What the evidence actually shows
Two quite separate pieces of research sit underneath the familiar claim, and nearly every version in circulation has fused them.
The first is a secondary analysis of a long-running Californian study of families, reported by the University of Minnesota in 2002. It found that taking part in household tasks at three or four years old was the best childhood predictor of a composite measure of doing well in the mid-twenties, and that starting as late as fifteen or sixteen appeared to backfire. The analysis covered 84 young adults; the adult outcomes were collected in a single brief telephone interview; and no peer-reviewed publication of it could be located. The composite included completing education, getting started on a career, relationships with family and friends, not using drugs, and IQ — which a telephone interview is a strange way to measure. The researcher's own family had been among those studied. An age-at-onset finding that fine is considerably more than a sample of that size can carry.
The second is real, and much stronger, and is the source of the word "Harvard" that usually travels with the claim. A 1981 paper in the American Journal of Psychiatry followed 456 men from an inner-city Boston cohort from about fourteen years old to forty-seven, with the childhood ratings made by assessors who did not know how the men had turned out. Childhood capacity to work predicted adult mental health, enjoyment of work and the capacity for close relationships better than social class, IQ or coming from a family with multiple problems did.
Three things about that paper are routinely lost. Its childhood measure was a composite of five indicators, regular household chores being one of them — the others were clubs or sport, school results relative to IQ, taking part in school activities, and general coping. Its outcome was mental health and relationships at forty-seven, not career achievement or earnings. And the men were white, working class, from one American city, and born around 1930; the cohort contained no girls at all.
What has been published since is weaker again. The most cited of the newer studies, from 2022, asked 207 parents in Australia about both their children's chores and their children's memory and self-control, at a single point in time. Self-care and family-care tasks went with higher scores, pet care did not, and the authors stated plainly that the design cannot show chores caused the difference. It was reported in the press as chores improving children's brains.
So the defensible position is this. Household tasks in childhood travel with better adult outcomes in two old observational datasets, one of which is weak. Nothing establishes that they produce them. A household that wants its children to help has better reasons available than this one.
Whether young children want to help
They do, within limits, and the limits are the interesting part.
A well-known series of experiments established that toddlers will help an adult who is visibly struggling — reaching unsuccessfully for a dropped object, unable to manage a cupboard door — without being asked, and that promised rewards did not increase it. Helping of this kind appears from around fourteen months and grows more frequent and more capable over the following year.
What it does not show is that toddlers want to do housework. The tasks are small instrumental favours in a laboratory, with an adult signalling need by look and effortful noise. Nothing in that work involves a floor being swept.
The picture has since been complicated, usefully. A 2022 experiment visited 83 infants aged six to nine months at home, weekly, for up to ten visits. Explicit encouragement and praise increased helping rather than suppressing it, and among the younger infants it increased tidying up specifically. Helping rose both with age and with the number of visits.
Set beside the finding that material rewards suppress helping, that is not a contradiction but a distinction: encouragement appears to build the behaviour, while payment appears to replace the reason for it. It does mean the cheerful version — children are natural helpers, so leaving them to it is enough — overstates what is known.
Rewards, praise and payment
Advice on rewards is confident in both directions and poorly supported in both.
The research usually invoked against them is the overjustification effect. In the original 1973 experiment, nursery-school children who already liked drawing were promised a certificate for doing it; afterwards they chose to draw less than children given nothing, and less than children given the same certificate unexpectedly. A later analysis of 128 experiments found a consistent negative effect of tangible rewards on later free choice, stronger in children than in adults.
Four qualifications turn that from a rule into something much narrower.
- Unexpected rewards showed no effect at all. The damage, where it occurred, came from the reward being promised in advance and made conditional on the task.
- The effect depends on there being motivation to undermine. The same authors noted that where a task holds little interest to begin with there is little to spoil, and the aim becomes getting the child to take on the reason for it rather than avoiding incentives. Most routine household work is exactly that sort of task for an older child.
- Praise is not the safe alternative it is presented as. Verbal praise did improve later free choice overall, but the improvement was clear only in adults; in children it was not statistically significant, and praise delivered in a controlling manner had an effect as negative as any reward.
- None of it was tested on chores. The experiments used drawing, puzzles and, in one case, toddlers handing back dropped objects. No study could be found that tested a reward against ordinary household work.
The one study that measured payment against real housework is observational. Between 2002 and 2004, researchers recorded about 11,000 observations across thirty middle-class dual-earner families in Los Angeles, each with a target child aged eight to ten. Children accounted for about 13 per cent of the household work observed; under 3 per cent of their own recorded activities were household tasks, against more than a quarter of their mothers'. Eighteen of the twenty-eight families for which it was recorded gave pocket money regularly, tied to chores in seven of them — and it bore no measurable relationship to how much the children actually did. The stronger predictor was whether the family employed a cleaner or a nanny, in which case the children did significantly less.
The same study found that chores were negotiated rather than assigned: in most of these households children routinely bargained over them, resisted, or simply refused. Its authors concluded that the earlier literature on chores and responsibility had been measuring what parents believed rather than what children did.
Thirty families in one American city twenty years ago is not a basis for a rule either. It is, though, the only direct evidence available, and it points away from payment doing the work it is assumed to do.
Which tasks, at which age
There is no United Kingdom-wide guidance on this. Nothing on the NHS website, nothing on GOV.UK, nothing from the Royal College of Paediatrics and Child Health and nothing in NICE guidance sets out which household tasks suit which age.
Two official sources come close, and both are chiefly worth knowing for what they reveal.
A children's occupational therapy service within NHS Tayside publishes an age-banded list, running from toys into a box and dirty clothes into the basket at two or three, through dusting and folding towels at six or seven, to cooking a meal and ironing at twelve and over. It carries no date, no review date and no source for any of it, and several of its examples — spraying off the patio, mowing the lawn, washing windows — read like a lightly reworded American list. Its one piece of general advice is sound, and is the only part that needs no citation: work alongside the child, and check they are safe with the task and the equipment before leaving them to it.
Parent Club, run by the Scottish Government, publishes an undated task list with no ages attached, recommends a chore chart with stickers, and states that children who help at home grow up more competent, confident and happy. That is the 2002 finding, passed on without attribution and without any of the qualifications that belong with it.
The practical conclusion is not that nothing is known, but that the ordering is observable rather than published. Which tasks a particular child can manage is a question about that child's reach, grip, attention and judgement, and it is answered by watching rather than by consulting a band. The absence of an authority also means that any list presented as official is somebody's opinion, whatever site it is hosted on.
Safety, and where the guidance runs out
No official British source states when a child can safely use a knife, an oven, a cleaning product, a lawnmower or a power tool. The Royal Society for the Prevention of Accidents, which is the obvious place to look, does not address children doing household tasks at all: its kitchen advice is to keep small children out of the kitchen where possible, and its cleaning-product advice is to keep chemicals in their original containers and out of reach.
What that body does publish is a picture of how children are hurt at home, and those figures deserve their dates. Falls are the commonest cause and account for a little under half of children's home accidents; suffocation, strangulation and choking account for most deaths under five; bath water causes the most severe scalds; and tens of thousands of children a year are treated after kitchen accidents, most of them under five. Most of these numbers come from publications of 2002, drawn from a home-accident surveillance system that has been discontinued, so they are unlikely ever to be refreshed.
Two things follow. The hazards that injure young children at home are overwhelmingly those of being in the room rather than of being given a job, which is an argument for working alongside a small child rather than for keeping them clear of tasks altogether. And for an older child and a sharp, hot or electrical tool there is no published threshold to consult: the judgement is an adult's, unaided, and it is better made deliberately than by default.
Where the law comes in
Helping at home sits outside child employment law, and the mechanism is worth knowing because it is narrow. Under section 30 of the Children and Young Persons Act 1933, unpaid work counts as employment where a person "assists in a trade or occupation carried on for profit", whether or not they are paid for it. Unpaid domestic help in a family home is neither a trade nor an occupation carried on for profit, so the restrictions on ages, hours and local authority permits do not reach it. The GOV.UK guidance on child employment makes no mention of helping at home, which is a silence rather than a confirmation.
Those restrictions are themselves in flux. Section 18 of the 1933 Act was rewritten by the Children's Wellbeing and Schools Act 2026 and is in force only for specified purposes from 29 April 2026, with the regulations it depends on still to come. Anything stated about minimum ages or permits for paid work needs checking against the legislation directly.
Two adjacent lines matter more for most households than the employment point does.
- Where tasks amount to caring for somebody, statutory rights attach. Under section 17ZA of the Children Act 1989, in force since 1 April 2015, a local authority must assess the support needs of a young carer — anyone under eighteen providing or intending to provide care for another person — on request, or where it appears they may have such needs, and must take reasonable steps to identify the young carers in its area. The line that matters in law is not the one between chores and employment but the one between chores and care.
- There is no minimum age for leaving a child alone anywhere in the United Kingdom. What exists instead is the offence of neglect, which bites where leaving a child alone puts them at risk. The NSPCC's position, as relayed by local authority family information services, is that children under three should never be left alone and that leaving any child under twelve alone is not recommended.
Who actually does it in British households
The best British evidence comes from time-use diaries kept by children aged eight to sixteen in 1975, 2000 and 2015 — nearly eight thousand diaries in all.
Adjusting for age, day of the week and household type, girls recorded about 13.6 minutes a day more housework than boys. The gap narrowed significantly after 1975 but did not close, and boys' own figures barely moved across the forty years. Older children did more than younger ones, by around ten minutes a day at eleven to thirteen and around eighteen minutes at fourteen to sixteen, and everyone did more on non-school days.
The larger change is in the total. Children's housework fell by roughly nine minutes a day between 1975 and 2000 and stayed down in 2015, and the fall was concentrated among fourteen to sixteen-year-olds — enough to flatten the difference between older and younger children. Whatever else is true, children in British households do noticeably less of this than their equivalents did in the 1970s, and teenagers account for nearly all of the change.
The adult pattern it sits inside has been measured separately: in 2015, men in the United Kingdom averaged sixteen hours of unpaid work a week and women twenty-six, with women doing more in every age band. That analysis covers nobody under sixteen, so it says nothing directly about children — but it is the arrangement they are watching.
What happens when a task is not done belongs with natural and logical consequences; how demands and warmth are combined in a household belongs with parenting styles.