Hostname: page-component-848d4c4894-8kt4b Total loading time: 0 Render date: 2024-07-03T15:18:43.491Z Has data issue: false hasContentIssue false

When do voters perceive intra-party conflict? A democratic life cycle perspective

Published online by Cambridge University Press:  29 August 2023

Tristan Klingelhöfer*
Affiliation:
Department of Political Science and European Forum, The Hebrew University of Jerusalem, Jerusalem, Israel
Jochen Müller
Affiliation:
Department of Political Science & Communication Studies, University of Greifswald, Greifswald, Germany
*
Corresponding author: Tristan Klingelhöfer; Email: tr.klinge@mail.huji.ac.il
Rights & Permissions [Opens in a new window]

Abstract

Political parties commonly experience internal disagreements. Recently, evidence is accumulating that outright internal discord makes a party much less attractive to voters. However, we do not understand well when citizens perceive a party to be internally conflicted in the first place. We here explain citizens’ perceptions from a democratic life cycle perspective: Factors related to the periodic conduct of elections induce higher levels of intra-party conflict and make it more visible to citizens. To test this argument, we combine survey data on citizens’ perceptions of political parties in Germany spanning 16 years with indicators moderating (the visibility of) intra-party conflict. The analysis shows that citizens perceive more internal conflict when parties are heterogenous, when they are governing, when election day is distant, and when electoral losses accumulate. This demonstrates the recurring patterns in citizens’ perceptions of political parties and suggests self-reinforcing dynamics between citizen assessments and election outcomes.

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - SA
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike licence (http://creativecommons.org/licenses/by-nc-sa/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the same Creative Commons licence is used to distribute the re-used or adapted article and the original article is properly cited. The written permission of Cambridge University Press must be obtained prior to any commercial use.
Copyright
© The Author(s), 2023. Published by Cambridge University Press on behalf of European Consortium for Political Research

Introduction

Although built around core ideological principles or some other kind of shared purpose, political parties in parliamentary democracies are complex organizations. Their tent invariably includes diverse actors with different opinions on policy and strategy. Strong incentives exist to keep any associated intra-party conflict at bay: Excessive institutionalization of conflict in party factions, for instance, can threaten the survival of the entire organization (Boucek, Reference Boucek2009). Less dramatically, party disunity frustrates office, policy, and vote-maximizing ambitions (Sjöblom, Reference Sjöblom1968; Bäck, Reference Bäck2008; Pedersen, Reference Pedersen2012; Greene and Haber, Reference Greene and Haber2015; Ceron, Reference Ceron2019). But despite party leaders’ best attempts, internal conflict with respect to programmatic, ideological, and tactical standpoints is part of the day-to-day reality of political organizations.

While there is a rich literature about party unity in the parliamentary context (e.g., Kam, Reference Kam2009) and an increasing interest in intra-party relations beyond the legislative arena (Gherghina et al., Reference Gherghina, Close and Kopecký2019: 649), we only have a piecemeal understanding of how citizens perceive, process, and evaluate party (dis-)unity. While long a staple of theorizing (Sjöblom, Reference Sjöblom1968: 242–244), what has become more and more empirically substantiated in recent studies is that outright internal conflict hampers the electoral strategies and fortunes of a party substantially (Greene and Haber, Reference Greene and Haber2015; Barrett, Reference Barrett2018; Lehrer and Lin, Reference Lehrer and Lin2020, Reference Lehrer and Lin2022; Lin and Lehrer, Reference Lin and Lehrer2021; Lehrer et al., Reference Lehrer, Stöckle and Juhl2022). However, we do not yet know which intra-party matters even register as conflict in the voters’ mind.

When do voters perceive parties to be internally conflicted? Our paper makes two advances to answering this question. First, we sharpen the concept of intra-party conflict. Specifically, we see it as characterized by hostile internal relations rather than the mere existence of different opinions inside the party organization. Second, factors related to the periodic conduct of elections regulate voters’ perceptions by, we argue, inducing higher levels of intra-party conflict and making it more visible to citizens. We thus take a long-term ‘democratic life cycle’ (cf. Müller et al., Reference Müller, Bergman, Strøm, Strøm, Müller and Bergman2008) perspective, focusing on variation throughout and across legislative terms. This approach complements recent explanations that have privileged party characteristics or the short-term campaign context. These studies have shown that centrist parties are perceived to be less united (Zur, Reference Zur2021) and that voters register the fallout of intense party scandals unless they are partisan identifiers (Plescia et al., Reference Plescia, Kritzinger and Eberl2021).

We draw on a unique collection of surveys that details how citizens rate political parties with respect to internal conflict and encompasses data from 16 years of German politics. Crucially, we bring this perceptual data together with external indicators that correspond to factors inducing differences in (the visibility of) intra-party conflict across the democratic life cycle. Our analysis reveals that political parties are perceived to be more internally conflicted (1) when there is greater intra-party heterogeneity of opinion, (2) when they are in government, (3) when elections are temporally distant, and (4) when they have accumulated electoral losses recently. Importantly, the democratic life cycle imprints in this way on partisans and the general public alike. Taken together with the finding that perceived intra-party conflict has electoral drawbacks (Sjöblom, Reference Sjöblom1968; Greene and Haber, Reference Greene and Haber2015; Barrett, Reference Barrett2018; Lehrer and Lin, Reference Lehrer and Lin2020, Reference Lehrer and Lin2022; Lin and Lehrer, Reference Lin and Lehrer2021; Lehrer et al., Reference Lehrer, Stöckle and Juhl2022), these results have far-reaching implications for party competition. Specifically, they suggest the potential of self-reinforcing dynamics between election outcomes and citizens’ views of intra-party matters.

The remainder of the paper proceeds as follows: The next section synthesizes a conceptualization of party (dis-)unity from the literature that distinguishes intra-party conflict from electorally less deleterious kinds of disunity. We then outline our theoretical perspective regarding the democratic life cycle. It identifies factors associated with the periodic conduct of elections that imprint in the perceptions of the electorate by increasing the absolute amount of intra-party hostility and by heightening its visibility. After presenting the data sources and operationalization of the variables, the fifth section of the paper presents the analysis regarding the effect of the democratic life cycle on intra-party perceptions. Subsequently, we test whether partisans are insulated from these effects or whether the democratic life cycle imprints on them as on the general public. The concluding section summarizes the findings of the paper, draws out their implications, and highlights avenues for future research.

What is intra-party conflict?

Given its multi-dimensionality, it is curious that intra-party conflict is often simply equated with intra-organizational ideological distance (Gherghina et al., Reference Gherghina, Close and Kopecký2019: 650), despite the fact that not every difference of opinion is actually something to fight about. We build on the work of Lehrer and Lin (Reference Lehrer and Lin2020: 784–786) and argue that it is particularly helpful to think of intra-party conflict as characterized by hostility in intra-party relations. Essentially, we see party (dis-)unity as two-dimensional (Fig. 1): In terms of substance, claims made by actors belonging to the same party organization regarding policy, ideology, personnel, and strategy can be homogenous or heterogeneous (cf. Sjöblom, Reference Sjöblom1968: 183). In terms of valence, such claims can be brought forth with harmony or hostility.

Figure 1. Dimensions of party (dis-)unity: substance and valence.

Taking these two aspects separately is helpful because it clarifies that – while heterogeneity in substance is likely going hand in hand with hostility in valence – the two dimensions are conceptually separable. Theoretically, parties can experience vigorous and passionate infighting whenever there are at least two alternatives with respect to policy, personnel, or strategy. Reversely, different party actors can propose a multitude of positions on an issue without this being understood and framed as an internal conflict at all. Indeed, this separation helps us make sense of divergent findings in the existing literature and thereby situate our investigation into the broader scholarly context:

A party organization that is united in terms of being both homogeneous and harmonious displays cohesion in Sjöblom’s sense: It is a party in which there is ‘a rallying of the members round the output decided by the party’ (Sjöblom, Reference Sjöblom1968: 183 emphases added; cf. also 201). Party leaders have used their resources and power in selecting among the alternative positions existing inside the organization in such a way that it invigorates the members and (re-)establishes harmony in internal relations (Sjöblom, Reference Sjöblom1968: 203). But there can be (1) strategic reasons for why reducing heterogeneity is undesirable or (2) organizational reasons for why reducing hostility is impossible.

  1. (1) Being disunited in the sense of sending ambiguous, tailored, or even inconsistent messages can be electorally advantageous (e.g., Downs, Reference Downs1957; Page, Reference Page1976; Tomz and Van Houweling, Reference Tomz and Van Houweling2008; Rovny, Reference Rovny2012; Somer-Topcu, Reference Somer-Topcu2015; Lo et al., Reference Lo, Proksch and Slapin2016; Bräuninger and Giger, Reference Bräuninger and Giger2018), particularly if parties are not too obvious in catering simultaneously to different audiences (Sjöblom, Reference Sjöblom1968: 86; cf. Snyder and Ting, Reference Snyder and Ting2002; Martin, Reference Martin2019; Tromborg, Reference Tromborg2021) or engaging in outright flip-flopping (Nasr, Reference Nasr2023; Nyhuis and Stoetzer, Reference Nyhuis and Stoetzer2021). The phenomenon of party leaders deliberately staging or highlighting intra-party heterogeneity to simultaneously appeal to different electoral constituencies is known as the broad-appeal strategy (Somer-Topcu, Reference Somer-Topcu2015).

  2. (2) Being disunited in the sense of openly feuding with one another, in turn, undermines the effectivity of a broad appeal profoundly (Lehrer and Lin, Reference Lehrer and Lin2020, Reference Lehrer and Lin2022; Lin and Lehrer, Reference Lin and Lehrer2021). What is more, outright hostility in intra-party relations is of course not a deliberate strategy but the expression of power struggles over the appropriate course of the party. Such intra-party conflict signals to voters that the party lacks competence (Greene and Haber, Reference Greene and Haber2015) and is unable to carry out the policy proposals it promised. That is, it becomes questionable how responsive and accountable the party can be to its principal (cf. Bowler et al., Reference Bowler, Farrell, Katz, Bowler, Farrell and Katz1999: 3; Müller, Reference Müller2000).

We analyze here when political parties are perceived to be internally conflicted by the electorate. It is in particular hostility in intra-party relations that has been shown to have highly deleterious electoral consequences for parties in parliamentary democracies, using both, observational (Greene and Haber, Reference Greene and Haber2015; Barrett, Reference Barrett2018; Lehrer and Lin, Reference Lehrer and Lin2020, Reference Lehrer and Lin2022; Lin and Lehrer, Reference Lin and Lehrer2021) as well as experimental data (Lehrer et al., Reference Lehrer, Stöckle and Juhl2022; see also Duell et al., Reference Duell, Lea Kaftan, Slapin and Wratil2023).Footnote 1 But it is not only at election time that intra-party conflict matters. We know that politicians nowadays constantly refer to polls, care about their parties’ image in the electorate, and respond even when elections are not around the corner. Hager and Hilbig (Reference Hager and Hilbig2020), for example, have shown how politicians fine-tune their speeches to public opinion also during the legislative term. Moreover, party leaders need to be alert to signs of conflict in their organization and become active, in particular when the conflict becomes so visible that regular voters start to take note. This is because it is not clear that all damage done can be repaired within the confines of an election campaign and because conflict can spiral out of control and threaten the survival of the organization as such (Sjöblom, Reference Sjöblom1968: 205; cf. Boucek, Reference Boucek2009). It thus becomes crucial to understand when voters perceive parties to be internally conflicted. Previous research on this question has focused on short-term campaign dynamics (Plescia et al., Reference Plescia, Kritzinger and Eberl2021) and differences by party types (Zur, Reference Zur2021).

Intra-party conflict perceptions through the democratic life cycle

We argue that there are predictable regularities in voters’ perceptions of intra-party conflict that stem from the periodic conduct of elections. Factors that vary through the legislative term, as well as across legislative terms, incentivize higher levels of intra-party conflict and increase its visibility to the electorate. We refer to this as a ‘democratic life cycle perspective’, appropriating a concept developed in the study of government coalitions (cf. Müller et al., Reference Müller, Bergman, Strøm, Strøm, Müller and Bergman2008). By ‘democratic life cycle’ we mean the recursive progression of phases of competition, election, and government, where each phase is connected but also each sequence impacts the next (cf. Müller et al., Reference Müller, Bergman, Strøm, Strøm, Müller and Bergman2008: 8–12): Citizens’ perceptions and evaluations crystallize as a function of the competition within and between parties. They are sharpened as elections approach and they manifest themselves in electoral results as well as the governments they enable. At the same time, the latter have a recursive effect on citizens’ evaluations as the democratic life cycle begins anew.

Specifically, our theory starts from the presumption that citizens receive the bulk of information on which they base their perceptions of political parties as organizations from reporting in the media.Footnote 2 Different constellations throughout the democratic life cycle (pertaining to the internal heterogeneity of a party, its participation in government, the proximity of an election, and past electoral performance) systematically influence the level of hostility displayed between politicians within the same party as well as the extent of media coverage on the party (with a conflict frame). Thereby, the four factors reverberate in the perceptions that citizens develop of the parties (see Fig. 2).

Figure 2. The democratic life cycle and intra-party conflict perceptions.

It is important to note in this respect that the distinction between hostile intra-party relations and media reporting only carries so far: In terms of the classic distinction made by Sjöblom (Reference Sjöblom1968), politicians belonging to the same party organization can feud with one another in the internal arena (e.g., at a party convention), in the parliamentary arena (e.g., in a legislative debate or in the cabinet), and in the electoral arena (e.g., in the communication with voters at rallies). More often than not, however, conflict only becomes clear across these different arenas (e.g., when a faction of the party criticizes the parliamentary party group or when a subnational leader criticizes the national party leadership on social media). In an important sense, then, intra-party conflict only becomes manifest because the media brings together behavior that is scattered through diverse forums (cf. Somer-Topcu, Reference Somer-Topcu2015: 848). That is, the media provides a venue for, and frames interactions as, conflict in the first place (cf. Stanyer, Reference Stanyer2003).

First, we expect that higher intra-party heterogeneity is positively related to higher levels of perceived intra-party conflict. While we have presented heterogeneity in substance and hostility in valence as conceptually distinct dimensions of actual party (dis-)unity above, in practice there is likely a positive relationship between the two. Divergent policy positions are expressed by politicians within and across the internal, the parliamentary, and the electoral arenas (Kam, Reference Kam2009; Greene and Haber, Reference Greene and Haber2015, Reference Greene and Haber2016; Proksch and Slapin, Reference Proksch and Slapin2015; Ceron, Reference Ceron2019). The resulting variance of policy positions is sometimes deliberate, as part of a broad-appeal strategy, but more often reflects the fact that politicians, activists, and members hold genuinely different visions of the party’s mission to transform society and the state based on their own experiences, norms, and characteristics. Heterogeneity breeds hostility when the issue is so salient that the different policies considered are seen as conflicting, attempts of the leadership to solicit loyalty or enforce discipline are weak, and politicians, then, because they believe in a specific solution, or for reasons of personal advantage, frame and publicize their disagreement as such. In other cases, organizational subunits inside the party contest their own elections and use them as occasions to gain traction for their own visions or signal responsiveness to specific electorates (Müller, Reference Müller2013).

Journalists readily read conflict into heterogeneity. In general, they find ‘intraparty struggles very newsworthy […] [and] often play a proactive role in reconstructing internal party strife for audience consumption […] thus reestablishing the significance of their contribution’ (Stanyer, Reference Stanyer2003: 85). In this, the media is aided by a party’s competitors who will always be quick to dramatize any internal programmatic variance and further the party’s image as internally conflicted for their own electoral benefit (Robertson, Reference Robertson1976: 52). Overall, we thus expect that:

HYPOTHESIS 1: The more heterogeneity the party organization exhibits, the more likely the voters perceive the party to experience internal conflict.

Despite high levels of party unity in legislative voting, government parties are often particularly likely to be associated with intra-party conflict (cf. Sieberer, Reference Sieberer2006). On the one hand, governing parties actually experience more public feuding. This is because they have to deal with unanticipated events and constraints out of their control. A lower budget than expected, for instance, frustrates the aspirations of those in office. Governing parties thereby regularly antagonize activists and factions who demand the party makes good on its promises.

The problem is particularly severe for parties which govern in coalitions. Reflecting the need for compromise, coalition partners meet each other somewhere in the middle (Warwick, Reference Warwick2001) with no party being able to realize its true policy preference, irrespective of external constraints to governance. Coalition governments thus hardly implement the pure vision that supporters and particularly activists were rooting for. There is thus often a conflict between the government (the parliamentary arena in Sjöblom’s scheme) and parts of the party organization (who articulate their dissatisfaction mostly in the internal arena).

Even if governing per se would not lead to more actual intra-party conflict, the conflict that exists in governing parties is subject to more media attention and interpretation: Information on government parties is more easily accessible (Baumgartner and Chaqués Bonafont, Reference Baumgartner and Bonafont2015: 271). Feuding inside governing parties makes for a better story – the stability of government is at stake after all (Baum and Groeling, Reference Baum and Groeling2009; Kane, Reference Kane2020). Finally, governing parties are held to a different standard by voters (cf. Fortunato, Reference Fortunato2021) and journalists; governing parties are expected to be particularly competent and professional (in contrast to opposition parties). Signs of internal conflict are paid special attention because they present a mismatch with these expectations.

HYPOTHESIS 2: Government parties are more likely to be perceived as internally conflicted by voters compared to opposition parties.

Party leaders have an interest in minimizing hostility in intra-party relations. This desire is particularly pronounced around election day because leaders know that internal feuding is electorally harmful (Sjöblom, Reference Sjöblom1968; Ceron, Reference Ceron2019) and that a united appearance is a central advantage in the negotiations to join government coalitions (Bäck, Reference Bäck2008). Around election day party leaders thus implore members to engage in team-like behavior; this is the moment in which it really matters for everyone to pull on the same string, particularly in the communication to voters (the electoral arena). What is more, calls for unity are especially likely to fall on fruitful soil at this time. Even notorious rebels and trouble-makers – those politicians that are well known for frequently challenging the leadership out of conviction or to advance their career – have incentives to tone-down their demands in order not to weaken their party electorally; after all, they seek to influence the course of their party, not to sabotage it. To buttress their effort of communicating a harmonious image in the electoral arena, leaders aim to keep conflictual issues out of the parliamentary and internal arenas. The latter is particularly important for securing support from the party base in the electoral arena, i.e., mobilizing members and supporters to hang posters, canvass, and engage with citizens in market places.

While actual intra-organizational hostility thus tends to be lower during election times, another mechanism that will make a difference for voters’ perceptions is, again, media reporting. During election campaigns and in the direct aftermath of elections, there is a genuine story about inter-party competition: Journalists and voters gravitate toward the ‘horserace’ (Iyengar et al., Reference Iyengar, Norpoth and Hahn2004) and the question of who will take over power, given that most electoral systems do not produce single-party governments on a regular basis. During a legislative term, in contrast, stories of inter-party competition are less ready to come by. The focus of the media is more on government-opposition divides and, crucially, intra-party dynamics.

HYPOTHESIS 3: Parties are more likely to be perceived as internally conflicted by the voters, the further away in time an election is.

Perceived intra-party conflict is associated with poor electoral performance in the short run (Greene and Haber, Reference Greene and Haber2015; Barrett, Reference Barrett2018; Lehrer and Lin, Reference Lehrer and Lin2020, Reference Lehrer and Lin2022; Lin and Lehrer, Reference Lin and Lehrer2021; Lehrer et al., Reference Lehrer, Stöckle and Juhl2022) but the relationship is not unidirectional in the medium to long-run. That is, poor electoral performance also prompts intra-party conflict (Greene and Haber, Reference Greene and Haber2016; Schumacher and Elmelund-Præstekær, Reference Schumacher and Elmelund-Præstekær2018). An electoral loss is a welcome opportunity for internal rivals to seek party change and prompt challenges to the current leadership (Harmel et al., Reference Harmel, Heo, Tan and Janda1995; Budge et al., Reference Budge, Ezrow and McDonald2010). Whatever discontents exist and may have been swept under the rug now surface, first and foremost in the internal arena. Additionally, even if there had been no genuine ideological disagreements before, electoral losses will prompt actors to question the course of the party for electoral reasons. Such processes are rarely quiet affairs; because so much is at stake, conflict regularly accompanies it. For these dynamics to unleash, trends will matter more than singular events. While a party might stay united after a single lost election, losses that accumulate will trigger demand for more radical change (Marx and Schumacher, Reference Marx and Schumacher2013) and the attendant public disagreements.

Poor electoral performance is particularly likely to find reflection in the voter’s mind because the failure of the respective leader(s) or main candidate(s) and the ensuing attacks of their internal opponents make for a particularly good story. Because the notion that electoral success requires a united appearance is widespread, intra-party conflict makes for a handy lens through which party performance can be interpreted. Especially when electoral losses are accumulating, journalists are likely to seek out intra-party conflict. The narrative is simply more interesting than a complex story that would relay, say, poor electoral performance to the disintegration of cleavages due to social change. The media is thus likely to report and the voters likely to take notice.

HYPOTHESIS 4: The worse a party’s recent electoral performance, the more likely the voters will perceive the party to be internally conflicted.

Research design

We test these hypotheses on the case of Germany, which is particularly apt for three reasons. First, the pioneering studies on the electoral effects of perceived intra-party conflict have focused on Germany (Greene and Haber, Reference Greene and Haber2015; Lehrer et al., Reference Lehrer, Stöckle and Juhl2022). The detrimental effect of intra-party conflict is thus well-documented and also widely appreciated by journalists and politicians themselves.

Second, Germany is a typical parliamentary system of government in which its political parties are complex membership-based organizations rather than loose networks of candidates and supporters (cf. Katz, Reference Katz and Caramani2017). In that sense, we expect our findings to apply directly to other parliamentary systems. Indeed, the finding that intra-party conflict undermines the effectivity of the broad-appeal strategy, established for Germany originally as well (Lehrer and Lin, Reference Lehrer and Lin2020), travels easily to other parliamentary and semi-presidential democracies (Lin and Lehrer, Reference Lin and Lehrer2021; Lehrer and Lin, Reference Lehrer and Lin2022; cf. Jung and Somer-Topcu, Reference Jung and Somer-Topcu2022).

Third, the Politbarometer, a monthly opinion poll contains, among numerous other questions (Wüst, Reference Wüst2003), an explicit measure of voters’ perceptions of intra-party conflict for each of the major parties (CDU, CSU, SPD, FDP, Greens, PDS/The Left, AfD) encompassing a period of 16 years.Footnote 3 These surveys thus provide us with the unique opportunity to test the hypotheses related to the democratic life cycle. Indeed, and curiously given the high importance that analysts, observers, and even politicians accord to it, explicit measurements of intra-party conflict perceptions are rare, cross-nationally but also for individual countries. The few explicit measures for intra-party conflict that exist (Greene and Haber, Reference Greene and Haber2015; Barrett, Reference Barrett2018; Plescia et al., Reference Plescia, Kritzinger and Eberl2021) are limited to a single time point or a couple of months at most. We bring this data on the perceptions of citizens together with objective data on intra-party heterogeneity and other situational factors related to the democratic life cycle.

Dependent variable

The Politbarometer question we use to operationalize the dependent variable reads: ‘What do you think: Is the [party] rather feuding or rather united with respect to important political issues?’Footnote 4 While of course not all ambiguity can be ruled out, this question should pertain in particular to perceptions of intra-party conflict rather than perceptions of heterogeneity. This is because the German zerstritten semantically strongly implies hostility in everyday language and therefore triggers the respective considerations in the heads of respondents (Zaller, Reference Zaller1992). Additionally, a cursory look at media reporting about parties shows that zerstritten is used exactly in the sense of vigorous disagreement rather than harmonious coexistence of different opinions.Footnote 5 A similar indicator for intra-party conflict is used by Greene and Haber (Reference Greene and Haber2015) as well as Plescia et al. (Reference Plescia, Kritzinger and Eberl2021).

We translate the two answer options into a binary measure that takes a value of 1 if the respondent regards the respective party as being internally conflicted. The item has been included quite regularly, throughout different legislative terms as well as different political constellations, as Fig. 3 shows. It covers the coalition between the Social Democratic Party and the smaller Green Party, coalitions between both big parties – the Christian Democrats and the Social Democrats, and the intermittent coalition between the Christian Democrats and the smaller Liberal Party.Footnote 6 In total, our analyses are based on the intra-party conflict perceptions of about 60,000 respondents.Footnote 7 Appendix A.2 details the summary statistics of the dependent and independent variables.

Figure 3. Number of respondents by survey and governing parties at the time.

Notes: For each survey the number of respondents who are included in the final regression model is shown.

Deriving policy position variance from election manifestos

We need a measure of intra-party heterogeneity to test HYPOTHESIS 1. In Germany, a country ranking high on regional authority (Hooghe et al., Reference Hooghe, Marks and Schakel2010), programmatic differences between the different Land branches of the parties provide an especially fruitful soil for intra-party conflict (Detterbeck, Reference Detterbeck2012). Indeed, the different Land branches are important players vying for influence over the position and strategy of the federal party (Bäck et al., Reference Bäck, Debus and Klüver2016). Their different policy positions are reflected in both houses of parliament, taken up in the surrounding media discourse, and thus shape voters’ perception of the party as a whole. Importantly, the divergence of the political orientation of a party’s Land branches is not merely an adaptation to the specific political context but reflects the fact that ostensibly non-territorial intra-party factions are stronger in certain Land branches than others. For instance, the CDU in North Rhine-Westphalia has been described as the left wing of the entire party because of its Christian-social traditions and because of being home to the Christian Democratic Employees’ Association (Solar, Reference Solar, Kost, Rellecke and Weber2010: 295).

Despite the interest in politics at the subnational level, only a few attempts have been made to identify the positions of the German parties’ Land branches. In particular, there is no longitudinal dataset based on expert survey placements or in the tradition of the Manifesto Project. To create a direct measure for intra-party heterogeneity, we thus draw on positions estimated on the basis of election manifestos with the help of the text scaling method Wordscores (Laver et al., Reference Laver, Benoit and Garry2003). This method essentially compares the word frequencies of unknown texts with the word frequencies of so-called reference texts whose positions are known. The more similar a text is to one of the reference texts in terms of the used words, the closer it will be placed to the respective reference score.

While Wordscores has been successfully validated in the case of German (sub-)national manifestos (Hjorth et al., Reference Hjorth, Klemmensen, Hobolt, Hansen and Kurrild-Klitgaard2015; Bräuninger et al., Reference Bräuninger, Debus, Müller and Stecker2019, Reference Bräuninger, Debus, Müller and Stecker2020), it is important to note that its assumptions and output have weaknesses (e.g., Lowe, Reference Lowe2008). For instance, when different words are used to describe the very same policy, the estimated position will be somewhat different. We believe that for our purposes, this is less of a problem than in other applications. In line with the logic underlying the Wordscores technique, it is politically important when actors describe the same policy in different ways because it is an expression of different ideological justifications of the policy or different approaches to communication. Given that we use the estimated positions to get an estimate of heterogeneity within the parties, it is less of a problem that the technique not only catches different policy proposals but also differences in justification or style.

In a first step, the results of expert surveys on parties’ positions (Bräuninger et al., Reference Bräuninger, Debus, Müller and Stecker2020) provided reference scores on a general left-right dimension that were assigned to manifestos for national elections. In a second step, each word stem that appeared in at least one of these manifestos received a score, based on its relative frequency. These word scores, in turn, allowed for the estimation of Land parties’ positions based on the word usage in more than 550 Land party manifestos. For the individual parties, the estimated positions cluster but they also reveal considerable intra-party differences that correspond to qualitative and anecdotal evidence. The Bavarian Christian Democrats (CSU), for instance, take far more conservative positions on the order of society than most regional branches of the CDU. Similarly, the East German branches of the Socialist party (PDS/The Left) take positions that are clearly more centric than their West German counterparts.

We then operationalize the independent variable of HYPOTHESIS 1 as the overall programmatic divergence across Land party branches.Footnote 8 The measure itself is simply the standard deviation of the Land party branches’ programmatic left-right positions at the time of the respective survey respondent’s interview.Footnote 9 Note that, while any given party manifesto is a collective proposition of the respective party branch, in many cases the Land party branches have incentives to differentiate themselves from each other as well as the national party as their respective election approaches (cf. Klingelhöfer, Reference Klingelhöfer2016). Attacking other parts of their own party is often what gives them publicity and makes the strategy of differentiation work in the first place. The measure changes whenever one of the 16 Land party branches publicly changes its position by publishing a new electoral manifesto. This happens before every Land election which are held at least every 4 or 5 years, though at different dates.

Operationalizing the remaining independent variables

To test HYPOTHESIS 2, we create a dummy variable that indicates whether a party governed at the national level at the time when the respondent was asked to evaluate its level of internal conflict. For HYPOTHESIS 3, we make use of the timing of national elections. Here, we straightforwardly operationalize the variable as the absolute number of days between the survey interview and the closest federal election. For example, 5 days before the election, the variable takes the value five. If a respondent is being interviewed 10 days after the election, the variable takes the value 10.Footnote 10 HYPOTHESIS 4 maintains that the relationship between perceptions of intra-party conflict and electoral performance is bidirectional if we abstract away from the short term. To measure electoral performance, we calculated for each party the share of elections in the last 5 years in which the party lost votes (on either the federal or the Land level).Footnote 11 This procedure takes the multi-level nature of German party politics seriously and directly corresponds to the idea that losses need to accumulate for parties to engage in the full-blown infighting most noticeable to voters.

The observations in our dataset are not independent from each other. First, respondents were asked to provide an assessment of intra-party conflict for several parties at the same time. This suggests the possibility of unmeasured characteristics of respondents that would result in systematically lower (or higher) levels of perceived internal conflict. Second, evaluations given in the same survey might be more alike because of survey design or the respective political context. Accordingly, we utilize logistic regression with standard errors clustered by respondents and fixed effects allowing for systematic differences between surveys. By controlling for surveys, we also take account of common shocks as well as possible differences between East and West Germany given that the samples are drawn from one part of Germany exclusively.Footnote 12 We additionally include party dummy variables in order to control for the correlation between the branches of one and the same party (HYPOTHESIS 1) and because certain kinds of parties are generally perceived to be more or less united than others (Zur, Reference Zur2021). The Social Democrats (SPD) represent the reference category in the full model.Footnote 13

Analyzing the sources of intra-party conflict perceptions

We find that voters’ perceptions of intra-party conflict are indeed highly dynamic and systematically respond to the democratic life cycle. First and foremost, there is a consistent relationship between intra-party heterogeneity and perceived intra-party hostility. In line with HYPOTHESIS 1, the results of our logistic regression models (Table 1) show that respondents are more likely to perceive a party as internally conflicted, when the party’s organizational entities exhibit greater variance in terms of advanced policy positions as well as the way policies are described, justified, and communicated. According to our theory, divergent opinions inside a party go hand in hand with higher levels of hostility: While sometimes heterogeneity reflects a coordinated broad-appeal strategy, it is often an expression of party actors struggling for the right course of party and country. The media picks up these disputes because of the attractivity of the conflict frame (Stanyer, Reference Stanyer2003) and the public receives the signals.

Table 1. Determinants of voters’ intra-party conflict perceptions

Notes: Coefficients from logistic regression models with standard errors clustered by respondent in parentheses. The dependent variable reflects the internal conflict perception that individual respondents have of the political party they are asked to evaluate. Dummy variables for surveys and West/East Germany included but not shown. Statistical significance: *P < 0.01, **P < 0.001.

If a party governs on the national level, voters are more likely to perceive internal conflict compared to when it is in the opposition, all else equal. Our theory argues that government parties have more inner-party tensions as activists are unhappy with the compromises the coalition or the situation dictate. Additionally, the media scrutinizes the state of government parties more in general. This supports HYPOTHESIS 2. Note that the effect is also quite substantial: According to our model, the likelihood of being perceived as internally conflicted is 0.44 for opposition parties and 0.64 for governing parties. Taken together with the previous finding that perceived intra-party conflict makes voters less likely to vote for a party, this effect might thus contribute to the general pattern of government parties losing in elections. It is therefore not simply specific compromises on public policies themselves that are likely to alienate voters. Rather, the mere opportunity to compromise appears to be costly (Fortunato, Reference Fortunato2021).Footnote 14 It is important to note here that any party needs to compromise on policy purity once it comes into office but that the effect is likely stronger for parties that govern in coalitions compared to parties that govern individually.

In line with HYPOTHESIS 3, respondents tend to perceive more conflict, the further away the survey was conducted from election day, all else equal. Party leaders’ attempts to minimize conflict (Sjöblom, Reference Sjöblom1968) and members’ willingness to put disputes on hold thus seem to be somewhat successful.Footnote 15 Before elections, a feuding appearance hurts the party’s electoral chances. Directly after elections, intra-party conflict is harmful because it hinders the winning parties’ bargaining power in coalition talks (Bäck, Reference Bäck2008). However, the effect is not overly large, signaling that the leadership is not entirely successful and that not every rift can be repaired or concealed within the few weeks of the campaign.Footnote 16

Finally, our analysis provides evidence of the reciprocal relationship between elections, intra-party conflict, and citizens’ perceptions thereof. It is not simply the case that intra-party conflict hurts a party electorally in the short run. In the medium to long run, lost elections impact the perceptions of voters, especially if they are recurring.Footnote 17 We find that, as the share of lost elections increases, the likelihood of perceiving internal conflict increases, all else equal. According to our theory, an accumulation of electoral losses signals to competitors of the leadership a chance to overturn it. Additionally, intra-party disputes are a plausible frame for the media to make sense of a party’s electoral losses.Footnote 18

Supporting explanatory approaches based on party types (Zur, Reference Zur2021), our analysis also shows that there are considerable differences between parties. CDU and FDP are generally perceived to experience less internal conflict than the SPD. At the same time the Greens are perceived to experience more internal conflict. It seems like their often turbulent past still manifests itself in public perception. Overall, there is a tendency for left parties to be perceived as more internally conflicted (the coefficient for the FDP and The Left wavers somewhat between models). The clear exception to the pattern is the right-wing AfD, which is perceived to experience considerably more internal conflict than any other party. The AfD was a very new party in consolidation when the voters rated its conflict in our sample. At the time, there was marked internal dispute about personnel and policy (Bieber et al., Reference Bieber, Roßteutscher and Scherer2018: 441).

Our theory presumes that the four factors associated with the democratic life cycle imprint themselves in voters’ perceptions by heightening intra-party hostility and by increasing its salience in media reporting. Unfortunately, there are no measures available that specifically assess the objective level of hostility in intra-party relations (analogous to our objective measure of heterogeneity within a party) or systematically quantify media reporting on intra-party conflict for the long time period we are studying. This means that we cannot directly test the posited theoretical mechanisms (see Fig. 2). However, we provide indirect support by repeating the analysis, distinguishing those respondents that are politically interested from those that are not with a dummy variable and interacting it with each of the main independent variables. If the four factors operate in the way our theory posits, it is the politically interested that should be particularly responsive: These are the citizens likely to consume reporting about politics, be attentive to intra-party relations, and therefore take note of more (reported) hostile interactions, as induced at different stages of the democratic life cycle. This is exactly what we find (reported in Appendix A.7): The interaction effects between political interest and the four factors are positive and three out of four are statistically significant.

The democratic life cycle and partisanship

In line with ideas about the centrality of partisanship in political thinking (Campbell et al., Reference Campbell, Converse, Miller and Stokes1960) and motivated reasoning (e.g., Zaller, Reference Zaller1992; Lodge and Taber, Reference Lodge and Taber2013), voters who identify with a political party tend to perceive that party as more united (Plescia et al., Reference Plescia, Kritzinger and Eberl2021) and consider only the relative intra-party conflict they perceive for the opponent parties in their voting decisions (see Appendix, Figure A.1.1). This is because team-like behavior and atavistic thinking is central to partisanship (Achen and Bartels, Reference Achen and Bartels2016). Partisans care about the reputation of their ‘team’ in a manner that is at least somewhat insensitive to objective conditions. Accordingly, they tend to actively seek information that portrays opposing parties as divided (Kane, Reference Kane2020). Conversely, there is a tendency among partisans to dismiss arguments suggesting internal conflicts within their own party. Instead, any visible and hostile disagreements are often rationalized as ‘healthy’, ‘necessary’, and ‘democratic’ intra-party discussions.

Do party-colored glasses obscure voters’ perceptions of intra-party conflict as related to the democratic life cycle or do the relationships we identified materialize among partisans as well? To address this question, we include a binary variable in our regression model indicating whether the respondent ‘leans’ toward the party whose level of internal conflict is being evaluated as well as interactions with the central independent variables. The ‘leaning’-question is the standard measure of partisanship in the European context (cf. Barnes et al., Reference Barnes, Jennings, Inglehart and Farah1988; Dassonneville and Grieb, Reference Dassonneville and Grieb2018). We summarize the results of the analysis with the help of graphs that display how likely it is that a respondent perceives a certain party as being internally conflicted (Fig. 4; see also Appendix A.8 and A.9).

Figure 4. Predicted probabilities of perceiving intra-party conflict.

Notes: Shaded area represents the 95% confidence interval. Rugs show the empirical variance of the respective independent variable.

We find that the party toward which the respondent leans is generally perceived to be less internally conflicted. Citizens’ partisanship thus greatly influences their interpretation of the parties’ internal status. In a cheerleading manner, they judge out-parties to be more internally conflicted, while suppressing signs that their party is internally conflicted or rationalizing objective conflict in the name of intra-party democracy.

That being said, the analysis also shows that the variables associated with the democratic life cycle have largely the same effects for those that lean toward the evaluated party as for those that do not. Partisans update their perceptions of intra-party conflict with the dynamics of the political life cycle in much the same way as those who identify with other parties or with no party. As Fig. 4 shows, even for the perception of partisans, it makes a great difference whether a party is heterogeneous, governs, is soon contesting an election, or suffers from a losing streak. On the individual level, partisans’ perceptual screen lowers the general level of conflict that is perceived but does little to dim the periodic substantial increases related to the democratic life cycle.Footnote 19 On the electoral level, the continuous decrease of partisanship (Dassonneville and Grieb, Reference Dassonneville and Grieb2018: 50), further hinders the bulwark party leaders can depend on to dim the deleteriousness of intra-party conflict.

Conclusion

When do citizens perceive a party to be internally conflicted? Complementing previous studies that have shown differences by party type (Zur, Reference Zur2021) and recorded how partisanship colors citizens’ assessments in a campaign context (Plescia et al., Reference Plescia, Kritzinger and Eberl2021), we focused on the variation in (the accessibility of) party in-fighting induced by the democratic life cycle. Specifically, we argued that perceptions of intra-party conflict are predictably related to the periodic conduct of elections.

To evaluate this claim empirically, we brought together survey data on citizens’ perceptions of political parties with an objective measure of intra-party heterogeneity and situational factors that, we argue, induce internal conflict or make it more publicly visible. All in all, the data spans 16 years of German politics. We found that political parties are more likely to be perceived as internally conflicted (1) when there is greater variance of policy positions advocated among their organizational subunits, (2) when they are governing, (3) when election day is temporally distant, and (4) when they have consistently lost recent elections. These effects exist for both in- and out-partisans, despite biases in favor of their favorite party.

Our results are important in that they not only uncover the long-term regularities in citizens’ evaluations of political parties but also in that they imply the strong potential of self-reinforcing dynamics. Political parties have a harder time at the polls when voters perceive them to be internally conflicted. Yet, sustained electoral losses are associated with more perceived conflict as well. According to our theory, this is because electoral losses induce internal conflict that is visible to citizens. If parties want to pull out of this vicious cycle, they somehow have to find a way to maintain or reestablish harmony in the face of electoral defeat. It stands to reason that this is more easily achieved with a ‘new beginning’ rather than the losing constellation of actors and issues. A change of the leadership after lost elections thus not only serves the ambitions of the rivals of the old guard but carries the potential of veiling past disputes, at least if the new settlement does not have too narrow support.

Other regularities of the democratic life cycle similarly come with ‘penalties’: Being in government will predictably increase the internal conflict citizens perceive. To the degree that this is due simply to increased media attention rather than actually increased conflict inside the party, it poses a serious and unfair disadvantage for government parties and should be added to the ‘costs of ruling’ (Stevenson, Reference Stevenson2002; cf. also Fortunato, Reference Fortunato2021). Note in this context again, that even supporters are not insulated from exhibiting the bias when their party is ruling. While they always rate out-parties as more internally conflicted than their own, partisans are swayed by the same forces as the general public. An additional danger is that partisans who perceive their party to be more internally conflicted (potentially because the media portrays it as such) might well start behaving themselves as someone who is part of that conflict.

Importantly, however, not every kind of disunity is automatically detrimental. While more intra-party heterogeneity is generally associated with perceived intra-party conflict, this does not negate the possibility of more strategic heterogeneity, as when parties aim to appeal broadly (Somer-Topcu, Reference Somer-Topcu2015). Our theory and analyses – while not in a place to formally assess the magnitude of the problem – suggest that party strategists need to be aware of the possibility that voters might ‘get it wrong’. For instance, some voters, and indeed some journalists (Stanyer, Reference Stanyer2003), might infer conflict from heterogeneity, even if there is no outright hostility present. The problem is confounded by out-partisans who profess to an upwardly biased level of hostility while they are exactly one of the target groups the broad-appeal strategy is designed to win over. In any case, those who envision a broad-appeal strategy need to ensure, in addition to distributing divergent positions in a way so that the heterogeneity is less visible (Tromborg, Reference Tromborg2021), that they are only increasing heterogeneity in the party’s communications with voters and not give any indication that there is increased hostility (Lehrer and Lin, Reference Lehrer and Lin2020; Lin and Lehrer, Reference Lin and Lehrer2021).

Promising avenues for future work pertain to (1) comparative work, leveraging different institutional setups and different timing with respect to the democratic life cycle, (2) an exploration of additional party and citizen characteristics – regarding, for instance, the salience of issues on which there is intra-party conflict or personality-based differences in the tolerance for conflict – as well as (3) a direct test of the theoretical mechanisms we posited. The latter requires both an analysis of how the media creates and portrays intra-party conflict and more experimental work that identifies how intra-party conflict perceptions form, as well as what exactly it is about internal conflict that turns voters off. Are voters confused about what they will get from a party when they perceive it as internally conflicted? Or do they reward teamwork for its own sake? Or are we merely witnessing a kind of self-fulfilling prophecy in which the media reports on intra-party conflict, interpreting it as deleterious for a party’s electoral chances, and voters merely take up the criterion? Indeed, much of the variance in citizens’ perceptions of political parties remains to be explained. As we have shown here, however, the patterns and dynamics induced by the periodic conduct of elections cannot be disregarded.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S1755773923000243.

Replication Package

Replication data will be made available on the Harvard Dataverse Repository at https://doi.org/10.7910/DVN/KJW6NV.

Acknowledgments

Precursory ideas and analyses leading to this paper were presented at CES and EPSA conferences as well as workshops in Berlin, Oldenburg, and Solstrand. For their insightful questions and comments during these discussions, we are particularly indebted to Nicole Rae Baerg, Nicolai Berk, Thomas Bräuninger, Adam Bonica, Christopher Claassen, Royce Carroll, Andrea Ceron, Mike Cowburn, Thomas Däubler, Marc Debus, David Fortunato, Simon Franzmann, Christina Gahn, Zac Greene, Michael Jankowski, Arndt Leininger, Roni Lehrer, Sandra León, Elin Monstad, Yvette Peters, Simona Piattoni, Carolina Plescia, Marius Sältzer, Arjan Schakel, Zeynep Somer-Topcu, Christian Stecker, Michaël Tatham, Markus Tepe, Elisa Volpi, and Thomas Zittel. The poignant feedback of reviewers and careful guidance of the editors also improved the paper substantially.

Financial support

No funding was received for conducting this study.

Competing interests

The authors have no competing interests to declare that are relevant to the content of this article.

Footnotes

1 We further bolster this emergent finding with our data in the online Appendix (A.1).

2 The exception are those citizens that are formal and at least somewhat active party members who can engage in direct observation. These represent a miniscule share of the electorate.

4 ‘Was meinen Sie: Ist die [Partei] in wichtigen politischen Fragen eher zerstritten oder eher einig?’

6 While Christian and Social Democrats both polled around 40% at the beginning of the time period, there was tendency toward fragmentation of the party system. More information on the general patterns of political competition can be found in Appendix A.1 as well as Debus (Reference Debus2007) and Faas (Reference Faas2015).

7 The sets of parties that respondents evaluate partially differ across surveys.

8 We provide a successful validation of our estimates of party positions with the Politbarometer data in Appendix A.3.

9 Positions derived from manifestos are assigned to Land party branches from 4 weeks before the respective election, i.e., the time when (new) positions are more offensively advertised and the public is more likely to take notice.

10 To allow for comparisons of the coefficients included in the regression models, values reflect the number of days divided by 360. Additional analysis (Appendix A.5) show that the results are robust for alternative operationalizations of the distance to elections.

11 The variable could take values between 0 and 1, where larger values mean that more elections were lost. We observe the highest value of 0.95 (see Appendix A.2) for the Greens, which in late 2002 had lost 20 out of the last 21 elections.

12 In three cases the samples include East and West Germans simultaneously. In order to avoid any biases, we treat them as separate surveys and include two dummy variables distinguishing the surveys that were carried out in 2013, 2014 and 2015.

13 Respondents were asked to evaluate the internal conflict in the CSU separately from its sister party, the CDU. The party is not included in the analysis of the effects of internal programmatic variation and the share of lost elections because we do not have meaningful measures of these concepts for the CSU (it only stands for election in Bavaria). This makes for a conservative test of HYPOTHESIS 1 and HYPOTHESIS 4 compared to treating the CSU as a branch of CDU – because of its particularly visible standing, distinct positions, and partly diverging electoral base. In the models that test only for the effect of government participation and the distance to elections, the CSU evaluations are included. Dropping them or adding an additional CSU dummy does not change the results (Appendix A.10).

14 Senior and junior coalition members are both (and to a similar degree) perceived to be more internally conflicted compared to opposition parties (cf. Klüver and Spoon, Reference Klüver and Spoon2020) (Appendix A.4).

15 Elections are not the only time when there is a particular interest in minimizing internal conflict. Assuming that party leaders want to remain in their position suggests that the same holds for party conventions, i.e., where the leadership has to be formally accountable to the members. In Appendix A.6, we show that parties are perceived to be more conflicted the closer in time a party convention or a leadership change is. In contrast to the four factors that we focus on in the main analysis, these variables are particularly endogenous. That is, often a convention is called or a previous leader deposed precisely because the level of visible conflict has become untenable.

16 In the German case, there are elections on multiple levels of the political system. In Appendix A.5, we test the effect of the distance to the next respective Land election. The coefficient is positive but not statistically significant. While Land party leaders generally have an interest in a united appearance of their Land branch as well as support from the national party for their campaigns, they in some cases also “look for a fight” with the national level (and thereby increase the perceived intra-party conflict in the overall party) to profit in the Land elections. We believe that the insignificant coefficient reflects these contrasting tendencies.

17 One might wonder if sustained losses, over time, make it less likely for a party to end up in government. Note in this context, however, that our operationalization includes electoral losses on the Land level. Additionally, in the German case, where minimal winning coalitions are the norm, it is not clear that electoral losses necessarily go hand in hand with being in the opposition. In the sense that electoral losses will divide the party and undermine its ability to enter government, however, it will ease the costs of ruling and thereby facilitate attempts at re-establishing harmony in the opposition.

18 An alternative interpretation would be that a string of losses confers upon the party a losing image and therefore leads citizens to perceive there to be more conflict without any actual increase in (the visibility of) intra-party hostility. While we cannot formally adjudicate the relative importance of the two perspectives with the data at hand, we see the findings of previous studies regarding how sustained electoral loss triggers intra-party conflict as indicating that our interpretation cannot be neglected (Marx and Schumacher, Reference Marx and Schumacher2013; Schumacher and Elmelund-Præstekær, Reference Schumacher and Elmelund-Præstekær2018).

19 Of the interactive terms with partisanship, the ones for policy position variance and electoral loss reach statistical significance. However, the difference in slopes is substantively meaningless, as Plots a and d also indicate.

References

Achen, Christopher H., and Bartels, Larry M.. Democracy for Realists: Why Elections Do Not Produce Responsive Government. Princeton, NJ: Princeton UP, 2016.CrossRefGoogle Scholar
Bäck, Hanna. “Intra-Party Politics and Coalition Formation: Evidence from Swedish Local Government.” Party Politics 14.1 (2008): 7189.CrossRefGoogle Scholar
Bäck, Hanna, Debus, Marc, and Klüver, Heike. “Bicameralism, Intra-Party Bargaining, and the Formation of Party Policy Positions: Evidence from the German Federal System.” Party Politics 22.3 (2016): 405–17.CrossRefGoogle Scholar
Barnes, Samuel H., Jennings, M. Kent, Inglehart, Ronald, and Farah, Barbara. “Party Identification and Party Closeness in Comparative Perspective.” Political Behavior 10.3 (1988): 215–31.CrossRefGoogle Scholar
Barrett, David J. The Effect of Observable Party Cohesion on Voter Choice in Liberal Democracies, Thesis, Trinity College Dublin, School of Social Sciences & Philosophy, Discipline of Political Science, 2018. <http://www.tara.tcd.ie/handle/2262/82676> (Accessed September 10, 2022).Google Scholar
Baum, Matthew A., and Groeling, Tim. “Shot by the Messenger: Partisan Cues and Public Opinion Regarding National Security and War.” Political Behavior 31.2 (2009): 157–86.CrossRefGoogle Scholar
Baumgartner, Frank R., and Bonafont, Laura Chaqués. “All News Is Bad News: Newspaper Coverage of Political Parties in Spain.” Political Communication 32.2 (2015): 268–91.CrossRefGoogle Scholar
Bieber, Ina, Roßteutscher, Sigrid, and Scherer, Philipp. “Die Metamorphosen der AfD-Wählerschaft: Von einer euroskeptischen Protestpartei zu einer (r)echten Alternative?.” Politische Vierteljahresschrift 59.3 (2018): 433–61.CrossRefGoogle Scholar
Boucek, Françoise. “Rethinking Factionalism: Typologies, Intra-Party Dynamics and Three Faces of Factionalism.” Party Politics 15.4 (2009): 455–85.CrossRefGoogle Scholar
Bowler, Shaun, Farrell, David M., and Katz, Richard S.. “Party Cohesion, Party Discipline, and Parliaments.” In Bowler, Shaun, Farrell, David M., and Katz, Richard S. (Eds.), Party Discipline and Parliamentary Government. Columbus, OH: Ohio State UP, 1999, pp. 322.Google Scholar
Bräuninger, Thomas, Debus, Marc, Müller, Jochen, and Stecker, Christian. “Party Competition and Government Formation in Germany: Business as Usual or New Patterns?German Politics 28.1 (2019): 80100.CrossRefGoogle Scholar
Bräuninger, Thomas, Debus, Marc, Müller, Jochen, and Stecker, Christian. Parteienwettbewerb in den deutschen Bundesländern. Wiesbaden, Germany: Springer VS, 2020.CrossRefGoogle Scholar
Bräuninger, Thomas, and Giger, Nathalie. “Strategic Ambiguity of Party Positions in Multi-Party Competition.” Political Science Research and Methods 6.3 (2018): 527–48.CrossRefGoogle Scholar
Budge, Ian, Ezrow, Lawrence, and McDonald, Michael D.. “Ideology, Party Factionalism and Policy Change: An Integrated Dynamic Theory.” British Journal of Political Science 40.4 (2010): 781804.CrossRefGoogle Scholar
Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E.. The American Voter. Unabridged Edition. Chicago, IL: U of Chicago P, 1960.Google Scholar
Ceron, Andrea. Leaders, Factions and the Game of Intra-Party Politics. Abingdon, UK: Routledge, 2019.CrossRefGoogle Scholar
Dassonneville, Ruth, and Grieb, Annika. “Partisanship and Political Sophistication During and Outside of Election Times: A Longitudinal Analysis Using the German SOEP Survey 1984–2012.” German Politics 27.1 (2018): 4465.CrossRefGoogle Scholar
Debus, Marc. “Bestimmungsfaktoren des Wahlverhaltens in Deutschland bei den Bundestagswahlen 1987, 1998 und 2002: Eine Anwendung des Modells von Adams, Merrill und Grofman.” Politische Vierteljahresschrift 48.2 (2007): 269–92.CrossRefGoogle Scholar
Detterbeck, Klaus. Multi-Level Party Politics in Western Europe. Basingstoke, UK: Palgrave Macmillan, 2012.CrossRefGoogle Scholar
Downs, Anthony. An Economic Theory of Democracy. New York, NY: Harper & Row Publishers, 1957.Google Scholar
Duell, Dominik, Lea Kaftan, Sven-Oliver Proksch, Slapin, Jonathan, and Wratil, ChristopherCommunicating the Rift: Voter Perceptions of Intraparty Dissent in Parliaments.” The Journal of Politics 85.1 (2023): 76–91.CrossRefGoogle Scholar
Faas, Thorsten. “The German Federal Election of 2013: Merkel’s Triumph, the Disappearance of the Liberal Party, and Yet Another Grand Coalition.” West European Politics 38.1 (2015): 238–47.CrossRefGoogle Scholar
Fortunato, David. The Cycle of Coalition: How Parties and Voters Interact Under Coalition Governance. Cambridge, UK: Cambridge UP, 2021.CrossRefGoogle Scholar
Gherghina, Sergiu, Close, Caroline, and Kopecký, Petr. “The Dynamics and Dimensions of Intra-Party Conflict: Introduction to the Special Issue.” Party Politics 25.5 (2019): 649–51.CrossRefGoogle Scholar
Greene, Zachary D., and Haber, Matthias. “The Consequences of Appearing Divided: An Analysis of Party Evaluations and Vote Choice.” Electoral Studies 37 (2015): 1527.CrossRefGoogle Scholar
Greene, Zachary D., and Haber, Matthias. “Leadership Competition and Disagreement at Party National Congresses.” British Journal of Political Science 46.3 (2016): 611–32.CrossRefGoogle Scholar
Hager, Anselm, and Hilbig, Hanno. “Does Public Opinion Affect Political Speech?American Journal of Political Science 64.4 (2020): 921–37.CrossRefGoogle Scholar
Harmel, Robert, Heo, Uk, Tan, Alexander, and Janda, Kenneth. “Performance, Leadership, Factions and Party Change: An Empirical Analysis.” West European Politics 18.1 (1995): 133.CrossRefGoogle Scholar
Hjorth, Frederik, Klemmensen, Robert, Hobolt, Sara, Hansen, Martin Ejnar, and Kurrild-Klitgaard, Peter. “Computers, Coders, and Voters: Comparing Automated Methods for Estimating Party Positions.” Research & Politics 2.2 (2015): 19.CrossRefGoogle Scholar
Hooghe, Liesbet, Marks, Gary, and Schakel, Arjan H.. The Rise of Regional Authority A Comparative Study of 42 Democracies. Abingdon, UK: Routledge, 2010.CrossRefGoogle Scholar
Iyengar, Shanto, Norpoth, Helmut, and Hahn, Kyu S.. “Consumer Demand for Election News: The Horserace Sells.” The Journal of Politics 66.1 (2004): 157–75.CrossRefGoogle Scholar
Jung, Jae-Hee, and Somer-Topcu, Zeynep. “United We Stand, Divided We Fall? The Effects of Parties’ Brexit Rhetoric on Voters’ Perceptions of Party Positions.” Journal of Elections, Public Opinion and Parties 32.3 (2022): 596614.CrossRefGoogle Scholar
Kam, Christopher J. Party Discipline and Parliamentary Politics. Cambridge, UK: Cambridge UP, 2009.CrossRefGoogle Scholar
Kane, John V.Fight Clubs: Media Coverage of Party (Dis)unity and Citizens’ Selective Exposure to It.” Political Research Quarterly 73.2 (2020): 276–92.CrossRefGoogle Scholar
Katz, Richard S.Political Parties.” In Caramani, Daniele (Ed.), Comparative Politics. Oxford, UK: Oxford UP, 2017, pp. 207223.Google Scholar
Klingelhöfer, Tristan. “Ensuring Consistency across Levels? The Delegation Model of Multi-Level Party Politics and Spanish Framework Manifestos.” Party Politics 22.4 (2016): 452–64.CrossRefGoogle Scholar
Klüver, Heike, and Spoon, Jae-Jae. “Helping or Hurting? How Governing as a Junior Coalition Partner Influences Electoral Outcomes.” The Journal of Politics 82.4 (2020): 1231–42.CrossRefGoogle Scholar
Laver, Michael, Benoit, Kenneth, and Garry, John. “Extracting Policy Positions from Political Texts Using Words as Data.” The American Political Science Review 97.2 (2003): 311–31.CrossRefGoogle Scholar
Lehrer, Roni, and Lin, Nick. “Everything to Everyone? Not When You Are Internally Divided.” Party Politics 26.6 (2020): 783–94.CrossRefGoogle Scholar
Lehrer, Roni, and Lin, Nick. Perceptions of Party Ambiguity and their Electoral Consequences. Mannheim, Frankfurt, and Taipei: Mimeo, 2022.Google Scholar
Lehrer, Roni, Stöckle, Pirmin, and Juhl, Sebastian. “Assessing the Relative Influence of Party Unity on Vote Choice: Evidence from a Conjoint Experiment.” Political Science Research and Methods, 2022, forthcoming.Google Scholar
Lin, Nick, and Lehrer, Roni. “Everything to Everyone and the Conditioning Effect of Intraparty Cohesion: A Replication in a Cross-National Context.” Party Politics 27.5 (2021): 909–16.CrossRefGoogle Scholar
Lo, James, Proksch, Sven-Oliver, and Slapin, Jonathan B.. “Ideological Clarity in Multiparty Competition: A New Measure and Test Using Election Manifestos.” British Journal of Political Science 46.3 (2016): 591610.CrossRefGoogle Scholar
Lodge, Milton, and Taber, Charles S.. The Rationalizing Voter. Cambridge: Cambridge UP, 2013.CrossRefGoogle Scholar
Lowe, Will. “Understanding Wordscores.” Political Analysis 16.4 (2008): 356–71.CrossRefGoogle Scholar
Martin, Danielle. “Party Ambiguity and Individual Preferences.” Electoral Studies 57 (2019): 1930.CrossRefGoogle Scholar
Marx, Paul, and Schumacher, Gijs. “Will to Power? Intra-Party Conflict in Social Democratic Parties and the Choice for Neoliberal Policies in Germany, The Netherlands and Spain (1980–2010).” European Political Science Review 5.1 (2013): 151–73.CrossRefGoogle Scholar
Müller, Jochen. “On a Short Leash? Sub-National Party Positions between Regional Context and National Party Unity.” Journal of Elections, Public Opinion and Parties 23.2 (2013): 177–99.CrossRefGoogle Scholar
Müller, Wolfgang C.Political Parties in Parliamentary Democracies: Making Delegation and Accountability Work.” European Journal of Political Research 37.3 (2000): 309–33.CrossRefGoogle Scholar
Müller, Wolfgang C., Bergman, Torbjörn, and Strøm, Kaare. “Coalition Theory and Cabinet Governance: An Introduction.” In Strøm, Kaare, Müller, Wolfgang C., and Bergman, Torbjörn (Eds.), Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe, Oxford, UK: Oxford UP, 2008, pp. 150.Google Scholar
Nasr, Mohamed. “Varieties of Ambiguity: How Do Voters Evaluate Ambiguous Policy Statements?Comparative Political Studies 56.6 (2023): 759–87.CrossRefGoogle Scholar
Nyhuis, Dominic, and Stoetzer, Lukas F.. “The Two Faces of Party Ambiguity: A Comprehensive Model of Ambiguous Party Position Perceptions.” British Journal of Political Science 51.4 (2021): 1421–38.CrossRefGoogle Scholar
Page, Benjamin I.The Theory of Political Ambiguity.” American Political Science Review 70.3 (1976): 742–52.CrossRefGoogle Scholar
Pedersen, Helene H.What Do Parties Want? Policy versus Office.” West European Politics 35.4 (2012): 896910.CrossRefGoogle Scholar
Plescia, Carolina, Kritzinger, Sylvia, and Eberl, Jakob-Moritz. “The Enemy Within’: Campaign Attention and Motivated Reasoning in Voter Perceptions of Intra-party Conflict.” Party Politics 27.5 (2021): 917–27.CrossRefGoogle Scholar
Proksch, Sven-Oliver, and Slapin, Jonathan B.. The Politics of Parliamentary Debate: Parties, Rebels, and Representation. Cambridge, UK: Cambridge UP, 2015.Google Scholar
Robertson, David. A Theory of Party Competition. London, UK: Wiley, 1976.Google Scholar
Rovny, Jan. “Who Emphasizes and Who Blurs? Party Strategies in Multidimensional Competition.” European Union Politics 13.2 (2012): 269–92.CrossRefGoogle Scholar
Schumacher, Gijs, and Elmelund-Præstekær, Christian. “Party Performance Explains Disagreement between Politicians and Their Parties.” West European Politics 41.2 (2018): 329–49.CrossRefGoogle Scholar
Sieberer, Ulrich. “Party Unity in Parliamentary Democracies: A Comparative Analysis.” The Journal of Legislative Studies 12.2 (2006): 150–78.CrossRefGoogle Scholar
Sjöblom, Gunnar. Party Strategies in a Multiparty System. Lund, Sweden: Studentlitteratur, 1968.Google Scholar
Snyder, James M. Jr., and Ting, Michael M.. “An Informational Rationale for Political Parties.” American Journal of Political Science 46.1 (2002): 90110.CrossRefGoogle Scholar
Solar, Marcel. “Nordrhein-Westfalen – das Erbe des politischen Katholizismus und der Mythos vom sozialdemokratischen Stammland.” In Kost, Andreas, Rellecke, Werner, and Weber, Reinhold (Eds.), Parteien in den deutschen Ländern: Geschichte und Gegenwart. München, Germany: C.H. Beck, 2010, pp. 275301.Google Scholar
Somer-Topcu, Zeynep. “Everything to Everyone: The Electoral Consequences of the Broad-Appeal Strategy in Europe.” American Journal of Political Science 59.4 (2015): 841–54.CrossRefGoogle Scholar
Stanyer, James. “Intraparty Conflict and the Struggle to Shape News Agendas: Television News and the Coverage of the Annual British Party Conferences.” Harvard International Journal of Press/Politics 8.2 (2003): 7189.CrossRefGoogle Scholar
Stevenson, Randolph T.The Cost of Ruling, Cabinet Duration, and the ‘Median-Gap’ Model.” Public Choice 113.1 (2002): 157–78.CrossRefGoogle Scholar
Tomz, Michael, and Van Houweling, Robert P.. “Candidate Positioning and Voter Choice.” The American Political Science Review 102.3 (2008): 303–18.CrossRefGoogle Scholar
Tromborg, Mathias W.Appealing Broadly While Appearing Unified: Resolving an Electoral Dilemma.” European Journal of Political Research 60.1 (2021): 131–52.CrossRefGoogle Scholar
Warwick, Paul V.Coalition Policy in Parliamentary Democracies: Who Gets How Much and Why?Comparative Political Studies 34.10 (2001): 1212–36.CrossRefGoogle Scholar
Wüst, Andreas M. (Ed.) Politbarometer. Opladen, Germany: Leske + Budrich, 2003.CrossRefGoogle Scholar
Zaller, John R. The Nature and Origins of Mass Opinion. New York, NY: Cambridge UP, 1992.CrossRefGoogle Scholar
Zur, Roi. “The Multidimensional Disadvantages of Centrist Parties in Western Europe.” Political Behavior 43.4 (2021): 1755–77.CrossRefGoogle Scholar
Figure 0

Figure 1. Dimensions of party (dis-)unity: substance and valence.

Figure 1

Figure 2. The democratic life cycle and intra-party conflict perceptions.

Figure 2

Figure 3. Number of respondents by survey and governing parties at the time.Notes: For each survey the number of respondents who are included in the final regression model is shown.

Figure 3

Table 1. Determinants of voters’ intra-party conflict perceptions

Figure 4

Figure 4. Predicted probabilities of perceiving intra-party conflict.Notes: Shaded area represents the 95% confidence interval. Rugs show the empirical variance of the respective independent variable.

Supplementary material: Link

Klingelhöfer and Müller Dataset

Link
Supplementary material: PDF

Klingelhöfer and Müller supplementary material

Klingelhöfer and Müller supplementary material

Download Klingelhöfer and Müller supplementary material(PDF)
PDF 566.8 KB