The Journal
Rotating chore schedules, and the jobs that should not rotate
Fixed ownership and rotation fail in opposite directions: one concentrates the knowledge and the resentment in a single person; the other asks for a handover every cycle that no study has priced. The evidence on both is thinner than the template grids suggest, and the decision is better made job by job than household by household.
What fixed ownership buys
The case for leaving each job with whoever does it best has a formal version. The economist Robert Pollak, in a 2002 paper expanded from a talk at a symposium in Gary Becker's honour, sets out the household production model in Becker's own words: households "combine time and market goods to produce more basic commodities that directly enter their utility functions". In the marriage models built on it, "the gains from marriage result solely from specialization". Pollak, who describes himself as one of Becker's most persistent critics, treats that as an auxiliary assumption rather than a finding. The everyday version is simpler: whoever is fastest at a job keeps it.
The speed is real, and it has been measured on office work fine-grained enough to time. Bradley Staats and Francesca Gino analysed two and a half years of transaction data from a Japanese bank's home-loan application line: 598,393 observations. Within a single day, repetition paid: each extra unit of work at the same stage cut completion time by about 0.035 per cent, or 5.7 seconds on a transaction of average length. A unit of work at a different stage helped too, but less — 0.012 per cent, or 2.0 seconds — and the gap was significant. Switching stage was itself associated with higher completion times.
Two limits. Across days the finding reversed: variety improved productivity, and workers with more variety were less likely to leave. And the study measures speed, not mistakes: quality there "is high in this context and shows little variation". Nobody has run the equivalent study on a kitchen.
Three costs that fixed ownership hides
The first is that the knowledge stops being written down. A job held by one person for two years stops being a procedure and becomes a habit: which cycle, which week the collection changes. None of it is secret; it is simply never said aloud.
The clearest measurement of it comes from software, where it has a name. Guilherme Avelino and colleagues, in 2016, estimated the truck factor — "the minimal number of developers that have to be hit by a truck (or quit) before a project is incapacitated" — across 133 popular GitHub projects, and found "the majority of our target systems (65%) have TF <= 2". That is code, not laundry, and the estimate is automated, endorsed by the projects' own developers in about half the valid replies: the shape of a problem, not a number about homes.
The second cost is that resentment attaches to particular jobs rather than to totals. Dan Carlson, writing for the Council on Contemporary Families in 2018 about his study with Amanda Miller and Sharon Sassler, reports that women doing the greater share of dishwashing "reported significantly more relationship discord". His proposed mechanism is a norm: "the more common it is to share a task, the more damaging to relationship quality it is for just one partner" to hold it. Those are American parents on low to moderate incomes, married and cohabiting, surveyed in 1992 and 2006, so the tasks named are dated. What survives is narrower: a division that balances on the total can still be wrong on the items.
The third cost shows up only in the week it matters. If one person owns the food shop, a fortnight of illness is not a gap in the rota but a gap in the only copy.
What rotation costs
Rotation is not the free correction the grids imply. Lisa Mlekus and Günter Maier pooled 56 studies covering 284,086 participants in Frontiers in Psychology in 2021, whose abstract opens by saying that although there are numerous publications on job and task rotation, "there is no consistent evidence of their effectiveness".
Rotation was significantly associated with job satisfaction (r = 0.27), organisational commitment (r = 0.16), career success (r = 0.31), labour flexibility (r = 0.32), general psychological health (r = 0.20), stress and burnout (r = −0.13), individual performance (r = 0.13) and productivity (r = 0.13). It was not significantly associated with work motivation, job involvement, competence development, speed of product development, innovativeness, financial performance or reduced turnover intention. Competence development is what rotation is sold on. The authors write that "many correlations were non-significant and small"; most of the pooled studies were cross-sectional, so none of it establishes cause.
The honest position on the handover is this. Nothing in that research measures what it costs to pass a household job from one person to the next. The nearest evidence is the switching cost within a single working day, on related office tasks, with quality unmeasured. Treat the handover as a risk to manage rather than a cost you can price — and the claim that a rotated standard drifts as what it is, an argument from how habits form rather than a measured effect.
Four questions to ask of each job
The evidence does not settle the household question, so settle it one job at a time.
How much of this job is knowledge rather than effort? Taking the bins out is effort. Knowing that the collection shifts after a bank holiday is knowledge. A job that is mostly effort rotates cleanly; a mostly-knowledge job rotates only once the knowledge is written down.
What does a bad week cost, and can it be undone? A poorly cleaned bathroom is recoverable. A missed renewal deadline is not. Recoverable jobs are where rotation is cheap: the price of a learner is one below-par week.
Does anyone actively mind doing it? Pollak and Michael Wachter object that people have direct preferences about how they spend their time — "if they like cooking and dislike cleaning" — which breaks the tidy model's assumptions. A job one person minds and another does not is the one that pays for itself when it moves.
How often does it come round? Rotation needs repetition. A weekly job gives the next person a turn soon enough to remember the last one. On a job that comes round twice a year, every turn is a cold start: all handover and no run.
The jobs that should not rotate
Four categories follow.
Jobs with a deadline and a penalty: insurance and tenancy renewals, tax filings, the boiler service. The failure is not a lower standard but a missed date, which cannot be made good next week by somebody else. These are the jobs that suit a permanent owner, and handing one over for good asks for more than the chore: "Why do I have to ask my partner to help? Asking is the job", elsewhere on this site, lists the five things that have to move with it.
Jobs where competence is genuinely asymmetric, and the asymmetry is not a habit. If one person is trained or qualified and the other is not, that is a fact about the job. "She is just better at it" usually is not: that is the tacit-knowledge problem wearing an explanation.
Jobs that come round once a year or less, where the handover is the entire job.
Jobs someone actively wants. But read preference off behaviour carefully, because behaviour is partly habit. Eurostat's harmonised European time-use data, collected between 2008 and 2015 across 18 countries including the United Kingdom, reports: "In 14 out of 18 European countries that conducted HETUS 2010, the participation rate of men in ironing is below 2 %", against 8 to 29 per cent for women. Participation rate there means the share of the whole population who spent at least ten minutes on the activity on an average day. Whole tasks sit at close to nobody. The American Time Use Survey's 2025 annual averages, for the civilian population aged 15 and over, show the unevenness running both ways: on an average day 23.8 per cent of men and 47.9 per cent of women did housework, while 11.2 per cent of men and 8.9 per cent of women did lawn and garden care. Neither says who in your household prefers what. Ask them.
Writing the handover down once
If a job is going to rotate, the handover is the part to build, written once rather than performed every cycle. Medicine is where that has been tested.
The evidence review behind the National Institute for Health and Care Excellence's guidance on emergency and acute medical care asked whether structured patient handovers between healthcare professionals improve outcomes. It found six studies — one non-randomised controlled, five before-and-after — with mixed results, every clinical outcome graded very low quality. NICE recommended structured handover anyway, on the committee's judgement rather than on strong evidence, and warned that it helps only if done properly rather than as a form to complete.
At household scale the written handover is five lines per job, kept with the job and not in a head: where the things are, what counts as finished, what usually goes wrong, when it is due, and who to ring if it goes wrong anyway. Written by whoever owns the job now, corrected by whoever takes the next turn.
Most households that run the four questions end up with a mixture, not a model: a short rotating list of frequent, disliked, recoverable jobs, and a fixed list of rare, deadline-bearing, knowledge-heavy ones with a written handover attached. That is a decision, not a failure to commit. The negotiation about who takes which half belongs to the companion piece on splitting chores with a partner.
Sources
- Carries both quotations in section 1's first paragraph and the fourth test's objection in section 4. Section 4 of this NBER working paper states that "The household production model postulates that households "combine time and market goods to produce more basic commodities that directly enter their utility functions" (Becker [1965])" — so the quoted words are Becker's, quoted by Pollak, and the piece now attributes them that way. The cover page states the paper "is an expanded version of a talk I gave at the University of Chicago on April 20, 2001 at the Phoenix Symposium in honor of Gary Becker" and is dated October 2002; it was issued by the NBER as a working paper rather than produced for it, and the piece no longer says otherwise. Section 9 states that "the crucial auxiliary assumption is that the gains from marriage result solely from specialization: marriage permits one spouse to specialize in the market and the other to specialize in home production". Section 4 also carries the Pollak and Wachter joint-production objection used as the third test: their cooking-and-cleaning example argues that "if individuals care whether they spend their time cooking or cleaning (e.g., if they like cooking and dislike cleaning), then the household technology exhibits joint production". The introduction reads "I have been one of Becker's most persistent critics -- indeed, I've spent much of my professional life criticizing Becker", which is why the body — not only this note — now names him as an interested party on first mention. PDF downloaded and the text extracted and read on 20 September 2026. — Robert A. Pollak, "Gary Becker's Contributions to Family and Household Economics", NBER Working Paper 9232, 2002
- Carries every figure and the quotation in section 1's second and third paragraphs. Harvard Business School working paper 11-015, the version of record being Management Science 58(6), 1141-1159 (2012). Abstract: "we use two and a half years of transaction data from a Japanese bank's home loan application processing line"; "over the course of a single day, specialization, as compared to variety, is related to improved worker productivity"; "when we examine workers' experience across days we find that variety, or working on different tasks, helps improve worker productivity"; "workers with higher variety are more likely to stay at the firm". The results section states that each additional unit of same-day stage-specific volume "decreases completion time by approximately 0.035%, or 5.7 seconds for a transaction of average length" while each additional unit of same-day other-stage volume "improves performance by approximately 0.012%, or 2.0 seconds for a transaction of average length", the difference being significant (chi-squared 184.45, p<0.001) — both kinds of same-day work improved speed, which is why the piece now says the different-stage unit helped too, but less. Separately, "task change (i.e., switching stages) is related to higher average completion times". Table 3 gives n = 598,393. The limitations section states that "quality is high in this context and shows little variation" — the basis for the piece saying the study measures speed rather than mistakes. The paper places itself inside an existing literature, citing "Adam Smith's (1776) pin factory and Frederick Taylor's (1911) brickyards to present-day factories in China, call centers in India, and fast food restaurants and banks in the United States", replicating "the findings of KC and Terwiesch (2009) in a non-healthcare setting" and supporting "the laboratory findings of Schultz et al. (2003)"; the piece therefore no longer claims this is the only setting in which the effect has been measured, only that it was measured on office work fine-grained enough to time. WebFetch could not parse the PDF, so I downloaded it and extracted the text with pypdf on 20 September 2026 and read the abstract, introduction, results and limitations. — Bradley R. Staats and Francesca Gino, "Specialization and Variety in Repetitive Tasks: Evidence from a Japanese Bank", Harvard Business School Working Paper 11-015, 2011
- Carries the truck-factor paragraph in section 2. Abstract, read on the arXiv record page on 20 September 2026: the truck factor "states the minimal number of developers that have to be hit by a truck (or quit) before a project is incapacitated"; the authors execute their approach "against a corpus of 133 popular project in GitHub"; "we find that the majority of our target systems (65%) have TF <= 2". The same abstract gives the validation result the piece uses as its caveat: surveying developers from 67 target systems, "in 84% of the valid answers we collect, developers agree or partially agree that the TF's authors are the main authors of their systems; in 53% we receive a positive or partially positive answer regarding our estimated truck factors" — which is why the piece says agreement came in about half the valid replies and labels the finding as the shape of a problem rather than a number about households. Published at the 24th International Conference on Program Comprehension (ICPC), 2016. — Guilherme Avelino, Leonardo Passos, Andre Hora and Marco Tulio Valente, "A Novel Approach for Estimating Truck Factors", 24th International Conference on Program Comprehension, 2016
- Carries the task-specific relationship-quality paragraph in section 2. A research brief by Dan Carlson of the University of Utah for the Council on Contemporary Families, dated 3 April 2018, summarising his Socius paper with Amanda Miller and Sharon Sassler comparing the 1992 and 2006 surveys of low- to moderate-income American parents in married and cohabiting couples — the population the piece now states exactly. Quotations verified on the page on 20 September 2026: women doing the greater share of dishwashing "reported significantly more relationship discord"; men who shared the shopping "not only reported greater sexual and relationship satisfaction"; men doing the majority of laundry reported "lower relationship and sexual satisfaction and more discord"; and "the more common it is to share a task, the more damaging to relationship quality it is for just one partner" to hold it. Cited in this form because the journal version at SAGE refuses automated requests (HTTP 403) and I could not open it; this brief is written by the study's lead author. The brief states no limitations of its own, so the dating and sample caveats in the piece are taken from its own description of the data (1992 and 2006, low- to moderate-income parents) and are stated as such. — Dan Carlson, "Not All Housework is Created Equal: Particular Housework Tasks and Couples' Relationship Quality", Council on Contemporary Families brief, 2018
- Carries the whole of section 3's first two paragraphs. Frontiers in Psychology, volume 12, article 633530, open access, read on 20 September 2026. Abstract, verbatim: "Although there exist numerous publications on job and task rotation from various disciplines, there is no consistent evidence of their effectiveness"; "The three-level meta-analysis on 56 studies (N = 284,086) showed that rotation was significantly associated with job satisfaction (r = 0.27), organizational commitment (r = 0.16), career success (r = 0.31), labor flexibility (r = 0.32), general psychological health (r = 0.20), stress/burnout (r = −0.13), individual performance (r = 0.13), and productivity (r = 0.13)" — the piece now prints that list in full rather than a selection from it, so that the two largest effects, labour flexibility and career success, are in front of the reader. The discussion states: "Although the relationships between rotation and its outcomes were positive on average, many correlations were non-significant and small", and that the authors "could, however, not find significant evidence for positive relationships between rotation and work motivation, job involvement, competence development, speed of product development, innovativeness, financial performance, and reduced turnover (intention)". The construct is named "Turnover (intention)" in Table 1 and hypothesis 5f, so the piece says reduced turnover intention rather than reduced turnover. The stated limitations include that "most included primary studies had a cross-sectional, correlational design", which "does not allow for conclusions to be drawn about causality". Nothing in the paper measures performance at the point of changeover or the cost of a handover, which is why neither the standfirst nor the meta description now states a handover cost as measured. — Lisa Mlekus and Günter W. Maier, "More Hype Than Substance? A Meta-Analysis on Job and Task Rotation", Frontiers in Psychology, 12, 633530, 2021
- Carries the Eurostat figures in section 5's fourth category. Statistics Explained article on the HETUS 2010 wave, read 20 September 2026. The page states that the data "was collected between 2008 and 2015 in 18 European countries, 15 of which are EU Member States and 3 are non-EU countries (Norway, Serbia and Turkey)", and the United Kingdom appears among the countries in its figures. Verbatim: "In 14 out of 18 European countries that conducted HETUS 2010, the participation rate of men in ironing is below 2 %", against a women's range of 8 % to 29 %; "in nine countries the participation rate of men doing laundry is below 3 %", against a women's range of 14 % to close to 40 %. The page defines participation rate as "the proportion of individuals who spent at least 10 minutes a day doing the activity among the whole population" — now given in the body immediately after the quotation, so that the 2 % reads as a share of the population on an average day rather than a share who ever iron; it is a population figure, not a within-household one, which is why the piece says neither dataset tells you who in a given household prefers what. — Eurostat, Statistics Explained, "How do women and men use their time - statistics" (HETUS 2010 wave), 2015
- Carries the American figures in section 5's fourth category. Table 1, "Time spent in primary activities and percent of the civilian population engaging in each activity, averages per day by sex, 2025 annual averages", read 20 September 2026. Percent of the population engaging in the activity per day, men then women: housework 23.8 and 47.9; lawn and garden care 11.2 and 8.9. The table also gives household activities overall at 74.9 and 86.7 and food preparation and cleanup at 57.9 and 72.2, not used in the piece. Population is the civilian population aged 15 and over, which the piece states. The table carries no separate rows for interior cleaning or laundry, so no such figures are claimed. — U.S. Bureau of Labor Statistics, American Time Use Survey, Table 1, 2025 annual averages, 2025
- Carries the medicine paragraph in section 6. NICE evidence review on structured patient handovers, from the guideline on emergency and acute medical care in over 16s, read on the NCBI Bookshelf on 20 September 2026. The review question is "Do structured patient handovers between healthcare professionals improve outcomes?"; six studies were included, one non-randomised controlled study and five before-and-after studies, across intensive care, neurological care, emergency department and internal medicine settings. All clinical outcomes were graded very low quality under GRADE, on grounds of study design, risk of bias and imprecision, and results were mixed: potential benefits in intensive care and internal medicine, no detected effect on readmission or avoidable adverse events in neurological care, and none on staff satisfaction or medication administration errors in the emergency department. The committee recommended structured handover despite the limited evidence and warned that it must be implemented properly rather than becoming a box-ticking exercise — paraphrased in the piece rather than quoted, because I read that point in the committee discussion summary and did not verify its exact wording character by character. — National Institute for Health and Care Excellence, "Structured patient handovers", evidence review for Emergency and acute medical care in over 16s: service delivery and organisation (NICE guideline NG94), 2018
Barriers is a household planner for tasks, money and receipts. A job carries someone's name, and whoever it lands on can accept it or say no with a reason. The task list is free.
Get Barriers