘procedural’ conceptions (polyarchy and rule of law), resulting in an ‘enlightened’ understanding of democracy. Norris’ understanding of citizens’ conceptions of democracy is thus a highly cognitiv- ist rather than a normative one: As citizens’ experience with democracy grows, false conceptions are increasingly corrected, ultimately leading to ‘enlightenment’ about what democracy means. The combination of rule of law and polyarchy does indeed constitute a kind of conceptual core of the concept of democracy, with anyone holding that ‘it is essential to democracy that the army takes over if the government is incompetent’ simply making a category mistake. However, we assert that citizens’ conceptions of democracy 4 Political Studies 00(0) are far more comprehensive than implied by Norris and entail normative evaluations that can be subject to reasonable disagreement even among democrats. Despite the attempts by Dahl and many others to specify empirical criteria that qualify a system of rule as democratic, the concept of democracy remains a normatively loaded and essentially contested concept. One of the reasons why the concept of democracy remains contested is that it is also an ‘appraisal concept’ with strong positive connotations (Connolly, 1993 [1974]: 29). Accordingly, even evidently non-democratic regimes nowa- days describe themselves as democratic, and ‘undemocratic’ appears to have become the ultimate disqualification of decisions and institutional orders. Given the normative appeal of the concept of democracy and its central relevance to political orders across the globe, it is little surprising that normative democratic theory is one of the most productive fields of political philosophy. Theories of democracy offer different interpretations and normative justifications of democracy and derive different demands and conclusions for institutional design from these. We suppose that citizens similarly hold different normative ideals of democracy and have different justifications available for them. While citizens’ normative ideas about democracy may be less formally developed and less encompassing than those elaborated on by political theorists, we expect them to be characterised by a considerable degree of consistency and coherence that makes it possible to explore them empirically and to iden- tify distinct conceptions behind citizens’ attitudes to democracy. The way in which we suggest to distinguish between a core concept of democracy that is subject to a far-reaching procedural consensus and the numerous and differing conceptions of democracy on which citizens can legitimately disagree is somewhat simi- lar to the way in which John Rawls distinguished the concept of justice from conceptions of justice: Thus it seems natural to think of the concept of justice as distinct from the various conceptions of justice and as being specified by the role which these different sets of principles, these different conceptions, have in common (Rawls, 1999 [1971]: 5).1 Put differently, the core is clasped into a second layer of norms and ideals, a layer that may differ much more between individuals than the core does. For example, take a group of three people who all agree that courts should treat everyone the same, that elections should be free and fair and that the media and opposition should be free to criticise the government. This set of assumptions is part of the conceptual core, or concept of democ- racy, and it is shared between the three. Apart from this, A holds a more comprehensive conception of democracy as deliberative democracy, while B would favour a more direct democracy and C supports supermajoritarian, consensus-oriented decision-making. These more comprehensive conceptions of democracy entail sets of assumptions that are not and need not be shared by each and every liberal democrat. Graphically, the relation- ship might be illustrated as a set of polycentric circles, each of which entails a different conception of democracy. An area of overlap between the circles represents the concep- tual core and sphere of consensus, whereas the areas with less or no overlap represent the sphere of disagreement. This distinction between a core concept of democracy and differ- ent conceptions of democratic decision-making is also familiar from classic work in the political culture tradition: Easton’s (1967) model of political support, for example, dif- ferentiates between support for ‘regime principles’ and ‘regime norms and procedures’. These levels of support are described by Dalton as follows: Landwehr and Steiner 5 Regime principles define the broad parameters within which the political system should function. At the broadest level, this involves choices about whether political relationships should be organised as a democratic, authoritarian, or other political form. […] A second major component of the regime consists of the norms of behaviour, which Easton called the operating rules or the rules of the game. These involve the specific rules or norms governing political action. […] democracy can take multiple forms that involve different assumptions about the role of the citizen, the political rights of individuals, the acceptance of dissent and political conflict, and other features of the political process (Dalton, 2004: 6). What is important is that in this article, we do not regard any of these specific concep- tions as more democratic than the others. Obviously, everyone regards their own concep- tion of democracy as the right one and has reasons to justify its validity claim. Political theorists may assess such justifications and discard some conceptions as less well-founded and thus less democratic than others on the basis of this assessment. Our article, by con- trast, regards citizens’ different normative conceptions of democracy as a sphere of rea- sonable disagreement between democrats, that is, between those who agree on the principles of polyarchy and the rule of law. Just as there are different ways to institution- alise a polyarchy – as a parliamentary or presidential, consensus or majoritarian, federal or centralist, more direct or more representative democracy – there are different norma- tive conceptions of democracy among citizens, for example, more aggregative or more deliberative ones. Previous research in the political culture tradition has tended to focus on support for regime principles or, more specifically, support for (electoral and/or liberal) democracy and democratic values, often with a particular focus on young democracies (e.g. Carlin and Singer, 2011; Fuchs and Roller, 2006; Kiewiet de Jonge, 2013; Norris, 1999). By contrast, there is relatively little comprehensive research on citizens’ normative orienta- tions towards specific types of liberal democracy and conceptions of good democratic decision-making within liberal democracies. The work that exists in this area tends to look at specific aspects of citizens’ normative conceptions of democracy in isolation such as orientations towards models of political representation and responsiveness (Bengtsson and Wass, 2010; Carman, 2006) support for direct (vs representative) democracy (Dalton et al., 2001; Webb, 2013) and/or stealth democracy (Bengtsson and Mattila, 2009; Coffé and Michels, 2014; Hibbing and Theiss-Morse, 2002), as well as populism (Spruyt et al., 2016). Interesting attempts to explore citizens’ complex attitudinal patterns with regard to democracy include the work by Dryzek and Berejikian who used Q-methodology to explore understandings of democracy in focus groups (Dryzek and Berejikian, 1993); Font et al. who distinguish support for participatory, representative and expert-based models of decision-making (Font et al., 2015); or Canache’s (2012) examination of sub- stantive conceptualisations of democracy in Latin American countries. To explore citizens’ normative conceptions of democracy more comprehensively, we have compiled a relatively long list of statements on which we expect opinions to vary. The resulting questionnaire was fielded via the GESIS Panel (GESIS, 2017) in Germany in 2015 (see section ‘Four Conceptions of Democratic Decision-Making’). The rationale behind our list of statements is that we expected disagreement to exist with regard to certain aspects of democratic decision-making. In particular, democrats may be expected to have different views on political representation and different expectations of their rep- resentatives. Some may support more inclusive, consultative, others more efficient, majoritarian styles of decision-making. Moreover, democrats may prioritise checks and 6 Political Studies 00(0) balances over governmental discretion or vice versa. Finally, they may have different attitudes where democratic self-determination and the protection of minority rights con- flict. We assume that citizens’ attitudes on these and possible other aspects are interde- pendent, constituting more or less consistent and coherent attitudinal patterns that can be rationalised as different normative conceptions of democracy – conceptions over which democrats disagree. However, we do not impose a specific dimensional structure of dem- ocratic disagreement a priori. Rather, we aim to explore the dimensionality leveraging the information contained in our longish list of statements in an inductive way. Four Conceptions of Democratic Decision-Making For our empirical analysis, we draw on an original item battery that was included in wave ‘cd’ (August–October 2015) of the GESIS Panel.2 Respondents were invited to rate their agreement with 17 normative statements (listed in Table 1 below) on demo- cratic procedures and decision-making on a 7-point scale ranging from ‘totally disa- gree’(=1) to ‘totally agree’ (=7).3 The first observation to take note of is that we observe substantial variation of survey responses on these items (see Figure S1 in the online appendix for a complete set of histograms). In a few instances, nearly all of the respond- ents tend to one side of the scale. But for most items, responses are spread all over the scales – showing that our list of statements taps into normative orientations towards specific democratic procedures over which citizens indeed disagree to a strong extent. Such variation is rarely observed in existing research on generalised support for liberal democracy in developed democracies (see above), and it constitutes the prerequisite for the first step of our empirical analysis: the analysis of latent factors that motivate responses to the individual items. To identify latent factors that structure responses on these single items, we utilise exploratory factor analysis. More specifically, we ran a principal-component factor analy- sis and rotated the solution via the promax procedure.4 In deciding for an oblique rotation method, we avoid making the restrictive and probably unrealistic assumption that the latent dimensions are orthogonal to each other, that is, completely uncorrelated. The anal- ysis resulted in four factors with an eigenvalue above 1. Using this usual eigenvalue threshold criterion, we chose a solution with four factors (also see the scree plot in Figure S2 in the online appendix). After rotation, we obtained the loadings displayed in Table 1. Most of our items load strongly on only one of these four factors. Four items load at least moderately on two factors. The correlations between these factors are moderate and below 0.3 in all cases (see Table S1 in the online appendix with the correlation matrix). From the factor analysis, we thus derive four distinct factors that drive agreement or disa- greement with controversial statements on democracy. We contend that these factors reflect aspects of disagreement over normative conceptions of democracy that can be rationalised and that are internally coherent in that they express a pattern of consistent and mutually reinforcing beliefs. Respondents’ positions on these dimensions are computed via the regression scoring method and used as dependent and independent variables in the analyses that follow below. To ease interpretation, we have standardised the scores such that they vary between 0 and 1 in the observed data for all analyses discussed below. Before turning to these, we describe the substance of the four latent factors we identify. We respectively label these four factors, considering their directionality, ‘trustee model of democracy’, ‘anti-pluralist scepticism’, ‘deliberative proceduralism’ and ‘populist major- itarianism’ and rationalise them as follows: Table 1. Results of Exploratory Factor Analysis. Factor 1 Factor 2 Factor 3 Factor 4 Explained variance 2.68 2.62 2.27 1.69 TRUSTEE1: The government should stick to planned policies even if a majority of 0.79 Trustee model citizens is against them. of democracy TRUSTEE2: Sometimes it is better when political decisions are taken behind closed 0.75 Landwehr and Steiner doors. TRUSTEE3: Members of parliament should follow their conscience even if a majority 0.57 0.46 of citizens happens to hold different opinions. TRUSTEE4: All bodies that take part in political decisions should meet in public. −0.51 TRUSTEE5: The government should change planned policies if a majority of citizens −0.44 no longer supports them. ANTIPLUR1: Struggles between different interest groups in our society damage the 0.71 Anti-pluralist common good. scepticism ANTIPLUR2: It would be better if important political decisions were taken by 0.66 independent experts rather than elected politicians. ANTIPLUR3: Important political decisions should be taken only with approval of all 0.61 affected groups. ANTIPLUR4: Important political decisions should be taken in deliberation and not 0.60 −0.41 through mere voting. ANTIPLUR5: In general, conflicts cannot be resolved through discussion and 0.48 negotiation. DELPROC1: One has to accept democratically taken decisions in any case, even if 0.73 Deliberative they conflict with own interests. proceduralism DELPROC2: In political decisions, the common good and not the own interest should 0.68 be the central focus. DELPROC3: It is important in democracy to understand why other people have 0.61 different opinions. DELPROC4: The government should develop policies in close dialogue with citizens −0.41 0.49 and affected groups. POPMAJ1: Minority rights must be protected from majority decisions. −0.70 Populist POPMAJ2: Majority decisions must apply, even if they curtail minority rights. 0.70 majoritarianism POPMAJ3: If there is a large majority in the population for a political decisions, this 0.42 indicates that the decision is correct. Results from a principal-component factor analysis after oblique promax rotation (Kaiser off, power three). Loadings below 0.4 suppressed from table. 7 8 Political Studies 00(0) The first factor reflects an understanding of democracy in which political labour is shared between citizens and elected politicians in a way that representatives are trusted to possess the expertise and experience to take decisions in the interest of the collective, whereas citizens exercise control only in periodic elections. This understanding resem- bles the classical distinction between the trustee model of political representation and its delegate (or mandate) counterpart. The essential difference, according to Dovi (2014), is that ‘[d]elegate conceptions of representation require representatives to follow their con- stituent’s preferences, while trustee conceptions require representatives to follow their own judgement about the proper course of action’. Behind such attitudes, we see not only a model of representation, but a distinct conception of democracy more generally in which the volatile opinions of the electorate are seen as impeding on good government. In this conception, publicity and transparency are rated low (see TRUSTEE2 and TRUSTEE4), as clandestine decision-making liberates politicians from excessive public scrutiny. Instead, respondents in whom this conception is strong tend to value what Majone labelled ‘temporal consistency’ and ‘credibility’ (Majone, 1996): rather than respond to changes in public opinion, representatives should ‘follow their conscience’ (TRUSTEE3) and ‘stick to planned policies’ (TRUSTEE1). In essence, people holding this conception willingly relinquish participatory rights beyond electoral suffrage and wish to surrender control to representatives – at least for the limited term of office. The second factor, anti-pluralist scepticism, entails both anti-pluralist and essentially sceptical attitudes. In the words of Mudde and Kaltwasser (2012: 152), the key point of pluralism is the assumption ‘that societies are composed of several social groups with different ideas and interests. […] Hence, pluralism takes for granted that it is impossible to generate something like a “general will” of the people’. Accordingly, pluralists accept, or even embrace, competition between interest groups, negotiations between competing interests and the notion that there are losers of political competition. In contrast, the anti- pluralism we identify here is characterised by a conception of democracy that somewhat resembles the attitudes Hibbing and Theiss-Morse (2002) described as ‘stealth democ- racy’. It rejects the idea that societies consist of different groups or classes with partly conflicting interests and norms. Instead, it seems to assume the existence of an independ- ent standard for ‘good’ or ‘correct’ decisions which are hampered by the influence of organised interest groups on politics. Given these assumptions, anti-pluralists reject com- petition and negotiations between interest groups (see ANTIPLUR1) and political antag- onism more broadly. They instead value consensus and harmony. Interest groups are thus viewed not as stakeholders or representatives of legitimate concerns, but as avaricious distributive coalitions in Mancur Olson’s sense (Olson, 1982). In keeping with this rejec- tion of competition between divergent interests and the belief in ‘correct’ decisions is the preference for expert decision-making (ANTIPLUR2), which is also an indicator for stealth democracy used by Hibbing and Theiss-Morse. In addition, however, the attitudes reflected here are characterised by a scepticism that becomes apparent in the insistence on consultation of all affected groups (i.e. demand for veto rights, see ANTIPLUR3), as well as a preference for deliberation over mere voting (ANTIPLUR4) paired with disillusion- ment on the possibility of conflict resolution through communication (ANTIPLUR5).5 The third factor captures attitudes on what we label deliberative proceduralism. This factor comes pretty close to the democratic ideals advocated by proponents of delibera- tive democracy (see, for example, Cohen, 1997; Habermas, 1996). Deliberative procedur- alists value rational discourse and public reasoning against the background of the common good. The priority assigned to a common good over own interests (DELPROC2) may Landwehr and Steiner 9 seem to resemble anti-pluralism, but in fact plays a different role here. In a conception of deliberative proceduralism, private material interests are to be banned from influencing decision-making – and they are to be banned from individual deliberation before even entering collective deliberation. Interests do play a role in deliberation, but as arguments to be considered in discursive interaction rather than motives for individual decisions. What is essential here is that the input to collective decisions, and therefore both to delib- eration and to voting and elections, should consist in judgements rather than subjective preferences (Estlund, 1990).6 In deliberation, all relevant arguments and reasons must not only be voiced but also be heard, which is why ‘in democracy, it is important to under- stand why other people hold different opinions’ (DELPROC3). Finally, the deliberative- proceduralist conception assigns strong authority to democratically taken decisions (DELPROC1) because deliberative democratic procedures are regarded as the way to collaboratively construct rational solutions that take into account all affected interests and relevant arguments. Substantial and procedural justice is thus not at odds in this ideal, but instead, one is realised through the other. The fourth and final factor characterises a populist-majoritarian conception of democ- racy. The key trade-off here is between adherence to the will of the majority and the protection of minority rights. On one side, populist majoritarianism is based on a strong belief in a will of the people that resembles Rousseau’s general will. This leads to a strong support for majority rule, even at the cost of minority protection (POPMAJ1 and POPMAJ2), and a belief in the wisdom of large majorities (POPMAJ3). In this concep- tion, losers in votes and elections have simply failed to identify the general will correctly and do not require particular consideration or protection. It may be assumed, however, that respondents understood minorities not (only) as electoral minorities, but instead as ethnic and cultural minorities. The rejection of minority rights sits easily with populist majoritarianism. After all, Rousseau’s ideal of the general will only make sense for a maximally homogeneous community. Cultural and ethnic plurality appears to rule out the identification of a single ‘will of the people’. Nativism and cultural reactionism are thus a complement to strong majoritarianism.7 Determinants of Citizens’ Normative Conceptions of Democracy Where do different normative conceptions of democracy come from? While they cer- tainly reflect and are reflected by discursive processes, there are likely to be determinants on the individual level as well. Only the latter can be explored on the basis of our survey data. In our analysis, we consider both pre-political individual characteristics and atti- tudes to politics. While attitudes to politics reflect and can be changed by socialisation and political experiences, pre-political characteristics tend to be more stable. If such pre- political characteristics were found to have effects on conceptions of democracy, some democratic disagreement would have to be viewed as irresolvable through political pro- cesses and thus as a fundamental condition of democracy. Among individual characteristics, personality is to a significant degree innate or devel- oped early in life and thus a basal one. Personality has been discussed as a determinant of political attitudes since the publication of Theodor W. Adorno et al.’s (1950) seminal book on the authoritarian personality. However, the indices used by Adorno et al. have been much criticised methodologically and they are not suitable for our purposes here since our goal is to explore disagreements between democrats rather than differences 10 Political Studies 00(0) between authoritarian and democratic personalities. In the models presented below, we will therefore make use of the Big 5-inventory which is well-established in psychology and enables reliable and valid measurements of five dimensions of personality (extraver- sion, emotional stability, conscientiousness, openness and agreeableness). Existing research shows correlations between personality on the one hand and political participa- tion (Dinesen et al., 2014; Gerber et al., 2011) and party preferences (Schoen and Schumann, 2007) on the other hand. Personality characteristics are also likely to shape citizens’ normative conceptions of democracy (Font and Alacron, 2011). We expected, for example, that openness to experience and agreeableness are negatively related to populist majoritarianism as individuals scoring higher on these personality traits are likely to be friendlier towards minority rights (for the connection between agreeableness and voting for populist parties, see Bakker et al., 2015). Higher openness to experience should also be associated with a preference for open discourse as entailed in the deliberative-proce- dural conception of democracy. Conscientious people might be expected to prefer more public scrutiny of elected representatives and thus to reject the trustee conception of democracy. A further individual property that is known to be of relevance for political attitudes is what psychologists describe as ‘cognitive style’. Politics is a cognitively highly demand- ing subject: it requires opinion-formation and decision-making under conditions of uncertainty, complexity and contingency. Cognitive closure is difficult to achieve in political questions and rarely lasting. It is thus little surprising that individuals with a high need for cognition, that is, ‘the tendency for an individual to engage in and enjoy thinking’ (Cacioppo and Petty, 1982), have a stronger interest in politics and are more willing to engage in it (Condra, 1992). We hypothesise that citizens with a high need for cognition are more likely to endorse a demanding and essentially cognitivist conception of democracy like deliberative proceduralism. At the same time, citizens with a high need for cognition should be less likely to hold the, arguably, simplistic belief in the existence of a single will of the people on which populist majoritarianism is based. We therefore expect a negative effect of need for cognition on populist majoritarianism. Somewhat related to this argument, past research has linked cognitive style, precisely need for cognitive closure, and authoritarianism (e.g. Chirumbolo, 2002) but we are unaware of studies that examine the link between cognitive style and preferences for democratic decision-making. Moreover, very general experiences and perceptions of politics are likely to be reflected in citizens’ conceptions of democracy. While some degree of frustration is inevitable in political engagement and confrontation with politics, people can see their own under- standing of it and politicians’ responsiveness in different lights, experiencing higher or lower levels of self-efficacy. We assume that people with a high degree of internal effi- cacy (who rate their understanding of politics high) are more likely to possess concep- tions of democracy that entail more and more direct citizen involvement, such as deliberative proceduralism and populist majoritarianism. Our reasoning is that those who feel internally efficacious should be more likely to think of themselves and citizens in general as being capable of directly engaging in politics. For the same reason, citizens who feel internally efficacious should be less supportive of the trustee model of democ- racy. Individuals with a perception of high external efficacy, who view decision-makers as responsive to their interests and opinions, seem by contrast more likely to adopt a trustee stance on political representation (see, similarly, Carman, 2006: 108) and democ- racy more generally. Low external efficacy is likely to result in sceptical, distrustful Landwehr and Steiner 11 attitudes and a demand for more control as reflected in anti-pluralist scepticism. In accordance with our expectations, past research has found that external efficacy is posi- tively related with support for the trustee model (Carman, 2006) and negatively with support for both direct democracy and stealth democracy (Bengtsson and Mattila, 2009). Finally, the classical socio-economic variables may be correlated with normative conceptions of democracy as well. We did not expect any strong effect for gender, but assumed income and education to be positively associated with the more elite-friendly conceptions of democracy, namely, the trustee model of democracy and deliberative proceduralism, and negatively associated with the more elite-sceptical or elite-challeng- ing conceptions of anti-pluralism and populist majoritarianism. In line with that, Bengtsson and Mattila (2009) report that education has a negative impact on support for direct and stealth democracy, and other studies report a positive relation between educa- tion and the trustee model of representation (Carman, 2006; Rosset et al., 2016). A fur- ther factor that is likely to be relevant in the German case is the fact that a considerable percentage of Germans were politically socialised in the communist-authoritarian sys- tem of the German Democratic Republic (GDR). A plethora of previous studies docu- ment East/West divides in the German political culture (e.g. Fuchs, 1999; Sack, 2016), for which we also control. We measured all these covariates through standard items and, therefore, just refer to Table S3 in the online appendix for detailed information here.8 This includes information on survey waves as we combine information from several waves of the GESIS Panel. All covariates have been standardised to range from 0 to 1 to allow for comparison of coef- ficient sizes. We analysed the expected associations via simple ordinary least squares (OLS) regression analyses using our four dimensions of democratic disagreement as dependent variables in turn. The results of these four models are displayed graphically in Figure 1 below, which show the point estimates for the coefficients and their 95% confi- dence intervals. The results mostly confirm our expectations. Openness to experience and agreeable- ness are negatively related to populist majoritarianism and positively related to delibera- tive proceduralism. Conscientiousness shows, as expected, a negative effect on the trustee model of democracy. Moreover, conscientious persons tend to be more supportive of the anti-pluralist sceptic conception than less conscientious individuals. Extraversion has a positive effect on anti-pluralist scepticism and a negative one on the trustee model of democracy. Neuroticism exhibits no statistically significant associations with any of the four conceptions of democracy. As expected, need for cognition is positively related with deliberative proceduralism and strongly negatively associated with populist majoritarian- ism. These findings so far support our general idea that basic psychological (and rather apolitical) attributes play a role when it comes to citizens’ relatively abstract political preferences over normative conceptions of democracy. Again in line with expectations, internal efficacy has a positive effect on deliberative proceduralism and a negative one on the trustee model, although we don’t find the expected positive association with regard to populist majoritarianism. External efficacy has a strong positive effect on the trustee model of political representation and a similarly strong negative effect on anti-pluralist scepticism, as expected. Furthermore, individuals who feel externally efficacious tend somewhat less towards deliberative proceduralism and populist majoritarianism. What is important is that the effects of internal and external efficacy are not always aligned. This is most clearly the case for deliberative procedural- ism. In what appears to be a theoretically sensible pattern, those with high external 12 Political Studies 00(0) Figure 1. Determinants of Conceptions of Democracy. efficacy are less supportive of deliberative proceduralism (potentially because they trust politicians to deliver good policies even without broad public deliberation), while those high on internal efficacy embrace it (possibly because they think of themselves and citi- zens in general as being capable of engaging in deliberation, as reasoned above). Regarding the socio-demographic controls, those with higher educational attainments tend to score higher on the deliberative-proceduralist conception of democracy and lower on anti-pluralist scepticism and populist majoritarianism (as expected); there is, however, no clear association with the trustee model. Income is positively associated with both of the more elite-friendly conceptions, that is, the trustee model and deliberative procedural- ism. Also as expected, income relates negatively to the more anti-elitist or elite-challeng- ing conception of anti-pluralist scepticism. However, contrary to our expectation, higher income goes along with increased, not decreased, populist majoritarianism. Men tend to be more in favour of the trustee model and less in favour of anti-pluralist scepticism. Living in the East of Germany is associated with a rejection of the trustee model of democracy as well as with sceptical anti-pluralist orientations. Older people are, on aver- age, more supportive of anti-pluralist scepticism and somewhat less supportive of popu- list majoritarianism. At the same time, the old tend to support the deliberative-procedural conception much more than the young do. Implications of Citizens’ Normative Conceptions of Democracy Normative conceptions of democracy may be partly driven by individual characteristics, but they are shaped and developed in discourses that engage ordinary citizens, politicians Landwehr and Steiner 13 and elected representatives, and the media. While it seems difficult to measure the effects of discourses on individual preferences, we may assume political parties and decision- makers to be to some degree responsive to citizens’ normative ideals and to offer quite specific institutionalisations of these. Where existing decision-making procedures cause frustration and dissatisfaction, they may offer democratic innovations that appeal to citi- zens’ expectations. Democratic innovations that are currently being advocated and that have partly been put to practice in recent years include direct democracy (referenda and citizen initiatives), deliberative citizen forums (deliberative polls, consensus conferences and planning cells) and non-elected commissions staffed with experts and stakeholders. Whereas from the perspective of the trustee model of representation democratic innova- tions seem neither necessary nor desirable, each of the other three conceptions identified (i.e. the positive poles of the dimensions) here seems to demand a specific kind of demo- cratic innovation of the presently mostly representative decision-making procedures in Germany: anti-pluralist scepticism would be consistent with a stronger emphasis on expert decision-making, deliberative proceduralism would require more deliberative and participatory decision-making, and populist majoritarianism seems to imply a preference for more direct democracy through referenda. Against this background, we expected citi- zens’ normative conceptions to be closely associated with preferences for concrete demo- cratic decision-making procedures. In order to test this expectation, we included another list of statements in the GESIS Panel (in the next wave, such that the problem of potential halo effects should be less of an issue). We asked for agreement on scales from 1 (‘totally disagree’) to 7 (‘totally agree’) with the following four statements: •• ‘There should be more referendums in Germany’. •• ‘There should be more dialogue procedures that involve citizens’. •• ‘There should be more expert commissions that are directly involved in important political decisions’. •• ‘Important political decisions should only be made by elected parliaments’. In order to see how preferences for these democratic decision-making procedures are related to the four dimensions of democratic disagreements, we estimated a set of OLS regressions with these four items consecutively used as dependent variables (re-scaled to range from 0 to 1 to ease interpretation). The results of the regressions, which additionally control for the two efficacy items as well as the demographic covariates from above, are shown in Figure 2 below. The results support our four core expectations: Those who adhere to the trustee model of democracy are satisfied with important political decisions being exclusively made by elected parliaments; anti-pluralist sceptics would like to see more expert commissions; those who score high on deliberative proceduralism want more citizen dialogues; and support for populist majoritarianism is positively associated with demanding more refer- enda. Unsurprisingly, adherence to the trustee model of democracy then also goes along with reduced demand for referendums and citizen dialogues (as alternatives to parliamen- tary decision-making). In addition to expert commissions, anti-pluralist sceptics also tend to be more supportive of referendums and citizen dialogues (as alternatives to traditional decision-making in representative democracy involving interest groups and politicians as representatives of different interests). Perhaps not that surprisingly, this analysis shows that the identified normative conceptions translate meaningfully into preferences for con- crete democratic decision-making procedures. As such, these findings also help to vali- date the results of our factor analysis. 14 Political Studies 00(0) Figure 2. Conceptions of Democracy and Preferences over Decision-Making Procedures. Finally, but most importantly, we expected different normative conceptions of democ- racy to be linked with political support for the regime. Following the core idea of the political culture paradigm, we assume that political support will depend on (perceived) congruence between the own conception of democracy and institutionalised norms and procedures. Considering the four factors identified above, we expected those with a trus- tee model of democracy to perceive the largely representative German system of govern- ment as more congruent with their own conception. Accordingly, political support should be positively linked with subscribing to the trustee model of democracy. As to anti-plural- ist scepticism, one would expect anti-pluralist sceptics to be more critical of a system of government that is characterised by open political competition and negotiations between competing interests and in which corporatist decision-making processes are dominated by organised interests. For this reason, anti-pluralist scepticism should affect political support negatively. Given that the German political system and culture offer sites for (mostly elite) deliberation and consultation, a certain degree of congruence may be given for those who adhere to deliberative proceduralism. However, as citizen involvement remains rather low and ineffectual, it also seems possible that expectations arising from deliberative proceduralism are frustrated. As a result, it seems unclear how deliberative proceduralism and political support are related. Populist majoritarianism, finally, is argu- ably incongruent with a political system that is characterised by strong protection of minority rights, strong federalism, little use of direct democracy, proportional representa- tion resulting in coalition governments and numerous veto-players. We therefore expected political support to be negatively associated with populist majoritarianism. To put these theoretical expectations to an empirical test, we look at two different established measures of political support at the regime level: trust in political institutions (and related authorities) and satisfaction with the way democracy works. Following Landwehr and Steiner 15 Figure 3. Conceptions of Democracy, Satisfaction with Democracy and Political Trust. Dalton (2004: 24), the former is an indicator of affective orientations towards political institutions of the regime and the latter a measure of evaluative orientations towards norms and procedures of the regime. We expected similar effects on both measures and use both to check the robustness of our findings. We measure trust in regime institutions and its central authorities via the mean of four questions that ask for trust in the German Bundestag, the German federal government, political parties and politicians on scales from 1 to 7. Satisfaction with the way democracy works in Germany is measured on a scale from 1 to 10. Both outcome measures were again re-scaled to range from 0 to 1 for the analysis. Our regressions control for the same set of socio-demographics used above. The results are displayed in Figure 3 and are similar for both measures of political sup- port. Our expectations are vindicated with regard to the trustee model of democracy, exhibiting a positive effect on support, and anti-pluralist scepticism, showing a negative effect on support. Subscribing to a deliberative-proceduralist conception of democracy goes along with higher levels of support. We fail to find unambiguous support for the expectation that populist majoritarians are less supportive of the regime, although the findings for political trust point in this direction. Discussion and Conclusion Our results show that while there is a strong consensus on core principles of democracy, democrats disagree about the premises, promises and specific institutionalisations of democracy. We have, in a first step, identified four distinct factors that drive citizens’ normative conceptions of democracy. We have described these factors as (1) a trustee model of democracy, (2) anti-pluralist scepticism, (3) deliberative proceduralism and (4) populist majoritarianism. All of these are generally compatible with support for liberal electoral democracy, but may result in different expectations and evaluations. In a second 16 Political Studies 00(0) step, we have explored the relationship between individual characteristics and these fac- tors, finding correlations with personality, cognitive styles, internal and external efficacy as well as socio-demographic characteristics. In a third step, we have shown that differing normative conceptions of democracy also result in different preferences for specific deci- sion-making procedures and in different levels of political support. Is the coexistence of different normative conceptions of democracy a threat to democ- racy? We are sure that in an open and pluralist society, it is not. On the contrary, it should be viewed as an essential characteristic of liberal democracy that the interpretation of essential normative concepts, including the concept of democracy itself, remains open to contestation and recalibration. Nonetheless, a lack of congruence between the own conception of democ- racy and institutionally entrenched norms can lead to disappointment and disaffection among citizens. In Germany, citizens adhering to the trustee model of democracy probably find a high congruence between their democratic ideals and institutional practice, as Germany is a representative democracy offering few opportunities for participation beyond elections. Accordingly, support for the trustee model of democracy results in higher politi- cal support among our respondents. Overall, our findings imply that those who advocate a trustee and/or a deliberative-proceduralist model of democracy tend to be more supportive of the existent regime while anti-pluralist sceptics tend to be less satisfied. Are the democratic aspirations and expectations of citizens whose normative concep- tion of democracy is at least in part incoherent with institutional reality to be thwarted? We do not think so. However, the procedural consensus upon which democracy rests can only be sustained and renewed when societies engage in processes of democratic meta- deliberation: when they deliberate about how, where and by whom collectively binding decisions should be taken (see Landwehr, 2015). Democratic meta-deliberation is based on the premise that institutions and procedures are not fixed once and for all, but that they are – within the scope of liberal democracy – challengeable and revisable. Democratic innovations are possible and they may be necessary to renew the procedural consensus. Democratic meta-deliberation does not necessarily require large-scale citizen involve- ment, although the British Columbia Citizens’ Assembly certainly constitutes an encour- aging example (see Warren and Pearse, 2008) of how the ‘rules of the game’ might themselves be democratically and deliberatively be changed. What is crucial, however, is that public discourses and political agendas are open to procedural issues, and that politi- cians and parties are responsive to demands for procedural reform. With regard to academic discourses and research, we would like to encourage scholars to look beyond support for the core principles of liberal democracy and explore what we have described as the second layer of citizens’ more comprehensive conceptions of demo- cratic decision-making and their implications. Funding The author(s) received no financial support for the research, authorship and/or publication of this article. Supplementary Material Additional supplementary information may be found with the online version of this article. Notes 1. In this distinction, Rawls follows H. L. A. Hart (2012 [1961]). 2. The GESIS Panel is an academic infrastructure for data collection that was launched in 2013 (see Bosnjak et al., 2017). Data are representative of the German-speaking population aged between 18 and 70 years Landwehr and Steiner 17 and permanently residing in Germany. The GESIS Panel uses a mixed-mode approach with respondents self-completing questionnaires either online or by mail in bi-monthly regular waves. Our items were included in wave ‘cd’ which was in the field between 12 August 2015 and 14 October 2015 and included 3615 participants. Some of the covariates we use in the later analyses were included in different waves of the panel, as described below. 3. Note that our original list included 18 items. We excluded the following statement from the analysis: ‘Most people do not know enough to make informed decisions about important political topics’. The item doesn’t align well with any others statistically and excluding it (marginally) improves the clarity of our dimensional solution. In hindsight, we noted that the item stands out from all others in that it is less about a preference for a specific way of democratic decision-making, but concerned with related prerequisites. 4. This procedure results in the listwise deletion of observations with missing values on at least one of the items. There are a few missing values on the statements, however, such that we still have 2484 valid obser- vations on the four factors (or for 68.7% of the 3615 participants in the respective wave of the panel). We alternatively recoded all ‘don’t knows’ to the midpoint of the scales and re-run our factor analysis. This resulted in a very similar solution (see Table S2 in the online appendix). 5. It may seem surprising that this factor also loads on the items demanding consent from all affected (ANTIPLUR3) and on the preference for deliberation to mere voting when it comes to important political decisions (ANTIPLUR4), which at a first glance seem to characterise a more deliberative attitude to democ- racy. Upon closer consideration, though, the correlations do make sense: A common and plausible criterion for the quality of decisions is Pareto superiority. While respondents are certainly not familiar with such technicali- ties, the idea that good decisions should hurt no one, but benefit everyone, is certainly more mundane. The insistence on consent from, and thus a veto right for, affected groups appears rational from a perspective of fundamental scepticism, as it ensures Pareto superiority and prevents any false positives in decision-making – even at the cost of political immobility. From another perspective, their rejection of political antagonism leads anti-pluralists to demand consent from all affected and to reject majority voting that would produce losers. 6. In line with valuing reasoned judgements as the basis for democratic decision-making, deliberative proce- duralists also tend to agree that ‘members of parliament should follow their conscience even if a majority of citizens happens to hold different opinions’ (TRUSTEE3 which also correlates with a trustee model of representation). At the same time, deliberative proceduralists want the government to ‘develop policies in close dialogue with citizens’ (DELPROC4). 7. Note that we do not want to imply that this dimension measures populism as such. What we describe as ‘populist majoritarianism’ covers only some aspects of broader definitions of populism. Mudde’s influen- tial definition views populism as a ‘thin- centred ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, ‘the pure people’ versus the ‘corrupt elite’, and which argues that politics should be ‘an expression of the volonté générale (general will) of the people’ (Mudde, 2007: 23). To clarify the difference between Mudde’s encompassing measure of populism on the attitudi- nal level and our measures of democratic disagreement, consider the populism scale by Akkerman et al. (2014). The items on their populism scale combine elements of, using our dimensions, a rejection of the trustee model of democracy, anti-pluralist scepticism and populist-majoritarian attitudes. In keeping these elements distinct, we stick to the solution of our factor analysis. 8. The online appendix also includes a table (Table S4) with descriptive statistics for all variables used in the analyses of this article. References Adorno TW, Frenkel-Brunswik E, Levinson DJ, et al. (1950) The Authoritarian Personality. New York: Harper & Bros. Akkerman A, Mudde C and Zaslove A (2013) How Populist Are the People? Measuring Populist Attitudes in Voters. Comparative Political Studies 47 (9): 1324–1353. Bakker BN, Rooduijn M and Schumacher G (2015) The Psychological Roots of Populist Voting: Evidence from the United States, the Netherlands and Germany. European Journal of Political Research 55 (2): 302–320. Bartels LM (2009) Unequal Democracy: The Political Economy of the New Gilded Age. Princeton, NJ: Princeton University Press. Bengtsson A and Mattila M (2009) Direct Democracy and Its Critics: Support for Direct Democracy and ‘Stealth’ Democracy in Finland. West European Politics 32 (5): 1031–1048. Bengtsson A and Wass H (2010) Styles of Political Representation: What Do Voters Expect? Journal of Elections, Public Opinion, and Parties 20 (1): 55–81. 18 Political Studies 00(0) Bosnjak M, Dannwolf T, Enderle T, et al. (2017) Establishing an Open Probability-based Mixed-mode Panel of the General Population in Germany: The GESIS Panel. Social Science Computer Review. Epub ahead of print 14 March 2017. DOI: 10.1177/0894439317697949. Cacioppo JT and Petty RE (1982) The Need for Cognition. Journal of Personality and Social Psychology 42 (1): 116–131. Canache D (2012) The Meanings of Democracy in Venezuela: Citizen Perceptions and Structural Change. Latin American Politics and Society 54 (3): 95–122. Carlin RE and Singer MM (2011) Support for Polyarchy in the Americas. Comparative Political Studies 44 (11): 1500–1526. Carman CJ (2006) Public Preferences for Parliamentary Representation in the UK: An Overlooked Link? Political Studies 54 (1): 103–122. Chirumbolo A (2002) The Relationship between Need for Cognitive Closure and Political Orientation: The Mediating Role of Authoritarianism. Personality and Individual Differences 32 (4): 603–610. Coffé H and Michels A (2014) Education and Support for Representative, Direct and Stealth Democracy. Electoral Studies 35 (1): 1–11. Cohen J (1997) Procedure and Substance in Deliberative Democracy. In: Bohman J and Rehg W (eds) Deliberative Democracy. Cambridge, MA: MIT Press, pp.407–438. Condra MB (1992) The Link between Need for Cognition and Political Interest, Involvement, and Media Usage. Psychology: A Journal of Human Behavior 29 (3–4): 13–18. Connolly WE (1993 [1974]) The Terms of Political Discourse. Oxford: Blackwell. Dahl RA (1971) Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press. Dalton RJ (1999) Political Support in Advanced Industrial Democracies. Oxford: Oxford University Press. Dalton RJ (2004) Democratic Challenges, Democratic Choices: The Erosion of Political Support in Advanced Industrial Democracies. Oxford: Oxford University Press. Dalton RJ, Bürklin W and Drummond A (2001) Public Opinion and Direct Democracy. Journal of Democracy 12 (4): 141–153. Dinesen PT, Nørgaard AS and Klemmensen R (2014) The Civic Personality: Personality and Democratic Citizenship. Political Studies 62 (S1): 134–152. Dovi S (2014) Political Representation. In: The Stanford Encyclopedia of Philosophy. Online version, Spring 2017 edition. Available at: https://plato.stanford.edu/entries/political-representation/ (accessed 17 July 2017). Dryzek JS and Berejikian J (1993) Reconstructive Democratic Theory. American Political Science Review 87 (1): 48–60. Easton D (1967) A Systems Analysis of Political Life. New York: John Wiley. Estlund D (1990) Democracy without Preference. The Philosophical Review 99 (3) 376–424. Ferrin M and Kriesi H (2016) How Europeans View and Evaluate Democracy. Oxford: Oxford University Press. Font J and Alacron P (2011) The Role of Personality in the Explanation of Preferences for Democratic Processes. In: Annual Scientific Meeting of the International Society of Political Psychology, Istanbul, Turkey. Paper presented at the AECPA Conference, 7–9 September 2011, Murcia. Font J, Wojcieszak M and Navarro CJ (2015) Participation, Representation and Expertise: Citizen Preferences for Political Decision-Making Processes. Political Studies 63 (S1): 153–172. Fuchs D (1999) The Democratic Culture of Unified Germany. In: Norris P (ed.) Critical Citizens: Global Support for Democratic Government. Oxford: Oxford University Press, pp.123–145. Fuchs D and Roller E (2006) Learned Democracy? Support of Democracy in Central and Eastern Europe. International Journal of Sociology 36 (3): 70–96. Gerber AS, Huber GA, Doherty D, et al. (2011) Personality traits and participation in political processes. The Journal of Politics 73 (03): 692–706. GESIS (2017) GESIS Panel – Standard Edition (GESIS Data Archive, ZA5665 Data file, version 17.0.0). Cologne: GESIS. Habermas J (1996) Between Facts and Norms. Cambridge, MA: Blackwell. Hart HLA (2012 [1961]) The Concept of Law. Oxford: Oxford University Press. Hibbing JR and Theiss-Morse E (2002) Stealth Democracy: Americans’ Beliefs about How Government Should Work. Cambridge: Cambridge University Press. Kiewiet de Jonge CP (2013) Political Learning and Democratic Commitment in New Democracies. PhD Thesis, University of Notre Dame, Notre Dame, IN. Landwehr C (2015) Democratic Meta-Deliberation: Towards Reflective Institutional Design. Political Studies 63 (S1): 38–54. Landwehr and Steiner 19 Luhmann N (1983) Legitimation durch Verfahren. Frankfurt am Main: Suhrkamp. Majone G (1996) Temporal Consistency and Policy Credibility: Why Democracies Need Non-Majoritarian Institutions. Robert Schumann Centre Working Paper, 96(57). Florence: European Union Institute. Mudde C and Kaltwasser CR (2012) Populism and (Liberal) Democracy: A Framework for Analysis. In: Mudde C and Kaltwasser CR (eds) Populism in Europe and the Americas: Threat or Corrective for Democracy. Cambridge: Cambridge University Press, pp.1–26. Norris P (1999) Critical Citizens: Global Support for Democratic Government: Global Support for Democratic Government. Oxford: Oxford University Press. Norris P (2011) Democratic Deficit: Critical Citizens Revisited. Cambridge: Cambridge University Press. Olson M (1982) The Rise and Decline of Nations: Economic Growth, Stagflation and Social Rigidities. New Haven, CT: Yale University Press. Rawls J (1999 [1971]) A Theory of Justice. Cambridge, MA: Belknap. Rosset J, Giger N and Bernauer J (2016) I the People? Self-Interest and Demand for Government Responsiveness. Comparative Political Studies. Available at: https://doi.org/10.1177/0010414015621078 (accessed 17 August 2017). Sack BC (2016) Regime Change and the Convergence of Democratic Value Orientations through Socialisation: Evidence from Reunited Germany. Democratisation 24 (3): 444–462. Schäfer A (2013) Liberalisation, Inequality and Democracy’s Discontent. In: Schäfer A and Streeck W (eds) Politics in the Age of Austerity. Cambridge: Polity Press, pp.169–195. Schoen H and Schumann S (2007) Personality traits, partisan attitudes, and voting behavior. Evidence from Germany. Political Psychology 28 (4): 471–498. Spruyt B, Keppens G and Van Droogenbroeck F (2016) Who Supports Populism and What Attracts People to It? Political Research Quarterly 69 (2): 335–346. Warren ME and Pearse H (2008) Designing Deliberative Democracy: The British Columbia Citizens’ Assembly. Cambridge: Cambridge University Press. Webb P (2013) Who Is Willing to Participate? Dissatisfied Democrats, Stealth Democrats and Populists in the United Kingdom. European Journal of Political Research 52 (6): 747–772. Author Biographies Claudia Landwehr is Professor of Public Policy at the Johannes Gutenberg University Mainz. She has previ- ously been a Schumpeter Fellow at Goethe-University Frankfurt-am-Main and a visiting fellow at the Australian National University and Harvard University. Her research focuses on contemporary theories of justice and democracy as well as empirical questions of democratic decision-making and institutional design. She is author of the book Political Conflict and Political Preferences (ECPR Press, 2009) and of a number of articles in international journals. Nils D Steiner is a Postdoctoral Researcher at the Chair for Comparative Politics at Johannes Gutenberg University Mainz, where he completed his PhD in 2015. His research focuses on political attitudes and behav- iour, on the one hand, and (international) political economy, on the other. His work has been published in international journals such as Electoral Studies, Journal of European Social Policy and West European Politics.