The gender gap in STEM exists because women are lost at every stage of the pipeline, from elementary school through mid-career, and no single factor explains it. Women make up 35% of the STEM workforce overall, but that number masks much wider gaps in specific fields: just 16% of engineers and 26% of computer and math professionals are women. The causes are layered, starting with early childhood stereotypes, reinforced by educational sorting, and compounded by workplace cultures that push women out.
Stereotypes Take Root Before Second Grade
Girls as young as six are already less likely than boys to associate being “really, really smart” with members of their own gender. That finding, from research on elementary students in grades one through four, captures how early the damage begins. Children absorb cultural messages about who is naturally gifted in math and science well before they have any real evidence to draw on. These beliefs shape which subjects feel like a natural fit and which feel like an uphill climb.
The consequences compound over years. Girls who internalize the idea that math talent is something boys “have” and girls “have to work for” are less likely to take advanced math and science courses in high school. By the time college applications roll around, the gap in preparation is already measurable. Boys score higher on science tests and complete more advanced coursework, and that pre-college readiness gap alone accounts for about 35% of the overall gender difference in STEM careers.
Women Leave the Pipeline at Every Stage
A common way to think about the gender gap is the “leaky pipeline,” where women drop out of STEM at specific transition points. But research shows there is no single stage where the leak happens. Women are lost before college, during college, and after college, in roughly equal measure.
Pre-college readiness explains 35% of the career gap. During college, men are far more likely to start in a STEM major, which accounts for another 26%. And after college, male STEM graduates are more likely than female graduates to actually enter STEM jobs, explaining the remaining 41%. That last number is striking: even among people who earn the same degree, men are significantly more likely to end up working in the field.
The variation across disciplines is enormous. In biological sciences, women earn over 60% of bachelor’s and master’s degrees and more than half of doctorates. But in computer science and physics, women earned only about 21% of doctoral degrees in 2018. The gap isn’t about whether women can succeed in science broadly. It’s concentrated in specific fields with distinct cultures and expectations.
Bias in Hiring and Evaluation
Even when women have identical qualifications, they face measurable disadvantages in how they’re evaluated. In comparative studies, human evaluators consistently rated male candidates higher than female candidates by about 0.15 standard deviations. That gap widens at the top: men were nearly 7 percentage points more likely to be ranked in the top 25% of candidates, and almost 8 percentage points more likely to land in the top 10%.
This pattern holds in academic hiring too. In a well-known double-blind experiment, both male and female science faculty evaluated applications for a lab manager position. Male applicants were significantly more likely to be hired, offered a higher salary, and given more mentoring opportunities than identical female applicants. The bias wasn’t limited to male evaluators. Women faculty showed the same patterns, suggesting these preferences are baked into professional culture rather than driven by individual prejudice.
Workplace Culture Pushes Women Out
Getting hired is only part of the problem. Staying is often harder. In surveys of STEM professionals, 45% of women said they had considered quitting their current STEM job in the past two to three years, compared to 23% of men. The reasons are consistent across studies: gender-based unfair treatment from managers and peers (reported by 35% of women versus 3% of men), limited access to mentors and networking, and a persistent sense that women need to be more qualified than men for the same positions.
More than half of women in STEM reported that leadership opportunities for men come with more resources, that organizations expect women to be more qualified than men for equivalent roles, and that women lack access to mentors compared to male colleagues. Social exclusion from male-dominated teams also narrows career opportunities over time, creating a cycle where fewer women in senior roles means fewer mentors for the next generation.
Caregiving expectations add another layer. A large majority of both women (82%) and men (62%) agreed that the caregiver stereotype forces women to choose between time-intensive careers and having a family more often than men face that tradeoff. In fields that reward long, inflexible hours, this pressure disproportionately affects women’s career trajectories during the years when promotions and tenure decisions happen.
The Gender Equality Paradox
One of the more counterintuitive findings in this area is the so-called gender equality paradox. Countries with higher gender equality scores actually show larger gaps in STEM enrollment, not smaller ones. In more gender-equal nations like Norway and Finland, women are less likely to pursue STEM degrees than in countries with lower gender equality, such as Algeria or Turkey.
The leading explanation is economic. In countries where financial security is less guaranteed, STEM careers offer a reliable path to a good income, motivating both men and women to pursue them regardless of personal preference. In wealthier, more equal societies, people have more freedom to choose careers based on interest, and existing cultural patterns steer women toward other fields. The paradox suggests that removing legal and economic barriers alone won’t close the gap if the cultural forces shaping preferences remain intact.
Race Compounds the Problem
The gender gap in STEM doesn’t affect all women equally. Engineering, where women’s representation is lowest at 16%, has particularly stark disparities for women of color. Black and Latina women are underrepresented across nearly every STEM category, facing both gender and racial bias in hiring, mentoring, and advancement.
Interestingly, Black women who hold their highest degree in engineering or computer science work in STEM occupations at relatively high rates (38% and 41%, respectively), suggesting that those who make it through the pipeline are determined to stay in the field. But the pipeline itself is narrower for them from the start, with fewer resources, less mentorship, and compounding stereotypes about both race and gender shaping their experiences from childhood onward.
Spatial Skills and the “Innate Ability” Question
Some discussions of the gender gap point to cognitive differences, particularly in spatial reasoning, as a potential biological explanation. Men do, on average, outperform women on certain spatial memory tasks, likely related to differences in how visuospatial working memory operates. But the picture is more nuanced than it first appears. These gender differences only show up at specific levels of task difficulty, not across the board. At easier and harder levels, performance is similar.
More importantly, spatial skills are trainable. Early exposure to spatial play, building toys, puzzles, and navigation games narrows or eliminates these gaps. The fact that boys are more likely to be given these kinds of toys and activities circles back to the same cultural patterns that shape stereotypes. What looks like a biological difference often reflects differences in childhood experience.