Contact in context: does intergroup contact function (better) in high-threat contexts?

ABSTRACT

Are interpersonal contacts effective in reducing group threat and prejudice in all contexts? Studies have found that the contact effect is stronger in regions and countries where large immigrant populations contribute to high levels of contextual group threat This paper tests the robustness of the observed positive association between contextual-level group threat and contact effects by identifying three methodological reasons for why the association might be spurious: (1) a potential omitted variable bias; (2) a potential self-selection effect; and (3) a potential ceiling effect. Using the 2002 European Social Survey, we find that the positive association is indeed robust. We interpret this as evidence that contact effects are strongest in those European contexts where the need for improved intergroup relations between immigrants and natives is most dire.

KEYWORDS:

• Contact hypothesis
• prejudice
• group threat
• intergroup relations
• European Social Survey
• multilevel modelling

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. ‘Contextual-level group threat’ is throughout the paper defined as group threat that is caused by contextual-level factors, such as high immigration levels and resource scarcity. It should be contrasted against group threat caused by individual-level factors, such as education and income. The distinction should not be confused with the distinction between ‘group threat’ and ‘individual threat’ that is common in the literature, and where the former term denotes that someone feels that a group to which he or she belongs is being threatened by an out-group while the latter term denotes that someone feels that an out-group constitutes a threat to themself or someone close to them (e.g. Rosenstein 2008). The terms we use both refer to the former kind of threat (i.e. group threat), but discriminate between different factors that cause the threat (contextual-level factors and individual-level factors). For example, low-educated individuals may be more prone than high-educated individuals to feel that their ethnic group is being threatened by immigrants, even though they do not personally feel more threatened by immigrants.

2. It should be pointed out that our dependent variable measures perceived group threat and not actual group threat.

3. Here, the expression ‘contact effects' refers to the association between contacts and group threat in our statistical models and should not be interpreted in a causal way. We have chosen to speak of ‘effects' rather than ‘associations' because it makes the text more readable.

Additional information

Funding

The authors gratefully acknowledge the financial support granted by the Swedish Research Council [grant number 2010–1469].