Under what conditions are voters likely to invoke gender-trait stereotypes and ethnic biases when they evaluate the performance of local political leaders? We test arguments from role congruity and social identity theories and employ intersectional analysis to explore voter reliance on gender-trait stereotypes and ethnic biases across four diverse Russian regions. We find that the structure of the regional economy matters for the reduction of gender-trait stereotypes, while spatial and labor market integration matters for the dissipation of ethnic favoritism in leader evaluation. Our findings have important implications for policy makers aiming to reduce political biases in diverse contexts. On the one hand, we encourage scholars to pay careful attention to stereotypes of underrepresented social groups and evaluate how they can translate into understanding the traits associated with good leadership. On the other hand, we demonstrate that the roots of social biases stem from complex structural phenomena, such as limited labor opportunities for women and minorities, and thus require transformative political and economic changes.