Support for the Populist Radical Right: Between Uncertainty Avoidance and Risky Choice

The last several decades have witnessed a structural change in politics toward cultural and identity conflicts, accompanied by the rise of populist radical right (PRR) parties. However, we know surprisingly little about the psychological or cognitive-motivational factors underlying PRR support. We claim that uncertainty avoidance (UA)—an epistemic avoidance motivation—represents a central motive because UA resonates with the PRR platform and precedes common predictors of PRR voting. Using data from the 2017 Austrian National Election Study, we found that UA was indeed indirectly associated with a higher likelihood of PRR voting. This association is because greater UA fostered right-wing sociocultural views, whereas associations with populist attitudes or expected government competence were more ambiguous. PRR parties appear to offer “certainty,” but as extreme parties, they also remain a “risky choice.” We conclude by discussing the contribution of a cognitive-motivational account to explain PRR voting.

Recent years have witnessed the rise of populist parties and like-minded leaders across the globe (e.g., Müller, 2016). In Western Europe, the rise of the populist radical right (PRR), which is characterized by a strict right-wing agenda fused with populist ideas (Mudde, 2007), has become of particular interest. These parties' electoral success is often attributed to broader societal transformations that can be subsumed under globalization and a shift towards more liberal values. In summary, those who perceive these transformations as disadvantageous or threatening make up the core of the PRR's constituency (Kriesi et al., 2006). PRR parties attract these voters by campaigning on a new or transformed sociocultural cleavage dimension that comprises divisions in values (authoritarian/traditionalist versus libertarian/progressive), identity (national versus trans-national), and political issues (such as immigration or European integration; see van der Brug & van Spanje, 2009;Kriesi et al., 2006). Eventually, PRR parties often portray themselves as the only ones representing the ordinary people and their interests against the proponents of globalization and cultural liberalism.
In this line of reasoning, support for PRR parties might be considered a psychological response to the instability, uncertainty, and complexity that arise from these societal transformations (see Hogg, Kruglanski, & van den Bos, 2013). At the same time, PRR parties also represent "risky bs_bs_banner prospects" (Steenbergen & Siczek, 2017, p. 123) because they challenge the status quo and might lack the capability to form a stable and competent government. While support for the PRR has received extensive scientific scrutiny, knowledge is limited about the cognitive-motivational processes underlying PRR support.
In this study, we examine the potentially ambiguous relationship between individual psychological differences in how people deal with uncertainty and voter support for the PRR. We maintain that uncertainty avoidance (UA) represents aspects of personality that pertain to epistemic needs to deal with uncertainty and ambiguity and argue that the PRR and its ideology matches these needs. This conjecture rests upon theories suggesting that people's political belief systems serve basic psychological functions, such as the need to attain epistemic certainty (Jost, Glaser, Kruglanski, & Sulloway, 2003) as well as approach and avoidance motivations (Janoff-Bulman, 2009). We thus contribute to a growing literature on psychological motives underlying political orientations more generally and factors underlying populism and PRR voting in particular.
We based our analysis on a large-scale election study in Austria, home to a prominent PRR party: the Freedom Party of Austria (Freiheitliche Partei Österreichs, FPÖ). Our data provide us with measures on epistemic avoidance motivations and populist beliefs previously unavailable to researchers. Overall, our results revealed that UA is indeed associated with voting for PRR parties. The radical-right ideological core-authoritarianism and nativism-clearly resonated with needs to manage (i.e., reduce) uncertainty. In contrast, the association between UA and populist attitudes as well as party competence turned out to be more ambiguous. We then discuss the implications of our results for research on PRR support and motivated social cognition.

UA and the PRR's Promise
While every individual seeks to reduce states of uncertainty, people differ in the degree to which uncertainty is accepted or avoided. We describe UA as deep-rooted individual differences in people's needs for certainty (Jost et al., 2003). UA is also referred to as (in)tolerance of uncertainty or ambiguity (see Budner, 1962;Frenkel-Brunswik, 1949), the need to manage uncertainty, such as the need for cognitive closure (Webster & Kruglanski, 1994), the need for order or low openness to experience (Jost et al., 2007). UA manifests in preferences and behavioral patterns: a desire for predictability and familiarity (e.g., regarding life events and social situations), preference for structure and repetitive tasks (e.g., in work tasks), and preference for simplicity and decisiveness in contrast to extended rumination (see Webster & Kruglanski, 1994).
So far, there is limited evidence on associations between personality traits, such as UA, and PRR voting. Existing studies on the Big Five traits suggest that, apart from low agreeableness, low openness to experience (Aichholzer & Zandonella, 2016;Bakker, Rooduijn, & Schumacher, 2016) and high conscientiousness (Aichholzer, Danner, & Rammstedt, 2018)-both related to needs for certainty and conformity-predict PRR voting in Europe. Chirumbolo and Leone (2008) further suggest that cognitive closure might be consequential for right-wing party preferences more generally. A study on voting for the United Kingdom Independence Party (UKIP), however, does not find any link with risk aversion (Steenbergen & Siczek, 2017). Overall, research in this field still lacks consistent hypotheses on how personality traits resonate with PRR beliefs and support.
We follow the literature on ideology as motivated social cognition, which suggests that our belief systems serve deep-rooted epistemic, existential, and relational needs (Jost, 2017;Jost et al., 2007). For PRR parties, Mudde (2007) identified three ideological features that might serve such needs: nativism, authoritarianism, and populism (i.e., PRR parties represent a form of "exclusionary populism;" see Betz, 2004). Based on the conjecture of motivated social cognitions and prior evidence, we argue that UA comprises two important and intertwined motivations regarding PRR voting: 1. UA is an epistemic avoidance motivation that represents people's preference for predictability and familiarity, simplicity and structure, and a heightened status quo bias. Furthermore, avoidance and approach are said to be central psychological motivations for adopting certain beliefs and behaviors (Janoff-Bulman, 2009). The core idea is that, as an epistemic avoidance motivation, UA resonates with ideological platforms that offer a way to reduce societal or political uncertainty and complexity. Indeed, a vast body of evidence suggests that right-leaning or conservative ideologies might represent a means of uncertainty reduction (Hibbing, Smith, & Alford, 2014;Jost, 2017;Jost et al., 2003Jost et al., , 2007. 2. UA entails the development of an exclusive identity, higher preference for a closed and homogeneous ingroup, and possible derogation of outgroups. This is an essential conjecture of uncertainty-identity theory, which states that social-category-based self-conceptualization, ingroup preference, and inclusive-exclusive intergroup attitudes are motivated by uncertainty reduction (see Hogg, 2000). This idea is supported by research showing that UA (or need for closure), as well as situational factors that induce the need for certainty (e.g., stress), go hand in hand with heightened group centrism (Kruglanski, Pierro, Mannetti, & De Grada, 2006). In a nutshell, developing an exclusionary social identity also represents a means to deal with uncertainty.
We, therefore, hypothesize: H1: UA entails two motivations that ultimately favor voting for PRR parties: epistemic avoidance and affinity for an exclusive identity.
In the following sections, we detail our theoretical framework depicted in Figure 1 by (1) analyzing how the PRR's ideological core resonates with UA, (2) exploring the more ambiguous relationship between UA and populism, (3) discussing the PRR as a risky prospect, and (4) considering behavioral consequences.

The PPR's Ideological Core
Nativism and authoritarianism represent the two ideological building blocks that constitute the party family's radical-right core. Their authoritarianism manifests in the promise of a return to traditional values, a desire for law and order, and preferences for punitive measures for criminals or deviants. Nativism, a form of exclusionary nationalism, comprises an exclusive and pronounced national identity, the protection of the nation's cultural identity, and policy preferences that express this exclusive and xenophobic nationalism (Mudde, 2007;Rydgren, 2007). As will be argued below, this ideological core serves an epistemic avoidance motivation and the formation of exclusive identities.
Nativist, especially anti-immigrant, sentiments serve as a mental reduction of societal complexity by applying (prejudiced) abstractions of social outgroups (Allport, 1954). Furthermore, as argued by Steenbergen and Siczek (2017), people seeking to avoid uncertainty are more likely to frame immigration as a challenge rather than an opportunity. UA also favors the development of an exclusive identity, which the PRR's national identity (e.g., Hooghe & Marks, 2004) provides. The PRR's agenda does not necessarily include support for leaving the European Union, but their nativism is often expressed in preferences for strong and sovereign nation-states as well as less European integration (Mudde, 2007). Because European integration brings about increasing political and cultural complexity, the PRR's focus on sovereign nation-states should add to its appeal for citizens who long to reduce uncertainty.
The PRR's authoritarian stance blends into the nativist worldview (Mudde, 2007) and appeals to individuals with greater UA in various ways. First, an individual's (right-wing) authoritarianism as an ideological attitude (RWA; Duckitt & Sibley, 2010) or "status quo conservatism" (Stenner, 2009) follows from motivations to seek group security, stability in societal order, and resistance to social change. These motivations are responsible for "intolerance of difference" and the goal to "minimize diversity" (Stenner, 2009) and, hence, favor an exclusionary social identity. Second, authoritarianism can be described as an avoidance-or prevention-based motivation in the interest of protecting one's group from harm (Janoff-Bulman, 2009). Third, facets of cognitive closure that tie in with the concept of UA, namely, the need for simple structure and predictability, relate very clearly to RWA (Cornelis & van Hiel, 2006;van Hiel, Pandelaere, & Duriez, 2004).
Overall, this ideological core represents the sociocultural dimension along which PRR parties mobilize voters (van der Brug & van Spanje, 2009;Kriesi et al., 2006). We, therefore, hypothesize: H2: Individuals scoring high in UA will more readily adopt radical-right core beliefs.

The Role of Populism
Populism is the third building block of the overall ideology, complementing the right-wing core beliefs (i.e., the thick ideology). Following Mudde's (2007) now famous definition, populism is a "thin-centered ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, 'the pure people' versus 'the corrupt elite,' and which argues that politics should be an expression of the volonté générale (general will) of the people" (p. 23). The populist message thus offers clear distinctions and simple explanations of how the political world operates (Mudde, 2004)-often also in simple language (Bischof & Senninger, 2018).
There are good reasons to believe that populism adds to the attractiveness of the PRR for people seeking certainty. First, its simplicity should speak to people high in UA. Furthermore, populism is in large part a story about group centrism: It emphasizes the people as a collective, their homogeneity, and the paramount importance of their views (Mudde, 2004;Müller, 2016). By allowing social identification with and assimilation to a group ("the people"), populism might offer a psychological means for uncertainty reduction.
However, populism is also a challenge to the status quo. By demanding wide-ranging sovereignty for the people, populism challenges established institutions in liberal democracy. Furthermore, disregard for political or societal elites is inherent to populist ideas (Mudde, 2004). This characteristic might dampen populism's appeal to individuals high in UA because they disproportionally prefer the status quo over the uncertainty of change, a form of "status quo bias" (e.g., Samuelson & Zeckhauser, 1988). In summary, the role of populism is ambiguous.
H3a: Due to its simplicity and group centrism, we expect populism to resonate with UA. H3b: Populism's challenge to the status quo, however, could alienate individuals longing for certainty. Now, how are the PRR's ideological components-the radical-right core and the populist beliefs-connected? In the case of the PRR, populist attacks on the status quo often allude to a return to an idealized past, which is centered around nativist and authoritarian ideas and contrasts an "uncontrollable" present (Rydgren, 2017). Therefore, the adoption of populist beliefs might be contingent on the adoption of the PRR's radical-right core ideology. There is no contradiction between UA and the radical right's populist demands for change.
H4: We, therefore, expect that UA is at least indirectly associated with more populist attitudesas far as they are driven by the radical-right core ideology.

The PRR as a Risky Prospect
Voters seeking certainty might be less willing to support parties characterized by uncertainty and change (Kam & Simas, 2012). As argued by Steenbergen and Siczek (2017, p. 123), PRR parties, although appealing to citizens' feelings of uncertainty, represent somewhat "risky prospects" that could repel people high in UA. As mentioned before, these parties' populist stance is an attack on the political status quo and, therefore, might be a source of uncertainty. The specific radical and populist stance of the PRR often makes them succeed in opposition to other parties, while these features might make their long-term success in government questionable (Heinisch, 2003). Therefore, uncertainty might also stem from worries about their ability to deliver stable and competent government because of their perceived inexperience, lack of expertise, and qualified personnel, or radicalismespecially in coalition governments.
Voters usually consider the past or anticipated performance, alongside positional aspects, when forming party preferences (Sanders, Clarke, Stewart, & Whiteley, 2011). This approach, however, is commonly neglected in cognitive-motivational models that focus exclusively on ideological congruence with parties. While it is open to debate whether voters' assessment of party competence is endogenous to their party preference or vice versa, it is important to note that voters might see positional distance and (possible) performance as related but distinct assessments.
H5: We should, thus, see a positive association between ideological closeness and competence assessments. H6: However, we might find a negative direct relationship between UA and the assessment of PRR parties' competence when, at the same time, considering the ideological closeness.
In summary, we would assume that voters high in UA favor PRR parties because their core ideology provides a means to deal with uncertainty. Following our argument that PRR parties also represent somewhat "risky prospects," higher UA, at the same time, might be a hindrance to voting for PRR parties. Thus, voters face a trade-off between uncertainty reduction by PRR parties and the risks associated with them.

Behavioral Consequences: PRR Voting
We argue that UA not only favors the adoption of the PRR ideology but eventually becomes consequential for people's voting behavior. Nonetheless, we expect UA to act only indirectly, mainly through ideological attitudes that are essential predictors for PRR voting (for similar models see, e.g., Aichholzer & Zandonella, 2016;Chirumbolo & Leone, 2010;Duckitt & Sibley, 2010).
H7: In sum, these ideological core beliefs make voting for the PRR more likely and ultimately mediate the effect of UA.
Research on the consequences of individuals' populist beliefs for their voting behavior is still in its infancy. Nevertheless, it has shown that populist ideas, antielitism, and distrust in politicians matter because they represent a common denominator that drives voting for any populist party (van Hauwaert & van Kessel, 2018;Rooduijn, 2018).
H8: Even though populism might play a subordinate role in the PRR's overall ideology, we expect it to be consequential for vote choice and to mediate the effect of UA as well. H9: Apart from these positional aspects, a positive assessment of the PRR's capability in government should enhance the likelihood of voting for the PRR.
Alternative explanations for PRR support, such as generational differences, formal education, a gender gap, or economic pressures, are accounted for by including sociodemographic and socioeconomic aspects as controls in our analytical model (see again Figure 1).

Data and Methods
We test our model with voter data collected in the context of the Austrian parliamentary election campaign in 2017. Austria hosts a very successful and "prototypical" PRR party (Mudde & Rovira Kaltwasser, 2013, p. 155), the Freedom Party of Austria (FPÖ). Starting in the mid-1980s, the FPÖ has built a strong populist stance (Luther, 2007). It now relies increasingly on nativist anti-immigrant sentiments and welfare chauvinism (Aichholzer, Kritzinger, Wagner, & Zeglovits, 2014;Ennser-Jedenastik, 2016).

Sample
The survey was conducted by the Austrian National Election Study and is available online (Wagner et al., 2018). Respondents were recruited using a commercial online access panel. The sample was a priori quoted to match known population distributions from census data in 2016. Hence, the sample was heterogeneous regarding age (Min = 16, Max = 85, Mean = 42, SD = 15), gender (49% male), and formal education.

Measures
The survey was designed as a panel study. Sociodemographic variables were collected in the first wave of the panel or upon panel (re-)entry for fresh respondents. Vote choice for the FPÖ was measured shortly after the election on October 15, 2017 (fifth wave); all other variables stem from the third and fourth wave. Supplemental analysis also made use of variables from Waves 6 and 7 (see Appendix A for the questionnaire items on UA; Appendices S1 and S2 in the online supporting information provide further details on all other measures and the panel waves).

Uncertainty Avoidance
To measure UA, we applied a rating-scale format that avoids the commonly used agree/disagree Likert-type format, because it is known to be prone to response biases. Portions of items were selected from established German ambiguity intolerance and cognitive closure scales (Dalbert, 1999;Schlink & Walther, 2007). The items describe very general behavioral tendencies and preferences unrelated to political beliefs. Each item uses directly opposed statements (e.g., "I can deal very well with unknown and unpredictable situations" vs. "I prefer situations that I am familiar with, that I know well"). 1 Overall, respondents were presented with six sets of oppositely worded self-descriptions and had to place themselves between them using a 6-point rating scale (see Appendix A).

Radical-Right Ideological Core
We estimated radical-right core attitudes as a second-order factor, which summarizes authoritarianism, anti-immigrant attitudes, rejection of European integration, and exclusive nationalism as first-order factors in a structural equation model.

Authoritarianism
Five Likert items were selected from a short German scale that taps into social attitudes described as right-wing authoritarianism (RWA; Aichholzer & Zeglovits, 2015), such as "Our society has to crack down harder on criminals for once."

Anti-Immigration Attitudes
We use six commonly used Likert items to measure anti-immigration attitudes (e.g., "Immigrants should adapt to Austrian customs and traditions.").

Anti-EU Attitudes
We make use of a standard survey question, asking whether unification should be pushed further (0) or has already gone too far (10) (Mean = 6, SD = 3.2).

Exclusive Nationalism
We measure exclusive nationalism by taking the simple difference of the strength (4-point scale) of attachment with the nation and Europe (i.e., exclusive nationalism = attachment nation − attachment Europe ). High scores, thus, represent exclusive national attachment (Mean = .6, SD = 1).

Populist Attitudes
Our populist attitude measure aims at capturing three interrelated subdimensions of populist attitudes: anti-elitism, belief in the homogeneity and virtuousness of the people, and a preference for wide-ranging popular sovereignty (Akkerman, Mudde, & Zaslove, 2014;Schulz et al., 2018). A set of six Likert items (adapted from Hobolt, Anduiza, Carkoglu, Lutz, & Sauger, 2016) was selected (e.g., "The people should make important political decisions, not politicians."). All items could be answered on a fully labeled 5-point rating scale (1 = agree completely, 5 = disagree completely).

Government Competence
People's expectations regarding the parties' performance in government was captured on an 11point scale ranging from 0 (would do a very bad job) to 10 (would do a very good job in government) (Mean = 4.8, SD = 3.4).

Support for a PRR Party
As our dependent variable, we use actual vote choice for the Freedom Party (1 = FPÖ, 0 = other; including all respondents who cast a valid vote). Of the respondents in the final sample, 32% reported a vote for the FPÖ. Second, we investigated the propensity to vote (PTV) for the FPÖ before the election, which is measured on an 11-point scale ranging from 0 (highly unlikely) to 10 (very likely) (Mean = 4.1, SD = 4).

Controls
In addition, we included the following sociodemographic variables as controls: age (in years), gender (1 = female), formal education (1 = at least upper secondary school-leaving certificate or Matura in Austria), and subjective income situation (1 = get along [very] well, 0 = get along with [great] difficulty).

Analysis
We analyzed our data using structural equation modeling, which, on the one hand, allows taking measurement errors that disturb the associations between the constructs into account and, on the other hand, allows studying the effect of UA on and mediation through ideological variables. Identifying mediational patterns is generally tricky; it relies upon the assumption of "sequential ignorability" or, in other words, no problems due to unmeasured confounding (Imai, Keele, Tingley, & Yamamoto, 2011). We, therefore, first, included variables known to drive populist support as well as sociodemographic controls in our models to make these assumptions more plausible. Second, sensitivity analyses proposed by Imai et al. (2011) were conducted, which lend support to the robustness of our main findings towards violations of these assumptions (see Appendix S4 in the online supporting information).
For our analysis, we used the software R (version 3.6) with the lavaan package (version 0.6-3) and calculated diagonally weighted least squares (DWLS) estimation for vote choice as the dependent variable and maximum likelihood (ML) estimation for PTV. Both models were computed with robust standard errors and mean-and-variance adjusted test statistics (Rosseel, 2012).

Results
We examined the impact of differences in UA on voting for a PRR party and how this effect is mediated by ideological core beliefs and populist attitudes as well as performance considerations. Figure 2 illustrates the structural equation model we used to test our hypotheses (for more detailed results, see Appendix S3 in the online supporting information). According to the CFI and RMSEA fit statistics, our specifications indicated a good fit to the data.
Our results suggested a significant overall effect of UA on PRR voting (β total = .11; see Table 1), thus corroborating Hypothesis 1. This total effect is distinct from the direct effect (β direct = −.06, n.s.) because it additionally includes all indirect effects (β indirect = .18; see Table 1). Regarding specific indirect effects, the association between voting for PRR parties and UA ran mainly through radical-right core attitudes rather than populist attitudes (see Table 1).
The substantial indirect effect through radical-right core attitudes resulted from both our Hypotheses 2 and 7 being confirmed by our model. The radical-right core-consisting of authoritarianism and nativism-was associated with the need to deal with uncertainty (β = .29). Furthermore, such attitudes were also highly consequential for PRR voting (β = .35). is not depicted. Entries are standardized regression coefficients. R 2 represents proportions of the outcome variables' variation explained by the model. The arrows' thickness in the structural model visually represents the effect size. 95% confidence intervals in square brackets. Two-tailed significance levels: *p < .05, **p < .01, ***p < .001.
As projected, voters' expectations about the FPÖ's performance in government were a strong predictor for people's vote choice (β = .51), thus confirming Hypothesis 9. In part, these performance expectations also mediated the effect of UA on voting. However, this happened mainly through the affinity for ideological core beliefs (β indirect = .11), which were associated with a better assessment of the FPÖ's prospects in government (β = .74; supporting Hypothesis 6). In turn, the direct effect of UA on the evaluation of government competence was negative (β = −.08; supporting Hypothesis 6). Controlling for ideological closeness, UA increased a person's doubt about the PRR party's ability to deliver a good and stable government. Based on our model and consistent with Hypothesis 8, populist attitudes were somewhat consequential for PRR voting in Austria (β = .12). However, the effect is small compared to people's rightwing beliefs and anticipated government performance. In line with our expectations, the way populist attitudes reflected individuals' needs to deal with uncertainty is more ambiguous (see Hypothesis 3a and 3b). Overall, our model yielded only a small positive but not statistically significant effect of UA on the adoption of populist attitudes (β total = .06). However, looking at the direct effect estimates, higher UA lowers the appeal of populist beliefs (β direct = −.12). Instead, a positive association between UA and populist attitudes was contingent on the connection with right-wing beliefs (β indirect = .18; confirming Hypothesis 4). 2 Note. Coefficients (β) are standardized regression coefficients from DWLS estimation (vote choice) or ML estimation (PTV). PRR = populist radical right, PTV = propensity to vote, UA = uncertainty avoidance, populism = populist attitudes, RR core = radical right ideological core, gov. compet. = (expected) competence of PRR party in government.
Since a binary measure of vote choice might blur certain associations, we reran our model with a larger sample predicting the propensity to vote (PTV) for the FPÖ instead (see Figure 3 and Table 1). Overall, the PTV model corroborates our findings for vote choice, although subtle differences exist. The model lends no support to Hypothesis 8; populist attitudes are not significantly associated with PTV (β = .02). Compared to actual vote choice, ideological core beliefs appeared less consequential, whereas performance expectations were more influential for PTV. Furthermore, the variables we included fully mediated the effect of UA on PTV (β direct = −.02, n.s.), for vote choice a small direct effect remained (β direct = −.06, p = .06).

Supplemental Analysis
Using an additional panel wave more than 16 months after the election (Wave 7, conducted February/March 2019) allows us to, on the one hand, scrutinize how stable the effect of UA on PRR support is over time and, on the other hand, compare the impact of UA to respondents' risk-taking propensity (see, e.g., Steenbergen & Siczek, 2017). The measure for UA is the same as before (Wave 3), whereas the measure for risk-taking propensity was included in the sixth panel wave (December 2017); all other indicators and the dependent variables stem from Wave 7. Detailed results from this supplemental analysis are presented in Appendix B.
The supplemental analysis proved that the association between UA and PRR support remained relatively stable over time. Even with data collected about 18 months later, those high in UA in Wave 3 were, on average, more likely to express support for the PRR (β total = .13 for prospective vote R 2 represents proportions of the outcome variables' variation explained by the model. The arrows' thickness in the structural model visually represents the effect size. 95% confidence intervals in square brackets. Two-tailed significance levels: *p < .05, **p < .01, ***p < .001. choice; β total = .10 for PTV). At this point, the FPÖ had been in government for more than a year. Two additional observations can be made from this fact: No significant association between UA and the assessment of the actual government competence existed when controlling for right-wing beliefs, as compared to how well the party might do in government. Also, populist (or antielite) attitudes were less intertwined with right-wing ideological beliefs (i.e., the proposed indirect pathway for UA was less important), whereas the direct association between UA and populist attitudes was still negative (see the PTV model).
Finally, the inclusion of risk-taking propensity in the model did not impair UA's explanatory power. Although risk-taking propensity correlated strongly with UA (r = −.51 in vote choice model; r = −.48 in PTV model), it did not significantly affect any of the outcomes.

General Discussion
This study was interested in individual differences regarding the need to deal with uncertainty in life, from which differences in the support for PRR parties may arise. That is not to say that every person is born as a right-wing populist or not, but certain predispositions, such as UA, make it more likely for an individual to lean toward a specific ideology. In particular, we hypothesized that these voters face a trade-off between reducing uncertainty through the PRR's ideological stance and the potential political risk that comes with these parties.
Our results indeed suggested that scoring higher on UA makes it more likely for the individual to vote for a PRR party, although our measure captured very general epistemic motivations unrelated to politics. We could also show empirically that the more fundamental concept UA goes beyond a mere willingness to take risks (cf. Steenbergen & Siczek, 2017). Our findings thus add to the literature on psychological antecedents of PRR voting, which thus far has focused mainly on the Big Five personality traits (see Aichholzer et al., 2018;Aichholzer & Zandonella, 2016;Bakker et al., 2016). Our results further suggested that people who tend to avoid uncertainty were, on average, significantly more inclined to embrace radical-right beliefs. The PRR's ideological core provides a clear match (elective affinity; see Jost, 2017) with the motives of this personality trait. Consequently, people high in UA were also (indirectly) more likely to support the PRR party FPÖ. Overall, these findings fit well with the literature on ideology as motivated social cognition, which suggests that right-wing ideology fulfills deep-rooted epistemic and existential needs (Jost, 2017;Jost et al., 2007).
It is quite surprising to see that, considering the commotion around populism in recent years, populist attitudes generally had only minimal impact on voting for the Austrian Freedom Party in 2017 and no direct association with its PTV (contrary to van Hauwaert & van Kessel, 2018;but cf. Stanley, 2011). While the decision to vote for a certain PRR party might be affected by a variety of context-specific factors such as campaign dynamics, this finding also reminds us that PRR parties such as the FPÖ should be considered foremost radical-right parties that are also populist-and not the other way around (Mudde, 2007, p. 26).
Individuals high in UA were not significantly more likely to exhibit populist attitudes, at least when the total effect is considered. Nonetheless, this association was more ambiguous. The reason for this ambiguity is that populism not only simplifies politics and provides a group identification via the "people," it also challenges the status quo and established authorities. People longing for certainty might be deterred from populist or radical ideologies that make a "demand for major transformation of the society, either toward some future vision or back to an idealized past" (Powell, 1986, p. 358). This concern could explain the finding that UA decreases populist attitudes when taking peoples' proximity to the radical-right core ideology into account. Hence, the association between UA and populist attitudes appears to be strongly contingent on the adoption of the radical-right host ideology. Seemingly, only the radical right's populist demand for change, which is presented as a restoration of order and return to traditional values, might provide means to deal with uncertainty, whereas radicalism or populism per se do not.
Nevertheless, there are some limitations to this study. Using data from Austria allowed us to study a prototype case for PRR parties, whereas the Austrian party system lacks actual left-wing populist parties. Therefore, it does not come as a surprise to find a powerful connection between populism and genuine rightist attitudes. It might be worthwhile to compare our results to other countries where being ideologically left-wing is generally associated with more populist attitudes, such as Spain (Rico, Guinjoan, & Anduiza, 2017) or Greece (Tsatsanis, Andreadis, & Teperoglou, 2018). Nevertheless, we believe that associations between UA and voting for a populist party are less plausible in party systems where populism connotes left-wing beliefs.
Given our results, we consider it worthwhile to further investigate the ambiguous prospects of PRR parties. On the one hand, they pretend to offer certainty in a complex and globalized world and, on the other hand, they remain a risky choice. Considering our findings on the role of anticipated government competence assessment and its negative direct association with UA, PRR parties' ability to reduce worries regarding their performance if elected into government might, in part, explain differences in their success.
Regarding the role of populism, future work should consider that it might attract voters looking for certainty as a communication style or framing strategy rather than as an ideology. According to extant research, UA is generally connected with individual differences in the preference for "simple" over "complex" things, such as text or art (see, e.g., Hibbing et al., 2014). Populist communication might, hence, be attractive to people high in UA through its use of easy (i.e., less complex) language (Bischof & Senninger, 2018) and frames. These issues thus deserve to be studied further.
Research on the endorsement of PRR parties and its association with personality traits has been very scarce so far. By showing that individuals with higher needs for UA tend to gravitate toward PRR parties rather than to other parties, we provide a novel contribution to understanding the rise of PRR populist parties and like-minded leaders. We can only speculate that adopting the PRR's ideological agenda in some way offers a way to deal with developments such as loss of national sovereignty or increasing cultural variety in a globalized world.

Supporting Information
Additional supporting information may be found in the online version of this article at the publisher's web site: Appendix S1. Further Measurements  Table S2. Description of Panel Waves and Collected Variables  Analysis  Table S4. Results from Sensitivity Analysis on Some of the Mediation Effects Figure S4.1. Sensitivity analysis of the ideological core's effect on the PTV for the FPÖ through government competence. Figure S4.2. Sensitivity analysis of the ideological core's effect on the PTV for the FPÖ through populist attitudes. Figure S4.3. Sensitivity analysis of uncertainty avoidance's effect on government competence through ideological core beliefs (PTV model). Figure S4.4. Sensitivity analysis of uncertainty avoidance's effect on populist attitudes through ideological core beliefs (PTV model). Figure S4.5. Sensitivity analysis of uncertainty avoidance's effect on the PTV for the FPÖ through ideological core beliefs. Figure S4.6. Sensitivity analysis of uncertainty avoidance's effect on the PTV for the FPÖ through government competence. Figure S4.7. Sensitivity analysis of uncertainty avoidance's effect on the PTV for the FPÖ through populist attitudes. Figure S4.8. Sensitivity analysis of the ideological core's effect on voting for the FPÖ through government competence. Figure S4.9. Sensitivity analysis of the ideological core's effect on voting for the FPÖ through populist attitudes. Figure S4.10. Sensitivity analysis of uncertainty avoidance's effect on government competence through ideological core beliefs (vote model). Figure S4.11. Sensitivity analysis of uncertainty avoidance's effect on populist attitudes through ideological core beliefs (vote model). Figure S4.12. Sensitivity analysis of uncertainty avoidance's effect on voting for the FPÖ through ideological core beliefs. Figure S4.13. Sensitivity analysis of uncertainty avoidance's effect on voting for the FPÖ through government competence. Figure S4.14. Sensitivity analysis of uncertainty avoidance's effect on voting for the FPÖ through populist attitudes.

Appendix B: Supplemental Analysis
Supplemental analyses make use of Wave 7 in the panel, which was conducted more than 16 months after the election. An important ingredient of the ideological core, the RWA measures were, however, only included in Wave 3 of the survey. So we had to rely on a single item on law and order, which was included in Wave 7. The reason is that UA as well as risk-taking propensity (measured in Wave 6) are considered to precede ideological attitudes. Yet, robustness checks with the Wave 3 measure of RWA did not differ substantially from the presented models. Figures B1 displays results on prospective vote choice in Wave 7; Figure B2 displays the same model with PTV. Appendix S5 in the online supporting information provides detailed results as well as direct and indirect effect sizes. is not depicted. Entries are standardized regression coefficients. R 2 represents proportions of the outcome variables' variation explained by the model. The arrows' thickness in the structural model visually represents the effect size. Ninety-five percent confidence intervals in square brackets. Two-tailed significance levels: *p < .05, **p < .01, ***p < .001. Figure B2. Structural equation model for the propensity to vote for a PRR party in Wave 7 (supplemental analysis). PRR Government Competence and Populism are allowed to covary, the correlation (−.43*** [−.49, −.37]) is not depicted. Entries are standardized regression coefficients. R 2 represents proportions of the outcome variables' variation explained by the model. The arrows' thickness in the structural model visually represents the effect size. Ninety-five percent confidence intervals in square brackets. Two-tailed significance levels: *p < .05, **p < .01, ***p < .001.