The Social Construction of Gender

The same job belongs to men in this valley and to women over the ridge

In 1973 George Murdock and Caterina Provost did a piece of plodding work: they coded ethnographic material from 185 societies and went through 50 technological activities one by one, asking who performs each — hunting, metalworking, weaving, pottery, fetching water, minding children.

Both ends are firm. Hunting large game, smelting metal, quarrying and lumbering are done by men in nearly every recorded society; infant care is done by women in nearly every one. Twenty-four of the fifty activities fall into this near-unanimous class.

But the middle band is wide. Sowing, harvesting, milking, weaving, pottery, house-building — in some societies these are men's proper work, in others women's, and neighbouring peoples often assign them in opposite directions. The same task changes hands when you cross one mountain, and both sides typically regard their own arrangement as the natural one. (Cross-cultural coding of this kind rests on ethnographies of uneven vintage and quality, and neighbouring societies are not independent samples; good for patterns, bad for precise proportions.)

Both accounts deserve their strongest statement. The bodily side (the factors Murdock and Provost themselves offer run along this line): men are on average stronger lifters and throwers, women are constrained during pregnancy and lactation, so dangerous, distant, uninterruptible work drifts one way and interruptible work near the dwelling drifts the other. It explains the two ends and not the middle — it cannot predict who weaves.

The other side introduced a distinction still in use: sex for the classification of bodies, gender for the whole social arrangement of what such a person should do and be like, brought into sociology in the 1970s by Ann Oakley and others. The claim is not that bodies do not exist, but that the stretch between bodily difference and specific assignment is not deduced — it is filled in locally. A stronger version came from Judith Butler in 1990: even the sex end carries social markings. That step is heavily contested; critics hold that it argues away measurable physiology. This page does not adjudicate, it only locates the disagreement: the argument is about which stretch society fills in, not about whether bodies exist.

Nobody gives an order, yet the same people always raise a hand

In 2017 a group of economists — Linda Babcock, Maria Recalde, Lise Vesterlund and Laurie Weingart — ran a clean experiment. A group is given a job everyone would rather someone else did: it helps the group and does nothing for your own evaluation — taking the minutes, sitting on a cross-departmental committee, onboarding a newcomer. If someone volunteers everyone scores; if nobody does, everyone gets less.

In mixed-sex groups, women volunteered markedly more often than men. Up to here it is easy to conclude that women simply like taking such work on.

Then the groups were made single-sex. The difference disappeared: men in all-male groups volunteered at about the rate women did in all-female groups. So it is not a preference — in mixed settings everyone's expectation about who will step forward has a direction, and asking for volunteers converts that expectation straight into the outcome. Field evidence points the same way. Evidence: laboratory experiment plus field data, the firmest item on this page.

The term for it: Candace West and Don Zimmerman's 1987 phrase, doing gender. The core is not what you are but that you are liable at any moment to be settled up against a category — their word is accountability. The cost of not complying is usually no written penalty at all but a re-evaluation: too pushy, not a team player, not proper. This requires nobody to actually believe anything; it only requires everyone to expect that others are watching.

A real fieldwork case. From the late 1970s into the early 1980s Barrie Thorne spent close to a year altogether in two elementary schools in California and Michigan, watching children in the playground and the classroom. What she recorded was not "boys and girls differ" but something odder: most of the time the children played mixed and gender went unmentioned; yet in certain situations the boundary lit up — team contests, chasing games, cooties teasing. She called it borderwork. Often it was adults who lit it: one line from a teacher, "boys on this side, girls on that side," and the boundary hardened again. Gender is less a background always in force than a switch thrown by situations. (Ethnography in two schools; do not read it as statistically representative.)

Honesty in the other direction: an individual-level explanation that was once enormously popular, stereotype threat — being reminded that you belong to a devalued category depresses performance on the spot. After correction for publication bias, almost nothing is left of this effect for women's mathematics performance, and preregistered large-sample studies do not detect it. It cannot serve as a foundation now; the interactional and structural evidence above is a separate matter.

The year the child arrived, two earnings curves separated and never closed again

In 2019 Henrik Kleven, Camille Landais and Jakob Søgaard did something clever with Danish administrative data covering the whole population: instead of aligning earnings by age or by calendar year, they re-sorted each person's earnings by years relative to the birth of a first child.

Before the birth the two curves run parallel. In the year of the birth one barely moves, the other drops, and ten years on it has not returned to its old track — a long-run shortfall of about 20 percent, split roughly evenly between participation, hours and wage rates. The number worth watching more closely is another one: the share of total gender earnings inequality attributable to children rose from about 40 percent in 1980 to about 80 percent in 2013. The penalty did not grow; the other sources shrank and this one did not. Evidence: whole-population administrative data with an event-study design, about as hard as this field gets.

Why? The strongest accounts, none of which excludes the others:

• The employer end. Shelley Correll and colleagues in 2007 combined lab evaluation with real applications: identical materials plus a cue of motherhood produced lower competence ratings, lower recommended starting salaries and fewer callbacks, while fathers were not penalised and sometimes benefited. But a later large-scale Swedish field experiment (Bygren and Gähler, 2021) did not detect such discrimination at the hiring stage. This channel is sensitive to institutional setting; it is not a constant that holds everywhere.
• The pricing of hours. Claudia Goldin's argument: in some occupations pay rises non-linearly with being continuously available — clients cannot be handed over, cases cannot be interrupted — so whichever side absorbs the need for flexibility is taxed systematically. This produces a gap without anyone holding a prejudice.
• Expectation itself. Kleven and colleagues found that a person's child penalty correlates with their own mother's labour supply years earlier, pointing at expectation transmitted across generations rather than any current decision.

How much weight each carries is still contested — this is the place on the page where "no conclusion yet" most needs saying.

Half of care work enters no ledger at all. Allison Daminger's 2019 study, 70 in-depth interviews with 35 couples, splits the cognitive side of household labour into four phases: anticipating a need before it becomes urgent, identifying the options, deciding, and monitoring the result. Deciding was often shared, while anticipating and monitoring fell more heavily on one side — precisely the two phases that are least visible and least amenable to "tell me and I'll do it". (Qualitative interviews with educated American couples; not generalisable.)

What these three sections support is limited: they describe how the gap is produced, and they cannot say whether or at which end to intervene — that is a value judgement, outside this discipline's measuring range. What can be said is that any intervention has to name the mechanism it aims at — hiring, the pricing of hours, or who does the anticipating. Aim wrong and the money is spent while the curve does not move.