Tuesday 16th July
Wednesday 17th July
Thursday 18th July
Friday 19th July
Download the conference book
Methodological challenges in the study of attitudes toward immigration |
|
| Convenor | Dr Oriane Sarrasin (Social Science Research Center Berlin) |
| Coordinator 1 | Dr Eva G. T. Green (University of Lausanne) |
The increasingly diverse flows of immigrants to Western countries during the last decades have provoked fierce societal debates. Based on data from international social surveys, a large body of research has sought to understand these debates by examining the attitudes of the native population toward various aspects of immigration (e.g., toward specific groups of immigrants, toward immigration policies). Despite its theoretical and societal relevance, the study of immigration attitudes encounters many methodological challenges prone to hamper substantive conclusions. First, unbiased comparisons of attitudes are generally known to require invariant measurement of scales across contextual units or groups. Because immigration attitudes involve a wide range of societal and cultural aspects, the understanding of questionnaire items is likely to differ across countries or groups. However, the extent to which this biases substantive findings remains unknown, as measurement invariance of immigration attitudes is rarely tested. Second, the use of data collected in different contextual units raises questions, such as which unit level (e.g., neighbourhood, region or country) is relevant for investigating the contextual determinants of immigration attitudes or to which extent attitudes in a given contextual unit depend on dynamics in neighbouring units. Yet another central methodological issue pertains to the criteria defining the populations whose attitudes toward immigration are studied. Indeed, many criteria, which differ in their level of strictness, can be relied on to define who belongs to the local population (e.g., being born in the country, having citizenship and/or ancestry), making the definition of group boundaries arbitrary. Using concrete examples, the present session aims at illustrating various methodological challenges--as outlined above but not limited to them--researchers face when studying attitudes toward immigration as well as the solutions they apply.
This paper analyzes upon the relation between immigration-related threat perceptions and the attitudes towards the integration of immigrants among natives in Luxembourg - the country with the highest proportion of immigrants in Europe. Furthermore, it investigates how this relationship is influenced by the intensity of friendship contacts with immigrants. Attitudes towards integration encompass two opposite integration dimensions and are measured by two composite scales: attitudes towards assimilation and attitudes towards multiculturalism. Also a composite measure of perceived group level threats is used. The European Value Study (EVS) from 2008 for Luxembourg using only the native residents and Structural Equation Modelling (SEM) analysis techniques are used to conduct the analysis.
The outcomes of our analyses reveal that general feelings of threat lead to less support for multicultural attitudes, whereas the opposite can be found for suppport for assimilation attitudes. Furthermore, it was found that more intense contact with immigrant friends reduces significantly the support for assimilation attitudes. However no effect of friendship contact was found for the threat perceptions and multicultural attitudes. Policy implications of these results are discussed.
Most research on immigration attitudes using large-scale survey data routinely excludes respondents with a migration background. However, the rationale for discarding the opinions of a share of the host society population is rarely discussed, and little is known as to whether including respondents' migration background does actually bias findings. The present study fills this gap by illustrating with multigroup analysis on two Swiss samples how to evaluate whether respondents' migration background affects both the measurement and the prediction of immigration attitudes. In our analysis, we compared attitudes of natives to those of respondents with a migration background (World Values Survey) and of two specific immigrant groups (longer established and culturally close vs. recent and culturally more distant immigrants; Monitoring Misanthropy and Rightwing Attitudes survey). In both samples measurement of immigration attitudes was highly similar across groups (i.e., partial scalar invariance was reached). These results indicate that attitudes can be compared or pooled without measurement biases. Further analyses revealed slight differences in means and predictions of immigration attitudes. Groups held similar stances toward immigration, with the exception of the culturally more distant immigrants expressing more positive attitudes. Moreover, attachment to Switzerland was found to predict immigration attitudes across all groups, albeit more strongly among natives. These finding indicate that although migration background should be carefully controlled for, satisfactory levels of invariance and similar means and predictions call for the inclusion of respondents with migration background.
Social trust is understood in social sciences as a key component for social cohesion, economic growth and political development of a society, as it is the “glue of social life”. Immigration caused by the economic miracles after World War II and by refugees from crisis areas generated a growing religious an ethnic divsersity in Western societies. The plurality of origins, denominations and cultures that increasingly form part of our society is frequently assumed to undermine social trust in current modern societies (Putnam). The purpose of this paper is to decompose for the first time empirically the impact of different factors that are assumed to be determinants of trust and distrust by using a factorial survey. This experimantal design carried out among students from Bilbao and Cologne consists in judging variing descriptions (vignettes) of fictitious persons acting in an from the researcher in advance defined situation. By using mulilevel analyses the implact of the described characters of the ficitious persons as well as the impact of respondent characteristics on trust can be estimated simultaneously. Besides that it also allows to answer the question of whether trust is higher a) among people that share the same characteristics and b) in denominational comparably more mixed Cologne than in denominatinal rather homogenious Bilbao.
That discrimination matters for individuals' life quality and that migrants are often targets of discrimination are truisms. But how much it matters, is less obvious. The paper charts some little traveled waters. It brings together insights from poverty and discrimination research in order to understand whether, how and to what extent the level of perceived discrimination in society affects the risk of in-work poverty of migrants in Germany. It proposes to answer the challenge of measuring discrimination against immigrants by using an aggregate measure of discrimination at each point in time in our observation period as it is experienced by a representative sample of migrants. Using longitudinal data provided by German Socio-Economic Panel Study (SOEP), a significant effect of discrimination on in-work poverty risks of second generation migrants is found.