Masculinity Norms, Occupational Choice, and Reallocation: Micro and Macro Evidence

Rachel J. Schuh (Federal Reserve Bank of New York) and Martina Uccioli (Nottingham) — “Masculinity Norms, Occupational Choice, and Reallocation: Micro and Macro Evidence,” presented by Schuh at the NBER Summer Institute, Gender in the Economy, on 28 July 2026 (Day 1, afternoon) as a fifteen-minute egg-timer talk. Questions were held for the evening reception, so there is no Q&A. Source note: written from the recording alone — there is no public draft, and the experimental half is a pilot of roughly 200 people that the speaker repeatedly and explicitly asked the room not to treat as final. Every number below is one she showed or said.

Two American trends, both familiar, rarely put next to each other.

Since the late 1960s, employment growth has been overwhelmingly concentrated in industries that were already female. Split industries into quintiles by their 1967 female share and plot cumulative employment growth: the top quintiles run away, the bottom ones barely move. This is not a historical curiosity — healthcare is still doing it.

Over the same period, male labor force participation fell from over 80 percent to under 70. The same holds for prime-age men and for the male employment-to-population ratio.

Schuh’s question is whether these are the same fact. It is not obvious that they should be. Female sectors might have grown simply because female labor supply grew, which need not touch men at all. And even if demand is genuinely moving, men could follow it.

They are correlated, and the correlation has a gradient

At the commuting-zone decade level, 1980 to 2010, growth of female-dominated sectors against the change in male EPOP gives a slope of −0.4751. Where the female sectors grew more, male employment fell more.

Then split commuting zones by whether their state has more or less progressive gender norms, as measured in the GSS:

Where norms are conservative, the relationship is more than twice as steep. Which is at least consistent with the proposed missing link: men have not moved into growing female jobs because doing so violates something.

What a masculinity norm is, and why the PSID can’t quite measure it

This is the conceptual contribution and it is worth stating carefully. Almost all large-scale gender-norms data measures what is appropriate for women — agreement with statements like “a preschool child is likely to suffer if his or her mother works.” Masculinity norms, following ongoing work Schuh attributed to Dias and coauthors, are the informal rules constraining men’s behaviour: “men should be aggressive and competitive to get ahead,” and — directly relevant here — “there is some work that is men’s and some that is women’s and they should not be doing each other’s.”

The observational evidence uses PSID layoff event studies (Callaway–Sant’Anna, individual, year and age fixed effects, no other controls), estimated separately for gender-traditional and gender-progressive men. And the measurement problem is severe enough that Schuh flagged it twice: the norms questions in the PSID are the women’s-behaviour kind, they sit in the caregiver supplement so they are typically asked of mothers, and the analysis therefore assigns the mother’s norms to the father. That rests on two assumptions — that general gender norms correlate with masculinity norms, and that there is assortative matching on norms — and it is, as she said, a noisy measure of a man’s beliefs.

With that caveat: after a layoff, progressive men return to employment slightly faster, though the estimates are noisy and not significant. The sharper result is about where they go. Progressive men are more likely to be in a female-dominated industry and occupation after a layoff. Traditional men are no more likely than before. So the employment recovery gap, such as it is, appears to run through willingness to take the growing jobs.

The pilot, with all appropriate hedging

The experimental half is meant to fix both problems — measure masculinity norms directly, and get something closer to causal. A survey on Prolific in the US, with a UK version going into the Understanding Society Innovation Panel, which will allow linking respondents to a long-running panel.

Two pieces of pilot evidence, both from about 200 people.

First, pluralistic ignorance. Ask people whether they agree with a masculinity statement, then ask what percentage of people of their gender and education level they think agree. The gap is large — about 30 percent agree that it is important for a man to take risks, but respondents expect over 50 percent to.

Everyone thinks everyone else is more traditional than they are, with wide dispersion in the misperception — which is the standard setup for a debiasing intervention with heterogeneous effects.

Second, willingness to pay. A hypothetical job-choice exercise offering free training and a guaranteed job at a given salary in occupations that differ in perceived gender. Men’s WTP to avoid a feminine occupation (nurse, teacher) relative to a gender-neutral one (loan officer) is about **$7,000** — the female job would have to pay $7,000 more to make them indifferent. More masculine men, by the elicited scale, have higher WTP.

Schuh named the obvious confound herself: respondents already have human capital invested in their current occupations, so some of that $7,000 is distance, not distaste. The repair is a separate module asking respondents to rank recommended occupations for their child, who has not yet made those investments.

And that module produces the finding I would most like to see survive the full sample. Respondents are willing to pay more to steer a son away from a feminine occupation than a daughter. Split by respondent gender, and women are willing to pay more than men to keep their sons out of these jobs.

If that holds up, the norm is not a thing men impose on themselves. It is a thing being enforced on boys, at the margin more vigorously by their mothers, about jobs that are where the employment growth is. The macro correlation at the top of this piece would then be, in part, a story about parenting.

(A talk that opens with fifty years of employment data and closes with two hundred people on Prolific is a strange shape, and Schuh knew it — she asked for feedback on which occupations to use, on the whole design, on everything. What the macro half establishes is that there is something to explain and that it varies with norms. What the pilot suggests is a price for it. Whether $7,000 is the number is a question for the next draft; whether the number is positive is not really in doubt.)