Trends in the thematic landscape of corporate social responsibility research: A structural topic modeling approach

Abstract

This study provides a bird’s eye view of research on corporate social responsibility (CSR) by text-mining 14,523 articles published in reputed scholarly journals spanning the period from 1991 to 2020. This is the first study on CSR to use an unsupervised text-mining method (structural topic modeling) thereby overcoming several limitations of methods used in previous analyses. Aside from the methodological novelty, this study contributes to three distinct but related strands of literature. First, it contributes to efforts made in the prior literature to characterize the CSR concept in a novel way. Second, this study contributes to the debate on the dimensions that constitute CSR. Additionally, it provides evidence of how the focus on particular dimensions has varied over time. Third, it contributes to the debate on instrumental versus injunctive CSR by providing concrete evidence that the literature has shifted from injunctive to instrumental CSR.