Employing Machine Learning to Deduce a Causal Link Between Corporate Social Responsibility and Financial Performance

Abstract

Incorporating Corporate Social Responsibility (CSR) into business strategy has become noteworthy, as that pertains to the discretionary initiatives a business entity undertakes to enhance its operational sphere’s societal and ecological circumstances. The evaluation of a company’s success is closely tied to its financial performance, which pertains to its capacity to produce earnings and enhance shareholder worth. The correlation between CSR and financial performance has been a topic of significant discourse, with numerous studies examining this subject from diverse perspectives. Nevertheless, establishing a causal association between the concepts above poses a formidable challenge, owing to the intricate nature of the association and the likelihood of extraneous variables exerting an effect. This study suggests a model for CSR based on Machine Learning (ML), referred to as the Machine Learning-based Corporate Social Responsibility Model (ML-CSRM). This model aims to establish a causal relationship between CSR and financial performance. The paper commences with a comprehensive exposition of CSR and its correlation with financial performance. The discourse delves into the intricacies of ascertaining the causal relationship between the two. The method under consideration utilizes Latent Dirichlet Allocation (LDA) and text clustering techniques to examine extant legal documents about CSR. The study presents the proposed method’s simulation results [Pearson’s Correlation Coefficient, precision, recall, Root-Mean-Squared Error (RMSE), and R-squared] and discusses the proposed approach’s outcomes. The outcomes of this research carry significant ramifications for both commercial enterprises and the broader community, given that they offer discernment into the correlation between CSR and economic outcomes.