CEO-narcissism and its asymmetric long- and short-run relations with environmental, social and governance performances: A panel examination of 125 companies in the USA

Abstract

By employing a new social-media-based CEO-narcissism index (CNISM), the study explores the dynamic association among CEO-narcissism (CN) and environmental social governance (ESG) performances, environmental performance scores (EPS), and a large set of controls for financial variables (FV) in 125 companies in the USA for 2010-2023. Novel Panel Quantile Autoregressive (PQARDL) model is employed to achieve an enriched empirical examination for asymmetric impacts of CN and EPS on ESG performances in the long- and short-run. Additional controls are included for robustness to their time-varying effects, including management quality, regulatory quality and industry effects along with bootstrapped standard errors. Afterward, Panel Granger-Causality (PGC) tests are utilized to identify causality dynamics among CN, ESG, EPS and FV. The empirical findings confirmed negative effects of CN with varying coefficients depending on different ESG quantiles, as well as positive effects of EPS, positive or negative effects from FV to ESG depending on FV measure including different financial performance measures in the long-run. Further, negative effects of CN in majority of quantiles of ESG in the short-run appear as a significant factor influencing ESG asymmetrically along with negative effects of EPS on ESG in line with relatively low association between ESG and EPS with important implications. The most negative effects of CN occur for companies and periods with lowest levels of ESG as CN upsurges come with further negative effects on ESG, however the ESG-reduction effect of CN diminishes at higher quantiles. In the short-run, the positive effect of CN dominates results, which highlights the contrast between the long- and short-run treatment. The results were tested and confirmed for robustness to reverse-causation and endogeneity, and reliability and measurement validity is affirmed with tests. Platform-choice were tested for correlations with the traditional CNIT index, and the results suggested efficiency of the new CNISM index in different platforms. PGC results further indicated significant causality from CN to ESG, bidirectional feedbacks between CN and ESG, and unidirectional causality from all FV to ESG. Important implications are suggested for policy makers, managers and for related empirical literature.