CC BY-NC-ND
176
Psychology. Journal of the Higher School of Economics. 2016. Vol.13. N 2. P. 256-272.
VALUES AND ATTITUDES TOWARDS IMMIGRANTS: CROSS-CULTURAL DIFFERENCES ACROSS 25 COUNTRIES
V.A. PONIZOVSKIYa
' National Research University Higher School of Economics, 20 Myasnitskaya Str, Moscow, 101000, Russian Federation
Abstract
Attitudes towards immigrants remain a relevant psychological outcome as they are related to prejudice, discrimination, and adaptation of migrants. Recent studies showed that basic human values could be used to explain considerable proportion of variance in attitudes towards immigrants and immigration. These studies reported that across cultural contexts the value dimension of self-transcendence is positively related to attitudes to migrants, and the value dimension of conservation is related negatively. In this paper I attempt to address some theoretical problems with universalistic interpretations of value-behavior relationships and propose a new method to identify culturally specific patterns of value-attitude relationships. Data from 25 countries collected in waves 5 and 6 of the European Social Survey (ESS) was used to assess feasibility of the proposed method. Accounting for cross-country variation significantly improved predictions of attitudes towards immigrants from individual values. Moreover, the proposed measure moderated the relationship between individual values and attitudes towards immigrants when tested against an independent data set. In line with past studies, the results indicate that overall, uni-versalism (a self-transcendent value) is the most positive predictor of attitudes towards immigrants, and security (a conservation value) is the most negative. Unlike previous studies, there is no support for universality of the pattern. A theoretical explanation for cultural variation is offered.
Keywords: attitudes towards immigrants, social values, content of values, culture, theory of basic human values, cross-cultural psychology, cognitive category, European social survey, discourse, meaning.
Introduction
In 2013, the number of forcibly displaced people worldwide has exceeded 50 million, for the first time in the post-World War II era (UNHCR, 2015).War in Ukraine and Syria, the continuing conflict in Afghanistan and
Iraq have led to increasing numbers of refugees and asylum seekers arriving in Europe. The number of permanent immigrants arriving in EU-15 countries have increased from 1.4 million in 1994 to 2.5 million in 2004 and remained relatively stable since then (OECD, 2006, 2015). At the same time
anti-immigrant attitudes remain widespread in Europe (Schneider, 2008; Semyonov, Raijman, & Gorodzeisky, 2006; Strabac & Listhaug, 2008). Moreover, it is often reported that the proportion of immigrants in a country is related to less favorable attitudes towards immigrants (Quillian, 1995; Schneider, 2008). In this light learning more about other psychological factors that influence attitudes towards immigrants is critical.
Understandably, attitudes towards immigrants remained a major focus of research in recent years. In sociology, attitudes towards immigrants were linked to socioeconomic and sociode-mographic factors such as education, income, and occupation (Coenders, Lubbers, & Scheepers, 2008; Quillian, 1995; Rustenbach, 2010). On a country level the proportion of immigrants in a country, the economic well-being of the majority population, and economic competition between the immigrants and members of the host society were used to explain cross-country variation in attitudes towards immigrants (Jackson, Brown, Brown, & Marks, 2001; Mayda, 2006; Scheepers, Gijsberts, & Coenders, 2002). Social psychologists considered numerous individual-level predictors of attitudes towards immigrant, of which perhaps the most influential were Right Wing Authoritarianism (Altemeyer, 1981), Social Dominance Orientation (Pratto, Sidanius, Stallworth, & Malle, 1994), Integrated Threat (Stephan & Stephan, 2000), and Social Identity (Tajfel & Turner, 1979).
Recently, the theory of basic human values (Schwartz, 1994) has been gaining ground as a framework for explaining formation and organization of atti-
tudes in both sociology and psychology. Human values are defined by Schwartz as trans-situational goals, desired end-states that motivate action and inform attitudes and opinions (Schwartz, 1992). The theory has been used in a number of studies of attitudes towards immigrants and cultural diversity (Davidov & Meuleman, 2012; Davidov, Meuleman, Billiet, & Schmidt, 2008; Sawyerr, Strauss, & Yan, 2005; Vala & Costa-Lopes, 2010). It was also shown that values can be seen as predictors of other relevant psychological constructs, such as Right Wing Authoritarianism, Social Dominance Orientation (Feather & McKee, 2008), and core political values (Schwartz, Capra-ra, & Vecchione, 2010). This suggests that individual differences in value priorities might play a fundamental, organizing role in the formation of political attitudes, including attitudes towards immigrants and immigration.
The findings of recent studies converge in identifying a universal pattern of relationships between values and attitudes towards immigrants, where conservation values (security, conformity, and tradition) are negatively related to acceptance of immigrants and immigration, while self-transcendence values (universalism and benevolence) are related positively across various countries and contexts (Davidov & Meuleman, 2012; Davidov et al., 2008; Vala & Costa-Lopes, 2010).
Differences in the patterns of relationships between values and attitudes are often conceptualized as differences in cultural meaning (e.g. Sagiv & Schwartz, 1995; Schwartz & Sagiv, 1995). A universal pattern of relations between values and attitudes towards immigrant would suggest that attitudes
towards immigrants express the same values in the same way across as many as 65 countries in which they were studied (Vala & Costa-Lopes, 2010). Considering the variation in socio-economic factors, sizes of immigrant populations, economic relationships between immigrant and host populations, ethnic and cultural composition of immigrants, and discourses about migration and migrants, this seems unlikely. Some recent studies did find significant cross-cultural differences in the relationship between values and attitudes towards immigrants (Da-vidov & Meuleman, 2012; Davidov et al., 2008), but despite their significance the authors urged them to "not be overrated" (Davidov & Meuleman, 2012).
In this article I attempt to address some theoretical problems with univer-salistic interpretations of cross-cultural data on the relationship between values and attitudes in general and attitudes to migrants in particular. I theoretically substantiate the use of multiple regression coefficients as proxies of cross-cultural differences in value-attitude relationships and proceed with a test of their predictive validity.
Theoretical framework
Schwartz's theory of basic human values (Schwartz, 1992) is currently the most prominent approach to the concept of values in social psychology. Synthesizing several past theories, Schwartz defines values as trans-situa-tional goals that vary in importance and serve as guiding principles in life. According to Schwartz, values motivate action, serve the interests of some social entity, act as standards for judging actions, and are acquired through
socialization into group values (Schwartz, 1994).
The motivating power of values is a critical component of the theory of basic values. Schwartz argues that values are cognitive representations of three universal human needs that have evolutionary significance: the biological needs of individuals, the need for coordinated social interaction, and the need for smooth functioning and survival of groups (Schwartz & Bilsky, 1987). These needs are translated into four higher-order motivational goals aligned on two dimensions: openness to change vs. conservation and self-enhancement vs. self-transcendence. Along these dimensions the original theory situates 10 basic values: security, conformity, tradition, benevolence, universalism, self-direction, stimulation, hedonism, achievement, and power (Schwartz, 1992). Values form a two-dimensional motivational continuum, where values that express similar motivational goals are compatible (e.g. conformity and tradition both express the higher-order motivational goal of conservation), and values that express opposing motivational goals are in conflict (e.g. conformity conflicts with self-direction). Hundreds of studies based on the theory of basic values provide strong evidence for near-universality of this conceptualization (Schwartz et al., 2012).
Schwartz argues that values motivate attitudes and behavior, and that people express values through value-consistent behavior. If a value is important for a person, they are more likely to act in a way congruent with the value: "people pursue security values by acting in a way promoting their personal safety, and they pursue hedonism
values by engaging in pleasurable activities" (Bardi & Schwartz, 2003). Further, Schwartz argues that the pattern of motivational congruities and conflicts that underlies the structure of values is also responsible for the relations among values and value-expressive behaviors, and that each value-expressive behavior is meaningfully related to all values (Bardi & Schwartz, 2003). In Schwartz's view, performing a security-congruent behavior or holding a security-congruent attitude is primarily motivated by the value of security, but also — albeit to a lesser degree — by values of tradition and conformity, since the serve a similar motivational goal, namely, conservation.
An important critique of Schwartz's interpretation of the relations between values and attitudes is that it is difficult to test whether respondents from different cultures interpret abstract values as having the same content or meaning (Peng, Nisbett, & Wong, 1997). In an attempt to address this issue Gregory Maio (2010) pointed to important differences between the functioning of values on the most abstract level — as guiding principles described by Schwartz — and the effects of values on specific behavioral choices. Maio problematized the proposition that highly abstract values directly motivate very particular behaviors, and studied the process of bridging the gap between abstract and concrete levels. He suggested thinking about values as mental representations, or cognitive hierarchies, where a Schwartzian highly abstract value is a category, and a particular attitude or a behavior is a member of the category, or an instantiation. Maio found that instantiations of values vary in the
elaboration of arguments in support of their relation to a specific value and in their typicality as a representation of a value, and that this variance has strong motivational consequences (Maio, 2010). For example, the relationship between values and attitudes strengthens when people explicitly describe their attitudes as value-expressive (Maio & Olson, 1995, 2000). Maio stressed that motivation is shared between values on the most abstract level, between values-as-categories. However, he never explicitly addressed the question whether motivation for specific instantiations of values is shared with neighboring values.
I argue that motivation is not shared on the instantiation level — specific attitudes and behaviors are not motivated by the whole "motivational continuum" of abstract values, but are motivated only through their connection to a specific, particular value — or, in Maio's terms, through being a member of a cognitive category.
From early childhood people are socialized into value-meaning of behaviors: "be kind and share your toys", "eat your vegetables and you'll grow strong", "clean up, don't be a slob". Previously meaningless behaviors acquire meaning through their connection with values. Until mother tells a child that sharing a toy is an act of kindness and reinforces it with praise, giving a toy away does not confer any advantage to the child — if anything, the child is forgoing an opportunity to play with their toy. Only after mother provides a child with discourse where sharing a toy is indicative of the child's value priorities, the behavior becomes meaningful. Sharing a toy and attitude towards sharing toys become members
of the cognitive category "benevolence". It is important to note that despite the fact that on the abstract level the value of benevolence shares some motivational power with the value of tradition, for example, the specific behavior (sharing toys) is only a member of the category "benevolence", and has no immediate connection to the category "tradition". It is also important that in a hypothetic situation where there is nobody to tell the child that sharing toys is nice, the child will probably not be able to develop an ethical argument for sharing toys on their own.
It is conceivable that the relationship between values and attitudes towards immigration gets established in a similar manner. This relationship, of course, is formed later in one's life, and it is likely that it is not the mother who provides the discourse that ties values to these attitudes, but the education system, the media, and personal communication with other people. One can encounter various discourses about immigration that can relate it to values of universalism (accepting immigrants expresses appreciation of equality, tolerance, and equal opportunities for all), security (accepting immigrants can increase crime rates and disturb social order), tradition (immigrants have no respect for our traditions and can change the traditional way of life in our society), and virtually any other value. However, it is important that discourse is a social product (Halliday, 1978), and people do not develop these discourses independently from their social environment. If previous theorizing holds, the formation of a relationship between a specific value and attitudes towards immigration is contingent on the exposure to a specific discourse.
Summarizing the theory above, I argue that a value is causally related to an attitude if, to the extent, and in the direction the two are tied through discourse. It is the discourse that enables the relationship between values and attitudes.
When bringing this theorization to the measurement level, I encounter difficulties with examining the relationship of interest using the methods used in contemporary values-attitudes research. Correlational studies of values and attitudes that are widely represented in the literature (e.g. Bardi & Schwartz, 2003; Feather & McKee, 2008; Sagiv & Schwartz, 1995) may confound two different relationships: 1) the relationship between a specific value and a specific attitude that gets established through exposure to discourse; and 2) the intercorrelations between abstract values discovered and described by Schwartz. Schwartz's theory describes motivation for holding an attitude or performing a behavior as shared between values neighboring the one it is primarily related to. In this view, sharing a toy is primarily motivated by the value of benevolence, but also, to the lesser extent, by values of universalism and tradition, because they all express higher-order motivational goals of self-transcendence and conservation. I follow Maio in arguing that motivations are only shared at the most abstract level — motivation to perform benevolent behaviors in general is shared with the motivation to perform traditional and universalistic behaviors, but performing a particular benevolence-related behavior (e.g. sharing a toy) is motivated by the value of benevolence separately from other values.
The second method that is widely used to assess the relationships between values and attitudes is regression analysis that employs higher-order value dimensions (e.g. (Davidov & Meuleman, 2012; Davidov et al., 2008; Schwartz, 2007). This approach does not allow the examination of links between specific values and attitudes since specific value scores are averaged into higher-order dimension scores. Davidov and Meuleman, for example, predict that in a European sample the value of universalism should be related to attitudes towards immigrants positively and strongly, the value of benevolence should be related positively, but weakly, and that there should be no cross-cultural differences in the pattern of relationships (Davidov et al., 2008). Since higher-order values were used for their analysis, the authors were not able to test these predictions.
I argue that using single values in a hierarchic regression could allow disentangling the two kinds of relationships described above. Such a regression could control for the intercorrelations between abstract values, and the resulting coefficients could be interpreted as culturally specific patterns of relationships between the ten values and the attitude of interest. This approach is generally discouraged, and to the best of my knowledge has never been used in studies of value-attitude relationships.
The arguments that are brought against such analyses are multi-collinearity problems arising from high correlations between specific values; and the fuzziness of borders between abstract values, where some items inevitably express elements of motivation relevant to neighboring values (Davidov et al., 2008; Schwartz et al.,
2012). However, since no multi-collinearity statistics on values were not reported in the literature, it is difficult to estimate the degree to which multicollinearity would hinder the interpretation of regression coefficients. Since I interpret single values as cognitive categories and not as representations of the higher-order motivational goals (in fact, I will be trying to control for the interrelatedness of values), the fuzziness of motivational borders on the abstract level should not affect my analysis.
Based on these theoretical considerations, I formulate the following hypotheses. Accounting for country-level differences (adding interaction terms with country-level regression slopes) in the relationship between values and attitudes towards immigrants will significantly improve the amount of variance in attitudes towards immigrants explained by the model (Hypothesis 1). Regression coefficients for countries obtained using the data at one time point will moderate the relationship between personal values and attitudes towards immigrants at a different time point in an unrelated sample (Hypothesis 2). A model that uses single values as predictors will perform significantly better than a model that uses higher-order values (Hypothesis 3). As an application of these theoretical hypotheses, I will also test a hypothesis that is based on my theorization about the differences in discourses on immigration between Western democracies and Eastern European post-communist countries. I expect that regression coefficients for universalism will be larger in Western European countries compared to Eastern European countries (Hypothesis 4).
Method
Data
To test the hypotheses, I used data from two waves of the European Social Survey (ESS), waves 5 and 6, collected in years 2010 and 2012, respectively. 25 countries that were represented in both waves of the survey were chosen for the analysis. For each of the countries, representative samples of population aged 15 years and older were collected in adherence to strict guidelines for cross-cultural social research developed by Harknes et al. (2003). The data was collected during face-to-face interviews. The following countries were included in the analysis (sample sizes for waves 5 and 6 are indicated in the parentheses): Belgium (BE) (1,704; 1,869), Bulgaria (BG) (2,434; 2,260), Switzerland (CH) (1,506; 1,493), Cyprus (CY) (1,083; 1,116), Czech Republic (CZ) (2,386; 2,009), Germany (DE) (3,031; 2,958), Denmark (DK) (1,576; 1,650), Estonia (EE) (1,793; 2,380), Spain (ES) (1,885; 1,889), Finland (FI) (1,878; 2,197), France (FR) (1,728; 1,968), United Kingdom (GB) (2,422; 2,286), Hungary (HU) (1,561; 2,014), Ireland (IE) (2,576; 2,628), Israel (IL) (2,294; 2,508), Lithuania (LT) (1,677; 2,109), Netherlands (NL) (1,829; 1,845), Norway (NO) (1,548; 1,624), Poland (PL) (1,751; 1,898), Portugal (PT) (2,150; 2,151), Russian Federation (RU) (2,595; 2,484), Sweden (SE) (1497; 1,847), Slovenia (SI) (1,403; 1,257), Slovakia (SK) (1,856; 1,847), and Ukraine (UA) (1,931; 2,178). The total size of the sample was 48,094 respondents for wave 5 and 50,465 respondents for wave 6 data.
Four sociodemographic characteristics were included: age, gender, education, and income. Education was measured as the number of years of full-time education completed, and income was measured as self-reported household total net income, with responses recorded on a 10-point scale from 1 ("1st decile") to 10 ("10th decile).
Ten basic human values were measured using the 21-item ESS human values scale. Items of the value scale are short verbal portraits of 21 persons, their life goals and priorities (e.g. "He thinks it is important that every person in the world be treated equally. He believes everyone should have equal opportunities in life"). The respondents are asked to identify how similar is the described person to themselves on a 6-point Likert-type scale, from 1 (very much like me) to 6 (not like me at all). Each of the ten Schwartz's values is measured with 2 items in the ESS, with the exception of Universalism that is measured with 3 items.
Scores for higher-order value dimensions were calculated based on single value scores. The score for conservation was calculated as the average of security, conformity, and tradition; for self-transcendence as the average of universalism and benevolence; for openness to change as the average of self-direction, stimulation, and hedonism; and for self-enhancement as the average of achievement and power.
Attitudes towards migrants were measured with 3 items from a migration module of the ESS questionnaire: whether immigrants are bad or good for the country's economy, whether country's cultural life is undermined or enriched by immigrants, and whether immigrants make the country a worse
or better place to live. The responses were recorded on a 10 point scale, where 0 corresponds to an extremely negative attitude or belief about immigrants, and 10 corresponds to an extremely positive attitude or belief. It is important to note that the word "immigrant" was not used in the wording of questions, as the meaning of that word varies across European countries. The wording used in the questionnaire is "people who come to live in [the country] from abroad" (Card, Dustmann, & Preston, 2005). The composite score for attitudes towards immigrants was computed as an average of the three items. Cronbach's alphas for all individual countries for waves 5 and 6 fell in the .77 — .91 range.
The distinction between Western and Eastern Europe was made in line with a previous study by Annabel Kuntz and colleagues (Kuntz, Davidov, Schwartz, & Schmidt, 2015). A dummy variable coded whether the country has ever been under communist rule.
Analysis
Before beginning the main analysis I tested whether multicollinearity affects the regression coefficients in question. In the selected data sets I ran a multiple regression that used ten single values as predictors of attitudes towards immigration. For both waves 5 and 6, variance inflation factor (VIF) did not exceed 1.83, and tolerance did not drop lower than .55 for any of the predictor values. These numbers fall well within the conventional guidelines of maximum VIF of 10 and minimum tolerance of .1 (Neter, Kutner, Nachtsheim, & Wasserman, 1996). Since multicolli-
nearity between values does not present a statistical problem in these datasets, I proceeded with my analysis.
I first performed a multiple regression accounting for correlations between individual values using multi-group analysis in AMOS and data from ESS wave 5. Ten single values were included as predictors, and the attitude towards immigration was the dependent variable. The results were 25 sets of 10 unstandardized coefficients, each representing the regression slope of a single value in a single country. For comparison purposes, a similar analysis was performed using scores for higherorder value dimensions.
Then, regression weights obtained in the previous step were included as country-level variables in the database for ESS wave 6. Their interactions with individual level values were computed and also included in the database.
Finally, I performed a hierarchical multiple regression using data from ESS wave 6, where the predictors were: age, gender, education, and income (step 1), individual values (step 2), country-level unstandardized regression weights (step 3), and the interaction term for individual values and corresponding country-level regression weights (step 4).
For comparison purposes, another analysis was run with the higher-order value dimensions. The predictors included were: age, gender, education, and income (step 1), individual higherorder value scores (step 2), country-level regression weights for higherorder value dimensions (step 3), and the interaction term for individual higher-order value scores and corresponding country-level regression weights (step 4).
Results
The coefficients resulting from the first regression analyses of wave 5 data can be found in Appendices A (for single values) and B (for higher-order value dimensions). Table 1 displays descriptive statistics from this analysis.
The results of the hierarchical regression that used data from wave 6 and included sociodemographic characteristics, individual-level single values, country-level coefficients obtained from wave 5, and the interaction terms to predict attitudes towards immigrants are summarized in Table 2. Sociodemographic characteristics accounted for about 6% of variance in the attitude, with all of them being significant, and education being the strongest positive predictor, in agreement with previous findings. Adding values significantly improved the prediction (up to 12% of variance), with universalism being the strongest positive predictor
of positive attitudes towards immigrants, and security being the strongest negative predictor, also supporting previously reported findings, and confirming the untested hypothesis by Davidov and colleagues (2008). Adding country-level coefficients and interaction terms further increased the amount of explained variance to 18% confirming Hypothesis 1. Seven out of 10 interaction terms contributed significantly to the prediction confirming Hypothesis 2. Interpreting individual contributions of country-level slopes, however, is hindered by the fact that they were treated by the regression as individual-level variables, resulting in unacceptable error levels.
For comparison purposes I performed a similar analysis using unstan-dardized regression coefficients for higher-order value dimensions derived from ESS wave 5 data. The results are summarized in Table 3. The first step of the hierarchical regression included
Table 1
Unstandardized regression coefficients for 10 values and attitudes towards immigrants
for 25 countries
Mean SD Minimum Maximum
Security —.27 .16 —.50 .09
Conformity —.05 .09 —.20 .18
Tradition —.12 .12 —.40 .11
Benevolence .01 .14 —.27 .24
Universalism .46 .32 —.06 .95
Self—Direction .03 .11 —.16 .27
Stimulation .04 .08 —.15 .16
Hedonism .01 .08 —.15 .13
Achievement .07 .09 —.09 .27
Power —.02 .10 —.26 .16
Note. Data was collected in ESS wave 5.
Table 2
Hierarchical regression models of the link between individual values and attitudes
towards immigrants
Model 1 Model 2 Model 3 Model 4
ß | SE ß 1 SE ß 1 SE ß 1 SE
Sociodemographic factors
Gender -.018*** .022 -.018*** .022 .000 .021 .000 .021
Age -.025*** .001 -.020*** .001 -.022*** .001 -.023*** .001
Income 087*** .004 077*** .004 .087*** .004 .087*** .004
Education .199*** .003 .155*** .003 .154*** .003 .152*** .003
Individual values
Security —170*** .013 -.134*** .013 -.068*** .026
Conformity -.018** .012 -.030*** .012 -.026*** .014
Tradition -.081*** .013 -.056*** .013 -.021** .017
Benevolence .035*** .017 -.007 .017 -.003 .017
Universalism .184*** .018 .151*** .018 .038*** .029
Self-Direction -.004 .014 -.002 .014 -.005 .014
Stimulation .011 .012 .016* .012 .002 .012
Hedonism .014* .011 -.019** .011 -.021** .011
Achievement .009 .012 .014* .012 .012 .014
Power -.075*** .012 -.025*** .012 -.015* .013
Country-level coefficients
Security slope .146*** .138 -.017 .383
Conformity slope -.146*** .159 -.144*** .501
Tradition slope -.041*** .173 -.181*** .444
Benevolence slope -.058*** .118 -.170*** .564
Universalism slope .098*** .079 -.439*** .240
Self-Direction slope -.055*** .143 - 141*** .502
Stimulation slope -.039*** .229 -.118*** .496
Hedonism slope -.022** .180 -.033 .503
Achievement slope -.104*** .171 -.123*** .435
Power slope .032*** .171 -.073*** .392
Interactions
Security interaction .166*** .074
Conformity interaction .001 .114
Tradition interaction .142*** .091
Benevolence interaction .108** .111
Universalism interaction .567*** .047
Self-Direction interaction .085*** .107
Stimulation interaction 079*** .119
Hedonism interaction .009 .117
Achievement interaction .016 .103
Power interaction .109*** .097
R2 .063 .124 .172 .180
Standard error of the estimate 2.12 2.05 1.99 1.98
*** p < .001, ** p < .01,* p < .05.
Note. All R2 changes are significant. Individual-level data was collected in ESS wave 6.
Table 3
Hierarchical regression models of the link between higher-order value scores and attitudes
towards immigrants
Model 1 Model 2 Model 3 Model 4
P 1 SE P 1 SE P 1 SE P 1 SE
Sociodemographic factors
Gender -.018*** .022 _ 027*** .022 -.015** .022 -.016** .022
Age -.025*** .001 -013* .001 -.019*** .001 -.019** .001
Income 087*** .004 079*** .004 .088*** .004 .088*** .004
Education .199*** .003 .163*** .003 .153*** .003 .154*** .003
Individual higher-order value scores
Conservation -.202*** .017 -.174*** .017 -.207*** .039
Self-Transcendence .187*** .019 .149*** .019 .096*** .035
Openness to Change .021** .016 -.003 .016 -.010 .017
Self-Enhancement -.071*** .013 -.035*** .014 -.036*** .014
Country-level coefficients
Conservation slope -.090*** .093 -.011 .344
Self-Transcendence slope .049*** .074 -.202*** .313
Openness to Change slope _ 027*** .105 -.074** .369
Self-Enhancement slope -.009 .114 -.078*** .415
Interactions
Conservation interaction .085** .074
Self-Transcendence interaction - 271*** .062
Openness to Change interaction -.046 .088
Self-Enhancement interaction - 071*** .104
R2 .063 .111 .129 .132
Standard error of the estimate 2.12 2.06 2.04 2.03
*** p < .001, ** p < .01,* p < .05.
Note. All R2 changes are significant. Individual-level data was collected in ESS wave 6.
sociodemographic characteristics and was identical to the corresponding step in previous analysis. Including individual-level scores for higher-order value dimensions improved the prediction from 6 to 11% of variance explained — an effect comparable to that of including single values. Including differences in slopes and interactions between individual-level scores and country-level slopes resulted in statistically significant increases of variance explained — to 13%. The resulting model, however, accounted for 29% less variance explained than the model based on
single values (13.2% as compared to 18%) supporting Hypothesis 3.
Finally, I performed separate regressions of single values onto attitudes towards immigrants for communist and non-communist countries using ESS wave 5 data. The unstandardized coefficient for universalism was .617 (p < .001, SE = .02) for countries with no former communist regimes, and .168 (p < .001, SE = .03) for countries that had former communist rule. The coefficients were significantly different (t = 12.45, p < .001) confirming Hypothesis 4.
Discussion
The major theoretical proposition of this paper — that values predict attitudes to immigrants if, to the extent, and in the direction in which the two are tied in a specific cultural context — was confirmed. Unstandardized regression coefficients for countries obtained at one time point did moderate the relationship between values and attitudes towards immigrant at another time point. This paper presents evidence contradicting previous findings that indicated no or minimal cross-cultural variation in the relationship attitudes towards immigrations (Davidov & Meuleman, 2012; Davidov et al., 2008; Vala & Costa-Lopes, 2010) and contributes to the growing body of literature on contextual differences in value-attitude relationships (e.g. Boer & Fischer, 2013; Grigoryan & Schwartz, 2016; Kuntz et al., 2015).
This study has direct implications for public policy. While it replicates findings from previous studies in that values can account for a sizable proportion of variance in attitudes towards immigrants, it also suggests that this relationship might not be uniform — while in some countries appeals to the value of universalism (equality, tolerance, and equal opportunity) might improve attitudes towards immigrants, in other countries such appeals may have no effect or even backfire.
Although my prediction that Western democracies will exhibit a stronger relationship between the value of universalism and attitudes towards immigrants than Eastern European
countries was confirmed, the aim of this paper was not to explain cross-cultural differences in the relationship between values and attitudes, but to advocate for a more sensitive measure of such differences, and to offer a theoretical justification for it. I interpret such cross-cultural differences as reflective of differences in value-meaning of attitudes towards immigrants, and that these value-meanings are formed through culture-specific discourse. Analysis presented in this paper does not test this theorization directly. While demonstrating predictive validity of unstan-dardized coefficients, this paper does not establish causality suggested by the theory. Further tasks include performing a multilevel analysis that would allow discerning the main effects of regression weights and the interaction effects, a cross-validation of the regression coefficients with quantitative measures of discourse, and a series of experiments that will test whether cross-cultural differences in slopes translate into differential effects of experimental manipulations of value accessibility.
Acknowledgements
The article was prepared within the framework of the Basic Research Program at the National Research University Higher School of Economics (HSE) and supported within the framework of a subsidy granted to the HSE by the Government of the Russian Federation for the implementation of the Global Competitiveness Program.
References
Altemeyer, B. (1981). Right-wing authoritarianism. Winnipeg: University of Manitoba Press.
Bardi, A., & Schwartz, S. H. (2003). Values and behavior: Strength and structure of relations. Personality and Social Psychology Bulletin, 25(10), 1207-1220.
Boer, D., & Fischer, R. (2013). How and when do personal values guide our attitudes and sociality? Explaining cross-cultural variability in attitude-value linkages. Psychological Bulletin, 139(5), 1113.
Card, D., Dustmann, C., & Preston, I. (2005). Understanding attitudes to immigration: The migration and minority module of the first European Social Survey (CReAM Discussion Paper No. 03/05). Retrieved from http://davidcard.berkeley.edu/papers/euroimmig.pdf
Coenders, M., Lubbers, M., & Scheepers, P. (2008). Support for repatriation policies of migrants comparisons across and explanations for European countries. International Journal of Comparative Sociology, 49(2-3), 175-194.
Davidov, E., & Meuleman, B. (2012). Explaining attitudes towards immigration policies in European countries: The role of human values. Journal of Ethnic and Migration Studies, 38(5), 757-775.
Davidov, E., Meuleman, B., Billiet, J., & Schmidt, P. (2008). Values and support for immigration: A cross-country comparison. European Sociological Review, 24(5), 583-599.
Feather, N. T., & McKee, I. R. (2008). Values and prejudice: Predictors of attitudes towards Australian Aborigines. Australian Journal of Psychology, 60(2), 80-90.
Grigoryan, L., & Schwartz, S. H. (2016). Values and attitudes towards cultural diversity in the Russian context. Unpublished manuscript, Department of Psychology, Higher School of Economics, Moscow, Russian Federation.
Halliday, M. A. K. (1978). Language as social semiotic. London: Edward Arnold.
Harkness, J. A., Van de Vijver, F. J. R., Mohler, P. P., & fur Umfragen, Z. (2003). Cross-cultural survey methods (Vol. 325). Hoboken, NJ: Wiley-Interscience.
Jackson, J. S., Brown, K. T., Brown, T. N., & Marks, B. (2001). Contemporary immigration policy orientations among dominant group members in Western Europe. Journal of Social Issues, 57(3), 431-456.
Kuntz, A., Davidov, E., Schwartz, S. H., & Schmidt, P. (2015). Human values, legal regulation, and approval of homosexuality in Europe: A cross-country comparison. European Journal of Social Psychology, 45(1), 120-134.
Maio, G. R. (2010). Mental representations of social values. Advances in Experimental Social Psychology, 42, 1-43.
Maio, G. R., & Olson, J. M. (1995). Relations between values, attitudes, and behavioral intentions: The moderating role of attitude function. Journal of Experimental Social Psychology, 31(3), 266-285.
Maio, G. R., & Olson, J. M. (2000). What is a value-expressive attitude. In G. R. Maio & J. M. Olson (Eds.), Why we evaluate: Functions of attitudes (pp. 249-269). Mahwah, NJ: Erlbaum.
Mayda, A. M. (2006). Who is against immigration? A cross-country investigation of individual attitudes toward immigrants. The Review of Economics and Statistics, 88(3), 510-530.
Neter, J., Kutner, M. H., Nachtsheim, C. J., & Wasserman, W. (1996). Applied linear statistical models (Vol. 4). Chicago: Irwin.
OECD. (2006). International Migration Outlook 2006. Paris: OECD.
OECD. (2015) . International Migration Outlook 2015. Paris: OECD.
Peng, K., Nisbett, R. E., & Wong, N. Y. C. (1997). Validity problems comparing values across cultures and possible solutions. Psychological Methods, 2(4), 329.
Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67(4), 741.
Quillian, L. (1995). Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review, 60(4), 586-611.
Rustenbach, E. (2010). Sources of negative attitudes toward immigrants in Europe: A multi-level analysis. International Migration Review, 44(1), 53-77.
Sagiv, L., & Schwartz, S. H. (1995). Value priorities and readiness for out-group social contact. Journal of Personality and Social Psychology, 69(3), 437.
Sawyerr, O. O., Strauss, J., & Yan, J. (2005). Individual value structure and diversity attitudes: The moderating effects of age, gender, race, and religiosity. Journal of Managerial Psychology, 20(6), 498-521.
Scheepers, P., Gijsberts, M., & Coenders, M. (2002). Ethnic exclusionism in European countries. Public opposition to civil rights for legal migrants as a response to perceived ethnic threat. European Sociological Review, 18(1), 17-34.
Schneider, S. L. (2008). Anti-immigrant attitudes in Europe: Outgroup size and perceived ethnic threat. European Sociological Review, 24(1), 53-67.
Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. Advances in Experimental Social Psychology, 25(1), 1-65.
Schwartz, S. H. (1994). Beyond individualism/collectivism: New cultural dimensions of values. In U. Kim, H. C. Triandis, Q. Kagitjiba^i, S.-C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method, and applications (pp. 85-119). Thousand Oaks, CA: Sage Publications, Inc.
Schwartz, S. H. (2007). Value orientations: Measurement, antecedents and consequences across nations. In R. Iowell, C. Roberts, R. Fitzgerald, & G. Eva (Eds.), Measuring attitudes cross-nationally: Lessons from the European Social Survey (pp. 161-193). London: Sage.
Schwartz, S. H., & Bilsky, W. (1987). Toward a theory of the universal content and structure of values: Extensions and cross-cultural replications. Journal of Personality and Social Psychology, 58(5), 878.
Schwartz, S. H., Caprara, G. V., & Vecchione, M. (2010). Basic personal values, core political values, and voting: A longitudinal analysis. Political Psychology, 31(3), 421-452.
Schwartz, S. H., Cieciuch, J., Vecchione, M., Davidov, E., Fischer, R., Beierlein, C.....Demirutku, K.
(2012). Refining the theory of basic individual values. Journal of Personality and Social Psychology, 103(4), 663.
Schwartz, S. H., & Sagiv, L. (1995). Identifying culture-specifics in the content and structure of values. Journal of Cross-Cultural Psychology, 26(1), 92-116.
Semyonov, M., Raijman, R., & Gorodzeisky, A. (2006). The rise of anti-foreigner sentiment in European societies, 1988-2000. American Sociological Review, 71(3), 426-449.
Stephan, W. G., & Stephan, C. W. (2000). An integrated threat theory of prejudice. In S. Oskamp (Ed.), Reducing rejudice and discrimination (pp. 23-45). Mahwah, NJ: Psychology Press.
Strabac, Z., & Listhaug, O. (2008). Anti-Muslim prejudice in Europe: A multilevel analysis of survey data from 30 countries. Social Science Research, 37(1), 268-286.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. The Social Psychology of Intergroup Relations, 33(47), 74.
UNHCR. (2015). Global trends: Forced displacement in 2014. Geneva: UNHCR.
Vala, J., & Costa-Lopes, R. (2010). Youth attitudes toward difference and diversity: a cross-national analysis. Analise Social, 45, 255-275.
Appendix A
Unstandardized regression coefficients for 10 values and attitudes towards immigrants (ESS wave 5)
Country SE CO TR BE UN SD ST HE AC PO
Belgium -.32 .02 -.04 -.27 .66 .05 .15 .04 -.09 .16
Bulgaria .04 -.04 -.31 .04 .06 -.13 .10 .07 .14 -.13
Switzerland -.40 -.09 -.13 .01 .46 -.02 .06 -.10 .09 .00
Cyprus -.13 .18 -.16 .12 .36 -.04 .08 .08 .13 -.06
Czech Republic -.22 -.05 .05 .12 .11 .06 .05 -.03 .09 .06
Germany -.50 -.12 -.09 .00 .89 -.02 .04 -.13 .04 .14
Denmark -.34 -.13 -.15 .02 .78 -.09 -.05 -.01 .21 .04
Estonia -.34 -.08 .00 -.06 .34 .15 .01 .12 .00 -.05
Spain -.34 -.13 -.08 -.22 .81 .09 .05 .00 .09 -.03
Finland -.29 -.04 -.19 .06 .74 .01 -.07 .04 -.02 .06
France -.50 -.20 -.09 .00 .73 .04 .03 .02 .02 .05
UK -.49 -.06 -.06 -.16 .77 -.03 .04 -.02 .27 -.06
Hungary -.13 -.04 -.06 .13 .30 .08 .05 -.03 .10 -.21
Ireland -.37 -.11 -.11 .07 .60 .14 .14 -.05 -.05 -.08
Israel .09 -.15 -.40 .14 .03 .21 -.09 -.01 .23 -.26
Lithuania -.06 -.05 .02 .14 -.01 .05 .16 .13 .01 -.05
Netherlands -.46 .04 -.09 -.15 .65 .11 -.05 .07 -.04 .10
Norway -.13 .11 -.30 -.24 .68 -.10 .06 -.15 .00 .02
Poland -.09 -.18 -.08 .01 .27 .06 .12 -.03 .13 .02
Portugal -.24 -.05 -.14 -.13 .30 .15 .11 -.06 .04 -.06
Russian Federation -.33 .11 .11 .06 -.06 -.16 .06 .08 .16 -.13
Sweden -.33 -.06 -.28 -.06 .95 .02 -.02 .00 .09 .05
Slovenia -.41 -.05 -.23 .24 .76 .05 -.15 .13 -.07 -.10
Slovakia -.24 .03 -.17 .17 .01 .27 -.02 .07 .09 -.15
Ukraine -.34 -.10 -.06 .12 .26 -.11 .10 -.09 .11 .08
Note. SE = security, CO = conformity, TR = tradition, BE = benevolence, UN = universalism, SD = self-direction, ST = stimulation, HE = hedonism, AC = achievement, PO = power.
Appendix B
Unstandardized regression coefficients for higher-order value dimensions and attitudes
towards immigrants (wave 5)
Country Conservation Self-Transcendence Openness to Change Self-Enhancement
Belgium -.29 .36 .23 -.02
Bulgaria -.36 .16 .14 -.06
Switzerland -.64 .48 -.04 .05
Cyprus -.01 .33 .16 .13
Czech Republic -.21 .26 .05 .12
Germany -.67 .91 -.13 .11
Denmark -.65 .93 -.18 .25
Estonia -.32 .21 .25 -.10
Spain -.52 .60 .14 .03
Finland -.47 .77 -.07 .01
France -.82 .74 .08 .00
UK -.58 .67 .03 .18
Hungary -.23 .46 .19 -.16
Ireland -.57 .66 .19 -.16
Israel -.58 .49 .00 -.11
Lithuania -.16 .18 .34 -.03
Netherlands -.46 .46 .10 -.02
Norway -.38 .55 -.23 .01
Poland -.36 .32 .13 .15
Portugal -.38 .11 .22 -.03
Russian Federation -.02 -.08 .10 -.05
Sweden -.66 .92 -.09 .13
Slovenia -.62 .96 .01 -.20
Slovakia -.37 .25 .26 -.04
Ukraine -.39 .26 -.03 .13
Vladimir A. Ponizovskiy — M.A., National Research University Higher
School of Economics.
E-mail: vladimir.ponizovsky@gmail.com
Ценности и установки по отношению к иммигрантам: кросс-культурные
различия в 25 странах
В.А. Понизовский"
' Национальный исследовательский университет «Высшая школа экономики», 101000, Россия, Москва, ул. Мясницкая, д. 20
Резюме
Поскольку установки по отношению к иммигрантам связаны с дискриминацией, предрассудками и успешностью адаптации мигрантов, исследование предикторов этих установок остается актуальной темой. Недавние исследования показали, что базовые индивидуальные ценности объясняют значительную часть дисперсии в установках по отношению к иммиграции и иммигрантам. Эти исследования демонстрируют, что в различных культурных контекстах ценности самопреодоления положительно связаны с установками по отношению к иммигрантам, в то время как ценности сохранения связаны отрицательно. В данной статье обсуждаются теоретические проблемы в универсалистичных интерпретациях связей между ценностями и установками. На данных из 25 стран, полученных в ходе Европейского социального исследования, была проведена проверка эффективности нового метода учета кросс-культурных различий. Статистический контроль кросс-культурных различий значительно улучшает предсказание установок по отношению к иммигрантам. Более того, в анализе независимой базы данных предлагаемая переменная модерировала связь между базовыми ценностями и установкой. Подтверждая ранее полученные результаты, универсализм (одна из ценностей самопреодоления) была самым сильным положительным предиктором установок по отношению к иммигрантам, а безопасность (одна из ценностей сохранения) — самым сильным отрицательным предиктором. В отличие от предыдущих исследований, универсальность отношений между ценностями и аттитюдами не подтвердилась. Предлагается теоретическое обоснование кросс-культурных различий.
Ключевые слова: установки по отношению к иммигрантам, социальные ценности, содержание ценностей, культура, теория базовых индивидуальных ценностей, кросс-культурная психология, когнитивные категории, европейское социальное исследование, дискурс, смысл.
Понизовский Владимир Антонович — магистр, Национальный исследовательский университет «Высшая школа экономики». Контакты: vladimir.ponizovsky@gmail.com
