A Question of Commitment: Investigating How Citizens Perceive Parties' Programmatic Responses to Competition

Abstract

Political parties vary in their responses to electoral challenges, including the rise of populist competitors. To address these challenges, they sometimes engage with peripheral issues located outside their ideological comfort zones, and at other times they adhere more closely to their core policies. Although these patterns are well-documented, voters' perceptions thereof remain under-examined. This article argues that voters evaluate parties' claims not just based on the direction of their policy engagement – positive or negative – but also based on the commitment behind these actions, distinguishing ideological commitment from strategic manoeuvres. Employing a pre-registered vignette experiment, the article shows that voters differentiate responses to core versus peripheral issues, regardless of their personal agreement with the policies. Populist attitudes further moderate these perceptions, as voters with such views are typically more sceptical of parties' motives, limiting the impact of party behaviour on perceived commitment primarily to non-populist individuals. This highlights the importance of perceived commitment in elections and the constraints parties face in responding to competition.

The rise of populist actors over the past decades has compelled political parties to address new voter demands and adapt to changes in their competitive environment (Kriesi 2014; Mudde 2004), leading parties to engage with policy issues often located outside their ideological ‘comfort zones’. Despite these pressures, long-standing policy goals and inert party–voter linkages often force parties to maintain a tight grip on their core policy issues – that is, those domains that are of primary importance for their voter mobilization (Budge 1994; Somer-Topcu 2009). Existing research has extensively documented how parties navigate these challenges and respond to changes in competition, by adapting their issue profiles or their issue positions (Vries and Hobolt 2020). These strategies commonly range from parties reinforcing their long-time policy positions and countering the claims of their opponents to alignment with and accommodation of their rivals' claims (e.g. Meguid 2005, 2008).

Given the far-reaching policy consequences of such decisions, a wealth of research has examined the policy strategies of parties in response to new competitors (Habersack 2024a; Hobolt and Vries 2015; Vries and Hobolt 2020; Wolinetz and Zaslove 2018), generally testifying to the ‘contagious’ policy impact (Abou-Chadi and Krause 2020; Habersack and Werner 2022; Meijers 2017; Spoon et al. 2014) and discursive influence (Breyer 2023; Hameleers and Vliegenthart 2020; Mudde 2004) of radical and populist challenger parties on their opponents. Furthermore, previous research has investigated the effectiveness of various programmatic strategies in keeping challenger parties at bay (Dahlström and Sundell 2012; Hjorth and Larsen 2022; Krause et al. 2022). However, there remains a dearth of knowledge when it comes to the question of how voters evaluate the sincerity of parties' actions and rhetoric.

While a rich body of literature focuses on vote choice and party sympathy to gauge the extent to which parties' responses to competition resonate with voters, a crucial aspect has remained underexplored: the level of commitment that these forms of policy engagement convey. Thus, to what degree do voters perceive parties' behavioural choices as ideologically genuine compared to strategically driven by electoral gain, and what influences these perceptions?

Although this issue is commonly highlighted during electoral debates, which often centre on political actors' motivations for specific policy statements, the literature has not sufficiently explored how voters themselves perceive these motivations. In this study, I argue that voters' perceptions of commitment are significantly shaped by whether policy adaptations concern core or mere peripheral areas of a party's ideological profile (Beyme 1985; Mair and Mudde 1998; Ware 1996). Specifically, when changes are made that affect core policy domains – those fundamental to a party's identity and traditional voter connections – voters are more likely to perceive these changes as a reflection of genuine commitment. This perception holds regardless of voters' personal policy convictions or the problem-solving competency they attribute to the party in the specific domain. Conversely, adaptations in peripheral policy areas, which are less central to the party's ideological profile, are more likely seen as manoeuvres induced by mere changes in parties' electoral strategies.

To test these claims, this study employs a vignette experiment included in a representative online survey fielded in Austria in March 2023 with a sample size of N = 2,013. The experiment confronts respondents with a semi-hypothetical electoral contest, fought between the respondent's struggling in-party and a rising challenger party which is portrayed as taking the exact opposite position on a selected policy issue. Exposing respondents to various approaches taken by their in-party in this scenario, I find that voters tend to view policy claims that reinforce their party's position in the ideological spectrum as more committed. However, I also find that, as expected, this perceived ideological commitment is contingent on whether core or peripheral policies are concerned. Independent of other factors such as the policy direction or justifications provided, core policy claims are seen as more committed than claims touching on peripheral issues.

These results suggest that parties often have limited means of fundamentally altering their policy positions and testify to voters' ability to discern between more and less committed forms of party behaviour. Populist attitudes exacerbate this issue, meaning that since voters who hold populist attitudes are generally wary of parties' motivations (Geurkink et al. 2020) and driven by ingrained ideas about ideological positions (Voogd and Dassonneville 2020) – that is, what needs to happen and who is to blame for why this change is not forthcoming – parties face a harder time justifying the need for positional adaption and addressing such voters.

The results further reveal that parties only have limited capacities to convince voters of their intentions through means of discursive justification. How a claim is justified does not fundamentally impact perceived commitment. This, however, comes as good news for representative democracy as voters evidently care first and foremost about policy direction and are unlikely to be deceived by mere rhetoric.

The findings speak to the literature on populism and party competition, and help explain why parties' intended signals and voters' perceptions of parties' policy stances often diverge (Adams et al. 2011). They also highlight how established and inert party–voter linkages limit a party's range of action when it comes to responding to new issues and new competitors. This is because parties tend to be ‘prisoners of national conditions’ regarding issue salience (Seeberg and Adams 2024: 14), and some issues will lend themselves more directly to voter mobilization, while others are best left untouched. Finally, the survey experimental design offers a methodological advancement over observational studies in this field of research. By employing a controlled experimental setup, this approach ensures that the survey respondents' assessment of ideological commitment is in fact attributable to the treatments administered rather than any other confounding variables that tend to influence voting decisions.

A commitment theory of parties' responses to competition

Signalling responsiveness or signalling commitment

Political parties are at the heart of representative democracy and a rich literature investigates how parties present themselves to voters (Budge 1994; Merrill and Grofman 1999). Much of this literature is grounded in spatial voting theories and the corresponding idea that voters will support a given party based on the perceived proximity between their own political preferences and the party's policy profile. Parties therefore position themselves on a range of issues so as to minimize the distance between themselves and their voters and maximize their electoral appeal (Downs 1957).¹ This can lead parties to diversify their campaign messages and to broaden their electoral appeal, promising ‘everything to everyone’ simultaneously in the most extreme case (Somer-Topcu 2015).

However, while spatial models of party competition commonly assume parties update their positions in the policy space as required in order to expand the ‘scope of conflict’ and carve out their fields of competence (Schattschneider 1975 [1960]), the question of how parties reconcile these strategic adaptations with their previous positions and ideological identities has been comparatively under-studied. This question is critical as existing research often presumes that parties possess the means of setting the agenda and shaping the public's issue perception through manipulations of positions and topic salience (Meguid 2005, 2008). Yet, their ideological origins constrain parties in how freely they can move in policy space and respond to emergent challengers. That is, parties adapt to their environments yet are simultaneously constrained in their capacities to fundamentally alter their policy positions on issues that are of core concern to their constituents (Budge 1994; Somer-Topcu 2009). This is because constant adaptation of policy positions risks undermining a party's identity and credibility with voters.

Saliency theory therefore holds that parties should emphasize issues that are ‘core’ to their ideological identity and de-emphasize others that are of more ‘peripheral’ nature (Wagner and Meyer 2014). According to Peter Mair and Cas Mudde (1998: 220), this ideological core is ‘a belief system that goes right to the heart of a party's identity and … address[es] the question of what parties are’. Often grounded in traditional cleavages (Lipset and Rokkan 1967) or new issues based on changes in value patterns in Western democracies (Inglehart 1977), these ideological cores are crucial for parties' position in the ideological spectrum, their connection to their voters and their role in government (see also Beyme 1985; Ware 1996). These ideological cores, reflected in the policies that parties prioritize, play a crucial role in shaping voter perceptions by helping them to associate specific issues with particular parties and to identify the party that best aligns with their interests (Meer and Damstra 2024; Walgrave et al. 2012).

Here, though, it is important to note that the concept of the ideological core, which parties and even to some degree party families may share, differs fundamentally from competence or associative ‘issue ownership’ (Petrocik 1996), which ceases to exist once it is contested – that is, once more than one party lays claim to the respective issue. For instance, as Tarik Abou-Chadi (2016: 421) states, ‘green parties’ issue ownership … of the environment issue is much higher than radical-right parties’ issue ownership of immigration’. However, even though radical-right parties may not own the issue, immigration is still inextricably linked with and therefore core to their ideological identity.

In aggregate, parties thus face a strategic dilemma. While rising competition and changing voter demands compel them to signal responsiveness to issues often located outside their ideological comfort zones, inert party–voter linkages and persisting organizational structures mean that they must concurrently maintain their established connections with voters and uphold a consistent public image (Bartolini and Mair 1990, 2007). Therefore, the Downsian and Rokkanian approaches to party competition have often been regarded as mutually exclusive (however, see: Koedam 2022; Rovny 2015). While one regards parties as actors who update their policy profile as electoral circumstances demand, the other stresses that parties ‘do not simply present themselves de novo to the citizen at each election; they each have a history and so have the constellations of alternatives they present to the electorate’ (Lipset and Rokkan 1967: 2). Consequently, parties often find themselves in a bind, unable to pivot away from established positions easily without risking the abandonment of issues critical to their identity and core supporters. Empirically, this leads parties to more readily adapt peripheral policy positions rather than shift their stances on issues that are central to their voter mobilization (Koedam 2022).

Although this balancing act between adapting to new realities and maintaining ideological consistency is a general feature of party competition, it is particularly critical in light of the rise of populism (Mudde 2004). Populist actors have gained momentum over the past decades by promoting a radical vision of democracy as a vehicle for the unmediated implementation of the ‘will of the people’ (Canovan 1999), often bypassing established democratic norms and processes (Urbinati 2019). This has increased the pressure on mainstream parties to adapt to evolving voter demands, increasing the tension between responsible government on one hand and responsiveness on the other (Bardi et al. 2014; Mair 2008). While ignoring the rise of populist challengers may thus appear tempting at first, it risks ceding ground and leaving populist actors with a competitive advantage in the long term. Given this tension, parties are left, as I argue, with two major pathways of responding to competitors.

First, a party may merely opt to signal general responsiveness to voters by manipulating political issues that are peripheral to its ideological identity. For instance, when engaging with competition in a positive manner, this responsiveness may manifest itself in a centre-right party accommodating the culturally conservative claims of a radical-right opponent – a policy position that is easily adopted. By the same token, if a party chooses to engage negatively with competition, it may do so by selectively countering a policy claim where the costs of doing so are low. Returning to the previous example, a centre-right party will find it easier to counter the economic nationalism of a radical-right competitor than its cultural claims. This is because centre-right parties themselves ‘endorse liberalization, but socially and culturally they tend to be nationalists and opposed to the opening up of borders’ (Kriesi et al. 2006: 927).

Second, a party may choose to signal to its voters a deeper ideological commitment. Unlike signals of responsiveness, which may be perceived as opportunistically aimed at electoral gain, signals of commitment are evidently driven by ideology, regardless of the specific policy direction. That is, whether reinforcing existing policy positions or fundamentally shifting them, these actions demonstrate a profound ideological conviction. For instance, a centre-left party might emphasize its long-standing dedication to left-wing economic policies, such as redistribution and welfare state expansion, to reinforce its position within the ideological spectrum. Conversely, it might fundamentally revise its economic policies in response to changes in the economic circumstances or due to a fundamental change of heart of its support base. For instance, in the 1990s, many social-democratic parties, including the British Labour Party under Tony Blair, shifted towards market-friendly policies as part of the governments they led, responding to globalization and pressures to maintain economic competitiveness.

Such signals carry significant risks, including giving rise to intra-party conflict (Bardi et al. 2014) and voter alienation (Van de Wardt 2015), but they also offer considerable rewards. Adapting peripheral policy positions whilst maintaining a firm stance on core issues may, taken together, appear inconsistent. By committing to policy change across peripheral and core policy issues, a party provides clear and coherent cues that can facilitate voting decisions especially for politically less sophisticated voters (Levendusky 2010; Petersen et al. 2010). Such radical and decisive policy shifts, which are less common than incremental adaptations, typically coincide with changes in a party's dominant faction or leadership (Harmel and Janda 1994; Heinisch 2016).

At this point, it is crucial to note that the extent of commitment parties express through their actions and rhetoric is conceptually distinct from the notion of credibility. According to Pierre Bourdieu (1981), actors accrue political capital by occupying a distinct ideological space, thereby setting themselves apart strategically from their inter- and intra-party rivals. Ideological positioning serves the purpose of carving out specific fields of competence, lending credibility to parties and their problem-solving capabilities in particular areas. Seen in this way, ideology and strategy are two sides of the same coin. However, this perspective diverges significantly from the notion of commitment, which refers to parties' intrinsic adherence to ideological principles and to their historical roots as opposed to electoral strategies induced by changes in parties' competitive environment.

To illustrate this distinction, consider the ongoing contest for working-class voters between social-democratic and radical-right parties in the Nordic region. Comparative studies show that Swedish parties within the Riksdag have long maintained a cordon sanitaire around the Sweden Democrats (SD), rooted in ideological principles (Salo and Rydgren 2021). However, this principled stance has been less consistent in other countries, such as in Denmark, the Netherlands or Finland. In Finland and Norway, social-democratic parties have initially attempted to reframe the radical right's cultural policies around failures of economic policy, thereby shifting the public's attention to emphasizing core issues pertinent to their own social-democratic voter bases (Salo and Rydgren 2021: 121). As radical-right parties grew more successful, social-democratic parties' strategies shifted to embrace overtly cultural issues such as immigration.

While these strategies might be interpreted, from a Bourdieusian perspective, as strategic battles for credibility with working-class voters, they point to substantial differences in approaches. Initially, conflicts centred on quintessential social-democratic themes such as redistribution and social welfare. Over time, however, mainstream parties have increasingly pivoted towards addressing cultural issues, traditionally associated with the rise of the radical right. In this context, credibility emerges from political contestation, as it relates to a party's competence to resolve specific policy challenges – often influenced by factors such as candidate traits and governing experience. By contrast, commitment pertains to the underlying motives with which politicians and parties enter into these contests. This also serves to demonstrate that commitment is conceptually distinct from the credibility voters may attribute to a party's policy stance or the perceived likelihood of a party fulfilling its campaign promises. It instead refers to the ideological, rather than electoral, motivations behind a party's policy messages and thus the depth and consistency of adherence to a policy direction. This distinction represents a critical yet underexplored aspect of how voters evaluate parties' policy claims and therefore constitutes the central focus of this study.

In sum, it is important to recognize that political parties' responses to competition transcend simple agreement or disagreement with their rivals' policies, as these responses also encompass the intent and commitment behind parties' messages and actions. This commitment, in turn, is deeply rooted in parties' ideological origins: for instance, when a liberal party collaborates with a centre-left party on sociocultural policies, the purpose and the message conveyed by this cooperation would differ fundamentally from those associated with a collaboration on economic policies and taxation.

Similarly, the effectiveness of accommodative and adversarial approaches will partially hinge on the policy domain these strategies revolve around. It is therefore critical to consider the multidimensionality of the policy space and thus to expand the typical focus on migration and multiculturalism (Dahlström and Sundell 2012; Hjorth and Larsen 2022; Krause et al. 2022; Spoon and Klüver 2020) to include other relevant policy areas around which parties mobilize. Against this backdrop, this study employs a controlled experimental design which not only maximizes the range of policy issues considered but, importantly, is also more suitable when it comes to illuminating the causal mechanisms of how parties' responses to competition reflect on voters' preferences and perceptions of ideological commitment.

Voters' assessment of parties' commitment

Within the literature on spatial party competition, the idea that voters support the party that best represents their interests is seen as a cornerstone of representative democracy. Political parties compete in a given ideological space and voters generally reward their parties for maintaining a firm ideological stance (White 2021). Voters tend to have a clear preference for parties fulfilling their campaign promises after the election (Born et al. 2018; Thomson and Brandenburg 2019) and remaining steadfast in coalition negotiations (Velden and Meijers 2023). This comes as no surprise and is the very reason why voters support a party in the first place. Remaining focused on a specific policy commitment creates certainty (Meyer 2013) and consolidates the ideological image of a party within the broader electorate, which is central to party competition and to party democracy (Bowler 1990; Petrocik 1996).

However, while value consistency certainly matters, voters also appreciate adaptability and ‘getting things done’, especially when it comes to more pragmatic issue areas (Tavits 2007). For instance, Kyung Joon Han (2017) finds that parties' positional shifts only hurt them electorally if these shifts occur along their primary ideological dimension. And even there, voters – at least those with high levels of political trust – are shown to tolerate parties that reverse their policy positions, if parties provide sufficient justification of their actions (Nasr and Hoes 2023). In a similar vein, even though a majority of voters may prefer parties that maintain a clear policy profile on normative grounds, voters also acknowledge the need for responsiveness to public opinion and expert advice in policymaking (Heinisch and Werner 2023). This tolerance for principled yet flexible action is encouraging for political parties, as they continually encounter new issues and changes in their competitive environment that demand prompt responses and pragmatism.

Since most studies on the effectiveness of parties' programmatic strategies such as policy accommodation largely rest on aggregate-level investigations of vote share changes, there remains a dearth of knowledge when it comes to the individual-level psychological mechanisms. Do voters even recognize policy shifts as such? Although previous research cast doubt on whether voters are even aware of parties' spatial policy adjustments (Adams et al. 2011), more recent empirical research reveals that voters indeed care about policy change and update their voting intentions accordingly (Ferland and Dassonneville 2021; Seeberg et al. 2017; Somer-Topcu et al. 2020), at least when the changes affect salient political issues (Plescia and Staniek 2017). Thus, while shifts along core policy issues seem to attract voter attention, smaller adjustments of peripheral issues alongside shifts in mere rhetoric are more likely to go unnoticed or at least do not affect vote choice. As a baseline claim, I therefore formulate the following hypothesis:

Hypothesis 1: Voters perceive in-party policy shifts towards a competitor as less committed than claims that reinforce previous policy stances. (Positive vs negative engagement)

Voters (may) hear the message, but what do they make of it? Spatial models of party competition suggest that voter support hinges on a party's agenda, but voters do not necessarily take a party's statements at face value (Fernandez-Vazquez 2019: 309). Recognizing the strategic nature of political campaigns, I posit that voters assess the motives behind policy claims, discerning whether they stem from ideological commitment or strategic positioning. Specifically, this article argues that the degree of commitment conveyed through a policy claim depends significantly on the issue domain itself. When parties address core policy issues, it signals strong commitment, regardless of other factors such as the direction into which the party is moving. Conversely, when parties emphasize peripheral policy issues, voters tend to view these actions as strategic manoeuvres, again independent of the direction of the policy claim.

To illustrate this logic, consider the following case. A pertinent example of far-reaching policy accommodation in the Austrian context has been the case of the Austrian People's Party (ÖVP) campaign in the 2017 national election (Heinisch et al. 2020). While the ÖVP aligned itself closely with the cultural positions and migration stances of the radical-right Freedom Party of Austria (FPÖ), it crucially also claimed migration constituted a fundamental threat to Austria's social system and economic sovereignty.² Even though the ÖVP could have chosen a cherry-picking approach to avoid the apparent clash with the FPÖ's policy in the standpoint it had adopted with its own core economic positions (i.e. economic liberalism), the party decidedly opted against such a manoeuvre. Whether one agrees or disagrees with the message in ideological terms, the move expressed commitment to policy change – and raises the key question of whether voters can discern between ideological commitment on the one hand and mere electoral strategy on the other.

Hypothesis 2: Voters perceive in-party policy claims that impact core policy areas as more committed than policy claims that impact peripheral issues. (Core vs periphery)

Two further conditions may moderate the discussed relationship, relating to both the party level and the voters' personal dispositions. For one, political parties, much like voters, understand the strategic and often pragmatic nature of elections. Aware that frequent changes in policy direction could undermine their credibility (Allgeier et al. 1979) and alienate voters (Adams and Somer-Topcu 2009), parties are more inclined to adjust their framing and discourse around a policy rather than the policy itself. In this vein, Mohamed Nasr and Emma Hoes (2023) demonstrate that although voters generally disapprove of positional adjustments, their tolerance can be significantly influenced by how parties justify the need for such action. Internal justifications are related to intrinsic policy motivations, but external justifications that centre on ongoing crises or expert advice are also frequent features of party competition and play a crucial role in moderating this tolerance (see also: Heinisch and Werner 2023; Jacobs 2024). Building on these insights, I propose that voters perceive parties' policy claims as more committed if these claims are supported by reference to internal deliberation processes within the party, rather than by external constraints such as responses to new competitors.

Hypothesis 3: External justifications reduce the extent to which voters perceive the core policy claims of their in-party as committed. (Internal vs external justification)

For another, I expect the degree to which voters hold populist attitudes to exert a crucial moderating effect. In ideational terms, populist attitudes are characterized by ‘the juxtaposition between a people-centered notion of political representation and the corrupt political elite’ (Geurkink et al. 2020: 248) and an understanding of politics as a moral struggle between these two antagonistic groups (Hawkins et al. 2018; Mudde 2004). Since acceptance of policy reversals often presupposes a given level of general trust in politics (or partisan identity), voters who hold populist attitudes are more likely to be wary of change. Consequently, when an in-party begins to mirror the policy claims of a competitor, voters who hold populist attitudes may perceive such an act as a betrayal of their support and evidence of elite collusion (Hameleers 2021). Thus, any shift that distances an in-party from its previously held positions and aims to narrow the gap to a competitor is likely seen as strategic behaviour rather than ideologically committed. This also implies that while populism generally demands greater responsiveness, it simultaneously often limits parties' leeway actually to deliver on such demands.

Hypothesis 4: Voters who hold populist attitudes generally regard positive engagement as a sign of low commitment, irrespective of the issue at stake and the justification given. (Populist attitudes)

Empirical strategy

This section provides a brief outline of the methodological approach and the design of the survey experiment, the aim of which is to illuminate how voters comprehend and judge parties' responses to competition.

Methods

Experiments provide a powerful means to explore the psychological mechanisms at play when voters respond to changes in conditions such as specific events in party competition. For this study, I employ a survey experiment conducted in March 2023 in Austria. This online survey experiment (N = 2,013), representative of the Austrian voting population with internet access, encompasses responses from Austrian citizens aged 16 and older – that is, the voting age in Austria. Additional population quotas were applied for gender, age and education levels. The survey was pre-registered via the Open Science Framework, and the full replication data along with extensive documentation are available on AUSSDA (Habersack 2024b).

The case of Austria presents an ideal setting for this analysis for several reasons. First, the use of the Austrian political system enables the creation of treatment manipulations that are both authentic and reflective of real-world scenarios. Second, Austrian voters have witnessed several shifts in political dynamics and alliances at both local and national levels (Heinisch et al. 2022), which enhances the credibility and relevance of the scenarios presented in the survey. At the time of data collection, the outcome of the 2024 national election and the FPÖ's eventual win of the largest vote share was anything but certain. Third, the Austrian party system has become increasingly differentiated over time, with smaller parties not only gaining parliamentary representation but also participating in government coalitions. Notable examples include The Greens–Green Alternative (GRÜNE) and the radical-right FPÖ, which not only significantly impact policy competition but also influence coalition dynamics in Austria. This multiparty context with political contests that unfold in a multidimensional issue space renders Austria particularly relevant for this study, though the theoretical implications likely extend beyond its borders.

The experimental design balances realism and hypothetical elements. Specifically, this means that the survey employs existing party labels and scenarios related to the national election (Nationalratswahl) in 2024, but also introduces a fictitious party by the name of ‘Besser Gemeinsam Österreich’ (Better Together Austria). This approach helps manage trade-offs inherent in survey experiments, such as balancing the hypothetical nature of scenarios – which reduces pre-treatment biases and increases the study's generalizability – with the need for realistic settings that improve control and respondent comprehension (Brutger et al. 2022).

The survey begins by assessing respondents' policy preferences before presenting them with the scenario of the 2024 election, where their in-party faces a new challenger party adopting a contrary stance on a selected policy issue. To minimize biases from pre-existing beliefs about party positions, the survey includes a series of pre-treatment questions, including voting propensity, as well as post-treatment checks to verify attention and to test whether the manipulation worked effectively.

Experimental design

Among other pre-treatment questions, respondents were asked to rate a diverse set of policy issues using an 11-point scale, allowing them not only to express their opinions but also to indicate the salience they attribute to each issue. This range of policy issues – from welfare policy and immigration to climate change and security – significantly broadens the scope of this study compared to existing research, which largely concentrates on immigration and assumes uniform voter responses across various policy domains. A central tenet of this study is that parties' responses to competitors are not merely about policy agreement or disagreement; rather, they span a continuum of commitment, ranging from steadfast opposition to complete accommodation. The position a party adopts on this spectrum is heavily influenced by the specific policy issue under consideration. Thus, this analysis necessitates a wider variety of policy issues than the traditional focus on migration policy. Moreover, the experimental design employed here addresses significant constraints inherent in observational studies. It effectively isolates the causal impacts of party behaviour on voter perceptions, which are often obscured by confounding factors in less controlled research settings.

The experimental component of the survey asks respondents to consider the case of the 2024 Austrian national election. At the time of data collection, this chosen case allowed for the presentation of a contest between the respondent's in-party and a fictional new challenger. The scenarios heightened the stakes by depicting the struggling in-party as competing closely with the emergent challenger party predicted to enter parliament. Following the pre-treatment questions, respondents participated in a 2 × 2 × 2 vignette experiment. This design varies the issue type (‘Core’), the direction of the in-party's policy claim (‘Engagement’) and the in-party's discursive justification of its actions (‘Justification’) as articulated by the party's secretary-general. Table 1 provides an overview of the eight vignette types, each including approximately 250 respondents.

Table 1. Overview of Vignette Types and Average Perceived Commitment

Note: Mean commitment on a 0–10 scale per treatment group.

Respondents were thus each presented with scenarios where their in-party responds to a challenger taking a policy position diametrically opposed to their own. These scenarios alternated the direction of the policy response and the discursive justification for such actions, namely whether the in-party referenced the new competitor or whether the policy position was portrayed as the outcome of extensive intra-party deliberation. For a detailed breakdown of the methodology and survey instrument, see Appendix B in the Supplementary Material, which includes full variable documentation and the German-language original questionnaire. For an example vignette translated into English, see Figure 1.

Figure 1. Example Vignette.

Next, please imagine the case of the National Council election in autumn 2024. The race is still completely open a few weeks ahead of the election date. Your vote matters. A new party named Besser Gemeinsam is running. The new party is said to have good chances of entering the National Council with a double-digit result.

The SPÖ [Social Democratic Party of Austria] is closest to you on many political issues, such as in the area of environmental policy. However, public opinion in Austria has shifted significantly recently. Many now believe that Austria cannot tackle climate change single-handedly and that it is more important at the present time to specifically support the economy rather than subsidize renewable energies. The newly competing Besser Gemeinsam represents this very standpoint. It is becoming apparent that now especially the SPÖ will lose votes to Besser Gemeinsam.

In a turn of events described as ‘unprecedented’ by political commentators, the SPÖ now changes its position on the matter. In a recent interview, the party's secretary-general stated, ‘What matters now is that we have internally agreed on the right course: this means that we recognize the need to change direction on this important matter’ (APA 25 August 2024).³

As has become firmly established in experimental research (Druckman et al. 2011), the survey then continues with a range of attention checks and manipulation checks designed to investigate the decision mechanisms involved in respondents' choices. These checks are crucial as they probe into respondents' evaluations of the in-party's policy shift, their attentiveness to the scenario and their perceptions regarding which existing party might be represented by the fictitious challenger.

After the treatment exposure, the survey progressed with a series of outcome and post-treatment questions. These are specifically designed to evaluate how respondents perceive their in-party's actions – whether as genuine or strategic. These perceptions constitute the main dependent variable. The survey concludes with inquiries into respondents' demographic characteristics and a debriefing message to ensure clarity and provide closure on the study's purpose.

Empirical analysis

Data

As previously noted, the sample is representative of the Austrian voting-age population with internet access, balanced across age, gender, education and geographic location, encompassing all nine Austrian federal states. To address minor imbalances, post-data-collection population weights were applied to the survey data. Detailed descriptive statistics of the sample are available in Section 1.1 of the Appendix.

Figure 2 visualizes respondents' positions across the nine two-sided policy issues corresponding to various domains: social, environmental, foreign, migration, equity, family, security, European and health policy. These self-placements on a 0–10 scale inform the vignette treatments and are crucial for determining the salience of core versus peripheral policy issues. Positions near the scale's midpoint indicate low salience, whereas positions towards the extremes suggest stronger and more definitive opinions in specific issue domains.

Figure 2. Density Plot of Policy Preferences on Nine Policy Dimensions, N = 2,013.

The figure illustrates the ideological distribution of respondents' positions along the mentioned policy domains, displayed as density plots on an 11-point scale. The density peaks indicate the levels of salience respondents attribute to each issue, with higher concentrations at the extremes reflecting stronger opinions. For example, a respondent placing her- or himself near the centre of the scale on the ‘family aid’ issue signals that this particular issue is of low perceived salience, whereas taking an extreme position on foreign policy (i.e. ‘support for Ukraine’ vs ‘maintain neutrality’) indicates high salience.

In the experimental setup, each respondent's most and least salient issues were identified as core and periphery respectively, and random assignment ensured that an equal number of respondents were presented with vignettes focusing on either a highly salient core issue or a less salient peripheral issue. This design introduced ideological variation and highlighted differing levels of issue salience among participants. Subsequent robustness and sensitivity analyses, discussed below, aimed to eliminate any potential biases that might arise from specific policy impacts on the observed treatment effects. For example, although the sample exhibits a slight skew towards pro-environmental stances, support for the welfare state and opposition to COVID-19 containment policies (Figure 2), it remains balanced in terms of party choice (Appendix, Figure A.2) and the general distribution of respondent characteristics across the eight vignette types (Appendix, Table A.7).

Results

Turning to the testing of H1–H4, I conduct a series of regression models using the perceived level of commitment of the in-party's action as dependent variable and the three vignette treatment dummies (core vs periphery; positive vs negative engagement; external vs internal justification) as main explanatory variables on the right-hand side of the equation. The models additionally include all pairwise interactions between the treatments. Model 1 contains only the main predictors, Model 2 includes the main predictors and all control variables, and Model 3 additionally features populist attitudes and their interaction with (positive or negative) engagement. The main findings are presented in the Appendix, Table A.2. The idea of the main model is summarized as follows:

$$\eqalign{{\rm Commitmen}{\rm t}_i = & \alpha _i + \beta _1\,{\rm Cor}{\rm e}_i + \beta _2\,{\rm Engagemen}{\rm t}_i \cr & \quad + \beta _3\,{\rm Justificatio}{\rm n}_i + \beta _4\,{\rm Cor}{\rm e}_i \times {\rm Engagemen}{\rm t}_i \cr & \quad + \beta _5\,{\rm Cor}{\rm e}_i \times {\rm Justificatio}{\rm n}_i + \beta _6\,{\rm Engagemen}{\rm t}_i \cr & \quad \times {\rm Justificatio}{\rm n}_i + \beta _7\,{\rm Cor}{\rm e}_i \times {\rm Engagemen}{\rm t}_i \cr & \quad \times {\rm Justificatio}{\rm n}_i + \varepsilon _i} $$

To recap, according to the baseline H1, positive engagement generally leads to low perceived commitment, whereas H2 states that core policy issues should reveal a positive effect on perceived commitment independent of the direction of engagement. H3 further qualifies this relationship and posits that external justifications with reference to other competitors will reduce the level of commitment, while H4 states that populist attitudes reinforce the negative effect of policy shifts due to greater sensitivity to and distrust related to positional adaptations among populist individuals.

Figure 3 displays the estimated effects of the three main predictors. In line with expectations, positive engagement with a competitor leads to lower perceived commitment, ceteris paribus. An in-party shifting and therefore adapting its previous position results in decreased commitment by 9 to 17 percentage points on the commitment scale. Even though the effect is weaker (six to seven percentage points), an in-party's manipulation of a core policy issue, no matter the direction, increases the level of perceived commitment. How this move is discursively justified does not fundamentally alter the perceived level of commitment. Neither on its own nor in interaction with other factors does the type of justification provided reach significance. This generally confirms H1 and H2 and runs counter to H3. Importantly, the findings are robust to introducing any covariates, including the crucial factor of whether or not somebody ideologically supports the in-party's behaviour vis-à-vis the challenger or not; which is moderately positively associated with commitment.⁴

Figure 3. Treatment Effects on Perceived Level of Commitment in Party Behaviour.

Turning to H4, I additionally investigate the effect of populist attitudes (see Figure 4). Contrary to H4, populist attitudes do not amplify the negative effect of policy shifts towards a competitor. Rather, populist attitudes generally lead to low levels of trust in the genuineness of policy claims, meaning that the direction of engagement plays no additional role. It is only among those with weak or no populist attitudes that direction of engagement actually matters and shows the effect on commitment described above in H1. Further exploratory analyses of the reverse effect revealed that the vignette to which respondents were exposed did not fundamentally increase or decrease their populist attitudes, as evidenced by Figure A.10 in the Appendix. The figure reports on a before/after comparison of populist attitudes and illustrates that the main differences in the distribution of populist attitudes are due to the two different operationalizations (geometric mean and the Goertzian approach), rather than any treatment effects.

Figure 4. The Moderating Effect of Populist Attitudes, 95% Confidence Intervals.

Finally, I explore whether the form of party behaviour described in the vignettes had any further downstream effects on respondents' voting intentions and intentions to participate in the 2024 national election (see Table A.4 and Figure A.6 in the Appendix). Neither the type of issue domain nor the way in which parties justified their policy stances had any significant impact on voting intentions or respondents' willingness to participate in the election. However, parties' policy shifts away from their previous positions did effectively reduce respondents' willingness to throw their support behind their in-parties by 5%. These findings suggest that voting decisions are often largely insulated from immediate reactions to new pieces of information and thus rather stable – which, after all, bodes well for democratic resilience. While short-term adjustments in party messaging can influence voter perceptions of parties' motives, they do not necessarily result in abrupt shifts in electoral outcomes, highlighting voters' capacity to reflect on parties' policy standpoints.

All findings are robust to a range of different model specifications. To begin with, the policy issues vary in their salience not only per respondent but also across the overall sample. Additional regression models were thus performed to account for the specific policy issue carried forward into the experiment (see Appendix, Table A.2). Additional investigations of the balance across vignette types revealed an about equal representation of in-parties (Appendix, Figure A.2) and of the nine policy issues in the vignettes as seen by the respondents (see Appendix, Figure A.3). Nevertheless, to rule out sensitivity of the findings to any specific segments of the sample, I ran additional models which iteratively removed a given in-party (Table A.5) and a given policy issue (Table A.6) at a time. All results are robust to any of these changes in model specification and data segmentation. More generally, a multinomial regression using the treatment types as dependent variable revealed that randomization effectively worked to create treatment groups without even minor imbalances (Appendix, Table A.7), which is impressive given the sample size and range of covariates.

In a series of further robustness checks, I used different operationalizations of key variables. First and foremost, I utilized an alternative approach of measuring the core/peripheral nature of policy issues. For the main models presented above, I relied on respondents' self-placement on the policy issue scale and randomly selected either an issue associated with the subjectively most extreme position on the scale (core) or an issue on which the respondent placed her- or himself closer to the centre of the scale (periphery) to carry forward into the vignette. This approach assumes that supporters of a given party are more likely to care about the core issues of said parties and that, therefore, the ideological cores of voters and parties tend to align. The effect of manipulating a core policy in the experiment should thus be analogous to the expected effect of a real political party manipulating some of its main policy messages.

Models 1b–3b in Table A.3 of the Appendix use an alternative measurement strategy that is in part based on the mean positions of each party's supporters (see Figure A.4) and in part informed by the wider literature on party families (Freeden et al. 2013; Langsæther 2023). Building on this, I classified the following policy areas as the Austrian political parties' core ideological policy issues and everything else as peripheral issues: SPÖ (social policy, equity policy), ÖVP (family policy, security policy), FPÖ (migration policy, European policy), GRÜNE (environmental policy, equity policy), The New Austria and Liberal Forum (NEOS) (European policy, equity policy), Communist Party of Austria (KPÖ) (social policy), People–Freedom–Fundamental Rights (MFG) (health policy), Peter Pilz List (PILZ) (environmental policy), Beer Party (BIER) (security policy) and Left (LINKS) (social policy).⁵

The above-described alternative and more deductive approach has evident drawbacks, as it leads to somewhat unbalanced groups, where peripheral issues outweigh core policy issues. While the main measure of core policies is balanced by design, the alternative measure results in about 24% of respondents who were exposed to an in-party's core policy position and 76% who were not. Remarkably, despite these drawbacks, the presented findings remain unchanged when using the alternative measure. Across all three models, the effects of the core policy variable remain as significant as in the main Models 1–3, while the effect sizes even increase.

Lastly, I changed the operationalization of populist attitudes to account for the non-compensatory nature of the three subdimensions, as discussed in Alexander Wuttke et al. (2020). For this part, see Section 1.6 of the Appendix. While for the main models I use the geometric mean of the factor scores derived from a three-factor Confirmatory Factor Analysis (CFA), the alternative method follows the Goertzian approach and takes the lowest value across the three dimensions as a respondent's level of populist attitudes. Effectively, this narrows populist attitudes down to the Manichean dimension, which on average features the lowest level of agreement. The alternative measure of populist attitudes thus, as expected, weakens the effects but does not fundamentally alter the findings.

Conclusion

This article has sought to establish the level of commitment party actions convey as an important additional dimension of voters' perspectives on party competition besides the perception of policy positions and the direction of policy claims. This perceived commitment relates to the question of where policy claims fall on a spectrum between strategic manoeuvring on one hand and ideological conviction on the other, and is conceptually and empirically distinct from agreement or disagreement with a given claim. While voters' perception of commitment in party politics may not directly sway vote choice due to the fact that voting decisions typically hinge on a range of additional factors, this new perspective does promise insights into why parties' communicated policy positions and voters' perception thereof occasionally diverge. Taking a closer look at commitment thus also helps contextualize why approaches to competition such as ‘policy accommodation' in some instances prove effective in keeping challenger parties at bay while backfiring in other scenarios. Thus, the central focus of this article has been to investigate whether voters are able to discern commitment in parties' policy responses and how the nature of the issue at stake – but also voters' personal dispositions in terms of populist attitudes – influence this ability.

To this end, this article utilizes a pre-registered 2 × 2 × 2 vignette experiment, presenting respondents with a semi-hypothetical scenario of an electoral contest fought between the respondent's struggling in-party and a newly emerging challenger. The experimental design systematically varies (a) the policy area at the centre of the contest, which constitutes either a core or a peripheral issue; (b) the nature of the in-party's engagement with the challenger, either positive or negative; and (c) the basis of the in-party's rationale for its actions, whether grounded in internal deliberation or justified by reference to the new challenger and external conditions. Through this methodological approach, the article not only provides new primary survey data from a representative sample of 2,013 eligible Austrian voters but also overcomes the inherent limitations of previous observational studies by facilitating robust causal inferences when it comes to the question of how specific forms of party behaviour reflect on voters.

Empirically, while the type of policy justification has no significant effect on the level of perceived commitment expressed in an in-party's actions, the best predictor of commitment is the direction into which a party moves (i.e. towards the challenger and away from its previous position or away from the challenger, reinforcing its previous position). However, the results also confirm that the nature of the issue itself influences perceived commitment: policy claims related to a party's core ideological identity tend to be associated with higher levels of perceived commitment (no matter how the ideological core is operationalized), irrespective of the direction into which an in-party moves or whether the voter approves of this move. Populist attitudes, in turn, fulfil an important moderating function as voters who hold such attitudes are generally more sceptical of the genuineness of parties' intentions. Their opinions are shaped by stronger and less malleable beliefs of what constitutes ‘good’ and ‘bad’ policy positions. Thus, the nature of the issue and the direction into which parties move only effectively influence perceived commitment among those with no or only weak populist attitudes.

The study acknowledges its limitations, including potential constraints on empirical breadth and the representativeness of the vignette scenarios. Though the timing of the data collection as well as the designing of the experiment around the real-world case of the Austrian political system ensured both plausibility and comprehension of the described scenarios, future research should investigate how well the findings travel and how other contextual factors influence voters' ability to discern commitment in party positions. Despite these limitations, the implications of this study may extend beyond the specific context of Austria. Similar political systems characterized by multiparty competition and dynamically changing alliances could see comparable voter behaviour. Additionally, the methodological approach ensured that any changes in perceived commitment were attributable to the treatments administered and as both the questionnaire and replication data are openly available, the design can be readily adapted to different settings to explore how cultural, socioeconomic and political factors might influence the generalizability of these findings. Therefore, while the immediate conclusions are drawn from the Austrian context, the underlying theoretical constructs about perceived ideological commitment in parties' rhetoric and actions are likely relevant also in other cases.

Furthermore, while this article primarily focuses on the how of parties' responses to competition and their impact on voters, subsequent studies should investigate more deeply the why and the when of these responses. For example, although previous research has indicated that factors such as electoral vulnerability (Abou-Chadi 2016; Adams 2012), changes in party leadership (Harmel and Janda 1994; Heinisch 2016) or a combination of both (Bale et al. 2010) may prompt parties to commit to policy change, further investigations are needed to investigate the constraining factors that might explain parties' resilience to policy change.

The results presented herein contribute to the literature on party competition and the role of populism within it. By highlighting the often-overlooked significance of perceived commitment in party policy responses, this research helps explain why there is frequently a discrepancy between the policy stances of parties and voters' perceptions thereof (Adams et al. 2011). That is, rather than accepting parties' policy positions as presented, voters assess these stances through the lens of their own expectations – and populist beliefs – as well as their past experiences with and images of these parties, enabling them to discern between genuine commitment and electoral strategy.

Moreover, the findings underscore the importance of parties' ideological identities for their capacity to adapt to changing competitive environments. For instance, they shed new light on why attempts to regain voters lost to challenger parties often appear futile (Vries and Hobolt 2020). This is because parties tend to be ‘prisoners of national conditions’ (Seeberg and Adams 2024: 14) when it comes to issue salience: while some issues may be closely tied to a party's ideological core, facilitating voter mobilization, others that are of more peripheral nature pose a significant challenge, limiting parties' ability to respond flexibly to changing political landscapes and voter demands. Contrary to previous research, according to which parties possess the means to influence public issue perceptions and set the agenda, this study therefore underlines that parties' ideological identities significantly restrict their manoeuvrability within the policy space. Taken together, this illustrates the relevance of the presented findings to political actors and suggests that, ultimately, the decision of whether and how much to respond to competition is a question of commitment.