How do you ask your partner to share the mental load?

It is 22:40. You are brushing your teeth and you remember the permission slip for Thursday, the dentist recall card that has been on the fridge since August, and the birthday present for Saturday’s party, which needs wrapping paper, which you are out of. None of it is hard. All of it is yours, because nobody else is keeping track.

So you pick up your phone and type the message a lot of people type: “Can you help more around the house?”

The reply comes back, kind and sincere: “Of course. Just tell me what you need.”

That reply is the whole problem in one line. To tell them what you need, you first have to notice it, work out what to do about it, hand it over, and later check that it happened. You asked for less work and were made the manager.

This piece is about the message that avoids that trap. The research on household labour is easy to find, and most of it points the same way. The hard part is the sentence, because a message about the mental load can easily land as an accusation, a verdict on someone’s character, or a scorecard you will both spend the evening disputing.

Before I go further, I should disclose that I co-founded Subtext, an app built for messages where the wording carries more risk than the content. I have an obvious stake in you agreeing that the sentence is the hard part, so weigh that, and check every study below yourself.

What works, in four points

• Ask for ownership of an outcome, including the noticing and the follow-up, instead of help with tasks.
• Open with one specific example, how it affects you, an acknowledgement of what your partner already carries, and one concrete request.
• Aim for a split you both call fair. In the research, feeling that the split is fair tracks relationship satisfaction more closely than an exact 50/50.
• Don’t expect any script to work every time. And if raising things with your partner isn’t safe, skip the conversation and get support first.

What the mental load actually is

The clearest model comes from the sociologist Allison Daminger. In 2017 she interviewed both partners in 35 Boston-area couples with young children, separately, 70 interviews in all, and asked each person to log every household decision they made over a day¹². From those interviews she split what she calls cognitive household labour into four parts. If you want words for explaining the idea itself before you ask for anything, that is in how to explain the mental load to your partner.

The physical task sits outside all four. Buying the boots is the visible part. Noticing they are too small, comparing three shops, choosing a pair and checking the fit is the part nobody sees.

What Daminger found in those couples was specific. Women did more of the cognitive work overall, and most of the gap sat in anticipating and monitoring. Deciding, the part she describes as most closely tied to power and influence, was shared about equally¹. In 26 of the 32 different-sex couples, the woman did more cognitive labour in total³.

That pattern explains why a household can look fair from the outside. The two of you decide together. One of you did all the thinking that led up to the decision, and will do all the checking afterwards.

Two cautions, because this study gets quoted as if it described everyone. The sample was small, affluent, highly educated, mostly white or Asian, and recruited in one city. Three of the 35 couples were not a cisgender man and woman (two couples of two women, one of a woman and a transgender man), and they were left out of the analysis by gender². Interviews show how the work divides and how couples explain it. They cannot tell you how common the pattern is.

What the newer research adds

Larger surveys have since put numbers on the gap. They mostly agree on the direction and disagree on the size.

• In the US, 322 mothers surveyed about three years after giving birth reported doing about 73% of the cognitive labour and 64% of the physical labour across 30 household tasks. The cognitive share was the more unequal of the two, and a bigger cognitive share went with more stress, burnout and depressive symptoms and worse relationship functioning⁴. The task list was adapted from the Fair Play card deck, and Fair Play’s creator is a co-author.
• In Germany, a probability survey of 2,675 people in couples found invisible mental work at least as unequal as visible housework and childcare. Men mainly handled three of 21 tasks: repairs, tradespeople and bills⁵.
• Among 2,133 US parents, mothers’ cognitive labour was the same whether or not they had a job and whatever they earned. The authors call this “gendered cognitive stickiness”⁶.
• Across ten European countries, in surveys run between 2004 and 2013, carrying more of the cognitive labour went with feeling too tired at work because of home, for women and not for men⁷.

Partners also see the same house differently. In the German survey, about half of men said mental work was shared equally, against about four in ten women⁵. In a 2026 study of 235 couples, bigger gaps between partners’ views of how equally the work was shared went with poorer relationship quality⁸. For the conversation, that means you will probably start from two different pictures, and both of you will think yours is accurate.

All of this is survey data at one point in time. It shows that the load and the stress travel together, not that one causes the other. A systematic review of the field also found studies with small or no gender differences, and noted that nobody has yet validated a way to measure mental labour⁹.

Why doesn’t “can you help more” fix the mental load?

Run “Can you help more?” through Daminger’s four parts and you can see where it breaks.

“Help” assumes there is an owner and a helper, and the helper waits to be told. So the person asking still anticipates the need, works out the options, often decides, hands over a task, and then checks that it happened. The only part that moved is the doing. Even if the helper does a great job, the asker is still the household’s project manager, now with one more person to manage.

A chore list does the same thing more formally. Someone had to notice what belongs on the list and write it, and someone will have to check it. Following up on delegated tasks is cognitive labour in its own right, and repeated reminders are how asking turns into nagging in the other person’s ears. That feeds a pattern couples researchers call demand and withdraw: one person pushes for change, the other pulls back. A meta-analysis of 74 studies found it consistently associated with worse relationship outcomes, whichever partner does the pushing¹⁰.

I could not find a study that tests the phrase “can you help more” itself, so this follows from the model, and no experiment backs it. Lists are not always wrong either. Some couples like an explicit list, and for some people, with ADHD or fatigue for example, loosely defined ownership is harder. A list works when it also says who owns the noticing and the checking.

This is the move from help to ownership in a single message. I wrote both versions and the note between them as an illustration. None of it is Subtext output.

Draft “can you help more around the house? i’m drowning and you never notice anything”

Reads as an accusation · “you never” invites a list of counterexamples · asks for help, so the managing stays with you

Rewritten “I’m stretched thin keeping track of everything at home, and most of it happens in my head where you can’t see it. Could you take over the kids’ appointments completely, from noticing they’re due to booking them? I’d like to stop being the one who reminds.”

Catching the first half of that shift, the words that read as blame, is the job Subtext is built for. It marks the words most likely to land wrong, names the feeling under them, and offers versions that keep your register. Which area to hand over is yours to decide.

What the research supports about fairness

This is where the evidence gets most useful for the conversation, and also weakest.

Feeling the split is fair tracks satisfaction more closely than the split itself. In a 2006 US survey of 487 couples with children, women who did most of the housework were less satisfied than women who shared it. When the researchers added both partners’ sense of fairness to the model, that gap shrank by 71% and was no longer statistically significant¹¹. A German panel that followed people through the birth of a child found that women’s satisfaction moved with changes in how fair the split felt, and not with changes in who actually did what¹². And in a 2026 US study of 3,455 people in same- and different-sex couples, how satisfied people were with the division predicted relationship quality better than time diaries of who did the work¹³.

Appreciation seems to soften unfairness. Across three samples of 2,193 people in the US and Canada during the pandemic, doing more of the housework, or seeing the split as unfair, predicted lower satisfaction, except among people who felt appreciated by their partner¹⁴.

Talking about it goes with better days. A diary study followed 157 Dutch working parents for eight days. On days when they talked more with their partner about household tasks, they reported a more equal split, felt it was fairer, and rated their relationship higher. Most of the effect did not carry into the next day¹⁵. This is the closest the research gets to the conversation itself, and it is still correlational. Good days might produce the talks, rather than the other way round.

Fair does not mean the same thing to everyone. In the German panel, a stable, fair arrangement went with lower satisfaction for men with traditional views¹². In urban China, Japan, South Korea and Taiwan, the link between doing more housework and lower marital satisfaction showed up for Japanese and Korean men and Korean and Taiwanese women, and not for the other groups¹⁶. In a US experiment, people rated a split as more acceptable when the partner doing less earned more or worked longer hours¹⁷.

Put together, nearly all of this is correlational. No randomised trial has changed how real couples divide the mental load and measured what happened to the relationship. The defensible conclusion is narrow. Aim for a split you both call fair and chose openly, whatever the ratio, and say thank you.

Gender matters, and it does not explain everything

In most studies of different-sex couples, women carry more of the noticing. But the default-noticer problem shows up without gender too.

Same-sex couples share housework and childcare more equally on average. A study of 160 US adoptive couples found exactly that, and also found that no group shared perfectly evenly, and that gaps in work hours and income predicted who did more¹⁸. A review of the research notes that same-sex couples still specialise when income, job flexibility or a new child push one partner into the default role¹⁹. In an Australian interview study of same-gender couples not living with children, one participant described a fair split as “not necessarily splitting everything 50/50”²⁰.

Daminger’s later work describes a quieter mechanism. Egalitarian-minded couples explained unequal outcomes with gender-neutral reasons, such as one partner’s personality or the demands of a job²¹. In a study of 44 couples, the belief that they made decisions together helped couples overlook the imbalances that remained²². So “I’m just better at remembering” deserves a second look before it settles anything. Sometimes it is true. Often it is how a habit gets locked in.

For couples where ADHD, autism, chronic illness or disability shapes who can do what, I could not find a single study that measures how the cognitive load is divided. Research on adult ADHD links it to more conflict-prone relationships²³, and an interview study of 13 women with ADHD partners is titled with the feeling many will recognise: “I wish it wasn’t all on me”²⁴. Everything I suggest below for these households, from written discussion to splitting by capacity, is design advice, and none of it has been tested.

The words people argue about

“Mental load” has no single agreed meaning, and the label itself can start a fight.

• Cognitive labour is Daminger’s term for the four-part thinking work.
• Mental load usually means more. One influential paper defines it as cognitive plus emotional labour, and describes it as invisible, boundaryless and never finished²⁵.
• Emotional labour originally meant managing your feelings as part of a paid job. Arlie Hochschild, the sociologist who coined it, objected in a 2018 interview to the term being stretched to cover household to-do lists²⁶.
• Relationship maintenance is the work of keeping the relationship itself going: positivity, openness, reassurance, shared social networks and shared tasks²⁷.

In the conversation, skip the label. “You don’t carry the mental load” invites a debate about what the phrase means. “I noticed the inhaler was running low, found the prescription, booked the refill and picked it up” describes something the other person can see.

What the communication research actually tested

There is a lot of advice on how to open a hard conversation. Less of it has been tested than it sounds.

The strongest evidence is for avoiding a harsh opening. John Gottman’s team filmed 130 newlywed couples arguing and followed them for six years. A harsh opening, and one partner refusing the other’s influence, predicted worse outcomes²⁸. The well-known claim about the first three minutes of an argument comes from a related study of 124 couples, which found that those minutes predicted which couples later divorced²⁹. Two cautions. The headline accuracy figures came from models fitted after the outcomes were known, and when another team tested the key findings in 85 couples from at-risk backgrounds, neither a negative start nor rejecting influence predicted separation³⁰.

Context changes what helps. In two studies of lower-income couples, a wife pushing while her husband withdrew predicted falling satisfaction for wives in the relatively better-off couples, while husband withdrawal looked adaptive in the most disadvantaged ones³¹. Advice that works in one household can backfire in another.

I-statements work better on paper than in real arguments. When 253 university students rated hypothetical conflict openings, first-person statements that also acknowledged the other person’s view came across as least hostile³². But when 20 young couples recorded themselves at home for two weeks, plain “I feel” statements were often deflected. The ones that helped showed awareness of the partner, or of the pattern between them³³. If you want more on this, how to tell someone they hurt you over text goes through the same research in detail.

Communication and satisfaction move together, and the line from technique to happiness is weak. In three long-term studies of 4,086 couples, negative communication rose and fell with satisfaction at the same time, without reliably predicting later satisfaction³⁴.

Nonviolent Communication is popular and thinly tested. An often cited review of its evidence is an unpublished master’s dissertation covering 13 studies, none of them randomised³⁵. The one randomised trial in couples I found is small: 32 distressed couples in India, testing a programme its own developers built on Nonviolent Communication, and I read only its abstract³⁶.

The tested part is small. Open without a sweeping criticism, and show you understand their side. Describing something specific, asking for one concrete thing and pausing when either of you floods are my additions. They are editorial, like every script in this piece.

How do you split the mental load without managing your partner?

Ownership means one person runs an outcome from noticing to done, and the other person stays out of it in between. For each area of the house, agree six things.

1. The outcome. What done well means. “The kids have clothes that fit for the season” is an outcome. “Buy clothes” is a task.
2. Who notices. Who keeps track of when it comes up, and how: a calendar, the school email list, a reminder.
3. Who plans. Who researches the options, and whether they act or come back with choices.
4. What they decide alone. A spending limit, brands, timing. Anything above the line, or anything that changes the other person’s week, gets a quick joint check.
5. When to check back. A short weekly review is a reasonable default. In between, the other person does not check, remind or redo.
6. Standards and backup. Which standards you share, where “different but fine” is allowed, and what happens when the owner is ill, travelling or buried at work.

Number five is the one I would expect to be hardest. If the old owner keeps checking, nothing moved. Expect the new owner to do things differently, and to let a few things slip at first. Raise it at the check-in, and resist taking the task back.

On how to divide the areas, the research gives no support for a universal 50/50. Divide by preference, skill, hours, health or season. The test is that both of you can call the split fair, you chose it openly, and either of you can reopen it.

Eight messages for eight situations

None of these has been tested, and I wrote all eight myself, without Subtext. They follow the better-supported principles above. Change the details, and keep the structure: something specific, what it does to you, an acknowledgement, one clear ask.

1. A cooperative partner

“Can we find twenty minutes this week to look at how we run the house? I think we both do a lot. The part wearing me down is the background work: noticing what’s coming up, working out what to do about it, and checking it happened. I’d like us each to take a few areas completely, from noticing to done, so neither of us has to manage the other.”

2. A partner who doesn’t see it

“Take school as one example. This month I saw the form deadline, found the trip money, reminded you about pickup and checked the form went in. You did the pickup, and that mattered. I was still running the whole thing in my head. Would you take school logistics for next term, including the noticing?”

The concrete list does the persuading here. It shows the noticing, the reminding and the checking without naming them, and it credits the part they did.

3. A partner who avoids hard conversations

“Nothing urgent and nothing you did wrong. I’d like to talk about how we split the planning at home. Twenty minutes, not a big sit-down. Would Sunday morning work, or would it be easier if I wrote it out and you answered when you’ve had time to think?”

Offering a time and a choice of medium lowers the pressure. It is an editorial choice, not a tested fix for avoidance.

4. A partner who gets defensive

“You do a lot here, and the bills and the car are yours. We don’t have to agree on who does more before anything changes. Could one area move over to you completely? I’d like you to take the kids’ medical appointments, from remembering they’re due to booking and following up.”

This separates two arguments people usually run together: who does more, and what changes next. You can make progress on the second without winning the first.

5. Mid-conversation, when it starts to go badly

“I think this has turned into a fight about who works harder, and I don’t want that. Can we stop for tonight and pick it up tomorrow after dinner? I’ll write down the two areas that matter most to me so it’s easier to talk about.”

If it has already turned into a fight, what to text after an argument covers the repair message.

6. A co-parent in another household

“To make the term easier for the kids, I’d like us to agree who owns what, start to finish. I suggest you take sports: registration, kit and deadlines. I’ll take school forms and medical appointments. We each put dates in the shared calendar by Sunday night. Big medical, school or money decisions stay joint. Happy to swap if another split suits you better.”

Businesslike on purpose. It splits routine ownership from the decisions that should stay shared, and it offers a swap so the other parent has something to say yes to.

7. When capacity or energy differs

“I don’t think either of us should be holding all of this in our heads. Could we pick one area and write down what done looks like, who notices when it’s due, what you can decide on your own, and how you’ll tell me early if you need backup? We can split things by what each of us can actually carry, even if the split ends up uneven.”

For households shaped by ADHD, chronic illness or disability. Writing the agreement down takes the memory load off both of you. If wording itself is the hard part, we wrote about tone checking for neurodivergent adults.

8. When it isn’t safe to raise it with them

This one is for a friend, a relative or a domestic abuse adviser, not your partner. Send it only from a phone and an account your partner cannot open. If you are not sure, ring from someone else’s phone or say it in person.

“Can we meet this week, just the two of us? There’s something I’d like to talk through in person.”

It says nothing about home on purpose, because anything written down can be read later. The section on safety below has helplines.

What they might say back

“Just tell me what you need.” Kind, and it hands the job straight back. Try: “That’s the part I’m trying to give away. Could you own the school stuff, including noticing what it needs?”

“I do plenty around here.” Probably true. “You do, and the car and the bills are yours. This is about one area I’m still running in my head. Could that one move to you completely?”

“You’re better at that stuff.” Maybe, and practice is how anyone gets better. “I’ve had more practice. I’d still like you to take it, and I’ll stay out of it while you work out your own way of doing it.”

“Why didn’t you just say something?” “I’m saying it now. What I’d like to change is being the one who has to say something every time.”

“Fine, I’ll do the shopping.” A good start. “Thank you. Does that include noticing what we’re out of and planning the week’s meals, or just the trip? I’m hoping for the whole thing.”

When raising it isn’t safe

Everything above assumes that disagreeing at home is basically safe. If your partner monitors your phone, controls money or where you go, punishes you for disagreeing, threatens you, or you feel afraid of them, a conversation about chores is not the priority, and confronting them is not a fix. Some abusive partners escalate when they feel they are losing control or access, which is why technology safety guides advise planning first and using a device the other person has never had access to³⁷.

Talk to someone trained in this before you say anything at home. If you are in immediate danger, call the emergency number: 911 in the US, 999 in the UK, 110 for the police in Germany, 112 anywhere in the EU.

• US: National Domestic Violence Hotline, 800-799-7233, or text START to 88788³⁸.
• UK: National Domestic Abuse Helpline, 0808 2000 247, free and open 24 hours. Men’s Advice Line, 0808 8010327. Galop, for LGBT+ people, 0800 999 5428.
• Germany: Hilfetelefon Gewalt gegen Frauen, 116 016, free, 24 hours, with interpreters in 18 languages. Hilfetelefon Gewalt an Männern, 0800 1239900.
• Anywhere else: Find A Helpline lists local services.

A privacy note that applies to every writing app, Subtext included. A draft or screenshot you paste in leaves your phone and goes to the providers that process it, and anyone who can open your phone can open the app. In Subtext, conversations delete themselves five days after you last use them, or 90 days if you pin one, and you can delete one immediately³⁹. On a shared or monitored phone, five days is still too long. Don’t draft anything about your safety there. If you want to compare how writing apps handle your data, I read thirteen privacy policies for that.

A conversation checklist

Before

• Is it safe to raise this? If you are afraid of the reaction, talk to an advocate first.
• Pick one or two areas, not the whole house.
• Write down recent, specific examples that cover noticing, researching, deciding and checking, as well as doing.
• Note what your partner already owns, so you can say so.
• Pick a calm moment, not the middle of a crisis or eleven at night.

Opening

• One specific observation, what it does to you, one concrete request.
• Acknowledge their view or what they already carry.
• Leave out “always”, “never”, character judgments and sarcasm.
• Ask for an outcome, including the noticing.

During

• For each area: the outcome, who notices, who plans, what they decide alone, when you check back.
• Allow different but acceptable standards.
• If either of you floods or goes quiet, pause and set a time to come back.
• Ask what they would like to hand to you.

After

• Write the agreement down where you can both see it.
• Review it after two to four weeks.
• When something slips, raise it at the check-in instead of taking the task back.
• Say thank you, more than feels necessary.

The part that is actually hard

Everything above is available to anyone who spends an afternoon reading. The harder part is the first sentence.

It has to say “I’m carrying more than you can see” without saying “you don’t care”, and ask for ownership without sounding like a performance review. It is easy to write a version too soft to change anything, delete it, write one too sharp, delete that too, and then say nothing for another month.

That gap is why we build Subtext. Paste your draft, or a screenshot of the thread it will land in, and it shows you which words could land as harsh, cold or passive-aggressive and the feeling sitting under them, then offers up to three versions that keep your register. When a message matters, it asks what you most want to get across before it writes anything. Then you copy the one you want and send it yourself. Try Subtext in your browserDownload on the App StoreGet it on Google PlayGet the app

I want to be clear about what it can’t do, because the mental load is exactly the kind of problem software likes to promise away. Subtext doesn’t remember the dentist, keep the shopping list, book appointments or decide whose turn it is. It can’t tell you whether your split is fair, and it doesn’t know your partner. No study has tested whether it improves relationships, and our research page says so. A tool that took over all the reminding would mostly move the manager role from one of you to an app that one of you maintains, and I looked at how far software gets in can AI reduce the mental load? The conversation is how ownership gets shared between two people. Subtext’s part is helping you ask for it clearly.

The one change I would try first is to ask for an outcome and put the noticing inside it. “Could you take the kids’ appointments, from noticing they’re due to booking them?” hands over the managing, which “can you help more” leaves with you.

A note on the numbers I left out

Three AI research reports fed into this piece, and several figures in them did not survive a check against the original papers.

“The person who delegates keeps 80% of the mental work.” One report attributed this to Daminger. I could not find it as a finding in her companion essay or her dissertation, and I could not read the full 2019 article. The dissertation does quote one respondent who estimated she does 80% of the cognitive labour. Another possible origin is a garbled version of her finding that women did more of the cognitive work in 26 of 32 couples, which is about 81%. Either way it is a different claim.

“96% of arguments end the way they start.” I could not trace this figure to a paper. The abstract and the university press release of the 1999 study it is usually attached to do not contain it. What the study reports is that the first three minutes of a filmed conversation predicted which newlywed couples divorced.

“Mothers say they carry 71% of the mental load, fathers say 45%, and that adds up to an impossible 116%.” The numbers are real, from a 2023 survey of 3,000 US parents, published in 2025. They are the share of 21 tasks each parent marked as “mostly me”, and the respondents came from different households, so the two figures were never meant to add up⁴⁰. Asked to estimate their own share, mothers said 78% and fathers 61%. These parents were not each other’s partners either, and some were single, so the two answers hint that people overestimate their own share and cannot show who does so more.

“Nonviolent Communication is empirically validated.” See above. The evidence base is small, mostly unpublished and non-randomised.

Gottman’s 83% prediction rate. I kept the direction of the finding and left out the number, because it came from a model fitted after the outcomes were known, and the finding did not hold up when another team tested it in a different sample.

There is one more gap. I could not find any trial that takes real couples, transfers ownership of household areas the way this piece describes, and measures what happens over months. If you try it, you are running the experiment yourself.

---

Think I read the evidence wrong, or that your household proves the opposite? Tell me on LinkedIn.

Samet Durgun is the co-founder of Subtext, an app that catches the emotional tone of your messages and rewrites them in your own voice. He’s based in Berlin.

Sources

Every link goes to the primary source where one exists. Checked 12 and 13 October 2026.

1. The Cognitive Dimension of Household Labor. Allison Daminger. American Sociological Review 84(4): 609-633, 2019. Peer reviewed, interviews. The source of the four-part model: 70 interviews with 35 Boston-area couples with a child under five, 2017. Small and affluent by design, so it shows how the work divides rather than how common the pattern is. I read the abstract only; the detail comes from the dissertation and the author’s own summary, both listed below.
2. Thinking Gender: The Cognitive Dimensions of Household Labor. Allison Daminger. Doctoral dissertation, Harvard University, 2021. Open access. Reuses the 2017 interviews and gives the methods in full: separate interviews, a 24-hour decision log, 32 couples of a cisgender man and woman plus three couples left out of the gender analysis.
3. How Couples Share “Cognitive Labor” and Why It Matters. Allison Daminger. Behavioral Scientist, 19 September 2019. The author’s own plain-language summary. Source of the 26 of 32 couples figure.
4. Cognitive household labor: gender disparities and consequences for maternal mental health and wellbeing. Aviv, Waizman, Kim, Liu, Rodsky and Saxbe. Archives of Women’s Mental Health 28: 5-14, 2025 (online July 2024). Peer reviewed, cross-sectional survey of 322 US mothers in different-sex relationships, fieldwork May to July 2023. Exact figures 72.57% cognitive, 63.64% physical. Funded by NSF and NIH. Tasks adapted from the Fair Play deck; co-author Eve Rodsky created Fair Play, and the conflict of interest statement reads none.
5. Who shoulders the mental work? Evidence on the distribution of mental work in different-sex couples in Germany. Hipp, Leshchenko and Sauermann. Social Problems, advance article, February 2026. Peer reviewed, probability survey of 2,675 coupled individuals in Germany, early 2023. Individuals, not couples.
6. Take a Load Off? Not for Mothers: Gender, Cognitive Labor, and the Limits of Time and Money. Weeks, Kowalewska and Ruppanner. Socius 11, October 2025. Peer reviewed survey of 2,133 partnered US parents, February to March 2023. A quota sample that approximates US parents, not a probability sample. Funded by the Australian Research Council.
7. The gendered division of cognitive household labor, mental load, and family-work conflict in European countries. Haupt and Gelbgiser. European Societies 26(3): 828-854, 2024. Peer reviewed. 19,782 employed people in different-sex couples in ten countries, Generations and Gender Survey fieldwork 2004 to 2013. Cognitive labour is measured with a proxy built from existing survey questions.
8. A dyadic examination of the division of household labor and relationship quality from a tri-load perspective. Coundouris and Henry. Scientific Reports, July 2026. Peer reviewed, early-access version. 235 UK-based heterosexual couples recruited through Prolific; authors at the University of Queensland. Cross-sectional. Single recent study, not yet replicated.
9. Gendered Mental Labor: A Systematic Literature Review on the Cognitive Dimension of Unpaid Work Within the Household and Childcare. Reich-Stiebert, Froehlich and Voltmer. Sex Roles 88: 475-494, 2023. Systematic review of 31 studies, 20 of them from the US. Notes the lack of validated scales and the dominance of white, educated, heterosexual samples. Open access.
10. A Meta-Analytical Review of the Demand/Withdraw Pattern of Interaction and its Associations with Individual, Relational, and Communicative Outcomes. Schrodt, Witt and Shimkowski. Communication Monographs 81(1): 28-58, 2014. Meta-analysis of 74 studies, 14,255 participants. A moderate association (r = .36), about equal in both gender directions. Correlational.
11. Division of Housework, Communication, and Couples’ Relationship Satisfaction. Carlson, Miller and Rudd. Socius 6, 2020. Peer reviewed, cross-sectional. 487 US heterosexual couples with children, 2006 Marital and Relationship Survey, lower-income households oversampled. Adding the man’s communication quality reduced the same gap by the same amount.
12. Changes in Perceived Fairness of Division of Household Labor Across Parenthood Transitions: Whose Relationship Satisfaction Is Impacted?. Hiekel and Ivanova. Journal of Family Issues 44(4): 1046-1073, 2023. Peer reviewed, German Family Panel (pairfam). Longitudinal but still observational.
13. Division of Housework, Sexual Intimacy, and Relationship Quality in Different-Sex and Same-Sex Couples. Smith, Bradbury and Karney. Journal of Marriage and Family, April 2026. Peer reviewed, open access. 3,455 people from the US National Couples’ Health and Time Study, September 2020 to April 2021. The link was stronger in different-sex couples.
14. Feeling Appreciated Buffers Against the Negative Effects of Unequal Division of Household Labor on Relationship Satisfaction. Gordon, Cross, Ascigil, Balzarini, Luerssen and Muise. Psychological Science 33(8): 1313-1327, 2022. Peer reviewed. Three samples, 2,193 people including 476 couples, US and Canada, during COVID-19.
15. Gender (in)equality at the kitchen table: A diary study on how parents’ coordination facilitates an equal task division and relationship quality. Riedijk, Aarntzen, van Veelen and Derks. British Journal of Social Psychology 63: 681-707, 2024. Peer reviewed, open access. Eight-day diary, 157 Dutch working parents in different-sex relationships, 1,235 daily reports. Individuals, not couples. Within-person associations, not causes.
16. Division of Labor, Gender Ideology, and Marital Satisfaction in East Asia. Qian and Sayer. Journal of Marriage and Family 78(2): 383-400, 2016. Peer reviewed. 3,096 married people, East Asian Social Survey 2006. Housework generally, not cognitive labour specifically.
17. Perceived Fairness of Couples’ Division of Housework: Evidence from a Multi-Factorial Experiment in the United States. Urbina and Tisch. Gender & Society 39(4): 590-629, 2025. Peer reviewed experiment. 1,502 US adults rated hypothetical couples. It tests judgments of fairness, not what happens in real households.
18. The Division of Labor in Lesbian, Gay, and Heterosexual New Adoptive Parents. Goldberg, Smith and Perry-Jenkins. Journal of Marriage and Family 74(4): 812-828, 2012. Peer reviewed. 160 US adoptive couples (55 lesbian, 40 gay, 65 heterosexual), interviewed before and after adoption.
19. “Doing” and “Undoing” Gender: The Meaning and Division of Housework in Same-Sex Couples. Goldberg. Journal of Family Theory & Review 5: 85-104, 2013. Analytic review of existing research, not a new study.
20. The management of cognitive labour in same-gender couples. McLean, Musolino, Rose and Ward. PLOS ONE 18(7): e0287585, 2023. Peer reviewed, open access, interviews. 16 participants in Australia without children at home, 2021. No specific funding. Small and qualitative.
21. De-Gendered Processes, Gendered Outcomes: How Egalitarian Couples Make Sense of Non-Egalitarian Household Practices. Daminger. American Sociological Review 85(5): 806-829, 2020. Peer reviewed, interviews with the 32 different-sex couples from the 2017 study.
22. The Myth of Mutuality: Decision-Making, Marital Power, and the Persistence of Gender Inequality. Wong and Daminger. Gender & Society 38(2): 157-186, 2024. Peer reviewed, open access. 112 interviews with 44 different-gender couples from two US studies.
23. Adult ADHD and romantic relationships: What we know and what we can do to help. Wymbs, Canu, Sacchetti and Ranson. Journal of Marital and Family Therapy 47: 664-681, 2021. Review. Does not measure the division of cognitive labour.
24. “I wish it wasn’t all on me”: women’s experiences living with a partner with ADHD. Zeides Taubin and Maeir. Disability and Rehabilitation 46(14): 3017-3025, 2024. Peer reviewed, interviews with 13 women. Small and qualitative.
25. The mental load: building a deeper theoretical understanding of how cognitive and emotional labor overload women and mothers. Dean, Churchill and Ruppanner. Community, Work & Family 25(1): 13-29, 2022. A theory paper, not a new empirical study.
26. The Concept Creep of “Emotional Labor”. Julie Beck. The Atlantic, 26 November 2018. Interview with Arlie Hochschild, who coined emotional labour in The Managed Heart (1983).
27. Maintenance Strategies and Romantic Relationship Type, Gender and Relational Characteristics. Stafford and Canary. Journal of Social and Personal Relationships 8(2): 217-242, 1991. Peer reviewed. The standard five-part model of relationship maintenance.
28. Predicting Marital Happiness and Stability from Newlywed Interactions. Gottman, Coan, Carrère and Swanson. Journal of Marriage and the Family 60(1): 5-22, 1998. Peer reviewed, observational. 130 newlywed couples followed for six years. The reported 83% accuracy came from models fitted after the outcomes were known.
29. Predicting Divorce among Newlyweds from the First Three Minutes of a Marital Conflict Discussion. Carrère and Gottman. Family Process 38(3): 293-301, 1999. Peer reviewed, observational. 124 newlywed couples, 17 of whom divorced. The often quoted 96% figure is not in its abstract or its press release; I did not read the full text.
30. Generalizability of Gottman and Colleagues’ Affective Process Models of Couples’ Relationship Outcomes. Kim, Capaldi and Crosby. Journal of Marriage and Family 69(1): 55-72, 2007. Peer reviewed. 85 married or cohabiting couples from at-risk backgrounds, followed for about two and a half years. The key Gottman predictors did not replicate.
31. Communication that is maladaptive for middle-class couples is adaptive for socioeconomically disadvantaged couples. Ross, Karney, Nguyen and Bradbury. Journal of Personality and Social Psychology 116(4): 582-597, 2019. Peer reviewed, observed behaviour. 515 low-income couples followed for 18 months and 431 newlywed couples in Los Angeles.
32. I understand you feel that way, but I feel this way: the benefits of I-language and communicating perspective during conflict. Rogers, Howieson and Neame. PeerJ 6: e4831, 2018. Peer reviewed experiment. 253 university students, 77% women, rating hypothetical scenarios. Predicted reactions, not real conversations.
33. Failure of I-statements for Mitigating Interpersonal Conflict in Arguments Between Young Adult Couples. Korobov. Studies in Media and Communication 8(2): 49-60, 2020. Discourse analysis of about 140 hours of talk recorded at home by 20 heterosexual couples aged 19 to 26. Small and qualitative.
34. Within-Couple Associations Between Communication and Relationship Satisfaction Over Time. Johnson and colleagues. Personality and Social Psychology Bulletin 48(4): 534-549, 2022. Peer reviewed. Three longitudinal studies in the US, Switzerland and Germany, 4,086 couples.
35. What is the impact of the application of the Nonviolent Communication model on the development of empathy? Overview of research and outcomes. Carme Mampel Juncadella. MSc dissertation, University of Sheffield, 2013. Unpublished, not peer reviewed. 13 studies, none randomised.
36. Efficacy of Emotion-Focused Couples Communication Program for Enhancing Couples’ Communication and Marital Satisfaction Among Distressed Partners. Vazhappilly and Reyes. Journal of Contemporary Psychotherapy 48(2): 79-88, 2018. Peer reviewed. A two-group randomised trial with 32 couples in Maharashtra, India, of a programme built on Nonviolent Communication. Abstract only read; small, and run by the programme’s developers.
37. Technology Safety Plan. National Network to End Domestic Violence, Safety Net Project. Practical guidance for survivors on devices, accounts and monitoring.
38. National Domestic Violence Hotline. The Hotline, US. Phone, text and chat, 24 hours. Its internet safety page covers browsing and device safety.
39. Subtext privacy policy. Subtext, dated 10 October 2026. Company source. Names every provider that processes messages, including Anthropic for text and Deepgram for voice, and states that submissions are not used to train AI models.
40. A typology of US parents’ mental loads: Core and episodic cognitive labor. Weeks and Ruppanner. Journal of Marriage and Family 87(3): 966-989, 2025. Peer reviewed. 3,000 US parents, quota sample, 2023, including single parents. The true source of the 71% and 45% figures.