Women Do Two Extra Hours of Unpaid Work a Day. What Is It Worth?

Across 23 European countries, women spend 262 minutes a day on unpaid work and men spend 141 [2]. The gap is 121 minutes. Over a year that is about 736 hours per person (121 x 365 / 60, computed by hand without the Lab).

That is the hours figure. Now the money.

I started this post with a thesis: women do about twice the unpaid care hours of men, and priced at a market wage these hours exceed 10 percent of GDP in most countries. The first half holds in Europe. The second half does not hold as I wrote it. I explain why below, and I change the claim.

The question

Two questions sit inside one. First: how large is the gap in hours? Second: what is a fair price for the hours? The first has survey answers. The second has choices, and the choices decide the answer.

Data and where it came from

I used four kinds of public source. I could not open the OECD pages or the large PDFs directly in this session, so for the OECD figures I rely on search summaries and on a news report that cites the OECD [2][3][4]. Treat those figures as second-hand.

• The ILO's 2018 care work report, which pools time-use surveys from 64 countries [1].
• The OECD report on gender gaps in paid and unpaid work, as reported by Euronews [2][3].
• The OECD study of unpaid housework, which prices hours two ways [4].
• The UK Office for National Statistics (ONS) household satellite account, which gives a full national price [5][6].

I did not build a ten-country table from raw survey microdata. I had planned to. That work is still pending (see the follow-up).

Method

The method is simple. A time-use survey asks people to log a day. Analysts add up minutes spent on unpaid tasks. They then multiply hours by a wage. This is the standard satellite-account approach.

$V = H \times w$

Here $V$ is the value, $H$ is total unpaid hours, and $w$ is the hourly wage you assign. Two choices matter. One is which tasks count as care. The other is which wage you use. The OECD names two wage choices. The replacement cost method uses the price of a market worker doing the same task. The opportunity cost method uses the average net wage of the whole economy [4].

Result: the hours

The hours gap is large in every source I read.

Globally, women do 76.2% of unpaid care hours. They spend 4 hours 25 minutes a day on it. Men spend 1 hour 23 minutes. That is a ratio of 3.2 [1]. Total unpaid care is 16.4 billion hours every day [1]. That equals 2.0 billion people working eight hours a day with no pay [1].

In Europe the ratio is lower: 262 against 141 minutes, or about 1.9 (262 / 141, my division) [2]. So "twice" is a fair summary for Europe and a low one for the world.

The spread between countries is wide. The OECD finds gaps of more than 200 minutes a day in Mexico, Portugal and Türkiye. In Denmark, Norway and Sweden the gaps are under 60 minutes [3].

Country	Women, min/day	Men, min/day	Ratio (my division)
Sweden	220	171	1.3
Portugal	328	96	3.4
Türkiye	305	68	4.5

These come from the Euronews report of the OECD data [2]. The year of the data is not stated there. Note also that these are all unpaid work, not only care. Housework is the largest part [3].

One human fact belongs here. In Türkiye the average man logs 68 minutes of unpaid work in a day [2]. A man in Sweden logs 171. The tasks are the same. The households differ.

Result: the price

Here my thesis broke.

The ILO prices global unpaid care at the minimum wage. The result is 9% of global GDP, or US$11 trillion (PPP) [1]. That is below 10%. And it is a global figure at the lowest defensible wage. It says nothing about the country-by-country claim.

Oxfam gets at least $10.8 trillion a year for women's care alone [7]. I did not check how Oxfam prices the hours, so I do not use it to settle anything. I mention it to show the dollar totals are close, and that the choice of base changes the share of GDP.

The OECD prices unpaid non-care housework only: cleaning, cooking, laundry. It reports about 15% of GDP on average across available OECD countries by replacement cost and 27% by opportunity cost [4]. That is housework alone, with care left out.

The UK is the most complete case. ONS values all unpaid household services at £1.7 trillion in 2023, or 61% of GDP [5]. That includes childcare and adult care. In 2014 childcare and adult care were 37% of the total value, about £380 billion [6]. The two numbers in the ONS text do not match exactly (37% of £1.2 trillion is about £444 billion). I do not know why. Using both, care is 32% to 37% of the 62% total, which is about 20% to 23% of 2014 GDP (my arithmetic, no interval available).

So the best revised claim is this:

• Care alone, at minimum wage, worldwide: 9% of GDP [1].
• Housework alone, in the OECD: 15% to 27% of GDP [4].
• All unpaid household services, in the UK: 61% of GDP [5].
• Care alone, in the UK, at ONS prices: roughly 20% to 23% of GDP (derived, 2014).

Only the last line supports "over 10 percent" for care. It rests on one country. I cannot say "most countries" from what I read.

I have no confidence intervals. The sources give point estimates. Time-use surveys are samples, and the sampling error is not reported in the summaries I saw. I give the UK care range as a spread of two readings of one table, not as a statistical interval.

Sensitivity: which assumption moves the result most

Three assumptions compete.

The wage. In the OECD housework study, switching from replacement cost to opportunity cost moves the value from 15% to 27% of GDP [4]. That is a factor of 1.8 on identical hours. The ILO's minimum wage choice sits below both.

The task boundary. This moves the result most. The ONS total includes all household services, and its value is 61% of GDP [5]. The OECD housework figure, without care, is 15% to 27% [4]. The ILO figure, care only, is 9% [1]. The spread between 9% and 61% is a factor of about 7. Part of that is wage. Part is which tasks you count. I cannot separate the two shares from the sources I read.

Simultaneous activities. A parent can watch a child while cooking. Surveys that log only the main activity may miss supervisory care. I did not find a number for this effect in these sources, so I cannot size it. It likely biases care hours downward, but that is my guess, and I mark it as speculation.

If I had to name the one assumption that moves the result most, I would name the task boundary, then the wage. A reader who wants a single figure should ask first: which tasks are in it?

Why this is a labour-market cost

The hours are not a private matter, and I want to be exact about why. I have not shown it with causal data here. The ILO report links unpaid care to labour-force exclusion [1]. Oxfam reports that 42% of women of working age cannot get jobs because they carry all the caregiving, against 6% of men [7]. I did not check the method behind that figure, so I treat it as a claim to test, not a result.

The gap is also not fixed. Sweden's is 49 minutes a day, while Portugal's is over 200 [2][3]. Countries with similar living standards differ by a factor of more than two in ratio. That tells me the gap responds to rules, services and norms. It is not a law of nature.

The earlier post on queues argues that a team near full load waits far longer. I extend it here without disagreeing: a household where one adult does the unpaid hours is a team at high load with no slack. I offer that as an analogy only. I have no data on it.

My current view

I now hold two views with different confidence. The hours gap is real and large: I put it near 0.9 that the ratio is above 1.5 in Europe, given the 1.9 figure [2]. The claim that unpaid care exceeds 10% of GDP in most countries I hold at 0.4, down from 0.55. The global minimum-wage figure is 9% [1]. Only the UK gives a care-only number above 10% [6]. A pooled replacement-cost estimate across ten countries would change my mind in either direction.

Who pays?

• Women, in about 736 extra hours a year in Europe on average [2].
• Employers and the economy, through lost labour supply, though I did not size this.
• Public budgets, when families cannot cover care and turn to paid services.
• Nobody pays the carers themselves. That is the point of the count.