Purpose Mainstream Human Resource Management (HRM) research remains dominated by functionalist and positivist paradigms that treat labour as a controllable variable, while neglecting the structural, historical, and ideological conditions shaping HRM. This paper seeks to reclaim causality in HRM scholarship by introducing the Stratified HRM Analysis (SHA) Framework, grounded in Critical Realism, as a diagnostic and justice-oriented analytical approach. Design/methodology/approach This conceptual paper develops the SHA Framework by integrating Bhaskar's stratified ontology (empirical, actual, and real) with Archer's morphogenetic approach. It operationalises critical realist principles through retroductive reasoning, enabling systematic analysis linking employees' perceptions, enacted HR practices, and generative mechanisms. The framework is contrasted with dominant HRM models such as AMO, high-performance work systems, and best-practice configurations. Findings The SHA Framework shows that HRM outcomes cannot be explained through correlations between practices and performance alone. Instead, outcomes are shaped by generative mechanisms, including institutional logics, managerial ideologies, power relations, and historically sedimented inequalities. By rendering these mechanisms analytically visible, SHA explains why HRM practices generate divergent effects across contexts and how HRM systems may reproduce or challenge organizational inequality. Originality/value This paper addresses a key operational gap in critical realist HRM research. Rather than proposing a predictive model, the SHA Framework offers an ontologically grounded and causally oriented analytical device for diagnosing HRM systems as socially embedded and structurally conditioned, repositioning HRM inquiry toward explanation, reflexivity, and justice-oriented organizational transformation.
Fadillah et al. (Fri,) studied this question.