The Second Shift: What Research Says About Housework and Paid Work

In 1989 the sociologist Arlie Hochschild put a number on the housework and childcare that wait at home after a paid workday: employed mothers in the US, she estimated, put in about an extra month of work a year once those home duties were counted. She called that unpaid work the second shift, and made it the title of her book.¹

Nearly four decades of time-use research has shrunk that estimate without making the problem go away. Later US studies found a much smaller gap in total work, but a stubborn gap in who does which kind of work. Worldwide, the gap in unpaid work is still wide: women do about three-quarters of all unpaid care work, the International Labour Organization (ILO) estimated in 2018, and in no country do men and women share it equally.²

For your own home, a useful measure is free time, especially on weekday evenings, and a one-week household log, set out in the closing section, puts that to the test. Housework also feeds into the broader research on work-life balance and how it is defined.

What the second shift means in sociology

In sociology, the second shift means the household labor still to be done at the end of a paid workday: cooking, cleaning, laundry and childcare. A sociology encyclopedia entry by Kristin Marsh credits the term to Arlie Hochschild’s 1989 book and notes that, as she and others established, women carry a disproportionate share of it.³

The phrase is newer than the observation: Wikipedia’s entry links it to the double burden, a term for similar observations in the mid-20th century.⁴ Hochschild’s contribution was to show it inside real homes, couple by couple.

Why does the word “shift” matter? Because it treats home duties as work with hours, not as a private favor. Picture two partners who both leave their jobs at the same time. One sits down; the other starts on dinner, then bath time, then the lunchboxes. Both had a full workday, but only one clocks in a second time. Once home tasks count as work, the question changes from “who helps?” to “who works how many hours?”

What Arlie Hochschild found in The Second Shift

Hochschild, working with Anne Machung, interviewed about 50 working couples in depth during the 1970s and 1980s and observed daily life in a dozen of their homes. They found that most wives still did most of the housework and childcare despite holding paid jobs, and that this second shift strained the couples.⁴

The couples described guilt, marital tension, and too little sleep and sexual interest.⁴ Hochschild also gave a name to the bigger pattern: a “stalled revolution”, in which mothers moved into paid work but jobs and husbands changed far more slowly.⁵

The book’s first case study, as the New York Times review retold it, is a couple she called the Holts. Both said they believed in sharing, but after failed attempts they settled on a deal: she took “upstairs” and he took “downstairs”. Upstairs turned out to mean shopping, cooking, cleaning, bills, laundry and most of the childcare. Downstairs meant the car, the garage and the dog.⁶ The couple considered the deal fair.

The method matters for how far the findings travel. This was qualitative researchqualitative research: Research that collects words, observations or images, through interviews, focus groups or fieldwork, rather than numbers. It is good at showing how and why people experience something, but it is not designed to say how common that experience is.Full entry in the glossary: rich interviews with a small group of families in one region of the US, the San Francisco Bay Area.⁶ Hochschild combined it with national time-diary data to estimate the extra month: about 15 more hours a week for employed mothers than for employed fathers.³ Later researchers described her study as a small sample of unknown generalizability.⁷

Labels that sound even can hide very uneven hours, so the Holts’ deal is best judged by the time each task takes and how often it comes round.

Is the second shift still true today? The extra-month debate

US time-diary research since the 1990s supports half of Hochschild’s picture. Counting paid and unpaid work together, employed mothers’ total workload came out much closer to fathers’ than an extra month a year. But the mix stayed lopsided: women did more of the unpaid work, men more of the paid work, and mothers had less free time.

The study

Moderate evidence

Milkie, Raley and Bianchi (2009), Social Forces

Using two representative samples of US parents with preschool children, the researchers compared mothers employed full time and married to a full-time worker with fathers. The mothers’ total workload of paid and unpaid work was greater than the fathers’ by about a week and a half a year, well short of an extra month. The mothers had less leisure and felt more time pressure, but did not get less sleep or take on more unpleasant types of unpaid work.⁸

The study tested Hochschild’s headline figure on representative US samples and found a gap of about a week and a half a year, well short of a month. Its main caveat is scope: these were American parents of preschoolers, and with the paper itself behind a paywall, nothing here goes past its published abstract.

A 2012 review by the same research group put the wider US finding plainly: the total work hours of men and women were similar, but women did a higher share of theirs as unpaid family care and men did more as paid work.⁷ So the second shift is real as a difference in the kind of work, more than in the total amount.

1. Paid work: in US time-diary studies, men put more of their working hours here
2. Unpaid housework and care: women put more of theirs here, so the mix differs even when the totals are close
3. Free time: where the gap often shows: in one US study, mothers employed full time and married to a full-time worker had less leisure than fathers

Why does the mix matter if the totals match? Paid hours earn income, pensions and promotions; unpaid hours do not. And free time is what people notice. To test it, compare evenings and weekends as well as working hours. If one partner’s free time keeps disappearing into chores, the second shift is alive in that home, whatever the totals say.

Do women still do more housework than men?

Yes, though the gap has narrowed. In US time-diary data analyzed by Suzanne Bianchi and colleagues, women’s weekly housework fell from about 30 hours in 1965 to about 16 hours by 2009-10, while men’s rose from about 5 hours to about 10. Women still did about 1.6 times as much housework as men at the end of that period.⁷

The authors tie the change to women’s rising paid work, which pressed women to shed housework and men, perhaps, to pick up some of the slack. Childcare moved the other way: US married mothers and fathers both spent more time on childcare in 2009-10 than in 1975, so housework shrank while care grew.⁷

The same story holds beyond the US, with a slowdown. Evrim Altintas and Oriel Sullivan of the University of Oxford pooled dozens of time-use surveys from 19 countries covering half a century, from 1961 to 2011. They found a general movement toward equality in core housework, but convergence slowed from the late 1980s in countries that were already more equal, while it continued faster where the split was less equal.⁹ Italian and Spanish women still did far more housework than women elsewhere.⁹

Myth

Now that men do more at home, the second shift has gone.

Fact

Men's housework has grown, but in no country the ILO studied do men and women share unpaid care work equally.

What does a slowing trend mean for a household today? Waiting for social change will not even out your own split. In Hochschild’s study, as the encyclopedia entry sums it up, mothers did the routine, everyday chores such as cleaning and laundry, while fathers took the less frequent jobs.³ Look first at tasks that recur every day, such as cooking and laundry, because that is where hours pile up.

How much unpaid work do women do compared with men worldwide?

Worldwide, women do about 3.2 times as much unpaid care work as men, according to the ILO’s 2018 report on care work, which drew on time-use surveys from 64 countries.² That is most of an afternoon for women against a little over an hour for men.

The ILO also counted paid and unpaid work together, and here the second shift shows on a global scale: women’s total working day was on average longer than men’s, by about three-quarters of an hour.² Change has been slow. In the countries with long-run data, the gender gap in unpaid care shrank by only 7 minutes a day between 1997 and 2012.²

National surveys reported to the UN show how much the gap varies. The table converts each country’s share of a 24-hour day spent on unpaid domestic and care work into hours, for people aged 15 and over.

Surveys differ in method, so read the table as a rough comparison. The OECD’s time-use database, which adds paid work and study to unpaid work, shows the same split as the US research: in countries such as the United States, Japan and Norway, women’s and men’s combined totals were within about 20 minutes a day of each other, while in Italy and India women’s totals were higher by roughly 90 minutes.¹¹ Where totals match, the second shift is about the mix; where they do not, it is about the amount too.

If you live outside these countries, your national statistics office probably runs its own time-use survey: it is the best local benchmark for your household.

Checking the second shift in your own home

The research suggests a household check in three steps: measure both kinds of work, compare free time, and then hand over whole jobs. The ILO’s own policy advice starts the same way at national level, calling on governments to recognize, reduce and redistribute unpaid care work.²

Statistics offices measure time use with diaries in which people record what they did across a whole day.¹¹ A home version works the same way: a recorded week gives both partners the same numbers to discuss, instead of two memories of who did what.

Free time is the fairest single yardstick. Milkie and colleagues found the strain showed up as lost leisure and time pressure rather than lost sleep, so a split that leaves one person with no evenings is unequal even if the hours of work match.⁸

A couple who both work full time might find this in their log. Their week log shows one partner doing the daily cooking and laundry and the other doing the weekend jobs. The totals are close, but the daily jobs eat every evening. Swapping two weekday dinners gives back free time where it is scarcest.

The planning behind the chores is a separate layer: the mental load of running a household rarely shows up in a diary, so a fair log of tasks is a floor, not the whole picture.

A one-week second-shift check

• Each partner logs paid work, unpaid housework and care, and free time for seven days
• Add up each category separately, then the totals
• Compare free time on weekday evenings and at weekends
• Mark which jobs come round every day and who does them
• Move one daily job, from start to finish, to the partner with more free time
• Repeat the log a month later and compare

The check is our suggestion, built from what the studies measured; no study has tried it out on couples. Single parents and people caring for a relative alone face the whole second shift without a partner to share it, and the same log can still show where outside help would save the most time.

The bottom line

Hochschild’s extra month was a 1989 estimate built on a small US study, and later US diary data put the gap in total work at about a week and a half a year for full-time employed mothers of preschoolers. The split itself has not gone away: men still do more paid work, women more unpaid work, and worldwide women’s working day is longer. Measure your own week, both kinds of work and free time, then move a whole daily job to the partner with more free time.