Skip to main content

10 July 2019

How platforms like LinkedIn shine a light on women's under-representation in STEM

Lisa Finnegan

LISA FINNEGAN: By analysing data on the workforce of today, we can find the gender gaps in STEM and seek solutions that work.

Women in STEM
How platforms like LinkedIn shine a light on women's under-representation in STEM

This is the first in new blog series on women in STEM. We'll be sharing further contributions throughout July from speakers who gave talks at the Women in Science and Engineering Conference 2019

After giving a talk at the Women in Science and Engineering Conference (WISE) on LinkedIn as a tool to address the shortfall of women in STEM, I wanted to share a summary of my remarks  from that day with a a wider audience.

I want to begin by making one thing clear: although I work at LinkedIn, I am a woman in the rather more traditionally female industry of Human Resources, and I don’t seek to lecture anyone within STEM or outside it. However, I am proud to represent a technology company that is committed to closing the gender gap in our industry, and thanks to LinkedIn’s Economic Graph and the many remarkable data scientists who work for us, I’m able to share some insights into the problem and its potential solutions.

So, what do I mean by “Economic Graph”? I would say that it's LinkedIn’s "secret sauce", as it gives in-depth, comprehensive insight into the workforce of today. When you enter your information onto your LinkedIn profile, where you work, where you study and so on, we anonymise and aggregate this information, and add it to a digital map of the global economy, as seen by our 630 million members. Through the Economic Graph, we can build an understanding of how education affects careers, how people move from industry to industry, and increasingly how women and men’s experience of work differs. 

At the WISE conference, I offered up the Economic Graph as a tool to understand the so-called "pipeline" problem - the idea that gender equality gaps we see now, are actually the fault of some earlier stage in career development or education.

Our view of those pipelines reveals three things:

  1. Female participation in STEM is rising - with an important exception - computer science

The share of female graduates continues to rise, and the share of women graduating in STEM fields is also rising – albeit not very quickly, from 30 % to 33% in just under a decade. But in Information and Computer Technology courses, the share of women is actually declining, down from 21% to 19% over that same period. At LinkedIn we can see where those “missing” women are going - it's to the Life Sciences. Today 62% of Graduates in Life Sciences are women, up from 55% ten years ago.

We’re not sure why this is, but our own female engineers tell us there’s a vicious circle underway - too few women in the field create too few examples for others, so there’s too few women managers to show leadership, and ultimately too few teams that women would like to join.

  1. Within IT there’s a particularly important area that’s even more underrepresented: AI

Last year we did some research with the World Economic Forum, looking specifically into this area, and found that only 22% of AI professionals globally are female, compared to 78% who are male.This gap is persistent and not closing.

There are clear reasons to worry about this - the objectives we set for AI will be influenced by the biases of those that create it. Only diverse teams, trained in how to be aware of and control for their unconscious bias, can ensure that AI will serve all parts of our community.

But it’s not just the people, but also the data sets that we use to train our AIs that will introduce biases of their own. Machine learning relies on feeding an algorithm with data, having it reach conclusions, and a human confirming if those conclusions are right or wrong. If the data set isn’t diverse, we automatically introduce a bias.

  1. The influence of AI on our jobs is going to make soft skills more important, and therein lies an opportunity

Our data shows “soft skills” are more in demand now than ever before. At the WISE conference, I challenged the very term  “soft skills” - there’s nothing soft about leadership and teamwork - but whatever they are to be called, we can see human-centred occupations rising in parallel with AI tools. We look towards a future where technically proficient workers combine their knowledge of AI tools with their soft skills, to create value in many different ways.

This is an opportunity for women across all fields - because that combination of hard and soft skills has traditionally been a strength for women in business, and it’s a chance for women in science and engineering to increasingly move into leadership positions, where that combination is going to be needed.

So, how can we get things moving to bring about the changes we want to see? If you have the chance to influence the next generation, work on encouraging more young women into the ICT industry. And, if you’re a woman in STEM - consider the opportunities in computer science, and in AI in particular. Whatever you do, value your so-called soft skills and speak up for their importance now and in the future.


Lisa Finnegan is Senior Director of Human Resources for LinkedIn EMEA and LATAM, working with a team that supports 3000 employees across nine offices.