An Occupation With No Exit Interview

It is written on forms. It has a pension category built specifically for it, a tax provision named after it, and two income thresholds that shape how many hours a person in it is willing to work in December. By any reasonable definition it is an occupation.

And it has no salary, no evaluation, no promotion, no colleagues, no handover, no retirement date and no exit interview. Nobody has ever sat down with a 主婦 at the end of twenty-two years and asked her what the role was like, what she would change about it, or what she intends to do next — which is a conversation every other occupation on earth provides, however badly.

This page takes it seriously as an occupation and then asks what an occupation with those particular properties does to the person holding it. It does not argue that she should have had a job instead. That is somebody else’s politics, and it would be the same instruction the culture already issues from the other direction. The position here is narrower and, I think, harder to disagree with: the choice should be made with the architecture visible, and a great many women made it with the architecture entirely invisible.

When people say housework is invisible they usually mean the physical labour, and the physical labour is actually the visible part: somebody can see a clean kitchen. The genuinely invisible portion is executive.

Remembering that the school needs the form by Thursday. Noticing that the youngest has outgrown his shoes before it becomes a problem on a Monday morning. Knowing which grandparent must be telephoned this month and which one telephoned last. Holding the running inventory of what is in the fridge against what is planned for four days. Tracking whose teeth are due, whose jacket is thin, which bill arrives this week, and what the husband is likely to be like this evening based on how he was on the phone at four.

None of that is visible when it is done and all of it is visible the moment it is not. Which produces the defining property of the role: it succeeds by leaving no trace. A week in which everything was anticipated looks exactly like a week in which nothing happened, and after some years of that a person can no longer tell the difference either — between having done an enormous amount and having done nothing at all.

There is a word for the load itself, borrowed into Japanese and increasingly used: 名もなき家事 — the household tasks with no name. Nobody lists them because they are not discrete enough to list: topping up the hand soap, noticing the towel needs changing, deciding what the family will eat on a day nine days from now, replacing the batteries before anybody discovers they are dead. They resist division precisely because naming a task is what makes it shareable, and these have no names — which is why a household that has genuinely split the visible chores can still have one person carrying the entire operating system.

And it does not switch off. Paid work has an end to the shift; this does not, which is why "but you were at home all day" is such a maddening sentence and such a difficult one to answer. The correct answer is that the house was staffed all day, and the staffing is the job, and the reason it does not look like one is that it was done well.

Japan has something close to a natural experiment in how policy shapes a woman’s week, and it happens every December.

A dependent spouse earning below certain thresholds is treated differently by the tax and social insurance systems. Crossing a threshold can mean enrolling in social insurance in her own right, paying premiums, and — because of how the increments work — a household ending up with less money after she earned more. So a very large number of part-time working women deliberately cap their own hours as the year end approaches. This is 就業調整: a nationally recognised, measurable behaviour in which the state’s architecture reaches directly into how many shifts a woman accepts in November.

Underneath sits the structure that matters more and is understood least: 第3号被保険者, the dependent-spouse pension category created in 1985, under which a spouse below the income threshold is enrolled in the national pension without paying premiums herself. It is one of the most consequential structures in Japanese women’s economic lives, and if you asked a hundred women it covers to explain what happens to their pension if they cross the line or if the marriage ends, very few could tell you.

And the thresholds have moved. Reform in 2018 raised and tapered parts of the spousal deduction, and further adjustment followed in the 2025 package. Popular understanding lags the law by years, so a meaningful number of women are currently capping their hours against a number that has already changed — which is a very specific and very fixable way to be poorer than necessary.

None of which is an argument for working more. It is an argument for knowing the actual numbers this year rather than the ones somebody mentioned at a school gate in 2016, because a decision made against a wrong number is not a decision.

Every structure described above is, taken individually, a support. Collectively they make one direction of travel cheap and the other expensive, and the expensive direction is out.

A woman who has been 第3号 for eighteen years has a pension record built on that basis, a career gap this shelf has already priced in another essay, and an earning capacity shaped by nearly two decades of deliberately staying under a line. If the marriage becomes something she wants to leave, those three facts arrive together, and they arrive at an age at which the labour market treats her worst.

So the single most useful thing on this page: 年金分割 exists. Since 2007, the employee pension record accrued during a marriage can be divided on divorce. It is not automatic in every case, there are procedures and time limits, and the pension office will explain it to anybody who asks. An enormous number of women in exactly the position it was designed for have never heard of it, and have concluded that leaving would mean old age with nothing — a conclusion that is frequently wrong and is doing a great deal of quiet work in a great many marriages.

To be clear, because this is a page about a real life and not an argument for a particular outcome: knowing what leaving would actually cost is not a plan to leave. It is the difference between staying and having nowhere to go — and a marriage in which one person knows she could leave and chooses not to is a different marriage from one in which she believes she cannot, even if the two look identical from outside and nobody ever mentions it.

Occupations supply things besides money, and their absence is felt even by people who did not want the money.

They supply a mirror: colleagues who reflect back a version of you that is not the domestic one. They supply evidence: a task completed, a thing shipped, someone saying that was good. They supply a boundary between shift and not-shift. And they supply an identity that survives being introduced at a party, which is a small thing until you have spent nine years watching people’s attention move on immediately after you answer the question.

Without those, two things happen slowly enough to be invisible while they happen. The first is that self-assessment loses its calibration: with no external signal at all, a person’s estimate of her own competence drifts, and it drifts downward, because the role’s only feedback is the absence of complaint. The second is that preference atrophies. Asked what she wants for dinner, she can usually answer for everybody else in the household and then pause, genuinely, at her own turn.

This shelf has an essay arguing that noticing preference is a skill that degrades without use. This is the occupational version of that finding, and it is why the advice at the end of this page is not about fulfilment or purpose, which are large words. It is about restoring a signal.

The economic arrangement produces a relational one, and it does so regardless of how decent either person is.

When one person earns all the money, even the most scrupulous couple ends up with an asymmetry in who explains a purchase, whose time is treated as having an hourly value, and whose tiredness is understood as having been earned. The words are rarely said. The pattern shows up in small things: which of them apologises for the cost of a thing, which of them says "let me check" before agreeing to a weekend, whose plans move when something conflicts.

This shelf has a page arguing that money and intimacy are not opposed and that the objection to paying for care is really an objection about something else. The domestic version of that argument is this: money is already in this marriage, structuring it, every day. The only question is whether the two people can look at it directly. A couple who can say out loud that the household runs on one salary and two jobs is in much better condition than one where that sentence would cause an incident.

And a small structural suggestion that does more than its size: money that arrives in her own account, unaccounted for, in an agreed amount. Not an allowance to be justified — an amount that is hers in the way a salary is somebody’s. A surprising number of long marriages improve measurably on that single change, because it converts a series of small negotiations, each of which cost her something, into one decision made once.

Deliberately, because this page is not making that argument.

One: find out this year’s actual numbers. The thresholds, what crossing one would mean for your household specifically, and — separately — what your pension record currently looks like. The pension office will tell you, free, and a 年金定期便 arrives with your own numbers on it and is thrown away unopened by an extraordinary number of people.

Two: find out what leaving would cost, once, even if you have no intention of going anywhere. 年金分割, what the household finances actually look like, where the documents are. This is not disloyalty. It is the difference between choosing and having no alternative, and it changes how a person stands in her own house.

Three: restore one external signal. Not a career — a context in which somebody who is not related to you responds to something you did. A class, a volunteer role, a regular group, anything with other people and a recurrence. The content genuinely does not matter. What matters is that an assessment of you arrives from outside the household, because it is the absence of any such signal that lets self-assessment drift.

And a note for the reader whose children have just left, because that is the point at which this arrives all at once rather than gradually. The role does not taper; it ends on a Tuesday in March, and the executive load that filled eighteen years of attention stops needing to be carried with no notice period at all. Almost everybody describes the first months as disorientation rather than freedom, and almost everybody assumes that reaction is shameful given that this is what they were working towards. It is not shameful and it is not permanent, but it is far easier to meet if somebody told you in advance that it was coming — which is the only reason this paragraph is here.

Four, and this shelf will keep saying it: one recurring block of time that is yours, in the calendar, defended like a shift. In this role more than any other, because this is the role with no shift boundary at all — the one where nobody else will ever draw the line, since the line has never existed and its absence is invisible to everybody including, after a while, you.

There is an obvious commercial move on this page and it should be named so that not making it is visible. A woman with unstructured time, low external validation and a household income is, on paper, an ideal customer for a service like this one.

Which is exactly why this page spends its length on pension records and income thresholds instead. If the thing missing is an external signal, a purchased hour does not supply it — a paid appointment is not a colleague, does not produce evidence of competence, and would be a poor substitute sold as a good one. The class, the group, the volunteer role are better answers to that specific absence, and they are cheaper, and this house would rather say so.

The narrow honest thing, for anyone who wants it, is different in kind rather than a cheaper version of the same. In a role defined by anticipating what everybody else needs, a couple of hours in which nobody needs anything, nothing is being managed and no one is being monitored is the structural opposite of the week — and for a person who cannot remember the last time she was in a room where she was not staff, that is worth something. It is not a mirror and it is not a colleague, and it should not be sold as either.

Twenty-two years, no handover, no evaluation, no exit interview. A role whose success condition is that nobody notices anything, performed for people who mostly did not notice anything, which means it was performed well.

The architecture was built before you got here and was never explained to you, and the December hours, the pension category and the cost of the exit were all decided by people who were not going to live inside them. Knowing the numbers does not oblige you to do anything with them. But a decision taken against a number you have actually seen is a decision, and everything before that was a default — and there is a considerable difference between a woman who chose this and a woman who was never shown the other column.