https://orcid.org/0000-0003-3008-9686
![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
Immigration has become an increasingly politicised issue in many European countries. While prior research emphasises the role of party cues and education in shaping public attitudes to immigration, we know less about the impact of these two factors when they place individuals under cross-pressures. This study explores how exposure to cross-pressures arising from the incongruence between party cues and the values one is expected to hold as a function of educational attainment affects immigration attitudes. The empirical analysis relies on cross-sectional analyses of public preferences in ten European countries and longitudinal analyses of public attitudes in the Netherlands. The results point to attitudinal moderation among party supporters with lower levels of education due to exposure to incongruent party cues. Highly educated party supporters appear to be insensitive to such incongruent cues.
Supplemental data for this article can be accessed online at: https://doi.org/10.1080/01402382.2021.1975447 .
Acknowledgements
Earlier versions of this paper were presented at the 2019 EPSA and ECPR Conferences and seminars at the European University Institute. The author would also like to thank the two anonymous reviewers for their helpful comments.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 The authors maintain that self-identified moderates are actually a mix of policy centrists and cross-pressured individuals, the latter not always being policy centrists.
2 A relevant question is what these strands of literature can tell us about how partisans react to congruent cues. The expectation regarding this scenario is less straightforward. On the one hand, prior work on the role of core values (e.g. Evans and Neundorf Citation2020) might suggest that attitudes are impervious to all cues, including the congruent ones. On the other hand, the above-mentioned research on social identities might suggest that reinforcing identities could lead to more extreme attitudes.
3 Data available online at: https://www.europeansocialsurvey.org/.
4 The percentages shown in Figure 1 correspond to column B in Table A14 in the Online Appendices. This table offers additional details regarding the differences between the initial country samples included in ESS (for the ten countries covered by this study) and the final sample used to estimate the main models in this study (see also main text).
5 Another potential issue is that experts could be influenced by public opinion. However, this should not be a major concern for this study, as it focuses on individual-level dynamics.
6 The final sample includes 26 negative shifts, 37 positive shifts and 2 instances where parties do not change their stance. The latter have been recoded as pro-immigrant cues since the party position is above the sample average.
7 Ideally, the analysis would incorporate also respondents who feel close to a given party but had voted for a different one in the past election. However, in that case it would be difficult to differentiate between instances where the voter merely moved to a party with a closer issue position or whether the voter’s attitudes are (also) influenced by the new party’s cues, as data on prior attitudes is not available. With these caveats in mind, several exploratory analyses are reported in the ‘Robustness checks’ section.
8 Table A4 in the Online Appendices provides information regarding shares of party switchers and non-switchers/stayers (see that table for details).
9 Social class is a categorical variable with eight groups. It has been constructed based on the syntax files provided by Daniel Oesch at: http://people.unil.ch/danieloesch/scripts/.
10 For the least educated, the difference between the mean attitudes of cross-pressured and non-cross-pressured individuals is 0.14 (p-value = 0.12). For highly educated, the difference is 0.10 (p-value = 0.29).
11 Gidron (Citation2020) uses these cut-off points to capture attitudinal consistence.
12 The data corresponds to the year 2010 (see the Online Appendices for further details).
13 LISS is a panel administered by CentERdata (Tilburg University, The Netherlands). More information about the LISS panel can be found at www.lissdata.nl.
14 These are: (1) it is good if society consists of people from different cultures; (2) legally residing foreigners should be entitled to the same social security as Dutch citizens; (3) there are too many people of foreign origin or descent in the Netherlands (reversed); and (4) it does not help a neighbourhood if many people of foreign origin or descent move in (reversed).
15 Table B5 in the Online Appendices reports results from a linear regression model of current attitudes on prior attitudes and partisanship, suggesting that both factors are strong predictors.
16 Figure 4 also indicates that, while the attitudes of highly educated supporters are on average more positive than those of the poorly educated exposed to similar party cues, within the high education group attitudes appear to be more positive in the scenario of restrictive party shifts (after accounting for prior attitudes). This difference is most discernible when party cues are weak (low magnitude) but disappears when party cues are strong. One explanation might be that it is more difficult for voters to be aware of small changes in party positions and to interpret them consistently.
Additional information
Funding
Notes on contributors
Alina Vrânceanu
Alina Vrânceanu is a Research Fellow at the European University Institute (Florence, Italy), Migration Policy Centre. Her research interests include political behaviour, public opinion, party competition and immigration politics. Her work has been published in Party Politics and the Journal of Elections, Public Opinion and Parties. [[email protected]]