Adopting restriction: how coalition governments respond to radical right parties

Ellger, Fabio; Klüver, Heike Article — Published Version Adopting restriction: how coalition governments respond to radical right parties West European Politics Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Ellger, Fabio; Klüver, Heike (2025) : Adopting restriction: how coalition governments respond to radical right parties, West European Politics, ISSN 1743-9655, Taylor & Francis, London, Vol. 48, Iss. 6, pp. 1299-1331, https://doi.org/10.1080/01402382.2024.2341352 This Version is available at: https://hdl.handle.net/10419/308005.2 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ West european politics 2025, Vol. 48, no. 6, 1299–1331 Adopting restriction: how coalition governments respond to radical right parties Fabio Ellgera and Heike Klüverb aWZB Berlin social science center, Berlin, Germany; bHumboldt-universität zu Berlin, Berlin, Germany ABSTRACT Radical right parties (RRPs) have gained representation in parliaments across Europe, but little is known about their impact on government policy. Even though RRPs tend to end up in opposition, they have considerable influence on policy making within coalition governments. One expectation is that coalition governments are tough on immigration to optimise voter support when being exposed to right-wing parties in parliament. Coalition negotiations temporarily reduce accountability and allow cabinets to adjust policy positions without bearing the costs associated with opportunistic behaviour. This argument is tested using novel data on pre-electoral policy positions and post-electoral immigration policies for coalition cabinets in 24 European democracies from 1980 to 2015. The findings reveal that governments shift to more restrictive immigration policies in face of RRPs. This article expands on prior research on the influence of the radical right by demonstrating its direct influence on coalition governments’ joint immigration policy plans. KEYWORDS Comparative politics; coalition government; radical right parties; immigration policy Do radical right parties influence coalition government policy? Right-wing and populist challenger parties have established themselves within a range of European party systems. While mainstream parties suffer dramatic electoral losses, radical right parties such as the German Alternative für Deutschland, the Danish Folkeparti or the Austrian Freiheitliche Partei Österreichs are on the rise. More recently, scholars started to investigate how these parties – albeit mostly being in the opposition – have affected the policy positions of established parties (Abou-Chadi and Krause 2020; Meguid 2005; Spoon and Klüver 2019) or the attitudes of voters (Bischof https://doi.org/10.1080/01402382.2024.2341352 © 2024 the author(s). published by informa uK limited, trading as taylor & Francis Group CONTACT Fabio ellger fabio[email protected] supplemental data for this article can be accessed online at https://doi.org/10.1080/01402382.20 24.2341352. this is an open access article distributed under the terms of the creative commons attribution license (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. the terms on which this article has been published allow the posting of the accepted Manuscript in a repository by the author(s) or with their consent. 1300 F. ELLGER AND H. KLÜVER and Wagner 2019). Relatedly, prior research also shows the impact of extreme parties on immigration policy change (Abou-Chadi 2016b; Folke 2014; Howard 2010; Schain 2006; Williams 2006). In addition, several studies explore the direct and indirect effects of the radical right on mainstream party positions (see Abou-Chadi and Krause 2020; Han 2015; Krause and Matsunaga 2023; van Spanje 2010). We contribute to this literature by conducting a novel test of the impact of RRP representation on changes in the joint policy agenda of coalition governments. Specifically, we show that coalition governments shift to more restrictive immigration policies in their joint agreements whenever radical right parties are represented in parliament during cabinet formation. All parties pay close attention to the distribution of voter preferences (Downs 1957). Once they recognise a preference shift among the electorate, parties adapt their positions in order to maintain electoral support (Adams et al. 2004; Ezrow et al. 2011). Scholars of party behaviour have examined this process of ‘policy optimisation’ and found varying strategies of adjustment. Adams etal. (2006) show that mainstream parties respond to shifts in the general population while niche parties pay special attention to preference changes among their core support groups. Klüver and Spoon (2016) similarly find that mainstream parties respond to issue priorities of the overall population while niche parties mostly respond to their own supporters. But open policy shifts do come at a cost, as parties face a range of restrictions which constrain their behaviour (Meyer 2013; Tavits 2007). Voters, and especially partisan supporters, closely monitor programmatic changes (Cox and McCubbins 1986; Ezrow et al. 2011) and have a desire for credible signals of policy stability. A party that changes its positions regularly and substantially has a hard time to convincingly commit to policy adherence in the future. Accordingly, Tavits (2007) shows that opportunistic shifts on salient policy issues result in electoral penalties. Political parties therefore try to strategically balance regular adaptation to preference changes in the electorate with avoiding to appear as ‘flip-floppers’. Research also shows that established parties pay attention to new contenders (Meguid 2005) and there exists a growing interest in the way emerging right-wing challenger parties provoke mainstream party reactions (see Abou-Chadi and Krause 2020; Han 2015). RRPs emerged in political arenas throughout Europe and built their success on a combination of anti-elitist rhetoric and nationalist policy positions (Kitschelt and McGann 1997; Mudde 2007; Williams 2006). These parties have also politicised new ‘cultural’ issues (Kriesi et al. 2006), with a special emphasis on anti-immigration policies (Dancygier 2010). When facing successful radical right challengers, mainstream parties often react with an accommodative strategy (Meguid 2005; van Spanje 2010). WEsT EuROPEAN POLiTiCs 1301 In this article, we focus on the influence of radical right parties during the formation of multiparty coalitions. Cabinet parties can use confidential coalition negotiations – a period of low accountability and clarity of responsibility1 – to optimise their positions on immigration without bearing the usual costs associated with policy change on salient issues. In the face of RRPs represented in parliament, cabinet parties propose more restrictive immigration policy positions in their coalition agreements. We test our argument based on a unique dataset covering 165 coalition formations in 24 West and East European party systems between 1980 and 2015. Our analysis makes use of a novel data that combines positions derived from coalition agreements with pre-electoral positions from party manifestos (Volkens et al. 2018). We provide more information on the origin of the agreements in Online Appendix Section 1.2. Our approach is unique since we combine party positions of government parties extracted from their pre-electoral manifestos with data on policy positions of coalition governments that originate from their post-electoral coalition agreements. The coalition agreements are usually negotiated shortly after the election and are published only a few weeks after voters have cast their ballot, which enables us to track policy change over an exceptionally short period of time. We can reasonably assume that position changes of government parties are barely affected by exogenous, time-varying confounders. And since the election manifestos and the coalition agreements are analysed using a similar coding procedure, we can moreover assess the magnitude and direction of policy adjustment during coalition negotiations on precisely the same ideological scale. To test our hypothesis on immigration policy optimisation during coalition negotiations, we analyse how much the coalition agreement positions (CAPs) deviate from pre-electoral expected positions (EPs) that we generate on the basis of Gamson (1961). We find that RRP presence in parliament, a symbol of support for extreme-right politics among the electorate, is a significant predictor of shifts towards more restrictive policies. A range of additional analyses show that this effect is not primarily driven by the size of radical challenger parties or their initial entry to parliament. We then explore how attributes of the incoming government moderate the effect. While the general ideological alignment of a cabinet or the duration of its formation does not appear to impact policy change, we find that minority cabinets, and cabinets which have lost electoral support, are more likely to shift towards restrictive policies. In a final series of robustness tests, we show that coalition cabinets exposed to RRPs in parliament are quite similar to those without right-wing pressures across a range of important covariates. Overall, the empirical analysis provides evidence showing that coalition cabinets confronted with RRPs in parliament adopt significantly more 1302 F. ELLGER AND H. KLÜVER restrictive immigration policy positions than those without similar exposure. Our findings have two main implications. First, we are going beyond prior research by highlighting the impact of RRPs on coalition cabinets. In particular, we provide evidence for the role of coalition negotiations as windows of opportunity for unpunished policy optimisation in face of pressures by the radical right. Rational, vote-seeking cabinet parties can accommodate to more restrictive positions while avoiding to appear opportunistic. Second, we show that the success of radical right parties does have an immediate effect on government policy. Compared to positions drafted just a few weeks or months before the elections, cabinets take on significantly harsher positions in their joint agreements when RRPs are represented in parliament. Overall, we contribute to the literature on policy positioning in contested multi-party systems and existing research on consequences of right-wing party success for immigration policy in advanced democracies. Radical right party performance and immigration policy change In this study, we test if the presence of radical right parties in parliament impacts the policy agendas of coalition governments. More specifically, we observe changes in immigration policy, a highly salient and contested issue area that is closely linked to radical right challengers (Dancygier 2010). Our argument builds on existing work that highlights the influence of the radical right on immigration policy. For example, Schain (2006) traces the development of the ‘Front National’ in France and shows how its electoral success causes realignment of existing parties. In a comparative framework, Howard (2010) showcases that the strength of far-right parties is the most relevant predictor of naturalisation processes across EU members. Relatedly, Abou-Chadi (2016b) shows that the politicisation and electoral competition of the immigration issue is a pivotal vetopoint for more progressive immigration policies. And in a quasi-experimental study on Swedish municipal elections, Folke (2014) provides a causal assessment of the influence of RRP representation on stricter immigration policy. Beyond analysing the impact of RRPs on general immigration policy, prior research has also demonstrated their effect on other parties. For example, van Spanje (2010), Abou-Chadi (2016a) and Han (2015) show accommodative behaviour and ‘contagion’ effects for mainstream parties and parliamentarians across established democracies. In a similar vein, Abou-Chadi and Krause (2020) implement a RD-design to provide causal evidence on the impact of RRPs on mainstream parties’ immigration positions, also finding accommodation effects among the establishment. WEsT EuROPEAN POLiTiCs 1303 Going beyond the assessment of radical parties’ impact on their competitor’s positions, Krause and Matsunaga (2023) show that accommodative strategies of the establishment do not pay off electorally. We contribute to this literature by shifting the attention away from individual party responses to reactions by entire coalition cabinets. For sake of comparability, our conceptualisation of RRPs follows established standards set by scholars in the field. We identify RRPs based on their critical stance towards mainstream immigration policy (see Abou-Chadi and Krause 2020; Roodujin etal. 2019). We conceive of radical and populist right-wing parties as challengers of established mainstream parties because they threaten them electorally (Abou-Chadi 2016b; De Vries and Hobolt 2020; Howard 2010). While mainstream parties suffer dramatic electoral losses in recent years, radical and populist right parties such as the German Alternative für Deutschland, the Danish Folkeparti or the Austrian Freiheitliche Partei Österreichs are on the rise. Radical right-wing parties exert pressure on mainstream parties by mobilising on the immigration issue and employing anti-establishment rhetoric to undermine mainstream party legitimacy (De Vries and Hobolt 2020; Mudde 1999; Schain 2006). One core claim is a demand for more restrictive migration laws (Brug & Van Spanje 2009). Often, these claims are combined with populist appeals and the usage of racist stereotypes (Mudde 2007). Thus, we argue that RRPs exert ‘ownership’ over the immigration issue by demanding to ‘go tough’ on immigration (see Abou-Chadi 2016b; Howard 2010; Schain 2006). Since immigration policy has become increasingly contentious and highly politicised in recent elections, radical and populist right-wing parties frequently benefitted from a mobilisation on this issue (Dancygier 2010; Kitschelt and McGann 1997). As challengers, they did so by attracting voters from established parties across the traditional left-right dimension (McGann and Kitschelt 2005; Rydgren 2008). We therefore argue that successful radical and populist right parties pose an electoral threat to coalition governments. The presence of RRPs signals to coalition parties that a sizable part of the electorate prefers more restrictive immigration policies. In an effort to appeal to these voters, we expect that coalition governments have strong incentives to adopt harsher policies on immigration. However, one central caveat we propose is that governments are not able to freely reposition themselves on such a salient issue. Flip-flopping will result in credibility losses and penalisation, especially by dedicated core supporters (Ezrow et al. 2011). In the next section, we lay out how the blurred lines of responsibility during confidential coalition negotiations can provide governing parties with an opportunity to update their policy positions without having to bear the negative consequences of position adjustment. After 1304 F. ELLGER AND H. KLÜVER developing our argument on why mainstream parties should make use of coalition negotiations to update their policy agenda, we conduct an empirical test. We provide evidence which shows that cabinets shift towards more restrictive positions if they have been exposed to a RRP in parliament. Policy optimisation during coalition negotiations We hypothesise that coalition parties will settle with more restrictive policies in their coalition agreements compared to the policy positions they proposed in pre-electoral campaign manifestos if they face RRPs in parliament. That is, because the coalition formation stage provides an ideal opportunity for government parties to adjust their policy agenda in light of recent election results without bearing electoral costs. We conceive of policy optimisation during the coalition cabinet formation as a three-stage process (see Figure 1). First, we assume that parties are rational, goal-oriented and purposeful collective actors interested in maximising their own electoral support (Downs 1957). Following the spatial proximity model, we then assume that the main instrument to maximise electoral support is strategic position-taking along a set of salient policy dimensions (Stokes 1963). The spatial proximity model also posits that political parties carefully analyse the distribution of voter preferences (Enelow and Hinich 1984). Voters are expected to support the party whose policy programme most closely aligns with their individual preferences on the issues they deem relevant. Since voter preferences change regularly over time (Converse 1964; Pedersen 1979), parties are incentivised to constantly monitor and Figure 1. strategic policy adjustment of coalition governments around elections. WEsT EuROPEAN POLiTiCs 1305 respond to shifts in the electorate (Adams et al. 2004; Schumacher, de Vries, and Vis 2013). Thus, we think of party competition as a dynamic system in which parties adapt their positions to changing voter preferences (see Stimson, Mackuen, and Erikson 1995; Soroka and Wlezien 2010; Adams et al. 2004). At the same time, political parties are also held accountable, especially by their core supporters. Ezrow et al. (2011) have contrasted the general electorate model, which focuses on vote maximisation among all voters, with the partisan constituency model, according to which parties pay special attention to the preferences of their closest followers. Political parties rely on a stable and committed group of activists for the maintenance of local party infrastructure, volunteering, and campaigning. According to the ‘core supporter model’ parties therefore are more responsive towards this small group and do not risk scaring them off (Cox and McCubbins 1986). Since these core supporters pay special attention to policies (Schumacher, de Vries, and Vis 2013), changes to the official party stances on salient matters like immigration can likely trigger attention and anger. Activists, who are typically driven by ideological motives, will abandon overly opportunistic parties (Bernhardt, Duggan, and Squintani 2009). Thus, open position adjustment can be a costly strategy that undermines electoral support. Political parties need to balance popular demand and core supporter ideology when defining their policy positions before elections, as depicted in stage one of Figure 1. They first draft election manifestos in which they lay out their ideological stance. These manifestos are the most encyclopaedic statements of parties’ policy programmes and set the guidelines for the ensuing campaigns, potential coalition negotiations and for future legislative activity (Budge 1987; Eder, Jenny, and Müller 2017). But parties have to create their manifestos in a context of incomplete information about actual voter preferences. Guessing how the positions found in the manifestos will resonate with the electorate is hard because the preceding election is only a poor indicator of current voter preferences (Somer-Topcu 2009). Polls are used by political elites for decision making (e.g. Pickup and Hobolt 2015), but are at best a crude indicator of voter preferences. They deliver an estimate of voting intentions at the time of their execution that comes with sampling error, measurement error, and is influenced by subjective weighting decisions (see Hudson et al. 2004). Recent evidence moreover shows that increasing electoral volatility in multi-party systems and the emergence of new political actors, such as RRPs, have further elevated ambiguity over final election outcomes (Hobolt and Tilley 2016; Prosser and Mellon 2018). 1306 F. ELLGER AND H. KLÜVER It is only through election results – shown at stage two of Figure 1 – that parties receive complete information about voter preferences. Therefore, we conceive of election outcomes as information signals on voters’ preferred policy positions (Abou-Chadi and Orlowski 2016; Budge 1994). All political parties receive direct feedback on the popularity of their pre-electoral policy propositions. In an environment without transaction costs, we would expect all parties to immediately optimise positions based on their performance at the ballot. But parties are constrained since open change of policy positions can undermine their credibility and support. We argue that participation in coalition negotiations, the third stage highlighted in Figure 1, constitute an exception from this rule. In multiparty systems, more than one party typically forms a coalition government since no single party gains a legislative majority. Coalition governments are composed of at least two political parties that share executive offices (Strøm, Müller, and Bergman 2008: 6). Governing in coalitions can be intricate. Even though coalition parties join forces to gain control over executive offices, they typically pursue different policy objectives. Before entering government, coalition parties therefore engage in lengthy coalition negotiations that typically last a few weeks to bargain over the allocation of ministerial portfolios as well as over their joint policy agenda that is then written down in a public coalition agreement (Klüver and Bäck 2019; Strøm, Müller, and Bergman 2008). These negotiations are often labelled as a ‘black box’ since the bargaining process is highly confidential. We argue that these coalition negotiations provide an excellent opportunity for cabinet parties to optimise their positions in face of RRPs. Cabinet parties can profit from being responsive by aligning their propositions to preferences of the general electorate and accommodate to RRP positions without bearing the costs associated with open position switching. This is because coalition negotiations create a context of low accountability during which position switching cannot be traced back to individual actors. Position switching during negotiations is also appealing as it allows parties to enshrine their compromise in their joint coalition agreement. The agreements have implications for actual policy making throughout the government’s term in office. Even though coalition agreements are not legally binding, they define policy priorities of a cabinet and constrain the behaviour of each coalition party. While coalition negotiations are secret, the coalition agreements are publicly released, which enhances the compliance as shirking will lead to public blaming and shaming. Accordingly, Thomson (2001), Moury (2011) and Schermann and Ennser-Jedenastik (2014) show that election pledges to which coalition governments commit WEsT EuROPEAN POLiTiCs 1313 EP R Lsqs ij nij ij ij j ni log log log () =+ +        = = = ∑∑ 1 1 05 05 601 . .∗jj i j ij ij ij qs qs qs s ++ () ++ ()         608 0 5 602 607 0 5 . . −∗ log (4) In Online Appendix Section 6.3, we confirm that these expected positions are highly correlated with positions weighted by a cabinet’s portfolio allocation (Correlation = 0.91; p < 0.001). In an additional robustness check in Online Appendix Section 6.3, we also replicate our main analysis using these portfolio-weighted positions. There, we show that utilising seat shares instead of the portfolio distribution does not change our results. Following Gamson’s law (Gamson 1961), we assume that the expected position (EPs) derived from manifestos are strong predictors of the coalition agreement positions (CAPs). Both measure the same phenomenon, immigration policy, on exactly the same scale based on a comparable content analysis of political texts which are usually drafted just a few months apart. Given the short time span we are also confident to rule out a multitude of exogenous confounders. Bivariate regressions presented in Online Appendix Section 5.1 provide strong empirical support for our assumption. The expected position on immigration policy that we scale based on the pre-electoral manifestos is a major predictor of the actual immigration position that coalitions adopt in their coalition agreement. Operationalisation of independent variables We rely on two independent variables in order to test our theoretical argument that coalition governments adopt more restrictive immigration policies when RRPs are present in parliament. First, the main analysis utilises the dummy variable RRP Parliament, which captures if a radical right party has gained representation at a given election. Obtaining parliamentary representation constitutes a strong signal aimed at more established parties (Bischof and Wagner 2019). As a consequence, mainstream parties are facing RRPs in their daily parliamentary work. Once elected into the legislature, right-wing challengers also receive additional resources and media attention which increases the threat that they constitute for established parties (Abou-Chadi and Krause 2020). Therefore, we expect coalition governments to respond by adopting more restrictive immigration policies. Empirically, we find that right-wing challengers gain representation in 61% of all coalition formation instances observed in our sample. This ensures that there is sufficient variation within the sample, suggesting that not every cabinet formation we observe occurs with exposure to a radical party in parliament. We consider parties 1314 F. ELLGER AND H. KLÜVER that display strong nationalist and anti-immigration positions as right-wing challengers. The classification is based on prevalent work in the field, most importantly by Mudde (2007) and Roodujin et al. (2019). For Eastern Europe and recent elections, we also consulted election reports and related studies (e.g. Bustikova 2014). Online Appendix Section 1.1 lists all RRPs covered in our analysis. Additionally, we add the second independent variable RRP Vote Share to some models. It is a continuous measure of the combined vote share of all radical right parties. We utilise the continuous measure of RRP prevalence to assess if exposure to larger right-wing challengers has a more pronounced effect on immigration position change. This would imply that it is not the mere presence of challengers, but their electoral dominance that should trigger policy optimisation during negotiations. We do not have strong ideological priors on the expected results. That is because Abou-Chadi and Krause (2020) have already produced convincing causal evidence that small RRPs, just above the electoral threshold, can influence positions of mainstream parties. Certainly, a larger RRP should signal more support for anti-immigration propositions and thus could induce larger changes during the negotiations. On the other hand, a large RRP might reduce the uncertainty about the appeal of anti-immigration propositions among the general electorate prior to the election day. That is because a right-wing challenger polling at around 15% or more of the popular vote is very likely to be represented in parliament after the elections. If uncertainty about the appeal of tough positions on immigration is low prior to the election, establishment parties could already account for them in their pre-electoral party programmes. This could reduce their potential for adaptation during the cabinet formation period. Data and estimation Our analysis investigates the effect of radical right parties on coalition government’s immigration policy change during their formation. We aim to observe if cabinets that form while being exposed to RRPs in parliament shift to more restrictive immigration positions than other coalition cabinets. We therefore compiled a novel dataset based on preand post-electoral positions for 165 coalition cabinets across 24 Western and Eastern European democracies between 1980 and 2015. For the selection of countries, we followed the established standard in coalition research (see Andersson, Bergman, and Ersson 2014). The countries in our sample include Austria, Belgium, Bulgaria, Czech Republic, Denmark, Estonia, Finland, Germany, Greece, Hungary, Iceland, Ireland, Latvia, Lithuania, WEsT EuROPEAN POLiTiCs 1315 Luxembourg, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Sweden, and the United Kingdom. This country sample is characterised by variation in several institutional features, which strengthens the external validity of our findings. With regard to the collection of coalition agreements, we build on the work by Klüver, Bäck, and Krauss (2023). The authors have compiled an extensive dataset on the policy content of coalition agreements across Europe. Crucially, they have assessed the documents for coherence and comparability. For example, the authors ensured that each agreement fulfils certain criteria (i.e. the discussion of specific policy plans for the legislature or the public backing by all cabinet parties) and their country experts were trained to ensure validity and reliability. To this data, we add the information on election results and the representation of RRPs from ParlGov (Döring and Manow 2018), as well as pre-electoral manifesto positions based on the Manifesto Project (Budge and Klingemann 2001; Volkens et al. 2018). To test our hypothesis on immigration policy optimisation during coalition negotiations, we then assess how the post-electoral coalition agreement positions deviate from the pre-electoral expected positions that we generate on the basis of Gamson (1961). By studying the difference between expected positions based on pre-electoral election manifestos with actual government positions adopted in coalition agreements, we measure policy change within an extremely short time span. This not only makes it possible to restrict the influence of time-varying confounders. The following equation represents our main specification with a differenced outcome variable: ∆ Position RRPParliament ict ct ict ict,, , ,, ,, =+ ++αβ β ε 12 X (5) We are interested in the change in immigration policy position of government i in country c at year t as the dependent variable. This is represented by ΔPositioni,c,t, which is the difference between a government’s pre-electoral expected position (EP) and the post-electoral coalition agreement position (CAP). Note that positive values indicate a change towards more restrictive, and negative values indicate a move towards less restrictive policy positions during cabinet formation. The dependent variable thus captures the dynamics of the position change rather than the absolute position at any given time point. The key explanatory variable of interest is RRP Parliament, a dummy variable that indicates the presence of RRPs in parliament. The coefficient β1 on this variable captures the average difference in immigration policy position change associated with the presence of an RRP in parliament, relative to when an RRP is not present, holding other variables constant. 1316 F. ELLGER AND H. KLÜVER To account for invariant confounding across countries and time trends, we add ac,t, representing country (c) and cabinet formation year (t) fixed effects, respectively, in our specification. These fixed effects control for unobserved heterogeneity that may affect immigration policy positions, holding constant cross-country differences and time trends such as the general political climate, institutional factors, or macroeconomic conditions. By doing so, we ensure that our coefficient of interest, β1, captures the within-country, within-year variation attributable to the presence of radical right parties (RRPs), thereby making our estimates more robust to omitted variable bias. In addition to the state and year fixed effects, we also include Xi,c,t, a vector of cabinet-specific control variables. This includes the absolute number of cabinet parties (as a measure of the complexity of negotiations), a dummy that denotes minimum-winning coalitions (which are the most common form of coalitions, instead of minority or super-majoritarian cabinets) and the national unemployment rate during the year of cabinet formation as a measure of general economic conditions. Even though our estimation strategy relies on the analysis of a broad set of available data on coalition formation in Europe, the time series cross section analysis is demanding due to the limited number of country-cases and time points. The results therefore can come with a degree of statistical uncertainty. Finally, all standard errors are clustered by country and descriptive statistics are presented in Online Appendix Section 2. Empirical analysis In Table 1, we present the first empirical test of our argument. We run a series of regression models that rely on a differenced outcome measure and include country and year fixed effects and clustered standard errors by country. The dichotomous RRP Parliament dummy acts as the independent variable. For the differenced outcome variable, positive values denote a change towards more restrictive policy positions, while negative values indicate a change towards more progressive positions on immigration. Again, these changes typically occur over a few weeks around election day. We run separate regressions for the full sample of cabinets (n = 165) and a restrictive sample (n = 132) that excludes cabinets which did not explicitly state policy positions in either party manifestos or the coalition agreement. For all models of Table 1, we observe that the entry of radical right parties into government is associated with a move to more restrictive policy positions during the coalition negotiation phase. In three of four models, this association is significant above conventional levels of WEsT EuROPEAN POLiTiCs 1317 statistical uncertainty. Importantly, the effect of radical right party presence in parliament on shifts to more restrictive immigration policy is highly robust once we restrict the sample to those 132 cabinets that did cover migration policy in both their election programmes and joint agreements. The coverage of immigration by establishment parties in their policy platforms indicates a heightened salience of this issue. In terms of effect magnitude, we find policy shifts of 0.4 to 0.7 points on the Kim/ Fording scale – with a theoretical maximum of 2 points – to be quite substantial. A maximum change of 2 points would constitute a change from the most extreme permissive pre-electoral position to the most extreme restrictive post-electoral agreement position. Again, we observe this change over the short time span between drafting electoral programmes and negotiating the joint coalition agreement. In a second step, we add RRP Vote Share, the vote share of all radical right parties, as an additional independent variable to test if position change becomes more pronounced once radical challengers gain larger vote shares. According to the results presented in Table 2, we find that the size of RRPs does yield larger policy shifts. If anything, bigger challengers appear to lead to marginally less restrictive change. However, most estimates are not significantly discernible from zero. Why would mere exposure to RRPs, but not their size influence a cabinet’s immigration plans? Providing conclusive evidence on this finding is beyond the scope of our article. However, one plausible explanation could be that establishment parties try to cater towards RRP supporters as long as they remain marginal competitors. The accommodative behaviour of mainstream cabinets might be a viable strategy as long as the radical competitors are small (Meguid 2005). Once RRPs have established themselves in the party system and become major competitors for offices, Table 1. policy change for rrps in parliament. Dependent variable: position change (KF) position change (log) (1) (2) (3) (4) rrp parliament 0.422* 0.724*** 0.749 1.350** (0.221) (0.176) (0.614) (0.523) MWc −0.170 −0.246 −0.307 −0.505 (0.133) (0.145) (0.330) (0.403) no. cabinet parties 0.0041 0.061 0.129 0.329 (0.078) (0.081) (0.234) (0.251) unemployment −0.024 −0.027 −0.052 −0.080 (0.019) (0.023) (0.058) (0.084) Fixed effects country + Year country + Year country + Year country + Year observations 165 132 165 132 sample Full restr. Full restr. adjusted r20.319 0.399 0.182 0.232 note: *p < 0.1; **p < 0.05; ***p < 0.01. 1318 F. ELLGER AND H. KLÜVER mainstream parties could opt for differentiation in order to sharpen their own policy profiles. In line with this proposition, Abou-Chadi and Krause (2020) have presented convincing causal evidence on the accommodative behaviour of mainstream parties in response to small RRPs that barely entered parliament around the electoral threshold. Cabinet characteristics and the influence of RRPs After showing that coalition governments change their position on immigration policies when facing radical challengers in parliament, we now explore what cabinet characteristics could influence the effect. In particular, we focus on four coalition attributes that could shape the degree of policy influence by RRPs. First, we assess whether the electoral performance of a cabinet moderates the influence of RRPs on policy change. One could expect that cabinets which have lost votes are prone to pressures from the radical right. We thus create the binary variable Electoral Loss to indicate if cabinet parties have on average lost votes compared to the previous election. A value of one shows that the cabinet has experienced an electoral loss. Next, we focus on the general ideology of a cabinet as a potential moderator. Here, one could assume that especially centre-right governments are prone to pressures from radical parties because they are more likely to lose supporters to the far right. To differentiate between left and right-leaning governments, we create the binary variable Left Cabinet. We measure cabinet ideology based on the weighted average position on the general left-right scale (or ‘rile’-scale) of each cabinet parties’ manifesto. Cabinets with an average ideology score lower than 0 are coded as being Table 2. rrp size and policy change. Dependent variable: position change (KF) position change (log) (1) (2) (3) (4) rrp parliament 0.465* 0.919*** 0.784 1.732*** (0.246) (0.184) (0.671) (0.605) rrp Vote share −0.006 −0.020** −0.004 −0.039 (0.010) (0.009) (0.023) (0.026) MWc −0.172 −0.247* −0.309 −0.508 (0.133) (0.138) (0.332) (0.393) no. cabinet parties 0.004 0.087 0.129 0.379 (0.079) (0.083) (0.235) (0.259) unemployment −0.026 −0.031 −0.053 −0.088 (0.020) (0.026) (0.061) (0.091) Fixed effects country + Year country + Year country + Year country + Year observations 165 132 165 132 sample Full restr. Full restr. adjusted r20.319 0.435 0.175 0.240 note: *p < 0.1; **p < 0.05; ***p < 0.01. WEsT EuROPEAN POLiTiCs 1319 left cabinets. Third, we suggest that the length of the coalition negotiations themselves could impact the influence of RRPs on policy changes. We thus have researched the length between the election date and government formation, measured in days. We have been able to retrieve information on the formation duration for 161 out of 165 cabinets in the full sample and 128 of 132 cabinets in the restricted sample. In the next step, we conduct a median split to create the binary indicator Long Negotiation. Longer negotiation periods should exert cabinet parties with a longer time period of implicit pressure from the radical right while providing them with more opportunities to shift their policies. Finally, we focus on the coalition type itself. Minority cabinets do not possess the parliamentary strength to implement legislation alone, thus making them rely on parts of the opposition for policy making. This could make them more prone to shift towards restrictive policies in order to establish support along the line of immigration policy. We replicate the prior OLS regression models, but interact each of the binary cabinet characteristics with the RRP Parliament dummy variable. We measure policy shifts on the Kim/Fording-scale and again include cabinet control variables, country and year fixed effects in the models. The standard errors are clustered by country. The results are presented in Table 3. The first two models paint an inconclusive picture of the moderating effect of cabinet’s electoral performance. In the full sample, the significant interaction terms suggest that losing cabinets are more likely to adjust their position. However, in the restricted sample, the coefficient of the interaction is smaller and lacks statistical significance. We interpret this as suggestive, but not conclusive evidence for the proposition that cabinets that did lose electorally are more prone to accommodative strategies in light of the radical right. For the remaining three cabinet-specific moderating variables, the analyses provide more consistent results. Looking at cabinet ideology, the insignificant interaction terms showcase that left-leaning cabinets are neither more nor less likely to amend their position in the face of RRPs. The direction of the effect is negative, suggesting that left-leaning cabinets shift towards more progressive positions compared to right-leaning coalitions. While the direction of the interaction effect aligns with our theoretical intuition, the estimates are not statistically significant. The same is true for the findings regarding government formation duration. Lengthy negotiations in face of RRPs are not associated with more or less intense policy position changes. The coefficients in models five and six are close to zero and lack statistical precision. Finally, we observe positive estimates for the interaction between RRP presence and the minority cabinet dummy. Minority cabinets, which are more reliant on opposition party support, do shift to significantly more restrictive 1320 F. ELLGER AND H. KLÜVER Table 3. rrp influence by cabinet characteristics. Dependent variable: policy change (KF) (1) (2) (3) (4) (5) (6) (7) (8) rrp parliament:lost Votes 0.40* (0.21) 0.21 (0.24) rrp parliament:left cabinet −0.30 (0.24) −0.49 (0.32) rrp parliament:long Formation 0.11 (0.19) −0.11 (0.16) rrp parliament:Minority cabinet 0.55** (0.20) 1.13** (0.45) rrp parliament 0.31 0.67*** 0.57** 0.95*** 0.38 0.81*** 0.42* 0.73*** (0.23) (0.19) (0.26) (0.24) (0.33) (0.25) (0.21) (0.16) lost Votes −0.12 (0.13) −0.05 (0.16) left cabinet 0.37** (0.15) 0.42* (0.21) long Formation −0.19 (0.12) −0.003 (0.15) Minority cab. −0.74*** (0.22) −1.55*** (0.54) MWc −0.14 −0.24 −0.16 −0.24 −0.14 −0.23 −0.44** −0.57** (0.14) (0.15) (0.13) (0.14) (0.14) (0.15) (0.18) (0.21) no. cab. parties −0.01 0.05 0.01 0.07 0.001 0.04 −0.03 −0.02 (0.08) (0.09) (0.08) (0.08) (0.08) (0.09) (0.09) (0.08) unemployment −0.02 −0.03 −0.03 −0.02 −0.02 −0.02 −0.02 −0.03 (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) Fixed effects country + Year country + Year country + Year country + Year country + Year country + Year country + Year country + Year observations 165 132 165 132 161 128 165 132 sample Full restr. Full restr. Full restr. Full restr. adjusted r20.33 0.40 0.34 0.42 0.28 0.36 0.35 0.48 note: *p < 0.1; **p < 0.05; ***p < 0.01. WEsT EuROPEAN POLiTiCs 1321 positions whenever RRPs are represented in parliament. This suggests that RRPs are especially influential of government policies when the coalition cabinet itself has no strong electoral mandate. Overall, the empirical analyses provide ample support for our theoretical argument. Coalitions perceive the presence of right-wing challengers as a signal of refusal for their prior positions on immigration policy. They immediately react by adopting a tougher stance on immigration in comparison to those governments without similar exposure. In addition, we present complementing evidence on the role of RRP size and their influence conditional on cabinet characteristics. In the auxiliary analyses, we find that larger RRPs do not have stronger influence on the degree of policy change than smaller RRPs in parliament. This aligns with prior research showing that even small radical right challenger parties can influence their competitors once they gain representation. With regard to the cabinet characteristics, we find some evidence that those coalitions under electoral are switching towards more restrictive policies. In particular, minority coalition cabinets and cabinets that lost votes show accommodative behaviour in face of RRPs. Further robustness and validity checks In the last empirical section, we summarise supporting evidence on the relationship between radical challengers and coalition governments’ policy change on immigration during negotiations. We probe if cabinets with and without exposure to RRPs are comparable within our sample. We conduct two tests showing that both groups do not differ along relevant characteristics. First, Figure 2 visualises the similarities of their pre-electoral positions. The plots depict a coalition cabinet’s immigration position across three time points: The post-electoral coalition agreement position CAP, the pre-electoral expected position EP(t0), and the expected position at the prior election EP(t-1). Crucially, we observe no differences across both groups when moving from EP(t-1) to EP(t0). Only in the coalition agreement positions that are drafted after the elections, significant differences between positions taken on by cabinets with and without RRP exposure occur. This pattern holds for the Kim-Fording scale on the upper pane, as well as the log scale on the lower pane of the figure. Second, we explore if the cabinets diverge along a set of other relevant dimensions. We therefore conducted a range of t-tests to check their balance on potentially relevant covariates. Table 4 summarises the results. For each variable, we showcase the average value for cabinets with and without RRP presence in parliament. The balance column then indicates if significant differences across both groups exist according to a two-tailed t-test (cutoff at p < 0.05). 1322 F. ELLGER AND H. KLÜVER Figure 2. immigration policy positions across cabinets. note: the figure shows trends in immigration positions for cabinets with rrps in parliament in green and cabinets without exposure to rrps in red. We observe that there are no significant differences across both groups for the pre-electoral eps at the current election (t0) or at the preceding election (t-1). significant differences only emerge for post-electoral coalition agreement positions. WEsT EuROPEAN POLiTiCs 1329 Kriesi, Hanspeter, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, and Timotheos Frey (2006). ‘Globalization and the Transformation of the National Political Space: Six European Countries Compared’, European Journal of Political Research, 45:6, 921–56. Laakso, Markku, and Rein Taagepera (1979). ‘Effective Number of Parties a Measure with Application to West Europe’, Comparative Political Studies, 12:1, 3–27. Laver, Michael (1998). ‘Models of Government Formation’, Annual Review of Political Science, 1:1, 1–25. Lowe, Will, Kenneth Benoit, Slava Mikhaylov, and Michael Laver (2011). ‘Scaling Policy Preferences from Coded Political Texts’, Legislative Studies Quarterly, 36:1, 123–55. Martin, Lanny W., and Georg Vanberg (2014). ‘Parties and Policymaking in Multiparty Governments: The Legislative Median, Ministerial Autonomy, and the Coalition Compromise’, American Journal of Political Science, 58:4, 979–96. McGann, Anthony J., and Herbert Kitschelt (2005). ‘The Radical Right in the Alps: Evolution of Support for the Swiss SVP and Austrian FPÖ’, Party Politics, 11:2, 147–71. Meguid, Bonnie (2005). ‘Competition between Unequals: The Role of Mainstream Party Strategy in Niche Party Success’, American Political Science Review, 99:3, 347–59. Meyer, Thomas M. (2013). Constraints on Party Politics. Colchester: ECPR Press. Moury, Catherine (2011). ‘Coalition Agreement and Party Mandate: How Coalition Agreements Constrain the Ministers’, Party Politics, 17:3, 385–404. Moury, Catherine, and Arco Timmermans (2013). ‘Inter-Party Conflict Management in Coalition Governments: Analyzing the Role of Coalition Agreements in Belgium, Germany, Italy and The Netherlands’, Politics and Governance, 1:2, 117–31. Mudde, Cas (1999). ‘The Single-Issue Party Thesis: Extreme Right Parties and the Immigration Issue’, West European Politics, 22:3, 182–97. Mudde, Cas (2007). Populist Radical Right Parties in Europe. Cambridge, UK; New York: Cambridge University Press. Müller, Wolfgang C., and Kaare Strom (2003). Coalition Governments in Western Europe. Oxford: Oxford University Press. Pedersen, Mogens N. (1979). ‘The Dynamics of European Party Systems: Changing Patterns of Electoral Volatility’, European Journal of Political Research, 7:1, 1–26. Pickup, Mark, and Sara B. Hobolt (2015). ‘The Conditionality of the Trade-off between Government Responsiveness and Effectiveness: The Impact of Minority Status and Polls in the Canadian House of Commons’, Electoral Studies, 40, 517–30. Powell, G. Bingham (2000). Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven: Yale University Press. Powell, G. Bingham, and Guy D. Whitten (1993). ‘A Cross-National Analysis of Economic Voting: Taking account of the Political Context’, American Journal of Political Science, 37:2, 391. Prosser, Christopher, and Jonathan Mellon (2018). ‘The Twilight of the Polls? A Review of Trends in Polling Accuracy and the Causes of Polling Misses’, Government and Opposition, 53:4, 757–90. Roodujin, Matthijs, Stijn van Kessel, Caterina Froio, Andrea Pirro, Sarah de Lange, Daphne Halikiopoulou, Paul Lewis, Cas Mudde, and Paul Taggart (2019). ‘The PopuList: An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe.’ http://www.popu-list.org. 1330 F. ELLGER AND H. KLÜVER Rydgren, Jens (2008). ‘Immigration Sceptics, Xenophobes or Racists? Radical Right-Wing Voting in Six West European Countries’, European Journal of Political Research, 47:6, 737–65. Saalfeld, Thomas (2008). ‘Institutions, Chance and Choices: The Dynamics of Cabinet Survival in the Parliamentary Democracies of Western Europe (1945-99),’ in Müller Wolfgang C, Strøm Kaare and Bergman Torbjörn (eds.). Cabinets and Coalition Bargaining: The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, 327–68. Schain, Martin A (2006). ‘The Extreme-Right and Immigration Policy-Making: Measuring Direct and Indirect Effects’, West European Politics, 29:2, 270–89. Schermann, Katrin, and Laurenz Ennser-Jedenastik (2014). ‘Explaining Coalition-Bargaining Outcomes: Evidence from Austria, 2002-2008’, Party Politics, 20:5, 791–801. Schofield, Norman, and Michael Laver (1985). ‘Bargaining Theory and Portfolio Payoffs in European Coalition Governments 1945-83’, British Journal of Political Science, 15:2, 143–64. Schumacher, Gijs, Catherine E. de Vries, and Barbara Vis (2013). ‘Why Do Parties Change Position? Party Organization and Environmental Incentives’, Journal of Politics, 75:2, 464–77. Somer-Topcu, Zeynep (2009). ‘Timely Decisions: The Effects of past National Elections on Party Policy Change’, Journal of Politics, 71:1, 238–48. Soroka, Stuart N., and Christopher Wlezien (2010). Degrees of Democracy: Politics, Public Opinion, and Policy. Cambridge: Cambridge University Press. Spoon, Jae-Jae, and Heike Klüver (2019). ‘Party Convergence and Vote Switching: Explaining Mainstream Party Decline across Europe’, European Journal of Political Research, 58:4, 1021–42. Stimson, James A., Michael B. Mackuen, and Robert S. Erikson (1995). ‘Dynamic Representation’, American Political Science Review, 89:3, 543–65. Stokes, Donald E. (1963). ‘Spatial Models of Party Competition’, American Political Science Review, 57:2, 368–77. Strøm, Kaare, and Wolfgang C. Müller (1999). ‘The Keys to Togetherness: Coalition Agreements in Parliamentary Democracies’, Journal of Legislative Studies, 5:3–4, 255–82. Strøm, Kaare, Wolfgang C. Müller, and Torbjörn Bergman (2008). Cabinets and Coalition Bargaining: The Democractic Life Cycle in Western Europe. Oxford: Oxford University Press. Tavits, Margit (2007). ‘Principle vs. Pragmatism: Policy Shifts and Political Competition’, American Journal of Political Science, 51:1, 151–65. Tavits, Margit (2008). ‘The Role of Parties’ past Behaviour in Coalition Formation’, American Political Science Review, 102:4, 495–507. Thomson, Robert (2001). ‘The Programme to Policy Linkage: The Fulfilment of Election Pledges on Socio-Economic Policy in The Netherlands, 1986-1998’, European Journal of Political Research, 40:2, 171–97. Van Spanje, Joost (2010). ‘Contagious Parties: Anti-Immigration Parties and Their Impact on Other Parties’ Immigration Stances in Contemporary Western Europe’, Party Politics, 16:5, 563–86. Volkens, Andrea Werner Krause, Pola Lehmann, Theres Matthieß, Nicolas Merz, Sven Regel, and Bernhard Weßels (2018). The Manifesto Data Collection. Manifesto Project (MRG/CMP/MARPOR). Version 2018b. WEsT EuROPEAN POLiTiCs 1331 Warwick, Paul V., and James N. Druckman (2001). ‘Portfolio Salience and the Proportionality of Payoffs in Coalition Governments’, British Journal of Political Science, 31:4, 649. Warwick, Paul V., and James N. Druckman (2006). ‘The Portfolio Allocation Paradox: An Investigation into the Nature of a Very Strong but Puzzling Relationship’, European Journal of Political Research, 45:4, 635–65. Williams, Michelle Hale (2006). The Impact of Radical Right-Wing Parties in West European Democracies. New York: Palgrave Macmillan US.