Opportunism vs legacy conservation: the impact of retiring CEOs on ESG controversies

Purpose This study investigates how chief executive officers’s (CEOs’) impending retirement is associated with corporate environmental, social, and governance (ESG) controversies, aiming to shed light on the theoretical tension between opportunistic motives (based on agency theory) and legacy conservation motives (based on reputation theory). It seeks to illuminate the micro-level behavioral correlates of ESG risk events during leadership transition periods. Design/methodology/approach Utilizing a sample of Chinese A-share listed firms from 2017 to 2023, this study employs fixed-effects regression models. CEO retirement is measured as the final two years of tenure; ESG controversies are measured using reverse-coded Wind ESG controversy scores. The analysis examines the main association, moderating conditions, plausible pathways, and heterogeneity, supported by extensive robustness checks including propensity score matching and two-stage least squares. Findings We find that CEO retirement is positively associated with corporate ESG controversies, a pattern more consistent with the opportunism perspective than with the competing legacy-conservation prediction. This association is attenuated by higher CEO age, positive external expectations, and effective internal control, but strengthened by financing constraints. Mechanism tests are consistent with plausible pathways involving lower substantive ESG investments and more frequent corporate violations/speculative behaviors. The association is more pronounced in non-state-owned enterprises and industries with weak market competition. Originality/value This study expands the antecedents of ESG controversies by introducing the perspective of CEO career dynamics, moving beyond prior focus on institutional or governance factors. It provides nuanced empirical evidence consistent with the “opportunism” perspective in the CEO retirement literature, while indicating the situational dependency of the observed association through key boundary conditions.