Sunday, January 6, 2013

Gender of Directors: The Critical Mass is Critical

There is lots of excitement about "Does the Gender of Directors Matter?" a paper by Miriam Schwartz-Ziv.  I think the reporting about this paper has been a bit misleading and has missed the novel and interesting contribution made.

Cosider Matt Yglesias title, "Boards with more women on them are more aggressive, and their companies perform better."  Alas, Miriam didn't find that more women on boards or attending board meetings improves oversight or performance.  She found that have AT LEAST three women at a meeting and on a board increased oversight and performance.  The difference is important, as reported in the paper, currently no relationship exists between gender and performance, "Rhode and Packel (2010) survey more than two dozen studies examining the association between the gender composition of boards and financial performance, and conclude that no robust and consistent relation has been documented between these two variables."

If the author had taken her eleven firm sample (note 11 firms!) and just entered another data point into the findings above, it wouldn't be all that interesting.  Instead, the author confirms the findings reported above.  The fraction of women on the board has no effect on performance or board meetings, "No significant linear or U-shaped relation is documented between the gender composition of boards and the variables measuring board activeness" (table 5).

So why is it interesting?  Because the author adds a big caveat.  If at least three women (and to a lesser extent at least three men) attend a meeting, the board is much more active in oversight that it otherwise would be.  Also, boards with at least three women serving have higher ROEs and higher profit margins (the better more aggressive performance).  This result doesn't hold with increasing the percent of women or even with at least two women attending (p. 17).

The author argues that for women's participation to be valuable that there must be a critical mass of at least three women and offers some theory to support this view.  It's an interesting result, but I think lots of caution is important.  The study only uses a peculiar data set of eleven Israeli firms.  We have lots of female board participation in Norway and due to legislation a somewhat natural experiment of female board participation.  I'd like to see the results hold up out of sample before I got very excited about them.  It's true when Yglesias says, "This is important not only for what it reveals about gender, but for the fact that it suggests that firms are being systematically mismanaged," but that's also a reason to be initially skeptical, until we have a lot of evidence to the contrary.

Most importantly, the critical contribution of the paper is that a critical mass is required for female board participation.  Let's make sure we emphasize that in our reporting of the study.  This interpretation changes the class of policies we should consider useful in improving board performance.

No comments:

Post a Comment