ESRA 2019 Draft Programme at a Glance


Between generalisation and specificity: attitudes towards immigrants and ethnic minority groups 2

Session Organiser Dr Aneta Piekut (Sheffield Methods Institute, University of Sheffield)
TimeFriday 19th July, 13:00 - 14:00
Room D24

Despite the fact that many surveys ask about various categories of immigrants, attitudes towards immigrant targets are often analysed jointly in academic studies, after aggregating a few measures into a composite index or through modelling a common latent variable. This might help in improving construct validity, but might also pose challenges in data interpretation. Immigrant groups and various minority ethnicities are not homogenous and differ in terms of time of arrival, origin, dominant socio-demographic profile, occupations and public discourses towards them, hence they can mobilise differential attitudes. As such, by using generalisation procedures we might be losing quite a lot of valuable information about what the public thinks. At the same time, some public opinion polls only ask about opinions towards a broad category of ‘immigrants’, which brings a question of measure reliability, since respondents may be thinking about very different immigrant or ethnic groups, while answering this kind of questions.

In this session, we would like to explore diversity in opinions towards immigrants and minority ethnic groups. This session will specifically discuss whether the same theoretical mechanisms commonly applied in studies on attitudes towards a general category of immigrants, like contact hypothesis, integrated threat theory etc., work in the same way in case of all immigrant groups. We welcome papers unpacking differences in attitudes towards immigrant groups differing in ethnicity, race, age, gender, socio-economic status, religion or region of origin.

Keywords: attitudes towards immigrants, construct validity, measurement reliability

The Differential Effects of Symbolic and Realistic Threat on Ethnic Prejudice toward Minority Groups in Germany

Mr Marcus Eisentraut (GESIS - Leibniz Institute for Social Sciences ) - Presenting Author
Dr Alexander Jedinger (GESIS - Leibniz Institute for Social Sciences )

Past research often focused on prejudice only as a general characteristic of ingroup people, which limits the understanding of the genesis of prejudice toward specific outgroups. In the present study, we examine the relative impact of symbolic compared to realistic threat perceptions on hostility toward different minority groups in Germany (Muslims, refugees, Sinti and Roma, foreigners). Based on recent theorizing, we argue that the primacy of symbolic and realistic threat in explaining prejudice is outgroup-specific because various minority groups challenge the economic, cultural and security-related interests of the majority group in different ways.

To test our hypotheses, we use data from two waves of the GESIS panel which draws from representative samples of the adult population in Germany. The results show that realistic threat plays a more important role in shaping negative attitudes toward Sinti and Roma, a minority group which the majority population often associates with laziness and the fraudulent acquisition of welfare benefits. In contrast, symbolic threat is more important in predicting prejudices against Muslims whose group stereotypes include the rejection of Western and democratic values. Attitudes toward refugees and foreigners, however, are equally strong related to realistic and symbolic threat.

Overall, we show that threat perceptions are differentially related across minority groups in Germany. Our results question the exclusive focus on a “general prejudice factor” approach and support a group-specific perspective on ethnic prejudices.


The role of contextual factors on the effect of contact on generalized and specific prejudices towards national minorities: a cross-national approach

Mr Tomislav Pavlović (Institute of social sciences Ivo Pilar) - Presenting Author
Professor Renata Franc (Institute of social sciences Ivo Pilar)

Contact hypothesis refers to human tendency to exhibit less negative prejudices (both generalized and specific) towards members of specific social out-groups they spend time with, which was generally confirmed by many studies. However, the question of relative predictive importance of contact in explaining generalized and specific prejudice, as well as role of contextual macro-level factors, is not so frequently investigated. This study: (1) tested the significance of contribution of contact to the explanation of generalized as well as specific prejudices towards national minorities (Roma, Jews, Muslims) over and above other common predictors, (2) tested the differences in predictiveness of contact in the context of specific macro-level indicators of inter-group equality and wellbeing (operationalized by national homogeneity, Human Development Index, GINI index, risk of violating human rights and socialist historical legacy), (3) discussed differences between linear and interactive relationship of contact as a determinant of generalized prejudice and prejudice towards specific national minorities. Multi-level models were applied on the survey data collected within EU FP7 MyPlace project (16935 participants aged 15-25, from 14 European countries, 50.2% females), joined by macro-level data collected from most recent national censuses and publicly available databases. The results suggested that (1) more contact is related to less (generalized and specific) prejudices, even in the context of other relevant individual variables, (2) socio-cultural context plays an important role in determining the extent of prejudices and effectiveness of contact, (3) linear and interactive effects of contact may vary with respect to targeted prejudice. Although this study did not provide any causal inferences, it indicated usefulness of specific prejudice measures and relevance of macro-level context for understanding effect of contact on generalized and specific prejudice towards national minorities.


Does complete measurement equivalence between groups and over time mean that measured concepts have the same meaning in populations?

Professor Jaak Billiet (CeSO, KU Leuven) - Presenting Author
Dr Cecil Meeus (Erasmus University Rotterdam)
Professor Bart Meuleman (Leuven University)

ABSTRACT: Longitudinal analysis of an invariantly measured latent variable “perceived ethnic threat” (PET) shows that averages of PET fluctuate over seventeen general surveys in Flanders (Belgium) in the 1991-2016 period. PET was found to be significantly lower in 2014 and 2016 than in each of the previous surveys. The highest score of PET, indicating the strongest negative attitude toward immigrants, was observed in 2010, the aftermath of the economic-financial crisis. The largest drop in PET was observed between the end of 2010 and 2014, and decreased still somewhat in 2016. This was unexpected since the 2014 data collection period started some months after the outset of the European refugee crises (Summer 2014), and in 2016 the data were collected after the terrorist attacks in Paris (November 13, 2015) and Brussels (March 22, 2016).
The unexpected findings indicating a drop in PET in two independent surveys were an occasion to analyse the trends in the four observed indicators measuring PET. The two items that expressed cultural (symbolic) threat did not evolve the same way as the two other (realistic) threat items. Therefore, two latent variables were specified. Analysis showed that the trends of the two latent variables (cultural and realistic threat) evolved indeed quite differently after 2003. Compared with the latent scores of realistic threat, the latent scores of cultural threat increased steady and faster until 2010. These findings suggest that measured concepts that pass statistical tests for measurement invariance can nevertheless hide important substantive nuances. This unexpected finding is checked against the trends found in the cultural and welfare threat items in ESS data from Belgium and its neighbouring countries (BE, DE, FR, GB, NL) in 2002-2017. We discuss what can be learned from both a substantive and a methodological point of view.