Female directors and agency costs: evidence from Chinese listed firms

PurposeThe purpose of this research is to examine whether board gender diversity reduces the agency costs of firms in the context of Chinese listed firms.Design/methodology/approachThis paper uses a large sample of 23,340 firm-year observations of Chinese listed companies during 2004–2017. The authors use ordinary least squares regressions as the primary methodology with a wide range of methods to control for endogeneity and to check robustness, including the fixed-effect method, instrumental variable approach, lagged gender diversity measures, propensity score matching, Blau index, Shannon index and industry-adjusted measures of agency costs.FindingsThe evidence reveals that the participation of female directors in corporate board reduces agency costs, which correlates with conflicts of interest. Moreover, gender-diverse boards are more effective in state-owned enterprises (SOEs), in which agency issues are more severe. Female directors also provide better monitoring roles in more-developed areas. Finally, corporate boards that have a critical mass of female directors have a greater tendency to reduce agency costs as compared to their token participation. Overall, all findings support the validity of agency theory.Practical implicationsThis study shows the economic benefit of female directors in the boardroom by reducing agency costs and by improving firms' governance structure. Regarding the government, which is gradually introducing board gender diversity policies, this study provides valuable pragmatic information for Chinese regulators on this issue.Originality/valueThis study extends the literature by providing evidence that gender diversity in boardroom matters for shareholders' wealth maximization. It provides novel evidence that a critical mass of female directors is more effective in reducing agency costs compared to a single female on the board, and that the effect of gender diversity varies in relation to ownership structure and region.