European commissioners are thinking of demanding quotas for the number of women on company boards. This raises some issues.

Issue one is about selection effects. Imagine - which shouldn't be hard - a world in which there are very few women directors. A company which appoints one is therefore, by definition, unusual - perhaps in being more forward-thinking. And the woman it appoints is also likely to be unusual, perhaps especially talented. The data might then show that companies with women directors do well. But this is because of a selection effect that could go into reverse if all firms were compelled to appoint women, some of whom might be mediocre.

However, selection effects needn't work positively. They can be adverse. Let's grant that there's some evidence that women - on average - would help companies make better decisions by virtue of their gender, say because they are less overconfident, more risk-averse or have higher ethical standards. (Whether these differences are innate or due to social constructs is irrelevant for my point). It does not follow from this that companies that appoint women would be better-run. This is because the women it appoints might not have these feminine virtues but might instead have gotten to the top by virtue of being more ballsy than men. If this is the case, "free" appointments of women might actually reduce the amount of feminine attitudes on boards. By contrast, quotas - insofar as they compel boards to appoint feminine women - might actually enhance performance.

These competing selection effects might explain why studies of the effect of women directors upon corporate governance have been so ambiguous; some find a positive effect, some a negative (pdf), and some a U-shaped one. The question is: which selection effect would predominate if quotas are introduced?

The second issue is: why quotas? There are many possible ways of improving corporate governance: decentralizing management; having more worker-directors; preventing reckless takeovers; preventing narcisscists, sociopaths or psychopaths from becoming CEO; encouraging shareholders to exercise proper control, and so on. Why, of all the possible ways of improving corporate governance, should quotas for women be top of the list?

Thirdly, if there should be quotas to ensure boardroom diversity, why confine them to women? Why not quotas for ethnic minorities or gays?

One possibility is that women bring something to boardrooms that other minorities don't, by virtue of the gender differences I've mentioned. But as I've said, there's no guarantee that women appointees would necessarily possess such characteristics. If you want less overconfidence and reckless risk-taking on boards, why not mandate appropriate psychometric tests rather than worry about the shape of genitalia?

Which brings me to a final issue. I suspect that what we're seeing here is not so much an attempt to improve corporate governance as a paradox of equality - the paradox that, by the time a group is powerful enough to press for equality, the need for such equality has diminished.

The time when women most needed legal protection in the workplace was in the 60s and 70s, when they were "dollybirds" to be raped at will. But women then lacked the political power to achieve their rights. Only now are they are politically strong enough to demand laws to ensure their boardroom representation - and yet their economic position is strong enough that they can (up to a point) achieve such representation without the need for law.

I don't say all this to argue against gender quotas; anything that frustrates careerist men has something to be said for it. But let's not think this is much about corporate governance.