September 18, 2014
In Defence of Diversity Measures
I recently attended two tech conferences (EMF Camp and JS Conf EU), both of which had successfully made a conscious effort to reach out and attract women speakers. Most people I’ve spoken to agree that attempts to increase diversity are a good thing. Inevitably however, there are some that immediately cry ‘positive discrimination!’. I find myself trying to combat the same old misconceptions time and time again. I’d like to try and dispel some of them here.
Conflating terms: positive discrimination, quotas, and diversity targets
People often use the phrase ‘positive discrimination’ when they mean something else entirely. Positive discrimination, otherwise known as affirmative action, is the process of; given two equal candidates; preferring the one who is usually disadvantaged by discrimination. This is different to quotas, where a certain number of places are reserved for disadvantaged minorities. This is different again from diversity targets, which as they describe, are a target, not a mandate. Targets often involve simply trying to attract a wider, more diverse range of people to apply for a role, with no preferential treatment after that stage. For brevity, I shall group these under the term ‘diversity measures’. You may take issue with one kind of diversity measure and not with another, but let’s get our definitions straight from the start.
“Diversity measures are inherently unfair”
What is unfair is the irrational prejudice faced by marginalised groups which prevent them from having the same opportunities as others. Diversity measures are attempts to level a very uneven playing field. At first glance, it may appear that they give marginalised groups an unfair advantage. In reality, they are an imperfect attempt to counteract the huge unfair advantages afforded to privileged groups. In terms of gender, numerous studies have shown that we all have biases that cause us to subconsciously favour men. This is compounded by the fact that the tech community is predominantly made up of white men, and people are more likely to favour others like themselves. This is institutionalised sexism. It is irrational, and makes it possible for jobs and speaker spots to be given to men over more deserving candidates.
Perhaps the least controversial approach to evening the playing field is to simply try to widen the pool of applicants to include marginalised groups. When it comes to public-speaking, women are much less likely to put themselves forward than men due to lack of confidence and under-estimation of their ability. JS Conf EU took a very successful approach to this simply by emailing women they believed had interesting things to say, but wouldn’t put themselves forward without encouragement. They then anonymised all submissions, ensuring both that their subconscious biases would not affect their choices, and that they could not be accused of choosing women ‘just for quotas’.
“It results in a drop in quality”
In fact, when the JS Conf EU team analysed the submissions, they found that while women submitted just 10% of talk proposals, they actually made up 25% of the successfully chosen talks. This means that a woman was around twice as likely to submit a talk of the required high quality than a man. This is not to say that women are better than men, it is explained by the theory that women tend to put something forward only when they are absolutely convinced of its quality, whereas men are more likely to take a chance on something sub-par.
This highlights another misconception around these measures, which is that they will result in a drop in quality. In fact, the reverse is usually true. This post uses game theory to illustrate how institutionalised sexism results in an overall drop in quality. Because of our biases, men are more likely to get a position they don’t deserve simply because they are men, yet ironically in the tech industry it is largely women who get asked ‘how do you feel about the fact that you probably got this job because of your gender?’
Subjectively, I found (and many others agreed) that both JS Conf EU and EMF Camp had a more welcoming atmosphere, a greater range of topics and a higher quality of speakers than others that have not made the same efforts around diversity.
“There just aren’t enough qualified women around”
When it comes to tech events, organisers often claim they simply can’t find enough qualified women to speak. Due to the fact that women are vastly outnumbered in the industry, this is not such an extraordinary claim. So a seemingly uncontroversial way to solve this problem would be to create a handy list of accomplished women speakers, making the lives of organisers much easier. However, for those who don’t recognise structural inequality, even this is unacceptable:
I can only imagine why someone would be so against such an inoffensive project to create a balanced, diverse community. The BBC has clear diversity targets for BME and disabled recruitment, but I do not in any way feel threatened. In fact I am looking forward to a higher quality work force because of them. Perhaps those who react defensively are realising they may finally have to give up their unfair advantage, and are worried that when judged against their peers of other genders, races, or disabilities; they will be found wanting.
“Women don’t like diversity measures, they’re patronising”
Because of the many misconceptions around these concepts, it’s easy to see why women would reject them. They don’t want people to believe they are only there as a token, or have been chosen over a more qualified man. But as we’ve discussed, the problem lies not with the concept, but with the fact that people perceive it incorrectly. Of course, even with all the facts there will still be some women who reject these measures because guess what, women are not a homogenous mass who all think alike, but are, in fact, people. But a few anecdotal cases should not outweigh the mounting evidence that doing nothing about diversity will make the problem worse for everyone.
“I just want the best person, regardless of gender”
Exactly. So do we. That’s precisely why we implement measures to try to mitigate the irrational biases that cause us to favour certain groups for no good reason.
“We should be blind to gender!”
Yes, in a perfect world, perhaps we should. But we are a long way off that utopia. For the nth time, tech is not a meritocracy. I address this and related myths here. One day, we can hope, none of these measures will be needed. But at the rate things are going, we won’t get to that stage without some kind of intervention. If we want more women in the industry, we need more women as role models and more women doing the selecting. These measures may not be perfect, but at the moment they constitute a best effort approach to combat institutionalised sexism.
Updated Jul, 21 2020