Scholarly article on topic 'Contextual perceived group threat and radical right-wing populist party preferences: Evidence from Switzerland'

Contextual perceived group threat and radical right-wing populist party preferences: Evidence from Switzerland Academic research paper on "Sociology"

0
0
Share paper
Academic journal
Research & Politics
OECD Field of science
Keywords
{""}

Academic research paper on topic "Contextual perceived group threat and radical right-wing populist party preferences: Evidence from Switzerland"

Research Article

Contextual perceived group threat and radical right-wing populist party preferences: Evidence from Switzerland

Research and Politics January-March 2016: 1-7 © The Author(s) 2016 DOI: 10.1177/2053168016635670 rap.sagepub.com

Carl C. Berning

Abstract

Existing studies suggest that perceived group threat is an important influence on radical right-wing populist party preferences. However, most have focused on perceived group threat at the individual level, overlooking the ideological climate. I examine how an ideological climate of group threat perception as a contextual factor can shape individual preferences for radical right-wing populist party preferences. I argue that above and beyond personal perceived group threat, the prevalence of local perceived group threat exerts a normative influence on personal preferences. Using voting preferences for the Swiss People's Party, I employ multilevel structural equation modeling to examine the theoretical model. I find clear evidence for a contextual effect of perceived group threat on individual-level Swiss People's Party preferences.

Keywords

Radical right-wing populist parties, perceived group threat, multilevel structural equation modeling, ideological climate, subnational comparison

Introduction

Radical right-wing populist (RRP) parties have been successful in many European countries over the last two decades, and there has been considerable research on the sources of their appeal to voters. A core selling point of RRP parties is their strong opposition to immigration (Mudde, 2007). Micro-level determinants explaining individual preferences for RRP parties have been well documented, and scholars agree that perceived group threat - generally defined as expected negative consequences from immigration - is a major predictor (Ivarsflaten, 2008). However, previous research on local variation in RRP party preferences has mostly been limited to structural determinants, adopted from cross-national research (Arzheimer and Carter, 2006), and there has been little attention to the possible role of an ideological climate as an antecedent of anti-immigrant attitudes and a characteristic of the social context (e.g. Christ et al., 2013; Sarrasin et al., 2012). In this article, I examine whether an ideological climate of local perceived group threat has an independent effect on

individual RRP party preferences, beyond individual group threat perceptions.

My contribution to the literature is both theoretical and empirical: I extend previous research on RRP parties by developing an explanation of how the social environment can contribute to individual RRP party preferences. This explanation explicitly acknowledges that individuals rely on one another for guidance, and that the prevailing attitudes of families, circles of friends, among colleagues, or other social encounters affect individual political opinions (Huckfeldt et al., 2005: 21ff.). I provide empirical evidence for the link between an ideological climate of perceived group threat and support for one

Department of Political Science, Johannes Gutenberg University of Mainz, Germany

Corresponding author:

Carl C. Berning, Department of Political Science, Johannes Gutenberg University of Mainz, Jakob-Welder-Weg 12, 55099 Mainz, Germany. Email: berning@politik.uni-mainz.de

Creative Commons NonCommercial-NoDerivs CC BY-NC-ND: This article is distributed under the terms of the Creative Commons Attribution- NonCommercial-NoDerivs 3.0 License (http://www.creativecommons.org/licenses/by-nc-nd/3.0/) which permits non-commercial use, reproduction and distribution of the work as published without adaptation or alteration, without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).

of the most successful RRP parties in Western Europe, the Swiss People's Party (SVP).

In this study I focus on differences between local contexts across Swiss districts. The focus on a single country enables me to rule out cross-national differences and maintain socioeconomic and structural homogeneity. Swiss districts provide an appropriate case to test how the ideological climate of perceived group threat influences individual support for RRP parties for two reasons: the SVP has consistently opposed immigration; and there are large differences in SVP support between districts (Kriesi et al., 2005; Lubbers, 2000).

Theoretical framework

People who feel threatened by immigration support RRP parties in order to express their concerns about immigration and to influence legislation in line with their policy preferences (Mughan and Paxton, 2006). Generally, a preference for RRP parties is associated with other types of motivations, such as authoritarianism or political distrust (Mudde, 2007). Nevertheless, none of these predictors appear to confound the impact of perceived group threat as the key attitudinal motivation (Ivarsflaten, 2008).

The theoretical framework of perceived group threat as the principal attitudinal motivation for RRP party preferences stems from group threat and group conflict theory (Lubbers et al., 2002; Rydgren, 2008). On the one hand, threat arises from competition over economic or political resources, i.e. realistic or materialistic threat. On the other hand, people perceive group threat to intangible goods, i.e. symbolic or cultural threat (Blalock, 1967; Sherif et al., 1961). The two sets of explanations are not mutually exclusive, and previous conceptualizations of perceived group threat suggest a common theoretical framework (Stephan and Renfro, 2002).

Previous research has found significant support for the rationale described above. However, individual information and concerns about immigration are derived not only from personal experiences, but also social encounters (Huckfeldt et al., 2005). In fact, "individuals are embedded in everyday environments that provide normative and ideological reference knowledge guiding their thinking about societal phenomena such as immigration" (Green and Staerkle, 2013: 876). Individuals learn about their environment's threat perception either through interpersonal communication or observation (Marsh, 2002). If the prevailing level of group threat perception is taken as normative reference (Blau, 1960; Huckfeldt and Sprague, 1995), the social climate will induce voters to support RRP parties in line with the prevalence of concern about immigrants and immigration in their environment. The contextual effect can thus be conceptualized as "social resonance, since the underlying intuition is one of reinforcement of a property possessed by the individual

through repeatedly encountering the same property in the environment" (Sprague, 1982: 101).

Thus, we need to consider that the prevailing attitude of the proximal social environment, i.e. families, friends, colleagues, or other social encounters, can exert an independent effect on individual RRP party preferences. The social-environmental threat perception is a collective property, which is conceptually and statistically more than the sum of individual attitudes (Blau, 1960; Welzel and Deutsch, 2012). An independent effect of an ideological climate must entail more than just the accumulated individual perceptions of immigrants as a threat to in-group interests. The ideological climate is an aggregate level concept of predominant attitudes, and is expected to influence individuals beyond internalized beliefs (Green and Staerklé, 2013; Sarrasin et al., 2012).

Collective social-environmental threat perception, or the ideological climate, affects individuals' decision to vote for RRP parties in several ways. A key mechanism is that the group threat perception of the social environment provides cues about the average support for RRP parties in the area, and can persuade individuals to vote for RRP parties in compliance with the social norm. The group threat perception of the social environment also supports the legitimacy of RRP party preferences and thus encourages voting for RRP parties. The social climate of perceived threat also raises the salience of individuals' threat perceptions and thus reinforces the effect of individual perceived threat on the support for RRP parties. Contextual effects are not unique to the link between RRP party preferences and perceived group threat, but they are especially fruitful in analyzing RRP party support given the consensus on threat as a major attitudinal predictor (Ivarsflaten, 2008).

An alternative argument is that average group threat perception will increase individuals' underlying threat perceptions as opposed to through the mechanisms outlined above. Such an argument would imply that the average group threat perception simply reflects the sum of individual perceived group threat. Moreover, residential self-selection might lead to apparent contextual effects in party preferences. Thus, it is important to control for socioeconomic and demographic variables that simultaneously influence where people live and the degree to which they support RRP parties when trying to evaluate the impact of contextual perceived group threat.

Data and method

I use data from the Swiss Electoral Study (SELECTS, 2011). SELECTS is based on a national representative sample of Swiss residents eligible to vote in the national elections. The data include relevant measures as well as geo-codes for the respondents' municipalities. An analysis

of average perceived threat across municipalities for these data would be problematic, as 60% of the municipalities only have information for one respondent. Hence, I aggregate municipalities to districts, an administrative unit smaller than cantons, resulting in 145 distinct second-level units. Twenty-eight respondents are excluded due to missing values for all analytical variables, resulting in a sample size of Ntadv = 4363 individuals nested in Ndis = 145 districts, with a mean of 29 respondents per district.

Measures

Dependent variable

I use a single item to assess RRP party preference. The respondents were asked: Please indicate the probability of voting for the SVP, when 0 represents a very small probability and 10 is a very large probability.

Independent variables

Perceived group threat is measured with four items on a 5-point Likert-type scale with 1 totally agree to 5 totally disagree. The respondents were asked if they agree that there are too many Muslim immigrants in Switzerland (Q1), that migrants exacerbate the job market situation (Q2), that Swiss culture is vanishing due to immigration (Q3), and that there is violence and vandalism due to young immigrants (Q4). These measures correspond to items in related research on perceived group threat (Scheepers et al., 2002; Schlueter et al., 2013). All items are rescaled so that higher values imply higher perceived group threat.

Control variables

I introduce the following potentially confounding factors. Since women are less likely to prefer RRP parties (Givens, 2004), I use a dichotomous measure for gender with female as the reference category. Existing studies show that RRP parties do better among voters who are vulnerable towards out-group competition, e.g. younger working-class individuals with little education (Betz, 1994; Coffe and Voorpostel, 2010). Therefore, I include age in years in my model as well as education, assessed in eight categories, ranging from a value of 1 indicating compulsory education only (as the reference category) up to a value of 8 corresponding to University level education. Further, I control for social class with a simplified Goldthorpe classification (manual worker, self-employed, routine nonmanual worker and professionals, with not in paid work as the reference category). Moreover, since the SVP started as a grassroots party with strong support in rural areas (Kriesi et al., 2005), I include a measure capturing the difference between rural and urban areas (1 rural municipality as reference; 2 isolated town/city; 3 other

municipality in an agglomeration; and 4 central city in an agglomeration).

Finally, preferences for RRP parties and perceived threat might be related to district-level characteristics. I introduce the percentage offoreign population, measured by the proportion of permanent residents with a citizenship other than a European Union country and the unemployment rate as potential sources for inter-group competition.1 Following previous research, I further control for the linguistic classification of the place of residence, with 1 German and 0 other linguistic areas.

Analysis

I test the postulated influence of environmental perceived group threat on the self-rated probability to vote for the SVP, controlling for confounding variables, by a two-level structural equation model for complex sample designs, with individuals nested within districts.2 This methodology generates more appropriate standard errors, taking the nested data structure into account and corrects for sampling and measurement error (Marsh et al., 2009). All analyses are based on full-information maximum-likelihood estimates (Enders and Bandalos, 2001), with standard errors that are robust to non-normality and non-independence conducted in Mplus 7 (Muthen and Muthen, 1998-2012). My core interest is the contextual effect of perceived group threat, which is given mathematically by the difference of the between and within effects (Raudenbush and Bryk, 2002: 139).

To assess the goodness of fit, I provide the following commonly used indices: x2/df, root mean square error of approximation (RMSEA), and comparative fit index (CFI) (Bentler, 1990; Boomsma, 2000; Hu and Bentler, 1999). Models withx2/df < 5, RMSEA < 0.06, and CFI > 0.95 are considered to have a good fit to the data. Model comparisons are based on Satorra-Bentler scaled differences of x2-values and degrees of freedom (Satorra and Bentler, 2001). The model fit indices above are not available for random slope models.

Results

The results of the measurement model for perceived group threat indicate a good fit to the data (x2/df = 2.88, RMSEA = 0.021, and CFI = 0.996). As described earlier, I measure perceived group threat as a latent variable on the individual and district level. To ensure that I measure the same construct, I need to establish cross-level metric invariance (Ludtke et al., 2008). I find that not constraining the factor-loadings to be equal across levels did not significantly decrease the model fit (A/2 = 1.46, Adf = 3,p > 0.05).

The results of the structural models are presented in Table 1. Model 1 presents the results of the random intercept model. The overall model fits well to the data with x2/ df = 4.11, RMSEA = 0.027, and CFI = 0.958. On the

Table 1. Structural model (unstandardized regression coefficients (b) and standard errors (SE)).

SVP preferences

Model 1

Model 2

Person-level

Male Age

Compulsory education Basic vocational training Vocational education Diploma school High school

Higher vocational training Vocational university University Not in paid work Manual worker Routine non-manual Self-employed Professional Rural municipality Isolated town/city Agglomeration Central city Perceived group threat District-level

Foreign population

Unemployment

German

Perceived group threat Contextual effect

Perceived group threat Cross-level interaction

Perceived group threat

(district) x perceived group

threat (person)

0.703 -0.015 Ref. -0.278 0.155 -0.285 -1.140 -0.245 -1.601 -1.556 Ref. 0.612 -0.063 0.839 -0.291 Ref. -1.162 -0.180 -0.679 2.189

0.034 -14.528 0.128 3.912

0.136*** 0.006**

0.223***

0.247***

0.225***

0.192** 0.162 0.251** 0.161

0.489* 0.154 0.148*** 0.138***

0.025 8.820 0.245 0.775***

0.700 -0.014 Ref. -0.184 0.141 -0.265 -1.170 -0.263 -1.660 -1.660 Ref.

0.594 -0.090 0.844 -0.305 Ref. -1.186 -0.153 -0.643 2.268

0.039 -12.028 -0.050 3.951

0.138*** 0.006*

0.228***

0.244***

0.235***

0.190** 0.164 0.251** 0.159

0.465* 0.133 0.140*** 0.134***

0.021 6.991 0.192 0.567***

0.542**

0.621***

68411.374 68775.086

68375.983 68733.314

N = 4363 nested in 145 districts; clustered SE; *p < 0.05 **p < 0.01 ***p < 0.001; all coefficients of individual-level control variables are total effects (i.e. the sum of direct and indirect effects via perceived group threat); ref.= reference category; AIC = the Akaike information criterion; BIC = the Bayesian information criterion.

individual level, the results show that perceived group threat has a strong and positive effect on the probability to vote for the SVP. At the district level, I find a significant positive effect of the average group threat perception on the average SVP voting probability. In line with the theoretical expectation, the contextual effect, i.e. the difference of the between and within effect, is positive and significant. In other words, the average perceived group threat in Swiss districts has a positive effect on the individual preference for the SVP.

I proceed by testing the moderating effect of the average group threat perception on the link between individual perceived group threat and SVP voting preferences. Model 2 reports the results of the random slope model. As expected, I find a significant positive cross-level interaction, where

the effect of perceived group threat on SVP voting probability is stronger in districts with higher average perceived group threat. Figure 1 summarizes the results of model 2 for the focal variables. The rectangles depict observed variables and the ellipses indicate latent variables. The arrows between the shapes illustrate paths and the black dots indicate the random slope (S) respectively random intercepts.

In all analyses I included a set of control variables to account for potential confounding factors. The results are generally in line with previous research. I find that men are more likely to prefer the SVP. Furthermore, the probability of voting for the SVP decreases with age. Explanations of social class predicting RRP preferences are congruent with my findings. People from the working class and the so-called petit bourgeoisie have a higher probability of voting for the

Figure 1. Multilevel structural equation model (model 2). All coefficients are statistically significant (p< 0.001). Control variables and residual variances are not shown.

Between level (districts)

Qi Q2 Q3 Q4

Within level (individuals)

2.268 1 A M t RRP party

preference

SVP and those with higher education have a lower probability of voting for the SVP. The findings reveal that respondents living in a city are less likely to vote for the SVP. At the district level, I find no significant effects of the control variables on the average probability to vote for the SVP.

Discussion

In this study, I examine to what extent the local group threat perception, i.e. the ideological climate, influences individual RRP party preferences. To test this relationship, I draw upon survey data from Switzerland. My research strategy offers an alternative theoretical explanation and provides empirical evidence that contributes to a deeper understanding of how the local context affects individual RRP party preferences.

I disentangle the effect of perceived group threat on RRP party preferences to between and within components, and show that an ideological climate of perceived group threat has substantial influence on individual RRP party preferences, over and above individual group threat perceptions. In other words, a person living in a district with a higher collective perceived group threat has a higher probability, ceteris paribus, of voting for the SVP, in comparison with a resident of a district where perceived group threat is not as socially prevalent. Moreover, I find that the group threat perception of the social environment moderates the individual-level relationship between perceived group threat and the probability of voting for the SVP.

The results presented here have a number of limitations. My findings may not generalize beyond the Swiss context, but previous comparative research shows that the electorate

of RRP parties in Western Europe seems to be motivated by the same factors. Whether separate national conditions may moderate the effect of an ideological climate on RRP party preferences deserves further scientific attention. An ideological climate effect diffuses with social interaction; therefore, one can speculate that preferences for RRP parties will be less affected by the group threat perception of the social environment in socially disintegrated, atomized societies than in more integrated societies. It would be helpful if future cross-national surveys on RRP party preferences cover geo-codes at the local level.

Another issue refers to the underlying mechanism, i.e. how the prevailing group threat perception affects individual RRP preferences. I have no direct information on the quantity of actual interactions or on the degree to which concerns about immigration play a major role in these. I am also bounded by cross-sectional data, and can only estimate inter-individual correlations, but my argument is in line with previous research regarding the temporal order of perceived group threat and RRP party preferences (Berning and Schlueter, 2016). It might be beneficial to complement the findings presented in this study with other research designs. For example, network analysis can allow researchers to directly measure information flows and provide a more comprehensive empirical representation of the underlying mechanisms. Additionally, one could conduct a content analysis of news reports on RRP parties and related topics. Previous research suggests that media attention can effectively mobilize the RRP electorate (Boomgaarden and Vliegenthart, 2007).

These limitations aside, my findings contribute to a better understanding of how local ideological contexts can

affect individual RRP party preferences. Existing evidence is mostly limited to political and economic conditions, omitting the potential effect of the prevailing attitudes in a local environment. This study shows that differences in the ideological climate of perceived group threat appear to be relevant in accounting for cross-regional variation of individual RRP party preferences. In sum, the average perception in a given local environment that immigrants are a threat for in-group interests constitutes a social property with an independent effect on individual RRP preferences, beyond individual threat perceptions.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

1. I also estimated models with an alternative measure of foreign population, i.e. the percentage of permanent residents without a Swiss citizenship. This did not alter the findings presented here.

2. Choosing the appropriate cluster level is crucial in multilevel analysis, for substantial and statistical reasons. The theoretical framework assumes social interaction, which is less likely in a larger area such as a canton. To account for the cantonal nesting of districts, I report standard errors clustered in the 26 cantons.

Carnegie Corporation of New York Grant

The open access article processing charge (APC) for this article was waived due to a grant awarded to Research & Politics from Carnegie Corporation of New York under its 'Bridging the Gap' initiative.

References

Arzheimer K and Carter E (2006) Political opportunity structures and right-wing extremist party success. European Journal of Political Research 45(3): 419-443. Bentler PM (1990) Comparative fit indexes in structural models.

Psychological Bulletin 107(2): 238-246. Berning CC and Schlueter E (2016) The dynamics of radical right-wing populist party preferences and perceived group threat: A comparative panel analysis of three competing hypotheses in the Netherlands and Germany. Social Science Research 55: 83-93. Betz HG (1994) Radical Right-Wing Populism in Western Europe.

London: Macmillan, Houndmills. Blalock HM (1967) Toward a Theory of Minority-Group

Relations. New York: Wiley. Blau PM (1960) Structural effects. American Sociological Review 25(2): 178-193.

Boomgaarden HG and Vliegenthart R (2007) Explaining the rise of anti-immigrant parties: The role of news media content. Electoral Studies 26(2): 404-417.

Boomsma A (2000) Reporting analyses of covariance structures. Structural Equation Modeling 7(3): 461-483.

Christ O, Asbrock F, Dhont K, et al. (2013) The effects of the intergroup climate on acculturation preferences of immigrants. Journal of Psychology 221(4): 252-257.

Coffe H and Voorpostel M (2010) Young people, parents and radical right voting: The case of the Swiss People's Party. Electoral Studies 29(3): 435-443.

Enders CK and Bandalos DL (2001) The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling 8(3): 430-457.

Givens TE (2004) The radical right gender gap. Comparative Political Studies 37(1): 30-54.

Green EGT and Staerkle C (2013) Migration and multiculturalism. In: Huddy L, Sears DO and Levy J (eds) Oxford Handbook of Political Psychology. Oxford: Oxford University Press, pp.852-889.

Hu L and Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling 6(1): 1-55.

Huckfeldt R and Sprague J (1995) Citizens, Politics and Social Communication: Information and Influence in an Election Campaign. Cambridge: Cambridge University Press.

Huckfeldt R, Johnson P and Sprague J (2005) Individuals, dyads, and networks: Autoregressive patterns of political influence. In: Zuckerman AS (ed) The Social Logic of Politics : Personal Networks as Contexts for Political Behavior. Philadelphia: Temple University Press, pp. 21-48.

Ivarsflaten E (2008) What unites right-wing populists in Western Europe? Re-examining grievance mobilization models in seven successful cases. Comparative Political Studies 41(1): 3-23.

Kriesi H, Lachat R, Selb P, et al. (2005) Der Aufstieg der SVP. Acht Kantone im Vergleich. Zürich: NZZ-Verlag.

Lubbers M (2000) Expert Judgement Survey of Western-European Political Parties 2000. Nijmegen, The Netherlands: Department of Sociology, RU.

Lubbers M, Gijsberts M and Scheepers P (2002) Extreme right-wing voting in Western Europe. European Journal of Political Research 41(3): 345-378.

Lüdtke O, Marsh H, Robitzsch A, et al. (2008) The multilevel latent covariate model: A new, more reliable approach to group-level effects in contextual studies. Psychological Methods 13(3): 203-229.

Marsh M (2002) Electoral context. Electoral Studies 21(2): 207217.

Marsh HW, Lüdtke O, Robitzsch A, et al. 2009. Doubly-latent models of school contextual effects: Integrating multilevel and structural equation approaches to control measurement and sampling error. Multivariate Behavioral Research 44(6): 764-802.

Mudde C (2007) Populist Radical Right Parties in Europe. Cambridge: Cambridge University Press.

Mughan A and Paxton A (2006) Anti-immigrant sentiment, policy preferences and populist party voting in Australia. British Journal of Political Science 36(2): 341-358.

Muthen LK and Muthen BO (1998-2012) Mplus User's Guide. 7th ed. Los Angeles, CA: Muthen & Muthen.

Raudenbush SW and Bryk AS (2002) Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, CA: SAGE Publications.

Rydgren J (2008) Immigration sceptics, xenophobes or racists? Radical right-wing voting in six West European countries. European Journal of Political Research 47(6): 737-765.

Sarrasin O, Green EGT, Fasel N, et al. (2012) Opposition to anti-racism laws across Swiss municipalities: A multilevel analysis. Political Psychology 33(5): 659-681.

Satorra A and Bentler PM (2001) A scaled difference chi-square test statistic for moment structure analysis. Psychometrika 66(4): 507-514.

Scheepers P, Gijsberts M and Coenders M (2002) Ethnic exclu-sionism in European countries: Public opposition to civil rights for legal migrants as a response to perceived ethnic threat. European Sociological Review 18(1): 17-34.

Schlueter M, Meuleman B and Davidov E (2013) Immigrant integration policies and perceived group threat: A multilevel

study of 27 Western and Eastern European countries. Social Science Research 42(3): 670-682.

SELECTS (2011) Befragung der Wählerinnen und Wähler nach den Wahlen [Dataset]. Lausanne, Switzerland: FORS.

Sherif M, Harvey O, White JB, et al. (1961) Intergroup Conflict and Cooperation: The Robbers Cave Experiment. Norman, OK: The University of Oklahoma.

Sprague J (1982) Is there a micro-theory consistent with contextual analysis? In: Ostrom E (ed) Strategies of Political Inquiry. Beverley Hills, CA: SAGE Publications, pp. 99-121.

Stephan W and Renfro L (2002) The role of threats in intergroup relations. In: Mackie D and Smith E (eds) From Prejudice to Intergroup Emotions. New York: Psychology Press, pp.191-207.

Welzel C and Deutsch F (2012) Emancipative values and non-violent protest: The importance of 'ecological' effects. British Journal of Political Science 42(2): 465-479.