Board gender diversity heterogeneity and sustainability. A systematic literature review on multi-dimensional board diversity

Purpose – This systematic review examines how within-gender heterogeneity in board gender diversity, variation in female directors' observable and underlying attributes, relates to corporate sustainability outcomes, captured as ESG performance and ESG reporting. Design/methodology/approach – Guided by stakeholder–agency theory and an adapted Milliken and Martins' (1996) framework, we systematically searched EBSCO Business Source Complete and Web of Science. We included 70 English-language, peer-reviewed archival studies that operationalise board gender diversity through multiple female-directors attributes and report associations with ESG performance and ESG reporting. We synthesise findings using vote counting and report simple summary statistics for the most frequently studied attributes. Findings – Evidence is most consistent for female non-executive directors, who are robustly associated with stronger ESG performance across governance systems and regulatory settings. Results for female executive directors are more context-dependent and frequently mixed. For other heterogeneity dimensions, the evidence base remains comparatively thin and inconclusive, particularly for ESG reporting, reflecting both limited study volume and measurement heterogeneity. Practical implications – Boards and regulators should treat gender diversity as a multi-attributed governance resource rather than a headcount metric. Emphasis on the composition and authority of female directors appears most consequential for sustainability performance, while disclosure-related effects require more consistent measurement and stronger identification. Originality/value – This review's contribution lies in systematically operationalising and coding within-gender heterogeneity by separating observable from underlying attributes and mapping them distinctly to ESG performance versus ESG reporting, thereby clarifying where evidence is comparatively consistent and where key gaps remain.