A meta-analysis of board attributes and real earnings manipulation across global contexts
Abstract
This study quantitatively analyzes previous empirical findings on the relationship between board attributes and real earnings management, employing a meta-analysis approach. It synthesizes 45 empirical studies comprising 207 effect size estimates following the PRISMA guidelines. The study uses a random effects model and meta-regression to examine how various board attributes influence real earnings management across different contextual and methodological settings. The results show that specific characteristics, such as board independence, board size, gender diversity, and nationality diversity, are effective governance mechanisms for reducing management opportunism. In contrast, CEO duality has a positive relationship with real earnings management and raises governance concerns related to board independence and effectiveness. The meta-regression results show that the adoption of IFRS, sample size, and publication year significantly moderate the relationship between board attributes and real earnings management. Furthermore, subgroup analysis revealed variations between legal system, real earnings management proxy types, statistical estimation models used, and the number of control variables used in each empirical study. The findings have practical implications for policymakers and companies to strengthen the implementation of IFRS and to optimize the composition of the board of directors to improve the quality of financial reports.