Female bosses raise profits by how much? Really?

How's this for a provocative headline at the Daily Mail:

More female bosses mean more profits: Companies whose boards are made up of at least a third women make 42 per cent more

That seems very unlikely, so I checked it out. The source of the statistics is a group called Catalyst. One mission of this group is to promote women to company boardrooms. That doesn't necessarily mean the data is false, but it does raise the problem of potential bias.

I looked for some other research on the issue. What I found does not support the Catalyst claims. For instance, there was research conducted when the Norwegian government forced companies to make their boards at least 40 per cent female:

Amy Dittmar, associate professor of finance at the University of Michigan’s Ross School of Business has recently analysed the impact of the Norwegian decision...Dittmar and Ahern’s study found that when a board had a 10% increase in the number of women, the value of the company dropped. The bigger the change to the structure of the board, the bigger the fall in returns.

An even better alternative source of information on this issue is a paper by two Harvard University sociologists summarising the research to date. According to the paper most research has found that adding female board members does not improve company performance:

Analysts have explored the effects of board diversity on both profitability and stock valuation.

The overall pattern of findings across the several dozen studies that have been published to date tends to support the view that gender diversity inhibits performance (p.10)

The Harvard sociologists do recognise that the Catalyst research connected female board members to profitability:

Perhaps the best publicized study linking board diversity to profitability is Catalyst’s comparison of over 500 leading U.S. firms between 2001 and 2004.

However, they criticise this research for not considering the possibility of reverse causation. In other words, does appointing women lead to higher profits or do higher profits lead to the appointment of women?

According to the Harvard sociologists those research projects which do examine the issue of reverse causation either find that there is no difference in profitability when female board members are appointed or that there is a decline in profits. Here are some of the research projects they list:

• Zahra and Stanton find no effect generally, and some evidence of a negative effect, among large American firms in the 1980s.
• The Scandinavian countries were leaders in promoting board gender diversity. A recent study shows no effect of gender diversity on stock performance in a sample of 443 Danish firms.
• Smith, Smith, and Verner use panel data on 2500 Danish firms to explore several performance measures. Female outside directors show negative effects, though female inside directors show positive effects.
• In their 2009 study, Adams and Ferreira use panel data between 1996 and 2003 on 1939 large American firms. Theirs is possibly the most sophisticated, and transparent, analysis published to date. While they find that boards with more women do better at monitoring firms, they also find negative effects of women board members on both Tobin’s q and ROA (return on assets). In particular, they find positive gender diversity effects in OLS models, but two different techniques for handling endogeneity (fixed effects, and fixed effects with instrumental variables) produce negative and significant effects (for profits and stock value) and a third produces negative but non-significant effects for both outcomes. (pp.11-12)

The conclusion? This:

Taken together, these studies are consistent with the idea that firms that are having good runs are more likely to appoint women, but that once appointed, women have neutral or negative effects on performance.

Several studies address this directly. Farrell and Hersch examine a sample of 300, Fortune-500 firms between 1990 and 1999, showing that firms with strong profits (ROA) are more likely to appoint female directors but that female directors do not affect subsequent performance. Adams and Ferreira find that Tobin’s q, but not ROA, predicts the appointment of female directors but, as noted, female directors have subsequent negative effects. They conclude: "Although a positive relation between gender diversity in the boardroom and firm performance is often cited in the popular press, it is not robust to any of our methods of addressing the endogeneity of gender diversity." (p.12)

The last bit is an academic way of saying "don't believe what you read in the papers on this issue". The problem is that the Catalyst statistic is a useful one for those pushing for boardroom quotas, so you'll be reading it over and over.