Perceived discrimination among migrants in Germany: Does social capital moderate harmful effects on mental health?

• Published in: Social science & medicine (1982) Vol. 370; p. 117854
• Main Authors: Biddle, Louise, Stacherl, Barbara, Heidinger, Ellen
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.04.2025
• Subjects: Adult; Cross-Sectional Studies; Discrimination; Female; Germany; Humans; Male; Mental health; Mental Health - statistics & numerical data; Middle Aged; Migration; Perception; Refugees; Refugees - psychology; Refugees - statistics & numerical data; Social Capital; Transients and Migrants - psychology; Transients and Migrants - statistics & numerical data; Germany; Discrimination; Refugees; Migration; Social capital; Mental health
• ISSN: 0277-9536; 1873-5347; 1873-5347
• DOI: 10.1016/j.socscimed.2025.117854

The harmful mental health effects of perceived discrimination for migrant populations are well established. The potential buffering effect of regional-level social capital, however, has not previously been explored. Using data from the German Socio-Economic Panel (SOEP; 2009–2018) we apply multilevel models to assess the effect of frequent or infrequent perceived discrimination on mental health (MCS subscale of the SF-12) adjusting for individual- and regional-level confounding in a cross-sectional research design. We assess interaction with general social capital (civic engagement, electoral participation, generalised trust) and migrant-specific resources (proportion of non-nationals). We juxtapose non-refugee migrants (n = 13478) with refugees subject to mobility restrictions (n = 5558) to account for bias introduced by selective mobility into regions. In the non-refugee sample, we confirm the negative effects of discrimination experiences on MCS (ß frequent: −3.74, 95%-CI: [−4.40; −3.09]/ß infrequent: −1.88, 95%-CI: [−2.24; −1.52]). Moderation analyses among non-refugees show no buffering of general social capital, but a negative interaction effect of the proportion of non-nationals with experiences of discrimination (ß frequent: −0.54, 95%-CI: [−0.87; −0.21]/ß infrequent: −0.68, 95%-CI: [−1.29; −0.07]). Analyses among refugees also demonstrate a negative effect of discrimination on MCS with stronger overall effects than in the migrant sample (ß frequent: −6.24, 95%-CI: [−7.44; −5.05]/ß infrequent: −3.56, 95%-CI: [−4.25; −2.86]). In the refugee sample, the effect is exacerbated by the proportion of non-nationals (ß frequent: −1.70, 95%-CI: [−2.84; −0.56]) and buffered by generalised trust (ß: 0.87, 95%-CI: [0.13; 1.60]). When faced with experiences of discrimination, levels of community trust seem to buffer the negative mental health impacts for refugees. We do not find buffering effects of social capital on the mental health of non-refugee migrants, possibly due to established social networks. Furthermore, our results suggest that the proportion of non-nationals within a community may have a negative reinforcing impact on the relationship between discrimination and mental health for both migrant samples. As this dynamic is currently underexplored in the German context, more community-based research is needed to develop appropriate policy interventions. •Negative effect of discrimination on refugee and non-refugee migrant mental health.•Generalised trust shows positive buffering in refugee, but not in migrant sample.•Proportion of non-citizens in area has negative buffering effect in both samples.