Understanding differences in attitudes to immigration: a meta-analysis of individual-level factors

Introduction

Immigration is among voters’ top concerns when asked about the main challenges for their country or other political entities such as the European Union (e.g.Commission, 2019a,b). Public attitudes toward immigration are becoming part of a new political cleavage (Kriesi etal., 2012;Hobolt, 2016), particularly in the after math of the so-called “migration crisis”. Consequently, explaining the reasons for individual differences in attitudes to immigration has attracted increased scholarly interest. Various hypotheses regarding factors affecting attitudes to immigration have been proposed, resulting in often highly correlated determinants, making it difficult to assess which of these are truly relevant. In this article, weconduct a meta-study asking which individual indicators are consistently found to influence attitudes to immigration within the broad social science literature. Meta-analyses are essential for formally structuring and summarising the scholarly state-of-the-art on a topic. They also play a crucial role in explaining the origins of the heterogeneity of research results to academics who are non-experts in the field, policymakers, and practitioners. Our paper complements the influential review papers on attitudes towards migration (CeobanuandEscandell, 2010 ; Hainmueller and Hopkins, 2014; Mayda, 2006) by providing a quantitative meta-analytical overview. Moreover, it is also worth highlighting that these reviews were done several years ago and a lot of new insights have emerged from the literature since then.

We systematise the knowledge regarding attitudes to immigration across various social science fields and cover the thirty top-ranked journals for each discipline across economics, political science, sociology, psychology, and migration/ethnic/demographic studies published between 2009 and 2019. From these, we select all 350 articles that quantitatively analyse the determinants of attitudes to immigration. After dropping all articles covering attitudes to immigration out of scope, we evaluate information from 140 academic articles and 1185 estimates in total. We thus provide an encompassing review of the research regarding attitudes to immigration published across different social science fields during the past decade.

Out of the 150 different types of attitudes to immigration that we have encountered in the literature, we focus on the two groups of dependent variables that are the most relevant and the two most commonly surveyed, which capture preferences towards migration policy (e.g., preferred levels of immigration) and views about immigrants’ contribution to society. These two dependent variables complement eachother as they measure two concepts-preferences regarding levels of immigration and opinions regarding the effect of immigration. We identify the relevant factors affecting attitudes to immigration based on statistically representative samples from all over the world.

Our approach focuses on eight individual-level indicators, namely age, gender, education, income, occupational and unemployment status, as well as respondents’ minority background and the type of area (urban versus rural) they live in. When selecting individual independent variables, we followed other reviews of public opinion on migration and only focused on the widely used determinants (see Ceobanu and Escandell (2010); Hainmueller and Hopkins (2014); Drazanova(2022)) rather than understudied factors such as, for example, disgust sensitivity (Aarøe, Bang Petersen and Arceneaux, 2017). After this initial pre-selection, we further reduced the number of independent variables, as some (for example subjective economic well being, political affiliation, etc.) had too few observations for a meta-analysis. Our study’s main objective is to assess recent empirical evidence on which of these individual-level factors are consistently (positively or negatively) linked with attitudes to immigration.

Read the full working paper.