THRIVING after the FALL
Book Two PAIRING after the FALL

PAIRING · Part II · The Campfire Contract

The numbers and sources

Chapter 7: The Second Shift That Wasn’t

The narrated edition tells this chapter with a handful of figures. Here is everything behind it: each figure with its study, its research document and its row in the log, the case for and the case against, where the sources disagree, and what is still open.

  • 69 open
  • 35 to check first

Read about the chapter · Its 69 rows in the log

Numbers to carry

The figures the chapter is built on, each with its research document and its row in the log.

  1. 450 to 500 minutes of total work a day for both sexes, women about 60 per cent of the unpaid and 40 per cent of the paid: MTUS, 1.2 million diary days, 30 countries (Gershuny and Sullivan, Annual Review of Sociology, 2019, per CO-02). Rows C07-05, C07-06, C07-09.
  2. Gender Care Gap 43.4 per cent, 1 hour 16 minutes a day, down from 52.4 per cent in 2012/13, and mothers of under-sixes 10.5 fewer paid hours a week: Destatis ZVE 2022, 20,000 people (per CO-02). Rows C07-02, C07-11, C07-12, C07-13.
  3. Husband not in full-time work: divorce hazard about 30 to 33 per cent higher, 3.3 against 2.5 per cent a year, post-1975 cohort, PSID 1968 to 2013, more than 6,300 couples (Killewald, American Sociological Review, 2016, per CO-02 and CO-06, which differ). Rows C07-25, C07-26.
  4. Cycle effect on exercise performance: 0.06 of a standard deviation, credible interval minus 0.16 to 0.04, 78 studies (McNulty et al., Sports Medicine, 2020, per CO-05). Row C07-43.
  5. Mean cycle 29.3 days, SD 5.2, 13 per cent exactly 28 days, 24 per cent of ovulations on day 14 or 15: 612,613 cycles (Bull et al., npj Digital Medicine, 2019, per BI-04 and CO-05). Rows C07-36, C07-37.

Every figure

One row for each figure or finding in the data edition's chapter, whether or not the narrated edition kept it. Layer says whether it belongs to physiology, behaviour or culture. Status is how the finding stands in the research: robust, contested, open or fallen. It is not yet the log's verdict.

FigureWhat it measuresStudy (authors, year, venue, N)DocumentRowLayerStatus
Diary in ten-minute slices, with columns for main activity, secondary activity and who was presentDesign of the German time-use diary (ZVE 2022)Destatis ZVE 2022 method. CO-02 says diaries are "generally" in 10-minute intervals and does not say so for the ZVECO-02C07-01behaviouropen
20,000 people in Germany filled in the diary in 2022Sample of the ZVE 2022Destatis ZVE 2022 results page. Persons, households or diary days unclear (Destatis reports about 10,000 households)CO-02C07-02behaviouropen
50 dual-earner couples, interviews and observationHochschild's sample for The Second ShiftHochschild and Machung 1989, Viking. 50-couple figure sourced to a 2015 retrospectiveCO-02C07-03behaviouropen
An extra month of twenty-four-hour days a yearHochschild's arithmetic of the double dayHochschild 1989CO-02C07-04behaviourfallen (not found in the diary archive)
Collecting since 1961, more than 1.2 million diary days, 30 countriesSize of the Multinational Time Use StudyGershuny and Sullivan 2019, Annual Review of Sociology 45:301 to 320CO-02C07-05behaviourrobust
50 years, 14 rich countries, total work 450 to 500 minutes a day for both sexes, barely moved in four decadesTotal of paid and unpaid work by sexGershuny and Sullivan 2019, Annual Review of Sociology. CO-02's citation is a 2018 CTUR working paperCO-02C07-06behaviourrobust
Gershuny and Sullivan "work from that archive"AffiliationCentre for Time Use Research (UCL, formerly Oxford), general knowledge, not in the documentCO-02C07-07noneopen
Leisure gap in total hours is an artefact of early measurement or particular sub-samplesWhy earlier studies found a second shiftGershuny and Sullivan 2019, as glossed by CO-02CO-02C07-08behaviourrobust
Women about 60 per cent of unpaid and 40 per cent of paid work, men the reverseShare of each kind of work by sexGershuny and Sullivan 2019 (CO-02 table says men perform more than 60 per cent of paid work)CO-02C07-09behaviourrobust
Paid work 250 to 300 minutes a day, unpaid 180 to 250 and converging slowly, routine tasks mostly with women in every country, core housework fell sharply for womenConvergence by kind of workGershuny and Sullivan 2019, CO-02 table. One source is a Council on Contemporary Families blogCO-02C07-10behaviourrobust
Gender Care Gap 43.4 per cent, women over 18 do 1 hour 16 minutes more unpaid work a day, nearly 30 hours a week against 21German unpaid-work gap, 2022Destatis ZVE 2022 press release, BMFSFJ noticeCO-02C07-11behaviourrobust
Gender Care Gap 52.4 per cent, 1 hour 27 minutes, in 2012/13German unpaid-work gap, a decade earlierDestatis ZVE 2012/13, via a bpb datasheetCO-02C07-12behaviourrobust
Mothers of children under six do about 10.5 hours a week less paid work than women without children. Fathers keep full-time jobs and trim leisure a littleEffect of young children on paid hours by sexDestatis ZVE 2022. Comparison group to checkCO-02C07-13behaviourrobust
Working mothers in every rich country report more exhaustion, time famine and work-family conflict than working fathersCross-national subjective burdenCO-02 framing, Greenhaus and Beutell 1985 for the construct, no primary for the cross-national claimCO-02C07-14behaviouropen
402 mothers, 291 fathers, dual-earner, random pager (experience sampling). Multitasking 48 against 39 hours a weekDensity of parents' timeOffer and Schneider 2011, American Sociological Review 76:809 to 833, 500 Family Study, N = 693CO-02C07-15behaviourrobust
Mothers' multitasking was childcare on housework and came with distress. Fathers' was leisure or paid work on something and did not. Fragmented time restores lessContent and cost of multitaskingOffer and Schneider 2011. The fragmentation claim is from an Oxford thesis in CO-02CO-02C07-16behaviourrobust
70 interviews, 35 couples. Four stages (anticipate, identify, decide, monitor). Deciding shared almost equally, anticipating and monitoring done by the womanCognitive labour and its allocationDaminger 2019, American Sociological Review 84:609 to 633CO-02C07-17behaviourrobust
322 mothers of young children. Cognitive-labour split more skewed than physical. Tracked depression, burnout, relationship functioning. Cross-sectionalMental-health correlates of cognitive labourAviv et al. 2024, Archives of Women's Mental HealthCO-02C07-18behaviourcontested (one cross-sectional survey)
Men primarily responsible for outdoor tasks and repairs in 80 per cent of couples (laundry 71, cleaning 66, cooking 58 per cent women, in the Preface)Who is responsible for which taskFReDA 2021 to 2023, BZgA Forum 1/2023SY-06C07-19culturerobust
Couples who begin symmetrical almost invariably revert to an unequal split at the first birth, "on both sides of the old border"Reversion at the first birthBiB FReDA policy brief, 9 July 2025. SY-06 does not split the finding by East and WestSY-06C07-20culturerobust (regional split open)
About one woman in five in a "female-overburdened" household is thinking about leavingSeparation thoughts by division of labourBiB FReDA policy brief 2025SY-06C07-21culturerobust
Fixed-effects models: a man's life satisfaction rises with more hours of routine domestic work and fewer of paid workMen's satisfaction and the division of labourFReDA 2021 to 2023, about 42,000 observations, authors not named in the documentSY-06C07-22behaviourrobust
"The German family panel" names both FReDA (2021 to 2023) and pairfam (Johnson 2016, 1,315 couples)NamingSY-06 for FReDA, CO-02 for pairfamSY-06, CO-02C07-23noneopen
Interruptibility, not strength, decided task allocationCampfire Contract cross-referenceBrown 1970, American Anthropologist, chapter 2 row C02-06chapter 2C07-24culturerobust
More than 6,300 couples, PSID 1968 to 2013. Married before 1975: wife's full-time work raised divorce risk. After 1975: no effect of her job or incomeDivision of labour and divorce by cohortKillewald 2016, American Sociological Review 81:696 to 719CO-02, CO-06C07-25behaviourrobust
Husband not in full-time employment: divorce hazard about 30 per cent higher (CO-02) or 3.3 against 2.5 per cent a year, 33 per cent (CO-06)Breadwinner norm in divorce riskKillewald 2016CO-02, CO-06C07-26behaviourcontested (report-to-report)
SOEP 1991 to 2012. West German wives raise housework the year after their income passes the husband's. East German wives do notHousehold response to a wife's income by regionLippmann, Georgieff and Senik 2020, Economic Journal 130:1445 to 1470SY-06C07-28culturerobust
Nearly 45 per cent of employed German women part-time against about 11 per cent of men. Older women draw 53 per cent of older men's pension incomePart-time rates and the pension gapGerman Politics 2021 article and an Ageing and Society paper, years and definitions not givenCO-02C07-29behaviouropen
About 4,500 couples, NSFH 1992 to 1994. Traditional split had more sex. Husbands doing more cooking, cleaning, shopping reported less, "about 5.2 times a month"Housework and sexual frequency, early ninetiesKornrich, Brines and Leupp 2013, American Sociological Review 78:26 to 50CO-02C07-30behaviourcontested (figures garbled in the document)
MARS 2006, 932 couples. Equal-sharing couples 6.8 times a month, half a time more than traditional, twice as often as man-does-mostHousework and sexual frequency, 2006Carlson, Hanson and Frisco 2016, Journal of Marriage and Family 78:975 to 995CO-02C07-31behaviourrobust
Era, not error: surveys of 1992 to 1994 and 2006, fourteen years apart, opposite signsReading of the two surveysCarlson et al.'s argument per CO-02, arithmetic from C07-30 and C07-31CO-02C07-32behaviourrobust
1,315 pairfam couples followed over time. Hours and a 50-50 split did not predict sexual frequency. Her perceived fairness and his judged contribution didFairness and sexual frequencyJohnson, Galambos and Anderson 2016, Archives of Sexual Behavior 45 (journal not named in the document)CO-02C07-33behaviourrobust
Same-sex couples divide more equally, not 50-50, by time, resources and preference. Talk about it explicitly. Report arrangements fairer even when hours unequalSame-sex couples as control groupGoldberg, Smith and Perry-Jenkins 2012 or 2013, JMF. Umberson, Thomeer and Lodge 2015, JMFCO-02, CO-06C07-34behaviourrobust
"Fairness paradox": objective inequality predicts dissatisfaction poorly, perceived unfairness predicts it wellPerceived justice in family labourMikula 1998, chapter. CO-02's first citation for it is bogus ("Google Sports Data")CO-02C07-35behaviouropen (phrase and year to confirm)
612,613 cycles from 124,646 (CO-05) or 124,648 (BI-04) users of the Natural Cycles appSample of the cycle-length studyBull, Scherwitzl et al. 2019, npj Digital Medicine 2:83CO-05, BI-04C07-36physiologyrobust (user count differs by two)
Mean cycle 29.3 days, SD 5.2. 13 per cent exactly 28 days. 24 per cent of ovulations on day 14 or 15. Luteal phase about 12.4 days and stable, follicular variesCycle variabilityBull et al. 2019. The 13 per cent is sourced in BI-04 to a blog and a gynaecologists' pageCO-05, BI-04C07-37physiologyrobust
A woman may ovulate on day 12 one month and day 20 the next, so calendar counting misassigns hormonal stateIllustration of cycle variabilityBI-04's own gloss, no primaryBI-04C07-38physiologyrobust (illustration)
Fertile-day preference for masculine faces did not survive salivary hormone measurementOvulatory shift, chapter 5 cross-referenceJones et al. 2018, Psychological Science, N = 584. Chapter 5 row C05-58BI-04C07-39behaviourfallen
Daily saliva: oestradiol predicts desire up, progesterone downHormones and sexual desire within the cycleRoney and Simmons 2013, Hormones and BehaviorBI-04C07-40physiologyrobust
26,000 diary entries, mid-cycle peak in in-pair and extra-pair desireDesire across the cycleArslan et al. 2021, journal not named. Chapter 3 row C03-43BI-04C07-41behaviourrobust
88 women, 68 completed a second cycle. Four points per cycle, hormones measured. First-cycle progesterone and working-memory association gone in cycle two, called a false positive. No consistent link to oestradiol, progesterone or testosteroneCycle and cognition, self-replicatedLeeners et al. 2017, Frontiers in Behavioral Neuroscience 11:120CO-05C07-42physiologyrobust (the first-cycle effect fallen)
78 studies (73 in the network analysis), women not on hormonal contraception. Early follicular against rest, ES minus 0.06, 95 per cent credible interval minus 0.16 to 0.04. Evidence rated low quality. Advise against population-wide rulesCycle and exercise performanceMcNulty et al. 2020, Sports Medicine 50:1813 to 1827CO-05C07-43physiologyrobust
Overlap about 98 per cent for d = 0.06Conversion of the effect sizeNarrator's arithmetic from the chapter 4 formula, row C04-06chapter 4C07-44behaviourrobust
Luteal progesterone raises core temperature about 0.4 degrees C and blunts the nocturnal fall. REM down. Light sleep and spindles up. Subjective sleep quality drops in the last week before the periodCycle and sleepBaker and Driver 2007, Sleep Medicine Reviews. The documents cite a different PMC paper and a sleep-clinic blogBI-04, CO-05C07-45physiologyrobust (primary to check)
17 studies, pooled PMS prevalence 47.8 per cent, self-report inflated by recallPrevalence of premenstrual syndromeDirekvand-Moghadam et al. 2014, Journal of Clinical and Diagnostic ResearchBI-04C07-46physiologyrobust
PMDD prevalence 3 to 8 per cent (BI-04) or about 5.5 per cent (CO-05)Prevalence of premenstrual dysphoric disorderAn Authorea preprint and a PMC primer. Check DSM-5 and Hantsoo and Epperson 2015BI-04, CO-05C07-47physiologycontested (report-to-report)
Diagnosis needs two cycles of daily charting because recall exaggeratesPMDD diagnostic standard (C-PASS)Eisenlohr-Moul et al. 2017, American Journal of PsychiatryBI-04C07-48physiologyrobust
Leuprolide in women with and without PMDD. Symptoms vanished. Add-back of oestradiol or progesterone at ordinary doses returned symptoms in the PMDD group onlyCause of PMDDSchmidt, Nieman, Danaceau, Adams and Rubinow 1998, New England Journal of Medicine 338:209 to 216BI-04C07-49physiologyrobust
Peter Schmidt's group at the National Institute of Mental Health in BethesdaPlaceBI-04 names no place. From the BI digest's scenes list and general knowledgeBI-04C07-50noneopen
22 women with PMDD. Suppression, single-blind placebo month, three months steady add-back. Symptoms returned in month one and subsided in months two and three while hormones stayed highPMDD as sensitivity to change, not levelSchmidt, Eisenlohr-Moul et al. 2017, American Journal of Psychiatry 174:980 to 989CO-05, BI-04C07-51physiologyrobust
Women with PMDD have the same hormone levels as controls. The brain responds to the rate of changeMechanism of PMDDSchmidt 1998 and 2017BI-04, CO-05C07-52physiologyrobust
SSRIs act within a day or two in PMDD rather than the weeks they take for depressionTreatment onset in PMDDBI-04's citation is a paper on allopregnanolone, not the primary for onset timeBI-04C07-53physiologyopen (primary to find)
Testosterone peaks in the early morning, falls through the day, moves with the seasonMale daily and seasonal rhythmSvartberg et al. 2003, JCEM 88:3099 to 3104 (seasonal). Brambilla's JCEM diurnal paper may be 2009. Chapter 3 row C03-19CO-05C07-54physiologyrobust (attribution to check)
565 people. Women peak 18:00 to 24:00 whatever their chronotype. Men more often ready in the morning. Only strong-lark women show a second peak 06:00 to 09:00Preferred time of day for sex by sex and chronotypeRandler et al. 2014 and 2017, Chronobiology InternationalCO-05C07-55behaviourrobust (which paper carries N = 565 to check)
Whether the women's evening peak is endocrine or the shape of the working day is unknownOpen question on the evening peakCO-05's own open question, no primaryCO-05C07-56behaviouropen
MCTQ asks when you sleep on free days, corrected for weekday sleep debt. Chronotype normally distributed. Drifts late through adolescence, peak lateness 19.5 years in women, 21 in men, then advances. "Social jetlag"Chronotype and social jetlagCited as Roenneberg et al. 2006, Current Biology. Likely merges three papers: Roenneberg, Wirz-Justice and Merrow 2003 (MCTQ), Roenneberg et al. 2004 Current Biology (19.5 and 21), Wittmann, Dinich, Merrow and Roenneberg 2006 Chronobiology International (social jetlag)CO-05C07-57physiologyrobust (citation to disentangle)
Roenneberg "and colleagues in Munich"PlaceInferred from the Munich ChronoType Questionnaire, not in the documentCO-05C07-58noneopen
Chronotype similarity predicts relationship satisfaction, sexual satisfaction and sleep quality across dating, newlywed and older couplesChronotype matching in couplesRandler 2014, Sprajcer et al. 2022, Díaz-Morales and Parra-Robledo 2019CO-05C07-59behaviourrobust ("gap matters more than direction" to check)
12 couples aged 18 to 29, four laboratory nights. Limb movement about 1.5 times higher together. REM about 10 per cent higher, 23 against 21 per cent of the night. Longer unbroken REM stretches, 22.0 against 13.4 minutesCo-sleeping and sleep architectureDrews et al. 2020, Frontiers in Psychiatry 11:583. Partly cited to a mattress retailer's blogCO-05C07-60physiologyrobust (a direction, not a weight)
89 first-time parents. Mothers' caregiving network also active in primary-caregiving fathers. Fathers' connectivity rose with hours of childcareCaregiving brain by roleAbraham, Feldman et al. 2014, PNAS 111:9792 to 9797. Chapter 8 rows C08-11, C08-12CO-02, FA-01C07-61physiologyrobust (the narrated edition leaves it to chapter 8)
Partners cannot detect fertilityFertility cuesGangestad 2019, Schleifenbaum 2022 (a Göttingen dissertation)BI-04C07-62behaviourrobust
"The eastern half of the panel, where the same bodies tip differently"East German mothers' earnings after a first birthBoelmann, Raute and Schönberg 2025, AEJ Applied (70 against 45 per cent of pre-birth earnings), social-security data, not FReDASY-06C07-63culturerobust (attribution to reword)
Lactation as the one bodily bottleneck no agreement movesPhysiological constraint on the division of labourSY-06 physiology paragraph, no primarySY-06C07-64physiologyrobust
"Cycle syncing" is a productMarketing trendCO-05 layer 3, no primaryCO-05C07-65cultureopen
A cooler bedroom in the luteal weekPractical consequence of the temperature riseMechanism from C07-45. CO-05's "16 to 18 degrees C" is from a mattress retailer and is not in the textCO-05C07-66physiologycontested
241 participants, women 10 ms faster and 25 per cent fewer errors while menstruatingReaction time across the cycleLikely Ronca et al. 2024, Neuropsychologia. BI-04 sources it to a Facebook post. Not in the data edition's text, only its working notesBI-04C07-67physiologyopen
First names supplied from general knowledge for 21 authorsAttributionNo primarynoneC07-68noneopen
Items in the data edition's lists that do not appear in its text (Hakim against Frodermann 2020, Schoep presenteeism, OR-04 shift work, Folke and Rickne, Coates and Herbert, Gettler 20.5 per cent)HousekeepingNoneSY-06, OR-04C07-69noneopen

The case for and the case against

The case the chapter makes

  • The second shift as a gap in total hours is not in the diaries. Men and women do the same total of paid and unpaid work, about eight hours a day, across fifty years and fourteen rich countries. Gershuny and Sullivan 2019, Annual Review of Sociology, MTUS (CO-02, rows C07-05, C07-06, C07-08).
  • What differs is the kind of minute. Women do about 60 per cent of the unpaid and 40 per cent of the paid, and the routine tasks sit with women in every country in the archive. Gershuny and Sullivan 2019 (CO-02, rows C07-09, C07-10).
  • Germany's care gap is closing slowly, and the routine part slowest. Destatis ZVE 2022 against 2012/13 (CO-02, rows C07-11, C07-12).
  • Mothers of young children drop about ten paid hours a week, fathers do not, and the routine tasks fall into those hours. Destatis ZVE 2022 (CO-02, row C07-13).
  • The interruptible minute costs more than the whole one. Offer and Schneider 2011, American Sociological Review, pager study of 693 parents (CO-02, rows C07-15, C07-16).
  • Housework has four stages and the two invisible ones, anticipating and monitoring, stay with the woman even where the chores are split. Daminger 2019, American Sociological Review, 35 couples (CO-02, row C07-17). First evidence that this tracks maternal depression and burnout: Aviv et al. 2024, 322 mothers (CO-02, row C07-18).
  • The tasks that went to women are the interruptible ones, which is the Campfire Contract's oldest clause. FReDA (SY-06, row C07-19), Brown 1970 (chapter 2, row C07-24).
  • Couples revert at the first birth almost invariably, and men in fixed-effects models are happier with more routine domestic work and fewer paid hours. FReDA (SY-06, rows C07-20, C07-21, C07-22).
  • The breadwinner requirement survived the requirement that she stay home. Killewald 2016, American Sociological Review, PSID (CO-02, CO-06, rows C07-25, C07-26).
  • Perceived fairness, not hours, predicts satisfaction and sex. Johnson, Galambos and Anderson 2016, pairfam (CO-02, row C07-33). Mikula 1998 (CO-02, row C07-35). Carlson, Hanson and Frisco 2016 against Kornrich, Brines and Leupp 2013 (CO-02, rows C07-30 to C07-32). Same-sex couples as the control group (CO-02, CO-06, row C07-34).
  • The cycle is measurable in the blood and all but invisible in the memory test and on the stopwatch. Bull et al. 2019 (rows C07-36, C07-37), Leeners et al. 2017 (row C07-42), McNulty et al. 2020 (row C07-43), all CO-05 and BI-04.
  • What survives is sleep and the minority with premenstrual dysphoric disorder, which is sensitivity to change in hormone level, not level. Baker and Driver 2007 (row C07-45), Schmidt et al. 1998 and 2017 (rows C07-49, C07-51, C07-52), BI-04 and CO-05.
  • Men run on a daily clock, the sexes' preferred hours for sex differ, and the couple's shared chronotype gap matters more than either. Randler 2014 and 2017 (row C07-55), Roenneberg (row C07-57), chronotype matching (row C07-59), Drews 2020 (row C07-60), all CO-05.

The case against

  • The cost of kinds rests on thin instruments: 35 couples, one cross-sectional survey of 322 mothers, one pager sample of 693 parents. Nothing in the multinational diary archive measures it, which CO-02 lists as its own open question (rows C07-15, C07-17, C07-18).
  • Against it, 1.2 million diary days, equal totals and core housework falling sharply for women across four decades. The second shift fell and the mental load is a name for the residue. Gershuny and Sullivan 2019 (CO-02, rows C07-05, C07-06, C07-10).
  • The reversion at the first birth is script, not body. West German wives raise their housework the year after their income passes their husband's. East German wives do not. Lippmann, Georgieff and Senik 2020, Economic Journal, SOEP 1991 to 2012 (SY-06, row C07-28).
  • The caregiving brain follows the ledger rather than writing it: the mothers' network lights up in primary-caregiving fathers and connectivity rises with hours of childcare. Abraham, Feldman et al. 2014, PNAS, 89 parents (CO-02, FA-01, row C07-61). The narrated edition leaves this to chapter 8.
  • The two clocks are a rounding error. Exercise performance moves by 0.06 of a standard deviation with an interval that includes zero (McNulty 2020, row C07-43). Working memory moves by nothing that survives a second cycle (Leeners 2017, row C07-42). Partners cannot detect fertility (Gangestad 2019, Schleifenbaum 2022, row C07-62). The sex-timing data are 565 questionnaires (row C07-55), the co-sleeping data 12 couples (row C07-60), and CO-05 cannot say whether the female evening peak is endocrine or timetable (row C07-56).
  • What predicts satisfaction and sex is perceived fairness, which no hormone sets and no diary measures. Johnson, Galambos and Anderson 2016, 1,315 couples (CO-02, row C07-33).
  • Hakim's preference theory against Frodermann 2020, listed as contested in the data edition but not argued in its text (SY-06, row C07-69).

Where the sources disagree

  • Whether the mental load is measurable beyond interviews and one survey. CO-02's own open question (rows C07-17, C07-18).
  • Killewald 2016 divorce hazard for a husband not in full-time work: about 30 per cent (CO-02) against 3.3 against 2.5 per cent a year, 33 per cent (CO-06). The data edition carried the range. The narrated edition says "about a third" (row C07-26).
  • Kornrich 2013 against Carlson 2016 on housework and sex: era, not error. The Kornrich figures ("5.2 times a month alongside wives reporting 5.6") are garbled in CO-02's text extraction and PDF. The press release reported about 1.5 fewer times a month for husbands in egalitarian households (rows C07-30 to C07-32).
  • Whether the female evening desire peak is endocrine or timetable. CO-05 open question (row C07-56).
  • Hakim's preference theory against Frodermann 2020 (SY-06, row C07-69).
  • Bull 2019 user count: 124,646 (CO-05) against 124,648 (BI-04). The narrated edition says "hundreds of thousands of cycles" (row C07-36).
  • PMDD prevalence: 3 to 8 per cent (BI-04) against about 5.5 per cent (CO-05). The narrated edition says "about one woman in twenty" (row C07-47).
  • Whether the luteal temperature rise justifies a cooler room. The mechanism is Baker and Driver 2007. The bedroom figure of 16 to 18 degrees C in CO-05 is sourced to a mattress retailer and is in neither version (rows C07-45, C07-66).
  • McNulty 2020: 78 studies in the pairwise analysis, 73 in the network analysis. The data edition used 78 (row C07-43).
  • Gettler 2011 testosterone drop: 20.5 per cent (CO-02) against 26 and 34 per cent (BI-02, BI-06). Not used in this chapter. Logged for chapter 8.
  • Hochschild's sample: "50 dual-earner couples" in CO-02, sourced to a 2015 retrospective, no second document (row C07-03).
  • The reversion at the first birth "on both sides of the old border": SY-06 does not split the FReDA finding by region, and the data edition's paragraph end ("absent in half of one country") conflated it with Lippmann's income-threshold response. The narrated edition keeps the reversion unsplit and gives the East and West finding only as Lippmann's (rows C07-20, C07-28).
  • "The German family panel" named both FReDA and pairfam in the data edition and still does in the narrated edition (row C07-23).

Open questions

The data edition's marks for checking, all removed from the narrated chapter and carried here.

  1. Slice length of the German ZVE diary. CO-02 says diaries are "generally" in 10-minute intervals, not that the ZVE is. Row C07-01. The narrated edition says "thin slices".
  2. Gershuny and Sullivan's affiliation, "who work from that archive". Row C07-07. The narrated edition says "who have spent their working lives in that archive", which rests on the same inference.
  3. Killewald's hazard: about 30 per cent (CO-02) against 33 per cent (CO-06). Row C07-26.
  4. Kornrich 2013 figures (5.2 times a month) garbled in CO-02 and in the PDF. Row C07-30.
  5. The London ISEH 241-person menstruation reaction-time study, sourced in BI-04 to a Facebook post. Not in either edition's text. Row C07-67.
  6. Schmidt 1998 at the NIMH in Bethesda: the place is from the BI digest's scenes list, not BI-04. Row C07-50. The narrated edition says "a team at the National Institute of Mental Health" and drops Bethesda.

Further open points from the verification rows and the data edition's working notes.

  • Whether 20,000 in the ZVE 2022 is persons, households or diary days. Row C07-02.
  • Whether the Roenneberg citation merges three papers (the MCTQ 2003, the adolescence peak 2004, social jetlag 2006). Row C07-57. The narrated edition names Roenneberg for the questionnaire and the term and gives no year.
  • Whether Mikula used the phrase "fairness paradox" and in which year. CO-02's first citation for it is bogus. Row C07-35. The narrated edition does not name him.
  • Which Randler paper carries N = 565 and the exact time windows. Row C07-55.
  • Which measure gives 53 per cent for the German pension gap, often reported at 46 to 49. Row C07-29.
  • Whether "the eastern half of the panel" in the data edition's In practice is Boelmann, Raute and Schönberg (social-security data) rather than FReDA. Row C07-63. The narrated edition says "the East German mother" and names no panel.
  • Whether Brambilla's diurnal paper is 2003 or 2009. Row C07-54.
  • Whether the Preface's FReDA paragraph (its cut order item 2) moves here. The narrated edition refers to the Preface's figures and does not restate them.
  • Whether to name the Lois 2020 cohort finding (attitudes stalled for West German women born after 1966 and men born around 1956). Left out of both editions.
  • Whether "Scherwitzl" should carry a first name (Elina Berglund Scherwitzl, general knowledge). The narrated edition does not name the authors.
  • First names supplied from general knowledge for Gershuny, Sullivan, Offer, Schneider, Daminger, Killewald, Kornrich, Brines, Leupp, Carlson, Hanson, Frisco, Bull, Leeners, McNulty, Schmidt, Randler, Roenneberg, Mikula, Abraham, Feldman. Row C07-68. The narrated edition names only Hochschild, Gershuny, Sullivan, Daminger, Kornrich, Brines, Leupp and Roenneberg.
  • Items in the data edition's lists that its text never used (Hakim against Frodermann, Schoep presenteeism, OR-04 shift work, Folke and Rickne, Coates and Herbert, Gettler 20.5 per cent). Row C07-69.

The documents

  • CO-02 Division of labour in the modern couple: the diaries, the care gap, cognitive labour, divorce, pensions, housework and sex.
  • CO-05 Cycle-aware and rhythm-aware living: the cycle's measured effects, PMDD, the male daily tide, chronotype, co-sleeping, the app data.
  • BI-04 The menstrual cycle and mind: cycle variability, the ovulatory shift's fall, sleep, the PMDD mechanism, presenteeism.
  • OR-04 Work design and female physiology: consulted for presenteeism and shift work, both left to chapter 11 in this draft.
  • SY-06 Germany, East and West: the FReDA household mechanisms and the household response to a wife's income only. The national story is chapter 13's.
  • CO-06 Control-group structures: not on this chapter's list, consulted for the Killewald discrepancy and the Australian census figure.
  • the CO and FA digest, the BI digest, the OR and PS digest: the distilled sheets.

Its documents in the Library

160 162 164 209 212 213 215 223 239