Real Earnings Management and Quality of Corporate Governance: A Meta-Regression Analysis

The goal of this study is to meta-analyse the results of 146 primary research studies examining how three categories of corporate governance – ownership structure (OS), gatekeepers (G), and management characteristics (MC) – affect real earnings management (REM). The application of meta-regression (MRA), which is a method for combining and analysing data from multiple studies, makes it possible to link effect sizes from primary studies into a single summary effect and to explain the variability among these effect sizes. The mean values of the partial correlation coefficients between corporate governance categories and REM are not statistically significant. However, the results of heterogeneity analysis show that family ownership and insider shareholding can be positively associated with REM. The analyst coverage seems to have a mitigating effect on REM, while external auditor quality does not. The study identifies several moderator variables that explain some of the structural and methodological heterogeneity in empirical results. The results on REM utilisation support the opportunistic view rooted in agency theory. The findings obtained confirm the need for further operationalisation of variables and an expanded analysis of various components of corporate governance.