Geopolitical Shocks, Perceived Discrimination, and Immigrant Political Attitudes: A Moderated Mediation Study of Migrants in Russia

Abstract Drawing on intergroup‑threat and social‑identity theories, this study analyses survey data from 212 Central Asian male labour migrants residing in Russia in the post‑2022 Russia‑Ukraine‑war context. It explores how perceived geopolitical shocks are associated with migrants’ perceived discrimination, political attitudes towards the Russian regime, and their subjective social identification with host‑country society. A moderated‑mediation analytical framework was constructed and tested via SPSS and the PROCESS macro. The empirical results demonstrate that perceived discrimination exerts a statistically full mediating effect in the link between perceived geopolitical shocks and both pro‑regime political attitudes and Russian‑host‑society identity. Specifically, migrants who report higher levels of perceived discrimination tend to express more positive evaluations of the Russian political regime and display stronger subjective belonging to Russian society. Russian‑language fluency marginally amplifies this indirect mediating pathway ( p = .061), whereas duration of residence in Russia yields no statistically significant moderating influence. Four mutually non‑exclusive interpretive frameworks are deployed to unpack this counter‑intuitive empirical pattern: threat‑triggered intragroup cohesion, realignment of social‑comparison reference frames, social‑desirability response bias embedded within authoritarian survey contexts, and survivorship bias generated by selective out‑migration of dissatisfied migrants. Important limitations must be emphasized: cross‑sectional research design, low internal reliability of the exploratory geopolitical‑shock measurement scale (α = .58), and high sample homogeneity (92% male respondents, 92% of Central‑Asian ethnic origin). Therefore, all empirical outputs should be treated as hypothesis‑generating rather than conclusive causal evidence. Reverse causality and unmeasured confounding variables cannot be statistically ruled out in the current analytical setup. This study stresses the necessity of refined scale development and longitudinal research designs to clarify whether observed correlations reflect genuine psychological processes whereby discrimination paradoxically reinforces regime support, or alternatively stem from measurement artifacts and contextual response biases.