Does the presence of independent and female directors impact firm performance? A multi-country study of board diversity

For example, Brickley et al. (1997), Luan and Tang (2007), Florackis and Ozkan (2009), Kim and Lim (2010), Jackling and Johl (2009), and Pombo and Gutiérrez (2011) report a positive relationship between the percentage of independent directors and firm performance. In contrast, Hermalin and Weisbach (1991), Barnhart and Rosenstein (1988), Bhagat and Black (2002), Vafeas and Theodorou (1998), Klein (1998), and Arosa et al. (2010) find that the presence of independent directors does not increase firm value. Moreover, Agrawal and Knoeber (1996), Bebchuk and Cohen (2005), and Shan and McIver (2011) find that independent directors decrease value. Faleye et al. (2011: 177) report that intense monitoring by independent directors may negatively affect firm value, thus “suggesting that the costs of weak advising outweigh the board’s monitoring.” Nguyen and Nielsen (2010) find that the death of an independent director decreases shareholder value. Kang et al. (2007) report mixed results. Still other studies find no significant differences.

Erhardt et al. (2003), Carter et al. (2003), Campbell and Mínguez-Vera (2008, 2010), Carter et al. (2010), Kang et al. (2010), Gul et al. (2011) and Mahadeo et al. (2012) Lükerath-Rovers (2013), Ntim (2013), among others, report a positive relationship between gender-diversified boards and firm performance. Sun et al. (2011) find no association between gender diversity of independent audit committees and the ability to constrain earnings management. Similarly, Kang et al. (2007) find no relationship. The lack of consistent findings may be due to prior studies’ limited and non-harmonized measures of firm performance and lack of control variables. A meta-analysis of over eighty studies finds some support for gender (Post and Byron 2015). A more recent stream of research examines female CEOs and Chairs, finding a positive relationship to performance (Peni 2014).

This model does not include the variable percentage of female directors and the interaction term because they are highly correlated and it is impossible to interpret their segregated effects. The implication of dropping the percentage of female directors from this model is that we do not consider the individual effect of this variable on firm performance, which then leads to an omitted variable problem. The instrumental variable regression model (GMM) solves this problem and the estimated coefficients remain robust.

We thank a reviewer for this suggestion.