The <3% Problem: What the Pipeline Math Actually Shows
I have a party trick I pull out when I need a room to pay attention.
I ask them to do some math with me.
About 18% of baccalaureate degrees in engineering are awarded to women. Not 50%. Not anywhere close. Eighteen percent. For every hundred engineers graduating in this country, eighteen are women and eighty-two are men. We’ll start there.
Now: of those women who graduate with engineering degrees, roughly half don’t enter or don’t stay in engineering roles. Some go into adjacent fields. Some make deliberate pivots. But a significant portion leave because the profession doesn’t accommodate them: the culture, the hours, the assumption that a serious engineer has no serious obligations outside of work. So now we’re at nine percent.
Of the women who do enter the workforce as practicing engineers, 43% leave full-time STEM employment within four to seven years of having their first child. Not because they stopped being engineers. Because the system – corporate America broadly, and engineering specifically – was not built to hold them while they were also doing everything else that falls to them. The second shift. The invisible third shift. The cognitive load that runs constantly in the background and never shows up on a performance review. So now we’re at roughly four and a half percent.
And then there’s the senior level. The attrition that happens not from parenthood but from two decades of operating in environments that reward conformity to a professional standard that was built around someone who doesn’t look like you. The “chilly climates.” The unchallenged bias. The promotion that went to the man who was less qualified but more visible. The authority that has to be re-earned with every new stakeholder, every new project, every new room.
Half of what’s left.
2.25%.
That is half of a half of a half of an initial 18%. That’s what’s left of the women twenty years after graduation. And before anyone says it: this is not because the women who remain are the only ones who were tough enough. It’s because the system filters based on tolerance for structural disadvantage, not technical ability.
The Four Barriers
The research on why women leave engineering is extensive and consistent. The barriers identified across decades of study aren’t mysterious. They are not about women lacking confidence, not leaning in, or not being interested. They are structural. Here’s what they actually are.
1. Hostile and exclusionary workplace cultures. “Chilly climate” is the academic term. What it means in practice: environments where women are talked over, undermined, excluded from informal networks, and held to different standards than their male peers. More than half of female engineers report experiencing sexual harassment at work. The women who stay in these environments do so at a cost that compounds over time, and eventually, for many of them, the cost exceeds what the career is worth.
2. Inadequate support for caregiving responsibilities. Engineering has historically been one of the least tolerant professions for caregiving responsibilities of any field studied. Demanding hours, expectations of constant availability, and the absence of flexible structures don’t push out women because women can’t handle demanding work. They push out women because women are carrying a second and third shift that men in the same roles typically are not. The math stops working. I know this math personally. I have done it many times.
3. Lack of visible authority and role models. When you can’t see it, it’s hard to build toward it. The absence of senior women in engineering isn’t just a representation problem, it’s an ongoing signal to every mid-career woman in the field about what the ceiling actually looks like. Visible female leadership matters for retention, not just recruitment. And organizations that actively authorize their senior women engineers – not just title them, but publicly back them – retain them at higher rates.
4. Absence of sponsorship and equitable promotion pathways. High-potential women are overmentored and undersponsored. The research on this has been consistent since at least 2010 and nothing meaningful has changed. Mentoring provides guidance. Sponsorship provides access: to the stretch assignment, the high-visibility project, the promotion conversation. Without it, women develop the skills and don’t get the opportunities. The pipeline doesn’t just leak. It has a valve, and the valve is who has a sponsor.
Why This Matters Beyond the Numbers
I want to be precise about something.
I am not asking anyone to feel bad about these statistics. I am asking organizations to recognize them as operational data: information about a system that is underperforming relative to its inputs.
You are graduating engineers. They are leaving. The ones who leave are disproportionately women, and they are leaving at predictable points in their careers for predictable and addressable reasons. That is not a diversity problem. That is a systems problem. And systems problems have engineering solutions.
What those solutions look like, specifically, in pharmaceutical and life sciences, in manufacturing, in the organizations doing the work that keeps medicines safe, is what I’m writing about next week.
June 23 is INWED (and also the date I am traveling to Phoenix for the EPIC National Pageant). The final post in this arc goes live then.
See you there.
Rebecca, PE
Founder, PT&P Solutions LLC | Mrs. Florida Emissary 2026 | Co-host, The Athena Factor
This is the second post in the INWED 2026 arc.
[Post 1: The Hard Hat and the Sparkly Hat Are the Same Tool]
[INWED June 23: What “Too Much” Costs.]
