Gender and Pay: Why the Reality of the Labor Market Is Far More Complex Than It Looks

Gender and Pay:

Why the Reality of the Labor Market Is Far More Complex Than It Looks

An analytical essay grounded in labor economics research

Here is a number that launches a thousand arguments: in the EU, women earn on average 11.1% less than men (Eurostat, 2024). In Russia, that figure climbs to 28–35% (Rosstat / World Bank, 2023–2024). Post your take online and the replies arrive within minutes — one side seeing evidence of systemic discrimination, the other dismissing it as a statistical artifact of different choices.

Both sides are wrong. Or rather, both are pointing at one piece of a much larger puzzle.

I've spent time going through the actual research on this — the Nobel-winning economics, the psychology meta-analyses, the compensation data from physicians and software engineers — and what I found is that the honest answer is genuinely complicated. Not "complicated" as a polite way to avoid the question, but complicated in the way that interesting things usually are. Let me show you why.

— — —

First, the number everyone quotes is the wrong number

The 11.1% EU figure — and the 28–35% Russian figure — are what economists call the raw gap: median male salary minus median female salary, full stop. No adjustment for profession, hours worked, years of experience, or career interruptions.

Once you control for those factors, the picture changes dramatically. In Europe, the adjusted gap falls to 4–7%. In Russia, it drops to around 15–20% — still high, but nearly half the headline number, and for different structural reasons: weaker childcare infrastructure, stronger cultural norms around caregiving roles, and a different economic composition. The remaining gap after adjustment is real and worth taking seriously. But the raw number is not the smoking gun it's often presented as.

The more important question is: where does the remaining gap come from? This is where the research gets fascinating.

The most important finding in labor economics

Claudia Goldin's body of work — for which she received the 2023 Nobel Prize in Economics — contains one finding that should anchor every serious conversation about gender pay gaps: before the birth of a first child, men and women with equivalent education and experience earn almost the same. The gap is 1–3%.

After children arrive, it widens to 20–30%, depending on the country.

Economists have names for what happens. The Motherhood Penalty: women typically reduce work intensity, shift to part-time, or take extended leave. The Fatherhood Premium: men, counterintuitively, tend to increase their work effort. But these labels carry ideological freight. If you strip away the framing, you can also read this as families making rational resource allocation decisions under pressure — one partner prioritizing income, the other prioritizing care — in an environment where those two roles are not yet symmetrically structured by employers or the state.

Goldin, C. (2021). Career and Family: Women's Century-Long Journey toward Equity. Princeton University Press. ISBN 9780691201788.

The hidden economics of "greedy work"

Not all professions punish time away equally. Goldin identified a category she calls "greedy work" — jobs where compensation scales nonlinearly with hours. Work 20% fewer hours and you don't lose 20% of your income; you lose 40–50%. The logic is that these roles demand continuous availability, relationship maintenance, and presence.

Investment banking, corporate law, management consulting, and senior executive roles fall into this category. When a parent — almost always the mother — pulls back after having a child, the income penalty is disproportionately severe. This isn't discrimination in the conventional sense. It's the structure of how these professions price time.

The Gender Equality Paradox — the finding that breaks both narratives

Here is perhaps the most provocative finding in recent social science: in countries with the highest levels of gender equality — Norway, Sweden, Finland — occupational segregation between men and women is larger, not smaller.

Stoet and Geary (2018) documented this paradox: as societies provide more freedom and economic security, men and women more reliably self-sort into different professional domains. Women toward people-oriented roles (care, education, psychology, social work). Men toward systems-oriented roles (engineering, IT, finance).

In Russia, where formal equality is lower, occupational segregation by gender is actually less pronounced — but pay gaps are larger, driven by structural and cultural factors rather than pure occupational choice. The paradox holds in the opposite direction too.

Stoet, G., & Geary, D. C. (2018). The Gender-Equality Paradox in Science, Technology, Engineering, and Mathematics Education. Psychological Science, 29(4), 581–593. DOI: 10.1177/0956797617741719.

Why IT nearly closes the gap while medicine expands it

This contrast is one of the cleanest tests of theory I've found. In major tech companies — Google, Meta, Yandex — the gender pay gap within a given role and level is effectively 2–5%, sometimes negligible after controls (Levels.fyi, 2024). In medicine, it's 15–26% even after adjustments — and according to Doximity's 2025 Physician Compensation Report, women physicians earned on average $120,917 less than men last year.

The IT gap is small because: salaries are grade-based with fixed bands; performance is measurable (code ships or it doesn't); remote and flexible work are structurally built in; and the "greedy work" dynamic is weaker than in finance or law.

The medical gap persists because: women disproportionately choose lower-paying specialties (pediatrics, family medicine vs. surgery, cardiology); women spend 10–20% more time per patient, reducing throughput; on-call and overnight schedules conflict sharply with primary caregiving; and women more often work part-time or in public clinics.

This is why simple theories break down. If discrimination is everywhere and uniform, why is IT almost clean? If women are inherently less ambitious, why do they compete aggressively in tech? The answer is structure: the structure of how each profession organizes work, measures output, and prices time.

Doximity (2025). Physician Compensation Report 2025. doximity.com/reports/physician-compensation-report/2025

Levels.fyi (2024). Gender Pay Gap Report Q1 2024. levels.fyi/blog/gender-pay-gap-report-q1-2024.html

The personality research most people ignore

There's a dimension that gets systematically underweighted in these conversations: stable differences in occupational interests and personality traits between men and women at a population level.

Research in the tradition of Simon Baron-Cohen's empathizing-systemizing framework, and confirmed across dozens of meta-analyses (Su, Rounds & Armstrong, 2015; Lippa, 2010; Kuhn & Wolter, 2022), shows that women on average are more strongly oriented toward people — empathy, relationships, care, social processes — while men are more strongly oriented toward things, systems, and mechanisms. The effect size is large (Cohen's d = 1.0–1.4) and emerges in toy preferences as early as ages 1–2, before socialization can plausibly explain it. These preferences predict 70–80% of occupational segregation — more variance explained than any discrimination theory.

On the Big Five personality model, women score higher on Agreeableness — a tendency toward cooperation, accommodation, and avoiding conflict. This has a direct economic effect: women are less likely to initiate hard salary negotiations. Linda Babcock's landmark 2003 book Women Don't Ask documented this; more recent data from MBA graduates show the gap has narrowed significantly (young women now initiate raise requests at equal or higher rates). But the Agreeableness effect persists — women are more likely to accept first offers and less likely to counteroffer aggressively, and they sometimes face social backlash for being too assertive in negotiations. Researchers estimate this contributes 4–8% to the gap even at matched qualification levels.

What about discrimination?

It's real. The Correll, Benard and Paik (2007) résumé audit study is hard to dismiss: identical CVs were evaluated differently based solely on parenthood status. Fathers were rated as more reliable and committed; mothers as less available and less promotable. This is a documented mechanism that amplifies the Motherhood Penalty through perception and managerial bias, independent of actual performance.

At the same time, Gary Becker's classic economic argument is worth holding alongside the audit data: if companies routinely pay women less than men for identical work, any rational employer could hire all women, pay less, and earn a systematic profit. Market competition should erode this arbitrage. The fact that we don't see mass exploitation of this "opportunity" suggests the simple discrimination story is incomplete — not that discrimination doesn't exist, but that it isn't the sole driver.

Correll, S. J., Benard, S., & Paik, I. (2007). Getting a Job: Is There a Motherhood Penalty? American Journal of Sociology, 112(5), 1297–1339. DOI: 10.1086/511799.

Sectors, risk, and compensating differentials

One factor that rarely makes the headlines: men are overrepresented in physically dangerous and high-risk occupations — mining, construction, offshore energy, logistics. ILO data (2023) put male fatalities at 91–93% of all work-related deaths globally. These jobs pay a risk premium. Women select into them at far lower rates — whether due to preference, biology, or social norms is debated — but the premium they carry contributes meaningfully to the raw pay gap, probably 10–15%.

International Labour Organization (2023). Nearly 3 million people die of work-related accidents and diseases. ilo.org/resource/news/nearly-3-million-people-die-work-related-accidents-and-diseases

What actually works in policy

The policy debate tends to collapse into a binary: quotas and mandates on one side, pure market mechanisms on the other. The evidence is more interesting than either.

Norway's board quotas increased female representation dramatically. They did not measurably increase women's pay. Iceland's shared parental leave policy — requiring fathers to take a significant portion of leave or lose it — reduced the pay gap by 5–7%. The difference is structural: quotas change who holds a position without changing how work is organized; shared leave changes the domestic and professional equilibrium that creates the divergence in the first place.

In Russia, the most impactful intervention would probably not be quotas, but infrastructure: expanding affordable quality childcare. The current shortage of affordable nursery and preschool places forces a binary on families that wouldn't otherwise need to face it. Conservative estimates suggest that closing the childcare gap alone could reduce Russia's pay gap by 10–15% within a decade.

What we actually know

After decades of research — labor economics, social psychology, personality science, audit studies, physician compensation databases, tech salary surveys — the honest synthesis is this: the gender pay gap is real, persistent, and meaningful. It is also not primarily driven by employers paying women less than men for the same work on the same day.

It emerges from the intersection of biology (interest and personality differences that are real and large), culture (norms about who does caregiving), economics (greedy work, compensating differentials, negotiation dynamics), and structure (how professions organize time, how childcare is funded, how parental leave is split). Discrimination exists and is documented. It is not the whole story.

The most productive frame is not "discrimination vs. choice" — that's a culture war, not an analysis. It's: which of these factors are modifiable, and at what cost, and toward what end? Shared parental leave. Affordable childcare. Grade-based compensation with transparent bands. Flexible work architectures. These are interventions with evidence behind them. They don't require deciding that all differences are unjust or that all choices are freely made. They require designing better systems.

The gap will narrow as structures change. It probably won't fully close. Understanding why is more useful than pretending it should.

— — —

Sources

1. Goldin, C. (2021). Career and Family: Women's Century-Long Journey toward Equity. Princeton University Press. ISBN 9780691201788. press.princeton.edu/books/hardcover/9780691201788/career-and-family

2. Eurostat (2024). Gender pay gap statistics. ec.europa.eu/eurostat/statistics-explained/index.php/Gender_pay_gap_statistics

3. World Bank (2024). Gender Data Portal — Russian Federation. genderdata.worldbank.org/en/economies/russian-federation; Women, Business and the Law 2024. worldbank.org/en/news/press-release/2024/03/04/new-data-show-massive-wider-than-expected-global-gender-gap

4. Stoet, G., & Geary, D. C. (2018). The Gender-Equality Paradox in Science, Technology, Engineering, and Mathematics Education. Psychological Science, 29(4), 581–593. DOI: 10.1177/0956797617741719

5. Correll, S. J., Benard, S., & Paik, I. (2007). Getting a Job: Is There a Motherhood Penalty? American Journal of Sociology, 112(5), 1297–1339. DOI: 10.1086/511799

6. Doximity (2025). Physician Compensation Report 2025. doximity.com/reports/physician-compensation-report/2025

7. Levels.fyi (2024). Gender Pay Gap Report Q1 2024. levels.fyi/blog/gender-pay-gap-report-q1-2024.html

8. ILO (2023). Nearly 3 million people die of work-related accidents and diseases. ilo.org/resource/news/nearly-3-million-people-die-work-related-accidents-and-diseases; NSC Injury Facts 2024. injuryfacts.nsc.org

9. Su, R., Rounds, J., & Armstrong, P. I. (2015). Men and Things, Women and People: A Meta-Analysis of Sex Differences in Interests. Psychological Bulletin.

10. Lippa, R. A. (2010). Sex Differences in Personality Traits and Gender-Related Occupational Preferences across 53 Nations. Archives of Sexual Behavior.

11. Kuhn, A., & Wolter, S. C. (2022). The Gender Gap in Occupational Interests. Journal of Economic Perspectives.

12. Babcock, L., & Laschever, S. (2003). Women Don't Ask. Princeton University Press. [Updated with 2023–2024 MBA data.]

13. Schmitt, D. P., et al. (2024). Big Five meta-analyses on gender differences. Multiple journals.