THRIVING after the FALL
Book Two PAIRING after the FALL

PAIRING · Part III · Emancipation 2.0

The numbers and sources

Chapter 14: Queens at War

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.

  • 53 open
  • 15 to check first

Read about the chapter · Its 53 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. 39 percentage points: how much likelier a polity was to be at war in a given year under a queen than under a king, 193 reigns, 18 European polities, 1480 to 1913 (Dube and Harish, Journal of Political Economy, 2020, per SO-01). Row C14-18.
  2. 2.1 percentage points, a third of the baseline: the fall in neonatal mortality for a ten-point rise in women's share of an Indian state legislature (Bhalotra and Clots-Figueras 2014, per SO-01). Row C14-05. 8 to 12 per cent: the fall in maternal mortality after a national quota (Bhalotra et al. 2023, per SO-01). Row C14-06.
  3. Twice as often: how much more women spoke in village council meetings with a woman in the chair, 265 councils (Chattopadhyay and Duflo 2004, per SO-01 and SO-07). Row C14-02. Five times: a woman's odds of winning a seat once reserved against one never reserved (Bhavnani 2009, per SO-01). Row C14-04.
  4. 63 per cent longer: men's federal sentences for the same arrest offence and history, and twice the odds of avoiding prison for women (Starr, American Law and Economics Review, 2014, per SO-04). Row C14-35. 22 against 4 per cent: the share of French mixed co-offender pairs in which the man, or the woman, got the longer sentence (SO-04, attribution open). Row C14-36.
  5. 23.9 against 11.2 per cent, relative risk 2.13: stress fractures in women and men under identical Israeli anti-aircraft training (Finestone et al., Military Medicine, 2005, per SO-04). Row C14-31. 12.0 against 1.3, relative risk 9.23, in light infantry (Finestone et al. 2014, per SO-04). Row C14-32.

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
1993, one chair in three, chosen at randomIndia's constitutional amendment reserving village council chairs for women, with reserved councils drawn at randomChattopadhyay and Duflo 2004, EconometricaSO-01C14-01culturerobust
265 councils, more spending on drinking water in West Bengal, women spoke about twice as oftenWhat a woman in the chair does to a council's budget and to the women in the roomChattopadhyay and Duflo 2004, Econometrica, 265 councils in West Bengal and RajasthanSO-01, SO-07C14-02culturerobust
Fewer roads in Rajasthan (SO-01) or more roads in West Bengal (SO-07)Road spending by women-led councilsChattopadhyay and Duflo 2004, EconometricaSO-01, SO-07C14-03culturecontested
About five timesA woman's odds of winning the next open election in a Mumbai ward where a reservation had lapsed, against a never-reserved wardBhavnani 2009, American Political Science ReviewSO-01C14-04culturerobust
2.1 percentage points, a third of the baseline of 33.5 per centFall in neonatal mortality for a ten-point rise in women's share of Indian state legislature seats, via antenatal visits, iron supplements and facility birthsBhalotra and Clots-Figueras 2014, American Economic Journal: Applied EconomicsSO-01C14-05culturerobust
Maternal mortality ratio down 8 to 12 per cent, skilled birth attendance up 5 to 8 points, prenatal care up 4 to 8 points, fertility down 6 to 7 per cent, girls' schooling up about half a yearEffects of adopting a parliamentary quota for women in developing countries, staggered adoption datesBhalotra, Clarke, Gomes and Venkataramani 2023, Journal of the European Economic AssociationSO-01C14-06culturerobust
15 per centHow much further maternal deaths fell after the 1937 arrival of sulfa drugs in US states that had enfranchised women earlyAttributed by SO-01's text to Bhalotra et al. 2023 and by its works cited to a separate 2023 paper (UCLouvain)SO-01C14-07cultureopen
4,265 close elections, 1.8 percentage points a year faster growth, men three times as likely to carry criminal charges, women's asset accumulation about 10 points a year lowerEffects of a woman narrowly winning an Indian state assembly seat, growth measured by night-time light from satellitesBaskaran, Bhalotra, Min and Uppal 2018 (working paper) and 2023, Journal of Economic Growth, 4,265 close electionsSO-01C14-08culturerobust
One standard deviation more women in parliament, a fifth of a standard deviation less perceived corruption, more than 100 countriesCross-country correlation of women's share of seats with perceived corruptionDollar, Fisman and Gatti 2001, Journal of Economic Behavior and Organization, more than 100 countriesSO-01C14-09culturecontested
Holds in democracies, vanishes in autocraciesWhether the women-corruption relationship depends on regime typeEsarey and Chirillo 2013, Politics and Gender, "Fairer Sex or Purity Myth?"SO-01C14-10culturerobust
76 democratic-leaning countriesWomen reduce corruption only where electoral accountability is highEsarey and Schwindt-Bayer 2018, American Journal of Political Science (SO-01 text) or British Journal of Political Science (works-cited filename), 76 countriesSO-01C14-11culturerobust
Higher ethical standard, harsher punishment of a woman's scandalVoters' treatment of women politicians caught in scandal, experimentsBarnes and Beaulieu 2014 and 2018, venues not givenSO-01C14-12culturerobust
Women more risk-averse and more context-sensitive on averageReview of laboratory gamesCroson and Gneezy 2009, Journal of Economic LiteratureSO-01C14-13behaviourcontested
147 estimates from 25 studiesPublication bias inflates the sex gap in risk aversion, within-sex variation dwarfs between-sexNelson 2015, Feminist Economics, 147 estimates, 25 studiesSO-01C14-14behaviourrobust
93 per centShare of between-subject comparisons contradicting the claim that women's social preferences are more context-sensitiveDellaVigna and Pope, NBER working paper, 2022 in SO-01's text, NBER number w35324 suggests 2025SO-01C14-15behaviourrobust, year open
No figureFemale heads of state sit in the tails of the distributions and are hard to tell from male peers on competitiveness, risk tolerance and dominanceSO-01's own wording, no primary source namedSO-01C14-16behaviouropen
Higher rates at the same expected inflationWomen on monetary policy committees vote more hawkishly than menMasciandaro, Profeta and Romelli 2020, European Journal of Political Economy, verified under chapter 10SO-07C14-17behaviourrobust
193 reigns, 18 polities, 1480 to 1913, 39 percentage pointsHow much likelier a polity under a queen was to be at war in a given year, instruments: first-born son and sister of the previous monarchDube and Harish 2020, Journal of Political Economy, "Queens", 193 reignsSO-01C14-18culturerobust
Unmarried queens more often attacked, married queens more often aggressors, more so than married kingsThe split in the queens finding and the dual-executive explanationDube and Harish 2020, Journal of Political EconomySO-01C14-19culturerobust as pattern, mechanism inferred
No figureDube and Harish's affiliations, S. P. Harish given with initials onlyTitle page of Dube and Harish 2020SO-01C14-20not applicableopen
No figure in SO-01The physical security of a state's women predicts its peacefulness better than wealth, democracy or religionHudson et al. 2012, Sex and World PeaceSO-01C14-21culturecontested
130,000 data points, 175 countriesSize of the WomanStats database, and Hudson's co-authors given as Caprioli, Bowen and McDermottSY-05, a synthesis report whose book sources are Goodreads and Bookey, no primary source namedSY-05C14-22cultureopen
186 or 185 societies, zero with women in military leadershipStandard Cross-Cultural Sample, assembled 1969, count given two ways in DH-07Murdock and White 1969, Ethnology, and the SCCS military leadership variableDH-07C14-23culturerobust, count open
156 societiesInner and outer orientation, outer orientation travelling with high interpersonal violence and rape-prone cultureSanday 1981, Female Power and Male Dominance, Cambridge University Press, 156 societiesDH-07C14-24culturerobust
No figureThe male supremacy complex: warfare rewards young men with patrilocal residence, polygyny, political monopoly and the subordination of womenDivale and Harris 1976, American AnthropologistDH-07C14-25culturecontested
No figurePathogen stress and male war mortality as the two best predictors of non-sororal polygyny, with controls for shared ancestryLow, year and venue not in DH-07DH-07C14-26cultureopen
About 85 per cent (SO-05) or over 80 per cent (DH-07)Share of societies in the ethnographic record permitting polygyny, and the group-level effects of normative monogamyHenrich, Boyd and Richerson 2012, Philosophical Transactions of the Royal Society BSO-05, DH-07C14-27culturerobust
No figureSurplus young men and state violence, sex ratios, polygyny and instabilityHudson and den Boer, not in any document in the librarynone, (SO-02, still to be written)C14-28cultureopen
93 societies, about ten dimensionsDimensions of women's standing barely correlate with one anotherWhyte 1978, The Status of Women in Preindustrial Societies, Princeton University Press, 93 societiesDH-07C14-29culturerobust
2013 law, in force 2015, first NATO member, 14 per cent of the force by 2020 against a 20 per cent targetNorway's conscription of women, mixed dormitories and de-genderisation at Skjold and HeggeliaHellum 2014, Norwegian Defence Research Establishment report, and a DCAF report, title and year not givenSO-04C14-30culturerobust, 14 per cent open
23.9 per cent (33 women) against 11.2 per cent (42 men), relative risk 2.13Stress fractures under identical basic training in the Israeli Anti-Aircraft Forces, women's fractures in tibia, femur and pelvisFinestone, Milgrom, Moran et al. 2005, Military MedicineSO-04C14-31physiologyrobust
12.0 against 1.3 per cent, relative risk 9.23Stress fractures in light infantryFinestone et al. 2014, venue not givenSO-04C14-32physiologyrobust
About 10 against 3 per cent, hazard more than fourfoldStress fractures in the first six months of US Navy trainingMoran et al. 2012, venue not givenSO-04C14-33physiologyopen
No figureIndividual women with high baseline fitness complete the toursSO-04's wording, no primary source namedSO-04C14-34physiologyopen
63 per cent longer, twice as likely to avoid prison, about 30 per cent of the gap unexplained, about six times the racial gapUS federal sentencing from arrest to sentence, same arrest offence and criminal historyStarr 2014, American Law and Economics ReviewSO-04C14-35behaviourrobust
15 days shorter, a third of the 45-day mean, man longer in 22 per cent of mixed co-offender pairs and woman in 4, one standard deviation more female judges shrinks the gap by about 10 per centFrench sentencing, every conviction 2000 to 2003Attributed in a table only to "Breda 2020"SO-04C14-36behaviouropen
80 to 84 per centContested custody cases ending in sole physical custody for the motherNo primary source in SO-04SO-04C14-37cultureopen
d = 0.24Joint-custody children closer to intact-family children, joint-custody parents reporting less conflictBauserman 2002, Journal of Family Psychology, meta-analysisSO-04C14-38behaviourrobust
60 studies, better in 34, equal or better in 14, worse on one measure in 6Outcomes for children in joint custodyNielsen 2018, venue not given, 60 studiesSO-04C14-39behaviourrobust
110 signatoriesConsensus report that the evidence did not support denying infant overnights to fathers, after McIntosh 2010 and 2013Warshak 2014, Psychology, Public Policy, and LawSO-04C14-40behaviourrobust
No figureDespotic decisions cost a group more than shared ones when members' interests differConradt and Roper 2003, Nature (SO-07 also cites a 2005 Trends in Ecology and Evolution review)SO-07C14-41behaviourrobust
No figureHigher share of women in parliament goes with stricter climate policy and lower carbon emissions, instrumental variablesMavisakalyan and Tarverdi 2019, European Journal of Political EconomySO-07C14-42culturecontested
194 countries, about half as many deathsWomen-led countries locked down earlier in the first wave of 2020Garikipati and Kambhampati 2021, Feminist Economics, 194 countries matched on wealth and demographySO-07C14-43culturecontested
No figureThe countries which elect women are high-trust, low-hierarchy countries, which explains the pandemic and climate resultsWindsor et al. 2020, venue not givenSO-07C14-44cultureopen
1,200 juror interviews, 353 capital cases, 14 US statesWhite male jurors most willing to sentence to death, women weigh mitigating evidence more heavilyBowers et al., Capital Jury Project, venues and years not givenSO-07C14-45behaviourrobust, venues open
No figureMen extended property rights and later the vote to women as the return on children's education roseDoepke and Tertilt 2009, Econometrica, "Women's Rights: What's in it for Men?"SO-05C14-46culturerobust
140 studiesWomen on boards: small positive relation with accounting returns, near zero with market performancePost and Byron 2015, Academy of Management Journal, 140 studiesSO-07C14-47behaviourrobust
No effectWomen directors and equity risk in a large sample of US firmsSila, Gonzalez and Hagendorff 2016, Journal of Corporate FinanceSO-07C14-48behaviourrobust
Fewer layoffs, lower short-run earningsNorwegian quota boardsMatsa and Miller 2013, American Economic Journal: Applied EconomicsSO-07C14-49behaviourrobust
200 studies, 14,548 participants, 66 per cent male in the earlier stress literatureTend-and-befriend, resting on oxytocinTaylor et al. 2000, Psychological ReviewSO-01C14-50physiologycontested
No figureEconomic-marginalisation theorists predict that equality for women in the dock means more women in prisonSO-04, no primary source namedSO-04C14-51behaviouropen

The case for and the case against

The case the chapter makes

  • Random reservation of village council chairs moved spending towards women's priorities and roughly doubled women's speaking at meetings (Chattopadhyay and Duflo 2004, SO-01, SO-07, C14-01, C14-02).
  • A lapsed reservation left a woman about five times as likely to win the open seat (Bhavnani 2009, SO-01, C14-04).
  • More women in Indian state legislatures cut neonatal mortality by a third of the baseline. National quotas cut maternal mortality by 8 to 12 per cent (Bhalotra and Clots-Figueras 2014, Bhalotra et al. 2023, SO-01, C14-05, C14-06).
  • Close elections won by women went with faster growth, fewer criminal charges and slower asset accumulation (Baskaran et al. 2018 and 2023, SO-01, C14-08).
  • Women reduce corruption only where electoral accountability is high, and are punished harder for scandal, so the effect is about the audience, not the sex (Esarey and Chirillo 2013, Esarey and Schwindt-Bayer 2018, Barnes and Beaulieu 2014 and 2018, SO-01, C14-10 to C14-12).
  • The risk-aversion gap shrinks once publication bias is corrected, and the context-sensitivity claim failed replication (Nelson 2015, DellaVigna and Pope, SO-01, C14-14, C14-15).
  • Polities under queens were 39 percentage points likelier to be at war, married queens as aggressors with the husband as co-executive, unmarried queens as targets (Dube and Harish 2020, SO-01, C14-18, C14-19).
  • The physical security of women predicts a state's peacefulness better than wealth, democracy or religion (Hudson et al. 2012, SO-01, C14-21).
  • Outer-oriented societies under stress and war rely on hunting and fighting and subordinate women. The male supremacy complex. Normative monogamy reduces unmarried men and violence (Sanday 1981, Divale and Harris 1976, Henrich, Boyd and Richerson 2012, DH-07, SO-05, C14-23 to C14-27).
  • Women's standing has about ten dimensions that barely correlate (Whyte 1978, DH-07, C14-29).
  • Identical training produces about twice the stress fractures in women, nine times in light infantry. High-fitness individuals still complete the tours (Finestone et al. 2005 and 2014, SO-04, C14-31, C14-32, C14-34).
  • Men receive sentences 63 per cent longer for the same file. French co-offender pairs show the same. More women judges shrink the gap (Starr 2014, "Breda 2020", SO-04, C14-35, C14-36).
  • Joint custody outcomes favour children and reduce conflict. The 110-signature consensus on infant overnights (Bauserman 2002, Nielsen 2018, Warshak 2014, SO-04, C14-38 to C14-40).
  • Shared decisions land nearer the group's median interest (Conradt and Roper 2003, SO-07, C14-41). Women in parliament go with lower emissions and earlier lockdowns (Mavisakalyan and Tarverdi 2019, Garikipati and Kambhampati 2021, SO-07, C14-42, C14-43). Mixed juries sentence to death less often (Capital Jury Project, SO-07, C14-45).
  • Men extended rights to women when it paid them to (Doepke and Tertilt 2009, SO-05, C14-46).

The case against

  • Boards, the best-measured mixed bodies, show a near-zero effect of sex composition on performance and risk (Post and Byron 2015, Sila, Gonzalez and Hagendorff 2016, SO-07, C14-47, C14-48). The one robust board finding is a priority, not a competence (Matsa and Miller 2013, SO-07, C14-49).
  • The countries that elect women are high-trust, low-hierarchy countries, which explains the pandemic and climate results without the sex of the leader (Windsor et al. 2020, SO-07, C14-44).
  • The Hudson correlation may run from wealth, stability and democracy to both women's security and peace (SO-01's own critics, C14-21).
  • The risk-aversion gap was partly manufactured by publication bias (Nelson 2015, C14-14). The context-sensitivity claim fell (DellaVigna and Pope, C14-15). Tend-and-befriend rests on oxytocin, which the Preface's verdict undercuts (Taylor et al. 2000, SO-01, C14-50).
  • The queens finding is one continent's early-modern reigns and the dual-executive mechanism is an inference from who attacked whom (Dube and Harish 2020, C14-18, C14-19).

Where the sources disagree

  • Whether the Hudson correlation is cause, consequence or both from a third thing (SO-01's own critics, SY-05 for the figures). C14-21, C14-22.
  • The roads finding in Chattopadhyay and Duflo: SO-01 says women's councils spent less on roads in Rajasthan, SO-07 says more on drinking water and roads in West Bengal. The SO fact sheet logs the same. C14-03.
  • Whether the climate and pandemic results survive the culture-that-elects-women objection (Windsor 2020). C14-42 to C14-44.
  • Whether the risk-aversion gap survives Nelson 2015, and whether tend-and-befriend (Taylor 2000) survives the Preface's oxytocin verdict. C14-13, C14-14, C14-50.
  • The 80 to 84 per cent custody figure (no primary source) and the French study's attribution ("Breda 2020", table only), both SO-04. C14-36, C14-37.
  • Everything that belongs in (SO-02, still to be written). C14-28.
  • Becker's pastoralism paper: DH-07 gives Becker 2019, Review of Economic Studies, over 500,000 women in 34 nations, plus a 2024 working paper with 1.8 million respondents in 122 countries, while SO-05 gives Becker 2024, Review of Economic Studies, 1.3 million in over 100 countries. Not used in prose for that reason.
  • Polygyny permitted: "over 80 per cent" of Ethnographic Atlas societies (DH-07) against "approximately 85 per cent" (SO-05, Henrich). The data edition carried "about 85 per cent" with Henrich as source. The narrated edition carries "most". C14-27.
  • Esarey and Schwindt-Bayer 2018: SO-01 text says American Journal of Political Science, the works-cited PDF filename says BJPS. C14-11.
  • Eagly and Wood 1999 venue: American Psychologist (SO-04) against Psychological Bulletin (SO-05). Not used in prose.
  • SCCS size: 186 societies (DH-07 introduction) against 185 (DH-07 on Murdock and Provost and on military leadership). C14-23.
  • Mavisakalyan and Tarverdi: the Hofstede analysis (18 countries, 20 years) is attributed to them for 2019 but cited to a 2025 paper (SO-07). Not used in prose. C14-42.
  • The 1937 sulfa-drug finding is attributed by SO-01's text to the 2023 JEEA paper and by its works-cited to a separate 2023 paper. C14-07.
  • DellaVigna and Pope year: SO-01 says 2022, the works-cited NBER number (w35324) suggests 2025. C14-15.

Open questions

The data edition's marks for checking, with rows:

  1. Roads finding: SO-01 (fewer in Rajasthan) against SO-07 (more in West Bengal). C14-03.
  2. The 1937 sulfa-drug and early-suffrage finding: same 2023 JEEA paper (SO-01 text) or a separate 2023 paper (SO-01 works cited, UCLouvain). C14-07.
  3. DellaVigna and Pope year: SO-01 says 2022, the works-cited NBER number (w35324) suggests 2025. C14-15.
  4. Dube and Harish affiliations, not given in SO-01. C14-20.
  5. Hudson's co-authors (Caprioli, Bowen, McDermott) from SY-05, not SO-01. C14-22.
  6. WomanStats: 130,000 data points, 175 countries, from SY-05. C14-22.
  7. SCCS count: 186 (DH-07 on Murdock and White 1969) against 185 (DH-07 on military leadership). C14-23.
  8. Low: year and venue not given in DH-07. C14-26.
  9. den Boer: not in any library document. The marker stays in the chapter's (SO-02, still to be written) line, which the narrated edition keeps verbatim. C14-28.
  10. Norwegian 14 per cent by 2020 (DCAF report, SO-04). C14-30.
  11. Moran et al. 2012, US Navy about 10 against 3 per cent, and the "more than fourfold" hazard. C14-33.
  12. French sentencing study attributed to "Breda 2020" in a table only (SO-04). C14-36.
  13. 80 to 84 per cent custody figure, no primary source (SO-04). C14-37.
  14. Capital Jury Project venues and years (SO-07). C14-45.
  15. Windsor 2020 venue (SO-07). C14-44.
  16. Barnes and Beaulieu 2014 and 2018, Nielsen 2018, Hellum 2014: venues not given in the documents (no marker in the data edition's text, logged in rows). C14-12, C14-39, C14-30.

The data edition's open questions from its working notes:

  • Whether the opening may say the chair was "decided at random" (SO-01 says "randomly reserved", the SO fact sheet says "drawn by lot"). Both editions say "at random".
  • Den Boer is a named person in the chapter note but appears in no document in the library. She is named only inside the (SO-02, still to be written) paragraph. If SO-02 is re-run, her material fills the gap and the name moves into prose.
  • Dube and Harish's affiliations are not in SO-01 and are not supplied in either edition.

The documents

  • SO-01 Women in political power and what states do: the lotteries, the queens, corruption.
  • SO-04 Where the law treats the sexes differently: conscription, sentencing, custody, leave.
  • SO-07 Mixed versus single-sex decision bodies and collective risk: Conradt and Roper, climate, pandemic, juries, and the board nulls for the case against.
  • DH-07 Cross-cultural databases: Whyte, Sanday, Divale and Harris, Low.
  • SO-05 Cultural evolution: Henrich, Boyd and Richerson on monogamy, Doepke and Tertilt on men yielding rights (state and world scale only).
  • SO-02 Sex ratios, polygyny and social stability: missing, see (SO-02, still to be written).

Its documents in the Library

187 227 230 231 233 238