You are on page 1of 17

Californias Proposition 8: What Happened, and What Does the Future Hold?

Patrick J. Egan New York University Kenneth Sherrill Hunter College-CUNY

Commissioned by the Evelyn & Walter Haas, Jr. Fund in San Francisco. Released under the auspices of the National Gay and Lesbian Task Force Policy Institute.

CaliforniasProposition8: WhatHappened,andWhatDoestheFutureHold?

PatrickJ.Egan NewYorkUniversity patrick.egan@nyu.edu KennethSherrill HunterCollegeCUNY kenneth.sherrill@hunter.cuny.edu January2009


2009PatrickJ.EganandKennethSherrill. ThisreportwascommissionedbyEvelyn&WalterHaas,Jr.Fund,afoundationwhose missionincludestheadvancementofthecivilrightsofgaysandlesbians.Wegratefully acknowledgethefoundationssupport.WealsoacknowledgeDavidBinderResearch, whichconductedthesurveyofCaliforniavotersanalyzedinthisstudy,andPeterFrase andtheCenterforUrbanResearchatTheGraduateCenter,CUNY,whichdevelopedand analyzedprecinctleveldataontheProposition8vote.Weremainfullyresponsiblefor theanalysisandinterpretationofresultsherein.


CaliforniasProposition8: WhatHappened,andWhatDoestheFutureHold?
OnNovember4,2008,CaliforniavotersapprovedProposition8whicheliminatedthe rightofsamesexcouplestomarrybya52to48percentmargin.Inthisstudy,we examinethreequestionsaboutthisresultusinganewsurveyofCaliforniavotersas wellasprecinctlevelelectionreturnsanddemographicdata.First,weexplorethe characteristicsofvotersthatwereassociatedwithsupportforandoppositionto Proposition8.Wefindthatvoterspartyidentification,ideology,religiosityandagehad amuchbiggerimpactonthevotethanothervotercharacteristics.Second,weexamine theAfricanAmericanvoteforProposition8.Weprovideevidenceshowingthatwhile AfricanAmericanssupportedProposition8morethanvotersasawhole,theydidnot dosointheoverwhelmingnumberssuggestedbyoneexitpoll.Weshowthatblack supportforProposition8canlargelybeexplainedbyAfricanAmericanshigherlevels ofreligiosityacharacteristicstronglyassociatedwithoppositiontosamesex marriage.Finally,weexaminehowCaliforniansopinionshaveshifteddramatically towardsupportofmarriageequalityovertheshorttimebetweentheKnightInitiative in2000andnow,andexploretheimplicationsofthischangeforthefuture.

PROPOSITION8:EXPLAININGTHEVOTE

VotersupportforProposition8splitmostsharplyalongthelinesofage, religiosity,andpoliticalviews.

Table1displaysfindingsfromapollofCaliforniavotersconductedbyDavidBinder Research(DBR)betweenNovember6thand16th,2008.Thesurveyincluded1,066 respondentsselectedatrandomfromstatevoterregistrationlists,includingan oversampleof266AfricanAmerican,Latino,andAsianAmericanvoters. 1 Participants wereaskedaseriesofquestionsaboutProposition8,aswellasbasicquestionsabout theirdemographicbackground,religion,politicalviews,andothercharacteristics.The sampleintheDBRsurveywaslimitedtothosewhoreportedvotingintheNovember4 generalelection,anditsmarginoferrorwas3percentagepoints(althoughthemargin isgreaterforanalysesofsubgroupswithinthesample).

1DataareweightedtorepresentthedemographiccharacteristicsofCaliforniavoters.

Table1.TheVoteonProposition8

(%of voters)

%votingYes onProposition8 52

Total Sex (46%) Men (54%) Women Age (17%) 1829 (21%) 3044 (38%) 4564 (23%) 65+ Race/Ethnicity (68%) White (7%) AfricanAmerican (14%) Latino/Hispanic (7%) Asian AttendanceofReligiousServices (45%) weekly (12%) monthly (14%) holidaysandspecialoccasions (29%) hardlyever PartyIdentification (45%) Democratic (18%) Independent (34%) Republican PoliticalIdeology (37%) Liberal (27%) Moderate (36%) Conservative HaveLesbian/GayFamilyorFriends (26%) No (74%) Yes

54 49

45 48 47 67

49 58 59 48

70 48 44 30

30 53 81

22 51 82

60 49

Source:DBRSurveyofCaliforniaVotersforEqualityCalifornia,November616,2008

AsshowninTable1,conservativesandRepublicanswerethemostlikelytosupport Proposition8:82%ofconservativesand81%ofRepublicansvotedinfavorofthe measure.Peoplewhoattendedreligiousservicesweekly(70%)andthoseoverage65 (67%)alsoapprovedProposition8bysubstantialmajorities.Menwereslightlymore likelytosupportProposition8(by54%)thanwomen(49%).Majoritiesofthoseunder age65opposedProposition8.AfricanAmericanandLatinovoterssupported Proposition8toagreaterdegree,58%and59%respectively,thandidwhitesand Asians.Accordingtothissurvey,blackssupportfortheballotmeasurewasmuch lowerthanreportedbyElectionDayexitpolls.(Anextensivediscussionaboutthis issueappearslaterinthisreport.)
3

ThevoteonProposition8waspolarizedtoaremarkabledegreealongthelinesofparty identification,ideology,andreligiosity.Thelargestdividefully60percentage pointswasbetweenconservativesandliberals(8222).Asimilarlylargegap(51 points)existedbetweenRepublicansandDemocrats.Byacommonlyusedmeasureof religiosityfrequencyofattendanceatreligiousservicesthemostreligious(those attendingservicesweekly)favoredProposition8by40percentagepointsmorethan theleastreligious(thosewhohardlyeverattendservices). ThreequartersofCaliforniavotersreportedknowinghavingfriendsorfamilymembers whoarelesbianorgay. 2 Fiftyonepercentofthesevoterscastballotsagainst Proposition8.Amongthosewhodonotknowanygaypeopleverywell,60%supported theamendmentand40%opposedit. 3 Significantdifferencesbetweenpopulationgroupsremainaftercontrolling forallvotercharacteristics. OnequestionthatarisesafterexaminingTable1iswhetherthedifferencesseen betweenpopulationgroupspersistafterholdingallothervotercharacteristicsconstant. WeanswerthisquestionwiththemultivariateanalysesshowninTable2,whichassess theextenttowhichvotercharacteristicshadindependentassociationswiththevoteon Proposition8aftercontrollingforothervariables.Thenumbersinthetableare estimatesofthedifferenceintheproportionvotingyesonProposition8among groupsintheelectorate.Foursuccessivemodelsareestimated,eachincluding additionalvariables.Entriesinthetablethataremarkedwithasterisks(*)identify votercharacteristicsthatweresignificantlyassociatedwiththevoteonProposition8 aftercontrollingfortheothervariablesineachmodel. AsshowninTable2,mostofthedifferencesfoundamongpopulationgroupsinTable1 persistinthemultivariatecontext.Acrossallmodels,men,oldervoters,themore religious,Republicans,andconservativeswereallsignificantlymorelikelytosupport Proposition8thanwomen,youngervoters,thelessreligious,Democratsandliberals.

2PersonalknowledgeoflesbiansandgayswasmeasuredwiththequestionDoyouhaveanyfriends,family members,orpeopleyouknowwellwhoarelesbianorgay,orinasamesexcouple? 3Oneadditionalvariableeducationhasbeenfoundtobestronglyassociatedwithsupportforsamesex marriage(see,e.g.,Egan,Persily&Wallsten2008).Unfortunately,theDBRsurveydidnotincludeaquestion aboutrespondentseducationlevelsandsoweareunabletoanalyzetherelationshipbetweenthisvariableand supportforProposition8.

Table2.MultivariateAnalysisoftheVoteonProposition8

Variable gendercomparisoncategory:female male race/ethnicitycomparisoncategory:white AfricanAmerican Latino Asian mixed/other agecomparisoncategory:age5064 1829 3039 4049 65+ frequencyofattendanceofreligiousservices comparisoncategory:attendmonthly weeklyormoreoften holidaysonly hardlyever personalknowledgeofgaysandlesbians comparisoncategory:havegayfriendsorfamily nogayfriendsorfamily partyidentificationcomparisoncategory:Independent Republican Democratic ideologycomparisoncategory:moderate conservative liberal samplesize %ofvotescorrectlypredictedbymodel

I .12* .08 .10* .00 .10 .08 .10 .04 .21* 1,052 58%

II .14* .02 .07 .00 .12 .06 .11* .02 .21* .26* .01 .13* 1,052 66%

III .13* .01 .07 .02 .12 .06 .11* .02 .20* .26* .00 .12* .08* 1,052 66%

IV .11* .25* .24* .02 .08 .08 .21* .10 .14* .22* .06 .06 .02 .15* .23* .20* .22* 1,052 76%

Cellentriesarefirstdifferencesderivedfromprobitanalyses.Theyareestimatesofthedifferenceintheprobability ofsupportingProposition8betweenvotersinthespecifiedcategoryandthoseinthevariablescomparisoncategory. Coefficientsmarkedwithasterisks(*)indicatecategoriesestimatedtobesignificantlydifferentfromthecomparison categorywithatleast95%confidence.Comparisoncategoriesarethevariablesmodes(inthecaseofgender, race/ethnicity,andknowledgeofgaysandlesbians)ormedians(inthecaseofage,attendanceofreligiousservices, partyidentification,andideology). Source:DBRSurveyofCaliforniaVotersforEqualityCalifornia,November616,2008

Twovariablesareexceptionsinthattheireffectsdonotpersistacrossallofthemodels: (1)raceandethnicity,and(2)personalknowledgeofgaysandlesbians.Theanalysis showsthatAfricanAmericansandLatinoswerestrongersupportersofProposition8 thanothergroups(ModelI),butnottoasignificantdegreeaftercontrollingfor religiosity(ModelsIIandIII).Thatis,muchofthestrongersupportfoundfor Proposition8amongthesegroupsisexplainedbytheirincreasedlevelsofattendance ofreligiousservices.ThedistinctivenessofblacksandLatinosreemergesoncewe controlforpartyidentificationandideology(ModelIV).Personalknowledgeoflesbians andgaymenwassignificantlyassociatedwithoppositiontoProposition8(ModelIII), buttheeffectofsuchknowledgedisappearsoncewecontrolforpartyidentificationand ideology(ModelIV). Partyidentification,politicalviews,religiosity,andagecontributedtothe votemorethanrace,gender,orpersonalknowledgeofgaysandlesbians. InTable3,weassesshowimportanteachvotercharacteristicwasincontributingtothe voteonProposition8.Wedothisbymultiplyingtheestimatedeffectofeach characteristic(theentriesinTable2,ModelIV)bythepercentageofvoterswiththat characteristic.Thismeasurethusaccountsforboththeprevalenceofacharacteristic andthestrengthofitsassociationwiththevote.Forexample,asshownonthetoprow ofTable3,45.9percentofCaliforniavotersaremale.InTable2,ModelIVweseethat menwere11percentagepointsmorelikelytovoteyesonProposition8thanwomen. Thecontributionofthecategoryofthisvariabletotheultimateresultistherefore(.459 x.11=.049),or4.9percentagepoints.Thustheimpactofgenderwastoaffectthevotes of4.9percentofCaliforniansonProposition8.Invariableswithmultiplecategories,we sumupthemagnitudesoftheseeffectsoverthecategoriesofeachvariabletocalculate thepercentageofvotesthatwereaffectedbythevariable. 4

4Thesepercentagesaretheshareofvotespredictedtochange(inonedirectionoranother)ifthevariablewere

tohavenoimpactonhowvotersdecidedonProposition8.Equivalently,thefiguresarealsothepercentageof votesonProposition8thatwouldhaveshiftedinonedirectionoranotherinthehypotheticalcircumstance whereallvoterssharedthecomparisoncategory.Thisanalysisisbasedonameasurecalledlevelimportance describedbyAchen(1982).

Table3.EstimatesoftheImpactofVoterCharacteristics ontheProposition8Vote

direct effectof charact eristic (A) .11 .25 .24 .02 .08 .08 .21 .10 .14 .22 .06 .06 .02 .15 .23 .20 .22 %of %ofvoters votes with affected charact by eristic charact (B) eristic |AxB| 45.9% 7.0% 13.7% 6.2% 3.2% 16.6% 11.8% 18.1% 22.6% 42.8% 13.0% 27.5% 25.8% 33.0% 45.5% 34.0% 35.4% 4.9% 1.8% 3.3% 0.1% 0.3% 1.3% 2.4% 1.8% 3.2% 9.5% 0.7% 1.6% 0.6% 4.8% 10.3% 6.8% 7.9% %of votes affected by variable 4.9% 5.5% 8.7% 11.8% 0.6% 15.2% 14.6%

Variable

gendercomparisoncategory:female male race/ethnicitycomparisoncategory:white AfricanAmerican Latino Asian mixed/other agecomparisoncategory:age5064 1829 3039 4049 65+ frequencyofattendanceofreligiousservices comparisoncategory:attendmonthly weeklyormoreoften holidaysonly hardlyever personalknowledgeofgaysandlesbians comparisoncategory:havegayfriendsorfamily nogayfriendsorfamily partyidentificationcomparisoncategory:Independent Republican Democratic ideologycomparisoncategory:moderate conservative liberal

Source:DBRSurveyofCaliforniaVotersforEqualityCalifornia,November616,2008

Figure1displaystherelativesizeoftheimpactofeachvariableonthevote.The lengthsofthebarsinFigure1correspondtothepercentageofvotesaffectedbyeach variable.Asshowninthisfigure,partyidentificationandideologyhadsubstantial impactsontheultimateresult:thetwovariableseachaffectedanestimated15percent ofthevote,acontributionaboutthreetimesthesizeofraceandgender.Religiosity (whichaffectedthevotesof12percentofCalifornians)andage(9percent)alsohad substantialeffects.Despitetheintenseattentionplacedonraceandethnicityasfactors indeterminingthevoteonProposition8,thisvariableonlyaffectedaboutsixpercentof thetotalvote.

Figure1.TheImpactofVoterCharacteristics ontheProposition8Vote
partyidentification 15.2

ideology

14.6

religiosity

11.8

age

8.7

race/ethnicity

5.5

gender

4.9

knowledgeofgays/lesbians

0.6

5 10 15 percentageofvotesaffectedbycharacteristic
Source:CalculationsinTable3.

AFRICANAMERICANSANDPROPOSITION8

HerewepresentdataindicatingthatwhileAfricanAmericansdidsupportthemeasure athigherratesthanvotersasawhole,wehavestrongreasontothinkthattheirsupport wasnotashighasthatestimatedbytheNationalElectionPool(NEP)exitpoll(70 percent).AnalysisofthefullrangeofdataavailablepersuadesusthattheNEPexitpoll overestimatedAfricanAmericansupportforProposition8bytenpercentagepointsor more.Furthermore,muchofAfricanAmericanssupportforProposition8canbe explainedbythefactthatblackstendtobemorereligiousthanCaliforniansasawhole. SurveysconductedjustbeforeandjustafterElectionDayfoundmuchsmaller differencesinsupportforProposition8betweenAfricanAmericansand votersasawholethandidtheNEPexitpoll.TheNEPresultshouldthusbe treatedasanoutlierthatoverstatesblacksupportforProposition8. AsshowninFigure2,twosurveysconductedjustbeforeElectionDay(byFieldand SurveyUSA)foundinsignificantdifferencesinsupportforProposition8between AfricanAmericansandCaliforniansasawhole.Twosurveysconductedintheweeks followingElectionDayfoundsimilarresults.Onaverage,thedifferenceinsupport betweenAfricanAmericansandallvotersinthesefoursurveyswasjusttwopercentage points.TheNEPexitpollfindingthatblacksupportforProposition8was18points higherthanCaliforniansasawholeismostlikelyanoutlier,aresultthatisvery differentthanwhatconcurrentdatatrendssuggesttobethecase.

Figure2.SupportforProposition8 inPreElection,ExitPoll,andPostElectionSurveys

%supportingProp.8
0 20 40 60 80

Field(10/23)

47 53 48 50 52 70 51 58 49 41

SurveyUSA(10/30)

NEPExitPoll(11/4)

DBRforEqualityCalifornia(11/11)

SurveyUSA(11/19)

AllVoters AfricanAmericans

source:Authorstabulationofpollingdata.Percentagescalculatedincludeonlythoseexpressingapreference.

WeanalyzedprecinctlevelvotingdataonProposition8fromfiveCaliforniacounties Alameda,LosAngeles,Sacramento,SanDiego,andSanFranciscothattogether comprise66percentofthestatesAfricanAmericanpopulation.Bymergingthesedata withestimatesoftheprecinctsracialandethnicmakeup,wewereabletoassessthe precinctlevelrelationshipbetweenvoterdemographicsandsupportforProposition8.

EvidencefromprecinctlevelvotingreturnssuggeststhatAfricanAmerican supportforProposition8wasintherangeof57to59percent.

Figure3.PrecinctVoteforProposition8andAfricanAmericanPopulation inFiveCaliforniaCounties
100

% of precinct voting yes on Prop. 8

75

50

25

0 0 25 50 75 100 % of precinct voters who are African American

Source:PeterFraseandtheCenterforUrbanResearchatTheGraduateCenter,CUNY

Figure3depictsthisrelationshipwithascatterplotinwhicheachprecinctis representedbyapoint.Thefigurealsoincludesalinecalledarunningmean smootherthatindicatesthepatterntakenonbythedata.Asseeninthefigure,aslight butunmistakablerelationshipexistsbetweentheproportionofaprecinctsvoterswho areAfricanAmericanandsupportforProposition8.Also,wenote,thatprecinctswith veryfewblackvoters(shownonthelefthandsideofthefigure)supportedProposition 8atlevelsaboutashighasthoseprecinctswithmanyblackvoters(shownontheright handside).Thatis,supportforProposition8wasgreatestinprecinctsthataretheleast raciallydiverse.

10

Whileitisdifficulttomakepreciseinferencesaboutindividualvotersfromaggregate data,twostatisticalanalysesthatweemployedgeneratedestimatesofAfrican AmericansupportforProposition8of57percentand59percent.Theestimateof57 percentisderivedfromecologicalanalysisconductedusingtheEzIsoftwareprogram (Benoit&King1999).Inthisanalysis,acontrolwasaddedforthepercentageofeach precinctestimatedtobeLatino.Theestimateof59percentisbasedonamuchsimpler approach,knownasGoodmansecologicalregression(Goodman1953).Ratherthan beingtreatedasdefinitive,theseestimatesshouldbeconsideredashelpingto corroboratetheindividuallevelfindingsdiscussedearlierinthissectionofthestudy. 5 Muchofthedifferenceamongracialandethnicgroupsinsupportfor Proposition8isexplainedbyvaryinglevelsofreligiosity.

AsshowninFigure4,AfricanAmericansaremorereligious(asmeasuredbyfrequency ofattendanceatreligiousservices)thananyotherracialorethnicgroupofCalifornia voters.Asawhole,43percentofCaliforniansattendreligiousservicesatleastonceper week.TheshareofAfricanAmericansattendingserviceswiththisfrequencyismuch higher:57percent.ThisdifferenceinfrequencyofattendancebetweenAfrican Americansandtherestofthepopulationisstatisticallysignificant. 6 AsshowninFigure5,controllingforfrequencyofreligiousattendancehelpsexplain whyAfricanAmericanssupportedProposition8athigherlevelsthanthepopulationas awhole.AmongCalifornianswhoattendworshipatleastweekly,supportfor Proposition8wasnearlyuniformacrossallracialandethnicgroups.Amongthosewho attendworshiplessthanweekly,majoritiesofeveryracialandethnicgroupvotedno onProposition8.Thedifferencesthatremainamonggroupsarenotstatistically significantatthe95%levelofconfidence.

5TheprecinctleveldataweredevelopedandanalyzedbyPeterFraseandtheCenterforUrbanResearchatThe

GraduateCenter,CUNY.EstimatesofthepercentofprecinctvoterswhoareAfricanAmericanandLatinowere obtainedfromtheCaliforniaStatewideDatabaseattheInstituteofGovernmentalStudies,UCBerkeley.Wenote thattheecologicalanalysisisaworkinprogress:futureresearchwillincludeadditionalcontrolvariables.We alsonotethatProposition8faredlesswellinthefivecountiesanalyzed(wherethemeasurewasapprovedby 48.3%ofvoters)thaninCaliforniaasawholealthoughthispartiallyreflectsthefactthatAfricanAmericans aremorelikelytoresideinliberalcountiesthanconservativeones. 6Thepvalueassociatedwiththistestislessthan.001.

11

Figure4.FrequencyofAttendanceofReligiousServices byRaceandEthnicity

Asian

40

AfricanAmerican

57

Latino

47

White

42

20 40 %attendingreligiousservicesweekly

60

Source:DBRSurveyofCaliforniaVotersforEqualityCalifornia,November616,2008

Figure5.SupportforProposition8 byFrequencyofAttendanceofReligiousServicesandRaceandEthnicity

White AfricanAmerican worshipatleastweekly Latino Asian 66

70

74 68

White AfricanAmerican worshiplessthanweekly Latino Asian

36 48 46 33

20

40 60 %supportingProp.8

80

Source:DBRSurveyofCaliforniaVotersforEqualityCalifornia,November616,2008

12

THEFUTURE
Since2000,voteinsupportofmarriageequalityhasgrownbynearlyten percentagepointsatrendfoundamongvirtuallyeverydemographicgroup inCalifornia.

Eightyearsago,CaliforniansapprovedProposition22,astatewidebanonthe recognitionofsamesexmarriages,by61to39percent.(Theproposalisalsoknownas theKnightInitiative,afterthelatestatesenatorWilliamPeteKnight,who spearheadedthemeasure).Thetwoinitiativesprovideauniqueopportunitytoassess overtimechangeinCaliforniansvotesonmarriage,whichwedoinTable4.

Table4.ChangeinCaliforniansVotesonMarriageEquality,20002008

Totals Gender Men Women Age 1829 3044 4564 65+ Race/Ethnicity White Black Latino/Hispanic Asian PartyIdentification Democratic Independent Republican Ideology Liberal Moderate Conservative Religion Protestant Catholic Jewish
*

%voting"Yes" onProp.22 (LATimes,2000)* 61

%voting"Yes" onProp.8 (DBR,2008) 52

change, 20002008 9 7 8 13 8 16 1 9 4 6 11 13 5 1 7 6 2 9 4 7

61 57

54 49

58 56 63 68

45 48 47 67

58 62 65 59

49 58 59 48

43 58 80

30 53 81

29 57 84

22 51 82

75 59 24

66 55 17

LATimesexitpolldata.Availableathttp://www.latimes.com/media/acrobat/200810/43119888.pdf

13

Table4comparesdatafromtheLosAngelesTimesexitpollonProposition22in2000 withtheresultsfromtheDBRsurvey. 7 Thetabledisplayssupportforthemarriage bansin2000and2008acrossdemographicgroupsincludinggender,age, race/ethnicity,partyidentification,ideology,andreligiouspreference.Innearlyevery instance,thevotesofCalifornianshaveshiftedinthedirectionofsupportformarriage equality.Menandwomenhavemovedtowardapprovalofsamesexmarriagetoa nearlyequaldegree(by7and8percentagepoints,respectively).Supporthasincreased amongeveryagegroupunderage65,acrossallracialandethnicgroups,andamong Protestants,CatholicsandJewsalike.ThestatesDemocrats,independents,liberalsand moderateshaveallshiftedinthedirectionofmarriageequality,aswell.The comparisonsrevealthreegroupsthatmaybeconsideredtobeholdoutsregardingthe movetowardapprovaloflegalizingsamesexmarriageinCalifornia:Republicans, conservatives,andthoseaged65orover.Votingpatternsamongthesegroupshave remainedstaticoverthepasteightyears,andallremainstaunchlyopposedto legalization. 8

Figure6.VotesonMarriageEqualityinCaliforniain2000and2008, byBirthCohort
75 %votingforsamesexmarriageban

50

Prop.22(2000)

25

Prop.8(2008)

0 1920 1940 1960 yearofbirth 1980

Sources:Proposition22:LATimesExitPoll,February2000 Proposition8:DBRSurveyofCaliforniaVotersforEqualityCalifornia,November616,2008
7Tomakethecomparison,weusedatafromtheDBRsurveyratherthanthe2008NEPexitpollto

maintainconsistencywiththeotherfindingsreportedthroughoutthisstudy.The2008NEP estimatesarebroadlysimilartothoseoftheDBRsurvey(excepting,ofcourse,thediscrepancy discussedearlierinestimatesofAfricanAmericansupportforProposition8).Notallofthevariables intheDBRsurvey(includingattendanceatreligiousservices)wereavailableforcomparisoninthe 2000L.A.Timespoll. 8Forthemostpart,thesefindingsechothoseidentifiedbyLewis&Gossett(2008)intheiranalysisof opinionsurveysofCaliforniansadministeredbetween1985and2006. 14

Figure6displaysanotherwaytoconsiderhowCaliforniansvotesonmarriageequality havechangedovertime.ThefigureplotssupportforProposition22andProposition8 amongthoseinthesamebirthcohortthatis,thoseborninthesametimeperiod.By doingthis,weareabletoassesstheextenttowhichCalifornianshaveactuallyshifted theirvotesovertime. ThetrendsshowninFigure6areilluminating.Thefigureshowsthatvirtuallyno changeinthevotehasoccurredamongthosebornbefore1940(thatis,thosewhoare currently68yearsorolder).ButtheBabyBoomgenerationthosebornbetween 1940and1960hasshiftedsubstantiallytowardvotingformarriageequalityinjust eightyears.Duringthisperiod,thevoteforsamesexmarriagegrewamongboomers by13percentagepoints,andaslimmajorityofthisgenerationvotedagainst Proposition8.Amongthosebornbetween1960and1982,thevoteinfavorof legalizingmarriagehasalsoincreased,from49%in2000to53%in2008. Thefiguresuggeststhattwofactorsagingandgenerationalreplacementmayhasten thearrivalofadaywhenamajorityofCaliforniansvotesinfavorofmarriageequality. AsCaliforniansbornfromtheBabyBoomandafterwardhaveaged,theyhavebecome moresupportiveoflegalizingmarriage,anditispossiblethatthistrendwillpersist amongtheyoungestCaliforniansastheygrowolder.Atthesametime,thenewest votersentertheelectoratewithmuchmoresupportiveviewsonsamesexmarriageand otherrightsforgaypeoplethanthosewhotheyreplace.

CONCLUSION
ThisstudyshowsthatthecharacteristicsthatshapeAmericansviewsonmany importantpoliticalissuesincludingpartyidentification,ideology,andreligiosity playedtheirexpectedstrongrolesindeterminingthechoicesofindividualvoteson Proposition8aswellasthefinalstatewideresult.Thedifferencesseenamongracial andethnicgroupsinsupportforProposition8werealmostcertainlymorenarrowthan indicatedbytheElectionDayexitpollconductedbytheNEP,andwebelievethese differencesdonotmerittheamountofattentiontheyhavereceived.Wehopethatthis studyshineslightonaphenomenonthatwillultimatelybeseenasmoreimportantto thefutureofmarriageequalityinCaliforniaandnationwide:achangeinvoter sentimenttowardsupportforlegalizingsamesexmarriagewhoseswiftpaceisrarely seenonanyissueinAmericanpolitics.

15

REFERENCES
Achen,Christopher.1982InterpretingandUsingRegression.NewburyPark,CA:Sage. Benoit,Kenneth&GaryKing.1999.EzI:AnEasyProgramforEcologicalInference. Availableathttp://gking.harvard.edu/stats.shtml. Egan,PatrickJ.,NathanielPersily&KevinWallsten.2008.GayRights.InPublic OpinionandConstitutionalControversy,eds.NathanielPersily,JackCitrinand PatrickJ.Egan.NewYork:OxfordUniversityPress. Goodman,L.A.1953.Ecologicalregressionandthebehaviorofindividuals.American SociologicalReview18:66364. Lewis,GregoryB.&CharlesW.Gossett.2008.ChangingPublicOpiniononSameSex Marriage:TheCaseofCalifornia.Politics&Policy36:430.

ABOUTTHEAUTHORS
PATRICKJ.EGAN(Ph.D.,UniversityofCalifornia,Berkeley)isassistantprofessorof politicsandpublicpolicyatNewYorkUniversity,wherehespecializesinpublic opinion,politicalinstitutions,andtheirrelationshipinthecontextofAmericanpolitics. HeiscoeditorofthevolumePublicOpinionandConstitutionalControversy,whichwas publishedbyOxfordUniversityPressin2008.EganservedasanAssistantDeputy MayorofPolicyandPlanningfortheCityofPhiladelphiaunderformerMayorEdward Rendell.HewasavisitingscholaratPrincetonUniversitysCenterfortheStudyof DemocraticPoliticsin200607. KENNETHSHERRILL(Ph.D.,UniversityofNorthCarolina,ChapelHill)isprofessorof politicalscienceatHunterCollege,CUNY,wherehehastaughtsince1967.Aspecialistin publicopinion,politicalparticipation,votingandelections,hehasbeenstudyingthe LGBTrightsmovementsince1972.HisbookwithMarcWolinsky,GaysandtheMilitary (PrincetonUniversityPress,1993)wonhonorablementionfortheGustavusMagnus PrizefordistinguishedbookonhumanrightsintheNorthAmericas.In1977,hewas electedDemocraticDistrictLeaderinNewYork's69thAssemblyDistrict,succeeding JerroldNadlerandbecomingNewYork'sfirstopenlygayelectedofficial.Heserveduntil 1985.

16

You might also like