Overview

This chapter compares gender gaps at college entry, STEM participation, and pay in the English study, then turns to the OECD’s 2022–23 survey of US adults aged 16–65. In the English estimates, including both skills reduces the STEM gap by roughly one-third but leaves the reported pay gap slightly larger; stronger verbal skill and field sorting work in opposite directions. The US data associate numeracy with labour-force participation and wages and show that over-qualification and field mismatch are linked to lower pay and life satisfaction. These are different, adjusted associations, not one causal story.

Learning outcomes

Gender gaps in college entry, STEM participation, and pay respond differently to the same measured skills.
Including both skills reduces the reported English STEM gap by roughly one-third but does not close the pay gap.
In US adults, skill value is associated with job use: over-qualification and field mismatch coincide with lower pay or wellbeing.

Keywords

math–verbal dividegender gapsSTEMnumeracymismatchwellbeing

When Skills Meet Gender, Adult Life, and Jobs

The math–verbal divide becomes most useful when it is applied to unequal pathways rather than reduced to a difference in average scores. This chapter examines how the two skills relate to gender gaps in college entry, STEM participation, and pay, then extends the question into adult life through US evidence on proficiency, wages, wellbeing, and job mismatch. The practical judgment is that the value of a skill depends on the outcome being sought and on whether education and work give that skill a chance to be used.

That perspective changes what counts as an explanation. A gender gap in college attendance is not the same as a gender gap in STEM, and neither is the same as a pay gap. Likewise, a high adult proficiency level does not automatically become a high wage if a person’s qualification or field does not match the job. The pathways identified in the English cohort study and the broader adult patterns reported by the OECD point in the same analytical direction, but they remain different kinds of evidence.

Gender gaps are different at different stages

The English study examines three margins separately: college enrolment, undergraduate STEM participation among college entrants, and log pay. In the baseline specification, women are 7.5 percentage points more likely than men to enrol in college. When verbal skill is added to the comparison, the female coefficient changes to a 3.1-point disadvantage. Adding mathematics skill alone changes the enrolment difference little; including both skills leaves a small female coefficient of −0.6 points.

The pattern suggests that women’s stronger verbal performance accounts for much of their observed college-enrolment advantage in the study’s design. But it does not mean verbal skill is the only cause of college access, or that gender differences in schooling can be read directly from one score. The regression accounts for the measured skills and listed demographic and school factors; it does not exhaust differences in resources, expectations, institutions, family circumstances, or later decisions.

The STEM margin looks different. The raw female disadvantage is about 19.9 percentage points. Controlling for verbal skill alone leaves it at about 19.4 points, and controlling for mathematics alone leaves it at about 19.2 points. When both are included, the gap falls to 13.3 points—a reduction of roughly one-third. The combined skill measures therefore explain a meaningful part of the observed STEM gap in this specification, even though neither skill by itself accounts for much of the difference.

Why does the joint comparison matter? STEM participation is not determined only by absolute mathematics performance. A student’s relative strength in verbal work can make HASS or other alternatives more attractive, while mathematical performance can make STEM a more compatible choice. When both domains are considered, the model captures more of that comparative profile. This is not the same as saying that women who do not choose STEM lack the ability to succeed there, or that the remaining gap is unexplained in any broad sense. It says that measured differences in the two skills are part of the sorting pathway.

Pay brings a third pattern. The raw female log-pay gap is −0.356.

Published by DeepMox

FAQ

Why do the English college and STEM gaps respond differently to skills?

They respond differently because college entry, STEM participation, and pay are separate outcome margins, and the joint skill profile captures different sorting relationships at each stage.

What does the US numeracy association with wages show?

Among employed US adults, a one-standard-deviation increase in numeracy is associated with 11% higher wages after adjustments; it is not a causal return.

How does job mismatch affect the value of a qualification?

For US workers, over-qualification is associated with 19% lower average pay and a 12-percentage-point lower likelihood of high life satisfaction than for similarly educated peers in well-matched jobs.

Previous9 minMath–Verbal Divide: Mathematics and Earnings

Next12 minMath–Verbal Divide: From Skill to Outcome

Math–Verbal Divide: Gender, Skills, and Jobs

When Skills Meet Gender, Adult Life, and Jobs

Gender gaps are different at different stages

FAQ