25 June
By Rose

Gender Pay Gap Data Hack


The event was a collaboration between several groups supporting women and minorities in the technology sector, and kindly hosted by Buzzfeed UK. The main organising groups were the AI Club for Gender Minorities, Codebar - a non-profit which matches volunteer mentors with (mainly) female students who are learning to code, and Ellpha - a data organization committed to reducing the gender pay gap using data-driven insights.

The data hack was a really fun day, which basically involved getting together with a load of other people and trying to find insights from various datasets concerning the gender pay gap. We were lucky enough to have a large dataset scraped from Indeed, the job search company, which was given to us directly by the company (it’s actually illegal to scrape data from them without their express permission). Two representatives from Indeed even flew all the way over to the UK from their headquarters in Austin, Texas in order to participate in the hack! We also had various other datasets, including company reports and text files with extracted data from company gender pay gap reports, a Twitter dataset of tweets from this year with relevant gender pay gap hashtags, an 'employee happiness' survey from the Personal Group with information on the gender happiness gap in different industries, company’s House data containing details of each company's past board members, and of course the government gender pay gap dataset, which is freely available for download from GOV.UK.



We split into several groups, each group taking a different approach to the data, and working with different 'datasets'. I joined the text analytics group, and we looked at combining the Indeed job descriptions with the government gender pay gap data, and carrying out textual analysis methods to try to find out whether certain ways in which a job description was written correlated positively or negatively with the gender pay gap. For example, we wondered whether companies which explicitly stated their commitment to equality and diversity in the job description actually had a smaller gender pay gap that companies which do not. Other groups worked on modelling differences in the gender pay gap across different areas of the UK, the gender pay gap across the charities sector, whether the composition of board members influenced the gender pay gap, and one group built a chatbot which helps you to negotiate a fair salary, based on your sector and role average.

All in all, it was a super fun day, and we discovered a lot of insights surrounding the gender pay gap. Hopefully there will be more opportunities in the future to use these insights to make technology a more diverse and welcoming place. We also hope that we can use some of these insights in our own push for equality at Solirius, as we go forward with our pledge to the Tech Talent Charter.