Unintentional bias isn’t less bias

2015-04-02

I was somewhat amused a few days ago by a Twitter exchange between Curt Rice, who will soon become the Rector of Oslo and Akershus University College of Applied Science and is a staunch supporter of gender equality in academia, and the author Ole Asbjørn Ness. Rice has criticized science prizes for sexism in the past, and noted in a tweet that the Abel Prize, which is given for mathematics, has never honored a woman. Ness retorted “What a stupid remark. Do you honestly believe the committee is sexist?”

My answer: Yes. Yes, it is.

This answer doesn’t mean the committee is explicitly stating that they don’t want to give women the prize. It is an admission that inherent biases against women abound in academia.

A new VR report A Gender Neutral Process? investigating the VR funding application review procedures contains many examples of unintentional gender bias against women applicants, including criticising women more often than men for lacking ‘independence’ and excusing men’s poor performance more than women’s. Inherent bias has been discussed as the primary reason women leave science. Of course, these are not new problems. Several years ago, linguistic analysis of academic letters of recommendation showed real differences in the way that males and females are represented.

All this matters because it comes down to money: women are awarded fewer grants. In VR’s summary of the 2014 awards, they stress the percentage equality of awarded projects—15% of the women’s applications were funded and 15% of the men’s were funded. The graphs they include also stress equality along percentage lines. But if you look carefully, you see that there were twice as many applications submitted by men. That means that twice as many grants ended up awarded to men!

Why is it assumed that the same percentage of submitters in each gender would be equally worthy of grants? Studies, like this one from Germany in 2012, have repeatedly pointed out that women tend not to apply for as many grants as men. The VR report recommendations were criticized in an editorial yesterday for not taking this into account. Women typically have lower-level appointments or fixed-term contracts and tend to de-value their own expertise, so they opt-out of applying; whereas men tend to apply regardless of employment status. Perhaps those women who do apply are actually much more qualified than their male counterparts. Because of differences in application behavior, grant success rate equality is not enough! Implicit sexism pervades the grant application and funding process.

I do not want to imply that anyone is intentionally not funding women, but just because the bias is unintentional doesn’t make it harmless. In fact, it makes it more detrimental because people may not be able to see that they are biased in their assessments. These biases need to be put out in the open if they are ever going to be overcome.