Abstract
Purpose This paper aims to examine whether corporate social responsibility (CSR) is associated with firms' earnings quality (EQ) and how this association is context-specific. The authors consider specific institutional differences in strength of corporate governance (CG) attributes, quality of law enforcement and level of investor protection found between Anglo-American, European and South-Eastern Asian CG models to test the impact of above country-level factors on this association. Design/methodology/approach To test the association between CSR and EQ, the authors consider EIRIS (Ethical Investment Research Service) (2018) CSR issues of sustainability indicators as proxy to capture CSR. Following Rezaee and Tuo's (2019) study, the authors classify EQ into innate earnings quality (IEQ) and discretionary earnings quality (DEQ). The authors investigate the innate (discretionary) EQ as to refer to firm's inherent operating uncertainty (earnings management). Several dependency models for panel data applying the generalized method of moment (GMM) estimator of Arellano and Bond (1991) are ruled based on archival data of 4,206 non-financial international listed firms over the period 2012-2017. Findings Univariate and GMM multivariate cross-country analyses show that CSR is positively associated with EQ and that this association is more pronounced for firms within countries where good CG tools and higher investor right protection are preserved. The authors interpret the findings as evidence that the CSR-EQ association is shaped by the degree of monitoring role played by institutional features at the country level. The results are robust to a battery of robustness tests. Originality/value The originality of this research is twice. On the one hand, it examines whether CSR is a reflection of manager's ethical opportunistic behavior resultant on earnings quality derived from a firm's innate traits. On the second hand, it tests whether CSR is a reflection of discretionary earnings quality manifested by earnings management behavior. This paper is the first to support that institutional features significantly matter when investigating the association between CSR and EQ.