As part of our commitment to diversity and to ensure we follow best practice, we have now opted to publish an annual diversity report covering our recruitment and team make up.
We currently employ 13 people; three women, and ten men. Four of the team are not developers. Below I only discuss the developer team. In the management team, we have an exact 50:50 gender split of two women and two men.
Last year for our front-end developer job opening we had 35 applicants, of which 7 were women. So a 5:1 ratio of men to women.
During that recruitment drive, we hired a woman, Claire. But truth is, we could just as likely have employed a man. It was just that Claire was the best person for the job, with a great attitude to learning and development.
Key is, with our team size it’s very hard to give statistically relevant statistics.
For our team of nine developers, we would expect approximately 1.8 developers to be female based on the ratio of applicants. We have one. But statistically, we can apply the chi-squared test to see if there’s a bias. The result? The chi-square statistic is 0.3802. The p-value is .537471. This result is not significant at p < .05. In other words, we can’t state with any certainty that we’re fair or not.
Can we do better?
What we need to look at is how many women it is possible to get applying for purely technical jobs. We do our best with gender-neutral advertising (as confirmed by LGG!) and we have policies that allow for as much diversity as possible in our team — including giving staff control over the hours they work, an allowance for home working once they have proven their work ethic, accessible offices and family-friendly policies. In supporting LGG we help to promote the idea of working in a tech field as being attractive. Our director is also a STEM Ambassador, who visits schools to promote the industry to all — in fact, David Coveney makes a point in targeting his time at state schools in deprived areas of the region.
We could also look at whether there is any other underlying bias. For example, are women actually better than men and then men more willing to lie about their skills, resulting in fewer women making it to the shortlist? I spoke about how CV shortlisting can be troublesome in a blog post on the Liverpool Girl Geeks site.
We believe it’s important that the language used in job descriptions doesn’t show unconscious bias, that can make jobs sound off-putting to women. As part of working with Liverpool Girl Geeks, we discovered that we already performed well on this score. That means that we don’t put off men or women from applying. You can check your own job adverts using the Gender Decoder.
One of the biggest problems for a small company like ours is to publish any genuinely useful information due to the small sample size we’re working with. Statistically, a developer team of our size with zero women, based on job applications, would not reveal any statistically significant signs of bias. If, however, we employed five women, then it *would* suggest a strong gender bias.
The same applies to almost all areas. We have three people who are not British citizens. That’s largely representative of our applications from qualifying applicants. But again, non-UK born citizens in Liverpool account for around 7% of people. That means we should have around 0.63. Except we actually have three. That’s because if you’re a growing company you will hire from the available talent base, which is more likely to be made up of immigrants and people new to the local workforce. That also suggests that although we have one woman developer, there aren’t any signs of a big growth in the number of women entering the tech workforce over the past five years, which is when we’ve really been growing. If there were, you could actually expect a disproportionate number of women in our company as compared to older, more established firms — and we don’t appear to be substantially different there.
In terms of employment of people with disabilities, again, it’s a numbers game. Of applicants with disabilities, we only know of a few people out of hundreds of applicants who have a disability (although it’s worth noting that not all disabilities are visible, and we don’t ask). In terms of the applicants we have received, statistically, it would be more likely for us to hire someone with a chronic health condition than a disability, and we know we have at least two employees in that category.
In total, out of a team of 13, we employ three women, three non-British nationals, four people with Irish heritage, nobody with known Asian, African, American or Australian heritage. All staff have some tech capability, although administrative staff don’t code very often, all have been taught to do so and have a deeper understanding than most.
The numbers suggest that our recruitment and processes aren’t actively discriminatory. We can also state proudly that we do our best to be as inclusive as possible. The wider a range of people we employ, the better we get at servicing the rich diversity of humanity on this planet. That can only be good, right? I’d say we’re representative of the dev community as a whole (go to a conference and see what it looks like), but that we can still do better and work harder to bring more people into what is a rewarding and growing industry.
P.S. No salary stats are published – with our relatively small numbers it would reveal potentially sensitive information. Let’s just say that top to bottom, with a formal pay scale, the difference between our lowest earner and our highest earner is currently about 2x.