The Contingent Effects of Sexism in Primary Elections

• Author: Danny Hayes,Jennifer L. Lawless
• DOI: http://doi.org/10.1177/10659129211043134
• Published date: 01 December 2022
• Date: 01 December 2022

Article

PoliticalResearchQuarterly

2022,Vol.75(4)1021–1036

©TheAuthor(s)2021

Articlereuseguidelines:

sagepub.com/journals-permissions

DOI:10.1177/10659129211043134

journals.sagepub.com/home/prq

TheContingentEffectsofSexismin

PrimaryElections

DannyHayes

1

andJenniferL.Lawless

2

Abstract

AlthoughthelandscapeforfemalecandidatesinU.S.politicshasimproved,researchcontinuestofindthatmanyvoters

possesssexistattitudes.Werelyonastandardpoliticalcommunicationframeworktohelpreconcilesexisminthe

electoratewithincreasinglyfavorableoutcomesforwomeninprimaryelections.Basedontwonationalsurveyex-

periments,wefirstdemonstratethatintheabsenceofgenderedcampaignrhetoric,sexismisaweakpredictorof

supportforfemalecandidatesonbothsidesofthepoliticalaisle.Wethenshow,however,thatwhenamalecandidate

attemptstoactivatesexismamongvotersbyattackingafemaleopponent,genderattitudesbecomemoresalient—but

nottothewoman’sdisadvantage.InaDemocraticprimary,genderedattacksbackfireandleadtoasignificantboostin

supportforthefemalecandidate.OntheRepublicanside,amalecandidatedoesnotfacethesamebacklash,butthe

attacksdoverylittletodepresshisfemaleopponent’ssupport.Whilethepersistenceofhostileattitudestowardwomen

hasslowedthemarchtowardgenderequalityinsociety,ourexperimentalresultssuggestthatsexismexertsonly

contingenteffectsinprimaryelectionsandnotsystematicallytofemalecandidates’detriment.

Keywords

sexism,primaryelections,genderstereotypes,campaignrhetoric

Inthelastdecade,thenumberofwomenholdingelective

officeintheUnitedStatesincreasedsubstantially.Be-

tween2011and2021,femalerepresentationgrewby58%

inCongress,38%amongstatewideelectedofficials,and

30%instatelegislatures.

1

Nearlyone-thirdofthe44

womenevertooccupythegovernor’smansionserved

duringthelastdecade.

2

Andin2020,KamalaHarris

becamethefirstwomanelectedvicepresident.Although

Democratsaccountformuchofthesegains,thenumberof

RepublicanwomeninCongressalsoincreased

markedly—by81%inthelast10yearsalone.

Thisdevelopmentisnotablegiventhepersistenceof

sexismwithintheelectorate.Surveyscontinuetofindthat

manyvoterspossessattitudesthatsuggesthostilitytoward

equaltreatmentforwomen.Forinstance,ina2020poll,

nearlyone-thirdofU.S.respondentssaidthat“when

womendemandequalitythesedays,theyareactually

seekingspecialfavors.”Askedif“womenwhocomplain

aboutharassmentoftencausemoreproblemsthanthey

solve,”47%ofRepublicans,26%ofindependents,and

15%ofDemocratsagreed.

3

Scholarsoftenpointtothepolarizednatureof

Americanelectoralpoliticstoexplainwhythecontinued

existenceofsexismisnotatoddswiththeriseinwomen’s

representation.Becausevoters’partisanattachmentshave

grownsostrong,theywillalmostalwaysvotefortheir

party’snominee;candidatesex(oranyothercharacter-

istic)exertsonlyminimalinfluenceonvoters’attitudes

(Brooks2013;Dolan2014;HayesandLawless2016;

Teele,Kalla,andRosenbluth2018).Putsimply,even

unabashedsexistsfinditmoreappealingtocastaballot

forafemaleco-partisanthantocrosspartylines.

Butpartisanshipcannotexplainwhywomenofboth

partiesnowwinprimaryelectionsinrecordnumbersas

well.

4

Afterall,genderandsexismshouldplayastronger

roleincampaignsinwhichvoterscannotrelyonparty

cues(Hayes2011;HayesandLawless2016;Kingand

Matland2003;McDermott1997).Whenrunningagainst

menofthesameparty,womentheoreticallyshouldhavea

hardtimewinningoversexistvoters.Andthefactthat

1

GeorgeWashingtonUniversity,DC,USA

2

UniversityofVirginia,VA,USA

CorrespondingAuthor:

DannyHayes,DepartmentofPoliticalScience,GeorgeWashington

University,MonroeHall,2115GStreet,NWWashington,DC20052,

USA.

Email:dwh@gwu.edu

malecandidatescouldexploitsexisminanefforttodrive

downsupportfortheirfemaleopponentswouldseemto

poseanotherformidablechallenge.Yet,inprimary

electionsforbothCongressandstatelegislatures,women

arejustasableasmentosecurepartynominations.

Inthisarticle,werelyonanexperimentalapproachto

helpreconcilesexistattitudesintheelectoratewithin-

creasinglyfavorableoutcomesforwomeninprimary

elections.Basedontwonationalsurveyexperiments,we

showthatsexismplaysanuancedroleincontestsforparty

nominations.Wefirstdemonstratethatintheabsenceof

genderedcampaignrhetoric,sexismisaweakpredictorof

supportforfemalecandidates.Wethenshow,however,

thatwhenamalecandidateattemptstoactivatesexism

amongvotersbyattackinghisfemaleopponent,gender

attitudesbecomemoresalient—butnottothewoman’s

disadvantage.InaDemocraticprimary,wherethevast

majorityofvoterspossess“non-sexist”attitudes,gen-

deredattacksbackfireandleadtoasignificantboostin

supportforthefemalecandidate.OntheRepublicanside,

wheresexismismoreprevalentamongvoters,amale

candidatedoesnotfacethesamebacklash.Buteven

amongthemostsexistGOPprimaryvoters,theattacksdo

verylittletodepresshisfemaleopponent’ssupport.Al-

thoughthepersistenceofhostileattitudestowardwomen

hasslowedthemarchtowardgenderequalityinsociety,

ourexperimentalresultssuggestthatsexismexertsonly

contingenteffectsinprimaryelectionsandnotsystem-

aticallytofemalecandidates’detriment.

HowSexismCouldAffect

PrimaryElections

Previousworkonfemalecandidatesinprimaryelections

hastakenanumberofapproachestoidentifythepotential

effectsofsexism.Someresearchershaveassessed

women’sfundraisingandwinratesinprimariestode-

terminewhethertheyfacedisadvantagesorbias(e.g.,

Anastasopoulos2016;Barnes,Branton,andCassese

2017;Burrell2014;HassellandVisalvanich2019;

KitchensandSwers2016;LawlessandPearson2008;

PalmerandSimon2008;PearsonandMcGhee2013;

Thomsen2019;2020).Othershavefocusedonlatent

genderstereotypesamongvotersthatcouldinfluence

attitudestowardfemalecandidates(e.g.,Bauer2017;

CasseseandHolman2018;KingandMatland2003;

Lawless2004;SanbonmatsuandDolan2009).Andstill

othershaveexploredthewaysthatsexistattitudesshape

howvoterssearchforinformationaboutcandidates,often

tofemalecontenders’disadvantage(Ditonto2019;

Ditonto,Hamilton,andRedlawsk2014).

Forthemostpart,however,scholarshavenotapplieda

standardpoliticalcommunicationframeworktostudy

howsexismaffectsfemalecandidates’fortunesinprimary

campaigns.Theclassicalmodelofpersuasion(e.g.,

Hovland,Janis,andKelley1953)suggeststhatpeople’s

politicalopinions—suchasvotechoice—typically

emergefromacombinationoftheirownpriorbeliefs

andtheinformationtheyareexposedtoinpublicdis-

course(e.g.,McGuire1968;Zaller1992).Thismeansthat

inelections,people’sexistingattitudes,orpredispositions,

donotdeterministicallydictatetheirchoices.Rather,

thoseattitudesareactivatedwhenvotersencounterrel-

evantcommunicationsinthecampaignenvironment

(Bartels1988;HillygusandJackman2003;Sidesand

Vavreck2013).Thecoreinsightofthisbodyofresearchis

that“campaignmessages...worktheirinfluencein

concertwithvoters’prevailingpredispositionsandsen-

timents”(IyengarandSimon2000,158).Bysimulta-

neouslyconsideringbothcampaignmessagesandvoters’

keypredispositions,thisframeworkallowsforthederi-

vationofarichersetofexpectationsregardingtherole

sexismwillplayinprimarycampaignsinvolvingmenand

women.

First,weexpectthatsexismwillnotautomatically

influencesupportforfemalecandidates.Instead,itshould

playitsstrongestrolewhencampaignrhetoricmakes

gender-relatedattitudessalient.Wegroundthisexpecta-

tioninpriorresearchthatsuggestsvoters’beliefsabout

genderstereotypesshapetheirsupportforfemalecan-

didatesonlywhencampaignactivitybringsthosecon-

siderationstomind(Bauer2015;Hayes2011).For

instance,inBauer’s(2015)experiment,respondentsex-

posedtonewsstoriescontaininggenderedlanguagewere

morelikelytojudgefemalecandidatesinstereotypical

waysthanweresubjectswhosawcoveragewithoutthat

language.Thesefindingsareconsistentwithstudiesar-

guingthatinsomecasesagenderedcampaignenviron-

mentcancreateaseriesofstrategicchallengesforwomen

(Bauer2017;CasseseandHolman2018;Dittmar2015;

KrupnikovandBauer2014;Windett2014).Incontrastto

anearlierlineofworkpositingthatcandidatesexisa

chronicallyinfluentialfeatureofvoters’choices(e.g.,

Sanbonmatsu2002),theemergingconsensusisthatthe

influenceofgenderattitudesis“highlyconditionaland

[dependent]onthetypesofinformationvotersreceive

overthecourseofacampaign”(Bauer2015,705).

Althoughresearchthatdirectlymeasuressexismin

primarycampaignsissparse,analysesofthepresidential

racebetweenHillaryClintonandDonaldTrumpare

consistentwiththisperspective.Numerousstudiesofthe

2016electionfindstrongcorrelationsbetweensexismand

votechoice(Bock,Byrd-Craven,andBurkley2017;

Bracic,Israel-Trummel,andShortle2019;Knuckey2019;

Schaffner,MacWilliams,andNteta2018;Sides,Tesler,

andVavreck2020).Butthatrelationshipappearsdueat

leastinparttothehighlygenderednatureofthecampaign.

1022PoliticalResearchQuarterly75(4)