Retention and the Cross-Generational Pipeline
November 01, 2017
A few years ago, Ben Horowitz gave a talk called The Future of Humankind is Dependent on Technovation Girls. Horowitz claims that "if you educate a girl in the developing world, you educate five people – on average...because if you educate one girl, then she will educate at least four other people through the course of her life" [1
]. He believes the same is true for the technology industry; that if you teach one girl computer programming, she will teach others and the ratio of women in the field will improve.
This is a nice idea, but what if the exact opposite is happening?
What if women are leaving the tech industry before they can educate other women? Or, worse, what if women in the tech industry are so unhappy that they are actively convincing their friends and daughters NOT to join.
This is exactly what's been happening in our field. The attrition of women from the technology industry, namely software engineering, over the past few decades is a significant contributor to the lack of women in the field today. And, while my research focuses mainly on women, I believe many of the same things are true for people of color. Let me explain why women are leaving, how that impacts the next generation's pipeline, and how we can go about fixing it.
Women are Leaving Tech
Women are leaving tech earlier and at significantly higher rates than men. As Rachel Thomas says in her blog post from 2015, “If you think women in tech is just a pipeline problem, you haven’t been paying attention” [2
]. Attrition is one of the main reasons why we have such a gender discrepancy in our field, and I'm already starting to see this in my generation.
Currently, my female peers are opting out of the engineering workforce at alarming rates. They're pursuing career paths other than software engineering or engineering leadership because they see no future at their current jobs. By the time women reach the mid-level point in their technology career, 56% of them will have left - twice the rate that men are leaving. What's even more depressing is where those women are going. These women aren’t just leaving your company, they’re leaving the industry. Most of them are leaving for jobs in other industries and taking a pay cut to do so [3
]. Despite the myth that women leave the workplace to have children, only 1 in 4 women who leave cite "spending time with their family" [4
]. Instead, over half the women who leave cite bad work environments, a lack of flexibility, and no advancement opportunities as their reason for leaving.
There's even been a rise of female entrepreneurs starting companies because they are fed up with the existing tech work cultures. Imagine that, women find starting a company to be less emotionally taxing than simply working as an engineer at existing companies.
And if women are leaving tech, then they can't do what Ben Horowitz hopes they will: educate and bring 4 other women into the field during the course of their careers.
The Pipeline Won't Solve Our Problems
Attrition has an effect on the next generation's pipeline. Much of the focus on diversity has been about getting women and people of color into the industry, but that focus misses many of the major problems that exist today. Blaming the pipeline and waiting 30 years for the next generation to solve gender diversity is irresponsible at best and a malicious avoidance of responsibility at worst. It ignores the fact that women are connected across generations, and that one of the main reasons we don't have a pipeline today is because we didn't retain women in the last generation. And one of the most significant things that companies and managers can do to improve diversity in our industry is to retain the women and people of color they work with today.
Cross Generational Retention
In college, I did original research on how and why women got started in computer programming for my honors thesis called Women in Computer Science
. One of my key findings is that most women in computer programming are "First Generation Women." These are women convinced to go into the field by a man; a father, brother, friend, or significant other who is male. However, there were shockingly few "Second Generation Women;" women who were convinced to go into the field by a mother, sister, friend, or significant other who is female. Given that the percentage of women in Computer Science in the 80s was 35%, more than double what it is today, why are there so few women convinced to go into the field by a woman [5
]? Let's take a look at how this can happen. While we might have had a higher percentage of women in the field in the 80s, if the work environments were toxic and they all left unhappy, then they wouldn't tell their friends, daughters, and nieces to go into the field. A friend of mine is one of the few second-generation women that I know, and her mom was a programmer during the 80s. However, fed up with the industry, she left at 29 to teach calculus at a local college. Flash forward to my friend, who went into the field despite her mother's warnings, and now at 30, she is considering opting out of the field as well.
Women significantly impact other women
There is a growing body of evidence that the race and gender of instructors and mentors significantly impact the performance and retention of their students [6
]. I.e. a black professor has a significant and positive impact on the performance of her black students. Which means that the happiness of a previous generation in a field can significantly impact the next generation - in either direction. I found this to be true in a survey I recently conducted asking people in my network: "How did you get into computer programming?". Over 500 people responded, and what I found was there were relatively few differences in the reasons why women and men entered the field. People come in through school, gaming, encouragement from teachers, and a general interest in technology. However, one of the few statistically significant differences is that women are much more likely to be influenced to go into the field by a woman
than men are [7
In other words, if women have a more powerful impact on other women, then the retention of women today will impact the pipeline of women tomorrow.
How do we get more second-generation women?
It's simple: we retain the women we already have.
Now, you might think there's some sort of magic trick to retaining women in our industry. Interestingly enough, the reasons that women leave are roughly the same as men. Getting more second generation women is not as difficult as you would think. It's the same formula that companies currently use to try to retain their male developers. There are three major things that ALL people want in their job: good management, upward mobility, and benefits and job flexibility.
Bad managers are the #1 reason people leave a company [8
]. Poor management is an epidemic in the tech industry (more on that some other time). The first, most basic thing companies can do to improve management is to train their people managers. Management, like programming, is a skill that requires practice. If you don’t know where to start with your training, Google did a research study on what makes an effective manager
. While Google itself is not a pinnacle of good people management in the technology industry, the research here is very thorough.* *Here's the short summary of Google's findings .
People want upward mobility. No one is going to stay at a company where they have no future, and currently, women are promoted at drastically lower rates than their male peers. My next post is entirely dedicated to this topic so I won’t go into a lot of depth here, but it is essential to have a clear, well-documented promotion process. Having no process for promotions is unacceptable (see my post on The Null Process) because a lack of documentation and process allows biases to run rampant. No matter how imperfect your system for promotions, having a process that is documented and accessible to every member of your team is critical to making promotions more egalitarian.
Life is crazy and unexpected and full of so many things other than work. Friends, family, children, significant others. Every person at your company has something they are dealing with outside of work. Benefits that help employees stay healthy and give them the flexibility to deal with life are a huge component of employee happiness and retention.
Women, for example, can’t help being the gender that bears children. Companies should be able to support the perpetuation of the human race while still retaining valuable female employees [11
]. Parental leave of 4 months should be an industry standard, and companies like Vodafone are even adding on transition plans for new mothers after they return from maternity leave [12
]. Similarly, many people will have an aging parent at some point, so having a different kind of parental leave will be necessary so people can take care of their families in times of need [13
Ben Horowitz wasn’t wrong that the future of humankind is dependent on Technovation girls. The thing he missed is that it’s not just about educating the next generation; the single most important thing you can do for diversity is to retain the current generation. You can retain any employee by valuing them equal to their work, training managers at your company, and giving employees the benefits and work flexibility to manage life outside the office. I wish we valued retaining underrepresented groups as much as we valued hiring them. The future of the technology industry depends on us retaining the women and people of color we already have.
- Horowitz, Ben. Ben Horowitz: The future of humankind is dependent on Technovation girls. Iridescent, youtube.com, 2012.
- Thomas, Rachel. If you think women in tech is just a pipeline problem, you haven't been paying attention.. Medium, 2015
- Ashcraft, Catherine and Sarah Blithe. Women in IT: The Facts. National Center for Women & Information Technology (NCWIT), 2010.
- Fouad, Nadia. Women Engineers: A National Study of Attrition and Persistence. studyofwork.com, 2011.
- Planet Money. When Women Stopped Coding. NPR, 2014.
- Gershenson, Seth. A law school instructor like me: How a professor’s race and gender can impact student performance. brookings.edu, 2017
- Heddleston, Katherine. How did you get into computer programming? (raw data). Google sheets, 2016.
- Schwantes, Marcel. Why Do People Quit Their Jobs, Exactly? Here's the Entire Reason, Summed Up in 1 Sentence. inc.com, 2017.
- Bryant, Adam. Google’s Quest to Build a Better Boss. NyTimes, 2011.
- Scheider, Michael. Google Employees Weighed In on What Makes a Highly Effective Manager (Technical Expertise Came in Last). inc.com, 2017.
- Florentine, Sharon. Lack of Parental Leave Drives Employee Turnover. cio.com, 2014.
- McGregor, Jena. An unusual new policy for working mothers. WashingtonPost.com, 2014.
- Sloan Center on Aging and Work. EMPLOYERS NEED » INCREASE RETENTION AND REDUCE TURNOVER. Boston College.
Want emails notifying you of new content? Subscribe here.