Assistant Professor & Ex. Co-Chairman, School of Management, Huazhong University of Science and Technology (HUST), 1037 Luoyu Road, Wuhan, Hubei 430074, P. R. China.
Department of Business, Administration Faculty of Business and Economic
Integrating CSR Practices for Sustainable Business Performance: A Three-Stage Data-Driven SEM-fsQCA-ML Approach
Purpose: Despite the increasing integration of CSR with Sustainable Development Goals (SDGs), the relationship between CSR dimensions and business sustainability remains underexplored, particularly in developing economies. Moreover, the ultimate aim of this study is to examine the multidimensional impact of philanthropic, environmental, social, ethical, legal, and economic—Corporate Social Responsibility (CSR) practices on Sustainable Business Performance (SBP) in manufacturing small and medium-sized enterprises (SMEs) within Bangladesh's manufacturing sector, utilizing a three-stage data-driven approach combining Partial Least Squares Structural Equation Modeling (PLS-SEM), Fuzzy-Set Qualitative Comparative Analysis (fsQCA), and Machine Learning (ML) to uncover predictive patterns, causal pathways, and strategic CSR configurations for long-term sustainability.
Design/methodology/Approach: Data was collected from 510 manufacturing SMEs across six industrial sectors in Dhaka and Chattogram. The study employed PLS-SEM to assess the symmetric relationships between CSR dimensions and SBP, fsQCA to identify asymmetric causal pathways, and ML algorithms (Random Forest, AdaBoost, and XGBoost) to predict CSR’s impact on SBP. Feature selection methods (ANOVA, F-Score) were applied to optimize model accuracy, and Sobol’s sensitivity analysis identified the most influential CSR dimensions driving SBP.
Findings: The results confirm that economic (β = 0.513, p = 0.002), environmental (β = 0.419, p = 0.013), and ethical (β = 0.401, p = 0.001) CSR dimensions have the strongest positive effects on SBP. Legal compliance and philanthropic engagement also significantly enhance sustainability outcomes, while social CSR has a moderate impact. fsQCA analysis reveals that multiple configurations lead to high SBP, underscoring the role of CSR as a strategic enabler rather than a singular determinant. Machine Learning models further validate the predictive power of CSR dimensions, with environmental CSR emerging as the most influential factor (normalized importance = 1.000).
Practical & Social Implications: This research offers practical insights for SME owners, policymakers, and stakeholders by highlighting the strategic value of CSR in enhancing financial, social, and environmental sustainability. Policymakers should strengthen regulatory frameworks and incentives to encourage SMEs to adopt structured CSR strategies. Businesses should integrate CSR beyond compliance, treating it as a long-term investment in sustainability rather than a cost.
Originality/Value: This study spreads CSR research by integrating PLS-SEM, fsQCA, and ML into a unified analytical framework, offering both predictive and explanatory insights into CSR’s role in sustainable business performance. The study also underscores the role of ML-based CSR analytics in helping businesses optimize CSR strategies by prioritizing high-impact sustainability initiatives.
Keywords: Corporate Social Responsibility (CSR), CSR Practices, SEM-fsQCA-ML Approach, Small and Medium-sized Enterprise (SMEs), Sustainable Business Performance (SBP).