Ownership structure and real earnings management: A meta-analysis

This article examines the relationship between ownership structure and real earnings management (REM) by applying a meta-analysis of 808 effect size estimates from 49 primary studies. We explore how institutional investors, family owners, and insiders relate to REM across the literature, assess publication selection bias, and identify factors contributing to the heterogeneity of reported results. The findings reveal publication selection bias in studies on institutional investors; after correcting for this bias, their relationship with REM appears to be insignificant. Family ownership similarly shows no significant mean effect, while insider ownership exhibits a consistent, though small, negative association with REM. Heterogeneity analysis indicates that the mitigating role of family and insider ownership is context-dependent, appearing stronger in Western country settings, consistent with cultural differences across institutional environments, while legal and macroeconomic factors explain little of the variation in reported estimates. Results are also sensitive to methodological choices, particularly the measurement of REM and controls for endogeneity. The findings of this meta-analysis provide insights for policymakers and researchers aiming to better understand the conditional relationship between ownership and earnings quality.