Board Diversity: Should We Trust Research to Inform Policy?by Daniel Ferreira

Corporate Governance: An International Review



Board Diversity: Should We Trust Research to

Inform Policy?1

Daniel Ferreira*

R esearch on board diversity has received much attentionrecently. This is a natural consequence of the fact that female representation on corporate boards has risen to the top of policy agendas. It is, however, far from obvious whether and how such research is useful for policy discussions. This is our own fault. We, the researchers, often do not explain what we are trying to achieve. We are not always clear about the limitations of our work, and we often tolerate misinterpretation of our work by policymakers and the media.

In this short piece, I summarize what we have learned from the board diversity literature, and what is useful (and not useful) for policy debates and design. I organize my comments around three questions.



The answer is: not a lot. It is very hard to disentangle “diversity effects” from the effects of other individual and group characteristics that correlate with measures of diversity.

Board directors, as a group of people, are highly nonrepresentative of the general population. For example, there is no reason to believe that female board members have the same personality traits as those observed in the general population. Thus, findings of “gender effects” on boards are unlikely to be generalizable beyond the board. This problem is an example of what some researchers call lack of external validity.

External validity issues notwithstanding, we can still learn something. In particular, we may learn about what kind of behavior survives many rounds of selection and selfselection. As an example, consider the evidence that, in the general population, women are more risk-averse than men (see, e.g., Sapienza, Zingales, & Maestripieri, 2009). Because of selection, it does not follow that female directors should also be more risk-averse than male directors. In fact, in a sample of Swedish directors, Adams and Funk (2012) find that female directors are more risk-loving than male directors.




Here I think we can be more optimistic. As I have argued elsewhere (Ferreira, 2010), there are six broad conclusions we can draw from the empirical literature on board diversity: (1) Firms appear to choose directors for their characteristics, and different types of firms choose different levels of director heterogeneity; (2) firms choose directors as a means to deal with the external environment; (3) CEOs and top executives appear to prefer directors who are similar to themselves; (4) social networks and commonality of backgrounds appear to affect director appointments and the dynamics of the board; (5) directors from minority groups perceive their minority status as a hindrance to their work as a director; (6) minority directors may serve interests other than those of shareholders.

My work with Renée Adams (Adams & Ferreira, 2009) reveals some of the interactions between gender and governance. We find evidence that female directors are more independent (from management) than male directors. Crucially, this is true for both nominal and de facto independence. We find that women have better attendance at board meetings, are more likely to sit on monitoring committees, and are more likely to force CEO departures after poor stock price performance. In short, female directors are more likely to be tough monitors of CEOs.

Our interpretation is not that independence is a female trait. We cannot rule out the possibility that the observed independence of female directors is explained by other unobserved characteristics, such as different social and busi*Address for correspondence: Daniel Ferreira, Department of Finance, London School of Economics, Houghton Street, London WC2A 2AE, UK. Tel: (+44) 20 7955 7544;

E-mail: 1

Corporate Governance: An International Review, 2014, ••(••): ••–•• © 2014 John Wiley & Sons Ltd doi:10.1111/corg.12092 ness networks of female and male directors. It is important to keep this in mind, especially when interpreting findings that show an association between board gender diversity and firm performance. Because we cannot disentangle “gender effects” from the effects of independence, we interpret our findings as suggesting that board gender diversity might be a better proxy for board independence than conventional measures, which are based on regulatory definitions of independence.

As we have argued elsewhere (Adams & Ferreira, 2007), independence comes with costs and benefits, and there is thus no reason to expect the average firm to benefit from adding more women to its board. In fact, our evidence also indicates that the link between firm performance and board gender diversity is tenuous. The estimated effects of board diversity vary substantially across empirical specifications and methods. Such effects are also heterogeneous, i.e., they depend on firm characteristics. Some firms appear to benefit from adding women to the board, while others would probably experience a decline in performance. This should come as no surprise, given that gender diversity is strongly related to board independence. The literature has taught us that board independence has no obvious impact on firm performance (see the survey by Adams, Hermalin, & Weisbach, 2010).

There is a fascination in the management and economics literature with estimating the impact of female directors on firm performance and profitability. This is understandable. But this literature is often too quick to jump to strong conclusions on the basis of the flimsiest of evidence. Establishing causality requires strong assumptions. In many academic studies and policy pieces, the usual disclaimer that “correlation does not imply causation” often comes in small print. Econometric black boxes do not solve this problem either. State-of-the-art identification strategies often look superficially convincing, but they rarely deliver the goods.