CSR DISCLOSURE AND PROFITABILITY ON TAX AVOIDANCE WITH INDEPENDENT COMMISSIONERS AS MODERATING VARIABLE: EVIDENCE FROM BANKING COMPANIES IN INDONESIA

Abstract

This study examines the effects of Corporate Social Responsibility Disclosure and profitability on tax avoidance, with independent commissioners serving as a moderating variable in banking companies listed on the Indonesia Stock Exchange during the 2021–2024 period. Drawing upon Agency Theory, Legitimacy Theory, and Stakeholder Theory, this study investigates how corporate sustainability practices, financial performance, and governance mechanisms influence corporate tax behavior. A quantitative research design was employed using an unbalanced panel dataset consisting of one hundred and twenty-nine firm-year observations from thirty-three listed banking companies. The data were analyzed using the Fixed Effect Model with robust standard errors to obtain consistent statistical inference under heteroscedasticity and within-panel serial correlation. The findings reveal that CSR disclosure has a positive and significant effect on the Corporate Effective Tax Rate, indicating lower levels of tax avoidance among firms with more extensive CSR disclosure. Profitability, measured by Return on Assets, has no significant effect on tax avoidance. Furthermore, independent commissioners significantly moderate the relationship between CSR disclosure and tax avoidance by weakening the positive effect of CSR disclosure on Corporate Effective Tax Rate, whereas they do not significantly moderate the relationship between profitability and tax avoidance. The model explains forty-nine-point four seven percent of the variation in Corporate Effective Tax Rate. These findings highlight the importance of integrating transparent CSR disclosure with effective corporate governance to encourage responsible corporate tax behavior. This study contributes to the literature by providing recent empirical evidence from Indonesia's banking sector during the post-pandemic period using a moderated panel data regression approach.