Keywords: values; social values; public opinion; welfare; social policy
No man is an island, and this interconnectedness with others guides all aspects of human experience (Christakis and Fowler, 2009). This mutuality is most clearly seen in modern politics in welfare policies that are designed to redistribute public funds to provide security and promote equality. Given vast country-level variations in approaches to the principle of state responsibility, it is often suggested that individual-level welfare attitudes are, at least to some extent, driven by contextual influences. So far the literature on the role of contextual factors on attitudes towards welfare has focused on institutional factors, arguing that support for government provision of social services is conditioned by the type of welfare regime and incentives provided to individuals under different welfare policies (Arts and Gelissen, 2001; Castles and Mitchell, 1992; Korpi, 1980; Svallfors, 1997). However, institutions are not the only contextual factor shaping individual attitudes (Lawson, 2008). For example, citizens of a particular country share notions of justice, solidarity and reciprocity norms resulting from their collective history, identity and reference points; in turn, these social-level values manifest themselves in individual attitudes (Davidsson and Marx, 2013; Hall, 1986; Mau, 2004; Rothstein, 1998; Van Oorschot, 2007). This link between social values and individual-level welfare opinions remains under-theorised and rarely addressed empirically, despite the recent ‘cultural turn’ in the study of welfare attitudes (Dion and Birchfield, 2010; Van Oorschot, 2010; Van Oorschot et al., 2008).
This article attempts to fill this gap by studying the effect of social values on attitudes towards social welfare services and support for welfare spending, building on insights from social and political psychology, particularly Shalom Schwartz’s (2004) theory of cultural values. Using multi-level modelling on data from the most recent waves of the European Social Surveys (ESS) (2008) and International Social Survey Programme (ISSP) Role of Government module (2006), we show that individuals in nations that emphasise egalitari- anism and embeddedness are more supportive of social insurance and increased welfare spending, and tend to evaluate welfare policies more positively.
Social values also condition the effect of ideological predispositions on social welfare preferences, such that certain ideological orientations are more susceptible to social envi- ronmental influences. We show that the welfare attitudes of right-wing identifiers are more affected by embeddedness values than those of their left-wing counterparts, making right-wing identifiers in countries ranking high on embeddedness quite supportive of government responsibility. Overall, our findings highlight the multilayered and interactive character of political attitudes, as informed by personal predispositions and broader societal contextual influences.
Social Values and Individual Attitudes
The effect of institutional or economic context on attitudes towards welfare and redistribution is well recognised. Guided mostly by Gosta Esping-Andersen’s worlds of welfare paradigm (Esping-Andersen, 1990), it has been suggested that different types of welfare regime produce different cleavages among socio-economic groups, leading to differences in attitudes towards welfare and redistribution among these groups (Arts and Gelissen, 2001; Castles and Mitchell, 1992; Svallfors, 1997). For example, since Nordic countries have social democratic welfare regimes offering universal entitlements and comprehensive social security benefits, all groups are expected to be supportive of redis- tribution. In contrast, in liberal welfare regimes such as the United States and United Kingdom, social security systems only deliver benefits to those who are in absolute need; so support for social welfare is expected to be highest among the most disadvantaged, such as the unemployed or the disabled. Thus, institutional context is expected to influence individual attitudes towards welfare through the conditioning of individual self-interest.
However, studies have provided, at best, mixed empirical support for the regime hypotheses (Arts and Gelissen, 2001; Jaeger, 2006; Svallfors, 1997; but see Linos and West, 2003), and some empirical findings even run counter to its predictions (Bean and Papadakis, 1998; Gelissen, 2000). In addition, while many studies show that the effect of some individual-level variables on welfare attitudes varies cross-nationally, they do not directly test whether these variations are due to the effect of welfare institutions or some other contextual factor. Thus, for example, Christian Larsen (2008) suggests that welfare regimes actually shape the way people perceive the poor or poverty – that is, whether poverty is due to luck or laziness – rather than conditioning self-interest. The potential role of social values as a fundamental contextual influence in shaping individual attitudes towards redistribution has been understudied (Dion and Birchfield, 2010, p. 331). Below, drawing on research in social and political psychology, we discuss social values and derive hypotheses regarding how they influence individual attitudes.
Social or cultural values may be defined as values representing a society’s shared ideas about what is good, right and desirable (Schwartz, 1999, p. 25). Social values have been found to be increasingly important in explaining cross-national variations in political outcomes, such as institutions (Greif, 1994) and their performance (Licht et al., 2007;
Putnam, 1993), economic development (Tabellini, 2010) and public policy (Arikan, 2010; 2011; Jacobs, 1992; Lockhart, 2003).
Social values have evolved over long periods of time and function as the broadest and most fundamental context for social interaction (Johnston and Klandermans, 1995, p. 4). The dominant social values in a society offer accessible frameworks for thinking about political issues by defining the relevant information for the debate, and the applicable principles and standards for evaluating events (Nelson et al., 1997; Reese, 2001, p. 11). In addition, they define social and political problems and indicate solutions to them (Gamson, 1992), while influencing how the media and elites frame issues (Hertog and McLeod, 2001).
Even when individuals do not fully adopt the values their society emphasises, their mental schemas and ways of thinking, acting and behaving are affected by them (Smith et al., 2006). Individuals tend to use terminology and refer to principles central to their society’s shared values when discussing policy preferences, even if their personal value orientations conflict with those of the society (Feldman and Zaller, 1992). Despite sharp societal disagreements, citizens of the same country may show more than a little consensus on certain issues since they share a common moral vocabulary (Bellah et al., 1985). Social values also mould individuals into certain ways of thinking (Triandis, 1994) as they serve as a standard for judging events (Smith et al., 2006). For example, both the American public and the media tend to attribute poverty to lack of individual effort and laziness, thereby placing responsibility on the individual, while citizens of most European countries tend to blame contextual factors like luck or social and economic conditions (Alesina and Glaeser, 2004; Iyengar, 1991; Semetko and Mandelli, 1997, p. 206). Similarly, political actors in different countries emphasise different concerns when considering the proper role of government in the economy (King, 1973).
Individuals are also exposed to contextual influences through their social networks and interactions with others (Huckfeldt et al., 2004), so become aware of opinions reflecting socially dominant value orientations (DiMaggio, 1997). Most people are neither suffi- ciently well informed nor cognitively active enough to develop counter-arguments to these influences (Iyengar, 1991), tending instead towards conformity and identification with in-group norms (Tajfel and Turner, 1979). Consequently, they end up retaining nearly all the information that reaches them (DiMaggio, 1997), making social values an important influence on individual attitudes. The next section discusses the types of social value that are relevant in influencing welfare attitudes.
Egalitarianism, Embeddedness and Support for Welfare Policies
Scholars from various disciplines have constructed different theories and instruments for mapping and comparing national cultures (Hofstede, 1980; Inglehart, 1990; Inglehart and Welzel, 2005; Triandis, 1994). Here, we use Schwartz’s theory of cultural values since it provides a comprehensive and fine-tuned characterisation of those social values relevant to understanding social welfare attitudes.
Schwartz’s value theory posits that social value dimensions reflect how societies resolve basic issues in regulating human activity (Schwartz, 2006, p. 138, p. 140). He identifies three major universal human requirements, and a number of values that reflect various ways of satisfying these needs (Schwartz, 2004; 2006; Schwartz and Ros, 1995). The first issue confronting human societies is the nature of the relationships between the person and the group, that is, to what extent people are embedded in their groups as opposed to being autonomous. Embeddedness characterises societies in which the individual is not autono- mous but embedded within the collectivity with responsibility for fulfilling the group’s goals. Embedded societies emphasise the values of social order, respect for tradition, security, obedience and wisdom, an emphasis celebrated, for instance, in Confucian-influenced coun- tries (Bond, 1996) and Middle Eastern nations (Schwartz, 2004). In contrast, autonomous societies focus on the values of broad-mindedness, curiosity, creativity, pleasure and an exciting life, which are often emphasised in many West European countries (Ester et al., 1994; Schwartz, 2004).
We expect embeddedness to be positively related to support for social insurance since this dimension also stresses taking care of members of the collectivity. This is in accord with the framing of welfare policies as part of a social duty towards the poor and the needy, and as a matter of collective virtue and promoting a collective good life (Freeden, 2003, pp. 14–5). Welfare policy may also be seen as a way of integrating the individual into an organic entity (Esping-Andersen, 1990, p. 40), as opposed to promoting individual autonomy and self-reliance. Further, social insurance and welfare policy often represent a tension between two important values: liberty and security (Freeden, 2003, p. 15). Consequently, individuals socialised in settings that stress a collectivity’s responsibility for the security and well-being of others may be more willing to accept government intervention in the economy to provide for the society’s needy groups. Since societies that emphasise embeddedness prioritise group responsibility towards others, and stress security and obedience, we expect these values to be associated with greater individual support for welfare policies (H1).
The second societal problem Schwartz considers is guaranteeing that people behave in a responsible manner that preserves the social fabric, which corresponds to the egalitarianism vs. hierarchy dimension. Egalitarian societies seek to induce people to recognise each other as moral equals, emphasising the internalisation of in-group cooperation, and to socialise people to be concerned with everyone’s welfare. Equality, social justice, responsibility, help- fulness and honesty are the types of value egalitarian societies emphasise, such as Western European democratic welfare states (Ester et al., 1994). In contrast, hierarchy values socialise people into accepting hierarchical roles as legitimate and complying with the duties and obligations attached to their roles, breeding values such as social power, authority, humility and wealth. A heavy emphasis on hierarchy and a rejection of egalitarianism is also known to characterise some Asian and Latin American societies (Schwartz, 2004).
Since the goal of welfare policies is to ensure equal opportunity and safety nets to those who are in need, egalitarianism, with its emphasis on equality, social justice and respon- sibility, is expected to be highly relevant in influencing attitudes towards social insurance policies that aim at enhancing the welfare of the society’s members. In egalitarian societies, individuals are exposed to frames emphasising equality, social justice, concern for and responsibility towards others, and thus are expected to be more supportive of the welfare state and increased spending on social welfare programmes (H2).
The final societal issue is regulating how people manage their relations to the natural and social worlds, represented by the harmony vs. mastery dimension. Since this dimension concerns the individual and societal relationships with nature and is not directly associated with caring for the well-being of others, we do not expect it to affect welfare attitudes.
Social Values, Ideology and Welfare Attitudes
We next argue that social values influence the relationship between individual predispo- sitions and attitudes towards redistribution and welfare. Individuals’ pre-existing inclina- tions such as ideology, socio-economic background or subjective beliefs that are crucial in shaping attitude formation may be in line or clash with dominant social orientations, and the social context might also affect the way individual characteristics shape opinion formation (Arikan, 2010; Ben-Nun Bloom and Levitan, 2011). That is, although egali- tarianism and embeddedness are expected to lead to pro-welfare attitudes, we expect these values to influence certain types of individual more than others.
Social values reinforce the way ideological predispositions influence attitudes towards welfare (Arikan, 2010; Arikan and Ben-Nun Bloom, 2013). When individuals are exposed to social influences that reinforce already existing beliefs and considerations, the end result is to strengthen their position regarding the attitude in question.
Ideological predispositions are key predictors of welfare attitudes, as the left–right ideological divide is conceptually and empirically closely related to the debate about welfare (Jaeger, 2008; Lipset and Rokkan, 1967), with economic equality being the core issue (Bobbio, 1996, p. 60). While the meanings of left and right vary across contexts and time, they still have significant power to explain economic policy positions, with the majority of people and political parties locating themselves and political parties along this continuum (Mair, 2007).1
Right-wing ideology generally does not support principles of fairness, supporting instead the rhetoric of individual achievement and restoration of social hierarchy. Right- wing framing of the welfare debate frequently focuses on government waste and individual responsibility, considering it unnecessary to reduce social status differences (e.g. Jacoby, 1994). However, right-wing orientations are also compatible with the rhetoric of patrio- tism and cultural pride. Thus, framing poverty or welfare so as to emphasise the interconnectedness of the country’s citizens and collective responsibility, or alternatively in terms of security, so that avoiding poverty and providing social insurance becomes a matter of promoting social interests and order (Freeden, 2003; Taylor-Gooby, 1983), may have a particular effect on how right-wing identifiers think about welfare, while reinforcing already existing predispositions.
Since embeddedness values emphasise obedience and the shared faith in collectives, which are compatible with right-wing orientations, we expect them to be associated with welfare frames that reinforce the effect of right-wing ideology, thereby increasing right- wing and, to an extent, centrist identifiers’ support for welfare policy. Thus, in a social context where social order, responsibility and collective well-being are emphasised, right- wing and moderate identifiers may, to a larger extent, consider social insurance as pro- moting social order and traditional values, and strengthening national identity. Therefore, we expect right-wing identifiers to be more supportive of welfare services in societies that are more embedded.
Left-wing ideology, which has historically been associated with promoting welfare, social justice and equality, stresses redistribution of wealth and income, and calls on government to take more economic responsibility (Piurko et al., 2011). Social justice and provision of social insurance is a core concern of left-wing identifiers in almost all countries (Mair, 2007), and left-wing identifiers are almost universally united by the view that government should assume full responsibility for society. Given these already very strong individual-level predispositions, additional framing in terms of societal-level egali- tarianism or embeddedness may not be able to increase left-wing identifiers’ support for redistribution further.
Thus, we expect embeddedness orientations to affect right-wing and centrist identifiers more than left-wing identifiers. Specifically, we predict right-wing and moderate indi- viduals to be even more supportive of welfare policies with increased societal-level embeddedness. On the other hand, social values are hypothesised to have less impact on left-wing identifiers: individuals who lean more towards the left will be less susceptible than right-wing identifiers to the effect of social values in terms of issues concerning welfare and social insurance (H3).
We used data from the fourth wave of the ESS (2008) and the ISSP Role of Government (2006) module. While the ESS comprises a large pool of European nations, the ISSP comprises a diverse pool of advanced industrialised nations as well as some developing countries.2 The range of countries included in the samples and the slightly different wordings for the dependent variables increase the robustness of our findings.
Three dependent variables were constructed. The first is support for government responsibility in providing social welfare services, coded using six ESS items (answered on a 0–10 scale) asking whether or not it should be the government’s responsibility to: (1) provide a job for everyone; (2) ensure adequate health care for the sick; (3) ensure a reasonable standard of living for the elderly; (4) ensure a reasonable standard of living for the unemployed;
- ensure sufficient childcare services for working parents; (6) provide paid leave from work for people who temporarily have to care for sick family
We used multiple group confirmatory factor analysis (MG-CFA) to test for the internal consistency of the items and differential item functioning (Reise et al., 1993; Van de Vijver and Leung, 1997). We first ran a baseline model in which all factor loadings for each scale were constrained to be equal in all countries, while factor means, variances, covariances and residuals were freely estimated for each country (CFI = 0.957, TLI = 0.946, RMSEA = 0.095). Next, we evaluated the modification indices and introduced several modifications to the models by relaxing the measurement invariance constraints for certain items. We then saved the factor scores from the final partial invariance model (CFA = 0.987, TLI = 0.980, RMSEA = 0.058), in which items that performed differently across countries were given different weights to ensure a common measurement scale (Reise et al., 1993).
The ISSP Role of Government Survey also has a number of items that tap support for government responsibility in providing welfare services. We used six items (measured on a five-category Likert scale) asking whether it should or should not be the government’s responsibility to provide: (1) a job for everyone; (2) health care for the sick; (3) a decent standard of living for the elderly; (4) a decent standard of living for the unemployed; (5) financial help to university students from low-income families; and (6) decent housing for those who cannot afford it. We also used MG-CFA in order to correct for potential item biases across nations (full invariance model, CFA = 0.832, TLI = 0.918, RMSEA = 0.172; final partial invariance model CFI = 0.920, TLI = 0.956, RMSEA = 0.126). Again, factor scores from the final partial invariance model were saved to use as dependent variables.
The second type of dependent variable came from the ESS and measures the extent to which individuals believe that welfare services are beneficial for the society, that is, the positive moral consequences of the welfare state (see Van Oorschot, 2010). Respondents were asked (on a five-point Likert scale) how much they agreed or disagreed that social benefits and services in their country: (1) make people lazy; (2) make people less willing to care for one another; and (3) make people less willing to look after themselves and their family. The final partial invariance model had excellent fit statistics (CFA = 0.997, TLI = 0.995, RMSEA = 0.045).
From the ISSP dataset, we also constructed a support for increased welfare spending depen- dent variable, using three five-category items asking respondents whether they would like to see more or less government spending on health, old age pensions and unemployment benefits. Again, the items were coded so that higher values represent higher support for welfare spending. Factor scores from pooled CFA models (CFI = 1.000, TLI = 1.000, RMSEA = 0.000) were used in the subsequent analysis.
Higher values of the dependent variables indicate greater support for government provision of social services, greater support for increased spending on welfare and more positive attitudes towards welfare services. To ensure they were all on the same metric, all the dependent variables were standardised to vary between 0 and 1.
Measuring Social Values
Since individual and social value orientations are conceptually different and refer to different levels of measurement (Fischer et al., 2010; Schwartz, 2006), social values cannot be inferred from the aggregate predispositions of a cultural group, that is, by simply taking the mean of group members on a particular item, or using within-society correlations (Hofstede, 1980). Instead, between-country correlations calculated from mean values of each variable for each society are used to assess the dimensions of social value orientations using either exploratory factor analyses (Hofstede, 1980; Inglehart and Welzel, 2005) or multidimensional scaling (Schwartz, 2004; 2006).
Schwartz derived social value structures using countries as units of analysis, although the assumption that a nation shares common attitudes and values may be problematic since it assumes that a degree of consensus exists among the nation’s citizens. For example, in some societies, Kathleen Dowley and Brian Silver (2005) have found disparities in inter- personal trust, achievement motivation and confidence in institutions across ethnic and/or linguistic minorities and majority groups. Yet while a degree of within-nation cleavage
may exist, it is still possible to talk about nationally meaningful value orientations. Forces encouraging integration, such as a common dominant language, political and educational systems, the mass media and national symbols, combine to produce substantial sharing of values (Hofstede, 1980; Smith et al., 2006). Despite significant ethnic diversity, a country’s institutions push its inhabitants towards greater cultural unity (Smith et al., 2006, p. 56). As a result, a nation’s various ethnic groups often produce similar profiles on psychologically relevant measures vis-à-vis other nations’ citizens (Smith et al., 2006). In fact, Schwartz’s (2004, p. 57) comparison of both within- and between-country cultural distances4 shows that cultural distances between samples from different countries are almost always greater than those between samples from the same country, suggesting that nations may be taken as meaningful cultural units.
Schwartz’s egalitarianism and embeddedness scores are taken from Amir Licht et al. (2007). The descriptive statistics by nations shown in Table S1 in the online Appendix typically reflect the theoretically expected profile. Overall, the scores suggest that West European culture emphasises egalitarianism and is low on embeddedness, as expected from a region of developed, democratic states (Ester et al., 1994). Countries such as Norway, Denmark, France, Germany and the Netherlands, and some Southern European countries – specifically Spain and Portugal – rank high on egalitarianism and low on embeddedness. Some post-communist countries such as Hungary, Poland, Slovenia and the Czech Repub- lic which have, according to some researchers, rejected the rhetoric of authoritarian communist regimes (Schwartz and Bardi, 1997), score among the lowest on egalitarianism, while others, particularly Estonia and Poland, are also the most embedded countries, together with a number of developing countries such as Cyprus, the Philippines and Turkey. Next, consistent with analyses of Confucian culture (Bond, 1996), Confucian- influenced countries in the sample, specifically Taiwan and Japan, reject egalitarianism, scoring the lowest on that scale. However, while Taiwan also scores high on embeddedness, Japan falls in the mid-scale. Generally, English-speaking countries such as the US and the UK lie mid-scale on both value orientations, being more clearly distinguished by two other value dimensions not studied here, mastery and individualism (Schwartz, 2004).
The models also controlled for age, gender, marital status, having children, level of education (measured by number of years of schooling), income,5 self-placement on the left–right scale,6 interpersonal trust, level of religiosity, satisfaction with the country’s economy, and dummy variables for being retired, permanently disabled, unemployed or having union membership. All independent variables except age and years of education were coded to vary between 0 and 1 (See Table S2 in the online Appendix).
It has been suggested that welfare regimes reflect enduring, deeply rooted social values (Esping-Andersen, 1990; Van Oorschot, 2010), so certain types of welfare state could be associated with some of these core social values. It is therefore possible that the dominant values of a society are closely related to its welfare regime type, which may also influence individual attitudes towards social insurance. However, the similarity in mean egalitarian- ism and embeddedness scores across countries with each of the three types of welfare state identified by Esping-Andersen, or the Southern type of welfare state or post-communist nations, suggests that certain types of welfare regime are not necessarily distinguishable by their dominant value orientations (See Table S3 in the online Appendix). Only the post-communist nations included in the study are distinguishable from other types of welfare regime, in that their mean levels of embeddedness are higher, while their mean egalitarianism is lowest. Nevertheless, this study controlled for the effect of welfare institutions using a combined index of social security laws that indicates the generosity of benefits for the elderly, disabled and sick (Botero et al., 2004), public social spending as percentage of GDP, measures of de-commodification, and a benefits generosity index to account for the welfare regime type (Esping-Andersen, 1990; Scruggs and Allan, 2006).
National economic conditions such as unemployment levels (Blekesaune, 2007; Blekesaune and Quadagno, 2003) and income inequality (Dion and Birchfield, 2010) are also known to influence individual support for redistribution and welfare policy. There- fore, we also controlled for GDP per capita, levels of inflation and unemployment, as well as communist legacy.
To test the effect of cultural-level values on welfare attitudes, national-level egalitarianism and embeddedness scores were submitted to a multi-level regression analysis with a set of individual-level predictors explaining support for government responsibility (Model 1a, using the 2008 ESS data, and 1b, using the 2006 ISSP data), opinion on social welfare services (Model 1c, using the 2008 ESS data) and support for increased welfare spending (Model 1d, using the 2006 ISSP data). Table 1 presents the results of this analysis.
Figure 1: The Conditional Effect of Embeddedness and Ideology on Attitudes towards Welfare
The findings present strong evidence for the effect of social values on individual attitudes towards support for social welfare policies. In accord with Hypotheses H1 and H2, both embeddedness and egalitarianism had positive and statistically significant effects on the dependent variables, indicating that, ceteris paribus, the more egalitarian and/or embedded a country is, the more likely its citizens are to support government provision of social insurance and spending on social welfare services, and to believe in the positive moral effects of welfare policies. However, in Model 1c, the standard error of the coefficient for egalitarianism was just too high to justify rejecting the null hypothesis.
Next, we were interested in the conditional effects of the social value dimensions. For this, we re-ran the same multi-level models appearing in Table 1, adding a cross-level interaction of social values and individual-level ideology. Table 2 presents the coefficients for these four models with cross-level interactions.
The cross-level interaction of national-level embeddedness and individual-level ideology was positive and statistically significant in three models, the exception being Model 2d, where the dependent variable was support for increased welfare spending. Figure 1 plots the conditional effect of country-level embeddedness by individual-level ideological orientation. The predicted levels of support for government provision of welfare services (the upper left panel for ESS data and the upper right panel for ISSP data) and positive evaluations of welfare policy (bottom left) as embeddedness increases are presented for extreme leftists (thin black line), moderates (dark grey line) and extreme right-wing identifiers (light grey line).
As expected, individuals from different ideological orientations were all positively affected by societal emphasis on embeddedness, but in accordance with H3, this effect was not uniform across all individuals. The results indicate that moderate to right-wing identifiers are much more affected by social values emphasising embeddedness than left-wing identifiers. All else being equal, the difference in predicted level of support for government responsibility for left-wing identifiers in the least and most embedded societies is about 4 per cent of its range (0.04 on a scale of 0 to 1), while for extreme right identifiers this difference is about 15 per cent of the range (ESS 2008 data), with the increase being about 9 per cent for moderate identifiers. Findings from both datasets thus show that moderate and right-wing identifiers are more prone to the effect of embeddedness while left-wing identifiers are less affected as predicted by H3.
The conditional effect of embeddedness was even more dramatic for positive evaluations of welfare policy (see lower left-hand panel of Figure 1). While there was no difference in the level of support of left-wing respondents across nations with differing levels of embeddedness, moderate and right-wing identifiers’ belief in the beneficial effect of welfare policies increased the more their society emphasised embeddedness. In addition, in highly embedded cultures, the predicted level of positive evaluations of welfare policies was higher for moderate and right-wing identifiers than for left-wing respondents.
We ran a number of models to ensure that the effects of our key level 2 variables are robust when controlling for other sources of contextual influence. Full results are presented in the online Appendix (Table S4), while key findings are presented below.
Random Intercept Models (Models 1a to 1d). We were first interested in testing the robustness of the effects of social values when integrating additional level 2 covariates. Each institutional and economic control variable was first submitted to multi-level analysis as the single level 2 control, before being run together with the two value orientations in the same model in an alternative specification.7 This showed that unemployment, inflation and GDP per capita all had positive effects on government responsibility and welfare spending when they were the only level 2 regressors, although their statistical significance usually disappeared when social values were added. Communist legacy did not have any statistically significant effects on the dependent variables. Social security laws, public social expenditure and Esping-Andersen and Scruggs and Allan indices were all statistically significant predictors of belief in government responsibility (in most cases, even when they were included with values in the same model), but not of support for welfare spending. In all models where government responsibility was the dependent variable, and in the majority of models that predicted welfare spending, both values retained their positive and statistically significant effects. These results were replicated using either egalitarianism or embeddedness as single level 2 predictors. The findings suggest that, while institutional structure and welfare regime affect individual attitudes towards social policy, their effects are less robust than the effect of social values.
Random Coefficient Models (Models 2a to 2c). We re-ran the models with multi-level interactions for government responsibility and evaluation of welfare services, adding each level 2 control one at a time. Controlling for unemployment, inflation, GDP per capita, post-communism and social security laws did not substantively change the results. Since some controls, such as the Esping-Andersen and Scruggs and Allan indices, reduced the level 2 degrees of freedom significantly, which leads to unreliable estimates, we did not control for the effects of these variables. These results were replicated using egalitarianism or embeddedness as single level 2 predictors.
Measurement of Social Values. Finally, some researchers prefer to use Schwartz values by subtracting the score of one type of value from the score of the value type constituting the opposite end of the dimension (e.g. egalitarianism minus hierarchy) so that the resulting score reflects the relative emphasis given to a particular value (Schiefer, 2013; Schwartz, 2006). Such scores correlate strongly with the scores used in the analyses (usually over 0.9) and models employing these measures yield the same results (Table S6A and S6B in the online Appendix). Therefore, we do not additionally present the results of models using these subtractive scores in this study.
Although Columbia School scholars have consistently stressed the effect of social context and ‘cross-pressures’ on the formation of political attitudes and behaviour (Berelson et al., 1954; Lazarsfeld et al., 1944), the role of social context has been largely neglected in the studies of attitudes. This study provides new empirical evidence for the key role of social values, an important component of social context, for welfare attitudes. Building on Schwartz’s theory of social value dimensions, we have shown that both embeddedness and egalitarianism increase support for social welfare, above and beyond other contextual influences, such as institutions or national economic conditions. Individuals in societies that emphasise equality and social justice, as well as social order and collective responsibility, are overall more supportive of government provision of welfare services, and tend to evaluate the societal effect of welfare policies more positively.
The results of this study suggest that support for social policy may be conditioned by different motivations. Individuals in egalitarian nations, such as Western European coun- tries, seem to be more supportive of welfare policies because these policies are believed to create equal outcomes for the members of that society. Members of embedded societies, such as post-communist nations or developing countries, are also more supportive of social welfare than those of autonomous societies, which seems mostly to be due to a concern for social order and security, and a sense of collective duty towards others in the society. These cross-national differences in societal motivations for supporting welfare policy deserve further scholarly attention. Thus, for example, egalitarian societies may be more supportive of policies redistributing income to all members of the society, while embeddedness could result in support for specific policies that uphold traditional social structures. In addition, embedded societies may be more opposed to policies that provide more autonomy to the individual, such as social assistance for single mothers, while rejecting social policies that assist groups seen as undeserving.
Our findings also suggest that social values may account for cross-national variation in elite or political party positions. For example, Sidney Verba et al. (1987) found substantive variation among political elites who come from parties with the same ideological orien- tations or those with similar interests (such as labour union leaders), with the more egalitarian Swedish political elites being more supportive of redistribution than American and Japanese elites. Stefan Svallfors (1997) notes that the Norwegian Conservative Party is much less right-wing than its Swedish counterpart Moderate Party, which may again be explained by the higher emphasis given to egalitarianism in Norway compared to Sweden. Future research could therefore focus on the effect of socially held values on cross-national differences in elite opinions and party positions.
So far, the literature on context effects on individual attitudes towards welfare has focused on the role of welfare institutions in shaping these orientations (Rothstein, 1998; Svallfors, 1997). Indeed, our findings confirm that variables such as communist legacy, social security laws and public social expenditure are generally significant predictors of welfare attitudes. However, we have also shown that socially shared values are an additional important influence on welfare attitudes, above and beyond institutional arrangements. In fact, some of the institutional variables lost their statistical significance as predictors of attitudes when we specified social values in the model, which leads to the tentative suggestion that their effect is due to shared variance with social values. Accordingly, future research should also consider the role of social and cultural values in affecting attitudes towards redistribution.
In addition, we have demonstrated that social values may even influence the way individual predispositions affect attitudes towards government responsibility. More specifi- cally, we found that moderate and right-wing identifiers are more influenced by embeddedness than left-wing identifiers. This also suggests a boundary condition for the effect of social values: when individuals have strong predispositions that are generally compatible with the values framework offered by their social environment, these societal dispositions affect their views. That is, cultural context influences the extent to which personal ideology determines individual support for redistribution and social welfare. Thus, in more embedded countries, even right-wing individuals do not seem to question the role of government in redistributing income, whereas less embedded nations are characterised by more variation in the redistributive attitudes of identifiers of different ideological orientations.
Individual attitudes, however, are not just shaped by an individual’s political ideology or the characteristics of political issues; they are also shaped by the broader social context, particularly by socially dominant value orientations. In addition, the effect of social context is not necessarily additive. Context rather constitutes a complex interplay with individual- level preferences, such that individual-level political ideology and social environment interact in moulding political attitudes. Therefore, a complete theory of political behaviour has to consider carefully the interplay across several levels of analysis. Future studies of welfare and social policy preferences should thus not only consider the role of personal predispositions and social environment, but also the interaction between them.
About the Authors
Gizem Arikan is an Assistant Professor in the Department of International Relations at Yasar University, Izmir, Turkey. Her research focuses on the effect of political culture and values on public opinion and policy. She also conducts research on religiosity and attitudes towards democracy. Gizem Arikan, Department of International Relations, Yasar University, Selcuk Yasar Kampusu, Agacli Yol 35–37, Bornova, Izmir 35100, Turkey; email: [email protected]
Pazit Ben-Nun Bloom is an Assistant Professor in the Department of Political Science at The Hebrew University of Jerusalem, Israel. Her research examines the role of morality, religiosity and values in political behaviour. Her broad research interests are in political psychology, comparative political behaviour and political methodology. Pazit Ben-Nun Bloom, Department of Political Science, The Hebrew University of Jerusalem, Mount Scopus, 91905 Jerusalem, Israel; email: [email protected]
Alesina, A. and Glaeser, E. (2004) Fighting Poverty in the US and Europe: A World of Difference. Oxford: Oxford University Press.
Arikan, G. (2010) Economic Individualism and Cross-National Differences in Redistribution. Unpublished PhD thesis, Stony Brook University.
Arikan, G. (2011) ‘Economic Individualism and Government Spending’, World Values Research, 4 (3), 73–95.
Arikan, G. and Ben-Nun Bloom, P. (2013) ‘The Influence of Societal Values on Attitudes towards Immigration’, International Political Science Review, 34 (2), 210–26.
Arts, W. and Gelissen, J. (2001) ‘Welfare States, Solidarity and Justice Principles: Does the Type Really Matter?’, Acta Sociologica, 44 (4), 283–99.
Bean, C. and Papadakis, E. (1998) ‘A Comparison of Attitudes towards the Welfare States in Different Institutional Regimes, 1985–1990’, International Journal of Public Opinion Research, 10 (3), 211–36.
Bellah, R., Madsen, R., Sullivan, W. M., Swidler, A. and Tipton, S. M. (1985) Habits of the Heart: Individualism and Commitment in American Life. New York: Harper and Row.
Ben-Nun Bloom, P. and Levitan, L. C. (2011) ‘We’re Closer than I Thought: Heterogeneity of Social Networks, Moral Messages and Political Persuasion’, Political Psychology, 32 (4), 643–65.
Berelson, B. R., Lazarsfeld, P. F. and McPhee, W. N. (1954) Voting: A Study of Opinion Formation in a Presidential Campaign. Chicago IL: University of Chicago Press.
Blekesaune, M. (2007) ‘Economic Conditions and Public Attitudes to Welfare Policies’, European Sociological Review, 23 (3), 393–403.
Blekesaune, M. and Quadagno, J. (2003) ‘Public Attitudes toward Welfare State Policies: A Comparative Analysis of 24 Nations’, European Sociological Review, 19 (5), 415–27.
Bobbio, N. (1996) Left and Right: The Significance of a Political Distinction. Cambridge: Polity Press.
Bond, M. H. (1996) ‘Chinese Values’, in M. H. Bond (ed.), Handbook of Chinese Psychology. Hong Kong: Oxford University Press, pp. 208–26.
Botero, J. C., Djankov, S., La Porta, R., Lopez-de-Silanes, F. and Shleifer, A. (2004) ‘The Regulation of Labor’, Quarterly Journal of Economics, 119 (4), 1339–82.
Castles, F. G. and Mitchell, D. (1992) ‘Identifying Welfare State Regimes: The Links between Politics, Instruments, and Outcomes’, Governance, 5 (1), 1–26.
Christakis, N. A. and Fowler, J. H. (2009) Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives. New York: Little Brown and Company.
Davidsson, J. B. and Marx, P. (2013) ‘Losing the Issue, Losing the Vote: Issue Competition and the Reform of Unemployment Insurance in Germany and Sweden’, Political Studies 61 (3), 505–22.
DiMaggio, P. (1997) ‘Culture and Cognition’, Annual Review of Sociology, 23, 263–87.
Dion, M. L. and Birchfield, V. (2010) ‘Economic Development, Income Inequality, and Preferences for Redistribution’, International Studies Quarterly, 54 (2), 315–34.
Dowley, K. M. and Silver, B. D. (2005) ‘Crossnational Survey Research and Subnational Pluralism’, International Journal of Public Opinion Research, 17 (2), 226–38.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.
Ester, P., Halman, L. and Seuren, B. (1994) ‘Environmental Concern and Offering Willingness in Europe and North America’, in P. Ester, L. Halman and R. de Moor (eds), The Individualizing Society: Value Change in Europe and North America. Tilburg: Tilburg University Press, pp. 163–81.
Feldman, S. and Zaller, J. (1992) ‘The Political Culture of Ambivalence: Ideological Responses to the Welfare State’, American Journal of Political Science, 36 (1), 268–307.
Fischer, R., Vauclair, C. M., Fontaine, J. R. J. and Schwartz, S. H. (2010) ‘Are Individual-Level and Country-Level Value Structures Different? Testing Hofstede’s Legacy with the Schwartz Value Survey’, Journal of Cross-Cultural Psychology, 41 (2), 135–51.
Freeden, M. (2003) ‘The Coming of the Welfare State’, in T. Ball and R. Bellamy (eds), The Cambridge History of Twentieth Century Political Thought. Cambridge: Cambridge University Press, pp. 7–44.
Gamson, W. A. (1992) Talking Politics. New York: Cambridge University Press.
Gelissen, J. (2000) ‘Popular Support for Institutionalised Solidarity: A Comparison between European Welfare States’, International Journal of Social Welfare, 9 (4), 285–300.
Greif, A. (1994) ‘Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualist Societies’, Journal of Political Economy, 102 (5), 912–50.
Hall, P. (1986) Governing the Economy: The Politics of State Intervention in Britain and France. New York: Oxford University Press. Hertog, J. K. and McLeod, D. M. (2001) ‘A Multiperspectival Approach to Framing Analysis: A Field Guide’, in S. D. Reese,
O. H. Gandy and A. E. Grant (eds), Framing Public Life: Perspectives on Media and Our Understanding of the Social World. Mahwah NJ: Taylor and Francis, pp. 139–62.
Hofstede, G. (1980) Culture’s Consequences: International Differences in Work-Related Values. London: Sage.
Huckfeldt, R., Mendez, J. M. and Osborn, T. L. (2004) ‘Disagreement, Ambivalence, and Engagement: The Political Consequences of Heterogeneous Networks’, Political Psychology, 25 (1), 65–95.
Inglehart, R. (1990) Culture Shift in Advanced Industrial Society. Princeton NJ: Princeton University Press.
Inglehart, R. and Welzel, C. (2005) Modernization, Cultural Change, and Democracy: The Human Development Sequence.
Cambridge: Cambridge University Press.
Iyengar, S. (1991) Is Anyone Responsible? Chicago IL: University of Chicago Press.
Jacobs, L. (1992) ‘Institutions and Culture: Health Policy and Public Opinion in the US and Britain’, World Politics, 44 (2), 179–209.
Jacoby, W. (1994) ‘Public Attitudes toward Government Spending’, American Journal of Political Science, 38 (2), 336–61. Jaeger, M. M. (2006) ‘Welfare Regimes and Attitudes towards Redistribution: The Regime Hypothesis Revisited’, European Sociological Review, 22 (2), 157–70.
Jaeger, M. M. (2008) ‘Does Left–Right Orientation Have a Causal Effect on Support for Redistribution? Causal Analysis with Cross-Sectional Data using Instrumental Variables’, International Journal of Public Opinion Research, 20 (3), 363–74.
Johnston, H. and Klandermans, B. (1995) ‘The Cultural Analysis of Social Movements’, in H. Johnston and B. Klandermans (eds), Social Movements and Culture. London: UCL Press, pp. 3–24.
King, A. (1973) ‘Ideas, Institutions, and the Policies of Governments: A Comparative Analysis’, British Journal of Political Science, 3 (3–4), 409–23.
Korpi, W. (1980) ‘Social Policy and Distributional Conflict in the Capitalist Democracies: A Preliminary Comparative Framework’, West European Politics, 3 (3), 296–316.
Larsen, C. A. (2008) ‘The Institutional Logic of Welfare Attitudes: How Welfare Regimes Influence Public Support’, Comparative Political Studies, 41 (2), 145–68.
Lawson, S. (2008) ‘Political Studies and the Contextual Turn: A Methodological/Normative Critique’, Political Studies, 56 (3), 584–603.
Lazarsfeld, P. F., Berelson, B. and Gaudet, H. (1944) The People’s Choice: How the Voter Makes up His Mind in a Presidential Campaign. New York: Columbia University Press.
Licht, A. N., Goldschmit, C. and Schwartz, S. H. (2007) ‘Culture Rules: The Foundations of the Rule of Law and Other Norms of Governance’, Journal of Comparative Economics, 35 (4), 659–88.
Linos, K. and West, M. (2003) ‘Self-Interest, Social Beliefs, and Attitudes to Redistribution: Re-addressing the Issue of Cross-National Variation’, European Sociological Review, 19 (4), 393–409.
Lipset, S. M. and Rokkan, S. (1967) ‘Cleavage Structures, Party Systems and Voter Alignments’, in S. M. Lipset and Rokkan (eds), Party Systems and Voter Alignments: Cross-National Perspectives. New York: Free Press, pp. 1–64. Lockhart, C. (2003) Protecting the Elderly: How Culture Shapes Social Policy. University Park PA: Pennsylvania State University Press.
Mair, P. (2007) ‘Left–Right Orientations’, in R. J. Dalton and H. D. Klingemann (eds), The Oxford Handbook of Political Behavior. Oxford: Oxford University Press, pp. 206–22.
Mau, S. (2004) ‘Welfare Regimes and the Norms of Social Exchange’, Current Sociology, 52 (1), 53–74.
Nelson, T. E., Clawson, R. A. and Oxley, Z. M. (1997) ‘Media Framing of a Civil Liberties Conflict and Its Effect on Tolerance’, American Political Science Review, 91 (3), 567–83.
Piurko, Y., Schwartz, S. H. and Davidov, E. (2011) ‘Basic Personal Values and the Meaning of Left–Right Political Orientations in 20 Countries’, Political Psychology, 32 (4), 537–61.
Putnam, R. D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton NJ: Princeton University Press. Reese, S. D. (2001) ‘Framing Public Life: A Bridging Model for Media Research’, in S. D. Reese, O. H. Gandy and A. E. Grant (eds), Framing Public Life: Perspectives on Media and Our Understanding of the Social World. Mahwah NJ: Taylor and Francis, pp. 7–32.
Reise, S. P., Widaman, K. F. and Pugh, R. H. (1993) ‘Confirmatory Factor Analysis and Item Response Theory: Two Approaches for Exploring Measurement Invariance’, Psychological Bulletin, 114 (3), 552–66.
Rothstein, B. (1998) Just Institutions Matter: The Moral and Political Logic of the Universal Welfare State. Cambridge: Cambridge University Press.
Schiefer, D. (2013) ‘Cultural Values and Group-Related Attitudes: A Comparison of Individuals with and without Migration Background across 24 Countries’, Journal of Cross-Cultural Psychology, 44 (2), 245–62.
Schwartz, S. H. (1999) ‘A Theory of Cultural Values and Some Implications for Work’, Applied Psychology: An International Review, 48 (1), 23–47.
Schwartz, S. H. (2004) ‘Mapping and Interpreting Cultural Differences around the World’, in V. Henk, J. Soeters and Ester (eds), Comparing Cultures: Dimensions of Culture in a Comparative Perspective. Leiden: Brill, pp. 43–73. Schwartz, S. H. (2006) ‘A Theory of Cultural Value Orientations: Explication and Applications’, Comparative Sociology, 5 (2–3), 137–82.
Schwartz, S. H. and Bardi, A. (1997) ‘Influences of Adaptation to Communist Rule on Value Priorities in Eastern Europe’, Political Psychology, 18 (2), 385–410.
Schwartz, S. H. and Ros, M. (1995) ‘Values in the West: A Theoretical and Empirical Challenge to the Individualism– Collectivism Cultural Dimension’, World Psychology, 1 (2), 91–122.
Scruggs, L. and Allan, J. (2006) ‘Welfare-State Decommodification in 18 OECD Countries: A Replication and Revision’, Journal of European Social Policy, 16 (1), 55–72.
Semetko, H. A. and Mandelli, A. (1997) ‘Setting the Agenda for Cross-National Research: Bringing Values into the Concept’, in M. McCombs, D. L. Shaw and D. Weaver (eds), Communication and Democracy: Exploring the Intellectual Frontiers in Agenda-Setting Theory. Mahwah NJ: Lawrence Erlbaum, pp. 195–208.
Smith, P. B., Bond, M. H. and Kagitcibasi, C. (2006) Understanding Social Psychology across Cultures: Living and Working in a Changing World. London: Sage.
Svallfors, S. (1997) ‘Worlds of Welfare and Attitudes to Redistribution: A Comparison of Eight Western Nations’, European Sociological Review, 13 (3), 283–304.
Tabellini, G. (2010) ‘Culture and Institutions: Economic Development across Regions of Europe’, Journal of the European Economic Association, 8 (4), 677–716.
Tajfel, H. and Turner, J. (1979) ‘An Integrated Theory of Intergroup Conflict’, in W. Austin and S. Worchel (eds), The Social Psychology of Intergroup Relations. Monterey CA: Brooks Cole, pp. 33–48.
Taylor-Gooby, P. (1983) ‘The Welfare State and Individual Freedom: Attitudes to Welfare Spending and to the Power of the State’, Political Studies, 31 (4), 640–9.
Triandis, H. C. (1994) Culture and Social Behavior. New York: McGraw-Hill.
Van de Vijver, F. J. R. and Leung, K. (1997) Methods and Data-Analysis for Cross-Cultural Research. Thousand Oaks CA: Sage. Van Oorschot, W. (2007) ‘Culture and Social Policy: A Developing Field of Study’, International Journal of Social Welfare, 16 (2), 129–39.
Van Oorschot, W. (2010) ‘Public Perceptions of the Economic, Moral, Social and Migration Consequences of the Welfare State: An Empirical Analysis of Welfare State Legitimacy’, Journal of European Social Policy, 20 (1), 19–31.
Van Oorschot, W., Opielka, M. and Pfau-Effinger, B. (2008) Culture and Welfare State: Values and Social Policy in Comparative Perspective. Cheltenham: Edward Elgar.
Verba, S., Kelman, K., Orren, G. R., Miyake, I., Watanuki, J., Kabashima, I. and Ferree, G. D. (1987) Elites and the Idea of Equality: A Comparison of Japan, Sweden, and the United States. Cambridge MA: Harvard University Press.
Additional Supporting Information can be found in the online version of this article at the publisher’s website:
Table S1: Schwartz values scores for countries included in ISSP and ESS datasets.
Table S2: Summary Statistics.
Table S3: Mean egalitarianism and embeddedness scores for different types of welfare regimes and postcommunist nations*.
Table S4: Alternative random intercept models*.
Table S5: Alternative random coefficient models**.
Table S6: (a) Random intercept models using alternative measure of social values.
(b) Random coefficient models using alternative measure of social values