The Career Gap Mystery

What the science actually says about who ends up where — and what it means for your coaching and DEI strategy

Download the free ebook: TruMind.ai/diversity

The Finding That Should Stop Every DEI Strategist Cold

Here is the most inconvenient fact in the gender-and-work literature: in the most gender-equal societies on earth — Scandinavia, the Netherlands, Finland — the occupational gap between men and women is larger, not smaller, than in less egalitarian countries. When economic pressure to take any available job is removed and individuals are free to follow their genuine preferences, men and women sort into different careers in greater, not lesser, proportions (Stoet & Geary, 2018, as cited in MacDonald et al., 2023).

If occupational gaps were primarily a product of structural coercion, they should shrink as structures equalize. They do not. This is not a minor anomaly. It is a signal that demands a better theory.

When women are free to choose, they choose differently from men — on average. That is not a problem to be solved. It is a fact to be understood.

The Puzzle That Refuses to Go Away

You have been here before. You open the corporate diversity report, note that women are underrepresented in engineering, quantitative finance, and technology leadership — and that men remain scarce in nursing, primary education, and social work. You greenlight another pipeline initiative. The gap persists.

Two data points sharpen the puzzle. A 2019 national survey of Americans' civic knowledge showed men outperforming women in every single state by a substantial margin on factual political and governmental content. A 2001 general intelligence study found men scoring higher on knowledge across every measured domain — with one striking exception: fashion knowledge, where the gap disappeared entirely. These are not anomalies. They fit a pattern the cognitive science literature has been documenting — and struggling to discuss honestly — for more than two decades.

Baron-Cohen's Framework: Empathizing vs. Systemizing

In 2002, Cambridge psychologist Simon Baron-Cohen introduced the Empathizing–Systemizing (E-S) theory to explain cognitive sex differences and the profile of autism spectrum conditions (Baron-Cohen, 2002). The theory proposes two orthogonal drives:

Empathizing is the drive to read minds, interpret emotions, navigate social dynamics — to respond to people with care and sensitivity (Baron-Cohen, 2003). Systemizing is the drive to extract the underlying rules that govern systems — mechanical, legal, financial, musical, civic (Baron-Cohen, 2009). Civic structures are systems in exactly this sense: constitutional rules, legislative procedures, governmental hierarchies. Fashion is not — it is social, contextual, trend-driven. The null result on fashion knowledge is the theoretical signature that distinguishes the systemizing account from a simple ability hypothesis.

Baron-Cohen identifies five brain types — Extreme E, Type E, Balanced B, Type S, Extreme S — with more females averaging toward Type E and more males toward Type S. The distribution is driven in part by fetal testosterone, which correlates positively with systemizing and negatively with empathizing (Baron-Cohen, 2010).

What Half a Million People Tell Us

The E-S theory is no longer a small-sample hypothesis. Greenberg et al. (2018) tested ten E-S predictions across 671,000+ individuals. Typical females averaged higher on empathy measures; typical males averaged higher on systemizing measures. Brain-type D-scores (the difference between EQ and SQ) accounted for 19 times more variance in autistic traits than biological sex alone — underscoring that cognitive style is the real organizing construct, not sex per se (Greenberg et al., 2018).

The vocational interest literature converges on the same conclusion. Su, Rounds, and Armstrong's (2009) meta-analysis of 503,188 respondents found men prefer working with things and women with people, producing a large effect size of d = 0.93. Men showed stronger Realistic (d = 0.84) and engineering (d = 1.11) interests; women showed stronger Social (d = −0.68) interests. Su and Rounds (2015) demonstrated that the Things-People dimension explains the gender composition gradient across every STEM sub-discipline, from engineering (fewest women) to social sciences (most women).

Critically, these are interests — not abilities. A woman with low Realistic interest is not less capable in a Realistic role if she chooses to enter one. But she is less likely to choose it, to persist in it without additional support, and to find it intrinsically motivating. This distinction matters enormously for how organizations interpret gap data.

Kaleidoscope Liberty: The n=1 Imperative

None of the above says anything determinate about the woman in front of you. Population distributions describe base rates; they do not assign individuals to categories. This is the central argument of Diversity Reboot.

Meet Kaleidoscope Liberty. She is a genuine Type S — a high-scoring systemizer who is keen on quantitative analysis, rule-extraction, and technical problem-solving. She belongs in a technical leadership role, not because diversity metrics require it, but because it matches her actual profile. She deserves coaching calibrated to who she is: rigorous, analytically demanding, and fully supportive of her atypical-for-women trajectory.

Support Kaleidoscope Liberty fully. Develop her rigorously. And do not be surprised when she is rare.

The Kaleidoscope Liberty framework names the ethical obligation: treat every individual as a unique configuration of traits, not as an exemplar of their demographic category. An organization that coaches a high-systematizing woman as if she needs confidence encouragement toward technical work is applying group-average stereotypes to an individual case — the precise error that rigorous psychometric assessment is designed to prevent.

What This Means for Diversity Interventions

Two diagnostic errors plague most organizational diversity strategies:

Pipeline-only thinking treats the gap as purely a supply constraint — if only we recruited more women into engineering. This swims against a strong psychological current. It may produce short-term demographic shifts while generating friction, disengagement, and attrition for individuals placed in roles incongruent with their intrinsic drives (Holland, 1997, as cited in MacDonald et al., 2023).

Discrimination-only attribution overstates demand-side barriers and understates supply-side interest differences. Vocational interests explain substantial variance in career choices. An audit built on discrimination-as-primary-cause will systematically misidentify the problem — and therefore design ineffective interventions.

The honest alternative is harder: acknowledge population distributions, invest in individual-level assessment, and let each person's actual profile drive their development trajectory.

3 Things to Do Differently on Monday Morning

Measurement at the Level That Matters

Population statistics are irrelevant to the individual in the coaching chair. What matters is that person's actual Empathizing and Systemizing profile, their Holland RIASEC interest configuration, their motivational substrate for knowledge acquisition. TruMind.ai's Rasch-calibrated scoring of leadership dimensions and ICF coaching competencies from session transcripts was built on exactly this premise: n=1 precision that group-level demographics cannot substitute for.

A coaching platform that treats a high-systematizing woman the same as a high-empathizing woman is not being equitable. It is being lazy.

Honesty as the Precondition for Effectiveness

The career gap is not a mystery once you are willing to read the full evidence base. It reflects, in substantial part, a robust, cross-culturally replicated, partially biologically-grounded difference in the distribution of systematizing and empathizing drives — a difference that expresses itself through vocational interests, knowledge acquisition, and occupational self-selection.

That evidence does not justify demographic discrimination. It does not preclude individual variation. It does not release organizations from the obligation to develop every person to their fullest capacity.

What it does do is set the intellectual precondition for diversity strategies that actually work: honest diagnosis, individual-level assessment, and interventions calibrated to the real distribution of human difference rather than a politically convenient version of it.

Download Diversity Reboot — Free | TruMind.ai/diversity

References

Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in Cognitive Sciences, 6(6), 248–254. https://doi.org/10.1016/S1364-6613(02)01904-6

Baron-Cohen, S. (2003). The essential difference: Men, women and the extreme male brain. Penguin.

Baron-Cohen, S. (2009). Autism: The empathizing–systemizing (E-S) theory. Annals of the New York Academy of Sciences, 1156, 68–80. https://doi.org/10.1111/j.1749-6632.2009.04467.x

Baron-Cohen, S. (2010). Empathizing, systemizing, and the extreme male brain theory of autism. Progress in Brain Research, 186, 167–175. https://doi.org/10.1016/B978-0-444-53630-3.00011-7

Barney, M. (2024). Diversity reboot. TruMind Press. https://trumind.ai/diversity

Greenberg, D. M., Warrier, V., Allison, C., & Baron-Cohen, S. (2018). Testing the empathizing–systemizing theory of sex differences and the extreme male brain theory of autism in half a million people. Proceedings of the National Academy of Sciences, 115(48), 12152–12157. https://doi.org/10.1073/pnas.1811032115

Holland, J. L. (1997). Making vocational choices: A theory of vocational personalities and work environments (3rd ed.). Psychological Assessment Resources.

Kidron, R., Kaganovskiy, L., & Baron-Cohen, S. (2018). Empathizing–systemizing cognitive styles: Effects of sex and academic degree. PLOS ONE, 13(3), e0194515. https://doi.org/10.1371/journal.pone.0194515

MacDonald, K. B., Benson, A., Sakaluk, J. K., & Schermer, J. A. (2023). Pre-occupation: A meta-analysis and meta-regression of gender differences in adolescent vocational interests. Journal of Career Development, 50(2). https://doi.org/10.1177/10690727221148717

Su, R., Rounds, J., & Armstrong, P. I. (2009). Men and things, women and people: A meta-analysis of sex differences in interests. Psychological Bulletin, 135(6), 859–884. https://doi.org/10.1037/a0017364

Su, R., & Rounds, J. (2015). All STEM fields are not created equal: People and things interests explain gender disparities across STEM fields. Frontiers in Psychology, 6, 189. https://doi.org/10.3389/fpsyg.2015.00189

Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in spatial abilities: A meta-analysis and consideration of critical variables. Psychological Bulletin, 117(2), 250–270.

Wakabayashi, A., Baron-Cohen, S., Uchiyama, T., Yoshida, Y., Kuroda, M., & Wheelwright, S. (2007). Empathizing and systemizing in adults with and without autism spectrum conditions: Cross-cultural stability. Journal of Autism and Developmental Disorders, 37, 1823–1832.