Value orientation and external political efficacy: assessing the relationship between traditional values, progressive politics and political responsiveness

Values are often used to explain phenomena associated with problematic political behaviours. For example, the election of far-right parties is often attributed to traditional values. The ‘Cultural Backlash’ thesis (Inglehart and Norris in Cultural backlash: trump, Brexit, and authoritarian populism. Cambridge University Press, Cambridge 2017a) is prominent in this field of research and explains such behaviour by a predominance of progressive issues as the result of Value Change. It is assumed that this causes traditional individuals to perceive the political system as less responsive, which can also be understood as low external political efficacy; however, there is little in the way of empirical research that tests the connection between one’s value orientation and one’s sense of external efficacy (Inglehart and Norris in Trump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash, 2016; Inglehart and Norris in Cultural backlash: trump, Brexit, and authoritarian populism, Cambridge University Press, Cambridge, 2017a; Inglehart and Norris in Perspect Polit 15(2):443-453, 2017b). This study aims to fill this gap by examining the effects of traditional values on perceptions of external efficacy, particular attention to the role of predominantly progressive politics. For this purpose, the relationship between values and external efficacy is brought into cross-national context in order to analyse the moderating effect of the degree of progressivity of the respective national political system. To test these assumption with multilevel mixed-effects model, this study uses European Social Survey data from 23 European countries and from the Manifesto Project. The results show that traditional values are negative associated with external political efficacy. They also show that this effect is partially intensified the more progressive the party manifestos are.


Introduction
Since the 1990s, value researchers have investigated political behaviours that directly challenge democracy, seeing it as a counter-reaction to the 'classic Value Change' from materialism to postmaterialism (e.g. Ignazi 1992;Inglehart and Baker 2000). The emergence of postmaterialist values changed social and political value systems and led to the rise of a new politics that emphasise non-material issues. This has potentially left conservative individuals feeling politically unrepresented. Recent elections of populist actors and far-right parties in Western democracies have revived this discussion (Inglehart and Norris 2017a, b;Schäfer 2021). In their prominent monograph 'Cultural Backlash', Inglehart and Norris (2017a) argue that Value Change led to predominantly progressive society and politics. As a result, traditionally minded individuals no longer feel their issues adequately represented, which can lead to e.g. the election of right-wing populist parties. Although this helps in explaining phenomena such as Trump's election or Brexit with traditional values, the link between those values and perception of being political irrelevant is rarely explored empirically. It is also still unclear whether the assumed predominance of progressive values in politics influences this relationship. Therefore, this study aims to fill these research gaps by investigating what effect traditional values have on feelings of political irrelevance and how the context of progressive political system in form of country's politics influences this effect. This research gap is relevant because understanding the connection between traditional values, progressive predominated societies and perceptions of political non-representation may help us better understand the roots of problematic political behaviour in European countries. We add to the literature by examining the association between traditional values and the perception of being political irrelevant in a cross-national comparison. This allows us to investigate the influence of different degrees of 'political progressiveness' on this relationship and to explore whether more progressive political contexts could lead to lower efficacy of traditional minded individuals. To represent the feeling of political irrelevance both theoretically and empirically, metrics for the concept of external political efficacy are our starting point. Defined as the perception of responsiveness of government and political actors to one's needs and preferences (Lane 1959;Craig et al. 1990), 'external political efficacy' is used as an indicator of this pessimistic attitude towards politics.
This study aims to answer two central questions: (1) What influence do traditional value orientations have on external political efficacy, and (2) how is this relationship affected by country's politics? Our main assumption is that politics in European democracies are becoming increasingly socially progressive and less economically oriented over the course of Value Changes (Inglehart and Norris 2017a: 45ff). Consequently, individuals with traditional values perceive political systems as less responsive, and thus, they report lower external political efficacy. Further, we theorise as to why the assumption of predominantly progressive politics is plausible and attempt to clarify its significance in the context of cleavages between progressive politics and traditional values.
To examine those relationships, this paper uses the 9th Round of European Social Survey data from 23 EU member states (ESS, 2018a), including the UK, and data from the Manifesto Project (Volkens et al. 2021a) as indicators for each country political party's ideological and economic orientations. Using a multilevel mixed-effects model, we first affirmed that respondents who reported traditional and security-oriented values were significantly and negatively associated with external efficacy. Furthermore, it can be shown that the more progressive and the less economically oriented the party politics on average, the stronger the negative effect of traditional values.

The importance of external political efficacy for democracy
The concept of political efficacy was first introduced by Campbell et al. (1954: 187) as the 'feeling that individual action does have, or can have, an impact upon the political process'. According to Lane (1959), there is a distinction between two dimensions of political efficacy: (1) internal political efficacy, understood as the self-attribution of the ability to understand politics, and (2) external political efficacy, defined as the perception of responsiveness of government and political actors to one's needs and preferences (see also Craig et al. 1990;Niemi et al. 1991). In this analysis, political efficacy is used to measure a respondent's perceptions of their political system's responsiveness.
A large body of literature not only shows that both dimensions can be seen as indicators of the health of a democracy (Craig et al. 1990;Karv et al. 2021), but also identifies it as a strong predictor of democratically desirable political behaviour, such as voting turnout (Gimpel and Lay 2005) and political participation (Scheufele et al 2006), as well as social and political trust (Niemi et al 1991;Wolak 2018) and satisfaction with democracy (DeHoog et al 1990). Referencing the social environment or political system as a potential influencing factor at the contextual level, additional studies show that an individual's social environment can influence their external efficacy (e.g. Karv et al 2021). Other findings link political efficacy to mechanisms of political alienation, such as Stoker (2006) and Hay (2007). This relationship is also supported by findings from Freie (1997), who report that external efficacy decreases when people feel politically ignored.

Low external efficacy as an outcome of value change?
Values are the general and consistent beliefs structuring the preferences of individuals, and they provide the foundations for worldviews and belief systems (Schwartz et al. 2014). Individual values thus form the fundamental basis for attitudes and the resulting behaviour in certain situations (Schwartz, 1997). This also (or perhaps especially) applies to political action. Several studies show that respondent values influence both political attitudes (e.g. Piurko et al. 2011) and political behaviour (e.g. Barnea and Schwartz 1998;Schwartz et al 2010).
A phenomenon that can be understood as EPE is a central component of the Cultural Backlash thesis (Inglehart and Norris 2017a), which is a continuation of Inglehart's (1977) classic theory of Value Change. According to this approach, the concept of political alienation and lack of responsiveness of the political system is of key importance in explaining phenomena such as the rise of populism or voting in favour of right-wing parties (Inglehart and Norris 2017a, b). Central to this is the cleavage between conservative (materialistic) and progressive (postmaterialistic) values, based on the classical theory of Value Change (Inglehart 1971(Inglehart , 1977(Inglehart , 1981Inglehart and Abramson 1999). According to this framework, older cohorts, materialistically influenced by the inter-and after-war period, were followed by younger generations that were more postmaterialistically oriented, due to the absence of economic threats. On the societal level, this led to a shift from conservative guiding values to a liberalisation of societal values, resulting in greater social diversity and the 'emergence' of new ways of life (Inglehart and Norris 2016).
The Cultural Backlash is understood as a counter-reaction by conservative individuals to the above-mentioned liberalisation processes and the resulting dominance of progressive values. The exact mechanism that leads to backlashing political behaviour is theoretically defined only vaguely by Inglehart and Norris and analysed empirically only rarely (2016, 2017a, b). In their analyses, the authors skip this step in their theoretically postulated causal chain and explain phenomena such as susceptibility to populism or the election of right-wing populist parties through individual values and associated socio-demographic characteristics. Equally unconsidered is the context of a society dominated as predominantly progressive, which is only outlined by a higher proportion of postmaterialists within the population. To close that gap, the role of political participation in the context of Value Change provides a valid entry point to link their Cultural Backlash hypothesis argumentatively with EPE. In this paper, we assume the process of marginalisation as described by Inglehart and Norris can be understood as lack of responsiveness of government and thus measured as a low EPE score.
Furthermore, intergenerational value shift rises the hypothesis that political participation is asymptotically distributed between materialists and postmaterialists (e.g. Inglehart and Abramson 1999). This follows from the definition of postmaterialists as individuals who, due to the absence of economic concerns, have more capacity for political activity and thus also for political interest. This results in a higher political participation of postmaterialistic oriented individuals and thus also a higher voter turnout (Dalton 2006;Norris 2002). If materialists are defined as a social group with little political interest and low political participation (Inglehart 1981), this effect is further reinforced by the fact that the reactivity of politicians and parties in democracies to groups that do not vote is relatively low (Crepaz 1990). This ensures that liberal issues find their way into the political process to a greater extent. These predominance of political participation by progressive individuals results in the presumption of strong effects on the prioritisation of political topics: the political elite in form of politicians and parties influences, which thematic priorities are set in public perception in connection with political decision-making (Wright 2015). Furthermore, the recipients of these thematic focuses are mostly people who are characterised by more liberal and progressive attitudes. This gives rise to the thesis that this interaction leads to a dominance of topics with progressive connotations on the political agenda of a modern society (Inglehart and Norris 2017a, b). That assumption is further supported by the issues consistency hypothesis, according to which materialist exhibited larger agenda-setting effects for materialistic issues than for postmaterialistic issues and vice versa (Valenzuela 2011).
As we have seen, the theoretical framework of the Cultural Backlash thesis is closely linked to individual socio-economic status. Although the relationship between traditional values and perceived political non-representation in the sense of Cultural Backlash has been little studied empirically, there is a growing body of literature that associates external political efficacy or political discontent with the election of right-wing populist parties. For example, Geurkink et al. (2020) show that EPE and populist attitudes are based on different underlying attitudes, but both are related to voting for populist parties. Krause and Wagner (2021) also show that EPE is a factor related to populist vote and find that the influence of EPE varies depending on the degree of establishment of populist parties in the political system. Spruyt et al. (2016), Magni (2017) and Roodujin et al. (2016, among many others) report similar findings regarding the relationship between EPE and populist vote. Parts of this literature also have a strong focus on socio-demographic characteristics (e.g. Kurer 2020) and seem to have large overlaps with this part of the value literature. To advance the literature on EPE, this study therefore leaves besides socioeconomic aspects such as age/cohorts and social class and focuses on the relationship between individual value orientation and the 'progressiveness' of the respective national political level.

The context of progressive societies and who forms them
The core purpose of this study is to examine the above-mentioned influence of a society that is assumed to be predominantly progressive on the relationship between values and EPE. We have learned that under the conditions of a predominantly progressive society and politics, individuals with traditional values may tend to have lower estimates of EPE. The next step is to understand these progressive societies and who forms them.
Assuming a society in the sense of the Cultural Backlash thesis, the core characteristic of such a society is the dominance of culturally and ecologically oriented issues. Traditional economically based issues lose relevance. However, it must be assumed that the population is not entirely progressive. It can also be assumed that progressiveness in political and cultural matters is due to social actors who are at least politically and culturally predominantly progressive. Taking a step back and looking at the actors who influence public discourse raises the question of the owners of political power: the elites. This necessity arises from the definition of elites as the holders of power resources as well as possibilities for influencing a society and its politics (Hoffmann-Lange 2018: 54). This raises the question whether there is a connection between members of elites in modern societies and individual postmaterialist attitudes. According to, for example, Helbling and Teney (2015) members of elites are characterised by, inter alia, a high level of education and a secure economic and social status. Combining this circumstance with the fundamental thesis of Value Change, it can be assumed that these socio-economic characteristics result in a progressive postmaterialistic attitude (Inglehart and Norris 2016). It is assumed that a high level of education and a higher professional position on the one hand favour progressive value orientations. On the other hand, this combination could also mean a higher level of influence on society. Therefore, in the following the combination of high education and high professional position is referred to as high socio-economic status (SES).
The contextual level is also conceptually linked to the elite idea. Since the previous argumentation outlined the political system as the context which is less responsive to traditionally minded people, it should then be assumed to be influenced by the political elite. This, in turn, is conceptualised by political parties, which significantly shape the political reality within a society. It can be concluded that decision-makers in political parties tend to be more progressive than average due to their high SES, which is reflected in their manifestos and leads to a higher proportion of progressive than conservative content in the political system. In addition to this progressive preponderance, it is also assumed that economic issues have moved into the background and are less relevant, if the argument of a materialistically shaped political environment applies. Moreover, it is also possible that political elites, irrespective of their individual values, pursue more progressive policies for strategic reasons in order to increase their vote share and thus their chances of winning elections in predominantly progressive countries.
Thus, the argumentation so far can be summarised with the following thesis: In the course of Value Change, Western societies have undergone a process that has led to an increase in progressive individual values through intergenerational Value Change. Since such progressive values are associated with more influential social positions and high political participation, it can be assumed that this has led to a predominance of progressive and liberal politics. This should be reflected above all in party manifestos. Due to this predominance, it can be assumed that conservative materialistic individuals feel politically excluded and therefore show lower EPE score.
Understood as an intergenerational process, Value Change as the emergence of predominantly progressive societies and politics has a temporal dimension. This study understands this predominance as a result of this process, whichfollowing the Cultural Backlash-describes a status quo of the circumstances in western democracies, which are related to individual values. In the view of this paper, the relationship between these progressive politics, individual values and EPE does not imply a temporal component. This implies the advantage that it can be analysed by cross-sectional data and does not require any longitudinal data perspective.
Finally, it remains to be clarified how traditional and progressive values are measured. Operationalising these values into survey questions is understandably the crucial question in almost every empirical analysis of values research. The most common concepts are Schartz's Basic Human Values and Inglehart's Postmaterialism Index (e.g. Marcos-Marne 2021). The latter is criticised for not making a clear distinction between attitudes (Datler et al 2013), which is what EPE is understood as here, and values. Therefore, the Basic Human Values (Schwartz 1992(Schwartz , 2012 are also used in this study. To maintain the links to the theoretical and empirical conceptualization of values according to the theory of Inglehart in this analysis, Wilson (2005) provides a valid way of integrating the concepts of materialism and postmaterialism as value sets into the conceptualization of Schwartz's Basic Human Values in an empirical study, on which we rely. Accordingly, this study defines Universalism and Self-Direction as values associated with postmaterialist attitudes, Security and Tradition as associated with materialism. Even though Tradition is not close to the 'Materialism Axis' in Wilson's findings, it can be assumed on the basis of the theoretical framework, and not least on the basis of the frequent use of the word 'traditional' up to this point, that it is relevant in the context of the marginalisation of people with traditional values.
Our theoretical assumptions can be summarised as follows: (1) Value Change led to progressive issues being over-represented in the political system, and (2) postmaterialists participate more in politics and have more power to influence society and politics. This is reflected in more progressive politics policy, while economic issues are losing relevance. Due to this incongruence between their own conservative values and the progressive predominance in politics, materialists perceive the political system as less responsive than postmaterialists.
Hypothesis 1 (H1) Measures associated with traditional and security-oriented values will decrease a respondent's reported external efficacy. This paper's thesis is that the more progressive a country's political issues are, the more the negative effect of traditional values on EPE should increase. Following H1, we predict the more progressive a political system is, and the less-focused it is on economic aspects, will be negatively associated with materialist scores on responsiveness.
Hypothesis 2a (H2a) More progressively predominated politics will strengthen the effects on external efficacy assumed in H1.

Hypothesis 2b (H2b)
Politics with stronger focus on cultural than economic issues will strengthen the effects on external efficacy assumed in H1.

Data and methods
This study uses individual level data from the European Social Survey Round 9 collected mainly between 1 September and 31 December 2018 across 30 European countries (ESS 2018a). The selection of countries includes all member states of the European Union, plus the UK. We selected the countries that are EU member states and for which party manifesto data from the Manifesto Project are available (see below). The sample used for the analyses includes altogether 23 countries: Austria, Belgium, Bulgaria, Cypress, Czech, Germany, Estonia, Spain, Finland, France, the UK, Croatia, Hungary, Ireland, Italy, Lithuania, Latvia, the Netherlands, Poland, Portugal, Sweden, Slovenia and Slovakia. To test the hypotheses of this paper, this study uses multilevel mixed-effects model. This approach has three advantages for the analysis of the data used here: it takes into account the multilevel structure of the data due to the context level of the 23 countries. 1 Furthermore, effects at the individual level can be tested (H1) and cross-level interactions between variables at the individual and context level can be modelled (H2a and H2b) 2 (Aguinis et al. 2013).

Dependent variable
To measure external political efficacy, the study used items from the short form of The Perceived Political Self-Efficacy Scale (PPSE-S). Respondents were asked two questions to assess their self-perceived EPE: 'How much would you say the political system in [country] allows people like you to have a say in what the government does?" and 'And how much would you say that the political system in [country] allows people like you to have an influence on politics?", both on a five-point likert scale, with the response options 'Not at all', 'Very little', 'Some', 'A lot' and 'A great deal'. Both variables were aggregated into one variable by mean (inter-item correlation r = 0.62; Cronbach's α = 0.77; Spearman-Brown coefficient SBC = 0.77).

Independent variables at individual level
To capture the Schwartz Basic Human Values, respondents were asked how similar a person described is to them. For each of the values of interest here, two or three descriptions of people to whom different things are important were asked: 'Important that people are treated equally and have equal opportunities', 'Important to understand different people' and 'Important to care for nature and environment' for Universalism (r = 0.35/0.29; α = 0.59; SBC = 0.54) and 'Important to think new ideas and being creative' and 'Important to make own decisions and be free' for Self-Direction (r = 0.28; α = 0.44; SBC = 0.44). Tradition was measured by 'Important to be modest and humble, not to stand out' and 'Important to follow traditions and customs' (r = 0.18; α = 0.30; SBC = 0.30) and Security by 'Important to live in a safe environment' and 'Important that the government is strong and provides security' (r = 0.42; α = 0.59; SBC = 0.59). 3 Following ESS recommendations, the 1 Another advantage of this method would be that it is relatively robust to the number of units at the context level. With 23 countries, this analysis lies above the threshold of 10 to 20 units needed to obtain correct estimates (Rabe-Hesketh and Skrondal 2008). 2 This method also allows the analysis of direct effects of the context variables on the dependent variable at the individual level. These are as well calculated and displayed in the results, but have no relevance to the hypotheses of this study. 3 Even though the reliability measurements of the value variables are very low, their reliability is nevertheless assumed here. On the one hand, the scales are based on two respective three variables, on the variables were aggregated by mean. The overall mean of all value variables was subtracted from each of these mean values to partialise out the influence of all other values.   (Saris et al. 2013). Other reliability measures of the Human Value Scale produce similar coefficients (Schwartz et al. 2015).
Footnote 3 (Continued) Figures 1, 2, 3, 4 and 5 show the averages of EPE and the four value variables for each country. As we can see, the EPE means vary between 3 and 1.6 (on a scale of 1 to 5), all of which are surprisingly low. Looking at Figs. 2, 3, 4 and 5, we see that progressive predominance is only partially reflected in the data. Security and universalism show quite high mean values, tradition and self-direction are somewhat lower. It is noticeable that tradition has the lowest mean values.

Independent variables at context level
For the context level, several data sources are used. First, 'progressiveness' of political parties operationalisation is based on party manifestos coding from the Manifesto Project (Volkens et al. 2021a). Manifesto Project Data contains collected and coded information of the content of party manifestos and derived party positions in several political issues. These data provide the quantitative emphasis of a total of 56 political issues by parties, measured in terms of percentage frequencies of emphasis in the overall manifesto. 4 We generated two variables from this data: the first one is intended to test H2a and captures the proportions of conservative and progressive content in the respective party manifestos. 5 The party manifestos from the last national parliamentary election prior to the ESS Round 9 are used as the sample. The individual codes are first weighted with the share of seats 6 won by the Fig. 5 Means of self-direction by country respective party in the last election. To generate the variable, the total conservative content is subtracted from the total progressive content for each country across all parties. These variable can be interpreted according to sign and amount: if the variable has a positive sign, there is a preponderance of progressive content in the respective party landscape. The higher the value of the variable, the stronger the predominance of conservative or progressive issues.
The second variable captures economic issues relevance in party manifestos. It is intended to measure the relationship between issues with cultural content and those with economic content to test H2b. The sum of all cultural issues, whether conservative or progressive, is added up and divided by the sum of economic issues. All codes included in this calculation are weighted per seat share in the same way as the previous variable. This calculated variable quantifies the relationship between cultural issues and economic issues in manifestos. 7 It can also be interpreted in terms of content: A value of less than 1 indicates a higher relevance of economic topics compared to cultural topics, vice versa for a value greater than 1. The more the value deviates from 1, the more unequally economic or cultural issues are represented in the party landscape, depending on the direction. In order to avoid bias in the estimators due to strong outliers in both variables (see above all Hungary), these variables are also logarithmised. The respective scores of the 23 countries for these two variables are shown in Figs. 6 and 7.
Using OECD (2021) and Eurostat (2020) data, two variables are used to capture economic uncertainty at the country level: the GINI coefficient and the unemployment rate. Again, the last data point before the ESS Round 9 survey was used for both variables. Table 1 displays the descriptives for all independent variables.

Control variables
We controlled for variables related to EPE (e.g. Grönlund and Setälä, 2007). These include political interest, trust in political system, in politicians and in the national parliament, and whether the respondent voted for the 'winner' in the last national election before Data collection of ESS Round 9. To operationalise birth cohorts as controls, the classification of Inglehart and Norris (2017a: 36f.) is used for relatedness. Based on the year of birth of each respondent, the cohorts were formed as follows: Interwar (1926-1945), Baby Boomers (1946-1964, Generation X (1958)(1959)(1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968) and Millenials (1980Millenials ( -1996. In addition, those born after 1996 are included as 'Post-Millenials'. Gender was a binary variable. As mentioned above, SES is quite relevant in the context of individual values. Since it is not the focus of analysis, it is included in the model as control variables. Therefore, educational level and respondent profession are used to capture SES and coded as a binary variables. ISCED levels 6-7 (higher education, from bachelor's degrees up to doctoral degrees) and ISCO levels 1-2 (managers and professionals) to represent high SES positions with high level of education and holders of economic leadership positions. Two further variables are included to capture economic insecurity as a component of low SES: the subjective perception of the respondent's household income insecurity and whether the respondent lives on social benefits. To capture differences in the political cultures or histories, a binary Western and Eastern Europe metric is included as a context level control variable. The average of the universalism and self-direction variable per country are also added as a control variables in each model.

Results
To test the hypothesised relationship between value orientations, EPE and the influence of predominantly progressive politics on this relationship, a multilevel mixedeffects model is estimated. The analysis is weighted with the poststratification and design weight provided by the ESS (2018b). Cross-level interactions and interactions on individual level are calculated to test the hypotheses H2a and H2b. The results are shown in Table 2. As variables of particular interest and due to the great relevance of the cross-level and interaction effects, random slopes are calculated for the four value variables. Robust standard errors are given in brackets. Model 1 implies all individual effects, Model 2 extends these to include context effects, and Model 3 adds cross-level interactions.
The results indicate that traditional and security-oriented values have indeed an influence on EPE. In Models 1 and 2, both Security (− 0.04) and Tradition (− 0.03) show a negative effect that is statistically significant (p < 0.001). The higher traditional and security-oriented variables, the lower the individual EPE. This supports the first hypothesis: Measures associated with traditional and security-oriented values will decrease a respondent's reported external efficacy. Therefore H1 can be confirmed.
The 'heart' of this analysis, the interactions between values and context in the form of party manifestos, can be found in Model 3. Among the two traditional value types, only Tradition shows a significant effect (p < 0.01 for both interactions) in the interaction with both manifesto variables (see Figs. 8, 9 for visualization). 8 H2a and H2b can thus be confirmed for this value type. It can be seen that the negative effect of Tradition on EPE becomes stronger the more progressive the content of the party manifestos are (− 0.01). The same pattern is found in the relevance of economic issues (− 0.07). The fewer economic issues the party manifests compared to cultural issues, the stronger the effect of Tradition on EPE. Thus, the central thesis of the Cultural Backlash of progressive predominance as a reason for political marginalisation of traditional individuals can be confirmed at least for this part of the political system represented by party manifestos. However, these effects are not significant for Security. Therefore, these results support hypothesis H2a, More progressively predominated politics will strengthen the effects on external efficacy assumed   in H1, and H2b, Politics with stronger focus on cultural than economic issues will strengthen the effects on external efficacy assumed in H1, only with regard to cultural traditional values, but not for security-oriented ones. At this point, we can only speculate on why this is the case. It seems reasonable to presume that Security is less culturally related than Tradition. Thus, one could assume that Tradition is associated with the conservative cultural part of materialism, while Security rather reflects the need for safety. To shed more light on this distinction, further research is needed. In the last step of the analysis, we will examine how the moderation effect of the 'progressiveness' also applies to different emphases of the party manifestos. So far, only the average of all party manifestos has been taken into account. In the following, a broader panorama of the context of party politics and this moderation effect will be examined in more detail. As we have assumed, political elites are a carrier of progressive predominance of the political system. If the effect were stronger for the party programmes of the governing parties, as the most influential political elite, than for the average, this would support this assumption. For this purpose, Model 3 is re-estimated with the party manifesto Fig. 8 Marginal effects for the cross-level interaction 'Progressive versus Conservative × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 3 as included in Table 2 Fig. 9 Marginal effects for the cross-level interaction 'Salience of Economic Issues × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 3 as included in Table 2 of the governing party (or the average in the case of governing coalitions) instead of the average of all party manifestos (see model 4). Following the assumption that elites are per se more progressive and/or adopt more progressive policies in order to maximise their electoral chances in a predominantly progressive society, it can be assumed here that the moderation effect is stronger since these party(ies) significantly determine national policy. In addition, another model is computed with the party manifesto of the 'most conservative' opposition party(ies), operationalised with the respective minimum of the Progressive versus Conservative and the Salience of Economic issues variables (see model 5). It would be misguided to assume that all parties are equally inclined towards progressiveness, and the more traditional the politics of these opposition party(ies), the smaller the above moderation effect could possibly be.
If only the party manifestos of the 'winning parties' are taken into account (see Table 3), the interaction effect with tradition is stronger Figs. 10,11,12,13,14,15 in the Appendix for visualization). This supports above all the assumption that political elites have a decisive influence on the political agenda (governing parties obviously more than opposition parties) and that their predominantly progressive orientation is therefore more relevant. The party manifestos of the most traditional opposition parties demonstrate the opposite. Here, this effect is much less pronounced. Nevertheless, it should be noted that although the moderation effect is very weak compared to the average model, it is still significant. Thus, it can be stated that this effect could apparently be weakened if 'traditional' parties are part of the national political level. The situation is similar with the distance model. The greater the distance between the governing parties and the most conservative opposition party the weaker the moderation effect. Since the three variants of the party manifest data considered here not only show a broader pattern of the mechanism investigated here, but also turn out in the expected direction, these results reinforce the confirmation of H2a and H2b.
Finally, it should be mentioned that the direction of the relationship between traditional values and external political efficacy is not empirically clear. Since only cross-sectional data are used here to test the hypotheses, it is also conceivable that the effect is in the opposite direction to that shown here. Empirically, low external efficacy could also lead to traditional and security-oriented values or causes traditional attitudes are strengthened. Re-running Model 1-3 with tradition and security as dependent variables and external efficacy as predictor, we find similar effects as shown in Table 2 and a significant positive effect of EPE. In this case, however, the moderating effect of the political context (in the form of an interaction effect of external efficacy and the party manifesto variables) demonstrated above cannot be identified (see Table 4 in the Appendix).

Discussion
What can we learn from this? First, the results show that linking individual values and external political efficacy makes sense in the context of the current discussion on political behaviour that challenges democracy. Tradition and Security as values associated with conservative worldviews show significant negative effects on EPE, understood as the responsiveness of the political system and its actors. This in turn could suggest that EPE might be included in future considerations in studies of the political phenomena addressed in relation to values. The results of this study suggest that this concept might be an important link in the relationship between individual values and political behaviour.
The core purpose of this study, linking values and EPE to the political context, has proved to be quite fruitful. The ideological orientation and the relevance of economic issues in politics show a significant effect on the relationship between traditional values and EPE. The higher the 'preponderance' of progressive politics over conservative politics, the stronger this effect. This strengthening is also found in the economic relevance in politics. The smaller the proportion of economic politics compared to cultural politics, the stronger the negative effect of traditional values on EPE. The argumentation of the Cultural Backlash thesis of perceived irrelevance in progressive political systems is reflected in the findings of this study. Nevertheless, this is only evident for traditional values, not for security-oriented ones. That this effect therefore only manifests itself in a cultural cleavage and that Security needs are not included in this can only be assumed at this point. It should also be noted that this study could not analyse the 'progressiveness' of the political system in depth. The party manifestos examined here should only be understood as an approximation. Further research is needed to analyse exactly which issues and values underlie this relationship between traditional values, EPE and progressiveness of political system. It is therefore worth analysing which elements and processes of a political system particularly influence the effect of traditional values on EPE and what role the existence of right-wing populist parties plays in this. In this context, it should also be mentioned that the analyses in this study are based on crosssectional data and thus only illuminate a 'section' of the underlying Value Change. Particularly with regard to the connection between low EPE and the election of right-wing populist parties, it seems worthwhile to focus on how the connection between EPE and traditional values shown here can in turn affect politics and thus the predominant values at the political level. Fig. 10 Marginal effects for the cross-level interaction 'Progressive versus Conservative-Winner × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 4 as included in Table 3 Fig. 11 Marginal effects for the cross-level interaction 'Relevance of Economic Issues-Winner × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 4 as included in Table 3 Fig. 12 Marginal effects for the cross-level interaction 'Progressive versus Conservative-Opposition × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 5 as included in Table 3 Fig . 13 Marginal effects for the cross-level interaction 'Relevance of Economic Issues-Opposition × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 5 as included in Table 3 Fig. 14 Marginal effects for the cross-level interaction 'Progressive versus Conservative-Difference × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 6 as included in Table 3 Fig. 15 Marginal effects for the cross-level interaction 'Relevance of Economic Issues-Difference × Tradition' from multilevel mixed-effects models predicting external efficacy. Notes: Calculation based on Models 6 as included in Table 3