Gender and the household economy
How the household went from the unit of production to a black box economics refused to count — and slowly, partly, back again.
The household was the economy
“The industrious revolution was a household-level change in behavior… households increased their supply of labor to the market and their demand for market-supplied goods, well before the Industrial Revolution.”
— Jan de Vries, The Industrious Revolution, 2008
For nearly all of economic history there was no “economy” standing apart from the home. The household was the unit of production, and women were full participants in it. To follow gender and the household across the centuries is to watch this one structure get written out of the discipline and, much later, partly written back in.
Start with the constraint everyone lived under. In what historians call the organic economy, energy came from land, muscle, wind, and water, and the binding limit on output was acreage and human and animal effort. Inside that ceiling, the household was the unit of production, not merely of consumption. It allocated its members’ labor across field, workshop, and hearth as a single decision: who spins, who tends the dairy, who minds the children, who walks the cloth to market. Production and reproduction were not two spheres. They were the same enterprise.
Women’s role in this enterprise was central, not marginal. Brewing, dairying, textile work, market-gardening, and the household’s daily interface with the market were women’s domains as much as men’s. There was no “domestic sphere” sealed off from economic life, because economic life ran through the household. Jan de Vries pushes this further: in the century and a half before steam, households across northwestern Europe chose to work harder and buy more in the market — an industrious revolution that treated the family as an active economic decision-maker, reallocating its own time toward market work long before any factory demanded it.
The apparatus here is not an equation. It is a way of seeing: the household as the optimizing unit, deciding how to deploy a fixed stock of hands and hours against the seasons. Hold onto that picture. The whole thread is the story of how the discipline lost it, and then spent a century rebuilding the tools to recover it.
The family economy on its own terms
Take the family economy at full strength, because the easy mistake is to treat it as a primitive arrangement waiting to be modernized away. It was nothing of the kind. It was a coherent, near-universal economic structure with its own optimizing logic. De Vries’s industrious household was not a passive backdrop to commerce; it was a producer that read prices, intensified its own labor, and bought more in the market when doing so paid. Women were not auxiliaries to this. They ran central lines of production and the household’s market interface. Measured by what it actually did, the pre-industrial household was a working firm.
None of which makes it a lost utopia. The family economy was also patriarchal, legally subordinating, and brutal in its labor demands — women’s central economic role came with little control over property or person. The point is not that it was good. The point is that it was coherent, and that it carried no separation between “work” and “home,” because none existed to draw. The family economy had no word for “the economy” as a thing apart from the household. And that missing concept — the economy as separate from the home — was about to be invented, in a way that would write half the household’s labor out of it.
Where this leaves us
For nearly all of economic history the household was the unit of production, and women’s economic role inside it was central. The family economy was coherent on its own terms — not a backwardness to be corrected, but a structure that worked. What it lacked was any line between work and home, because production and reproduction shared one roof. Industrialization was about to draw that line. And the first decision the new discipline would make, once production moved into the factory, was which side of the line still counted as “the economy.”
The family economy didn’t have a word for “the economy” because the economy was the household. Then the factory took production out of the home — and someone had to decide whether the work left behind still counted. The answer, when it came, was no.
The work that stopped counting
“If a man marries his housekeeper or his cook, the national dividend is diminished.”
— A. C. Pigou, The Economics of Welfare, 1920
Pigou wasn’t joking, and he wasn’t being cruel. He was pointing at a problem he had just helped create. The same work — cooking, cleaning, caring — counts when it is paid and vanishes when it is not. The household, once the unit of production, was about to become a black box the discipline would not open for half a century.
Two things happened together, and they reinforced each other. The first was historical. Industrialization pulled production out of the household and into the factory. The putting-out weaver who set his own hours in his own cottage became the mill hand who arrived at the factory bell. As production left the home, a new norm hardened over the nineteenth and twentieth centuries: the male-breadwinner model, the wage-earning husband paired with the unpaid household-managing wife. Labor split into two gendered categories — “productive” work meaning the (male) wage, “reproductive” work meaning the (female) unpaid household.
The second was conceptual, and it is the one that mattered for the apparatus. When national income was formalized into GDP — Kuznets in the 1930s, the System of National Accounts after the war — someone had to decide where the production boundary ran: which activities count as economic output. The boundary was drawn at the market transaction. National accounts measure what passes through a market because market prices are observable; a paid nanny’s wage is a number, an unpaid parent’s identical labor is not. By construction, that boundary excludes unpaid household and care work.
The SNA production boundary sorts activity into counted and uncounted:
Inside the boundary: market transactions, plus a short list of imputations the accounts make anyway (owner-occupied housing is valued as if rented; subsistence farm output is estimated). Outside the boundary: unpaid household production and care — cooking, cleaning, child-rearing, eldercare done for one’s own household.
Pigou’s paradox is this boundary’s reductio. Let $w$ be the housekeeper’s wage. Paid, the work adds $w$ to measured output. The instant the employer marries the worker and the same hours continue unpaid, measured output falls by $w$ — though not one meal fewer is cooked. The number moved; the economy did not.
GDP counts a transaction, not a task. A paid housekeeper’s wage shows up; the identical work, done for free by a family member, shows up as zero. The line wasn’t drawn to insult anyone — it was drawn where the prices are. But because the reproductive economy ran almost entirely off the books, drawing the line at the market made the larger share of all human labor disappear from the discipline’s headline number.
So the household didn’t just leave the factory floor. It left the apparatus. The unit of production for nearly all of economic history became, in the discipline’s headline measure and in most of its theory, a black box on the supply side of the labor market — a place workers came from, not a place where production happened.
The household-labor exclusion is one of several reasons GDP misleads as a welfare measure. This thread takes it as a single rung in the household’s story; the broader case — what GDP misses and what to do about it — is Is GDP a misleading measure of progress?
A real measurement logic, with a real cost
Give the exclusion its strongest reading. The home/work split and the production boundary were real developments with real logic, not a careless oversight. The male-breadwinner model was the lived nineteenth- and twentieth-century norm, not a strawman to knock down. And the GDP boundary was a consequence of measuring what is observable: market prices exist, the shadow value of an hour spent rocking a child does not. National accountants who drew the line at the market were not declaring household work worthless. They were drawing it where they could put a number on it.
The defensible logic still had an indefensible consequence. An entire economy of unpaid household and care work — historically the larger share of all economic activity — became literally invisible to the discipline’s headline measure and to most of its apparatus. Invisibility is not neutral. What a discipline doesn’t measure, it tends not to theorize, and what it doesn’t theorize, it tends to assume away. The household that the family economy made central became, once production moved out, a box the discipline simply did not open. The question that took half a century to get a serious answer was whether anyone could open it again.
Where this leaves us
The home/work split was real, and the production boundary that excluded household labor had a defensible logic: you can only measure prices you can see. But the consequence was that the reproductive economy disappeared from the discipline’s headline number and from its apparatus. The household became a black box on the supply side of the labor market — counted as a source of workers, never as a site of production. The next move in the thread is the obvious one nobody had made: could you point the standard economic tools at the box itself?
For fifty years after the factory took production out of the home, economics had almost nothing to say about what happened inside the household. Then a Chicago economist decided to open the box — and to treat the family as just another optimizing firm.
The family as a small factory
“The household is truly a ‘small factory’: it combines capital goods, raw materials, and labor to clean, feed, procreate, and otherwise produce useful commodities.”
— Gary Becker, “A Theory of the Allocation of Time,” 1965
Becker’s move was simple and radical. If the household is a producer, point the tools of producer theory at it. The black box opens, and inside it is a firm: choosing inputs, allocating time, responding to prices. For the first time, the discipline had something to say about what happened in the home.
The New Home Economics begins with a single reframe: households don’t consume market goods directly, they use market goods plus their own time to produce the things they actually want — meals, a clean home, healthy and educated children. From there the standard apparatus follows.
In “A Theory of the Allocation of Time” (1965), the household produces commodities $Z_i$ from market goods $x_i$ and time $t_i$:
$$Z_i = f_i(x_i, t_i)$$and maximizes utility over the $Z_i$ subject to a full-income constraint — time is not free, it is priced at the wage $w$ it could have earned. The opportunity cost of an hour spent cooking is the hour’s wage forgone. Specialization then follows from comparative advantage: the partner whose market wage is higher specializes in market work, the other in household production, exactly as two countries specialize in trade.
Treat the home like a workshop. It takes raw materials from the market (groceries, a stove) and labor (time), and turns them into the things the family lives on (dinner, childcare). Time spent on one is time not earning a wage or doing the other, so the household economizes: whoever gives up the most by staying home goes to work, and the household specializes — the same logic that says a country exports what it’s relatively best at.
In A Treatise on the Family (1981), Becker ran the same engine across the whole of family life: marriage as a market with gains from trade between partners, fertility as a quantity-quality choice (as the price of “child quality” rises, families have fewer children and invest more in each), the division of household labor as comparative-advantage specialization. One framework, applied to the most intimate decisions people make — and it generated testable predictions where there had been none.
In the history of economic thought this is a Chicago-school export: the “economic approach” pushed past markets into marriage, fertility, and the home. It won Becker the 1992 Nobel and earned its own name — the New Home Economics — but it has no settled chapter of its own in the standard lineage, which still files Becker mainly under human capital. The thread names it here directly: Becker 1965, 1981.
Becker has a second labor program. His human-capital work — schooling as investment, the Mincer wage equation — is the other half of his labor economics, and it belongs to a separate thread (Labor: Clark through the credibility revolution — forthcoming). This thread traces only his family economics; the two are genuinely distinct programs that happen to share an author.
A genuine achievement, and what it couldn’t see
Becker engaged at full strength, because the standard misreading is to set him up as a patriarchal foil. He was no such thing — he was the economist who brought the household back inside the discipline. Before him, economics had essentially nothing to say about fertility, marriage, divorce, or who did the housework; after him, all of it was analyzable. And the model predicted real patterns: fertility falling as the opportunity cost of women’s time rose with their wages, divorce rising, household specialization shifting with relative pay. This was field-creating work, and it is why the household-production framework is still the workhorse of the field. Whatever comes next builds on Becker; it does not bury him.
But look at what the apparatus made invisible. Becker’s household has a single utility function — the family as one optimizing agent, a unitary mind deciding how to allocate its time. That assumption erases the inside of the box he just opened: intra-household conflict, bargaining, and the unequal division of the household’s gains. Who decides? Who captures the surplus from the household’s “trade”? A single utility function cannot even pose the question. And the comparative-advantage rationalization of the gendered division of labor — she stays home because her market wage is lower — reads, to its critics, as naturalizing an arrangement that was structural and contested, not chosen on a level field. Becker opened the black box. He just assumed away the politics inside it.
Where this leaves us
Becker brought the household back inside the apparatus — a field-creating achievement that made marriage, fertility, and the division of labor analyzable for the first time, and the household-production framework survives as the workhorse it became. But the unitary-household assumption made the household’s internal politics invisible. A single utility function cannot represent two people with conflicting interests negotiating over who does the dishes and who takes the career. To see inside the box Becker opened, the next generation had to admit something he had assumed away: the household contains more than one agent, and half its work had still never been counted.
Becker’s household has one utility function — as if a married couple were a single mind. But anyone who has lived in a household knows it contains at least two people, and they don’t always want the same thing. What happens to the model when you let them disagree — and when you finally start counting the work the accounts left out?
Counting what was uncounted
“If women counted — if their unpaid work were valued, if it appeared in the figures that govern policy — the whole picture of the economy would change.”
— Marilyn Waring, If Women Counted, 1988 (paraphrase of the book’s argument)
Waring went back to the rung Stage 2 left open — the production boundary that wrote women’s work out of the accounts — and refused to accept it. Folbre attacked the rung Stage 3 left open — the unitary household that hid the conflict inside. Together they re-opened both. What follows is the story of how much of that critique the mainstream went on to absorb.
The critique opens the door. Nancy Folbre (Who Pays for the Kids? 1994; The Invisible Heart 2001) argued that the care economy — the raising of children, the tending of the old and sick — is real economic activity that the discipline systematically failed to count, and that Becker’s unitary household was a fiction that hid who actually bore its costs. Marilyn Waring (If Women Counted 1988) took the measurement side: the national accounts she traced literally excluded women’s unpaid work by design. The two together re-opened Stage 2’s GDP exclusion and attacked Stage 3’s single utility function.
The bargaining models walk through it. If the household has two people, model it as two. Manser and Brown (1980) and McElroy and Horney (1981) replaced Becker’s single utility function with a Nash bargain: each partner has their own preferences and a threat point — the welfare they could secure if cooperation broke down (an outside option). Lundberg and Pollak (1993) refined the threat point to non-cooperation within the marriage (the “separate spheres” default), not only divorce. The household’s allocation now depends on who has the better outside option — so a benefit paid to the mother and one paid to the father need not have the same effect, which the unitary model could never explain.
The cooperative household allocation maximizes the Nash product of each partner’s gain over their threat point:
$$\max_{\;} \; (u_A - d_A)\,(u_B - d_B)$$where $u_A, u_B$ are the partners’ utilities from the household allocation and $d_A, d_B$ are their threat points. Raise one partner’s outside option $d$ and the bargained allocation shifts toward them. The household is no longer one mind — it is a negotiation, and the terms of the negotiation are now part of the model.
Two people share a household, and each could walk away to something — a job, a separate life, a worse but survivable arrangement. The better your fallback, the more you can hold out for in the split of chores, money, and careers. Whoever has the weaker outside option ends up with the weaker hand. That is why “who earns the wage” and “who could leave” shape life inside the home — and why a single family utility function was always missing the most important thing about a household: that it contains a negotiation.
Goldin makes the structure visible. Claudia Goldin’s Career and Family (2021), recognized with the 2023 Nobel, showed that the modern gender gap is substantially a temporal and household structure. The penalty falls at parenthood (the “motherhood penalty”), and it is amplified by greedy work — jobs that pay disproportionately for long, inflexible, on-call hours, so the parent who takes the flexible role to manage the household forfeits not a little pay but a lot. Goldin also traced the U-shaped history of women’s participation: high in the agrarian household economy, collapsing with industrialization’s home/work split, climbing back across the twentieth century. The thread closes on its own first stage.
In the lineage these figures sit where the discipline absorbed its outside critics: the care-economy and feminist-economics strands run adjacent to Sen’s capabilities work in development economics and to the post-2008 pluralist turn, but Folbre, Waring, the bargaining theorists, and Goldin have no settled node of their own in the standard map. The thread names them directly: Folbre 1994, 2001; Waring 1988; Manser-Brown and McElroy-Horney 1980–81; Lundberg-Pollak 1993; Goldin 2021.
How much of the critique got absorbed
Take the feminist critique at full strength, because the lazy reading treats it as ideological complaint against rigorous economics. It was not a complaint. Folbre and Waring identified a real measurement-and-theory failure: the care economy is genuine economic activity the discipline didn’t count, and the unitary household was a false simplification that hid who paid for the family’s production. And the kernel got substantially absorbed — which is the whole point of the thread. The bargaining models are now standard in graduate household economics; the unitary household is no longer the research default. Time-use satellite accounts and the UN and OECD time-use programs now measure unpaid work, so the care economy is increasingly counted, even if off the headline. And Goldin’s 2023 Nobel “for having advanced our understanding of women’s labour-market outcomes” is the institutional confirmation that gender and the household are now mainstream labor economics, not a heterodox sidelight. The box that industrialization wrote out, and that Becker re-opened as a unitary black box, is now opened up, counted, and structurally theorized.
How complete is that integration? Here the disagreement is real, and it is a matter of degree rather than frame. The strong-integration reading says the critique has largely been answered: bargaining models, satellite accounts, and Goldin’s Nobel together cover the ground. Feminist economists press a “still incomplete” reading at full strength: care work remains outside headline GDP (satellite accounts are supplementary, not the number that governs policy); the unitary household persists in much applied work despite the bargaining literature; and the bargaining models themselves still assume self-interested agents with outside options, which the care-ethics critique (England, Nelson) argues misdescribes the very relationships care is built on. This is a degree-of-absorption argument inside a shared frame — both sides now agree gender and the household are economic territory, which is exactly what the thread has been tracing. For the wider beyond-GDP case, see Is GDP a misleading measure of progress?; whether feminist economics coheres as a research program alongside the rest of the heterodox left is a separate question taken up in Is there a coherent left-economics? — forthcoming.
The thread verdict
The household that the family economy made central, that industrialization wrote out of the apparatus, and that Becker brought back as a unitary black box, is now — substantially but incompletely — integrated: opened up by the bargaining models, counted by time-use satellite accounts, and structurally theorized by Goldin’s Nobel-recognized work. Economics long treated the household as a black box and made gendered, unpaid, and reproductive labor invisible, and it has partly but really corrected that omission.
Keep the two layers of that verdict apart. The integration narrative — family economy to GDP exclusion to the New Home Economics to bargaining, satellite accounts, and Goldin — is not seriously contested; it is the discipline’s own account of its trajectory, and the 2023 Nobel is its institutional confirmation. The question of how complete the integration is remains a live degree-of-absorption split, and the mainstream’s honest answer is partial-but-real: Becker’s framework was field-creating but incomplete, and the feminist critique’s real kernel — the invisibility of care work, the limits of the unitary household — has been substantially, not totally, absorbed. The integration is real and ongoing, neither finished nor cosmetic. Economics now sees inside the household in a way it spent most of its history refusing to.
Gender inequality is one structure inside the larger question of inequality and what economics can do about it. For the general policy question — and where gender inequality sits within it — see Is inequality a problem economics can solve?
For the gender wage gap as a result of the credibility revolution — the natural-experiment evidence rather than Goldin’s structural account — see Labor: Clark through the credibility revolution — forthcoming.
The household is also a site of relational and care capital, which connects to the trust-as-economic-and-social-capital thread (Is trust a form of capital? — forthcoming).
The thread, end to end
Follow the household across the centuries and you watch one structure move in and out of the discipline’s view, each reframing illuminating what the last one hid:
- The unit of production. For nearly all of economic history the household was the economy — production and reproduction co-located, women central — with no line between work and home.
- The black box. Industrialization split home from work, and the formalization of GDP drew the production boundary at the market — making unpaid household and care labor literally invisible to the discipline’s headline measure.
- The New Home Economics. Becker pointed producer theory at the household and brought it back inside the apparatus — field-creating, but with a unitary utility function that hid the conflict and bargaining inside.
- The correction. Folbre and Waring re-opened the exclusion and the unitary household; the bargaining models replaced one utility function with a negotiation; Goldin made the temporal, household structure of the gender gap visible — and won the 2023 Nobel for it.
What survives: Becker’s household-production apparatus, now the workhorse the bargaining models extend rather than discard. What changed: the household is no longer a black box, and unpaid and care work is increasingly measured (satellite accounts, time-use surveys) and theorized (bargaining models). Each reframing corrected the last — the exclusion buried the household, Becker dug it up, the critique opened it, and the integration counted what was inside. The omission was real, the correction is real, and it is neither finished nor cosmetic.
From here: Is inequality a problem economics can solve? for where gender inequality sits in the general question; Is GDP a misleading measure of progress? for the unpaid-labor critique of the headline number. Forthcoming threads pick up the adjacent arcs — the labor-market apparatus lineage, feminist economics within the heterodox left, the household as relational capital, and the urbanization that accompanied the home/work split. The thread also touches the Distribution and Labor roots of the topic map.