America Has a Huge STEM Workforce. Women Still Enter Its Core Tech Jobs at a Much Lower Rate

America does not have a small technology labour market. The National Science Foundation estimates that 37 million people worked in STEM occupations in 2024, around 26% of the US workforce. STEM workers also had higher median annual earnings and lower unemployment than workers outside STEM.

Yet the opportunity is distributed unevenly. Bureau of Labor Statistics analysis of workers aged 25 to 54 shows men consistently more likely to work in computer and mathematical occupations. Among workers aged 25 to 34 and 35 to 44, roughly 7% of employed men were in computer occupations, compared with around 3% of employed women in the same age groups.

The disparity survives across age groups

The pattern is not confined to one generation. Among workers aged 45 to 54, BLS data put 5.8% of employed men in computer occupations compared with 1.9% of employed women. Architecture and engineering show a similar direction, with roughly 3% to 4% of employed men aged 25 to 54 in those occupations compared with around 1% of women.

Those figures do not show discrimination by an individual employer. They show occupational sorting across the economy. The business question is why companies competing aggressively for technical talent still draw from a labour pool in which women are much less represented in several of the roles with the strongest links to software, data and engineering investment.

Hiring is only one point in the funnel

Employers often focus diversity efforts on the top of the recruitment funnel: more applicants, broader sourcing and changes to job descriptions. Those interventions can matter, but they cannot explain whether women who are hired remain in technical tracks, receive high-impact assignments, move into management or return successfully after career breaks.

A more useful corporate dashboard follows the whole funnel. Applicant-to-interview conversion, offer rates, starting pay, promotion rates, voluntary attrition and representation at senior technical levels reveal whether the constraint sits in sourcing, selection or progression.

The opportunity cost rises as technical wages rise

BLS occupational wage data illustrate why representation in technology matters beyond headcount. Software developers had mean annual wages of about $148,100 in May 2025, while data scientists averaged about $126,800. These are occupational averages rather than gender-specific salaries, but they show the size of the income pools attached to technical work.

When women are less likely to enter those occupations, the effect can accumulate through earnings, retirement saving, equity compensation and later access to management or entrepreneurship. For the economy, it also means a smaller effective talent pool in occupations that companies repeatedly describe as difficult to fill.

Our view: the US tech gap is increasingly a capacity problem

American Commerce Review's view is that employers should treat women's underrepresentation in core technical occupations as a labour-capacity issue, not simply a corporate reporting category. The US is investing heavily in AI, semiconductors, data centres and advanced manufacturing while simultaneously competing for engineers, developers and data specialists.

Expanding the pool of women entering and remaining in these careers will not solve every skills shortage. But leaving a persistent participation gap untouched while companies compete for scarce technical labour is economically difficult to justify.