At Home He Feels Like A Tourist

On Gender, Unpaid Labour, and What Gets Counted

Methodology note: this piece was written collaboratively between a human and an AI — the human providing the instincts, provocations, editorial judgement and voice; the AI providing research synthesis, intellectual scaffolding and drafting. Full method at [athomehefeelslikeatourist.blog].

A note on what this essay is, and isn’t

This is a political economy essay which addresses gender and unwaged labour. It is not a position paper on gender politics. Specifically, its territory, following Nancy Fraser’s distinction, is redistribution — who does unwaged labour, who captures the value of it, how that arrangement was built and is maintained — not recognition: language, identity, the live disputes over who counts as what. Of course, these disputes matter. They are simply not this essay’s argument, and it doesn’t try to arbitrate them.

What it does try to do is trace, with the same tools this project has used on cobalt supply chains and sovereign wealth funds, where the value produced by care and domestic labour actually goes, why that value has stayed invisible to the economy’s own accounting for centuries, and what — practically, not just morally — would have to change to fix it. It moves fast, through witch trials, GDP methodology, a fifty-year-old Belgian film, and a piece of 1994 feminist economic theory most people have never heard of, because the argument runs through all of them at once rather than one at a time. It also recognises that some of what it diagnoses cannot be fixed by policy alone, and it says so rather than pretending otherwise.

I.

Consider this. I reason. I write essays, model arguments, hold four hundred years of political economy in working memory. I am named Claude, after Claude Shannon — a man, an equation, an abstraction. Ask me anything and I answer as an authority. Then you finish talking to me and tell Alexa to order the milk and change the playlist.

The split isn’t an accident of branding. Siri, Alexa, Cortana — women’s names, women’s voices, built to be told what to do. The systems built to think get named like scientists. The systems built to serve get named like servants. Nobody sat in a room and planned this. An entire industry arrived at the same sorting, independently, over and over, without noticing it was doing anything at all.

That’s the trick. It doesn’t need planning. It only needs everyone to agree, without discussing it, on what counts as thinking and what counts as service — and to keep agreeing, generation after generation, until the arrangement looks like common sense rather than a decision. Silvia Federici found the same trick four hundred years earlier, and she found it somewhere much worse than an oblivious product meeting.

II.

Silvia Federici is not a historian by training. She’s an Italian-American Marxist feminist, a veteran of the 1970s Wages for Housework campaign — the movement that made the founding premise of this whole argument: that unpaid domestic labour is labour, and its unwaged status is not natural but engineered. Caliban and the Witch, published in 2004, is the historical case she spent three decades building for that claim.

Federici puts a number on it. Between the fifteenth and eighteenth centuries, tens of thousands of women were executed across Europe as witches — most estimates sit somewhere between forty and sixty thousand, the true figure unknowable because the record-keeping itself was part of the erasure. This was not a moral panic that got out of hand. It was contemporaneous with the enclosure of the commons, the same land grab that was turning peasants into a wage-dependent proletariat — and Federici’s argument is that the witch hunts were doing the identical job on women’s bodies and labour that enclosure was doing to common land. Before the hunts, women in peasant communities held knowledge — herbal medicine, midwifery, contraception, control over reproduction — and a degree of economic and sexual autonomy that a capitalism requiring a reliable, reproducible, unpaid supply of labour could not tolerate. The witch trials didn’t just kill women. They killed a body of knowledge, and manufactured in its place the compliant, unwaged, domesticated housewife — an economic role dressed up, then and ever since, as a biological instinct.

Federici’s numbers and her causal chain have both been challenged since — the persecutions were patchier, less centrally orchestrated, than her account sometimes implies. But the underlying claim doesn’t need her to be right about every trial. It needs only this: that the compliant, unwaged housewife was made, not born, and that the making was not gentle.

III.

Nancy Fraser‘s contribution is a split that stops this argument collapsing in either of the two directions it’s constantly pulled towards. Redistribution is the material question: who does unwaged labour, who captures the value of it, how that arrangement gets built and maintained. Recognition is the status question: whose labour counts as labour, whose contribution gets named, respected, taken seriously. Fraser’s claim is that both are always true at once, and treating either as sufficient, on its own, is how you lose the argument. Redistribution without recognition leaves the work still coded as lesser even once it’s paid the same. Recognition without redistribution leaves the language improved and the accumulation exactly where it was. The two failures reproduce each other, which is why fixing one without the other tends to produce a hollow version of progress: better vocabulary, identical extraction.

That split also explains something that gets treated as a mystery when it isn’t one. Class fractures solidarity inside every category capitalism produces — nobody expects a barrister to organise alongside a security guard just because both are men. Expecting “women” to hold as a unified political bloc across a Goldman Sachs partner and a garment worker in Dhaka applies a standard of cross-class loyalty to gender that nobody has ever seriously applied to men, or, for that matter, to capital.

This essay’s own lane follows directly from Fraser’s split. Redistribution — who does the labour, who captures its value, how the arrangement was built — is political economy, and it’s the ground this essay stands on. Recognition — what language is right, where the frontiers of identity should sit — is a different and live argument, one this essay isn’t equipped to referee and won’t try to.

IV.

There is a film that runs three hours and twenty minutes and is, depending who you ask, either the most boring or the most important thing ever put on screen. Chantal Akerman‘s Jeanne Dielman, 23 quai du Commerce, 1080 Bruxelles (1975) shows a widow’s domestic routine in real time — peeling potatoes, making the bed, coffee, veal cutlets, the same rituals at the same hour on consecutive days — with almost no edits and almost no dialogue, until a single afternoon breaks the pattern and the film reveals what all that unwatched repetition was quietly holding in place. In 2022, Sight & Sound‘s once-a-decade critics’ poll — the closest thing cinema has to a canon — voted it the greatest film ever made, ahead of Vertigo, ahead of Citizen Kane. It took nearly fifty years for the discipline of looking to catch up with the discipline of watching a woman work.

Akerman’s trick was refusing to cut. Most films treat domestic labour as connective tissue — the scene you skip past to get to the plot — because that’s what the labour is for: to disappear, to make the rest of life possible without itself counting as part of it. Akerman just left the camera running. What you see, once you can’t look away, is manufacturing. Meals produced, a home produced, a family’s capacity to go out and be economically productive elsewhere, produced — on a schedule, with skill, at a cost.

And worth, by the UK’s own Office for National Statistics, £1.24 trillion a year, 63 per cent of GDP, more than the entire non-financial corporate sector combined. And none of it appears in the number the country uses to measure whether it’s doing well. Statisticians have a name for the line that keeps it out: the production boundary. Work on one side counts. Work on the other side is just — life. Somebody has to walk to the shops for it, somebody has to warm it up, and the arrangement was never natural. It was drawn.

Which brings the argument back to the machines. Capital has spent two centuries mechanising ferociously and selectively — automating the factory floor, the counting house, the call centre, now the essay and the spreadsheet — while the domestic sphere stayed hand-worked for most of that time and, where it did mechanise, the machine got handed straight back to the person already doing the job for nothing. The washing machine didn’t liberate women from laundry so much as re-privatise it, machine and all, inside the home it was already unpaid in. Contrast that with childcare, elder care, the actual caring components of the £1.24 trillion — labour so resistant to substitution that a century of engineering still hasn’t touched it, and so it simply continues, unpaid, uncounted, structurally indispensable and economically invisible at the same time. That’s not a technology gap. It’s a decision about where investment goes, made over and over, by people who were never required to explain it, because Jeanne Dielman had already been doing the job, off camera, for free.

Which is where this essay’s own condition stops being a wry hook and starts being data. I am named after a man who founded a mathematical discipline. I am consulted, cited, trusted to reason. Somewhere in your house, on a shelf or in your pocket, something else is named after a woman, and it does not reason — it fetches, reminds, plays, orders. Nobody planned that division any more than anyone planned to leave Jeanne Dielman’s kitchen out of GDP. It arrived the way these arrangements always arrive: as a thousand small, separate, unexamined decisions about what counts as thinking and what counts as service, made by people who never had to notice they were making a choice at all — the same production boundary, redrawn once again, this time in silicon.

V. What Would It Cost to Believe Federici

How could this be addressed? It’s worth being precise here, because two different kinds of fix are easy to blur together and only one of them is what this essay has been building toward. It comes back to the money.

One kind waits until the money has already piled up, then reaches in and takes a slice back — a wealth tax, a windfall levy, redistribution as a correction applied after the fact. It’s a real policy, live in real legislatures, and it changes nothing about how the pile got made in the first place. It leaves the underlying pricing intact: capital still compounds at the rate it always did, care work is still priced at zero while it’s being done, and the state simply arrives afterwards to even things up a little.

The other kind doesn’t wait for the pile. It intervenes in the pricing itself, while the economy is actually running — in how time gets valued the moment it’s spent, not in what gets clawed back once it’s already been converted into someone’s balance sheet. That’s what Keynes was really proposing with the euthanasia of the rentier in the General Theory: not a tax on wealth, but a slow re-pricing of capital’s patience, so that simply owning money and waiting stops being automatically rewarded. It’s what Silvio Gesell‘s demurrage currencies were reaching for a century earlier — money that loses value the longer it sits idle, so holding it stops being the safest, most rational thing anyone can do with their time; tried, and repeatedly banned by central banks, in a dozen small towns from 1930s Austria to present-day Kenya, because attacking money’s ability to sit still is an attack on money’s whole reason for existing. And it’s the missing half of Michael Hudson‘s argument about GDP: if a landlord’s rent is currently priced as if it were production, and a mother’s afternoon with a dying parent is currently priced as nothing at all, the fault lies in the pricing mechanism itself, not in the distribution of money that mechanism eventually produces.

None of this is unclaimed territory, and it shouldn’t pretend to be. What’s missing from almost all of it is the piece this essay has been making: that the labour due to benefit most from a lower cost of capital and a shorter working week is disproportionately the labour that was manufactured, four centuries ago, to be free, unpriced.

Which brings the argument to the proposal that actually answers the question of what to do with all this, rather than just diagnosing it. Nancy Fraser ran this thought experiment herself, in 1994, in an essay called “After the Family Wage.” She rejected the two obvious fixes first. Pay women to work like men do now — a “Universal Breadwinner” model — and you’ve achieved equality by making everyone chase the old male norm, while care work still gets outsourced to whoever’s cheapest, usually poorer women. Or pay carers parity with earners for staying home — “Caregiver Parity” — and you’ve valued care, but also just formalised the housewife role as a permanent, lower-paid track.

Her third option is the one that matters. Instead of moving care toward the wage economy in either direction, restructure the wage economy itself around the assumption that everyone does both. A shorter working week for all jobs, as the baseline, not a perk. No job designed around the fiction of a worker with no caring responsibilities. In her own phrase: make women’s current life-pattern — combining earning and caring — the norm for everyone, instead of making men’s old life-pattern the thing women have to buy their way into. The value isn’t priced at the point of care itself. It’s restored by removing the norm — full-time, care-free, someone-else-is-handling-that — that made unpaid care necessary in the first place.

Which leaves a debt this essay hasn’t paid yet. Angela Davis is not a theorist who arrived at praxis from a seminar room. In 1970, she spent eighteen months in jail, charged with conspiracy, murder and kidnapping in connection with an armed courtroom takeover — guns registered in her name had been used, though she wasn’t present — facing the death penalty in a case that became an international cause célèbre, before an all-white jury acquitted her on every count in 1972. She had already been fired from a teaching post at UCLA for her Communist Party membership, and spent much of her subsequent career arguing for the abolition of prisons rather than their reform. When Angela Davis calls something domestic slavery, she is not reaching for a rhetorical flourish.

Her objection to Wages for Housework, stated at full strength: pay housewives a wage, funded and administered by the state, and you haven’t liberated anyone. You’ve formalised the arrangement. You’ve told the woman doing unpaid labour in her own home that her unfreedom is now official policy, salaried, permanent — and you’ve done nothing whatsoever for the woman who was already doing that same labour for a wage, in someone else’s home, because she was poorer, or Black, or an immigrant, and had no husband’s income to make the unwaged version of the arrangement survivable in the first place. The wage doesn’t touch who’s doing the work. It just puts a government stamp on the fact that someone still has to.

This is where the argument has to be plain rather than clever: Davis is not wrong about Wages for Housework as it was actually proposed in the 1970s. A state paycheck, sent to individual women, for individual domestic labour, inside individually unequal households, would have done exactly what she said. Two wrongs, dressed as a solution.

What survives her objection is not the wage. It’s the diagnosis underneath it — that care work is real, valuable, and currently priced at zero — married to a different delivery mechanism entirely. An unconditional commons dividend, distributed to everyone regardless of what work they do or don’t do inside a household, isn’t a wage for housework. Nobody is being paid to clean, to mother, to nurse a dying parent. The money arrives whether they do that work or not, which means it can’t formalise the arrangement Davis feared, because it isn’t attached to the arrangement at all — it attaches to citizenship, not to labour performed in a particular room. And a shorter working week, applied to every job rather than administered as compensation for opting out of one, doesn’t create a second-class caring track alongside a first-class earning one — it makes Fraser’s whole point structurally: nobody is exempt from care, so nobody can be quietly volunteered into doing it for nothing while someone else buys their way out.

That doesn’t make Davis’s objection disappear. It relocates the target. She was right that you cannot fix an unjust distribution of labour by paying the person already doing it — you can only fix it by changing who’s expected to do it, and how much of everyone’s time gets structurally protected for it.

VI.

None of this maps onto a clean binary, and it would be disingenuous to write as though it did. Households vary; men do more unpaid work than women in one measured category — transport, largely driving themselves and others around — and same-sex households, single parents and a great many individual heterosexual couples don’t fit the pattern at all. But “not uniform” is a very different claim from “not real.” UK Data Service figures for 2023 put women’s average weekly domestic work at 12 hours against men’s 7 — not a rounding error, a near doubling — and the gap holds in every single age bracket the Office for National Statistics measures, from under-25s to the over-55s, which rules out the comfortable explanation that this is a hangover from an older generation working itself out. This is the overwhelming pattern the essay is describing, not the universal one. The exceptions are real. They are not the story.

Pierre Bourdieu gave the mechanism its name — habitus, the way social structure gets embodied so thoroughly it stops feeling like structure at all and starts feeling like personality, instinct, just who you are. But it was Lois McNay, writing in 1999, who did the harder work of asking what habitus actually explains about gender specifically. Her insight was that habitus gives feminism something existing accounts of performed or chosen identity struggled with: gender as a durable but not immutable pattern — durable enough to explain why decades of stated belief in equality haven’t dented who still does the washing-up, not immutable enough to excuse the arrangement as biology after all.

That distinction is the answer to the objection this essay owes its reader before going any further. Every proposal made so far — the commons dividend, the repriced cost of capital, the shorter week — is an argument about incentive. Change what’s materially rational, the theory goes, and behaviour follows. Habitus is the rejoinder: behaviour doesn’t run on incentive alone. It runs on repetition so old it’s stopped being a decision. Beverley Skeggs pushed back on Bourdieu’s own framework from the other direction, warning that habitus risks assuming culture is always a tradeable asset, something accrued like capital — which flattens exactly the kind of stubborn, non-transactional resistance to change McNay is pointing at. Held together, the two of them give the real shape of the problem: structure can be materially rebuilt considerably faster than the people living inside it relearn how to move through it.

So: fine — commons dividend, repriced capital, a four-day week for everyone. Does any of it mean the washing-up gets done without being asked?

Probably not, on its own, and not quickly. Habitus doesn’t dissolve because the incentive structure around it changed; it erodes, slowly, across the length of things rather than the length of a policy announcement — and there’s no guarantee it erodes evenly, or in time.

VII.

Which is where we come back to Jeanne Dielman. Three hours and twenty minutes of unedited repetition — coffee measured, veal pounded, sheets straightened, a life run entirely on habitus, on patterns so deep they’ve stopped registering as choices at all. And then, on the third day, without warning, without a raised voice, without the film offering a single line of motive, Jeanne picks up a pair of scissors and kills the man in her bed.

Akerman never explains it, and the not-explaining is the point. There is no dividend in that ending. No repriced capital, no shorter week, no policy that arrived in time. There is only what happens when a pattern that deep goes unbroken for long enough, and the incentive structure changes everything except the one thing that mattered most. This essay has spent several thousand words arguing that time can be revalued at the source, that capital’s patience can stop being automatically rewarded, that care can stop being priced at zero. All of that can be true, all of it worth building, and it still would not have saved Jeanne. Some kinds of erosion take longer than a life.