THEMAKINGOFU.S.SUPREMECOURTMAJORITYOPINIONS By RachelA.Schutte ADISSERTATION Submitted toMichiganStateUniversity inpartialoftherequirements forthedegreeof PoliticalScienceŒDoctorofPhilosophy 2015 ABSTRACT THEMAKINGOFU.S.SUPREMECOURTMAJORITYOPINIONS By RachelA.Schutte Upondecidingacase,theSupremeCourtissuesanopinionthatcontainsbothadichoto- mousjudgmentandalengthyrationalejustifyingitsdecision.Theargumentsmadeintheopinion determinethescopeofpublicpolicyandhavebroadimpactsonthepoliticalsystem.Yetpre- ciselyhowthemembersoftheCourtinteracttoproducethecontentofthepolicycontainedwithin themajorityopinionremainsshroudedinmystery.Thisdissertationinvestigatesthecausesand consequencesofbargainingandaccommodationdynamicsinthecreationofU.S.SupremeCourt majorityopinions,arguingthat,thoughtheyareconstrainedbycontextualandinstitutionalfac- tors,certainjusticesarepivotalplayersinthecreationofpubliclaw.Itprovidesanempirical micro-foundationofjudicialbehaviortoinformtheoriesofintra-courtbargaininganditreveals hiddenaspectsoftheSupremeCourt'sopinionwritingprocess.FocusingontheBurgerCourt (1969-1985),theprojectdrawsevidencefromanoriginaldatasetthatcombinesdraftsofmajority opinionswithprivatememorandathejusticesusetocommunicateafteroralargumentsandbefore thepublicreleaseoftheCourt'sopinion. Itbeginsbyintroducingthemotivationforthestudyandbydetailingthecollectionand codingoftheoriginaldata,whicharethenusedtoprovidetheanalysesintherestofthedis- sertation.Theprojectpresentsthreetheoreticallyrelated,butempiricallyindependent,essaysto examinethedynamicopinionwritingprocess.Throughoutthework,noveldatarevealpreviously unknownaspectsofthebehind-the-scenesnegotiationsbetweenSupremeCourtjustices.Content analysisofthejustices'internalbargainingmemorandarevealthetopicsandextentofjustices' engagementwithoneanotherandthestrategiesemployedasbargainingmembersattemptto encethecollegialdevelopmentoftheCourt'sopinion. Buildingonexistingresearch,thisdissertationtheorizesthatjusticeswhositinkeymedian positionsŒoftheentireCourtandofthewinningcoalitionŒaremostlikelytotheshape andscopeofthelegalrulecontainedinthemajorityopinion.Itpresentsadditionalpredictions abouttheconstrainingofinstitutionalandcontextualfeatures.Theevidencesuggests thatwhileopinionauthorsarelargelyamenabletotheircolleagues,contrarytopopulardiscourse, itisnottheCourt'sfiswingjusticeflthatisprincipallybutratherthatthemedianmem- berofthemajoritycoalitionholdsparticularswayinthedevelopmentoftheopinion.Median membersofthewinningcoalitionsaremorelikelytosucceedinpersuadingtheauthortomake changestoopiniondraftsandarelesslikelytofacerequestsfromotherjusticeswhentheywrite themajorityopinion.Conversely,theshowthatwhentheCourt'smedianjusticewrites themajorityopinion,theothermembersofthewinningcoalitionaremorelikelytowriteseparate concurringopinionswithoutpriorengagement.Whilethesepivotalplayersarecertainly tial,theevidenceshowsthatthesuccessofajustice'sattemptstopersuadetheopinionauthorto alterthemajorityopiniondependsmainlyonthecontentofthenegotiationaswellasthetoneor frameoftherequest,andthatpastcooperationhasasizableimpactonajustice'swillingnesstoen- gagewiththeauthorduringthecraftingoftheopinion.Thedissertationfurtherinvestigateswhat happenswhencompromisescannotbestruckbetweentheauthorandabargainingjustice,and evidencethatithasconsequencesfortheloutputoftheCourtbyoccasionallysparking thepublicationofconcurringopinions. ForJonathan iv ACKNOWLEDGEMENTS Thisdissertationwouldnotexistwithoutthemanypeoplewhogenerouslygaveoftheirtime, wisdom,andresourcesthroughoutmyPh.D.education.Ioweadebtofmydeepestgratitudeto eachofthesepeoplewhocontributedtomyintellectualgrowthandthedevelopmentofthisproject inwaysbothlargeandsmall. RyanBlacktaughtaclassandchangedthetrajectoryofmycareer.Itisimpossibleto capturehowfortunateIamtohavefoundanincrediblementorwho,soearlyinhisowncareer, investedsuchenergyintoguidingthewayIthinkaboutresearch.Everystudentshouldbeso lucky.Ryan'sintellectualcuriosityandpassionforrigorous,thoughtfulscholarshipremainsan inspiration.Isimplycannotthankhimenoughforwhathehasdoneformeandformyresearch.It wasinhisclassroominthefallof2010thatIdiscoveredapassionforjudicialpoliticsandtheseed forthisdissertationwasplanted.Withhisencouragement,IattendedtheEmpiricalImplications ofTheoreticalModelsInstitutehostedbytheHarrisSchoolattheUniversityofChicagoin2011. Itwastherethatthisideabegantotakeroot,andIthanktheincredibleteachersandseminar participantswhochallengedmythinkingandgavemeimportantinsightsintothefoundationsof thisproject. TheDepartmentofPoliticalScienceatMichiganStateUniversityprovidedastimulating environmentandIamgratefulfortheintellectualandsupportthatmadethisprojectpos- sible.ChuckOstromhashadatremendousimpactonhowIapproachresearchdesignandanswer theoreticalquestions,andIthankhimforhisconsistentencouragementthroughout.MelindaGann Hall'sexperiencewithdatacollectionprovedtosetrealisticexpectationsandI'mindebtedtoher advice.EarlydiscussionswithReggieSheehanhelpedfocustheprojectandclarifymyquestions. Itwouldnothavebeenfeasibletoundertakethisinvestigationwithoutsubstantial supportandmuchofthematerialpresentedinthisdissertationisbasedonworksupportedbythe NationalScienceFoundationunderGrantNo.SES-1228458.Iamthankfulforthesupplemental v supportIreceivedfromaCoreyResearchEnrichmentFundAwardandtheMichigan StateUniversityDepartmentofPoliticalScienceMeritFellowship.Igratefullyacknowledgethe assistanceofAndrewGoldstein,KaitlinGriswold,SarahPrather,ElleRoss,andAdriannaSi- monelliforthecollectionoftheoriginaldatafromtheprivatepapersofformerSupremeCourt justiceshousedintheLibraryofCongressinWashington,DC.Ialsothankthereferencelibrarians intheManuscriptReadingRoomattheLibraryofCongressfortheirsteadfasteffortsandassis- tance.IfurtherthankKaitlin,Elle,andAdriannaforstayingonasresearchassistantstocontent analyzeandcodeallofthebargainingmemoranda,matchingthesedocumentswithchangesinma- jorityopiniondrafts.Likewise,thisresearchwouldnotbepossiblewithoutthepubliclyavailable SupremeCourtOpinionWritingDatabaseprovidedbyPaulWahlbeck,JimSpriggs,andForrest Maltzman,whoalsograciouslysharedtheirreplicationdata.IamparticularlygratefultoJim Spriggswhosharedthereplicationtoaidinthestatisticalanalysespresentedinthiswork. Iamfortunatetohaveamassedaverysmartgroupofcolleaguesandfriends,manyof whomhavehadahandinmyworkwhethertheywereawareofitornot.IanCook offeredinsightfulcommentsatvariousstagesandprovidedagoodsoundingboardformanyof myhalf-bakedideasandMelanieBowersofferedsupportinamyriadofwayswhenIneededit most.IamparticularlyindebtedtoDominikDuell,whoprovidedhelpwithstatisticalmodelsand graphicsatclutchmoments.Dominiktoleratedmymany(many,many)questionsandconsistently walkedthroughvariousargumentstoidentifylogicalws.Throughoutitsnumerousiterations, hereadpiecesofthisprojectanditalwayscameoutstrongerafterhiscarefulcriticism.Dominik's encouragementnottolosetheforestforthetreeshelpedkeepmeontrack,andIameternally gratefulforhissteadfastbeliefinthiswork. Beyondmyadvisorsandcolleagues,Ihavefromtheguidanceandfeedbackfrom anumberofpeopleatprofessionalconferences.EarlyversionsofChapters2and4werepresented attheannualmeetingsofMidwestPoliticalScienceAssociation(2013,2014),theAmericanPo- liticalScienceAssociation(2013),andtheWesternPoliticalScienceAssociation(2015)andIam especiallythankfultothediscussantsŒSaraBenesh,CliffCarrubba,ChuckCameron,andSteve vi KautzŒwhoofferedconstructivecriticismandtotheaudiencememberswhoaskedquestionsthat frequentlymyownthinkingabouttheresults. Whiletheseintellectualandsupportswereessentialtothecompletionofthis dissertation,Iowemysincerestthankstothenetworkoffamilyandfriendswhohavekeptme groundedandremindedmeofwhatistrulyimportantinlife.MygenerousfriendsŒKelli,Lindsey, andCarleen,amongsomanyotherintelligentandstrongwomenŒletmefitalkoutloudflaboutmy workandbroughtwine,laughter,andwisdomtoenrichmyworldallthewhile.Whilemydadwas notalivetoseemegrowintoadulthood,hisvoicecontinuestoremindmethatfican'tneverwill,fl whichhasencouragedmeinanindescribablewaythroughoutthisprocess.Mymomwalkedmiles andmileswithmearoundalakeandthroughthewoodsandinNorthernMichigan,asking probingquestionsaboutmywork,empathizingwitheveryfailure,andcelebratingeachsuccess. Iwouldhavebeenlostwithoutherconstantsupport.Mylittlebrother,Jonathan,showedupin Washington,DCtohelpwithdatacollectionwhenIranoutofmoneyandhasofferedcountless late-nightcounselingsessionsasIdoubtedmyabilitytothisthing.Myolderbrother,Kevin, andhishilariouschildrengavemewelcomedistractionsandbroughtmebacktomyroots. Myhusband,Lucas,andmysweetdog,Jake,punctuatedmyPh.D.educationandthisdis- sertationworkwithincredibleadventures.IappreciateLuke'sboundlesspatienceasIhibernated inmyofthinking,coding,analyzingandwriting.Hiswordsofafandnumerous actsofkindnessthroughoutthisprocesssimultaneouslykeptmebuoyedandanchored.Though hedidn'tmakeittotheend,ofallthosewhosawmethroughthis,Jakewasmymostconstant companion,layingatmyfeetandalwaysreadytorunandplaywhenIneededabreak.Notonly wouldthisprojectnothavecometofruitionwithoutthisnetwork,butmylifewouldhavebeenfar lessmagical. vii TABLEOFCONTENTS LISTOFTABLES ....................................... x LISTOFFIGURES ....................................... xi CHAPTER1INTRODUCTION ............................... 1 1.1BehindtheScenes:BargainingoverOpinionContent................1 1.2TheOpinionWritingProcess.............................4 1.3WhyMajorityOpinionContentMatters.......................6 1.4DataCollectionandCodingProcedure........................8 1.4.1DataCollection................................8 1.4.2CodingProcedure...............................9 1.4.2.1BargainingStatementTypes....................10 1.4.2.2BargainingStrategy........................14 1.4.2.3IdentifyingAccommodation....................15 1.5SummaryofChapters.................................16 REFERENCES.......................................20 CHAPTER2COMPROMISEANDCONTROLINTHECOLLEGIALDEVELOP- MENTOFU.S.SUPREMECOURTMAJORITYOPINIONS ........ 23 2.1ChapterAbstract...................................23 2.2Introduction......................................23 2.3TheOpinionWritingProcess.............................27 2.4TheoryandHypotheses................................28 2.4.1JusticePowerandOpinionControl......................29 2.4.2ContentofRequestsandtheLikelihoodofAccommodation........31 2.4.3MeansofPersuasion.............................32 2.4.4AdditionalCollegialandContextualConstraints...............33 2.5DataandMeasurement................................36 2.5.1QuantifyingConnecting Who Matterswith How TheyMatter..37 2.5.1.1WhatJusticesWant:TheContentofBargainingMemoranda...39 2.5.1.2HowJustices(attemptto)PersuadetheOpinionAuthortoChange43 2.5.1.3WhenOpinionAuthorsAccommodate:ADirectMeasure....47 2.5.2MeasurementofExplanatoryVariables....................51 2.6Results........................................53 2.7DiscussionandConclusion..............................59 REFERENCES.......................................63 CHAPTER3(DIS)ENGAGEMENTINTHESUPREMECOURT'SOPINIONWRIT- INGPROCESS ................................. 67 3.1ChapterAbstract...................................67 viii 3.2Introduction......................................68 3.3SelectionEffects...................................70 3.4DeterminantsofthePropensitytoBargain......................72 3.4.1PositionalBargainingPower.........................72 3.4.2Context,Politics,andCollegiality......................75 3.5Results:TheDecisiontoBargain..........................78 3.5.1DataandMeasurement............................78 3.5.2ModelandResults..............................82 3.6DiscussionandConclusion..............................86 REFERENCES.......................................91 CHAPTER4CONCURRINGOPINIONSINAFRACTUREDMAJORITY ....... 94 4.1ChapterAbstract...................................94 4.2Introduction......................................95 4.3ResponsestotheMajorityOpinionAuthor'sRefusaltoAccommodate.......99 4.4SeparateOpinionsandTypesofConcurrences....................102 4.5ConditionsthatIncreasedLikelihoodofConcurringOpinions...........105 4.5.1WhenConcurrencesEmerge:TypesofRequests...............105 4.5.2WhenConcurrencesEmerge:StrategyofNegotiation............108 4.6DiscussionandConclusion..............................110 REFERENCES.......................................113 CHAPTER5CONCLUSION ................................ 115 5.1TheSecretiveCreationofU.S.SupremeCourtMajorityOpinions.........115 5.2DiscerningJudicial.............................117 5.3MajorFindings....................................118 5.4Limitations......................................120 5.5AvenuesforFutureResearch.............................122 REFERENCES.......................................125 ix LISTOFTABLES Table2.1Aggregate-levelrequestcontent..........................40 Table2.2Logitmodelofmajorityopinionauthor'sdecisiontoaccommodate.......55 Table3.1Multinomiallogitmodelofmajoritycoalitionmember'sdecisiontobargain...84 x LISTOFFIGURES Figure1.1Absolutefrequencyofeachbargainingrequestbytype..............13 Figure1.2Absolutefrequencyofeachbargainingstrategybytype..............15 Figure2.1Barplotsofbargainingfrequencyandtypebyjustice..............42 Figure2.2Stackedbarplotsofrequestcontentbytone....................45 Figure2.3Barplotoftherelativefrequencyofthetonebargainingjusticesemploy.....46 Figure2.4Barplotsofthenumberofrequestsbyopinionauthorsandjustices' accommodationrates...............................49 Figure2.5Accommodateratebyauthor-bargainingjusticedyad..............50 Figure2.6Averagemarginaleffectsforstatisticallyvariablesontheproba- bilityofaccommodation..............................56 Figure2.7Predictedprobabilityanauthoraccommodateswhenthebargaineristhe MoMCorwhentheauthoristheCourtmedian..................57 Figure2.8Predictedprobabilityofaccommodationcontingentonthedegreeofamicus participation....................................58 Figure3.1Averagemarginaleffectsforstatisticallyvariablesontheproba- bilityofjoiningthemajorityopinion.......................85 Figure3.2Averagemarginaleffectsforallstatisticallyvariablesontheprob- abilityofwritingaconcurrencewithoutbargaining.............86 Figure4.1Barplotoftherelativefrequencyoftheconcurrencebytype...........105 Figure4.2Barplotsofproportionofconcurrencesresultingfromdifferenttypesofbar- gainingrequests..................................107 Figure4.3Nestedbarplotofdistributionoffailedaccommodationbyjusticeandresul- tantconcurringopinionaction...........................109 xi CHAPTER1 INTRODUCTION 1.1BehindtheScenes:BargainingoverOpinionContent In1968,CongresspassedalandmarkpieceoflegislationknownastheFairHousingAct.Acentral partofthispolicyoutlawedthediscriminatorypracticeofracialfisteering,flatacticusedbyreal estatebrokerstomanipulatethehousingmarketbyguidingprospectivehomebuyersexclusively tohousesinneighborhoodsalreadypopulatedbythebuyer'sracialorethnicgroup.Sevenyears afteritspassage,agroupofresidentsintheChicagosuburbofBellwoodbelievedthattwoChicago realestatebrokeragecontinuedthenow-illegalpracticeofsteeringpotentialhomebuyersto certainpartsoftheirvillageonthebasisofrace.Thissuspicionledagroupofsixraciallydiverse individuals,fourfromthesuspectedfitargetflareaandtworacialminoritymembersfromadjacent neighborhoodsbothwithinandoutsidetheVillagelimits,toposeasinterestedclients,engagingthe servicesofthetwobrokerageWhiletheinquiriesweremerelyarouse,thesixindividuals allegedthatthedid,infact,steerthemtodifferentareasofBellwoodonthebasisofrace. Citingeconomicandsocialinjury,thegroupadiscriminationsuitintheDistrictCourtfor NorthernIllinois.TheDistrictCourtdeterminedthat,sincetheindividualsweremerelytestersand notactualhomebuyers,anydiscriminationinjurywasindirectatbestand,adoptinglegalreasoning fromasimilarcaseintheNinthCircuitCourt,heldthattheindividualslackedstandingtosue.The CourtofAppealsfortheSeventhCircuitdisagreedwiththelegalreasoningofferedbytheNinth Circuitanddeclaredthatevenindirectvictimsofhousingdiscriminationwereentitledtosueunder provisionsoftheAct. ThecaselandedontheSupremeCourtdocketin1978,whereaconferencevoteestablished a7-to-2majority,afthelowercourt'sholdingwithregardstoinjuryandbroadlyinterpret- ingthelegislativeintentoftheFairHousingActtodeterminethattheindividuals within thealleged 1 targetareadidindeedhavestandingtosuitinfederalcourt.Asamemberoftheinitialconfer- encemajority,ChiefJusticeBurgerassignedtheCourt'sopiniontoJusticeLewisPowell,who,in turn,circulatedthedraftoftheopinioninFebruary1979.ThemedianmemberoftheCourt, JusticeHarryBlackmun,readilyagreedtoPowell'sproposedmajorityopinion,castingavoteto jointheverytdraftwithoutdelay.ThoughtheiracquiescencecamemoreslowlythanBlack- mun's,twoothermembersofthemajoritycoalition,theChiefandJusticeThurgoodMarshall,also ultimatelyjoinedPowell'sopinionwithoutanactiveattempttoswaythecontentcontainedtherein. However,thethreeremainingmembersoftheCourt'smajoritycoalitionactivelyengaged withPowell,requestingsubstantivechangesbemadetotheopinion.Inparticular,these membersoftheCourt'smajorityquibbledabouttheinclusionofFootnote24(Footnote25inthe publiclyreleasedopinion),whichdistinguishedbetweenthesixindividualrespondentsinthecase, separatingthefourtargetarearesidentsfromthetwonon-residents.Intheinitialopinion,Powell soughttoaftheSeventhCircuit'srulinginfavorofthefourBellwoodresidents,buttoreverse thestandingdecisionforthetwonon-residentssincenoharmorinjurywasallegedoutsideof thetargetneighborhood.ThechallengetoPowell'slegalreasoningcamefromJusticeByron White,themedianmemberoftheCourt'smajoritycoalitioninthisparticularcase,justfourdays aftertheinitialmajoritydraftcirculation.Inhisresponsememo,Whiteindicatedthathewasfiquite with[the]opinioninthiscasewiththeexceptionofnote24.flHeexpressedhisbeliefthat fitheconsiderationsgivingArticleIIIstandingtothevillagealsogivesufsubstancetothe allegationsofinjuryflfromnon-residents.ItwasWhite'sviewthatPowell'sopinionwas toorestrictive:justbecausethenon-residentmembersofthegrouphadnotallegedsocial oreconomicharmdidnotmeanthatthey couldn't doso.Inamemodatedsixdayslater,Justice WilliamBrennanpushedthispointfurther,suggestingthatsincethesuitallegedharmonlyin thetargetarea,thestandingissueforthoseresidingoutsideofthatneighborhoodwas notaquestionbeforetheCourtandrequestedthatPowellmakeitclearthattheCourt'sopinion didnotestablishwhethersuchpersonshavestandingtosueundertherelevantprovisionofthe FairHousingAct.Finally,onthesamedaythatBrennancirculatedhisrequest,JusticeJohnPaul 2 StevenslobbiedPowelltofisimplyomitn.24.fl Throughouttheopinionwritingprocess,JusticePowellsoughttoincorporatethesubstan- tiverequestsmadebybothJusticesWhiteandBrennan,butheelectednottoadoptStevens'most preferredoutcome,refusingtodeletethefootnoteentirely.Intheopinion,Powellmaintained adistinctionbetweentherespondents,butadoptedWhite'srequestanddidnotcreateprecedent toalegalruletodistinguishamongtherespondents.Instead,hereachedacompromise withBrennanandtookhissuggestiontoremandthequestiontothelowercourt.ThoughPow- ellretainedcontroloftheopinion,throughaninteractivenegotiation,thecontentofthemajority decisiontheviewsofothermembersofthemajoritycoalition. Thisshortcasestudy,whichalsoprovidesthemotivationfortheempiricalinvesti- gationinthenextchapter,highlightsheretoforeunanswerablequestionsabouttheopinionwriting process.Inparticular,Powelladoptedtwoofthethreesubstantiverequestsmadebyhiscolleagues, indicatingthatmembersofthemajoritycoalitionhavethepowertopersuadetheassignedopinion authortoalterthecontentoftheopinion.Moreover,asthisanecdotereveals,preciselywhose viewsareincorporatedintotheCourt'sopinionsmaybesurprising.Inthiscase,whilePowell hastilyadoptedthesuggestionprofferedbythemajoritycoalitionmedianmember(White),the slightlyright-of-centerJusticePowellalsoactivelyengagedinthenegotiationprocesswithanide- ologicallyextremeliberalmember(Brennan),whilevirtuallyignoringtheviewsofamoreproxi- mateideologicalally(Stevens).BecausetheCourt'sopinionsdeterminethescopeofpublicpolicy andthushavefarreachingimplications,whethercertainmembersoftheCourtaremore thanothersindetermininglegalpolicyhaslongbeenatopicofinterestforjudicialscholarsand legalexpertsalike.Consequently,avastliteraturehasgrowntoexaminethefactorsthat theCourt'sopinion(see,e.g.,EpsteinandKnight1998;Maltzman,SpriggsandWahlbeck2000). YetpreciselyhowthemembersoftheCourtinteracttoproducethe content ofthepolicycontained withinthemajorityopinionremainsshroudedinmystery. Asthestrategicmodelofjudicialdecision-makinghasgainedtractioninrecentyears,there hasbeenaclearpushtomovebeyondthesimpledispositionofthecase(i.e.,thedecisiontoaf 3 orreversethelowercourt'sdecision)totryunderstandthecontentoftheopinions.Despitethis evidentshiftinemphasis,todate,extantscholarshiplargelyfailstoactuallyexaminethecontent ofbargainingrequestsorthedegreesofaccommodation.Ultimately,byignoringcontent,Iargue thatimpreciseproxymeasuresforthelikelihoodofaccommodationmayservetounderminewhat wethinkweknowabouttheopinionwritingprocess.Byutilizingoriginaldatatoexaminethe iterativechangesinmajorityopiniondrafts,thisdissertationbeginstoopentheblackboxand lookintotheopinionwritingprocess,uncoveringifandhowcertainmembersoftheCourtshape judicialpolicy. 1 Combiningtheseoriginalopiniondatawithexistingdatafromthetreasuretrove thatistheSupremeCourtOpinionWritingDatabase(Wahlbeck,SpriggsandMaltzman2009),itis possibletoexaminewhoseexpressedpreferencesareincorporatedintothebodyoftheopinionand whosedemandslanguishunanswered.Throughcontentanalysisofthebargainingmemosdirected totheopinionauthor,thisanalysisshedslightonthedistinctivetypesofchangessought.Finally,I seektounderstandtheconsequencesofanopinionauthor'schoice not toaccommodatehisorher colleagues.,Iseektoanswerseveralfundamentalquestions:(1)Whichmember(s) ofthecourtthescopeandshapeofthemajorityopinion?(2)Whatdojusticesbargain aboutandwhattypesofrequestsareaccommodatedbytheopinionauthor?(3)When,orunder whatconditions,doessuchaccommodationoccur?(4)Whatmotivatesajusticetoseekchanges fromtheauthorintheplace?And,,(5)Whenanauthoroptsnottoaccommodateher colleague,howdoesthisbreakdowninthenegotiationprocessaffecttheCourt'soutput? 1.2TheOpinionWritingProcess Followingoralargument,oneoftheonlypubliccomponentsofacase,SupremeCourtjustices meetinprivateconferencetoestablishthedispositionalmajoritycoalition.Aftertheyhavepro- 1 ClarkandLauderdale(2010)introduceameasurethatallowsscholarstoestimatetheideolog- icallocationofopinions,providingdeeperinsightintoitscontent.UnlikeClarkandLauderdale,I amconcernednotwiththe ideologicallocationofthemajorityopinion,butIseektounder- standhowinstitutionalandcollegialfactorspushorpulltheopinion throughout thenegotiations betweenthejustices. 4 videdindividualassessmentsofthearguments,thejusticesvoteindescendingorderofseniority todeterminethemajorityviewoftheCourt.Thesenon-bindingconferencevotesareusefulto establishtheexpectedwinninganddissentingcoalitions,butpreviousresearchrevealsthatthere isoccasionallymovementfromthevotetothepublicreleaseoftheCourt'sdecision(e.g., Brenner1995;MaltzmanandWahlbeck1996). Theseniormemberofeachtentativecoalitionassignsoneofthejusticestowriteanopinion forthatalliance.RegardlessofhistenureontheCourt,thechiefjusticeisalwaysconsideredtobe themostsenior(Schwartz1993)andisresponsiblefortheopinionassignmentwhenhevoteswith themajority.Ifthechiefhimselfinthedissentingcoalition,thenextmostseniorassociate justiceŒbylengthofserviceŒisresponsibleformakingtheopinionassignment(Segal andSpaeth1993).Presumably,itisthejoboftheassignedmajorityopinionauthortosummarize theviewsofthewinningcoalitionbydescribingthelegalrationalebehindthedispositionalvote. Althoughthemajorityopinionauthorhastheprimaryresponsibilitytodrafttheopinion, inordertosetnationalpolicyandtobindlowercourtstoitsdecision,theauthormustnecessarily obtainthesignaturesofatleastfourothermembersofthebench.Becausenon-authoringjustices mustchoosetoendorsetheopinion(ornot),thereisaclearavenuefortheseactorstotryto thescopeanddirectionofthemajorityopinion.Asthemotivatingexampleattheoutset illustrates,themajorityopinionoftheSupremeCourtmayoftenbeanamalgamationofnumerous justices'viewsorvoices. Aftertheinitialopinionassignmentismadeduringconference,nothinghappensinacase untilthemajorityopinionauthorcirculatesadraftofthedecision.Oncethatinitialdraftis circulated,however,thejusticesmaydistributeseparateopinionsortoengageindirectlywith theauthortonegotiateoverthecontentcontainedtherein.JusticeWilliamBrennandescribedthe process,writingfiBeforeeveryonehasmadeuphismind[thereis]aconstantinterchange amongus...whilewehammerouttheformoftheopinionfl(Brennan1960,405,asquotedin Maltzman,Spriggs,andWahlbeck2000,9-10).Itistheprimaryaimofthisdissertationtounder- standhowthisbehind-the-scenesengagementmoldsthecontentoftheCourt'smajorityopinion 5 andhowitcontributestothecreationofseparateconcurringopinions. 1.3WhyMajorityOpinionContentMatters Theargumentsmadeinthemajorityopinionaresteepedinthenuanceoflawandhavethepower toaffectdecision-makinginlowercourts,toalterthereachofpublicpolicy,andtothe privatebehaviorofindividualcitizens.LawyerspouroverCourtopinionstoadviseclients,legal scholarsdissectitslanguagetounderstandthelaw,andpoliticiansemployCourtdecisionstoshape theirownagendas.AmericancitizensgrapplewiththeconsequencesofSupremeCourtopinions aspoliciesshiftandchangeunderthejustices'interpretationofwhatthelawis.Thearguments, rationales,andwordscontainedintheCourt'sopinionsarethusofcriticalimportance.Butwhere dotheycomefrom?Howisthelawmadeasitpassesthroughthechambersofnineunelected justices?Howdothejusticeswork(orfailtowork)togethertosettleimportantlegalquestions? Adherentstothetextbookpoliticalscienceapproachtojudicialdecision-makingfocuson thedispositionalvotesofthemembersoftheCourttosuggestthatbehaviorisasincere ofindividualpolicypreferences(SegalandSpaeth1993).Atthemostbasiclevel,thisapproach emphasizesthatjusticesareunhinderedbyelectoralpressureandarerelativelyunconstrainedby theotherbranchesofgovernmentandarethusfreefitobasetheirdecisions solely uponpersonal policypreferencesfl(RohdeandSpaeth1976,72,emphasisadded).Itisnowcommonplaceto suggestthatjusticesaresimplyideologicalactorsandtoascribetheirbehavioraccordingly. Despitewidespreadsupportfortheattitudinalmodel(see,e.g.,RohdeandSpaeth1976; SegalandCover1989;SegalandSpaeth1993,2002),lawscholarshavelonglamentedthefact thattheseearlyempiricalstudiesglazeovertheactualcontentoftheCourt'sopinions.Contrary toattitudinalmodels,thelegalapproachhaslongtheorizedthatcaseoutcomesarearesultofcase factsandestablishedlawandprecedent(Levi1949;Mendelson1963).Althoughthelegaland attitudinalapproachesappeartobecompletelydivergent,eventhestaunchestsupportersofthe attitudinalmodelconcedethatthelanguageofopinionsmatters.Take,forexample,Rohdeand 6 Spaeth(1976)whoacknowledgethattheCourt'spowercomesfromitsopinion,writing,fiItis themajorityopinionwhichlaysdownthebroadconstitutionalandlegalprinciplesthatgovernthe decisioninthecasebeforetheCourt,whicharetheoreticallybindingonlowercourtsinallsimilar cases,andwhichestablishprecedentsforfuturedecisionsoftheCourt.fl SinceCourtopinionsareespeciallyimportantintheAmericanpoliticalsystem,withinthe lasttwodecades,judicialscholarshaveshiftedfocustogobeyondthestudyofdispositionalout- comes.ThestrategicmodeladvancedbyEpsteinandKnight(1998)andMaltzman,Spriggsand Wahlbeck(2000),amongothers,arguesthatwhilejusticesaredrivenprimarilybytheirpolicy preferences,legalfactors(casefactsandprecedent),institutionalrulesandnorms,andcollegial characteristicsaffectdecision-makingonthehighestcourtintheland.Diggingbeyondthevotes, studiesinthisveinsuggestthatthemembersoftheCourtarenotasunconstrainedasweoncebe- lieved(e.g.,Clark2011).Thesurgeinscholarshipthatseekstounderstandhowopinionsaremade havebeguntorevealthebargainingprocess(Wahlbeck,SpriggsandMaltzman1998;Spriggs, MaltzmanandWahlbeck1999;Maltzman,SpriggsandWahlbeck2000),thesourcesofopinion content(Corley,CollinsandCalvin2011),andtheideologicallocationofmajorityopinion (ClarkandLauderdale2010;Carrubbaetal.2012).Whilethesearenotableadvances,theultimate goalhasnotyetbeenachieved.Thatis,scholarshavenotyetbeenabletofiexplain[]theactual contentofCourtopinionsfl(Maltzman,SpriggsandWahlbeck2000,154). Themostlimitationtoourunderstandingofthemechanismswhichproducethe opinionoftheCourthasbeentheunavailabilityofameasureofaccommodation.Withoutaccess tothebehind-the-scenesnegotiationsthattakeplaceafteroralargument,ithasnotbeenpossibleto ascertaineitherthecontentoftheintra-Courtnegotiationortodeterminewhichrequestsforchange theopinionauthorelectstoincorporateintotheCourt'smajorityopinion.Inthisdissertation project,Iredressthisdataythroughthecollectionoforiginaldatathatexplicitlyreveal thefullnegotiationprocessleadingtothepublicationofthemajorityopinion. Inthefollowingsection,Iprovideadescriptionofthedatacollectionandcodingprocedure. Thesectionprovidesasummaryoftheindividualchaptersofthedissertation. 7 1.4DataCollectionandCodingProcedure Thisdissertationservestwopurposes:Itilluminatesthebehind-the-scenesdevelopmentof Courtopinionsthroughtheuseofnoveldata.Second,thedatageneratedforthisprojectwill haveimplicationsforfuturescholarshipfarbeyondthescopeoftheimmediateproject.Thesenew datawillbemadepubliclyavailabletoprovidecompletedocumentationforallaspectsofopinion developmentontheU.S.SupremeCourtfrom1969-1985. 1.4.1DataCollection Themostcomprehensiveworkontheopinionwritingprocessattemptstoaccountforcollegialand institutionalfactorsusingproxymeasuresforanauthor'saccommodationdecision.Forexample, Wahlbeck,SpriggsandMaltzman(1998)suggestthatthesheervolumeofdraftscirculatedby theassignedopinionwriterisevidenceofattemptstoincorporatecolleagues'preferences.In laterwork,Maltzman,SpriggsandWahlbeck(2000)employeventhistoryanalysis,usingtime asaproxyforaccommodation.Theseapproachesareplausibleapproximationsofthebargaining process,buttheyfailtoaccountforwhetheractualchangesweremadetothemajorityopinion. Sincetheoutcomeofthissecretiveexchangedetermineslawandlegalpolicy,itisvitalthatwe understandhowthecollaborativeeffortofthejusticestheshapeoftheCourt'smajority opinion. Priorscholarshipimplementedcoarsemeasuresofaccommodationquitesimplybecause thedatatoaccountforchangestoopinioncontentdidnotexist.Toovercomethedatay thatlimitedpastanalyses,IspentthreemonthsinthearchivesattheLibraryofCongressinthe fallof2012,collectingoriginaldatathatexplicitlyrevealthefullnegotiationprocessleadingto thepublicationofthemajorityopinion.Workingalongsideveundergraduateresearchassistants, Iphotographedeverysinglepageofeverysinglemajorityopiniondraftfor17terms.Wedoc- umentednearly150,000pagesofarchivalmaterial,which,together,allowmetopeelbackthe curtainofsecrecytodetermineifandhowthemajorityopinionchangeswhenjusticesengagewith 8 theopinionauthor. IcombinethesenoveldatawiththerawarchivalimageshousedintheSupremeCourt OpinionWritingDatabase,whichprovidesdocumentationofthebargainingmemorandathatthe justicessenttotheopinionauthorforallcasesfromthe1969to1985term(Wahlbeck,Spriggsand Maltzman2009).Mydatasetiscomprisedof6,542majorityopiniondraftscirculatedacross2,305 cases.Activebargainingorengagementisobservableinapproximately50percentofthecases. Amongthesecases,thedatarevealnearly5,000uniquerequestsaddressedtotheopinionauthor. Ianalyzedandcodedtheidentityoftherespondingjusticeandthecontentofeachrequest bythemajorityopinionauthor. 2 MyresearchassistantsandIinvested450hoursinthearchivaldatacollectionprocessand havespentmorethan3,000person-hourstranscribingandcodingthesedata.Thedatacollection andcodingprocesseswerefundedprimarilybytheaNationalScienceFoundationDissertationIm- provementGrant(No.SES-1228458)withadditionalsupportfromaCoreyResearchEnrichment FundAwardandtheMichiganStateUniversityDepartmentofPoliticalScienceMeritFellowship. 1.4.2CodingProcedure Asakeytounderstandingthebargainingdynamicsanditsonthelanguagecontained withintheSupremeCourt'smajorityopinion,Ineededtoascertainthecontentofeveryrequest justicessolicitedthroughengagementwiththeassignedopinionauthor.Tothatend,Ireadevery singlebargainingmemorandumandcodedforthreedistincttypesofrequests:minorrevisions, legalchanges,andpolicychanges.Themajorityoftheserequests(45percent)focusonsubstantive challengestotheapplicationandtreatmentofprecedentintheopinion.Asecondtypeofrequest, comprising10percentofallbargainingstatements,asktheopinionauthortosomehowalterthe 2 TheempiricalchapterstofollowexcludedatafromOctoberTerms1969and1970sincethe Courtaltereditsinternaldocketingmechanism,raisingtheconcernthatmemorandafromthese earlycasesaremissing.Therearenootherreasonstosuspectthatthesetermsaresubstantially differentfromtherestoftheyearsincludedintheanalysis. 9 opinion'spolicycontent.Theremainingrequestsasktheauthortomakenon-substantive(i.e., wordchoice,grammatical,ortotheCourt'smajorityopinion. Afterestablishingthesethreemajorcategoriesbasedoninsightsfromextantscholarship, andinspiredbyBryanandJohnson(2012),Iusedanorganicapproach,drawnfromarandom sampleof250cases,todevelopacomprehensiveŒyetrelativelyparsimoniousŒsubcategorization ofeachtypeofrequest.Thatistosay,Idevelopedanewsubcategoryforeachnewtypeofrequest encounteredinthedata.Thisensuredboththatmycodingstrategywasnotonlytheoretically derived,butalsosufentlydetailedtoovercomechallengesofinter-codersubjectivity.Requests forminorrevisionsasktheopinionauthortomakeslight,non-substantivechanges(i.e.,word choiceorgrammaticalalterations)totheopinion.Bargainingstatementsthataskforlegalor policychange,however,asktheauthortosubstantivelyaltertheopinioncontent.Requestsfor legalchangefocusontheapplicationandtreatmentofprecedent,whilepolicychangeschallenge thescopeoflawsettledintheCourt'sopinion. Thememosthatthejusticescirculatetotheopinionauthorcontaindifferenttypesofinfor- mationorrequests.Theaimhereistodistinguishbetweenthetypesofrequestscontainedinwhat Icallfibargainingstatementsflorfibargainingmemos.flTogetherwiththreeresearchassistants,I codedeachbargainingstatementatthe request level,ratherthanatthedocument/memolevel.This meansthateveryrequestservesasasingleobservation(i.e.,asinglememothatseeksvedistinct changestothemajorityopinionwouldcountasveobservations). Foreachrequest,then,wecodedthebargainingcontentusingtheschemepresentedbelow. Thepercentofeachbargainingtypeisincludedinparenthesesbeforethedescriptionofthecontent. 1.4.2.1BargainingStatementTypes MinorRevisions 1 Œ(4.83%)Writing(i.e.,sentenceconstructionorwordchoice)orgrammaticalsuggestion 2 Œ(15.79%)Insertordeletenon-controversialtext;theserequestsaremoresubstantialthan 10 theminorrevisionsoftheprecedingcategory,butstillshouldnotsubstantivelychangethe legalorpolicyimplicationsofthedecision 3 Œ(12.22%)Clarifysectionofopinion 4 Œ(13.76%)Altertheopinionsoittheparticipation,orlackthereof,ofthebargain- ingjustice.ThesetypicallytaketheformoffiPleaseshowmeashavingtakennopartinthe decisioninthesecases.flItalsoincludesrequestswhereajusticesays(s)heonlyagreeswith certainsectionsoftheopinionandothersimilarrequests. 5 Œ(1.63%)Minordiscrepancyoffactthatthejusticerequestsoralternation, butdoesnotchangethepolicyorlegaloutcomeofthecase 6 Œ(0.51%)Indicationfordraftpreference(e.g.,whenmultipledraftsareincirculation,a justicesays(s)heprefersDraft2overDraft3) LegalChanges 7 Œ(2.43%)Expandthelegalruledevelopedinthecase(orelaborateonaparticularsection ofthecurrentcase) 8 Œ(6.79%)Limitormakethelegalrule(orapointorapplication)morenarrow 9 Œ(2.03%)Expandorstrengtheninterpretationofcitedprecedent 10 Œ(4.65%)Limitordistinguishinterpretationofprecedentintheopinion 11 Œ(4.28%)Add/citeadditionalprecedentorstatute 12 Œ(4.59%)Removeacitationtoprecedentorstatute 13 Œ(1.17%)Challengeconstitutionalinterpretation,wherethebargainingjusticetakesa broaderormoreinclusiveviewoftheConstitutionthandoestheopinionauthor 11 14 Œ(2.76%)Challengeconstitutionalinterpretation,wherethebargainingjusticestakesa narrowerormorerestrictiveviewoftheConstitution 15 Œ(10.37%)Requestachangetothelegalquestionansweredbytheopinion(i.e.,avoida legalissueorsettleanadditionalquestion) 16 Œ(1.26%)Clarifyinstructionsforlowercourtsand/orCongress PolicyChanges 17 Œ(5.07%)Changethecasedisposition;includingaskingthatthecasedbedismissedas improvidentlygrant(DIG),vacatedandremanded(GVR),orsummarilydecided 18 Œ(0.40%)Considerandaddressahypotheticalrelatedtothelegalquestion 19 Œ(2.62%)Broaden,strengthen,orclarifythediscussionofpolicyconsequences 20 Œ(0.90%)Changethediscussionofjurisdiction 21 Œ(0.22%)discrepancyoffact or oflowercourtopinionthat broadens the policyoutcomeorconsequencesinthecase(i.e.,theCourtexpandsreachofopinionby morebroadlyreadingalowercourtdecision) 22 Œ(0.31%)discrepancyoffact or oflowercourtopinionthat limits thepolicy outcomeorconsequencesinthecase(i.e.,theCourtnarrowsreachofopinionbymore restrictivelyreadingalowercourtdecision) 23 Œ(1.41%)RequesttheapplicationofadifferentlegalorConstitutionalstandardaltogether (thatis,arequestnottoalterthecurrentstandard,buttouseacompletelydifferentapproach toanswerthelegalquestion(s)athand) Eachdistinctcategorymakesuponlyasmallpercentageofthetotal.AsFigure1.1illustrates, somecategoriesmakemorefrequentappearancesthanothers.Ofrequestsforminorrevisions,a 12 majorityofstatementsasktheauthortosimplyinsertordeletetextthatdoesnotsubstantiallyalter themeaningoftheopinion. Figure1.1Absolutefrequencyofeachbargainingrequestbytype. Themodalcategoryinamongrequestsfocusedonalteringthelegalcontentoftheopinion arethosewhichseektolimitormorenarrowlytailortheinterpretationofthelegalruledeveloped intheimmediatecase.However,thecategoriesthatdealwithtreatmentofprecedentcompromise asizableproportionofrequestssenttothemajorityopinionauthor.Bargainingstatementsaimed todistinguishthecurrentlegalrulefromestablishedprecedentmakenearlyanequalappearance asbargainingmemosrequestingtheauthoraddorremoveanadditionalcitationtoprecedentor statute. Quiteunexpectedly,themostfrequentlyoccurringpolicychangebargainingstatements includerequeststoeitherchangethecasedispositionorapplyanentirelydifferentlegalrationale, whilenegotiationoverinterpretationoffactorlowercourtopinionsconstituteaninconsequential proportionofallbargainingstatements. 13 Usingthetripartitecategorization,thenextchapterexaminesthesedataatthejustice-level. 1.4.2.2BargainingStrategy Whilethedissertationaimstouncoverthecontentofthenegotiationsrelatingtothedevelopment ofitslegalrule,Iamalsointerestedinwhethercertainbargainingtechniquesaremoresuccessful (orinwhatcircumstancesdifferentapproachesaremostefAssuch,myassistantsandI codedthefitoneflusedforeachrequest.Theschemapresentedbelowcaptureshowajusticeframes hisorherrequest,buildingfromnon-threateningtothemostseveretypeofthreat: RequestTone 1 Œ(41.94%)Friendly/neutralsuggestion;thisisan unconditional joinwitharequestat- tached 2 Œ(28.00%)Suggestion/requestwithoutexplicitthreat;moreambiguousthanunconditional join 3 Œ(16.39%)Willjoin only ifaccommodated,butnatureofthethreatisunclear(i.e.,the justicedoesn'tsaywhattheconsequencewillbeotherwise) 4 Œ(10.75%)Willauthororjoinaseparate,concurringopinionifnotaccommodated 5 Œ(2.93%)Willdefectanddissentifnotaccommodated Theempiricalchaptersexploretheimpactofajustice'sstrategyonboththeauthor'sdeci- siontoadoptrequestsandonabargainingjustice'sdecisiontoretaliatewhennotaccommodated. ThedescriptiveillustrationinFigure1.2revealsthatjusticesmostoftenframetheirsuggestionsin friendlylanguage.Infact,asthedegreeofhostilitygoesup,theproportionofrequestsusingthose strategiesplummets. 14 Figure1.2Absolutefrequencyofeachbargainingstrategybytype. 1.4.2.3IdentifyingAccommodation Todeterminewhotheopinionauthoraccommodates,whattypesofrequestsaremetwithacqui- esce,andwhenchangesaremorelikelytooccur,Icreatedadirectmeasureofaccommodationin threesteps:First,IconvertedtheJPEGimagesofthemajorityopiniondraftstosearchabletext usingopticalcharacterrecognition(OCR)software.Second,Ideterminedthechangesin eachcirculateddraft.Todothis,Ifollowthepathofotherstudies(BlackandOwens2012;Corley 2008;Corley,CollinsandCalvin2011)andusethecommonplagiarismsoftwareWCopyFindto revealtheoverlapbetweentwoormoredocuments.Thismethodallowsmetoidentifythelocation andmagnitudeofthechangesfromtheinitialdrafttoeachsubsequentcirculationoftheopinion. Third,Iconnectedthechangeintheopinionwiththeexplicitrequestscontainedinmemoranda fromtheotherjusticesbymergingtheopiniondraftdatawithinformationfromtheSupremeCourt OpinionWritingDatabase.Thisapproachcreatesdyadsoftheopinionauthorandeachresponding justiceattherequestlevelforeachcase,allowingforthecreationofauniquebinarydependent variableusedintheprimaryempiricalcontributionofthisdissertation. 15 Armedwithnoveldataforanalysis,Iturnnowtodescribingtheindividualchaptersthat comprisethedissertation. 1.5SummaryofChapters Theoverarchinggoalofthisdissertationistoprovideanempiricalmicro-foundationofjudicial behaviortoinformtheoriesofintra-courtbargainingandtoimproveourunderstandingofthe SupremeCourt'sopinionwritingprocess. Thenewdatacollectedforthisprojectprovidethecoreforempiricalanalysisinthere- mainingchapters,whichcanbereadastheoreticallyrelated,butempiricallyindependentessays. CompromiseandControlintheCollegialDevelopmentofU.S.SupremeCourtMajorityOpinions , Chapter2,investigatesanauthor'sdecisiontoaccommodatehisorhercolleagues.Thenextchap- ter, (Dis)EngagementintheSupremeCourt'sOpinionWritingProcess stepsbacktoconsiderhow ajustice'sdecisiontoengagewiththemajorityopinionauthorintheinstanceshapesthebar- gainingprocessandtheoutputoftheCourt.Thefourthandsubstantivechapter, Concurring OpinionsinaFracturedMajority ,analyzeswhathappenswhennegotiationbetweentheauthor andhercolleaguesgoesawry.Inthesechapters,researchdesignandresultsareprecededbythe- orysectionsthatprovidepredictionsthataimtoguidetheinterpretationofresults.Achapter providesabriefconclusiontoconnectthecentralandpointtopossibleavenuesforfurther research. In CompromiseandControlintheCollegialDevelopmentofU.S.SupremeCourtMajority Opinions ,IprovideanempiricaltesttodeterminewhichmembersoftheCourtareactually entialinshapingthecontentoftheopinionbyexamininganauthoringjustice'sdecisiontoalter themajorityopinionasaresultofthatnegotiation.Myresultssuggestthatwhileopinionauthors arelargelyamenabletotheircolleagues,contrarytopopulardiscourse,itisnottheCourt'sfiswing justiceflthatisprincipallybutratherthatthemedianmemberofthemajoritycoalition holdsparticularswayinthedevelopmentoftheopinion.Thisresultrecenttheoreticalre- 16 tomodelsofjudicialbargaining.Moreimportantly,Ithatthesuccessofajustice's attemptstopersuadetheopinionauthortoalterthemajorityopiniondependsmainlyonthecon- tentofthenegotiationaswellasthetoneorframeoftherequest.Inparticular,opinionauthorsare easilypersuadedtoadoptsuggestionsforminorrevisionstoimprovetheclarityoftheopinionand, toalesserdegree,areresponsivetorequeststoalterthelegalprovidedintheopinion. Bycontrast,requeststochangethepolicyimplicationsortomodifybreadthoflawsettledinthe Court'sopinionaremetwiththemosthostilitybyopinionauthorsandarelessfrequentlyadopted. Ialsothathowjusticesframetheirrequestsaffectsthelikelihoodthattheauthorrevisesthe opinion.Whenabargainingrequestcarrieseitheranexplicitorimplicitthreattothestabilityof themajoritycoalition,theopinionauthorbecomesespeciallyactiveinthenegotiationprocessand workstoappeasethebargainingjustice. Inthesecondsubstantivechapter, (Dis)EngagementintheSupremeCourt'sOpinionWrit- ingProcess ,Itraceajustice'sinitialdecisiontoengageŒornotŒwithassignedopinionauthor. HereIdissectajustice'sresponsetotheinitialmajorityopiniondraftwhereajusticemayoptto joinitentirely,negotiatewiththeauthortomakechanges,orwriteaconcurringopinionfromthe outset.Iagainevidencethatthemedianmemberofthemajoritycoalitionholdsan positionincraftingtheCourt'sopinion.Theothersmembersarelesslikelytorequestalterationsto theopinionwhensuchajusticeistheassignedauthor.Thoughtheauthor'sidentitymaysuppress bargainingactivity,theresultsrevealthatthesinglegreatestpredictorofajustice'swillingnessto engagewiththeauthoristhedegreeofcooperationbetweenthejusticeandtheauthor.Whena justiceandtheauthorhaveahighdegreeofcooperation,ajusticeisfarlesslikelytoasktheauthor tomakechangestotheCourt'sopinion.Additionally,caseandcontextualfactorsplayagreater roleinapotentialbargainer'sdecision-makingcalculus.Justicesaremorelikelytoengagewiththe authorwhenthecaseislegallycomplex,butfreshmanjusticesareleastlikelytobargainacrossthe board.Thispaperpushesthequestionfurthertoconsiderhowfidisengagedfljusticesreacttothe majorityopinion.IthatwhentheCourt'smedianauthorstheopinion,ajusticeislesslikelyto requestchanges,butisfarmorelikelytopenaseparateconcurringopinion.Similarly,justicesare 17 moreapttoforgobargainingtowriteseparatelyinlegallycomplexcasesandastheirideological distancefromthemeanofthemajoritycoalitiongrows.Justicesaremorelikelytosimplyaskfor changeswithouttakingthesteptowriteaconcurringopinionasthewinningmarginincreases. Finally,Ithatjusticeswhoareexpertsinaparticularissueareaaremorelikelytoengagewith themajorityopinionauthorthanjusticeslackinginspecializedknowledge. Inthethirdinvestigativechapter, ConcurringOpinionsinaFracturedMajority ,Iconsider thereversedrelationshipbetweentheopinionauthorandtheotherjusticesontheCourt.Com- biningtextanalysissoftwarewithcontentanalysisofthewrittenexchangesbetweenjustices,I descriptivelyanalyzetheconsequencesoffailedorincompletenegotiation.Contrarytoexpecta- tion,myresearchthatnegotiationcontentisnotthesingularpredictorofconcurringopinion behavior.Rather,inlinewithintuitionderivedfromtheoriesofbargainingandcooperation,I thatjusticesaremoreapttoauthororjoinaseparateopinionwhentheengagedjusticehasem- ployedastrategythatinvokesamorehostiletoneduringnegotiationwiththemajorityopinion author.Thatis,concurrencesaremorelikelytoemergeifajusticethreatenedthemajorityopinion authorduringthebargainingprocess.Despitethefactthatthebargainingjusticethreatensthesta- bilityofthemajoritycoalition,theopinionauthorrisksafracturedmajorityandrefusestoappease hercolleagueinfavorofmaintaininggreatercontroloftheopinioncontent.Althoughbargaining justicesworktomaketheirthreatscredible,thisgambleisfrequentlyworththeriskasthefailure toplacateacolleague'srequestforchangetothemajorityopinionresultsinretaliatoryactionin onlyafractionofcases.Inshort,thedatasuggestthereisageneralreluctancetoauthorconcurring opinionstohighlightinstancesofnon-accommodation. ThechapterconcludesbyreviewingmycentralargumentsandAcknowl- edgingthelimitationsofthedissertationproject,theconclusionpointstopossibleapproachesfor futurescholarship.Inparticular,theresultspresentedinthelatteressaysraisenewpuzzlesand suggestclearavenuesforavenuesforresearch.Forexample,takentogether,theresearchpre- sentedheresuggeststhatanextlogicalstepistoexaminethedownstreamconsequencesoffailed negotiationduringtheopinionwritingprocessbyconsideringtheeffectsofconcurringopinions 18 oncomplianceinthelowerfederalappellatecourts.Todate,empiricalevidencesuggeststhata justiceismorelikelytowriteaconcurringopinionastheideologicaldistancefromthemajority coalitionmedianincreases.Descriptivedatarevealconcurrenceson both sidesofthecoalitional median,raisinganunexplainedquandaryconcerningtheofconcurrencesasasignalto lowercourts:ifthemedianmemberoftheCourtauthorsaconcurrenceinaprecedent-settingcase, itsignalsthatmedianisnotentirelywiththelegalruleestablishedinthemajorityopinion. Lowercourts,then,haveachoicetofaithfullyexecutethelegalruleandriskunfavorablereviewby theSupremeCourtortoadoptthelegalrulemostpreferredbythemedianmemberoftheSupreme CourtbyannouncingadecisionthatisatleastpartiallyinconsistentwiththeCourt'smajorityopin- ion.Infutureresearch,buildingoninsightsfromthisdissertationproject,Iwillexplicitlyconsider concurrencesasstrategicbehavior,ratherthansimplyasexpressiveactions,andIutilizeasimple spatialmodeltodemonstratehowsuchapatternmightencouragenoncompliancefromalower court,allowingexploitationofcoalitionaldivisionsamongmembersoftheSupremeCourt. 19 REFERENCES 20 REFERENCES Black,RyanC.andRyanJ.Owens.2012. TheSolicitorGeneralandtheUnitedStatesSupreme Court:ExecutiveandJudicialDecisions .NewYork:CambridgeUniversityPress. Brennan,WilliamJ.,Jr.1960.fiStateCourtDecisionsandtheSupremeCourt.fl PennsylvaniaBar AssociationQuarterly 31(June):393Œ407. Brenner,Saul.1995.fiFluidityinVotingontheUnitedStatesSupremeCourt:ABibliographic OverviewoftheStudies.fl LawLibraryJournal 87(Spring):380Œ386. Bryan,AmandaandTimothyR.Johnson.2012.ThePowerofthe(Red)Pen:Bargainingand StrategicAccommodationontheU.S.SupremeCourt.AnnualMeetingoftheMidwestPolitical ScienceAssociation. Carrubba,Cliff,BarryFriedman,AndrewD.MartinandGeorgVanberg.2012.fiWhoControlsthe ContentofSupremeCourtOpinions.fl AmericanJournalofPoliticalScience 56(2):400Œ412. Clark,TomS.2011. TheLimitsofJudicialIndependence .NewYork:CambridgeUniversityPress. Clark,TomS.andBenjaminLauderdale.2010.fiLocatingSupremeCourtOpinionsinDoctrine Space.fl AmericanJournalofPoliticalScience 54(4):871Œ890. Corley,PamelaC.2008.fiTheSupremeCourtandOpinionContent:TheofParties' Briefs.fl PoliticalResearchQuarterly 61(3):468Œ478. Corley,PamelaC.,PaulM.Jr.CollinsandBryanCalvin.2011.fiLowerCourtonU.S. SupremeCourtOpinionContent.fl JournalofPolitics 73(1):31Œ44. Epstein,LeeandJackKnight.1998. TheChoicesJusticesMake .Washington,DC:CQPress. Levi,EdwardH.1949. AnIntroductiontoLegalReasoning .Chicago:UniversityofChicagoPress. Maltzman,Forrest,JamesF.IISpriggsandPaulJ.Wahlbeck.2000. CraftingLawontheSupreme Court .NewYork:CambridgeUniversityPress. Maltzman,ForrestandPaulJ.Wahlbeck.1996.fiStrategicPolicyConsiderationsandVotingFlu- idityontheBurgerCourt.fl AmericanPoliticalScienceReview 90(3):581Œ592. Mendelson,Wallace.1963.fiTheNeo-BehavioralApproachtotheJudicialProcess:ACritique.fl AmericanPoliticalScienceReview 3(593-603). Rohde,DavidandHaroldJ.Spaeth.1976. SupremeCourtDecisionMaking .SanFrancisco:W.H. Freeman. Schwartz,Bernard.1993. AHistoryoftheSupremeCourt .NewYork:OxfordUniversityPress. 21 Segal,JefferyA.andAlbertD.Cover.1989.fiIdeologicalValuesandtheVotesoftheU.S.Supreme CourtJustices.fl AmericanPoliticalScienceReview 83:557Œ565. Segal,JefferyA.andHaroldJ.Spaeth.1993. TheSupremeCourtandtheAttitudinalModel .New York:CambridgeUniversityPress. Segal,JefferyA.andHaroldJ.Spaeth.2002. TheSupremeCourtandtheAttitudinalModel Revisited .NewYork:CambridgeUniversityPress. Spriggs,JamesF.II,ForrestMaltzmanandPaulJ.Wahlbeck.1999.fiBargainingontheUS SupremeCourt:Justices'ResponsestoMajorityOpinionDrafts.fl JournalofPolitics 61(2):485Œ 506. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.1998.fiMarshallingtheCourt: BargainingandAccommodationgontheUnitedStatesSupremeCourt.fl AmericanJournalof PoliticalScience 42(1):294Œ315. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.2009. TheBurgerCourtOpinion WritingDatabase .Washington,DC:GeorgeWashingtonUniversity,DeptofPoliticalScience. 22 CHAPTER2 COMPROMISEANDCONTROLINTHECOLLEGIALDEVELOPMENTOFU.S. SUPREMECOURTMAJORITYOPINIONS 2.1ChapterAbstract TheargumentscontainedwithinSupremeCourtmajorityopinionsdeterminelawandwieldau- thoritytoalterthescopeofpublicpolicy.Existingresearchpresentscompetinghypothesesabout bargainingpowerontheCourt,withmanymodelspredictingconvergencetothepreferencesofa singlemedianmember,whileotherssuggestthattheopinionauthorherselfretainscon- trol.Despitethesepredictions,datatodisentangleindividualmemberhave remainedelusive.Inthischapter,Iredressthisybypresentingevidencefromanovel datasetcomprisedofeachmajorityopiniondraftcirculatedamongmembersoftheU.S.Supreme CourtduringtheBurgerCourtera.IcreateadirectmeasureofaccommodationontheCourtthat matchesbargainingactionswithchangesinthecontentofthemajorityopiniondrafts.Myresults suggestthatwhileopinionauthorsarelargelyamenabletotheircolleagues,themedianmemberof themajoritycoalitionholdsparticularswayinthedevelopmentoftheopinionbywhen ajusticenegotiatesandbyaffectinganauthor'sdecisiontoadoptachange.Ifurthershowthatthe contentandtoneofthebargainingrequestthelikelihoodofaccommodation. 2.2Introduction Argumentsmadeinthemajorityopinionaresteepedinthenuanceoflawandhavethepowerto affectdecision-makinginlowercourts,toalterthereachofpublicpolicy,andtothe privatebehaviorofindividualcitizens.Foranopiniontohaveprecedentialauthority,atleastve membersoftheCourtmustsignontoasingleopinion,openingthedoorforactivebargaining betweentheopinionauthorandothermembersoftheCourt.Yetpreciselyhowthecollegial 23 Courtproducesthe content ofthepolicycontainedwithinthemajorityopinionremainsshrouded inmysterybecausetheopinionwritingprocessissecretive.Competingtheoreticalexpectations suggestthatdifferentmembersoftheCourtŒsimplybecauseoftheirrelativeideologicalpositions Œaremorelikelytoseetheirpolicypreferencestranslatedintolaw.Thischapterpresentsoriginal archivaldatathatopenstheblackboxtoprovideasystematicempiricalexplorationintotheintra- Courtbargainingprocess.Ithatthemedianmemberofthemajoritycoalitionhasparticular swayinthedevelopmentoftheCourt'sopinion,and,moreimportantly,thatthesuccessofa justice'sattemptstopersuadetheopinionauthortoalterthemajorityopiniondependsmainlyon thecontentofthenegotiationaswellasthetoneorframeoftherequest. ThescholarlyliteraturehasofferedthreedistinctanswersaboutwhichmemberoftheCourt hasthemostoverthecontentordirectionofthemajorityopinion. Medianjustice theo- riesequatetheSupremeCourttoalegislatureandcontendthatthelegalrulewilltheideal pointoftheCourt'smedianmember(Epsteinetal.2005;Hammond,BonneauandSheehan2005; Martin,QuinnandEpstein2005;Schwartz1992).Incontrast,othertheoriesfocusontheroleof theopinionauthorandeitherpositan authormonopoly outcome,wherethelegalrulethe preferenceoftheauthor(Hammond,BonneauandSheehan2005),oralessdeterminant author outcome,wherethelegalrulepartlytheviewsoftheauthor(LaxandCameron 2007).Finally,byemphasizingthedistinctroleofthedispositionalvote,morerecenttheoretical approachesfocusontheuniqueinstitutionalfeaturesoftheCourtandassertthatcontentofthe opinionmostfrequentlyconvergestotheviewsof medianmemberofthemajoritycoalition (Car- rubbaetal.2012).Mylargelysupportthelastaccountandprovideevidenceofhowthe medianmemberofthemajoritycoalitionarrivesatthispositionofpower. Whileeachofthesetheoreticalmodelsyieldspredictionsabouttheoutcomeofintra-Court bargaining,directempiricalevidencehasnotfollowedsuitandithasnotbeenfeasibletodetermine whomattersinthedevelopmentofthemajorityopinionorunderwhatcircumstanceswemaysee varyingdegreesofUsingtheprivatepapersofjusticesthatservedduringtheBurger Courteraallowsmetopeelbackthecurtainofsecrecythatenshroudstheopinionwritingprocess 24 and,indoingso,revealsavibrantnegotiationprocessbetweentheopinionauthorandtheother membersoftheCourt.Take,forexample,the1978caseof Gladstone,Realtors,etal.v.Village ofBellwood ,wheretheCourtwasaskedtosettleaquestionofstandingunderthefederalFair HousingAct.Afteroralargument,theconferencevotesuggestedaneasydispositionaldecision, establishinga7-to-2majoritythatessentiallyafthelowercourt'sdecisionwhichbroadly interpretedtheintentofthefederallegislationasitappliedtostandinginanti-discriminationsuits. ChiefJusticeBurger,amemberofthemajority,assignedtheCourt'sopiniontoJusticePowell, who,inturn,circulatedthedraftoftheopinioninFebruary1979.Themedianmemberofthe Court,JusticeBlackmun,agreedtoPowell'sproposedmajorityopinion,castingavotetojointhe veryrstdraftwithoutdelay.ThoughtheiracquiescencecamemoreslowlythanBlackmun's,two othermembersofthemajoritycoalition,BurgerandJusticeMarshall,alsojoinedPowell'sopinion withoutattemptingtogetPowelltoeditthedraft'scontent. However,thethreeremainingmembersoftheCourt'smajoritycoalitionactivelyengaged withPowell,requestingsubstantivechangesbemadetotheopinion.Inparticular,these membersoftheCourt'smajorityquibbledabouttheinclusionofasinglefootnote,whichdrew adistinctionbetweensomeoftherespondentsinthecase.Intheinitialopinion,Powellsought toaftheCourtofAppealsrulinginfavorofonlysomeoftherespondents,buttoreversethe standingdecisionforothers.ThechallengetoPowell'slegalreasoningcamefromJustice White,themedianmemberoftheCourt'smajoritycoalitioninthisparticularcase,justfourdays aftertheinitialmajoritydraftcirculation.Inhisresponsememo,Whiteindicatedthathewasfiquite with[the]opinioninthiscasewiththeexceptionof[thefootnote].flItwasWhite'sview thatPowell'sopinionwastoorestrictive.Inamemodatedsixdayslater,JusticeBrennanpushed thispointfurther,suggestingthatthestandingissueforthoseotherindividualswasnotaquestion beforetheCourtandrequestedthatPowellmakeitclearthattheCourt'sopiniondidnotestablish whethersuchpersonshavestandingtosue.Finally,onthesamedaythatBrennancirculatedhis request,JusticeJohnPaulStevenslobbiedPowelltofisimplyomitflthefootnote. Overthecourseofthreeadditionalmajoritydrafts,JusticePowellsoughttoincorporate 25 thesubstantiverequestsmadebybothJusticesWhiteandBrennan,whileentirelydisregarding Stevens'mostpreferredoutcomebyhisrefusaltoremovethefootnoteentirely.Intheopinion, Powellmaintainedhisdistinctionbetweentherespondents,butadoptedWhite'srequestanddid notofferaprescriptive,hardandfastlegalruletodistinguishamongtherespondents.Instead, heBrennan'srequestandremandedthequestiontothelowercourt.Theopinion contentisadirectculminationofnegotiationandcompromise. Thisparticularexampleraisesimportantquestionsabouttheopinion-writingprocess.In particular,thoughthe medianjustice (Blackmun)readilysignedontohisopinion,JusticePowell adoptedtwoofthethreesubstantiverequestsmadebyhiscolleagues,suggestingthattheopinion authorisamenabletoalterthecontentoftheopinionatthebehestofhiscolleaguesandundermin- ingtheideaofanalmighty authormonopoly .Moreover,preciselywhoseviewsareincorporated intotheCourt'sopinionsmaybesurprising.Inthiscase,whilePowellhastilyadoptedasuggestion fromthe majoritycoalitionmedianmember (White),theslightlyright-of-centerPowellalsoac- tivelyengagedinthenegotiationprocesswithanideologicallyextremeliberalmember(Brennan), whilevirtuallyignoringtheviewsofamoreproximateideologicalally(Stevens).Morecritically, however,thesinglefootnoteopenedthedoorforfurtherlitigationandshiftedthescopeofthe Court'sopinion.Inparticular,ratherthanafthelowercourtinpartandreversinginpart, andthereforeestablishingnationalprecedentonthestandingissue,theSupremeCourtduckeda potentiallystickyissueandremandedthecasebacktothelowercourt. Clearly,then,theCourt'sopinionsŒsometimessinglefootnotesŒhavefarreachingim- plicationsforAmericansociety.Inthischapter,Iutilizeoriginaldataonthechangesinmajority opiniondraftstounderstandhowthemembersofacollegialcourtarriveatajointopiniontoset legalpolicy.More,Iusecontentanalysisofthebargainingmemorandadirectedto majorityopinionauthorstoshedlightonthedistinctivetypesofchangessought.Ithentracethe treatmentoftheserequeststhroughtheiterativeopinionwritingprocess.Suchanapproachmakes itispossibletoexaminewhoserequestsareincorporatedintotheCourt'sopinionandwhoseare not.Simplystated,Ianalyzewhichmember(s)ofthecourtare inbargaining withthe 26 majorityopinionauthor,whichauthorsare abletobe ,what typesofrequestsareac- commodated bytheauthor,and,,when,or underwhatconditions ,accommodationoccurs. Ibeginbytracingtheopinionwritingprocessthattakesplacebehindtheclosed doorsoftheMarblePalaceinSection2.3.Next,inSection3.4,Ipresentestablishedtheories aboutwhichjusticescontrolthecontentoftheCourt'smajorityopinionandgeneratehypotheses toexaminetheconditionsunderwhichauthorsdecidetoaccommodatetheircolleagues.Ithenex- plainthedataandmeasurementinSection2.5beforepresentingtheresultsfromempiricalanalyses inSection3.5.Section3.6concludeswithadiscussionoftheresults. 2.3TheOpinionWritingProcess AftertheSupremeCourthearsoralargumentinacase,thejusticesmeetinprivateconference toestablishthedispositionalmajoritycoalition.Duringthisconference,indescendingorderof seniority,thejusticesdiscusstheirindividualunderstandingofthecaseandcastinitial,non-binding votestodeterminethemajorityviewoftheCourt.Importantly,whilethesevotesestablisha tentativewinningcoalition,thevotesinconferencearenotirrevocable.Indeed,extantscholarship showsthatthereisfromthesevotestothereleaseoftheCourt'ssignedopinions(e.g., Brenner1995;MaltzmanandWahlbeck1996).Afteroralargument,then,thejusticeshavethe opportunitytoconveytheiropinions,buthaveampleopportunitytotheseviewsthroughout theopinionwritingprocess. Oncetheinitialmajorityanddissentingcoalitionsareestablished,themostseniormember ofeachcoalitionassignsoneofthejusticestowriteanopinionfortheCourt.Asthemostsenior memberoftheCourt,ifthechiefjusticevoteswiththemajority,theopinionassignmentfallson hisshoulders(Schwartz1993).Intheeventthatthechiefjusticeisintheminority,themostsenior associatejusticeisresponsibleformakingtheopinionassignment(SegalandSpaeth1993).The assignedopinionauthoristaskedwitharticulatingtheCourt'slegalrationaletojustifyitsdecision inthecaseand,upondoingso,(s)hecirculatesaninitialdraftofthemajorityopinion. 27 Whilethisseemstobeatremendousamountofpowerinthehandsofasinglejustice, foranopiniontohaveprecedentialauthority,atleastvemembersoftheCourtmustsignonto asingleopinion.Thisopensthedoorforactivebargainingbetweentheopinionauthorandother membersoftheCourt.Thatis,afterthejusticesreceiveadraftofthemajorityopinion,they maychoosetoattempttopersuadetheopinionauthortoalterthecontentoftheopinion.Asthe Gladstone examplereveals,theoutputoftheCourtmayverywellbeaproductofnegotiation andcollaboration. Outofdeferencetotheassignedmajorityopinionauthor,theotherjusticestypicallywait fortheauthortocirculateadraftoftheCourt'sopinionbeforetheyofferanycompeting concurringordissentingopinions.Oncetheinitialdraftiscirculated,however,thejusticesarefree tocirculateseparateopinionsortoengageinbargainingwiththeauthor.InthewordsofJustice WilliamBrennan:fiBeforeeveryonehasmadeuphismind[thereis]aconstantinterchange amongus...whilewehammerouttheformoftheopinionfl(Brennan1960,405,asquotedin Maltzman,Spriggs,andWahlbeck2000,9-10).Iturnnowtotheorytoexaminehowthisficonstant interchangeflmoldstheoutputcontainedintheCourt'ssignedopinion. 2.4TheoryandHypotheses Morethanhalfofacenturyofpoliticalsciencescholarshipreliesonthebasicassumptionthat justicesaremotivatedbyadesiretoseetheirpolicypreferencestranslatedintolawandthatthis motivationdrivesthebehaviorthatweobserveduringtheopinionwritingprocess.Buildingon thatbasicpremise,IexaminehowjusticesŒsimplyonaccountoftheirideologicalorcollegial positionontheCourtŒthelikelihoodthattheirviewsareincorporatedintotheCourt's majorityopinion.Ifurtherconsiderhowthecontentandtoneofthebargaininginteractionbetween theauthorandhercolleaguesaffectsthedraftoftheCourt'sopinion.Finally,Iexaminehow otherinstitutional,contextual,andpoliticalfactorstheproductionofthelaw. 28 2.4.1JusticePowerandOpinionControl TherearethreedistinctformaltheoriesthatpredicttheideologicallocationofSupremeCourt majorityopinions,twoofwhichrelyontheideologicallocationofthejusticesandonethatexam- inestheuniqueroletheassignedauthorplays.Eachofthesetheoriesyieldsdifferenthypotheses aboutwhichmember(s)ofcontrolthecontentoftheCourt'sopinion. ThepredominantmodelofSupremeCourtdecision-makingistheso-called medianjustice theory,whichtakesitsformalrootsfromthemedianvotertheorem(Black1958).Thehistorical dominanceofthismodelisunsurprisingbecauseitssimplicityisintuitivelyappealing:tomain- tainastablecoalitionandsetprecedent,thepolicypreferenceofthemedianjusticewieldsgreat powersincehisorhervoteisnecessarytoachieveasimplemajority.Scholarscontendthatwhen weassumeauni-dimensionalpolicyspacewherejusticeshavesinglepeakedpreferences,thema- jorityopinionoftheCourtwillbewrittenattheidealpointofthemedianjustice(Epsteinetal. 2005;Hammond,BonneauandSheehan2005;Martin,QuinnandEpstein2005;Schwartz1992). Accordingtothistheory,Iexpectthat MedianJustice: themedianjusticewillwieldgreaterpoweroveropinion contentthanhercolleagues,whethersheistheauthororamemberoftheCourt's majoritycoalition . More,themedianjusticetheorysuggeststhatanauthorshouldbeespeciallykeento accommodaterequestsfromthemedianmemberoftheCourt.Atthesametime,ifthemedian justiceistheassignedauthor,thistheoryindicatesthatthejusticesshouldbelesslikelytocede controlandlesslikelytoaccommodaterequestsfromhercolleagues. Arguingthatmedianjusticemodelsarebasedontheoriesofthelegislature,whichassumes atwo-partysystemwithmajoritarianrule,AndersonandTahk(2007)suggestamoreapplicable modelaccountsforinstitutionalfeaturesuniquetotheCourt.Theirapproachpermitsotherjustices tooffercompetingopinions.Whilethemodelstillpredictsconvergencetoamedian,itisnolonger themedianoftheCourtthatwieldsinordinatebutratherthemedianoftheparticularle- 29 galissue.Similarly,Carrubbaetal.(2012)considertheroleconcurringopinionsplayinshaping thepolicyoutputoftheCourt.Thismodelpredictsthat,inmostcases,themajorityopinionwill convergetotheidealpointofthe medianmemberofthemajoritycoalition .Thistheory,then, predictsthat MedianMemberofMajorityCoalition: themedianmemberofthemajority coalitionshouldholdparticularcontrolovertheopinion'scontent . Inparticular,ifandwhenthemedianmemberofthemajoritycoalitionoptstoengageinbargaining withtheauthor,theauthorshouldbemoreaccommodatingtoherrequeststhanshewouldother- wisebe.Similarly,ifthethemedianmemberofthemajoritycoalitionistheassignedauthor,she maybelesswillingtheaccommodatebargainingrequests. Contrarytomodelsthatpredictconvergencetoamedian,othermodelsgivethemajority opinionauthorvaryingdegreesofcontroloveropinioncontent.Inparticular,Hammond,Bonneau andSheehan(2005)proposean authormonopoly modelthatsuggestspolicyisafunctionofthe preferencesoftheopinionwriterandtheCourtmedian.ThismodelreliesonstrictŒandunre- alisticŒassumptions,proposingbothedpolicyalternativesandthatthemembersoftheCourt willinglycedeficompletecontroloftheCourt'sagendatothemajorityopinionwriterfl(Hammond, BonneauandSheehan2005,111).Conversely,LaxandCameron(2007)developamodelof au- thor .,theycontendthatcirculatingacompetingopinioniscostlyinterms oflegalqualityaspersuasiveness,clarity,andcraftsmanship).Asaresult,thecontent oftheopinionisbycharacteristicsoftheassigningjustice,theassignee,thepotential counterwriter,andthecharacteristicsofthecase.Giventhemultiplicityofinterveningfactorsin themodel,thereisnoonesinglepredictedoutcomeacrosslargenumbersofcases. Inshort,theliteraturepositscompetinghypothesesaboutthemostmemberson theCourt,withmanymodelspredictingconvergencetoasinglemedianmemberofsometype. Otherapproachessuggestthattheopinionauthorhassomedegreeofcontrolovertheopinion content. 30 2.4.2ContentofRequestsandtheLikelihoodofAccommodation Thetheoriesofauthorsuggestthattheassignedauthorhasanimportantrolein thecontentoftheCourt'sopinion.Yetwhenthetdraftiscirculatedandthedoor openstonegotiation,thejusticesmayworktogethertocraftthelegalrule.Intuitively,we mightexpectthatthelikelihoodthatajusticeisabletothelanguageorcontentofthe majorityopiniondependsinnosmallpartonpreciselywhatitisthatthejusticeisaskingthe authortodo.Inotherwords,itisthecontentoftherequestthatdeterminesthemagnitudeof changenecessaryfortheauthortoadequatelyaccommodatetherespondingjustice.Thecontentof therequest,then,functionsasasignalofthedivergencebetweenthelegalpolicyexplicatedbythe opinionauthorandthemostpreferredoutcomeoftherespondingjustice.Additionally,identifying thecontentoftherequestilluminatesthepossibilityforstrategicinteractionbetweenthemembers ofthecollegialcourt.If,forexample,thebargainingmemosarelargelycomprisedofrequests toclarifyorsubtlyalterlanguageortogrammaticalerrorswithintheopiniondraft,itwould suggestthattheopinionauthorhasabilitytocontrolthepolicycontent.Conversely,if thebargainingmemospredominantlyattempttopersuadetheauthortomodifyalegalargument,or torelyonalternativeand/oradditionalprecedents,ortoadoptadifferentoutcomealtogether,then theevidencewouldbesuggestiveofastruggleforcontroloverthelegalandideologicalposition oftheopinion. Becauseminorchangestogrammarandwordchoicearerelativelyeasytoimplementwith alowopportunitycostfortheopinionauthor,Iexpectthat Content1: opinionauthorswillbemorelikelytoaccommodateminorrevisionrequests comparedtorequestsforsubstantivelegalorpolicychanges . Putdifferently,changestoeitherthepolicyortothelegalrulecontainedwithintheopinionare bothharderfortheauthortoimplementandmoredirectlyanaffronttotheviewsoftheauthor explicatedintheopiniondraft.Assuch,Iexpectthattheserequestsarelesslikelytobeaccommo- datedthansuggestionsforminorchanges.Iarguethatafurtherdistinctionmaybemadebetween 31 legalandpolicyrequests.Whilelegalrequestsmightbemiredinideologicalinterpretation,atleast atfacevalue,suchrequestsaredressedintheguiseoftryingtoestablishthecorrectinterpretation oflaw.Thus,Iexpect Content2: opinionauthorswillbeleastlikelytoaccommodaterequeststhatseekto alterthescopeorinterpretationofpolicy . Additionally,Iexpectthathowmuch,thequantityorthesheer numberofrequests ,acol- leagueasksforwilltheauthor'swillingnesstoaccommodate.Ifajusticeaskstheauthor tomakemultiplerevisions,itsubstantiallyaddstotheauthor'sworkload.Moreimportantly,per- haps,theauthormayfearthataccommodatingmultiplerequestsfromonecolleaguemayalienate othermembersofthemajoritycoalitionandwillthereforebemorereluctanttorevise theopiniondraft.Asaresult,Ihypothesize Content3: opinionauthorswillbelesslikelytoaccommodateasthenumberofrequests increases . 2.4.3MeansofPersuasion Onceajusticemakesthedecisiontoengagewiththeopinionauthor,undoubtedlyshehopesto inducetheauthortoadoptherrequestforchange.Bytheverynatureoftheinstitutionalrule,for amajorityopiniontobindlowercourtsandsetnationalprecedent,atleastvemembersofthe Courtmustsignontothemajorityopinion.Thisprovidesbargainingjusticeswithanimportant tooltoengenderaccommodation.Anopinionauthormustbalanceherdesiretokeeptheopinion asclosetoherpreferredoutcomeaspossiblewiththeneedtosecureatleastfouradditionalvotes. Thissuggeststhat,especiallyinminimumwinningcoalitions(discussedingreaterdetailbelow), anauthormaybeamenabletoaltertheopinionatthebehestofhercolleagues.Thisshouldbe particularlytruewhenabargainingjusticeinsomewayindicatesthathervoteisupforgrabs. Notably,bargainingjusticeshavemultipletoolsofpersuasionattheirdisposal.Themost 32 obvious,perhaps,istothreatentoretractthenon-bindingconferencevotefromthemajoritycoali- tionandtodefecttothedissent,therebyweakeningtheeffectofprecedent(Fowleretal.2007). However,asCarrubbaetal.(2012)argue,evenifthejusticesdisagreeaboutthelegalrule,defec- tionsfromthedispositionalvoteareextraordinarilyrare.Inmostcases,todefectentirely,ajustice wouldhavetoswitchfromavotetoafalowercourt'sdecisiontoreverseit.Sowhilethisis astrong-armedtacticavailabletothejustices,itmightnotbethemostcredible.Amorecredible meanstopersuadetheopinionauthormaybetosignalawillingnesstowriteorjoinaseparatecon- curringopinion.Althoughconcurrencesdonotcausethesamedegreeofdamagetothelongevity ofprecedentasdodissents(HansfordandSpriggs2006;Fowleretal.2007),itservestopublicly displaydisagreementontheCourt.Simplyput,fiajusticemustlearnnotonlyhowtoputpressure onhiscolleaguesbuthowtogaugewhatamountsofpressurearesuftobe`effective'and whatamountswillovershootthemarkandalienateanotherjudgefl(Murphy1964,asquotedin Maltzman,Spriggs,andWahlbeck2000,62).Iexpectthat PersuasiveThreats: justicesthatdeploythreatsaremorelikelytogettheirviews incorporatedintothemajorityopinionthanjusticesthatelecttorequestachangewithout indicatingaconsequenceforfailedaccommodation . 2.4.4AdditionalCollegialandContextualConstraints Currentliteraturepointstoadditionalcollegial,contextual,andpoliticalvariablesthatmight encethepropensityofajusticetoseekchangesfromtheopinionauthorand,inturn,the author'sdecisiontoadoptorrejectthatrequest.Notonlydoestheorysuggestthatopinionauthors mightbecognizantoftheideologicalorcollegialpositiontheabargaineroccupies,butitisnearly axiomaticinpoliticalsciencetoassumethattheideologicaldistancebetweenthebargainingduo undoubtedlyaffectsbehavior(e.g.,SegalandSpaeth1993).Thatis,theoreticallyspeaking,ajus- ticewhoisideologicallyopposedtotheassignedauthorhasthemosttogainifshethrowsherhat 33 intothenegotiationringandsucceedsingettingherpreferencestranslatedintolaw.Conversely, opinionauthorswillbemorereticenttoadoptrequestsfromideologicalopponents,whomayseek topulltheopinionfurtherfromtheiridealpoint.Iexpect,then,that IdeologicalDistance: anauthor'swillingnesstoaccommodatewilldecreaseasthe ideologicaldistancebetweentheauthorandbargainingjusticeincreases . Astheembodimentofpersonalpolicypreferences,ideologicaldivergenceprovidesanin- dicatorofhowcloselyalignedtheauthoriswithhercolleagues.Importantly,however,thejustices arealsolong-termcolleagues.Thatis,thejusticesservetogethertermaftertermandtheories ofreciprocityorcooperationsuggestthatitshouldnotbesurprisingtoseeatit-for-tatstrategy emergeinsuchanenvironment.Intheirseminalworkonthecollegialcourt,Maltzman,Spriggs andWahlbeck(2000)evidencethatpastcooperationaffectsthewayjusticesin- teract.AsMurphy(1964,53)suggests,pastcooperationbetweenjusticescanfibuildupareservoir ofgoodwillforlaterusefl(asquotedinMaltzman,SpriggsandWahlbeck(2000,73)).Asaresult, Iexpectthat Cooperation: whenusualalliesseekchangesfromanopinionauthor,theauthorwillbe morewillingtoaccommodatetherequest . Beyondthecollegialrelationship,justicesareaffectedbythecontextthatsurroundsthe case.TheinstitutionalfiRuleofFiveflsuggeststhatanindividualjustice'soverthecon- tentofopinionmaydependonthesizeofthewinningcoalition(Murphy1964;Rohde1972; Wahlbeck,SpriggsandMaltzman1998;Maltzman,SpriggsandWahlbeck2000).Inminimum winningcoalitions,wherethemajoritywasestablishedbyvenon-bindingvotesatconference, themajorityopinionauthorcannotaffordtoloseasinglevote.Withthisincreasedleverage,I expectthat 34 CoalitionSize: majorityopinionauthorswillbemotivatedtoappeasetherequestsfrom theircolleagueswhenthemajoritycoalitionissmaller . Justasjusticescannotcontroltheinstitutionalrulesthatgovernprecedent-settingauthor- ity,sotoojusticesareheldtoaedterm,creatingauniquewwenvironment.TheSupreme Court'sannualtermrunsfromOctobertoJune.Thus,casesthatarearguedearlyinthetermhave monthsbeforethejusticesneedtoproduceasignedopinion.Onthecontrary,whenacaseis arguedinMarch,thejusticeshaveonlythreeshortmonthstoproduceajudgmentandsettlethe law.Moreover,asthetermprogresses,thecaseloadbeginstopileup.AsJusticePowellwrote toinstructhislawclerks,fiAswemovedeeperintotheTerm,sayfromandafterFebruary,the numberofopinionscirculatedbyeachChambermultiplies,andtheproblemofkeepingabreast isaseriousonefl(asquotedinMaltzman,SpriggsandWahlbeck(2000,75)).Intheirvarious works,Wahlbeck,SpriggsandMaltzman(1998),Spriggs,MaltzmanandWahlbeck(1999),and Maltzman,SpriggsandWahlbeck(2000)demonstratethatjustices'workloadaffects bargainingactivity.Itfollowsthat Workload: opinionauthorswillbelesswillingtorevisethemajorityopinionto accommodatetheircolleaguesastheirworkloadgrows . Anaccumulationofevidencesuggeststhatnotallcasesareofequalimportancetothejus- tices(EpsteinandSegal2000;Maltzman,SpriggsandWahlbeck2000;UnahandHancock2006; Black,SorensonandJohnson2013).Black,SorensonandJohnson(2013)suggestthatjustices signalthatsomecasesmattermoretothemindividuallyduringoralargumentbyengagingmore activelyduringoralargument.Furthermore,justicesarenotimmunefromthelargersocietalcon- textandarecognizantofwhenthemediaorthepublicistunedintotheCourt'sdecision.The membersoftheCourtknowthatthelegalandpolicyconsequencesofsomecasesaresimplymore importantthanothers.Yetthepredictionfortheauthor'sresponseinsuchcasesisnotobvious.On onehand,theauthoringjusticewillbehighlymotivatedtokeepcontroloftheopinionsoastonot loseauthorshipofacriticalcase.But,ontheotherhand,insalientcases,Iexpectthat 35 Salience: theauthor'sownpolicypreferencesaremorethoroughlyentrenchedinsalient, lesseningthelikelihoodofaccommodation . Attheindividualjusticelevel,priorstudiessuggestthatjusticeexpertiseandcasecomplex- ityaffecttheopinionwritingprocess(e.g.,Brenner1984;BrennerandPalmer1988;Maltzman, SpriggsandWahlbeck2000),leadingtoadditionalhypotheses: Expertise: Authorswillbemorelikelytoheedrequestsfromjusticeswithgreater expertiseonaparticularissue . Beyondpersonalexperiencewithanissue,somecasesaresimplymorecomplicatedthanareoth- ers.Iexpectthat CaseComplexity: incaseswheremultiplelegalissuesareatplay,theopinionismore likelytobeaproductofactivedialoguebetweentheauthorandtheothermembersofthe majoritycoalition . Finally,theliteraturesuggeststhatneworfreshmanjusticesmaytaketimetoacclimate totheproceduresoftheCourt(Hagle1993;BrennerandHagle1996;Maltzman,Spriggsand Wahlbeck2000;Wahlbeck,SpriggsandMaltzman1998),whichsuggeststhat FreshmanAuthor: freshmanjusticesmaybemoredeferentialtotheircolleaguesand, subsequently,morelikelytoaccommodaterequests . 2.5DataandMeasurement Inthissection,Idescribethefailureofexistingempiricalapproachestotestthesetheoriesto establishtheneedforanewempiricalinvestigationintotheofintra-Courtbargaining. 36 2.5.1QuantifyingConnecting Who Matterswith How TheyMatter Todeterminewhomattersandtounderstandpreciselyhowtheymatter,itisnecessarytoexamine thedevelopmentofthemajorityopinionitself.Onelimitationtoourunderstandingof themechanismswhichproducetheopinionoftheCourthasbeentheunavailabilityofameasure ofaccommodation.Despiteconcretetheoreticalpredictions,ithasnotyetbeenempiricallyestab- lishedwhichmembersoftheCourtswaythecontentoftheopinion.Themostcomprehensivework onaccommodationbytheopinionauthorattemptstoaccountforcollegialandinstitutionalfactors usingrelativelycoarsemeasuresofaccommodation.Intheiroriginaltest,Wahlbeck,Spriggsand Maltzman(1998)suggestthatthesheervolumeofdraftscirculatedbytheassignedopinionwriter isevidenceofattemptstoincorporatecolleagues'preferences.Inlaterwork,thesameauthors employeventhistoryanalysis,usingtimeasaproxyforaccommodation(Maltzman,Spriggsand Wahlbeck2000).Theseapproachesarereasonableapproximationsofthebargainingprocess,but theyarewhollyincomplete.Theopinion-writingstageisquiteliterallywhenSupremeCourtjus- ticesdeterminewhatthelawis,whenthelegalruleiscontinuallyshapedandthroughactive negotiation,andwhendebateoverasinglefootnotecantakeweekstoresolve.Theendresultis anopinionthatcapturesthenuancesandcomplexityoflawandpolicy.Byfocusingonaggregate patternsofopinionwriting,previousempiricalapproachesprovideanoverlybroadgeneralization ofthedevelopmentofthemajorityopinion.Thatis,tofullyunderstandtheevolutionoflawandto understandtheideologicallocationofCourtopinions,wemustunderstandtheintricacies oftheopinion-writingprocess. Limitedbydataavailability,thecoarseproxymeasuresŒdraftvolumeandtimeŒglazeover thedyadicbargainingdynamicandfailtoactuallycapturewhenaccommodationoccurs.Whilethe datalimitationsareproblematic,theyarenotinsurmountable.Toredressthedataythat hasplaguedpastanalyses,Icollectedoriginaldatathatexplicitlyrevealthisprocess., byacquiringcopiesofeachmajorityopiniondraftcirculatedduringtheopinionwritingstageon theU.S.SupremeCourtduringtheBurgerCourtera,Iamabletodeterminehowthemajorityopin- 37 ionisalteredthroughoutthenegotiationswithothermembersoftheCourt.Thesedatacomplement theextensiveSupremeCourtOpinionWritingDatabase(Wahlbeck,SpriggsandMaltzman2009), whichcontainsinformationonallmemorandaandopiniondraftsthattheSupremeCourtjustices circulatedtotheircolleaguesbetweenthe1969and1985terms. 1 Whilethisexistingdatabasecon- tributesvaluableinsightintothenegotiationprocessbyprovidingdataonattemptsbyjusticesto thecontentoftheopinionthroughwrittenmemoranda,itlacksdetailsonthecirculation ofopinions,accountingonlyforthenumberofdrafts. TomergethenewmajorityopiniondraftdatawiththeexistingSupremeCourtOpinion WritingDatabase,Iconductedacontentanalysisoftherawdataandreadeachcirculatedbargain- ingmemoaddressedtotheopinionauthorwherethememoauthorposedanexplicitsuggestion orrequesttochangethecirculatedmajorityopiniondraft.Ithentracedthetreatmentofthatre- questthrougheachiterationofthemajorityopinion. 2 Thisapproachallowsmetodeterminethe substantivecontentandtoneofeachrequestandtodirectlyidentifyifandwhenthatrequestis addressedbytheopinionauthor.DuringthecourseoftheBurgerCourt,atotalof6,542majority opiniondraftswerecirculatedacross2,305uniquecases,withanaverageof2.8draftspercase. Therewere1,960casesdecidedbymajorityopinionfromOctoberTerm1971throughthecloseof 1 Theempiricalanalysisdropsthe1969and1970terms.Priorto1971,SupremeCourtcases didnotfollowaconsistentdocketingscheme.Asaresult,casesthatwerecarriedfromonetermto thenextwerere-docketedaswhollynewcases,makingitinfeasibletotracetreatmentofrequests fromonedrafttoanothersincethecasewasessentiallyfirestarted.flTheCourtadoptedaconsistent docketingschemefortherestoftheBurgerCourtera,whichisthefocusofthisanalysis. 2 LikethebargainingmemorandadataobtainedfromtheSupremeCourtOpinionWriting Database(Wahlbeck,SpriggsandMaltzman2009),theopiniondraftdatacomefromcase containedwithintheprivatepapersofformerjustices,including:HugoL.Black(Libraryof Congress),WilliamO.Douglas(LibraryofCongress),JohnM.Harlan(PrincetonUniversity), WilliamJ.Brennan,Jr.(LibraryofCongress),ThurgoodMarshall(LibraryofCongress),HarryA. Blackmun(LibraryofCongress),LewisF.Powell(Washington&LeeUniversity),andWilliam H.Rehnquist(StanfordUniversity).Morethan80percentofthedocumentscontainedwithinthe BurgerCourtOpinionWritingDatabasearelocatedintheLibraryofCongress.Inadditionto thesethepapersofByronR.White(LibraryofCongress),releasedinApril2012,provedto beaninvaluableresource.Indeed,morethant95%ofthemajorityopiniondatawerefoundinthe papersofJusticesWhiteandBrennan. 38 OctoberTerm1985,yetweobserveactivebargaininginjustover50percentofthetotal,oronly 1,031cases.Thebargaininginthesecasesyieldsatotal3,946uniquerequestsaddressedtothe opinionauthor,whichequatestoapproximatelyfourrequestsorsuggestionsforeachcasedecided intheseterms. 3 Thesenoveldataallowmetoexaminethreekeyaspectsofthebargainingprocessduringthe opinionwritingstagethathavethusfarbeenunknown:(1)thesubstanceorcontentofthememo authors'requesttochangethemajorityopinion,(2)thetoneorstrategydeployedbytheresponding justicesintheserequests,and,(3)whethertheopinionauthoraccommodatedthesuggestionand alteredtheCourt'smajorityopinionaccordingly. 2.5.1.1WhatJusticesWant:TheContentofBargainingMemoranda Theinterveningvariableofinterestisthe content ofthebargainer'srequest.Tothatend,Iread everysinglebargainingmemorandaandcodedforthreedistincttypesofrequests:minorrevisions, legalchanges,andpolicychanges. 4 Requestsforminorrevisionsasktheopinionauthortomake slight,non-substantivechanges(i.e.,wordchoiceorgrammaticalalterations)totheopinion.Bar- gainingstatementsthataskforlegalorpolicychange,however,asktheauthortosubstantivelyalter theopinioncontent.Requestsforlegalchangefocusontheapplicationandtreatmentofprecedent, whilepolicychangeschallengethescopeoflawsettledintheCourt'sopinion. AstheaggregatedatapresentedinTable2.1makeclear,requestsforminorrevisionscom- priseasizable44percentofallrequeststoalterthemajorityopinioncontent.Theprecisecontent 3 Theamountofbargainingvariesquitedramaticallyacrossthesamplecases,rangingfromone uniquerequestto26requestsinasinglecase.Becausebargainingmemoscouldcontainmorethan asinglerequest,theunitofanalysisistherequestratherthanthememo. 4 Afterestablishingthesethreemajorcategoriesbasedoninsightsfromextantscholarship,and inspiredbyBryanandJohnson(2012),Iusedanorganicapproach,drawnfromarandomsample of250cases,todevelopacomprehensiveŒyetparsimoniousŒsubcategorizationofeachtypeof request.Thatistosay,Idevelopedanewsubcategoryforeachnewtypeofrequestencountered inthedata.Thisensuredboththatmycodingstrategywasnotonlytheoreticallyderived,butalso sufdetailedtoovercomechallengesofinter-codersubjectivity. 39 ofthesememosrangefromsmallwordchoiceandgrammaticalsuggestionstolargerrequeststhat asktheauthortoclarifyasectionortomorecarefullykeylegalconcepts.Whatthesere- questshaveincommon,however,isthattheyaredevoidoflegalorideologicalovertones.Such requestsaresimplyaimedatimprovingthequality,clarity,andaccuracyoftheCourt'smajority opinion. RequestContentFrequencyPercent MinorRevisions1,75644.5 LegalChange1,77244.9 PolicyChange41810.6 Total3,946100.00 Table2.1Aggregate-levelrequestcontent Consider,forexample, Gilmorev.CityofMontgomery (1974),acivilrightscaseconcern- ingtheexclusiveuseofcityparksbysegregatedprivategroups.JusticeBlackmun,theassigned opinionauthor,bargainingmaneuversbothlargeandsmallfromseveralofhiscolleagues. JusticeRehnquistwasparticularlyactiveinthiscase,requestingnumerousminorrevisions.In additiontoaskingBlackmuntoaddanemphatictohighlightthecity'sdejure promotionofsegregation,Rehnquistwrites: IamindeedloathtosuggestcorrectionstoanacknowledgedmasterofEnglishusage, butsinceithasbeensaidthatevenHomernodded,Iofferthefollowing: (a)Wouldnotthedeletionofthewordfithatflinthethirdlineoftheparagraph makethesentencemoregrammatical? (b)Inthesentencebeginningwiththewordfimoreoverflinthesecondparagraphon thepage,doesfiassumeflcomecloserthanfioverlookfl? Althoughthiscategoryofrequestisnotanattempttoalterthelegalpolicyoftheopinion (andisthereforenotstrategic),itisnottodiscountitspotentialimportanceassuchcollegialcollab- 40 orationmayservetostrengthentheopinion,whichcould,inturn,downstreamtreatment oftheprecedentsetinthecase. Unlikerequestsforminorrevisions,legalchangesaresubstantiverequeststhatwouldalter thelegalimplicationsoftheCourt'smajorityopinion.ComprisingacriticalmassŒ45percent Œofallrequests,thisisthemostoft-usedbargainingchip.Thesetypesofrequestsoftenentail anappealtoprecedent:requestingthataparticularprecedent-settingcasebeaddedorremoved, questioninghowtheopinionauthorinterpretstheprecedent,orpushingtheauthortoreconsider howprecedentisusedasalegalbasistojustifypointsintheopinion.Otherlegalrequestsmight appealtodecisionsmadeinthelowercourt,sometimesdrawingexplicitlyfromthedistrictor circuitcourtopinions.Inshort,thesetypesofrequestswouldaffectthelegalimpactofthecase. Consider,again,thecaseof Gilmorev.CityofMontgomery .UnlikeRehnquist'srequests forminorrevisions,Brennanissuesasixpagemissivethatincludesnumerousconcretelegalre- quests.Heargues,forexample,thatBlackmunfimightproperlyfollowthereasoningflof Greenv. CountySchoolBoardofNewKentCounty ,adejuresegregationcasedecidedbytheCourtjustsix yearsprior.Similarly,JusticeWhitequestionsBlackmun'srelianceon MooseLodgeNo.107v. Irvis ascontrollingprecedentinthiscase,arguing,fiIdon'tthink Moose Lodgenecessarilycon- trols.Thequestionthereconcernedthesofthegrantofalicenseaspartofaregulatory program.Nopublicsubsidywasinvolved[...].flEachoftheserequestschallengedBlackmunto reconsidertheinterpretationandrelevanceofprecedentandcarriesnotablelegal Duetothepreeminenceoftheattitudinalmodelinjudicialdecision-making,itmightcome asasurprisethatrequestsforlegalchangefaroutweighthenumberofpolicyrequests.Yet,this isfullyinlinewithrecentworkthatprovidesevidencethatjusticesarenotsolelypolicy-driven actors,butareinsteadfrequently(andsometimescritically)constrainedbylaw(see,e.g.,Bailey andMaltzman2008;HansfordandSpriggs2006;RichardsandKritzer2002).Takenatfacevalue, thesedataseemtosuggestthatSupremeCourtjusticesactuallydocaretofigetitright,flandfre- quentlyprofferlegalrequestsandasktheopinionauthortoconsideralternativetreatmentsofthe law. 41 Figure2.1Barplotsofbargainingfrequencyandtypebyjustice Barplotsofthebargainingfrequencybyrespondingjustice(top)andtheproportion ofrequestsbytype(bottom). Finally,theremaining10percentofrequestsasktheauthortosubstantivelychangethepol- icycontentoftheCourt'sopinion.Thisrequesttypeisqualitativelythemostdiverse.Thesetypes ofrequestsmightasktheopinionauthortoconsideradifferenthypotheticalrelatedtothelegal question,tochangethescopeoftheopinion(i.e.,toestablishlimitsorbounds),ortosomehow reconsiderthedispositionofthecase(i.e.,askingthatthecasedbedismissedasimprovidently granted(DIG),vacatedandremanded,orsummarilydecidedinlieuofafull-blownopinion).Fi- nally,theserequestsmightraisequestionsabouttheinterpretationofactuallegislationorstatute 42 athand.InadditiontohisrequestthatBlackmunreconsiderthelegalprecedentin Gilmore ,White suggestsanalternativedispositiontothecase,notingthatwhileheisfiagreeabletotheremandfl, hewouldfiprefertoaftheDistrictCourtfl.Policyrequeststendtobeideologicalbynatureand thusheightenthestakesinthenegotiationprocess.In Gilmore ,Blackmunchosenottoaccommo- dateWhite'srequestsand,inturn,Whiteauthoredaseparateopiniontoexpresshisviews. Interestingly,whiletheaggregatedatapaintapictureoftheoverallbargainingbehaviorof thejustices,Figure2.1illustratesthatdisaggregatingrequestsbyindividualjusticesrevealssizable variationbothinthesheervolumeofrequestscirculatedbyeachjusticeandinthecomposition ofthoseinquiries.Thoughthedistributionofrequesttypelargelyfollowstheaggregatepatternof theCourtasawhole,JusticePowellisthemostbargainer,followedcloselybyBlackmun, whilethemajorityoftheotherjusticesparticipateatnearlyhalftherate.Attheotherendofthe spectrum,JusticeMarshallrarelyengagedinthebargainingprocessand,whenhechosetodoso, optedmorefrequentlytorequestminoreditorialrevisions. 5 2.5.1.2HowJustices(attemptto)PersuadetheOpinionAuthortoChange Justastherearedifferencesinthetypesofrequestsmadetotheopinionauthor,sotooarethere observabledifferencesacrossjusticesinthetacticsusedinnegotiationwiththeopinionauthor. Intheirworksonintra-courtbargaining,Maltzman,SpriggsandWahlbeck(2000)andWahlbeck, SpriggsandMaltzman(1998)arguethatSupremeCourtjusticesaremoreapttocomplywith colleaguesdemandsandsuggestionsiftherequestcarriesanassociatedthreattothestabilityof themajoritycoalition.However,becausethisworkimplementsonlyproxymeasuresforaccom- modation,itispossiblethatthesepreviouseitheroverstateorunderestimatethestrategic advantageofthreateningtheopinionauthor. 5 NotethatJusticeDouglaslefttheCourtandwasreplacedbyJusticeStevensinthe1975 termand,thus,onlyappearsinfourofthe15termsconsideredhere.Similarly,JusticeO'Connor replacedoutgoingJusticeStewartin1981andappearsonlyinthelastfewterms.Thishelps accountforthelesserfrequencyofrequestsreportedinthisanalysis. 43 Followingtheestablishedintuition,Iadoptasimilar,thoughslightlymorenuanced,mea- suretoaccountforthe tone oftherequestsandspeaktothetoolsofpersuasionthejusticesemploy throughouttheopinionwritingprocess.Thoughitcloselymirrorstheapproachdeployedinthe literature,ratherthanassesstheoveralltoneofeachbargainingmemo,Imeasurethetoneatthe request-level,whichallowsabargainertoutilizevariousmeansofpersuasioninasingledocument. Inparticular,Imeasuretonebydistinguishingbetweenfriendlyorneutralsuggestionsandmore ominousthreats.Icodedtoneonathreepointscale,fromnon-threateningtothemostseveretype ofthreat. 6 Iconsiderthetonetobe friendly orneutralifthejusticemakesarequestbutalsojoins theauthor'sdraftofthemajorityopinion.Ina1973memotoChiefJusticeBurger,theassigned authorin Levittv.CommitteeforPublicEducation ,JusticePowelljoinsthemajorityopinion,but entreatsBurgertomodifythecontent.Hewrites:fiPleasejoinmeinyouropinion.Ihavenoted thecommentssuggestingdeletion[...and]Iwouldbequitecontenttoseethisomitted,although- certainlyfrommyviewpoint-Iagreewiththesubstanceoftheviewyouexpress.flPowellgoeson tosuggestthatifBurgerwishestokeepthesection,hemaywishtoreviseittoclarifytheCourt's position.Powellgoessofarastore-writetheparagraphforBurgertoconsider.Yetitisclearthat evenifBurgerelectsnottoaccommodatePowell'srequests,Powellwilljointhemajority. NotonlydidtheChieffaceasuggestionwithunequivocalsupportfromPowellinthis case,butBurgeralsoreceivedmemorandawithvaryingdegreesofhostilityashisothercolleagues soughtchanges.Forexample,JusticeBrennanthoughtthattheopinionwenttoofarinitsholding, whichcouldberemediedbytheremovalofasinglesentence.Hewrites:fiThus,Iwouldmuch preferachangeintheabove-quotedsentencesoastoeliminatethissuggestion.Withsuchan alteration,Iwouldbehappytojoinyouropinion.flInsuchan ambiguous request,Brennanmakes itclearthathewillonlyjointhemajorityifBurgeraccommodateshisrequest,buthefailstonote theconsequenceofnon-accommodation. ThisisinstarkcontrasttoJusticeDouglas'smemorandainthesamecase,whichcarried 6 Ialsouseda5-pointscaleandtheresultsarerobusttothisalternative 44 anexplicit threat ofconsequenceifhisdemandswentunanswered.Inparticular,Douglassought topersuadeBurgerbyemployingathreattothecohesivenessofthemajoritycoalition,writing, fiIwouldprefertokeepourholdingsasnarrowaspossible.If,however,youkeeptheopinion[] initspresentform,pleasenoteattheendofeach:Mr.JusticeDouglasconcursintheresult.fl Inshort,thetoneissaidtobeathreatifthebargainingjusticeinanywaythattheshe maytakefurtheractionifherrequestisnotaccommodatedbythemajorityopinionauthor.Most frequently,threatsindicatethatthebargainingjusticemaycirculateaconcurrence,joinsomeone else'sopinion,orotherwisesignalsthatherfivoteflforthemajorityopinionisn'tastableorsecure one. Finally,becausethesedataarecodedattherequestlevelratherthanthatoftheentire memoranda,thereareinstanceswhenabargainingjusticeswitchestoneŒaskingforaminor change,forexample,verynicely,butthendemandingalegalchangewithmoreforce. Figure2.2Stackedbarplotsofrequestcontentbytone. AsshowninFigure2.2,thejusticesarefarmorelikelytoadoptafriendlyorneutraltone whenappealingtotheopinionauthor.Infact,nearlytwothirdsofallrequestsŒ65percentŒare ofacollegialnature.Justover20percentofappealsaddressedtotheauthoremployanambiguous threat,whiletheremaining15percentinvokeanexplicitthreattothestabilityorcohesivenessof 45 Figure2.3Barplotoftherelativefrequencyofthetonebargainingjusticesemploy. themajoritycoalition.Despitethepropensityforjusticestoremainaffable,weseethatjustices aremoreinclinedtoinvokeahostile,explicitlythreateningtonewhenthatjusticerequestsale- galorpolicychangeintheopinion.Weobservethat22percentofrequestsforalegalchange areaccompaniedbyanexplicitthreattothemajoritycoalition.Similarly,whenajusticerequests apolicychange,sheismorelikelytoindicatethathervoteisupforgrabs.Nearly25percent ofpolicyrequestsareframedasexplicitthreats.Aspreviouslynoted,thesetendtobethemost contentiousrequests;still,however,roughlythreequartersofthesepolicyrequestsstrikeanam- biguousorfriendlytone.Unsurprisingly,inonlytherarestofcasesŒlessthanvepercentŒdo threatsaccompanyrequestsforminorrevisions. 7 Despitethefactthattherequestsforaccommodationareoverwhelminglypositiveintone, Figure2.3illustratesthatsomejusticesŒlikeJusticesStevensandRehnquistŒaremoreproneto strikeathreateningtonethanothers.Forexample,whileStevensusesaheavyhandandattaches 7 Anythreatthataccompaniesaminorrequestis,ofcourse,unexpected.However,theseoddi- tiesoccurwhenarequestforaminorrevisionaccompaniesarequestforapolicychangeandthe entiretoneofthebargainingmemoishostile.Insuchacase,thebargainerindicatesthatshewill onlyjointheopinionifallofherdemandsaremet. 46 eitheranambiguousorexplicitthreattoroughly46percentofhisrequests,infrequentbargainer JusticeMarshallsitsattheotherextreme,apparentlymorereticenttothreatenthestabilityofthe majoritycoalition,deployingthistacticinonly12percentofhisrequests.Morefrequentbargain- erslikeJusticeBlackmunsitsomewhereinthemiddle,deployingahostiletoneinapproximately 25percentoftheirrequests. 2.5.1.3WhenOpinionAuthorsAccommodate:ADirectMeasure Despitetheirpathbreakingworkinthisarea,currentscholarshipoffersonlyanindirectmeasureof accommodation(Maltzman,SpriggsandWahlbeck2000;Wahlbeck,SpriggsandMaltzman1998). Byaccountingforthesheernumberofdraftscirculatedortheamountoftimethathaslapsed,such measurementtechniquesnecessarilyfailtocapturewhichrequestsareactuallyincorporatedinto theopinion.Atthemostbasiclevel,theseproxymeasuresmisstheiterativeandinteractiveprocess ofthemajorityopiniondevelopmentasmultiplerequestscanbesenttotheauthorbetweendrafts. Additionally,theauthormustbalanceadesiretoaccommodateonememberwiththepreferences oftheothermembersofthemajoritycoalition. Take,forexample,themajorityopinionissuedfor UnitedStatesv.Generes (1971),arather technicalfederalincometaxrefundsuitthataskedtheCourttodeterminetheappropriatestandard toapplytothemotivationofthetaxpayer,ashareholderandemployeeofthecompany,atthe timeabaddebtisincurred.Aftercirculatingadraftofthemajorityopinion,JusticeBrennan activelytriedtobargainwithJusticeBlackmun,whohadbeenassignedtheopinion.According tothememoBrennancirculatedtotheentireconference,heindicatedhiswillingnesstojointhe majorityopinion,agreeingthattheopiniontheappropriatestandard,butthathewould preferthesectionoftheopinionbeabandonedinfavorofremandingthecasetothelower courtforaretrial.JusticeWhitesecondedBrennan'ssuggestionandthreatenedaconcurrenceif Blackmundidnotremandthecaseforretrial.Insubsequentcirculationsofthemajorityopinion, Blackmunstruggledtomaintainacohesivemajoritycoalition,notingcontinualdiscordfromfrom bothWhiteandBrennan.Inprivatememorandum,Blackmunevenindicatedawillingnessto 47 followBrennan'ssuggestion.However,JusticesStewart,Burger,andMarshallpreferredthetext intheoriginalopinionanddidnotsupportthemoveforaretrial.Tomaintaincontrolofthe opinion,then,Blackmundecided not toaccommodateBrennanandWhite,favoringtoholdthe votesoftheothermembersofthemajoritycoalition.Asaresult,Whitepublishedaconcurrence, whichBrennanjoined. Generes highlightsseveralimportantfeaturesabouttherolecollegialandinstitutionalfac- torsplayinshapingthecontentcontainedinthemajorityopinion.ally,itdemonstratesthe emergenceofaconcurrencewhentheopinionauthorrefusedtoaccommodatetherequestsofhis colleagues.Moreover,theconcurrencefistoleflavotefromthemajoritycoalition.WhileBrennan saidhepreferredaretrial,heindicatedhisagreementwiththemajorityopinionuntilaconcurrence wascirculated.Inthiscase,becausehehadthesupportofapluralityofmembersonthemajority coalition,thepublicationofaconcurrenceactuallyallowedBlackmuntokeepthemajorityopinion closertohisownidealpolicypreference. Itispreciselythistypeofnuancethata direct measureofaccommodationisabletocap- ture.Tocreatesuchameasure,therearetwonecessarysteps.Theistodeterminethechanges ineachcirculateddraft.Todothis,IfollowCorley(2008),Corley,CollinsandCalvin(2011) andBlackandOwens(2012)andutilizeWCopyFind,afreeplagiarismsoftwarethatdetermines theoverlapbetweentwoormoredocuments.Thismethodallowsmetoidentifythelocationand magnitudeofthechangesfromtheinitialdrafttoeachsubsequentcirculationoftheopinion.How- ever,simplydeterminingthelocationandcalculatingthemagnitudeofthechangeisinsuf tomeasurehowtheopinionmovesduetoexternalThesecondstep,then,istoconnect thechangeintheopinionwiththeexplicitrequestscontainedinmemorandafromtheotherjus- ticesbymergingtheopiniondraftdatawithinformationfromtheBurgerCourtOpinionWriting Database.Thisapproachcreatesadyad(s)betweentheopinionauthorandeachrespondingjustice attherequestlevelforeachcase,allowingforthecreationofauniquebinarydependentvariable. Figure2.4containsdescriptivedataaboutthevolumeofbargainingactivityfacedbyeach authorandthesubsequentrateofaccommodationbyopinionauthor.Whatisreadilyapparentis 48 thatsomejustices,suchasBurgerandPowell,encounteredmorebargainingactivity thanJusticesStewartandDouglas,forinstance.Whilethisdisparitymayseempuzzling,there severalpotentialexplanationsforthispattern.First,thedatapresentedherearelimitedtoOctober Terms1971-1985.Thefourjusticesthatwerefacedwiththelowestoveralllevelsofbargainingall appearinthedataforalimitednumberofterms. 8 Second,thoughheauthoredmorethanhisshare oftheCourt'sopinions(12percentduringtheperiod),ChiefJusticeBurgerwasamemberofthe majoritycoalitioninthevastmajorityofcases,andthuslyresponsibleformorethan86percentof allmajorityopinionassignments.LikeBurger,JusticePowellauthorednearly12percentofthe Court'sopinionsinthetermsunderconsiderationhere,whichmayagainbegintoaddresswhywe seemorerawbargainingactivity.Whilethedescriptivedatabegintotellastory,itisn'tpossible tosystematicallyassessthedifferencesweobserve. Figure2.4Barplotsofthenumberofrequestsbyopinionauthorsandjus- tices'accommodationrates Barplotsofthetotalnumberofrequestsbyindividualmajorityopinion authors(left)andtherateatwhicheachjusticeaccommodatesrequests(right). Despitethequestionsraisedbythedifferentlevelsofbargainingpresentforthevarious opinionauthors,themostfascinatingconclusionfromFigure2.4isthatthereisnorealdiscernible 8 RecallthatJusticeStevensDouglas'seatin1975andO'ConnorreplacedStewartin 1981. 49 ideologicalorinstitutionalpatternintheratesofaccommodation.Therateofaccommodationby theChiefJustice,forexample,isvirtuallyindistinguishablefromthatofJusticeMarshall,who representedtheoppositeideologicalextremeandwasonlyinfrequentlyamemberofthemajority coalition.ItistemptingtospeculatethattheCourt'sideologicalmediansŒnamelyBlackmun (themedianfortwooftheterms),Stewart(twotermmedian),andWhite(theCourt'smost frequentmedian,occupyingthepositionineightterms)Œarelesslikelytoaccommodatethanare theirnon-mediancounterparts.ButJusticePowell,theCourt'smedianinthreeterms,doesnot thepattern.Inshort,thedescriptiverateofaccommodationraisemorequestionsthanitanswers. Figure2.5Accommodateratebyauthor-bargainingjusticedyad Plotofaccommodationratebyauthor-bargainingjusticedyad.Darkersquaresin- dicatehigherratesofaccommodation.Onbothaxes,thejusticesareideologically orderedfromleasttomostconservative. Althoughtheaggregateratesofaccommodationfailtoprovidenoteworthyvariationsor patternsinaccommodation,Figure2.5illustratesacloserlookattheindividual-levelauthor- bargainerdyads,revealingintriguingdescriptiverelationships.Particularly,itappearsthatthere 50 arenotsystematicpatternsofaccommodationbetweenjustices.Takeforexample,JusticesBren- nanandRehnquist,ideologicalopponentsanchoringeitherpoleofthespectrum.Brennanand Rehnquistdisplayverydifferentaccommodationbehavior.Inparticular,whenheiswritesfor themajority,Brennanisespeciallyamenabletorequestsmadebyhisideologicalallies,incor- poratingallofsuggestionsprofferedbystalwartliberalJusticesDouglasandMarshall,butheis leastlikelytoadoptrequestsmadebyideologicalopponents.AstheCourt'smostconservative member,Rehnquist,ontheotherhand,iskeentoaccommodaterequestsmadebyideologicalop- ponentsandappearstoworktoappeasethemedianmembersoftheCourtaswell.Heislesslikely, however,toadoptchangesmadebyhisclosestideologicalally.Whenheisthemajorityopinion author,Rehnquistdeniesapproximately75percentofthechangesChiefJusticeBurgerrequests. ThemoremoderateJusticeStewartofferssimilarlypuzzlingratesofaccommodation;heisespe- ciallyresponsivetorequestsmadebymembersatbothendsoftheideologicalspectrum,butisless amenabletorequestsbyothermoderates,likeJusticeWhite.ChiefJusticeBurgerappearstobe themosteven-handed;hisaccommodationratesvarylittleacrosstheboard. 2.5.2MeasurementofExplanatoryVariables Thekeyinterveningvariableofinterestisthebargainingjustice'spositionontheCourt.Asde- velopedintheprecedingsection,theorypointstoseveralmembersoftheCourtthatmightwield inordinateintheshapeofthemajorityopinion.Iftrue,weshouldexpectthatthesejus- ticesaremoreinbargainingwiththeopinionauthor.Conversely,iftheauthormaintains controlofthemajorityopinion,theauthor'srelativepositionontheCourtshouldaffectwhenac- commodationoccurs.ThesetofmodelspredictthemedianmemberoftheCourttobemost powerful(Hammond,BonneauandSheehan2005;Martin,QuinnandEpstein2005).Following standardpracticebyusingtheMartin-Quinndynamicideologyscores,adichotomousvariable, BargainerisCourtMedian iscodedas1ifarespondingjusticeisthemedianoftheCourtand 51 0otherwise(MartinandQuinn2002). 9 Morerecentscholarshipsuggeststhemedianofthema- joritycoalitiondeterminesthecontentofthemajorityopinion(Carrubbaetal.2012;Clarkand Lauderdale2010).Afterexaminingtheinitialcoalitionsestablishedattheconferencevote,Icode BargaineristheMedianoftheMajorityCoalition(MoMC) as1whenaparticularjusticeoccu- piesthispositionand0otherwise. 10 Becausethereistheoreticaldisagreementaboutthedegreeof controlthatrestsinthehandsoftheopinionauthorherself,usingthesamemeasurementstrategy, Iincludedichotomousvariablestocapture AuthorisCourtMedian and AuthorisMoMC . Toaccountfortheubiquitousofideology,Iincludetwoindicators. 11 The measuresthe ideologicaldistance betweenthebargainerandopinionauthor,whichissimplythe absolutevalueofthedifferencebetweenthejustices'Martin-Quinnideologyscores(Martinand Quinn2002).Toprovideasenseofideologicaldispersionwithinthewinningmajoritycoalition,I provideameasureof coalitionheterogeneity wheregreaterscoresindicategreatervariancewithin themajoritycoalition. Totapintohowreciprocityor cooperation affectsthenegotiationprocess,Iusethemeasure developedbyMaltzman,SpriggsandWahlbeck(2000,79),whichcapturesfithepercentageoftime thattheauthorjoinedaseparateopinionwrittenbythejusticeinthepreviousterm.fl Becausenotallcasesarecreatedequalandwemayreasonablyexpectthatjusticescare moreaboutsomecasesthanothers,Iimplementseveralmeasurestoproxyjusticeengagement. Thetwomeasurecasesalience.Tocapture politicalsalience ,Ideriveaveragesof 9 JusticeWhiteheldthemedianpositionmostfrequently,occupyingtheseatfromOT1971-174 andagainfromOT1980-1983.JusticeStewartwastheCourt'smedianinOT1975and1976before Blackmuntookthecenterpositionin1977and1978.Powelloccupiedthemedianpositioninthe remainingOctoberTerms:1979,1984,and1985. 10 Otherscholarssuggestthatthemedianmembershiftsbasedonthelegalissue(Anderson andTahk2007).Becausenoideologyscorescurrentlyexistattheissue-level,itisnotcurrently possibletodirectlytesttheimplicationsfromthemodel. 11 Themeasurementofmanyofthecollegialandinstitutionalfactorscomelargelyfromthe variousworksfromMaltzman,Spriggs,andWahlbeck(1998,1999,2000).Theauthorsgenerously providedreplicationdatafortheirstudiesandIwasabletointhegapswheretheirdatafellshort. Asaresult,mymeasurementschemedrawsdirectlyontheonesemployedinthesestudies. 52 amicuscuriaeparticipation.Foreachcaseaz-scoremeasuresthevolumeofamicusparticipation relativetotheaverageofallcasesinthatterm. Legalsalience ,ontheotherhand,isadummy variablethatcaseswhereprecedentisoverturnedorsubstantiallyaltered(Spaeth1994).At theindividuallevel,IemploytheMaltzman,Spriggs,andWahlbeckopinionratiobyissueareato accountforjustice expertise .Totapinto casecomplexity ,IagainfollowWahlbeck,Spriggsand Maltzman(1998)andcombinethreeindicatorsthattogethersignalthatmultiplelegalissuesareat playinacase.DerivedfromSpaeth(1994),theresultantsinglefactorwithaneigenvaluegreater thanoneisgeneratedbythenumberofissuesraised,thenumberoflegalprovisionsrelevanttothe case,andthepost-hocnumberofopinionsthatacasegenerates. Todetermineinstitutionalandpracticalconstraints, winningmargin ismeasuredasthe absolutevalueofvotesneededtomaintainthemajoritycoalitionandtheactualnumberofvotes received, workload accountsforthenumberofopenopinionassignmentsforeachjustice,for end- of-term denotesthenumberofdaysleftintheCourt'stypicalterm(i.e.,thenumberofdaysuntil thelastdayofJune),and freshmanauthor isadichotomousmeasurethatcapturesjusticeswho haveservedlessthantwofullyearsontheCourt. 12 2.6Results Theopinionauthor'sdecisiontoaccommodateisabinarydependentvariable,soIestimatealo- gisticregressionmodel.Recallthattheunitofanalysisisattherequestlevel.Thus,foreachcase opinion,therearemultipledyads,representingeveryrequestandresponseforeachbargainingjus- ticeandthecorrespondingopinionauthor.Ifajusticeengagesinmultipleroundsofbargaining withtheopinionauthor,orifthejusticesendsalengthymemototheauthorcontainingmultiple requests,thesamedyadappearsmultipletimeswithinthegivencase.Becausetheindependence 12 Notethattherewereonlytworeplacementsinthedataset.JusticeStevenswasafreshman justiceinthe1975and1976termsandJusticeO'Connorservedasafreshmanin1981and1982. 53 assumptionisviolated,robuststandarderrorsareclusteredatthecaselevel. 13 Theresultsofthe modeloffersupportforbothmodelsofmedianmemberopinionauthorsofallideologi- calstripesaresystematicallymoreresponsivetorequestsfromthemedianmemberofthemajority coalition,buttheidentityoftheauthormatterstoo.Ithatwhentheauthoristhemedianofthe entireCourt,sheislesslikelytoaltertheopinion.Moreimportantly,however,Ithatthecon- tentandtoneofthenegotiationarepowerfulpredictorsofhowtheopinionchangesthroughoutthe negotiationprocess.TheparameterestimatesforthismodelarefoundinTable2.2andtheaverage marginaleffectsforallstatisticallyinterveningvariablesareillustratedinFigure2.6. AsshowninFigure2.7,holdingallothervariablesattheirmeanormodalvalues,I thatwhenarevisionrequestisprofferedbyanon-coalitionmedianjustice,thereisanestimateda 0.70probabilitythatthejusticeissuccessfullyaccommodated.Thatprobabilityjumpsto0.77Œa 10percentincreaseŒwhenthemedianofthemajoritycoalitionasksforasubstantivechangein themajorityopinion. Yetthemedianmemberofthemajoritycoalitiontheoryisnottheonlytheoryofopinion controltogaintractionhere.Indeed,theresultssuggestthatthepositionofthemajorityopinion authorremainsrelevant.Inparticular,whentheassignedauthoristhemedianoftheentireCourt, thelikelihoodarequestismetwithacquiescencedeclinesbyroughly13%fromanestimated0.71 probabilitytoanestimated0.62probability. Contrarytomyexpectation,howmuchŒthesheermagnitudeofchangerequiredŒisre- questeddoesnotsignireduceanauthor'swillingnesstoadoptchangesrequestedbyher colleagues.Thecontentofnegotiation,ontheotherhand,anauthor'swill- ingnesstoalterthemajorityopinion.Theevidencesuggeststhatwhile who makestherequestand towhom therequestissentisindicativeofthelikelihoodofaccommodation,the what and how matter,too.Inparticular,allelseequal,comparedtorequestsforlegalchangesintheopinion (theomittedcategory)requestsforminorrevisionsincreasestheprobabilityofaccommodation 13 Theresultsarerobusttovariousspeciincludingclusteringbybargainingorauthoring justice. 54 Constant0 : 808 ( 0 : 209 ) BargainerisMoMC0 : 344 ( 0 : 169 ) BargainerisCourtMedian 0 : 054 ( 0 : 101 ) AuthorisMoMC0 : 106 ( 0 : 091 ) AuthorisCourtMedian 0 : 428 ( 0 : 180 ) MinorRevisionRequest1 : 262 ( 0 : 115 ) PolicyChangeRequest 0 : 284 ( 0 : 071 ) NumberofRequests0 : 072 ( 0 : 065 ) FriendlyTone 0 : 057 ( 0 : 160 ) ThreateningTone1 : 102 ( 0 : 098 ) IdeologicalDistance0 : 005 ( 0 : 003 ) CoalitionHeterogeneity 0 : 004 ( 0 : 005 ) Cooperation0 : 513 ( 0 : 471 ) PoliticalSalience 0 : 033 ( 0 : 011 ) LegalSalience 0 : 535 ( 0 : 070 ) Expertise0 : 039 ( 0 : 043 ) CaseComplexity0 : 129 ( 0 : 061 ) WinningMargin 0 : 098 ( 0 : 023 ) Workload 0 : 023 ( 0 : 009 ) End-of-Term0 : 000 ( 0 : 000 ) FreshmanAuthor0 : 075 ( 0 : 148 ) Observations3 ; 949 log L 2346 : 64 Table2.2Logitmodelofmajorityopinionauthor'sdecisiontoaccommodate Logitmodelofmajorityopinionauthor'sdecisiontoaccommodate.Robuststandarderrors clusteredatthecaselevelarereportedinparentheses.*denotes p < 0 : 05(two-tailedtest). 55 Figure2.6Averagemarginaleffectsforstatisticallyvariablesonthe probabilityofaccommodation Averagemarginaleffectsforallstatisticallyvariablesontheprobability themajorityopinionauthoraccommodateshercolleaguesandmakeschangesto opiniondraft.Baselinepredictedprobabilityis0.720.Whiskersrepresent95% intervals.Allvariablesheldatmeanormedianvalues,asappropriate. by0.26Œfromapredicted0.55probabilitytoa0.81probabilityŒandthelikelihoodofaccom- modationleapstomorethanafourinvechancethattheopinionauthoradoptsthesuggestion ofhercolleague.Alternately,asexpected,theresultsindicatethatrequestsforpolicychangeare indeedmorecontroversialthanarerequestsforlegalchanges.Whenabargainerlobbiestheau- thortomakesubstantivepolicychanges,theprobabilityofaccommodationdropsfromapredicted probabilityof0.71to0.65,nearlya10percentdecrease. Thetoneorframeoftherequestalsoaffectsthelikelihoodofaccommodation.Compared toambiguousorneutralrequests,theomittedbaselinecategory,friendlysuggestionsarenotstatis- ticallydifferent.Butwhenabargainingjusticethreatensthestabilityofthemajoritycoalition,the probabilityofaccommodationincreasesfromapredictedbaselineprobabilityof0.731toa0.89 56 Figure2.7Predictedprobabilityanauthoraccommodateswhenthebargaineris theMoMCorwhentheauthoristheCourtmedian Predictedprobabilityanopinionauthoradoptsachangeatthebehestofhercol- league,conditionalonwhetherthebargaineristhemedianofthemajoritycoalition (leftpanel)orwhethertheauthorholdsthemedianpositionoftheCourt(leftpanel). Theblackcircleisthepointestimateandthewhiskersrepresentthe95%simulated intervalaroundthatestimate.Allothervariablesareheldattheirmean ormodalvalues,asappropriate. probability,resultinginnearlya90percentchancethattherequestwillbeaccommodated.That is,theresultsindicatethatitmaybeforbargainerstoemploytheuseofthreatsinpar- ticularlycompellingrequestsassuchthreatsresultinabouta22percentincreaseinthelikelihood ofaccommodation. Thecaseindicatorsthatachievestatisticalinthemodellargelysupport expectationsderivedfromtheliterature,thoughseveralvariablesfailtoachievestatisticalsignif- icancewhileotherschallengetheresultsfrompriorapproachesthatstudytheauthor'saccommo- dationdecision.Forexample,unliketheresultspresentedbyMaltzman,SpriggsandWahlbeck (2000),Ithatcasecomplexitydoesnottheauthor'sdecisiontoadopt thoserequests.Conversely,salienceŒbothpoliticalandlegalŒaffectsthelikelihoodanauthor acceptsthechangessoughtbycolleagues,butintheoppositedirection.Usingtimeasaproxyfor accommodation,Maltzman,SpriggsandWahlbeck(2000)arguethatauthorsappeartobemore willingtoaccommodatetheircolleaguesinsalientcases.Itheoppositerelationshipexists; 57 Figure2.8Predictedprobabilityofaccommodationcontingentonthedegreeof amicusparticipation Predictedprobabilityofaccommodationcontingentonthedegreeofamicuspar- ticipation.Thickblacklinerepresentspointestimatesandtheshadedgrayregion depictsthe95%simulateddenceintervalaroundthatestimate.Allothervari- ablesareheldandtheirmeanormodalvalues,asappropriate. Figure2.8illustratesthatwhenpoliticalsalienceisheldtoitsaveragelevelofamicusparticipation, Iestimatea0.71predictedprobabilitythattheauthoraccommodatesarequest.Thisfallstoanes- timated0.64probabilitywhenmoreamicusbriefsare 14 Legalsaliencehasasimilareffect onthelikelihoodarequestisincorporatedinamajorityopiniondraft.Theestimatedprobability ofsuccessfulaccommodationdropsbymorethan17percentfroma0.72to0.60,whentheCourt overturnsorsubstantiallyaltersprecedent.Together,thesevariablessuggestthatauthorsarefar lesslikelytoaltertheopinioninimportantcases. Othercontextualfactorsbehaveasexpectedandshapetheprobabilitythatanopinionau- thoradaptsthemajorityopiniontotherequestsofhercoalitionmembers.Ataunanimous ornear-unanimousdecision,determinedduringtheconferenceheldfollowingoralargument,the 14 Theupperboundofthevolumeofamicusbriefisactually50morebriefsthanthe average.However,99percentofallcasesfallwithintherangesimulatedinFigure2.8. 58 authoraccommodatesata0.66predictedprobability.Whendispositionalvoteoutcomeisunclear andtheauthorneedstoattractatleastoneadditionalvotetosecureawinningmajority,thepre- dictedprobabilityincreasestomorethan0.77,arelativechangeof18percent.Inthemostunclear ofcases,wherenomajorityhasbeenestablishedandanauthorattemptstorecruitcolleaguesto establishamajoritycoalitionaftertheconferencevote,theabargainerhasevengreaterchancesof success:Iestimatea0.81probabilitythatanauthoraccommodatesarequestinsuchacase. Theauthor'sworkload,too,hasasizableimpactontheopinionwriter'swillingnesstodraft anewopiniontoincorporatethesuggestionsfromtheotherjustices.Whenanauthorhasnothing elseinthepipelineandisworkingentirelyonthemajorityopinionathand,Iestimatea0.74 probabilitythatarequestwillbeadopted.Yetasanauthor'sworkloadincreases,theprobabilityof accommodationdecrease.Atthesamplemaximum,whenanauthoringjusticeisjuggling20open opinionsofanykind,thepredictedprobabilitydropsto0.64,abouta15percentdecrease. 2.7DiscussionandConclusion WhileimportantinroadshavebeenmadeintoopeningtheblackboxoftheCourt'sopinionwriting processinthepasttwodecades,theoreticalquestionsaboutthebargainingandnegotiationprocess betweenSupremeCourtjusticesremain.Todate,scholarshaveprovidedonlyindirectevidence thatopinionauthorsfunctionasstrategicactors,selectivelychoosingwhentoaccommodatethe preferencesofhercolleagues.Thisanalysispicksupwheretheoriginalexaminationleftoffby usingoriginaldatatoanalyzetheactualcontentofthebargainingmemosandtodirectlyconnect theserequeststochangesmadeincirculatedmajorityopiniondrafts.Throughthecreationofanew measureofaccommodation,wearebetterabletodeterminewhoseviewsareincorporatedintothe Court'sopinionandtounderstandthecontentofdebateornegotiationbetweenthejustices. Formaltheoriesofopinioncontrolseektopredicttheideologicallocationofthe Court'smajorityopinionbasedonthestrategicallyadvantageouspositionofcertainmembersof thenation'shighestbench.Thischapterdoesnotattempttoplacetheopinioninideologicalspace, 59 butinsteadborrowsinsightsfromthosetheoriestounderstandwhenandwhycertainjusticesought tobemoreinshapinglaw.Isupportfortwoseparatemodelsthatpositcompeting hypotheses:onethatsuggeststhemedianoftheCourtwillretaincontrolandtheotherthatcon- tendsthattheopinionwillthepositionofthemedianofthemajoritycoalition.Theresults suggestthatbothmediansareplayersinthecollegialgame.Inparticular,asbargainers, medianmembersofthemajoritycoalitionarethemostsuccessfulinseeingtheirviewstranslated intolaw,whiletheCourt'smedianmember,astheopinionauthor,islesslikelytonegotiateover opinioncontent. Bydissectingtheauthor-bargainerdyad,explicitlyaccountingfortheauthor'sdecisionto alterthemajorityopinionattherequestoftheothermembersoftheCourt,thischapterprovidesa bridgebetweenthetheoreticalandempiricalaccountsoftheopinionwritingprocessand onthecollegialCourt.Thoughthefiswingjusticeflisnotanymorelikelytoseeherviews translatedintolawwhenshebargainswiththeopinionauthor,whensheistheassignedmajority opinionauthor,theevidencesuggeststhismedianmembermayexertmorecontrolovertheopin- ioncontent.Tounderstandhowamedianauthorarrivesatthispositionofpower,Ire-estimated thestatisticalmodeltoexaminetheinteractiveeffectbetweenthemedianauthorandthesizeof thewinningcoalition.Ithat,astheauthor,themedianeffectisespeciallynoteworthyincon- tentiouscases.Whilenon-medianjusticesare more likelytoadoptchangesfromtheircolleagues toattractvotesfortheopinionwhenaclearmajorityhasnotbeenestablishedatconference,I estimatethatmedianjusticesare less willingtodothesame.Thissuggeststhatwhenthepost-oral argumentconferencevoteisunclearandthemedianmemberoftheCourtoffersamajorityopinion, shemayrecognizeherstrategicpositionastheCourt'sswingjustice,writinganopinionclosest toherownidealpointwithouttheneedtocompromise.Inotherwords,theotherjusticesarenot abletoofferviablecompetingopinionsinafracturedmajorityandthemedian'spreferencesare especiallyrelevant.Infact,theinteractioneffectrevealsthatwhenthedispositionalvoteisclear andamajority(evenasimplemajority)isestablished,themedianauthor'sbehaviorisvirtually indistinguishablefromnon-medianauthors. 60 Notably,whentheswingjusticeisnottheauthor,heroverthemajorityopinion seemstodisappear.Inotherwords,theCourt'smedianisnomorelikelytotheauthor thannon-medianjustices.Conversely,theresultsrevealthatthemedianmemberofthemajority coalitionisinaparticularlyadvantageouspositionandismorelikelytobeaccommodated,regard- lessoftheidentityofthemajorityopinionauthor.Inshort,theresultsspeaktoboththeoriesof opinioncontrol:medianmembersofthemajoritycoalitionareparticularlypowerfulindetermin- ingthecontentoftheCourt'smajorityopinion,butthemedianoftheCourtexertsgreater overtheopinionwhenthemajoritycoalitionisnotdeterminedpriortothecirculationofamajority opinion. Beyondthecreationofadirectmeasureofaccommodationthatallowsustocriticallyas- sessjusticeonthepublishedopinion,thisworkmakesseveralcontributions.Forexam- ple,thebothchallengeandsupportextantscholarshipthatsuggestanimportantrolefor institutional,interpersonal,andcontextualfactors.WhilesomeexplanatoryvariablesŒlikework- loadŒaresimilartotheresultsreportedbyMaltzman,SpriggsandWahlbeck(2000),Spriggs, MaltzmanandWahlbeck(1999),andWahlbeck,SpriggsandMaltzman(1998),thereareimpor- tantdifferences.Forinstance,thesepreviousstudiesafargreaterideologicalin theaccommodationdecisionthanthisanalysissuggests.Iwouldarguethatbyfocusingonthe numberofdraftsorthelimitedtimeframefromthetolastdraftslikelyoveremphasizedthese considerations.Ideologymayhaveservedasaproxyforjustices'positionsofpower,whichthen failstohaveanindependenteffectwhenthesefactorsareexplicitlyaccountedfor.Inshort,amore nuancedapproachtoincorporatingtheactualcontentofrequestshedsnewlightonthestrategic interactionsbetweenthejusticesduringtheopinionwritingstage. WhiletheresultsofthisanalysissupportstrategicmodelsofaccommodationŒafterall, authors are morelikelytoaccommodaterequestsfromcertainmembers,especiallyinsalientcases Œitalsorevealsthecollaborativenatureofmajorityopiniondevelopment.Boththedescriptive dataandresultsfromthestatisticalanalysishighlightthecollegialand(mostly)friendlydynamic presentbetweentheopinionauthorandhercolleagues.Theespeciallyhighratesofaccommoda- 61 tionrevealthatopinionauthorsarelargelyamenabletosuggestionsfromothermembersofthe majoritycoalition,particularlythosethatservetostrengthentheopinionthroughminorrevisions andbythetreatmentandapplicationofthelawthroughlegalsuggestions. Thedescriptivedatatellusthatjusticesaretypicallyquitefriendlywhentheyappealto theopinionauthor.Yettheempiricalresultsrevealthatajusticeisevenmorelikelytosecurethe changesshewishestoseetotheopinionifshedeploysahostiletoneandthreatenstoupsetthe stabilityofthemajoritycoalition.Thisraisesapuzzleforfutureresearch:whatmakesthreatssuch aneffectivetoolofpersuasion?Or,whathappensifabargainingjusticeisnotaccommodated? Onepersistentaccountsuggeststhatconcurrencesarefrequentlyaresultoftheauthor'sfailureto negotiatewithmembersofthemajoritycoalition,butthisassertiondemandsevaluationandisa topicIrevisitinChapter4. Despiteitscontributions,thisanalysisisnotwithoutitslimitations.BecauseIrelyexclu- sivelyondatafromtheBurgerCourtera,excludingtheearliestterms(1969-1970),thedataare limitedtoonechiefjusticewhomadethemajorityofopinionassignments.Therewasalsolittle turnoverduringthisperiod,limitingthenumberoffreshmanjustices.Astheprivatepapersof morejusticesbecomeavailable,futureworkwillspeaktothegeneralizabilityoftheresultsby extendingtotheRehnquistCourtandbeyond. Ultimately,however,thepresentedhereadvanceourunderstandingofhowjus- ticesonacollegialcourtinteracttoreachadecisionthatcommandsthevotesofthemajorityof theCourtbyillustratingforthetimewhichmembersoftheCourtaremorelikelytosucceed innegotiationswiththeopinionauthor,whichauthorsaremostpersuadable,andbyuncovering evidencethatpowerisattributednotonlytothecontentoftherequestbuttothebargainer'swill- ingnesstodisrupttheunityofthemajoritycoalitionbythreateningtocirculateorjoinaseparate opinion. 62 REFERENCES 63 REFERENCES Anderson,RobertIVandAlexanderM.Tahk.2007.fiInstitutionsandEquilibriumintheUnited StatesSupremeCourt.fl AmericanPoliticalScienceReview 101(4):811Œ825. Bailey,MichaelA.andForrestMaltzman.2008.fiDoesLegalDoctrineMatter?Unpacking LawandPolicyPreferencesontheU.S.SupremeCourt.fl AmericanPoliticalScienceReview 102(3):369Œ384. Black,Duncan.1958. TheTheoryofCommitteesandElections .London:CambridgeUniversity Press. Black,RyanC.,MaronW.SorensonandTimothyR.Johnson.2013.fiTowardsanActor-Based MeasureofSupremeCourtCaseSalience:Information-SeekingandEngagementDuringOral Arguments.fl PoliticalResearchQuarterly 66(4):803Œ817. Black,RyanC.andRyanJ.Owens.2012. TheSolicitorGeneralandtheUnitedStatesSupreme Court:ExecutiveandJudicialDecisions .NewYork:CambridgeUniversityPress. Brennan,WilliamJ.,Jr.1960.fiStateCourtDecisionsandtheSupremeCourt.fl PennsylvaniaBar AssociationQuarterly 31(June):393Œ407. Brenner,Saul.1984.fiIssueSpecializationasaVariableinOpinionAssignment.fl JournalofPoli- tics 46(4):1217Œ1225. Brenner,Saul.1995.fiFluidityinVotingontheUnitedStatesSupremeCourt:ABibliographic OverviewoftheStudies.fl LawLibraryJournal 87(Spring):380Œ386. Brenner,SaulandJanPalmer.1988.fiTheTimeTakentoWriteOpinionsasaDeterminantof OpinionAssignments.fl Judicature 88:179Œ184. Brenner,SaulandTimothyM.Hagle.1996.fiOpinionWritingandtheAcclimationEffect.fl Polit- icalBehavior 18(3):235Œ261. Bryan,AmandaandTimothyR.Johnson.2012.ThePowerofthe(Red)Pen:Bargainingand StrategicAccommodationontheU.S.SupremeCourt.AnnualMeetingoftheMidwestPolitical ScienceAssociation. Carrubba,Cliff,BarryFriedman,AndrewD.MartinandGeorgVanberg.2012.fiWhoControlsthe ContentofSupremeCourtOpinions.fl AmericanJournalofPoliticalScience 56(2):400Œ412. Clark,TomS.andBenjaminLauderdale.2010.fiLocatingSupremeCourtOpinionsinDoctrine Space.fl AmericanJournalofPoliticalScience 54(4):871Œ890. Corley,PamelaC.2008.fiTheSupremeCourtandOpinionContent:TheofParties' Briefs.fl PoliticalResearchQuarterly 61(3):468Œ478. 64 Corley,PamelaC.,PaulM.Jr.CollinsandBryanCalvin.2011.fiLowerCourtonU.S. SupremeCourtOpinionContent.fl JournalofPolitics 73(1):31Œ44. Epstein,Lee,DanielE.Hoe,GaryKingandJefferyA.Segal.2005.fiTheSupremeCourtDuring Crisis:HowWarAffectsOnlyNon-WarCases.fl NewYorkUniversityLawReview 80(1):1Œ116. Epstein,LeeandJefferyA.Segal.2000.fiMeasuringIssueSalience.fl AmericanJournalofPolitical Science 44(1):66Œ83. Fowler,JamesH.,TimothyR.Johnson,JamesF.IISpriggs,JeonSangickandPaulJ.Wahlbeck. 2007.fiNetworkAnalysisandtheLaw:MeasuringtheLegalImportanceofSupremeCourt Precedents.fl PoliticalAnalysis 15(3):324Œ346. Hagle,Timothy.1993.fiFreshmenEffectsforSupremeCourtJustices.fl AmericanJournalofPo- liticalScience 37(4):1142Œ1157. Hammond,ThomasH.,ChrisW.BonneauandReginaldS.Sheehan.2005. StrategicBehaviorand PolicyChoiceontheU.S.SupremeCourt .PaloAlto,CA:StanfordUniversityPress. Hansford,ThomasG.andIISpriggs,JamesF.2006. ThePoliticsofPrecedentontheU.S.Supreme Court .Princeton,NJ:PrincetonUniversityPress. Lax,JeffreyR.andCharlesM.Cameron.2007.fiBargainingandOpinionAssignmentontheUS SupremeCourt.fl JournalofLaw,Economics,andOrganization 23(2):276Œ302. Maltzman,Forrest,JamesF.IISpriggsandPaulJ.Wahlbeck.2000. CraftingLawontheSupreme Court .NewYork:CambridgeUniversityPress. Maltzman,ForrestandPaulJ.Wahlbeck.1996.fiStrategicPolicyConsiderationsandVotingFlu- idityontheBurgerCourt.fl AmericanPoliticalScienceReview 90(3):581Œ592. Martin,AndrewD.andKevinM.Quinn.2002.fiDynamicIdealPointEstimationviaMarkov ChainMonteCarlofortheU.S.SupremeCourt,1953-1999.fl PoliticalAnalysis 10:134Œ153. Martin,AndrewD.,KevinM.QuinnandLeeEpstein.2005.fiTheMedianJusticeontheU.S. SupremeCourt.fl NorthCarolinaLawReview 83(5):1275Œ1322. Murphy,WalterF.1964. ElementsofJudicialStrategy .Chicago:UniversityofChicagoPress. Richards,MarkJ.andHerbertM.Kritzer.2002.fiJurisprudentialRegimesinSupremeCourt DecisionMaking.fl AmericanPoliticalScienceReview 96(2):305Œ320. Rohde,David.1972.fiATheoryoftheFormationofOpinionCoalitionsintheU.S.Supreme Court.fl MidwestJournalofPoliticalScience May(208-224). Schwartz,Bernard.1993. AHistoryoftheSupremeCourt .NewYork:OxfordUniversityPress. Schwartz,EdwardP.1992.fiPolicy,Precedent,andPower:APositiveTheoryofSupremeCourt Decision-Making.fl JournalofLaw,Economics,andOrganization 8:219Œ252. 65 Segal,JefferyA.andHaroldJ.Spaeth.1993. TheSupremeCourtandtheAttitudinalModel .New York:CambridgeUniversityPress. Spaeth,HaroldJ.1994. UnitedStatesSupremeCourtJudicialDatabase,1953-1996Terms .8th ed.AnnArbor:Inter-UniversityConsortiumforPoliticalandSocialResearch. Spriggs,JamesF.II,ForrestMaltzmanandPaulJ.Wahlbeck.1999.fiBargainingontheUS SupremeCourt:Justices'ResponsestoMajorityOpinionDrafts.fl JournalofPolitics 61(2):485Œ 506. Unah,IsaacandAnge-MarieHancock.2006.fiU.S.SupremeCourtDecisionMaking,Case Salience,andtheAttitutinalModel.fl LawandPolicy 28(3):295Œ320. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.1998.fiMarshallingtheCourt: BargainingandAccommodationgontheUnitedStatesSupremeCourt.fl AmericanJournalof PoliticalScience 42(1):294Œ315. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.2009. TheBurgerCourtOpinion WritingDatabase .Washington,DC:GeorgeWashingtonUniversity,DeptofPoliticalScience. 66 CHAPTER3 (DIS)ENGAGEMENTINTHESUPREMECOURT'SOPINIONWRITINGPROCESS 3.1ChapterAbstract Thischaptertakesastepbacktoconsiderajustice'sinitialdecisiontoengageŒornotŒwith assignedopinionauthor.Itdissectsajustice'sresponsetotheinitialmajorityopiniondraftwherea justicemayopttojoinitentirely,negotiatewiththeauthortomakechanges,orwriteaconcurring opinionfromtheoutset.Iagainevidencethatthemedianmemberofthemajoritycoalition holdsanpositionincraftingtheCourt'sopinion.Theothersmembersarelesslikelyto offercompetingopinionswhensuchajusticeistheassignedauthor.Thoughtheauthor'sidentity maybargainingactivity,theresultsrevealthatthesinglegreatestpredictorofajustice's willingnesstoengagewiththeauthoristhedegreeofcooperationbetweenthejusticeandthe author.Whenajusticeandtheauthorhaveahighdegreeofcooperation,ajusticeisfarlesslikely toasktheauthortomakechangestotheCourt'sopinion.Additionally,caseandcontextualfactors playagreaterroleinapotentialbargainer'sdecision-makingcalculus.Justicesaremorelikelyto engagewiththeauthorwhenthecaseislegallycomplex,butfreshmanjusticesareleastlikelyto bargainacrosstheboard.Thispaperpushesthequestionfurthertoconsiderhowfidisengagedfl justicesreacttothemajorityopinion.IthatwhentheCourt'smedianauthorstheopinion,a justiceislesslikelytorequestchanges,butisfarmorelikelytopenaseparateconcurringopinion. Similarly,justicesaremoreapttoforgobargainingtowriteseparatelyinlegallycomplexcases andastheirideologicaldistancefromthemeanofthemajoritycoalitiongrows.Justicesaremore likelytosimplyaskforchangeswithouttakingthesteptowriteaconcurringopinionasthewinning marginincreases.Finally,Ithatjusticeswhoareexpertsinaparticularissueareaaremore likelytoengagewiththemajorityopinionauthorthanjusticeslackinginspecializedknowledge. 67 3.2Introduction In Parhamv.J.R. (1979),theUnitedStatesSupremeCourtconfrontedalegalbattlethatweighed therightsofchildrenagainstparentalauthorityinthedecisiontocommitaminorchildtoamental institutionwhileattemptingtodiscerntheinterestofthestateandoftheappropriateroleforjudicial oversight.Asisclearfromtheirprivateconferencenotesandpapers,thejusticeswerewaryabout wadingintosuchacomplexandemotionallychargedcase.Aftertheinitialpost-oralargument conference,noclearmajoritycoalitionhadbeenformedandChiefJusticeBurgerproposedthat theysetre-argumentforthefimurkyflcasethatincludedanfiinadequateopinionflfromthelower districtcourt.Whenthatsuggestionfailedtogaintraction,apreliminaryvotegeneratedaninitial winningcoalitionandBurgerassignedtheCourt'smajorityopiniontohimself. Burger'sopinionquicklygainedthesupportofJusticeBlackmun,theCourt'smedianmem- ber,aswellassecuringvotesfromJusticesWhiteandRehnquist.However,JusticePowell,the medianmemberofthemajoritycoalition,optedtoengagedirectlywithBurgertotryto thescopeoftheCourt'sopinion.InhisbargainingmemotoBurger,hewrites:fiAsyouhave theresponsibilityfortheCourtopinion,Iwouldliketocommongroundifthisispossible.I couldnotjoinyourpresentdraft,although(apartfromsomeofthedictathatnodoubtyouintend tocondenseordiscard),Ihopewearenotirreconcilablyapart....[I]willawaityourseconddraft inthehopethatIcouldjoinallorsomepartsofyouropinion.fl TheChieftookPowell'ssuggestionsseriouslyandcirculatedaseconddraftoftheopinion, attachinganotetoPowelltosaythathefimadechangesinanefforttoaccommodateyourviews,fl whilearticulatingwhichofPowell'srequestswentunansweredinthemostrecentdraft.Powell wasnotandtriedtomakehisargumentsaboutthestate'sinterestonceagain.Burger pushedback,claimingthattoaccommodatePowellcompletelymeantthathewasfilosingother votesflandifPowellcouldnotjointheopinionfully,Burgerwouldfihavenochoicebuttoreturnto [his]basicposition.flInotherwords,Burgerthreatenedtoremovetheplacatingchangescontained inthesecondopiniondraftunlessPowellwouldgetonboard.Notably,however,Burger'sassertion 68 thatotherjusticeswerejumpingshipranghollow.Inhisprivatepapers,Powellquestionedwhich justiceswerechangingtheirvotesandthereisnorecordinanyofthejustices'privatepapersthat anyoftheothervoteswerePowellengagedwithBurgeragain,thistimeusinghisown majorityopinioninadifferentcase( Bellottiv.Baird (1979)),tohighlightwhathebelievedtobe inconsistencyinthedevelopmentofalegalrule.Hemadepointedsuggestionsconcerning sentencesandincludedlanguagethatwouldassuagehisconcerns.Burgeraccommodatedthese requestsandPowellsignedontothemajorityopinionwithoutfurthernegotiation. Unliketheotherjusticesinthemajoritycoalition,JusticeStewartneitherjoinedtheopinion attheoutsetnordidheattempttoBurgerthroughdirectengagement.Instead,Stewart seemedtorecognizethathisviewsweresofundamentallydistinctfromBurger'sthatheoptedto immediatelywriteaconcurringopiniontospellouthisviewsinthecase.LikePowell,Stewart foundhimselfintheideologicalcenterofthewinningcoalition.Yet,unlikePowell,hedisagreed withthelegalrulearticulatedintheCourt'scontrollingopinion.Intheend,heconcurredonlyin judgmentanddidnotsignontothemajorityopinionbeyondthedispositionalvote. WhydidPowellopttoengageBurgerwhileStewartwroteseparately?WouldBurger haveworkedtoaccommodateStewartifhehadmadeadifferentchoice?Aretherecase orjusticecharacteristicsthataffectthedecisiontobargain?Howmightthischoicethe majorityopinionauthor'saccommodationdecision?Perhapstherearesomejustices,likenotorious andloquaciousJusticePowell,whoalwaysmaketrouble.Or,perhaps,bargainingboilsdownto thestrategicpositionalpowerofaparticularjustice(i.e.,thosewhooccupykeymedianpositions orwhoareideologicallyproximate/extremetotheassignedauthor).InthischapterIgobacktothe beginningofthebargainingprocessand,inSection3.3,consideranecdotallywhyitmaymatter whodecidestobargain.Next,inSection3.4,Ipresenthypothesesonwhenandwhyjustices engageinbargainingwiththeopinionauthor.Ithenpresenttheresultsfromempiricalanalysesin Section3.5.Section3.6concludeswithadiscussionoftheresults. 69 3.3SelectionEffects WhenSupremeCourtjusticesmeetinconferenceupontheconclusionoforalargumentinacase, theyestablishnon-bindingmajorityanddissentingcoalitions.Themostseniormemberofthe winningcoalitionselectsanauthortowriteanopiniontothewillofthemajority.The justicesdispersetotheirseparatechambersandnothinghappensinacaseuntiltheauthorcirculates adraftofthemajorityopinion.Asthepreviouschapterestablished,theinitialcirculationofa draftofthemajorityopinionfrequentlykickstartsavibrantexchangebetweentheauthorandthe respondingjustices. Notably,afterreceiptoftheopiniondraft,theothermembersoftheCourtmustrespond totheauthorandtheyhaveseveralviablealternatives.Themostobvious,andleastrelevantfor purposesofthispaper,istosimplydisagreewiththedispositionandtheopinionbyauthoringor joiningthedissentingcoalition.Butmembersofthemajoritycoalitionarenotpowerless.While theymayopttosimplyjointheCourt'sopinionŒsendingafijoinderflmemorandumtotheau- thorŒtheymightseektochangethecontentoftheopinionbyengagingdirectlywiththeauthor. Alternately,respondingjusticesmightforgothebargainingprocessalltogetherbyimmediately authoringaseparateconcurringopinion. Thepreviouschapterfocusedentirelyontheauthor'sresponsetotherequestsprofferedby membersofthemajoritycoalitionandrevealedthatauthorsarelargelyamenabletochangessug- gestedbytheircolleagues.Thisraisesapuzzlingquestion:giventhehighrateofaccommodation, whydon'tjusticesengageinmorebargaining?Thereareseveralplausibleexplanationsforthis observation.First,itsuggeststheremaybeaselectioneffectatplay.Inotherwords,aresponding justicemightonlyopttoengagewiththeauthorifsheexpectsherrequesthasahighprobability ofaffectingchange.Second,thedepressedbargainingratemightsuggestthattherearecase-or factorsthatarecorrelatedwithbargainingandaccommodation.Forexample,per- hapstheresultsofthepreviouschapteraredrivenbythepropensityofcertainjusticesŒagain, likePowellŒtosendbargainingmemoranda.Finally,ratherthanengagedirectlywiththeauthor, 70 therespondingjusticemightopttosimplyauthoraseparateopinionrightoffthebatwithout tryingtoalterthecontentoftheCourt'smajorityopinion. Theseexplanationsdeservefurtherscrutinybecauseiftherearecaseorjustice factorsatplayduringtherespondingjustice'sdecisiontoengagewiththemajorityopinionauthor, itindicatesthatamajorityopinionauthor'sdecisiontoaccommodaterequestsfromhercolleagues iscontingentupontherespondingjustices'choicetoengageintheplace.Thatis,itiscritical toestablishiftroublingpatternsemergeatthebargainingstagetosuggestabiasintheauthor's accommodationdecision. Moreover,as Parhamv.J.R. (1979)illustrates,itseemsthattherearefundamentallydif- ferentmechanismsatworkwhenajusticesdecidestoneitherjointheCourt'sopinionnorengages withtheopinionauthorinhopeofsecuringachangetothelegallybindingmajorityopinion.It iseasyŒoreasierŒtounderstandwhyajusticewouldsimplyjointhemajorityopinionwithout engagingwiththeauthorifshefullyagreeswiththedraftoftheCourt'sopinion,butitisnot yetclearwhyajusticewouldforgonegotiationinfavorofimmediatelywritingseparately.Since theoptiontowriteseparatelyexists,itmightbethatthebargainingactivityweobserveisfunda- mentallyskewed.Forexample,thelastchapterfoundthatjusticesweremosttypicallycollegial andfriendlyintheirsuggestionstothemajorityopinionauthor.Thismightbedrivenbythefact thatthemostcontentiousdisagreementsnevermakeittothebargainingstage;theyarecaptured, instead,inseparateconcurringopinions.Thispaperattemptstoteaseoutthefactorsthatdrivethe decisiontobargainrelativetotheotherchoicesavailabletothemembersoftheCourtafterthe initialcirculationofthemajorityopinion.Indoingso,Iseektoexplainwhatpatternsweshould expecttoseeinajustice'sdecisiontobargaininsteadofjoinorwriteseparately. 71 3.4DeterminantsofthePropensitytoBargain 3.4.1PositionalBargainingPower Whiletheprimarypurposeoftheoriesofopinioncontrolaimtopredicttheideologicallo- cationoftheSupremeCourt'smajorityopinion,theyprovidealenstoexaminewhatweought toexpecttosee before thedecisioniseverpublished.Indeed,thesetheoriessuggestcompeting patternsaboutwhichjusticesareexpectedtoparticipateinthebargainingprocess.Inthissection, Idrawfromtwotheoriesofopinioncontroltodevelophypothesesconcerningajustice'sinitial responsetothecirculationofthemajorityopinionbytheassignedauthor. The medianjustice modelpositsthattheCourt'sopinionwillthepreferencesof thejusticewhoisfortunateenoughtooccupythemid-pointonthejustices'combinedideologi- calcontinuum.Thishasimportantimplicationsforbargainingandconcurringopinionbehavior. Assumingjusticeshavesinglepeakedpreferencesinauni-dimensionalpolicyspace,themedian justicemodelshowsthatthemajorityopinionwillconvergetotheidealpointofthemedianjustice (Epsteinetal.2005;Hammond,BonneauandSheehan2005;Martin,QuinnandEpstein2005; Schwartz1992).Namely,iftheauthoristheCourt'smedianmember,itsuggeststhattheauthor oughttobefreertowritethemajorityopinionatornearheridealpoint.Indeed,theprevious chapterfoundevidencethatwhentheCourt'smedianauthorstheopinion,sheislesslikelyto accommodatehercolleagues,whichsuggeststhattheremight,infact,begreaterpowerforthe justiceinthiskeystrategicposition. IfthemedianjusticewieldsinordinatepoweroverthelegalruledevelopedintheCourt's majorityopinion,itcouldsendanimportantsignaltotheotherjustices.Inparticular,thistheory suggeststhatrespondingjusticesshouldrecognizethattheyareunlikelytoaffectchangeinthe contentofthecontrollingopinion.Asaresult,thismodellendsthepredictionthat 72 PowerfulMedianAuthor: ininstancesofdisagreement,arespondingjusticewillbe lesslikelytoengageinnegotiationwiththemajorityopinionauthorandwillbemore likelytoimmediatelywriteaseparateconcurringopinionwhenthemedianjusticeisthe assignedauthor. Whilethemedianjusticemodelgivesaframeworkforhowpotentialbargainersweighthedecision toengagewhentheauthoroccupiesthekeyposition,italsosuggestsdirectionalityforhowtheme- dianjusticeŒshouldshebeanon-authoringjusticeinthemajoritycoalitionŒoughttorespond. Inparticular,thistheorysuggeststhat,inanticipationofaccommodation, PowerfulMedianBargainer: themedianjusticeshouldbemorelikelytoengagein bargainingwiththeopinionauthor,allelseequal. Inshort,thejusticesinthemajoritycoalitionarelikelytoconsiderboththeauthor'spowerrelative totheothermembersoftheCourtaswellastheirownstrategicbargainingposition. Asecondapproachgeneratessimilarpredictions,butarguesthatthekeyplayerinajudi- cialbargaininggameisnotthemedianoftheentirebench.Instead,themostpowerfuljusticein craftingthemajorityopinionisthe medianmemberofthemajoritycoalition (MoMC).Unlikethe medianjusticemodel,scholarswhoascribetothistheoryincorporatetheinstitutionalstructures uniquetotheSupremeCourt.Inparticular,AndersonandTahk(2007)andCarrubbaetal.(2012) arguethatbecausejusticesarefreetopenseparateopinions,thebenchisnotbeholdentothesin- glepeakedpreferenceofthemedianjustice.Contendingthatthedispositionalvoteisstableand thatjusticesareabletoexpresstheirdistinctpreferencesinconcurrences,thismodellendsanew perspectivetothebargainingprocess.Ihypothesizethat PowerfulMoMCAuthor: whenthemedianmemberofthemajoritycoalitionisthe assignedauthor,arespondingjusticewillbelesslikelytoengageinnegotiation. Similarly,theMoMCshouldrecognizeherstrategicadvantage,soIpredict 73 PowerfulMoMCBargainer: themedianjusticeofthemajoritycoalitionshouldbe morelikelytoengageinbargainingwiththeopinionauthor,allelseequal. Yetthereisanotherimplicationofthistheoryismarkedlydifferentfromthemedianjusticemodel. Priorevidencesuggeststhatthechiefjustice'sopinionassignmentdecisioniscontingentoncase context,whereheismorelikelytoassigntheopiniontoanideologicallymoderatejusticeinmini- mumwinningcoalitions(MaltzmanandWahlbeck1996;Maltzman,SpriggsandWahlbeck2000). Asaresult,thismightincreasethepropensityofrespondingjusticestoeschewbargaininginfavor ofwritingseparately,asIpredictinthe powerfulmedianauthor hypothesis.Inshort,whenthe Court'sauthorislocatedattheideologicalcenter,thissuggeststhattheremaybemoredisagree- mentintheideologicalwings.Toonesideoftheauthor,thedissentingcoalitionarticulatesits disagreementand,totheother,Iexpectanincreaseinconcurringopinionbehavior.Morefor- mally,Iexpect DivergencefromtheMedian: whenthemedianmemberofthemajoritycoalitionwrites theCourt'sopinion,ajusticewillbelesslikelytowriteaconcurringopinion. Putdifferently,whenthemedianmemberofthemajoritycoalitionauthorsanopinion,the expectationfromthesecondmodelpositsconvergencetoanon-medianideologicaloutcome(i.e., theideologicallocationofthelegalrulewillbetotheleftortotherightoftheCourt'smedian). Theresultantlegalrule,then,shouldinherentlybeideologicallymoreproximatetotheidealpoint ofmoreofthemembersofthewinningcoalitionanddepressthelikelihoodthatajusticewillpen aconcurrenceasaresponsetotheopinion.Onthecontrary,whenthemedianoftheentire benchwritesthemajorityopinion,Ianticipatethiswillincreasethelikelihoodthataresponding justicewillforgobargaininginfavorofwritingseparatelyattheoutset. 74 3.4.2Context,Politics,andCollegiality Establishedliteraturesuggeststhatcontextmatters.,therearevitalinstitutional,colle- gial,andpoliticalfactorsthatcouldmotivateorconstrainajustice'sresponsetothecirculation ofthemajorityopiniondraft.Asinthepreviouschapter,thevariablesthatmighttheau- thor'sdecisiontoaccommodaterequestsforhercolleaguesarethesamevariablesthatshould directtheresponsesoftheothermembersofthemajority.Forexample,inbothinstances,thepo- tentialbargainerandtheopinionauthormustbothacknowledgetheirstrategicpositionrelativeto oneanotherandtheotherjustices.Andinbothcases,morethanhalfacenturyofpoliticalscience scholarshipthatthe ideologicaldistance betweenthebargainingduoundoubtedlyaffectsbe- havior(e.g.,SegalandSpaeth1993).Thatis,theoreticallyspeaking,ajusticewhoisideologically opposedtotheassignedauthorhasthemosttogainifshethrowsherhatintothenegotiationring andsucceedsingettingherpreferencestranslatedintolaw.Ontheotherhand,justicesthatoccupy thefurthestideologicalpositionsfromtheauthorshouldrecognizethattheyareunlikelytosucceed innegotiationswiththeauthor.Therefore,Iexpect IdeologicalDistance1: ajusticewillbemorelikelytowriteaconcurrencethanto bargainasideologicaldistanceincreases . Conversely, IdeologicalDistance2: ajusticewillbemorelikelytojointhemajorityopinionwithout bargainingasideologicaldistancedecreases . Ideologicaldistancecapturesonekeyaspectoftherelationshipbetweenapotentialbar- gainerandtheauthor,buttheoriesofreciprocitysuggestthatbecausejusticesserveyearsŒoften decadesŒtogether,thereisampleopportunityfortheseindividualstoformalliancesandtode- veloptrustandcooperation.Ontheotherhand,justiceshavetheopportunitytopenalizedefectors throughtit-for-tatstrategies.Maltzman,SpriggsandWahlbeck(2000)evidencethatpastcoop- erationaffectsthelikelihoodthatajusticewillseekchangesfromtheopinionauthor. 75 Accordingly,Iexpectthat Cooperation: justiceswillbemorelikelytojointhemajorityopinionwithoutbargaining iftheyhaveahistoryofcooperation . Casecontextmayalsoshapebargainingbehavior.TheinstitutionalfiRuleofFiveflsug- geststhatanindividualjustice'soverthecontentofopinionmaydependonthe sizeof thewinningcoalition .Inminimumwinningcoalitions,wherethemajoritywasestablishedby venon-bindingvotesatconference,themajorityopinionauthorcannotaffordtoloseasingle vote(Murphy1964;Rohde1972;Wahlbeck,SpriggsandMaltzman1998;Maltzman,Spriggsand Wahlbeck2000).Withthisincreasedleverage,Iexpectthat CoalitionSize1: respondingjusticeswillbemorelikelytobargainwhenthewinning marginissmall,allelseequal . However,coalitionsizemayalsobeanindicatoroftheunofthemajoritycoalition.I expect,then,that CoalitionSize2: ascoalitionsizeincreases,respondingjusticeswilllesslikelytowrite concurringopinionsasastresponsetothemajorityopiniondraft . TheSupremeCourt'sannualtermrunsfromOctobertoJune,whichimpliesthatcasesar- guedearlyinthetermhavegreatertimefordeliberationandrevisionthancaseslaterintheterm.In theirvariousworks,Wahlbeck,SpriggsandMaltzman(1998),Spriggs,MaltzmanandWahlbeck (1999),andMaltzman,SpriggsandWahlbeck(2000)demonstratethatjustices' workload cantlyaffectsbargainingactivity.Itfollowsthat Workload: justiceswillbemorelikelytosimplyjointhemajorityopinionwithoutas theirworkloadgrowsorthetermendapproaches . Asnotedinthepreviouschapter,thereisacriticalmassofevidencethatsuggeststhatnot allcasesareofequalimportancetothejustices(EpsteinandSegal2000;Maltzman,Spriggsand 76 Wahlbeck2000;UnahandHancock2006;Black,SorensonandJohnson2013).Black,Sorenson andJohnson(2013)suggestthatjusticessignalthatsomecasesmattermoretothemindividually duringoralargumentbyengagingmoreactivelyduringoralargument.Furthermore,justicesare notimmunefromthelargersocietalcontextandarecognizantofwhenthemediaorthepublicis tunedintotheCourt'sdecision.ThemembersoftheCourtknowthatthelegalandpolicyconse- quencesofsomecasesaresimplymoreimportantthanothers.Iexpect,asMaltzman,Spriggsand Wahlbeck(2000,74)predict,thatfi[i]fajusticeviewsacaseasrelativelyunimportant,heorshe ismorelikelytoresistpayingthepersonalandworkloadcostsassociatedwithrespondingtothe author.flThetheoreticalpredictionisrelativelyclear.Iexpectthatas CaseSalience: ifthecaseisofgreaterimportance,thejusticeswillbemoreinclinedto trytothedirectionoftheCourt'smajorityopinionthroughengagementwith theauthor. Attheindividualjusticelevel,priorstudiessuggestthatjustice expertise and casecomplex- ity affecttheopinionwritingprocess(e.g.,Brenner1984;BrennerandPalmer1988;Maltzman, SpriggsandWahlbeck2000).Justiceswithapersonalinterestorexperiencewiththelawinagiven issueareamaybemorelikelytoengagewiththeauthor.Moreimportantly,perhaps,justiceswith issueexpertisemighthavemoretechnicalorspecializedknowledgeaboutacaseandbelievethat itistheirresponsibilitytostrengthenthemajorityopinion.Formally,Ihypothesize Expertise: justiceswillbemorelikelytobargainiftheypossessexpertiseinagiven case . Casecomplexitygenerateslessclearexpectations.Ononehand,itmightbereasonabletoexpect thatjusticeswouldnegotiatemorecarefullyoverthemyriadoflegalissuesraisedinespecially stickycases.Yet,ontheotherhand,complexcasesmightgenerateaspikeaconcurringopinions sincetherecouldbedifferentviableavenuesforansweringthequestionathand.Assuch,Iantici- patethat 77 CaseComplexity: justiceswillbelesslikelytosimplyjointhemajorityopinionasthe complexityofacaseincreases . Finally,wemightexpectthatotherjustice-levelcharacteristicsaffectthelikelihoodthat ajusticeengagesinthebargainingprocess.Extanttheoreticalandempiricalliteraturesuggests thatneworfreshmanjusticesmaytaketimetoacclimatetotheproceduresoftheCourt(Hagle 1993;BrennerandHagle1996;Maltzman,SpriggsandWahlbeck2000;Wahlbeck,Spriggsand Maltzman1998).TheorysuggeststhatjusticesnewtotheCourtwilltaketimetoadjusttothe inner-workingoftheCourtandmightthenbemorereticenttoengageinthebargainingprocess, whichindicatesthat FreshmanAuthor: freshmanjusticesmaybemoredeferentialtotheircolleaguesand, subsequently,morelikelytosimplyjointhemajoritywithoutbargaining . 3.5Results:TheDecisiontoBargain 3.5.1DataandMeasurement ThischapterrelieslargelyonthesamedataandmeasurementspresentedinChapter2,withseveral keydifferences.ThebargainingdatacomeprimarilyfromtheSupremeCourtOpinionWriting Database(Wahlbeck,SpriggsandMaltzman2009),whichcontainsarchivaldocumentationof allbargainingmemorandacirculatedafteroralargumentsandbeforethepublicreleaseofthe decision. 1 However,unlikeprevioustreatmentsofthebargainingstage(Maltzman,Spriggsand Wahlbeck2000;Wahlbeck,SpriggsandMaltzman1998),Idonotcountallmemorandaequally. 1 ThesedataareagainlimitedtoOctoberTerms1971through1985andcomefromcase containedwithintheprivatepapersofformerjustices,including:HugoL.Black(Libraryof Congress),WilliamO.Douglas(LibraryofCongress),JohnM.Harlan(PrincetonUniversity), WilliamJ.Brennan,Jr.(LibraryofCongress),ThurgoodMarshall(LibraryofCongress),HarryA. Blackmun(LibraryofCongress),LewisF.Powell(Washington&LeeUniversity),andWilliam H.Rehnquist(StanfordUniversity). 78 Priorstudiesdidnotexaminethecontentofthesememos,whileIeachmemoaccording tocategoriesdevelopedinacomprehensivecontentanalysis.Ithenisolatedonlystatementsthat actuallysolicitedchangestothemajorityopiniondraftandeliminateddissentingcoalitioncontent. Theanalysiscentersonacase-level(notrequest-level)decision. Aspreviouslynoted,bargainingonlyoccursinapproximatelyhalfofallcasesdecided bymajorityopinion.Thisraisesanimportantquestion:whattypesofcasesbecomesubjectto negotiation?Thus,thisanalysisaimstouncoverwhatmotivatesmembersofthemajoritycoalition tonegotiateinthetplace.Thedependentvariableinthemodelpresentedbelowcomes,in part,fromJusticeBrennan'sdocketsheets,whichrecordswhichjusticesvotewiththemajority duringthepost-oralargumentconference.Thetripartitedependentvariabletakesonthreedistinct valuestocapturethethreechoicesavailabletoarespondingjustice.First,ajusticemightelectto simplyjointhemajorityopinion.Ofthe9,867initialresponsestothedraftofthemajority opinionoverthe1,960uniquecasesinthedataset,approximately80percentoftheresponses weresimplefijoinderflstatements.Insucharesponse,ajusticesignalsherwillingnesstojointhe opinionaswritten.Inasecondchoiceavailabletorespondingjustices,comprising12percentof responses,justicesmayelecttoengageinbargainingwiththeauthor.Thelastavailableoption forrespondingjusticesistoforgonegotiationinfavorofwritingseparatelyfromtheoutset,which constitutestheremaining8percent. 2 Thedecisiontobargainisdbysimilarfactorsastheauthor'saccommodation decision.Themodelofthebargainingstage,therefore,includesmanyofthesameexplanatory 2 Thedependentvariableherefocusesonmembersofthemajoritycoalitionratherthanthe entireCourtfortworeasons:(1)Intuitively,Iexpectthatmembersofthemajoritycoalitionhave moreatstakeinshapingthecontentoftheCourt'sopiniontotheirpreferences.Thatis, thesearethejusticesthatagreewiththedispositionaloutcomeintheconferencevoteandtheir votesarekeytopublishedopinion.Onthecontrary,membersofthedispositionaldissentwork togethertoauthoraseparateopinionandreasonablyexpectthattheywillhavelittleon themajorityopinion.(2)Observationally,fromthecontentanalysisofthebargainingmemoranda, weseethatindeedwhendissentersengagewiththemajorityopinionauthoritisforoneoftwo primaryreasons.Theistoprovidenon-substantivesuggestionsforminorrevisionsandthe secondistoexplainthedissentasitrelatestothemajorityopinion. 79 variablesasinthepreviouschapter. Thekeyinterveningvariableofinterestisonceagainthebargainingandauthoringjustices' positionsontheCourt.Onceagain,todeterminethestrategicpositionofkeyplayers,Iusethe Martin-Quinndynamicideologyscores(MartinandQuinn2002)tocreateadichotomousvariable for BargainerisCourtMedian (codedas1ifarespondingjusticeisthemedianoftheCourtand 0otherwise).IuseBrennan'sdocketsheetstodeterminetheinitialcoalitionsestablishedatthe conferencevote,andcode BargaineristheMedianoftheMajorityCoalition(MoMC) as1whena particularjusticeoccupiesthispositionand0otherwise.Ifthemajoritycoalitionisanevennumber, mycodingstrategyallowsfortwomedianmembers.Iusethesamemeasurementstrategiesto includedichotomousvariablestocapture AuthorisCourtMedian and AuthorisMoMC . Toaccountfortheubiquitousofideology,Iincludetwoindicators. 3 The measuresthe ideologicaldistance betweenthebargainerandopinionauthor,whichissimplythe absolutevalueofthedifferencebetweenthejustices'Martin-Quinnideologyscores(Martinand Quinn2002). Inthisaccount,Iincludeasecondindicatordesignedtocaptureabargainer'sextremity fromtheothermembersofthemajoritycoalition. Coalitiondistance measuresthepotentialbar- gainer'sabsoluteideologicaldistancefromtheissueideologicalaverageofthemajority coalition(excludingtheauthor),sothatthescoreincreasesasthejusticemovesfurtherfromthe average. 4 Onceagain cooperation ismeasuredbyfithepercentageoftimethattheauthorjoineda separateopinionwrittenbythejusticeinthepreviousterm.fl(Maltzman,SpriggsandWahlbeck 3 Themeasurementstrategyofmanyofthecollegialandinstitutionalfactorscomefromthe variousworksfromMaltzman,Spriggs,andWahlbeck(1998,1999,2000).Theauthorsgenerously providedreplicationdatafortheirstudiesandIwasabletointhegapswheretheirdatafellshort. Asaresult,mymeasurementschemedrawsdirectlyontheonesemployedinthesestudies. 4 Theissueaveragescomefromthe12issueareasbySpaeth(1994).Inthe accommodationmodel,Iutilizeadifferentindicatortotapintothesameconcept,accountingfor thedispersionor coalitionheterogeneity wheregreaterscoresindicategreatervariancewithinthe majoritycoalition. 80 2000,79).Additionally,forthisanalysis,Icalculatedtheaverage rateofprioraccommodation , measuredbytheproportionofrequestsaccommodatedinthepreviousterm,bymajorityopinion authorforeachjustice. 5 Thelattermeasureisincludedinthebargainingmodelasameansto tapintojusticelearning.Thatis,Iexpectthatjusticesdeveloprationalexpectationsaboutthe likelihoodofaccommodationbasedontheirpastexperiencewiththatauthor.If,forexample,a justiceisrarelyaccommodatedbyanauthor,shemaybelesslikelytoinvestinthenegotiation processthansheotherwisewouldbe. Therestofthevariablesareaone-to-onematchwiththepreviouschapter,soIonlysum- marizethemhere.IusetheWahlbeck,SpriggsandMaltzman(1998)measureof politicalsalience thatderivesaveragesofamicuscuriaeparticipationandplaceseachcaserelativeto thataverage(i.e.,aboveaveragesuggestsacaseismoresalient). Legalsalience isasimpledummy variablethatcaseswhereprecedentisoverturnedorsubstantiallyaltered(Spaeth1994).At theindividuallevel,IemploytheMaltzman,SpriggsandWahlbeck(2000)opinionratiobyissue areatoaccountforjustice expertise .Totapinto legalcasecomplexity ,IagainfollowWahlbeck, SpriggsandMaltzman(1998)andcombinethreeindicatorsthattogethersignalthatmultiplelegal issuesareatplayinacase. Winningmargin isagainmeasuredastheabsolutevalueofvotesneededtomaintainthe majoritycoalitionandtheactualnumberofvotesreceived, workload accountsforthenumber ofopenopinionassignmentsforeachjustice, end-of-term denotesthenumberofdaysleftinthe Court'sterm,and freshmanauthor isadichotomousmeasurethatcapturesjusticeswhohaveserved lessthantwofullyearsontheCourt. 5 Icalculatedthisrateusingvariousrangingfromtheprevioustermtothreeor vetermsaverages,totheproportionofallpreviousterms.Theresultsarerobusttoall tionsandthevariableneverobtainsstatistical 81 3.5.2ModelandResults Thedependentvariableinthiscasecontainsmultiplenominalcategories,soIestimateamultino- miallogit.Sincethismodelsthelikelihoodthataresponsewillbechosencomparedtoanother alternative,whichservesasthereferencecategory,itproducestwosetsofestimates. 6 BecauseI ammostinterestedinwhatdrivesajustice'sdecisiontoengagewiththeopinionauthor,asopposed tosimplygoing-along-to-get-alongbyjoiningthemajorityoutrightortoimmediatelypenninga concurrence,Iusefibargainflasthecomparisongroup. 7 Itestforstatisticalsigniusing robuststandarderrorsclusteredonjustice. 8 Theresultsfromthemultinomialmodelprovideevi- dencethatpotentialbargainersconsidertheauthor'sstrategicposition,andthatthatcollegialand institutionalfactorsarethecriticaldriversofajustice'sdecisiontoengagewiththeopinionauthor. TheparameterestimatesfromthemodelarereportedinTable3.1andFigures3.1and3.2present theaveragemarginaleffectsforallstatisticallyvariables. Standingalone,anexaminationofthecovariatesintheoutcomeŒtojointhemajority Œsuggeststhatthereareveryfewimportantjustice,case,orcontextualvariablesthatdrivethe bargainingdecision,which,inturn,alleviatestheconcernforselectioneffectsintheprevious chapter.Forexample,whileideologicaldistancebetweenthepotentialbargainerandtheassigned majorityopinionauthorisstatisticallyasFigure3.1illustrates,thesubstantiveimpact issmall.Whenallvariablesareheldattheirmeanormodalvalues,theprobabilityajustice 6 Multinomiallogitmodeldependsontheassumptionoftheindependenceofirrelevantalterna- tives(IIA),whichmeansthatajustice'sdecisiontobargaininsteadofjoinshouldnotbe byherabilitytowriteaconcurrence.Whilethisraisesconcernbecauseajusticecouldlaterdecide towriteseparately(orchangefromwritingseparatelytobargainingorjoining,oranynumber ofcombinationstherein),theHausmanandSmall-Hsiaotestsprovideevidencethat,intheinitial responsedecision,theIIAassumptionis 7 Whilethecoefchange,itisarbitrarywhichcategoryservesasthebaseline.Theinfer- encesareidenticalwhenusingeitherfijoinflorfiwriteconcurrenceflasthereferencecategory. 8 Eachjusticeappearsinthesedatamultipletimes,butonlymakesasinglebargainingdecision inanygivencase.Tocontrolforthepossibilitythattheerrortermofonejustice'sdecisionis correlatedwiththeresidualssubsequentdecisions,itistheoreticallymostappropriatetoclusteron bargaineridentity. 82 joinstheopinionwithoutnegotiationstandsatnearly0.8.Thesignoftheexplanatoryvariableis asexpected:asideologicaldistanceincreases,ajusticeisstatisticallylesslikelyto jointheopinion.Yet,again,thisdoesnotchangetheexpectedoutcomeinameaningfulsense. Workload,too,whilestatisticallyissubstantivelyinconsequential. Priorcooperation,measuredasthefrequencythatajusticejoinedaseparateopinionbythe authorinthepreviousterm,isapowerfulpredictorofajustice'sdecisiontojoininsteadofbargain withtheauthor.Indeed,aspastcooperationincreases,ajusticebecomesfarlesslikelytoengage inbargainingandisalmostcertaintojointhemajorityopinionafterthecirculation. Figure3.1alsorevealsthatlegalsaliencedecreasesthelikelihoodajusticewillsimplyjoin, therebyincreasingtheprobabilitythatajusticeengageswiththemajorityopinionauthor,while freshmanjusticesareapparentlyreticenttobargain.Whenapotentialbargainerisinhertwo yearsonthebench,thelikelihoodthatshejoinsjumpsfrom80percenttonearly90percent. Figure3.2presentsmorecompellingevidencethatthereare,indeed,criticalcollegial,con- textual,andcasefactorsthatmotivateajustice'sdecisiontoengageinthebargainingprocess. Theresultslendsupporttoboththemedianjusticemodelandthemedianjusticeofthemajority coalitionmodel.AsIpredictedinthePowerfulMedianAuthorhypothesis,theresultsthat ajusticeislesslikelytobargainandmorelikelytowriteaconcurrenceasaresponsewhen theCourt'smedianistaskedwithauthoringthemajorityopinion.Whilethepredictedvaluesare small,therelativeincreasehereissubstantial.Thelikelihoodthatajusticewritesaconcurrence whenanon-medianjusticepensthedecisionstandsat7percent.Butwhenthemedianwritesthe majorityopinion,thisjumpstoalmost15percent,morethandoublingthechancethatajusticewill opttowriteseparately. Thesignandimpactisreversedwhentheauthoristhemedianmemberofthemajority coalition.Indeed,comparedtothedecisiontobargain,ajusticeisnowlesslikelytowritea concurringopinion.Theprobabilitydropsbyafaciallyslim0.035,buteffectivelyreducesthe probabilityajusticeelectstowriteaconcurrencefrom0.08to0.04,cuttingtheexpectedprobability inhalf. 83 VariableJoinWriteConcurrence Constant1.994 -1.224 (0.370)(0.630) BargainerisMoMC-0.098-0.388 (0.137)(0.246) BargainerisCourtMedian0.1200.341 (0.177)(0.305) AuthorisMoMC0.085-0.745 (0.170)(0.296) AuthorisCourtMedian0.1931.061 (0.208)(0.296) IdeologicalDistance-0.017 -0.001 (0.003)(0.006) CoalitionDistance0.0020.074 (0.007)(0.010) Cooperation1.123 -1.409 (0.525)(0.975) PastRateofAccommodation-0.0330.222 (0.403)(0.683) PoliticalSalience0.0040.032 (0.020)(0.032) LegalSalience-0.420 -0.459 (0.151)(0.270) Expertise-0.039-0.225 (0.058)(0.106) LegalComplexity-0.0190.290 (0.058)(0.087) WinningMargin-0.001-0.279 (0.040)(0.068) Workload-0.029 -0.059 (0.015)(0.027) End-of-Term-0.0000.000 (0.000)(0.001) FreshmanJustice0.427 -0.232 (0.214)(0.411) Observations9,867 log L -1667.23 Table3.1Multinomiallogitmodelofmajoritycoalitionmember'sdecisiontobar- gain Multinomiallogitmodelofmajoritycoalitionmember'sdecisiontobargain.Ro- buststandarderrorsclusteredonbargainingjusticeinthedataarereportedinparen- theses.*denotes p < 0 : 10(two-tailedtest). 84 Figure3.1Averagemarginaleffectsforstatisticallyvariablesonthe probabilityofjoiningthemajorityopinion Averagemarginaleffectsforallstatisticallyvariablesontheprobability ofjoiningthemajorityopinionwithoutbargaining.Baselinepredictedprobability is0.764.Whiskersrepresent95%intervals.Allvariablesheldatmean ormedianvalues,asappropriate. Severalothercollegialfactorsareworthhighlighting.Whileideologicaldistanceissub- stantivelycantwheninthedecisiontojoinorbargain,ideologicalconsiderationssurface inadifferentwayinthesecondsetofparameters.Whilethedistancebetweentheauthorand potentialbargainerremainsasajusticegrowsmoreideologicallyextremefromthe averageofthewinningcoalition,thejusticeisslightlymorelikelytowriteaconcurringopinion. Ontheotherhand,astheexpertisehypothesispredicted,justiceswhohavespecializedknowledge aboutacaseoranareaorlawarelesslikelytowriteseparatelyandmorelikelytomakesuggestions totheopinionauthor.Theexpectedprobabilitydropsamodest0.01,from0.08to0.07. Comparedtothedecisiontobargain,legalcomplexityincreasestheprobabilitythatajus- 85 Figure3.2Averagemarginaleffectsforallstatisticallyvariablesonthe probabilityofwritingaconcurrencewithoutstbargaining Averagemarginaleffectsforallstatisticallyvariablesontheprobability ofwritingaconcurrencewithoutbargaining.Baselinepredictedprobabilityis 0.087.Whiskersrepresent95%intervals.Allvariablesheldatmeanor medianvalues,asappropriate. ticewillsimplyskipstraighttowritingaseparateconcurrence.Asacasebecomesincreasingly complex,thelikelihoodthatajusticewritesaconcurrencewithoutattemptingtopersuade theauthortomakechangesjumpstoabove10percent.Finally,thesizeofthewinningcoalition expectations;asthewinningcoalitiongrows,ajusticeislesslikelytowriteseparately. 3.6DiscussionandConclusion Thischaptersoughttounpackthestageoftheiterativeopinionwritingprocessbyrevealing howjusticesrespondtothedraftofthemajorityopinion.Itwasmotivatedbyadesireto understandthedifferentchoicesavailabletothemembersofthemajoritycoalitiontoascertain 86 whetherthisseparatecase-leveldecisionaffectedthemajorityopinionauthor'sbehaviorincraft- ingtheCourt'sdecision.Ithat,asintheauthor'sdecisiontoaccommodationrequestsfrom thebench,respondingjusticesarestrategicactors.Inparticular,theauthor'sstrategicpositional advantagerelativetotheothermembersofthemajoritycoalitioneffectsbargainingbehavior.Un- likepreviousattemptstodiscernthemechanismsthatmotivatebargainingbehavior(Maltzman, SpriggsandWahlbeck2000;Wahlbeck,SpriggsandMaltzman1998),Ilittleevidencefor severalcontextualindicatorsandlimitedeffectsforseveralothers. Byincorporatinginsightfromgametheoreticmodelstoderiveempiricalimplications,for thetime,thejustices'relativebargainingpowerwasputtothetest.Theresultsshowthat apotentialbargainerisunlikelytoconsiderhisorherownpositionalpowerinthedecisionto engage.Thatis,inthejustice'sstresponsetothecirculatedmajorityopiniondraft,thereisnot evidencetosupporteitherthemedianjusticeorthemedianofthemajoritycoalitionmodel.This istruewhetherajusticeelectstojoin,tobargain,ortowriteseparately.Thejusticesoccupying keystrategicseatsarenomoreorlesslikelytodooneortheother. Yet,theresultsprovidestrongevidencethatpotentialbargainersdoconsiderthe author's pivotalposition,lendingsupporttoboththeoriesofopinionpower.Additionally,thisgiveslimited supporttoscholarshipthatpredictstheauthorretainsameasureofcontroloverthecontentand locationofthelegalruledevelopedinthemajorityopinion(Hammond,BonneauandSheehan 2005;LaxandCameron2007).Ithatjusticesaremorelikelytoforgobargainingandareless likelytosimplyjointhemajoritywhentheauthoristhemedianmemberoftheSupremeCourt.As Iexpected,whentheopinionislocatedintheideologicalcenter,theauthorfacesgreaterfracturing oftheopinionthroughdissentsandconcurrences.Bycontrast,whenthemedianmemberofthe majoritycoalitionauthorstheopinion,itisclosertotheidealpolicypreferencesofmoremembers ofthemajoritycoalition,andseparateopinionwritingiseffectivelydiminished.Thisresultaligns withthepatternsofconcurrencesnotedinCarrubbaetal.(2012).Itremainsafascinatingresult thatmeritsfurtherstudy.Inparticular,futureresearchmightexaminethedifferencesincontentof concurrencesthatarisewhenvariouspivotalactorsoccupytheauthorship.Forexample,whenthe 87 medianjusticewritestheopinion,aretheconcurringopinions(especiallythosewrittenwithout priorengagementwiththeauthor)moreideologicallyextreme?Whatconsequencesdotheseopin- ionscarryforlowercourtinterpretationofandcompliancewithprecedent?Doesitchangewhen themedianofthemajoritycoalitionauthorstheopinionandtheconcurrencesare,presumably, bothmoreextreme and moremoderate? Beyondthepivotalpositionofauthoringjustices,Iideologicalconsiderations bargainingbehavior.Justicesideologicallyproximatetotheauthoraremostapttosimplyjointhe opinionwithoutbargaining,whileengagementincreasesasdistancegrows.Asanactorbecomes increasinglyextremerelativetotherestofthemajoritycoalition,thepropensitytoforgobargaining andwriteseparatelyincreases.Becausethestrategicadvantageofpivotalplayersisexplicitly modeledandcapturesasubstantialdegreeofideologicaltheseparametersaremarkedly lesspronouncedthanpreviousapproaches. Unlikethepreviouschapterwhichfoundnoevidenceforreciprocityinanauthor'sdecision toworkwithhercolleagues,previouscooperationplaysastarringroleinthedecisiontoengage. Itisthesinglelargestpredictorthatajusticewilljointhemajorityopinionwithoutanattemptto alteritscontent. LikeMaltzman,SpriggsandWahlbeck(2000)andWahlbeck,SpriggsandMaltzman(1998), Ithatcertaincase-levelvariablessystematicallybargainingbehavior.Whilepolitical saliencefailstoachievestatisticaltheresultsexpectationsthatcasecomplex- ityandcasesaliencedictateajustice'sdecisiontoengageinthebargainingprocess.More cally,asthecasebecomesincreasinglycomplex,justicesarelesslikelytobargainandmorelikely towriteconcurringopinions.Inespeciallycomplexcases,atthesamplemaximum,Iestimatethat theprobabilityajusticewillwriteseparateisnearly20percent.Althoughthisisremarkable,given therarityofwritingseparatelyasaresponse,itisnotentirelysurprisinggiventhatthecontent analysisofbargainingmemorandarevealsthatjusticesaremostoccupiedwithdiscussionsoflegal issues.However,awordofcautionisanorder:becausemostcases(72percent)inthesample arenotlegallycomplex,theresultsprovidegreaterforthepredictedprobabilityatthe 88 lowerboundthantheupperbound(whereonlyafractionofonepercentofcasesare Thelegalsalienceofthecasealsoleadstosystematicchangesinbargainingbehavior. Whilebothsetsofparameterestimatesarestatisticallylegalsalienceisonlysubstan- tivelymeaningfulinthatitmarginallydecreasesajustice'swillingnesstojointheopinionoutright. theevidencepresentedinthepreviouschapter,justicesaremorelikelytoengagein collegialdevelopmentofthemajorityopinioninlegallysalientcases. Othermeasuresoftheofthenation'shighestbenchalsosuggestparticulartypes of(dis)engagement.y,inaccordancewithoneofthecoalitionsizehypotheses,asthe winningmargingrows,ajusticeislessliketoofferaconcurringopinionasanini- tialresponsetothemajorityopiniondraft.Quiteunexpectedly,andasacleardeparturefrom establishedliterature,thesizeofthewinningcoalitiondoesnotseemtoincreasethepropensityto bargainoverthedecisiontodisengageandjointheopinionoutright. Finally,theresultspointtoinstitutionalfactorsasmodestconstraintsonbargainingactivity. Contrarytoexpectations,workloadactuallyappearsto increase ajustice'swillingnesstorequest changestotheopiniondraftratherthanpassivelyjoin.Yet,toreiterate,whilethisisstatistically theexpectationchangesverylittle.Themarginaleffectsofthesecontextualvariables ontheprobabilitythatajusticebargainsarevirtuallyindistinguishable,evenattheupperandlower levelscontainedinthesample.Inoppositiontomyhypothesis,thereisalsonoevidencetosuggest thatjusticesareanymoreorlesslikelytobargainastheendoftermdrawsnear.Asexpected, freshmanjusticesarelesslikelytosolicitchangesfromtheauthorandmorelikelytopassively jointhemajorityopinion. Takentogether,theresultssuggestthatthedecisiontoengageshapestheoutcomeofnego- tiationbetweenanauthorandabargainingjustice.Becausejusticeshaveanopportunitytovoice theirownviewsthroughconcurringopinions,andbecausejusticeselecttodosowithincreasing frequencywhentheauthoroccupiesapivotalposition(themedian),whenthecaseisespecially complex,andwhenthejusticeismoreideologicallyremovedfromtherestofthewinningcoali- tion,itislikelythattheopinionauthorisshieldedfromthemostfundamentallydivergentviews. 89 Moreover,sincepastcooperationisastrongdeterminantofajustice'swillingnesstojointhema- joritywithoutnegotiation,itislikelythatauthoringjusticesdealprimarilywithrequeststhatseek tochangethemajorityattheperiphery.Thestepbacktoexamineajustice'sinitialdecisionto engageofferscriticalinsightintowhy,precisely,therateofaccommodationbyopinionauthors issoremarkablyhighand,simultaneously,offersanexplanationforwhywedonotseemore bargaining. 90 REFERENCES 91 REFERENCES Anderson,RobertIVandAlexanderM.Tahk.2007.fiInstitutionsandEquilibriumintheUnited StatesSupremeCourt.fl AmericanPoliticalScienceReview 101(4):811Œ825. Black,RyanC.,MaronW.SorensonandTimothyR.Johnson.2013.fiTowardsanActor-Based MeasureofSupremeCourtCaseSalience:Information-SeekingandEngagementDuringOral Arguments.fl PoliticalResearchQuarterly 66(4):803Œ817. Brenner,Saul.1984.fiIssueSpecializationasaVariableinOpinionAssignment.fl JournalofPoli- tics 46(4):1217Œ1225. Brenner,SaulandJanPalmer.1988.fiTheTimeTakentoWriteOpinionsasaDeterminantof OpinionAssignments.fl Judicature 88:179Œ184. Brenner,SaulandTimothyM.Hagle.1996.fiOpinionWritingandtheAcclimationEffect.fl Polit- icalBehavior 18(3):235Œ261. Carrubba,Cliff,BarryFriedman,AndrewD.MartinandGeorgVanberg.2012.fiWhoControlsthe ContentofSupremeCourtOpinions.fl AmericanJournalofPoliticalScience 56(2):400Œ412. Epstein,Lee,DanielE.Hoe,GaryKingandJefferyA.Segal.2005.fiTheSupremeCourtDuring Crisis:HowWarAffectsOnlyNon-WarCases.fl NewYorkUniversityLawReview 80(1):1Œ116. Epstein,LeeandJefferyA.Segal.2000.fiMeasuringIssueSalience.fl AmericanJournalofPolitical Science 44(1):66Œ83. Hagle,Timothy.1993.fiFreshmenEffectsforSupremeCourtJustices.fl AmericanJournalofPo- liticalScience 37(4):1142Œ1157. Hammond,ThomasH.,ChrisW.BonneauandReginaldS.Sheehan.2005. StrategicBehaviorand PolicyChoiceontheU.S.SupremeCourt .PaloAlto,CA:StanfordUniversityPress. Lax,JeffreyR.andCharlesM.Cameron.2007.fiBargainingandOpinionAssignmentontheUS SupremeCourt.fl JournalofLaw,Economics,andOrganization 23(2):276Œ302. Maltzman,Forrest,JamesF.IISpriggsandPaulJ.Wahlbeck.2000. CraftingLawontheSupreme Court .NewYork:CambridgeUniversityPress. Maltzman,ForrestandPaulJ.Wahlbeck.1996.fiMayItPleasetheChief?OpinionAssignments intheRehnquistCourt.fl AmericanJournalofPoliticalScience May(40):421Œ433. Martin,AndrewD.andKevinM.Quinn.2002.fiDynamicIdealPointEstimationviaMarkov ChainMonteCarlofortheU.S.SupremeCourt,1953-1999.fl PoliticalAnalysis 10:134Œ153. Martin,AndrewD.,KevinM.QuinnandLeeEpstein.2005.fiTheMedianJusticeontheU.S. SupremeCourt.fl NorthCarolinaLawReview 83(5):1275Œ1322. 92 Murphy,WalterF.1964. ElementsofJudicialStrategy .Chicago:UniversityofChicagoPress. Rohde,David.1972.fiATheoryoftheFormationofOpinionCoalitionsintheU.S.Supreme Court.fl MidwestJournalofPoliticalScience May(208-224). Schwartz,EdwardP.1992.fiPolicy,Precedent,andPower:APositiveTheoryofSupremeCourt Decision-Making.fl JournalofLaw,Economics,andOrganization 8:219Œ252. Segal,JefferyA.andHaroldJ.Spaeth.1993. TheSupremeCourtandtheAttitudinalModel .New York:CambridgeUniversityPress. Spaeth,HaroldJ.1994. UnitedStatesSupremeCourtJudicialDatabase,1953-1996Terms .8th ed.AnnArbor:Inter-UniversityConsortiumforPoliticalandSocialResearch. Spriggs,JamesF.II,ForrestMaltzmanandPaulJ.Wahlbeck.1999.fiBargainingontheUS SupremeCourt:Justices'ResponsestoMajorityOpinionDrafts.fl JournalofPolitics 61(2):485Œ 506. Unah,IsaacandAnge-MarieHancock.2006.fiU.S.SupremeCourtDecisionMaking,Case Salience,andtheAttitutinalModel.fl LawandPolicy 28(3):295Œ320. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.1998.fiMarshallingtheCourt: BargainingandAccommodationgontheUnitedStatesSupremeCourt.fl AmericanJournalof PoliticalScience 42(1):294Œ315. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.2009. TheBurgerCourtOpinion WritingDatabase .Washington,DC:GeorgeWashingtonUniversity,DeptofPoliticalScience. 93 CHAPTER4 CONCURRINGOPINIONSINAFRACTUREDMAJORITY 4.1ChapterAbstract ExistingresearchpresentscompetinghypothesesaboutbargainingpowerontheCourt,withmany modelspredictingconvergencetothepreferencesofasinglemedianmember,whileotherssug- gestthattheopinionauthorherselfretainscontrol.Whilethesemodelsareusefulto understandtheideologicallocationoftheCourt'smajorityopinion,onefeaturefrequently overlookedintheliteratureistheoccurrenceofconcurringopinions,whichallowjusticestoartic- ulatetheirdistinctviews.TheseseparateopinionshavebecomecommonplaceinCourtdecisions asroughly40percentofallmajorityopinionsareaccompaniedbyatleastoneconcurrence.The twosubstantivechaptersestablishedthatmajorityopinionauthorsarelargelyamenableto requestsfromtheircolleagues,butthatthesecompromisesmightbeanartifactoftheabilityof majoritycoalitionjusticestosimplywriteseparatelywhendisagreementrunstoodeep.Yet,I thatlessthanhalfofallconcurrencesarewrittenwithoutengaginginnegotiationwiththe majorityopinionauthor.Theremainingconcurrences,then,resultaftertheassignedauthorand themajoritycoalitionmemberfailtostrikeabargain.Inthischapter,Iexplorewhathappenswhen intra-Courtbargainingcollapses.Descriptivedatasuggestthatjusticesaremorelikelytowriteor joinconcurringopinionswhenthetoneofexchangewiththemajorityopinionauthorishostile.My initialresultssuggestthatthetypeofconcurrenceŒanditssubsequentchallengetotheauthority ofthemajorityopinionŒmaybebythebreakdownduringtheintra-Courtbargaining process. 94 4.2Introduction In1982,theU.S.SupremeCourtconfrontedanobscenitychallengethattestedthelimitsoftheFirst Amendment.ThecasearosewhenaNewYorkCitybooksellersoldtwotoundercoverpolice ofthatcontainedfootageofunderageboysengaginginsexualactivity.Thestoreowner,Paul Ferber,wasconvictedundertheNewYorkobscenitylawthatmadeitillegalforanindividualto fipromote[...]anyperformancewhichincludessexualconductbyachildlessthansixteenyearsof age.flFerberappealedhisconviction,whichwasoverturnedbytheNewYorkCourtofAppeals onthegroundsthatthelawwasoverbroadandcould,unintentionally,bannon-obscenematerials thathadsomeorsocialvalue.TheSupremeCourtdisagreedwiththeCourtofAppeals andunanimouslyupheldNewYork'sobscenitylaw,arguingthatthestatehadavestedinterest inprotectingchildrenfromsexualexploitationandthatallinstancesofchildpornographywere outsideoftheboundsofprotectionprovidedbytheFirstAmendment. SimplytakingtheCourt'sunanimousdecisionatfacevaluewouldsuggestthatthejus- ticeswereinagreementaboutthescopeoftheFirstAmendmentandtheConstitutionalityofthe lawinNewYork.However,acloserlookattheopinionshowsthatwhilethedispositionalvote wasindeedunanimous,theCourtwasdividedaboutthescopeofthelegalrule.JusticeWhite deliveredtheopinionoftheCourtandwasjoinedinfullbyonlyChiefJusticeBurgerandJus- ticesPowell,Rehnquist,andO'Connor.WhilesheagreedwithWhite'slegalrational,O'Connor tooktheadditionalsteptoaconcurringopiniontofistressthattheCourtdoesnotholdthat NewYorkmustexcept`materialwithseriousliterary,oreducationalvalue'[...It]merely holdsthat,eveniftheFirstAmendmentshelterssuchmaterial,NewYork'scurrentstatuteisnot sufoverboardfl( Ferberv.NewYork (1982)).JusticeBrennan,joinedbyJusticeMarshall, aspecialconcurrenceinwhichheagreedwiththedisposition,butofferedentirelydifferent legalLikeBrennan,JusticeStevensanopinionconcurringonlyinthejudgment whilearticulatinghisownviewsconcerningFirstAmendmentobscenitystandards.Finally,Justice Blackmunnotedthathe,too,concurredonlyinthejudgment,butheoptednottoexplainwhy. 95 Alookintothejustices'privatepapersrevealsarichhistoryintheopiniondevelopment inthiscase.Notably,JusticesBrennan,Marshall,Blackmun,andStevenswereallinitiallyinthe dissentingcoalition.YetafterByronWhitecirculatedthedraftoftheCourt'smajorityopinion, Brennanwrotetohiscolleaguesstating:fiIadvisedyouonMarch30thatI'dundertakethedissent intheabove.Now,however,thatI'vereadByron'scirculation,I'minclinedtothinkthatheisright bothonthemeritsandtheoverbreadthquestion.flUltimately,hedecidedtoconcurinthejudgment whileprovidingaseparateopiniondetailinghisviews.JusticeMarshall,Brennan'smostfrequent ally,joinedthisseparateopinion.ThoughtheywerepersuadedbyWhite'sopinion,neitherjustice solicitedchangesfromWhite,optingtowriteseparatelywithoutbargaining.JusticesStevensand Blackmuntookasimilartact:changingtheirvotesandconcurringopinionswithoutengaging withWhiteoverthecontentofthemajorityopinion. Thisbehaviorisinstarkcontrasttotheotherjusticesinthemajoritycoalition.TheChief JusticeandJusticePowellimmediatelyjoinedWhite'smajorityopinion,butJusticesRehnquist andO'ConnorengageddirectlywithWhiteinthedevelopmentoftheCourt'sopinion.O'Connor acknowledgedthatWhitehadcirculatedafipersuasiveopinionflbutsheremainedfitroubledbythe exceptionforvisualdepictionsthathaveseriousliterary,artistic,political,orscientvalue.The pointsyoumakeonpp.8-15[...]havethesameforcenomatterthe`value'ofthedepiction.In addition,anexceptionforpictureswithseriousliterary,artistic,politicalorvaluewill requirethecourtstomakejustthesortofcontent-baseddistinctionstheFirstAmendmentabhors. [...It]isunnecessarytocreatethisexceptioninordertoprotectlegitimateworksthatareunlikely toharmchildren.flO'Connor'smemoprovidedadditionaldetailstosupportherargumentand sherequestedthatWhiteeliminatetheexceptionformaterialsthatfihaveseriousvalue,stressing insteadtheneedtocategoriesofprohibitedconduct.fl Rehnquistpennedhisowntwo-pagebargainingmemorandumtoWhite,notingthefisame areaofconcernflastheissuesbyO'Connor.Heindicatedthathe,too,wouldlikefito seethisrequirement[forseriousliterary,artistic,politicalorvalue]eliminatedaltogether fromthetest.flRehnquistprovidedWhitewithchangestotroublinglanguagethatwould 96 alleviatehisconcernsandconcludedthatfiotherthanthepointsmadeabove,Iammorethanhappy tojointhedraftaswritten.fl BeforeheevenreceivedthesecondmemofromRehnquist,WhitesetouttoaddressO'Connor's concerns.Inalettersenttotheentireconference,hewrotebacktohertoindicatethatheficrossed thisbridgeinthecourseofpreparingthecirculatingdraft,butIamperfectlywillingtoreconsider. Iamrecirculatingeliminatingthe...flparagraphsthatwereparticularlytroublesometoher.He continued,fiThismaysatisfyagooddealofyourconcern,althoughitmaymakethedraftconsid- erablylessattractiveinotherquarters.flWhite,then,soughttoaccommodatemostofO'Connor's requests,whichinturnalsoaddressedtheissuesraisedbyRehnquist,whilearguingthathethought itimpossiblefitoavoidcase-by-caseadjudicationsthatwillbecontent-oriented,anymorethanit ispossibletoavoidsuchcasesintheobscenitycontextunder Miller and ParisAdult .fl 1 White'srevisionwasenoughtopacifyRehnquist,whosubsequentlyjoinedtheopinion. O'Connor,ontheotherhand,wasnotShewrotetoWhiteagain,stating,fiIcontinueto betroubledbytheconcernsexpressedinmylettertoyou.I,therefore,plantocirculateabrief partialconcurrencewhichsimplyreservesjoiningyouastotheexceptionforworksofserious socialvalue.flThroughmultipleiterationsofherconcurrence,O'Connorthelanguageso thatheropinionwasfinotinconsistentflwiththemajorityviews,butinsteadputparticularemphasis onacertainpoint. Ferberv.NewYork (1982)isillustrativeofseveralkeyfeaturesofSupremeCourtbargain- ingdynamics.First,itshowsthedifferent types ofconcurrencespublishedbyjustices.Atthemost generallevel,aconcurringopinionisaseparatewrittenstatementinwhichajusticeagreeswith thedispositionand,frequently,someorallofthelegalprovidedinthemajorityopin- ion,butthejusticeelectstoprovidefihisviewsforthecaseorhisreasonsforconcurringfl(Black 1 Millerv.California (1973)wasalandmarkcasedecided10yearsearlierinwhichtheCourt obscenityasanythingthatwasfiutterlywithoutsociallyredeemingvalueflorthatlacksfise- riousliterary,artistic,politicalorvalue.flInthisdecision,theCourt'smajoritydeveloped athree-prongtestforlowercourtstoestablishwhetherthisstandardhadindeedbeenviolated. Decidedthesameyearas Miller , ParisAdultTheatrev.Slaton (1973)upheldastate'sinjunction againstshowingobsceneinpublicmovietheaters,evenifviewerswereallconsentingadults. 97 1991,200).Therearefurtherdistinctions,however,betweenregularandspecialconcurrences. Intheformer,asshownbyJusticeO'Connorin Ferber ,ajusticeagreeswiththemajority'sdeci- sionand(mostof)thebutelectstoexpressherviewsseparately.Thelatter,special concurrences,areliketheoneswrittenbyJusticesBrennan,Stevens,andBlackmunwherebythe justiceagrees only withthedispositionalvoteanddoesnotsubscribetothelegalrule.Second, itshowsthevariationinthelengthandbreadthofconcurringopinions.Blackmun,forexample, notedhisconcurrencewithoutwhilealloftheotherjusticesauthoredseveralpagesof texttosupporttheirarguments.Third,theconcurringopinionsunderscorethatevenfiunanimousfl decisionsaresometimesmoredivisivethantheyappearonthesurface.Fourth,itrevealsthatsome concurrencesresultwithoutengagementbetweenthemajorityopinionauthorandanothermember ofthemajoritycoalition(aswasthecasewithBrennan,Stevens,andBlackmun)whileothercon- currencesresultafteranauthorfailstofullyaccommodatetheviewsofabargainingjustice(like O'Connor). AsMaveety(2005,139)observes,ficoncurringvoicesproducethelegaldebatethatfurthers theintellectualdevelopmentofthelawontheSupremeCourt.flTheviewsarticulatedinthese separateopinionscanhavemeaningfulimpactsontheinterpretationofprecedentbylowercourts,it canagendasettingandpolicy-makinginCongress,andcanimpactthedebateaboutlegal rulesinthelegalcommunity.Becauseoftheirpotentialtohavelonglastingimpacts,understanding whereconcurrencescomefromisanimportantstepinunderstandingpreciselyhowthese opinionsaffectthebroaderpoliticalworld. Inthelastchapter,Ifoundthatjusticesaremorelikelytowriteconcurrenceswithout engagingwiththeassignedmajorityopinionauthorwhentheSupremeCourt'smedianmember authorsthemajorityopinion,whenthejusticeisfurtherfromtheideologicalmeanofthewinning coalition,whenthedecisionisespeciallycontentiousanddecidedbyaminimumwinning(5-4) vote,andwhenthecaseisparticularlycomplexandraisesalargernumberoflegalissues.The decisiontoforgobargainingisilluminating;itsuggeststhatjusticesrecognizethatsometimesthe divisionswiththerestofthemajoritycoalition,orattheveryleasttheassignedauthor,areso 98 fundamentalthatitismoreeftosimplywriteseparatelythantotrytopersuadetheauthor tomakechangestothemajorityopinion.Yet,Ithatthedecisiontowriteseparatelyfromthe outsetexplainsapproximately45percentofallconcurringopinions.Theremaining55percent, then,result after ajusticehasengagedwiththemajorityopinionbyrequestingchangestothe Court'sdecision.Inthispaper,Iseektounderstandthefactorsthatmotivateajusticetowrite separatelyafterhavingengagedwiththeauthor. Thischapterproceedsasfollows:inSection4.3,Iprovideanecdotalevidencetoillustrate justicesresponsestoamajorityopinionauthor'sdecisionnottoadoptsomeoralloftherequests offeredduringthebargainingstage.Section4.4offersatypologytoaccountforthediversitywithin separateopinions.Section4.5presentsdatatoshowwhichjusticeswriteseparatelyandanalyzes theconditionsunderwhichconcurringopinionsappear.Section4.6concludeswithadiscussion oftheresultsandprovidesavenuesforfurtherresearchintothislittleunderstoodphenomenon. 4.3ResponsestotheMajorityOpinionAuthor'sRefusaltoAccommodate Takentogether,asestablishedinthepreviouschapters,thedescriptivedataandstatisticalmodels revealthatauthorsworkhardtofosteracollegialspiritinthedraftingoftheCourt'smajorityopin- ion.Bythetimeitisreleasedtothepublic,majorityopinionsaremostfrequentlyacollaborative effort,havingundergonerevisionsastheauthoradoptssubstantiverequestsfromthe othermembersofthemajoritycoalition.Yetdespitethefactthattheaccommodationratestandsat anaverageofnearly70percent,almostathirdofrequestsgounansweredbythemajorityopinion author.Thisraisesanimportantquestion:Whathappenswhenajusticeis not accommodated? Onepersistentaccountsuggeststhatconcurrencesarefrequentlyaresultoftheauthor'sfailureto negotiatewithmembersofthemajoritycoalition. Thenotionthatconcurrencesresultfromthefailureoftheauthortoincorporateacol- league'sdemandsiscertainlynotnew.Indeed,morethanyearsago,Murphy(1964)argued thatajusticehastwotoolsathisorherdisposaltosanctiontheopinionauthor:changingvotesor 99 authoringaseparateopinion.InherexaminationofconcurringopinionsontheSupremeCourt, Corley(2010,41)beginstoaddressthisveryquestion,notingthatifafijusticeisaccommodated, thejusticewillsilentlyjointhemajorityopinion;however,ifthejusticeisnotaccommodated,the justicemaywriteorjoinaconcurrence.flProvidingaqualitativeanalysis,Corleyusestheprivate papersofJusticesBlackmunandMarshalltoexplorebargainingandaccommodationfrom1986- 1989.SheevidencethatBlackmunandMarshallactivelyincorporatedsuggestionsfrom othermembersoftheCourtintheopinionstheyauthored.However,whentheyrefusedtodoso,it frequentlyspurredthebargainingjusticetowriteorsignontoaconcurringopinion. Thisqualitativeassessmentgainsadditionaltractionhere.Indeed,aswasthecasein United Statesv.Generes (1971),Blackmuntradedhismostpreferredoutcometomaintaincontrolofthe majorityopinion.Ashefoughtforcontrolinthismundanefederalincometaxcase,split themajoritycoalition.Blackmunrequestsfromeverymemberofthemajoritycoalition andashesoughttoaccommodatesome(particularlythosefromBrennanandWhite),heriskedthe votesoftheothermembersofthemajoritycoalition.Toregaincontrol,Blackmunaccommodated Stewart,Burger,andMarshall,and,asaresult,Brennanwroteaconcurrencethatchallengedthe policyimplicationsofthedecision.Inthiscase,theconcurrencewasveryclearlyadirectresult ofnon-accommodation.HadBlackmunbeenabletoincorporatethemyriadofrequestsfromthe othermembersofthemajoritycoalition,noconcurringopinionwouldhavebeenpublished.In otherwords,noneofthejusticeshadwrittenseparatelywithouttryingtheirhandsatshifting thedirectionoftheCourt'smajorityopinion. Likewise,theCourt'smajorityfounditselftobefracturedin Nixonv.Fitzger- ald ,anespeciallycontentiouscasefrom1981thataskedtheCourttodeterminethescopeofthe President'simmunity.Fromtheoutset,thoughhehadthepowerofopinionassignmentandelected toappointJusticePowelltothetask,theChiefJusticesoughttoforcefullystrongarmhisviews intothemajorityopinion.Innofewerthanveseparatebargainingrequests,Burgermadeitclear thathewouldn'thesitatetopullhisvotefromthenarrow(5-4)dispositionalmajoritycoalition, especiallyifPowellwenttoofaraddressingtheconcernsofJusticeByronWhite.Inhisresponse 100 toPowell'smajorityopiniondraft,Burgerpromptlyresponded:fiPut`hard,'yourchoiceis myvoteorByron's!flWhilePowellandWhiteultimatelyendedinopposingcoalitions,Burger continuedtoworrythatPowellwastooconcernedwithWhite'sperspective.Afterthecirculation ofthemajorityopiniondraft,Burgertriedhishandagain,writing,fiLetByron`rant'onthis pointbutdon'tfallintothetrapofansweringhimandrenderingtheopinionunacceptabletome.fl PowellrecognizedhewaswalkingatightropeandsoughttoplacatetheChiefwithout losinganymorevotes(andhencelosingtheabilitytosetprecedentaltogether).Throughoutthe opinionwritingprocess,PowelldealtwithrequestsfromeverymemberoftheCourtasthemajority coalitionshiftedunderhisfeet.FacingcontinualpressurefromtheChiefandseekingtosecure thecriticalvote,Powellworkedevenhardertoaccommodaterequestsfromhiscolleagues, eventuallypersonallyappealingtoJusticeO'Connortoexplainwhyheadoptedonerequestfrom Burgerandtoaskforhersupport: Ihaveacceptedyourtwosuggestedsentencesasexplanationsoftheopinion'sap- proachtothequestionspresented.[...]Thelastsentencehasbeenaddedtomake clearthatourholdingfairlycanbecalled`constitutional'Œamatterofgreatconcern totheChiefJustice.Asperhapsyousense,Iamanxioustoconcludeworkonthis opinionŒbutonlywithaCourt.[...]Iparticularlyappreciateyourinterestandsup- port,andyoursuggestionshavebeenconstructive.Iverymuchhopeyouwillrenew yourjoin...itisnecessarytohaveasolidCourt. Ultimately,Powell'seffortstobalancetherequestsofhiscolleaguespaidoffandhesecureda minimumwinningcoalition.Despitehispersistentbadgering,Powellchosenottoincorporate Burger'srequeststothefullestextent.Thus,whileBurgerstillhadahandinshapingtheCourt's majorityopinion,someofhisrequestsfellondeafears.Asaresult,Burgerissuedaconcurring opiniontoclarifyhispositionandtosuggestthattheCourthadnotgonefarenoughinestablishing theConstitutionalquestion. Thisexampleisnotablenotonlybecauseitagainservestoillustratethecomplexityofthe 101 iterativeopinionwritingprocess.Butitencapsulatesthedualityofthreateningthestabilityofthe majoritycoalition.Ononehand,Powellmostcertainlyactivelyengagedinthenegotiationprocess withtheChiefJustice,incorporatingmanyofhissuggestionsintotheCourt'smajorityopinion. Yet,Powellwouldnotbebulliedintoacceptingapositionwithwhichhedidnotagree.Burger, then,hadachoicetomake:Ifhisthreatsweretocarryanyweight,heoughttowriteaconcurring opiniontosanctionPowell'srefusaltoaccommodate. IturnnowtoinvestigatethecircumstancesunderwhichmembersoftheCourtchooseto bringtheirdisagreementswiththeCourt'smajoritytolightbytakingtheadditionalsteptopublish aconcurringopinion. 4.4SeparateOpinionsandTypesofConcurrences Asinthepreviouschapters,Iuseoriginalmajorityopiniondraftdatatogetherwiththebargaining memorandacontainedwithintheSupremeCourtOpinionWritingDatabasetoteaseoutthedown- streamconsequencesofnon-accommodation.Afterreadingeachrequestposedtothemajority opinionauthor,Iascertainedwhethertheauthoracceptedthisrequestandchangedthemajority opinion. 2 Whilethebargainingmemorandaandmajorityopiniondraftsshedlightonthenegotia- tionprocessduringtheopinionwritingstage,itfailstorevealtheconsequencesofthebreakdown innegotiationbetweentheauthorandtherespondingjustice. Todeterminehowabargainerproceedsafteranauthorfailstoaccommodateherrequest, usingLexis,Icollectedeverywrittenopinion(majority,concurring,anddissenting)published duringtheBurgerCourt.Forthisanalysis,Iisolatedtheinstancesoffailedorincompleteaccom- 2 LikethebargainingmemorandadataobtainedfromtheSupremeCourtOpinionWriting Database(Wahlbeck,SpriggsandMaltzman2009),theopiniondraftdatacomefromcase containedwithintheprivatepapersofformerjustices,including:HugoL.Black(Libraryof Congress),WilliamO.Douglas(LibraryofCongress),JohnM.Harlan(PrincetonUniversity), WilliamJ.Brennan,Jr.(LibraryofCongress),ThurgoodMarshall(LibraryofCongress),HarryA. Blackmun(LibraryofCongress),LewisF.Powell(Washington&LeeUniversity),andWilliam H.Rehnquist(StanfordUniversity). 102 modationandtracedthesubsequentactionŒorlackthereofŒofthejusticewhoserequestwent unanswered.Ithenselectedarandom3-termsample(OctoberTerms1978,1983,and1984)to determinewhathappenedwhenanauthordecidedagainstincorporatingachangeprofferedbyan- othermemberofthemajoritycoalition.Thereare431casesdecidedduringtheseterms,butwe observeactivebargaininginlessthan60percentofthetotal,oronly242cases.Thissampleyields atotal753uniquerequestsaddressedtotheopinionauthor,whichequatestoapproximatelythree requestsorsuggestionsforeachcasedecidedintheseterms. Inthissample,220bargainingrequestswerenotaccommodated.Yettheslightedjustices optedtofracturethemajoritycoalitionbywritingorjoiningaconcurringopinioninfewerthan halfoftheseinstances;inonly90observationsdidtherespondingjusticeelecttoexpresstheviews thatwerenotaccommodatedbythemajorityopinionauthorbyjoiningorwritingaconcurrence. Inoneinstance,ajusticeactuallyreversedhissupportofthemajorityopinionandswitchedhisdis- positionalvotetosignadissentingopinion.Interestingly,whiletheslightedjusticesonlypursued retaliatoryactionin40percentofthecasesoffailedorincompletenegotiation,thedescriptive datasuggestthatthefailuretocompromisedoesindeedincreasethepropensityforajusticeto authororjoinaseparateopinion.Inthetermsunderconsideration,therewere124separatecon- currenceswritten;fully70writtenindirectresponsetothemajorityopinionauthor'srefusalto adoptasuggestion. Inadditiontomerelyassessingwhetherornotthesejusticesauthoredorjoinedseparate opinions,tosystematicallyassesstheimpactoftheconcurringopinionsandtodeterminehow deeplythebargainersplitwiththeCourt'smajority,IfollowedCorley(2010)andconducteda contentanalysisoftheseseparateopinions.UsingthetypologyderivedfromWitkin(1977)and Ray(1990),Icodedtheconcurrencesintosixcategories: (1)An expansiveconcurrence attemptstobroadenoraddtoscopeofthelegalrule explicatedinthemajorityopinion.Insuchaconcurrence,theauthorexpressesbelief thattheCourtshouldhavedonemoreorgonefurther. 103 (2)A limitingconcurrence istheoppositeoftheexpansiveinthattheauthoringjus- ticewishestorestricttheCourt'sopinioninsomeway.Inthistypeofconcurrence,the justicethinkssomepartsofthemajority'sopinionwereunnecessaryorthattheCourt wenttoofarinitsreasoning. (3)An emphaticconcurrence servesasavehicleforajusticetoclarifyorexpound uponaparticularpointarticulatedinthemajorityopinion,withoutseekingtosubstan- tivelychangetheinterpretationofprecedent. (4)Ina reluctantconcurrence ,ontheotherhand,theauthoringjusticemakeshisdis- pleasurewiththeCourt'smajoritywellknown.Inthistype,thejusticeexplainsthat shehasonlyjoinedtheCourt'smajoritybecauseshefeltcompelledtodosobyprece- dentorŒperhapsironicallyŒtoallowtheCourttospeakwithavoiceonan importantlegalissue. (5)A doctrinalconcurrence iswhatwetypicallyrefertoasaSpecialConcurrence, whereaconcurringjusticeagreeswiththeoutcomeinthecase,butnotthelegaljus- intheCourt'smajorityopinion.Theseconcurrencestendtobelongerand lookmoreliketraditionalopinionssincetheauthoringjusticeoffersherown,alternate forthedecision.And,, (6)An unnecessaryconcurrence is,asitsnameimplies,anopportunityforajusticeto simplystateheragreementwiththemajority,withoutofferinganyfurtherexplanation ofwhyherviewsdivergefromthepublishedopinion. AsshowninFigure4.1,40percentofallconcurrencesareemphaticinnature,doinglittle toaffecttheinterpretationandapplicationoftheCourt'sopinion.Yet,together,doctrinaland limitingconcurrences,bothwhichchallengethelegalruleestablishedinthemajorityopinionalso makeupacriticalmassoftheconcurrencesthatresultofabreakdowninthenegotiationprocess, comprisinganother40percent.Fewerthan10percentofconcurrenceseektoexpandthereachof theCourt'sopinionandjustover10percentprovidenoadditionalinsight.Therewerenoinstances 104 Figure4.1Barplotoftherelativefrequencyoftheconcurrencebytype. ofreluctantconcurrencesinthissample. 4.5ConditionsthatIncreasedLikelihoodofConcurringOpinions 4.5.1WhenConcurrencesEmerge:TypesofRequests Recallthatbargainingjusticespursueretaliatoryactionbywritingaseparateopinionlessthan halfofthetimewhenthemajorityopinionauthorrefusestoaccommodate.Insomecases,a bargainingjusticemightoptnottowriteaconcurrencebecausetherequestwasinconsequentialor notenoughtomeritaseparateopinion.Indeed,therearemanyexampleswherejustices makeasuggestionandindicatefromtheoutsetthattheywilljointheopinioneveniftheyarenot accommodated.SuchwasthecaseinRehnquist'sbargainingmemorandumtoWhitein NewYork v.Ferber (1982).RehnquisttookissuewithWhite'sfiuseofthe[word]`works'"andsuggested thattheopinionlanguagebechangedtoamoretailoreddescriptionofchildpornography. Hegoesontonote,however,fiIamperfectlywillingtoleaveittoyourjudgmentastowhetheritis possibleordesirabletoobtainamajorityforthatproposition.flRehnquistrequestedachangethat 105 Whitedidnotadopt,butRehnquistjoinedwithoutfurtherdelayanyway. Inothercases,however,sometimesajusticeexpressesadesiretowriteseparately,but choosesnottoforthegoodofthemajorityopinion.Suchwasthecasein UnitedStatesRailroad RetirementBoardv.Fritz (1980)whentheCourtarguedthatthelegislatureneededonlypassthe rationalbasistestinitsestablishmentofsocialandeconomicacts.Rehnquistwastaskedwith authoringthemajorityopinionandPowellimmediatelysoughtsizablechanges.WhileRehnquist adoptedsomeofPowell'srequests,herefusedtoaltertheopiniontoallofPowell'ssugges- tions.Powellwroteatwopagememotoarticulatehisdispleasurewiththeoutcome,butultimately decidedtofijoin[Rehnquist's]opiniontoassurethatyouhaveaCourt.flInnosmallpart,Powell wasmotivatedtojointhemajoritywithoutwritingseparatelybecauseRehnquisthadexpressed fearthatlowercourtswouldfipickandchooseflamongrelevantprecedentsifPowellpublisheda separateopinion. Thus,althoughjusticeshavemotivationnottowriteseparately,itisabundantlyclearthat therearecaseswherejusticestaketheadditionalsteptopenaconcurrence.Justasthecontent ofabargainingrequestthelikelihoodofaccommodation,itisreasonabletoexpectthat sometypesofrequestsmattermoretoabargainingjustice,providingthejusticewithheightened motivationtowriteorjoinaconcurringopinionifherviewsarenotadequatelyaddressedby themajorityopinion.Recallthethreebroadcategoriesofrequesttypesinthepreceding chapters:minorrevisionsasktheauthortoclarifylanguageormakeothernon-substantivechanges totheopinion,requestsforlegalchangedealexplicitlywithConstitutionalorstatutoryprovisions, whilerequestsforpolicychangeexpressconcernsforimplementation,interpretation,andthescope ofthelegalrule. AspresentedinChapter2,thedatarevealthatmajorityopinionauthorsarelesslikelyto adoptsuggestionsforpolicychangeandmostlikelytoincorporateminorrevisionsatthebehest oftheircolleagues.Ifthecontentoftherequestfunctionsasasignalofthedivergencebetween thelegalpolicyexplicatedbytheopinionauthorandthemostpreferredoutcomeoftheresponding justice,weshouldcorrespondinglyexpectthatabargainingjusticeismorewillingtoinvestthe 106 timeandeffortintoauthoringaseparateopinionwhenthejusticeseekschangestothepolicy implicationsoftheCourt'smajorityopinion.AsFigure4.2reveals,justicesarefarmorelikely toauthoraseparateopinionwhentheirpolicyconcernsarenotalleviatedbythemajorityopinion author.Oddly,nearly25percentofconcurringopinionsoccurafteramajorityopinionauthor refusestoadoptasuggestionforaminorrevision. Thispuzzlingissomewhatreconciledthroughanexaminationofwhattypeofcon- currenceresultsineachinstance.Minorrevisionstypicallyrequestclarityorsuggestemphasis onaparticularconcept.Accordingly,astherighthandpanelofFigure4.2shows,whenminor revisionsareignoredbytheopinionauthor,theresultantconcurrenceiseitherunnecessaryand addsnofurthersubstancetotheCourt'spublishedopinion,emphaticinthatishighlightskeytext withoutdisagreeingwiththemajorityopinion,or,inasmallminorityofcases,itseekstoexpand aparticularaspectofthelegalrule.Inthethreecaseswhereaminorchangewasnotadoptedand adoctrinalconcurrencewasthefiresult,fltheseparateopinionhasnothingtodowiththerequested change.Instead,thesearetherareinstanceswhereajusticebothbargainsandwritesseparately, nearlyinconjunctionwithoneanother. Figure4.2Barplotsofproportionofconcurrencesresultingfromdifferenttypesof bargainingrequests Barplotsoftheproportionabargainingjusticejoinedorwroteaconcurringopinion bytypeofbargainingrequest(left)andproportionoftypeofconcurrencewritten byrequestcontent(right). 107 Morethanathirdofallconcurringopinionsthatarewrittenafteraninitialattempttonego- tiatewiththeopinionauthorareemphaticinnature.Thisisastrikingpatternsinceitindicatesthat, mostcommonly,theseconcurrencesdonotseektolimitoraltertheinterpretationofprecedent. Instead,theseconcurrencesaremostoftenexpressiveinnatureandprovideanopportunityfora justicetopubliclyclarifywhichaspectsofanopinionsheviewsasmostvital.Lessthanaquarter ofconcurrencesinthesecasesseektolimitthereachofprecedentsetbythemajority,whileless than10percentexpressapreferenceforabroaderapplicationofthelegalrule.Another10percent areunnecessaryandcontributelittletoourunderstandingoftheindividualjustice'sreasoningor tothedevelopmentoflaw. Doctrinalconcurrencesmakeuptheremainingcategoryandofferthemostdistinctpattern ofjudicialbehavior.Moststrikingly,andleastsurprisingly,doctrinalconcurrencesareprimarily aresultoffundamentaldisagreementsbetweenmajorityopinionauthorsandbargainingjustices regardingtheinterpretationoflaw.Doctrinalconcurrencesarethoseinwhichthejusticesdis- agreeaboutthelegalrule,soitfollowsthatwhenmajorityopinionauthorsrefusetomakelegal changesconcerningConstitutionalinterpretation,forexample,justicesaredriventowritespecial concurrencesthatexplainadifferentlegalrationaletojustifythedispositioninacase. 4.5.2WhenConcurrencesEmerge:StrategyofNegotiation Whilefailuretoaccommodatearequestclearlydoesnottranslateimmediatelyintoapublicly fracturedmajoritythroughthecirculationandpublicationofaconcurringopinion,Figure4.3re- vealsthatJusticePowellwasrighttobeconcernedin Nixon whenChiefJusticeBurgercontinually threatenedthestabilityofthemajoritycoalition.Itisabundantlyclearthatwhenjusticesthreaten toretaliateagainstnon-compliantmajorityopinionauthors,theyfollow-up. ThelargegraybarinFigure4.3providesaproportionaldisaggregationoffailedaccom- modationbybargainingjustice.Thereareelevenjusticesservinginthethreetermsrandomly selectedforanalysisandthereissizablevariationintheproportionofrequestsunmetbythema- jorityopinionauthor.Whilethisdisparitymayseempuzzlingatblush,thereseveralpotential 108 Figure4.3Nestedbarplotofdistributionoffailedaccommodationbyjusticeand resultantconcurringopinionaction Nestedbarplotwherethelargegraybarsprovidethedistributionoffailedaccom- modationbybargainingjustice.Thesmallnestedbarsaresharesofthelargebars. Thesmallbarsdepicttherateatwhichjusticestakeaconcurringopinionaction whentheythreatenedtodoso(blackbar)duringtheopinionwritingstageand whentheydidnotprovidesuchanadvancewarning(whitebar). explanationsforthispattern.First,thedatapresentedherearedrawnfromonlythreeterms.Jus- ticeO'ConnorjoinedtheCourtin1981,replacingoutgoingStewart.Thus,inthecurrentsample,I capturebehaviorforStewartinonlyasingleterm.AsafreshmanjusticeontheCourt,O'Connor exhibitedgreaterrestraintintheopinionwritingprocess(Hagle1993;BrennerandHagle1996; Maltzman,SpriggsandWahlbeck2000;Wahlbeck,SpriggsandMaltzman1998)andwasless likelytoengageinbargainingintheplace.Asideologicalextrememembers,Marshalland Douglaswererarelyinthemajoritycoalitionandthusinfrequentlyengageddirectlywiththeopin- ionauthor. Thesmallerwhiteandblackbarsaresharesofthelargegraybaranddepicttheratesat whicheachjusticewroteaconcurrencewhentheythreatenedtodosoifnotaccommodated(black 109 bar)andwhentheywroteaconcurrencewithnoadvancewarning(whitebar).Theseshareslook smallsimplybecauseasingleconcurrenceaccountsforseveralinstancesofnon-accommodation withinasinglebargainer-authordyad.Whatismostinterestingisthatwhilenojusticesought retaliationagainstuncompromisingmajorityopinionauthorsmorethantwothirdsofthetime,if ajusticethreatenedtogopublicwiththeirdisagreementwiththeCourt'smajority,theywerenot launchingblankmissiles.Thispatternisconsistentforeveryjusticeinthesample.Forinstance, Marshall never authoredorjoinedaseparateopinionunlesshethreatenedtodosoandJustices StevensandRehnquistfollowedthroughwiththeirthreatswithremarkableconsistency. Thesedescriptivedataindicatethat,althoughabreakdowninthenegotiationprocessin- thecohesivenessoftheCourt'spublishedopinionswritlarge,justicesarefarmoreaptto authororjoinaseparateopinionwhenthetoneofnegotiationwiththemajorityopinionauthoris exceptionallyhostile. 4.6DiscussionandConclusion Thischapterattemptstodiscernwhereandwhyconcurringopinionsoriginate.Addingevidenceto thequalitativeassessmentprovidedbyCorley(2010),Ithatmorethanhalfofallconcurring opinionsareaconsequenceofafailednegotiationsbetweenabargainingjusticeandamajority opinionauthor.Ipushtheanalysisfurtherbybringingdatatoshedlightonboththetypesof requeststhatmostfrequentlyprovokeaconcurrenceandtheresultanttypeofconcurringopinion. Ithatmostconcurrencesthatstemfromfailednegotiationdolittletoaffecttheinterpretation ofprecedentsetbytheCourt'smajorityopinion,thoughlegalandpolicydisagreementsaremore likelytoleadtoconcurrencesthatprovideanadditionalperspectiveforlowercourts,Congress, andthelegalcommunity.Thedatashowthatnotonlyarejusticesmorekeentopursueseparate opinionwritingwhentheyhavesubstantivedisagreements,butthatthetoneofthenegotiation matters.Whenajusticethreatenstowriteseparatelyiftheyarenotaccommodated,theyaremore likelytofollowthroughthantheyotherwisewouldbe. 110 However,thisposesanewpuzzle:somejustices,likeBrennanandChiefJusticeBurger, invokethreatsfrequentlyintheirbargainingmemorandaandyetfollowthroughlessthanhalfthe time.Ingame-theoreticterms,thisisexactlywhatshouldbeexpected.Ifathreatiscredible,the activitythatgivesrisetothethreatshouldchange,andthereshouldbenoneedforactuallycarrying thethreatout.Ofcourse,intheimperfectrealworldsomeonemightnottakeathreatascredible, andignorethethreat,inwhichcasethejusticeshouldprobablyact,ifhe/sheeverwantsherthreats tobetakenseriously.Thatis,itisnecessaryforjusticestoacton some threats,buttheyshouldnot havetoineveryinstance.Accordingly,threateningbehaviorwithoutafollowingenactmentofthe threatshouldbethenormalobservedpattern. Thedataaretoolimitedtofullyunderstandthedynamicsatwork.Infutureresearch,I aimtoexaminetheratioofthreatstoactualfollow-through.Ifsomeonehastokeepactingon theirthreats,thatwouldseemtobeevidencethateitherthepersonbeingthreatened(inthiscase, themajorityopinionauthor)doesnotbelievethethreatiscredible,orthatpersonbelievesthey canabsorbtheficostflofhavingthethreatcarriedout.Inotherwords,thismeansthateitherthe persondoingthethreateningisweak(thecostofthefollowthroughbythatjusticeislow),orthe personbeingthreatenedisverystrong(again,inthattheybelievetheycanwithstandtheficostflof defection).Istheratioofthreats/actionhigherforChiefJustices?Whataboutthemedianjustice orthemedianmemberofthemajoritycoalition?Aretheretrendsovertimeforajustice's credibility?Itwouldalsobeinformativetolookatjusticelearningovertime.Doesthefrequency ofthreatsŒandthreatscarriedoutŒdecreaseasjusticesupdatetheirbeliefsabouttheircolleagues credibility?Howdoesthisaffectthebargainingdynamicsbetweensetsofactors?The sampleprovidesaglimpseintothebehind-the-scenesmechanismsatwork,butfutureresearch shouldextendourunderstanding.Additionally,itwouldbeinterestingtocomparethedifferences inconcurrencesthatareimmediatelydraftedwithoutengaginginbargainingwiththeopinion author(asinChapter3)andthosethatresultfromnon-accommodation(asseeninthischapter). Whilethisanalysisislimited,itstillprovidesimportantinsightsintothedownstreamconse- quencesoftheopinionwritingbargainingprocessbyrevealingpatternsofbehaviorinthebehind- 111 the-scenesbargaininggame.Inparticular,Ithatwhenjusticesrequestachangethatwould alterthepolicyimplicationsofthemajorityopinionandtheauthorrefusestoadopttherequest, theengagedjusticeismorelikelytowriteseparately.Whileaseparateopinionresultsonaverage in40percentofallunaccommodatedrequests,whenpolicynegotiationsfail,concurrencesresult inmorethan50percentofcases.Notably,andperhapsunexpectedly,unaccommodatedpolicyre- questsdonotfrequentlyresultindoctrinalconcurrences.Instead,thesetypesofrequestsaremore likelytogeneratelimitingconcurrencesthatrestricttheapplicationofthelegalrulebythe majorityopinionoremphaticconcurrencesthatmerelyemphasizeaparticularaspectoftheCourt's opinion.Onlyfourintenunansweredchallengestothemajorityopinionauthor'sinterpretationof precedent,statutoryregulations,orConstitutionalprovisions,conversely,generateseparateopin- ions,butthesefailuresinnegotiationaremorelikelytoresultindoctrinalconcurrenceswhereby thebargainingjusticeoptstoendorseonlythemajority'sdispositionaldecisionwhileprovidingan entirelydifferentlegalrationaleforreachingthesameconclusion. Tobeclear,despitethecontentofengagementthebargainingmemorandum,whenama- jorityopinionauthorfailsorrefusestoaccommodatehercolleague,ifajusticepursuesretaliatory actionasMurphy(1964)predicts,themostlikelyoutcomeisanemphaticconcurrencewhich neitherseekstolimitorexpandthereachoftheCourt'smajorityopinion.Indeed,whennegoti- ationovercontentisunsuccessful,wearemostlikelytoobserveconcurringopinionslikeJustice O'Connor'sin Ferberv.NewYork (1982)thatareexpresslynotatoddswiththemajorityopin- ionandserveprimarilyasavehicleforjusticestoclarifytheirpositionrelativetothe majoritycoalition. 112 REFERENCES 113 REFERENCES Black,HenryCampbell.1991. Black'sLawDictionary .St.Paul,MN:WestPublishingCompany. Brenner,SaulandTimothyM.Hagle.1996.fiOpinionWritingandtheAcclimationEffect.fl Polit- icalBehavior 18(3):235Œ261. Corley,PamelaC.2010. ConcurringOpinionWritingontheU.S.SupremeCourt .NY:SUNY Press. Hagle,Timothy.1993.fiFreshmenEffectsforSupremeCourtJustices.fl AmericanJournalofPo- liticalScience 37(4):1142Œ1157. Maltzman,Forrest,JamesF.IISpriggsandPaulJ.Wahlbeck.2000. CraftingLawontheSupreme Court .NewYork:CambridgeUniversityPress. Maveety,Nancy.2005.fiTheEraoftheChoralCourt.fl Judicature 89:138Œ145. Murphy,WalterF.1964. ElementsofJudicialStrategy .Chicago:UniversityofChicagoPress. Ray,LauraKrugman.1990.fiTheJusticesWriteSeparately:UsesoftheConcurencebythe RehnquistCourt.fl UniversityofCalifornia,DavisLawReview 23:777Œ831. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.1998.fiMarshallingtheCourt: BargainingandAccommodationgontheUnitedStatesSupremeCourt.fl AmericanJournalof PoliticalScience 42(1):294Œ315. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.2009. TheBurgerCourtOpinion WritingDatabase .Washington,DC:GeorgeWashingtonUniversity,DeptofPoliticalScience. Witkin,BernardErnest.1977. ManualonAppellateCourtOpinions .St.Paul,MN:WestPublish- ingCompany. 114 CHAPTER5 CONCLUSION 5.1TheSecretiveCreationofU.S.SupremeCourtMajorityOpinions WhenhewasavisitortotheUnitedStatesintheearly19thCentury,deTocqueville(1835)fa- mouslyobservedthatfiScarcelyanypoliticalquestionarisesintheUnitedStatesthatisnotre- solved,soonerorlater,intoajudicialquestion.flAlmost200yearslater,thatstillseemstohold true.Inafundamentalway,courtsareatthecenterofAmericanlife(McGuire2012).Thejudi- ciaryplaysafoundationalroleinpolicyconcernsrangingfromdomesticissues,likeparentaland genderrights,totaxandlabordisputes,tocriminalcases,toimmigrationandenvironmentalregu- lation,to,recently,resolvingquestionsofterrorism.Inrecentterms,togetherwithcomplexissues likecampaignregulationandquestionsabouttheextentofseparationbetweenthechurch andstate,majorsocialissueslikeafveaction,reproductiverightsprotection,andmarriage equalityhaveappearedonthedocketoftheU.S.SupremeCourt.This,ofcourse,isjustasmall samplingoftheissuesthatcomebeforethetopcourtintheUnitedStates. Particularlyatthehighestlevelsofthejudicialsystem,wherejusticeshavegreaterdiscre- tionindecisionmaking,thecourtplaysamajorroleinshapingpublicpolicy.BecausetheSupreme CourthassuchanimportantonAmericanpoliticallife,itisimperativetounderstandhow theCourtoperates,andhowninelife-tenured,unelectedmembersofthebenchmakedecisions.Yet essentiallyuntilthepublicationof TheCollegialGame (Maltzman,SpriggsandWahlbeck2000), mostpoliticalscienceresearchfocusedonexplainingjusticesdispositionalvoteswhileentirely disregardingtheactualoutputorcontentoftheCourt'sopinion(e.g.,RohdeandSpaeth1976; SegalandCover1989;SegalandSpaeth1993,2002).Theintroductionofthestrategicmodel (e.g.EpsteinandKnight1998)generatedapushforsocialscientiststotakeseriouslyhowjustices' policypreferencesareconstrainedbyinstitutionalandcasecharacteristics,legalprecedent,and 115 othercollegialfactorsuniquetotheSupremeCourt.Forthelasttwodecades,then,manyscholars haverespondedbyinvestigatingtheprocessesthroughwhichthesepoliticalactorscreatepublic law(Wahlbeck,SpriggsandMaltzman1998;Spriggs,MaltzmanandWahlbeck1999;Maltzman, SpriggsandWahlbeck2000)throughanexaminationofthingslikethesourcesofopinioncontent (Corley,CollinsandCalvin2011)andbymappingtheideologicallocationofmajorityopin- ions(ClarkandLauderdale2010;Carrubbaetal.2012;LauderdaleandClark2015).Thesestudies haveundoubtedlycontributedtoincreasedunderstandingofthemechanismsthatdrivejudicialbe- havior,yetintheparagraphoftheirseminalwork,Maltzman,SpriggsandWahlbeck(2000, 154)acknowledge,fi[I]tisimportanttonotethatwehavenotmadethestep:explainingthe actualcontentofCourtopinions.Thistaskweleavetothefuture.fl Itispreciselythiscallthatmyresearchseekstoanswer.Inthisdissertation,Icreateacom- prehensivedatasetthatexpandstheSupremeCourtOpinionWritingDatabase(Wahlbeck,Spriggs andMaltzman2009)toprovidearchivaldocumentationofallbehind-the-scenesengagementbe- tweenthejustices,includingeachpageofeverymajorityopiniondraftcirculatedtotheconference duringtheBurgerCourtera(1969-1985).Forthetime,itispossibletoopentheblackbox oftheCourt'ssecretivenegotiationstoplainlyexaminetheopinionwritingprocess,uncoveringif andhowcertainmembersoftheCourtshapejudicialpolicy.Myresearchcombinestextanalysis softwarewithhandcodingofbargainingmemorandaandsubsequentaccommodation,enabling metoshowwhoseexpressedpreferencesareincorporatedintotheopinionbymatchingexplicit negotiationorrequestactionswithchangesinthecontentofthemajorityopiniondrafts.Compre- hensivecontentanalysisofbargainingmemorandarevealboththecontentofthenegotiationand howjusticesframetheirrequests.Combiningthesenoveldatawithinformationoninstitutional features,justicecharacteristics,andcasefactors,myanalysesprovidesnewinsightintothedevel- opmentoflegalopinions.Byprovidinganunprecedentedlookintotheopinionwritingprocess, myworkfurthersourunderstandingofthecollegialinteractionsthatproducetheCourt'sopinion. Additionally,myworkconsidersboththecontextthatgeneratesnegotiationdynamicsand assessestheimpactofanauthor'sdecisionnottomodifythemajorityopinion. 116 5.2DiscerningJudicial Thedataamassedforthisprojectnotonlyallowsforanovelinvestigationintothecollegialdevel- opmentofSupremeCourtmajorityopinionsbypreciselyaccountingforthecontentofnegotiations andtracingthetreatmentofeachrequestthroughtheiterativewritingprocess,butitalsodeparts fromexistingscholarshipbyderivingnewempiricalimplicationsfromdivergenttheoriesofju- dicialInakeysense,myapproachbridgesformaltheoreticmodelsthatpredictthe ideologicaloutcomeofCourtopinionstoempiricalscholarshipthathasfocusedoninstitutional, case,andcollegialfactorsthattheopinionwritingprocess. Theexaminationoftheopinionwritingprocesswasmotivatedbytheoreticalmodelsthat offercompetingpredictionsabouttheoutcomeofthelegalrulecontainedinthemajorityopinion. Themedianjusticemodelcontendsthatthelegalrulewillrethepreferencesofthecentermost justice(Epsteinetal.2005;Hammond,BonneauandSheehan2005;Martin,QuinnandEpstein 2005;Schwartz1992).Itisnotastretchtoassume,consequently,thatthemedianjusticewill wieldinordinatepowerinthedevelopmentoftheopinion,bothintheauthor'streatmentofre- questssoughtfromengagedcolleaguesandinapotentialbargainingjustice'sinitialresponseto themajorityopiniondraft.Borrowingfromthismodel,Ianticipatedthatmedianauthorsshould recognizetheirstrategicpositionandwouldbelesslikelytoaltertheopinion.Ontheotherhand, ifthemedianjusticeoptstobargainwhen(s)heisnotthemajorityauthor,Iexpectedtheopinion writertoworkhardertoaccommodateherrequests.Itfollowedthatthemedianjusticemightrec- ognizeherpowerrelativetotheotherjusticesandactuallybemoreengagedinnegotiationwith theauthorthanotherjustices. Asecondmodelarguesthatbecausedispositionalvotesarequiteconsistentlyed,the legalruleinthemajorityopinionshiftstopreferencesofthemedianmemberofthemajority coalition(Carrubbaetal.2012).Liketheimplicationsfromthemedianjusticemodel,Iderived hypothesesthatpointedtoaparticularlyroleforthispivotalactor.Inshort,ifthis modelisamoreaccuratedepictionofthecreationofthemajority'slegalopinion,Ianticipated 117 evidencewouldamasstohighlightauniquestrategicadvantagethatwouldimpactbothanauthor's accommodationdecisionandajustice'schoicetoengagewiththewriterintheplace. Isituatedthesehypotheseswithinmodelsthatcontainednumerouscontrolsforalterna- tiveexplanationsthathavebornefruitfulresultsinpreviousworks(e.g.Maltzman,Spriggsand Wahlbeck2000;Spriggs,MaltzmanandWahlbeck1999;Wahlbeck,SpriggsandMaltzman1998). Whilelessrigorouslytheoreticallyderived,thesevariablesplausiblycapturecriticalconstraintsto ajustice'spropensitytoact.Forexample,accountingforajustice'sworkloadorthenearnessto theendoftheCourt'stermarereal-worldfactorsthatIexpectedwoulddepressbargainingactivity and/oranauthor'swillingnesstomakeyetanotherchangetothemajorityopinion.Ianticipated thatthelegalandpoliticalimportanceofthecasewouldincreasebargaining,butmightcausean authortobelesswillingtoadoptrequests.Asthetheoriesofjudicialuencepointtopower relativetotheotheractors,Ilikewiseexpectedthatpastcooperativeorotherwisereciprocalrela- tionshipsbetweenjusticeswouldconstrainandshapebehavior.Asaresult,inadditiontothekey interveningvariables,themodelsinChapters2and3aretoincludeahostofcontextual variables. 5.3MajorFindings Fromadescriptivestandpoint,myresearchprovidesinsightintothesometimesspiritedbehind- the-scenesexchangebetweenthejusticesduringtheopinionwritingprocess.Ishowthatjustices engageinlivelynegotiationsoverlegalcontent,butarealsoconcernedwiththeclarityofthe languagecontainedinthemajorityopinion.Onlyinfrequentlydojusticesexplicitlydisagreeabout thepolicyimplicationsofaparticularopinion.Moreover,thecontentanalysisofbargainingmemos indicateahighdegreeofdeferenceandrespectfortheauthor.Mostfrequently,requestsdirected totheauthorarefriendly,neutral,andrespectful.Byandlarge,justicesarereticenttothreatenthe stabilityofthemajoritycoalition. TheempiricalanalysisinChapter2revealsthatwhileauthorsarekeentoadoptthebulkof 118 suggestionsofferedbyotherjustices,thetypeandtoneofrequestsarethemostcriticalpredictors ofanauthor'swillingnesstoaccommodate.Suggestionsforminorrevisionsarealmostalways adopted,whiletherelativelyrarerequeststoalterthescopeofpolicyinthelegalrulearemost oftenignored.Theresultsrevealthatdespitethejustices'seeminglynaturalafforniceness, threateningtojoinorauthoraconcurringopinionalmostalwaysspursthemajoritywriterinto acquiesce. Theevidencepresentedthetwosubstantivechapters(2and3)indicatethatwhen weaccountforboththecontentoftheexchangeandforthepivotalpositionofthebargaining- authoringjusticedyad,theofinstitutional,case,andcontextualvariablesarelesspro- nouncedthanpreviousresearchsuggests(withonenotableexception:myresearchrevealsthat pastcooperationisthemostimportantdeterminantofajustice'sdecisiontoengageinnegotiation withthemajorityopinionauthor).YetIdonotoverwhelmingevidenceforjustonemodelof justiceInstead,theresultspointtoimportantpatternsforboththemedianoftheentire Courtandthemedianmemberofthemajoritycoalition. InChapter3,Ithatwhentheauthoristhemedianofthemajority,sheislesslikelyto receiverequestsforchangefromtheothermembersoftheCourt'sdispositionalmajoritycoalition. Instead,respondingjusticesaremorelikelytoforgoengagementwiththeauthorandinsteadopt towriteaconcurringopinionfromtheoutset.Thiseffectisreversedwhenthemedianmemberof themajoritycoalitionwritestheopinionfortheCourt:respondingjusticesarelesslikelytowrite separatelyandmorelikelytoengagedirectlywiththeauthor.Becausethisjusticeoccupiesthe middleofthecoalition,regardlessofitsideologicaldispersionorheterogeneity,itmaybethatthe authorsimplyoffersanopinionthatisideologicallymoreproximatetoagreaternumberofjustices inthewinningcoalitionandseparateopinionwritingplummetsasaresult. WhileChapter3revealsthatthesepivotalplayersarenomoreorlesslikelytoengagewith theopinionauthorthanotherjustices,Chapter2providesevidencethatwhenthemedianmember ofthemajoritycoalitionasksforchangestothecontentofthelegalion,she'smorelikely tosucceedinseeingherviewstranslatedintolaw.Thisdoesnotholdtrueforthemedianofthe 119 entirebench,however.InsteadIthatthemedianjusticeleadstoalternateinterestingpatterns intheaccommodationdecision.Notably,thoughthefiswingjusticeflisnotaparticularlypowerful bargainer,whensheistheauthor,theresultsshowthatthemedianismorereticenttochangethe opinioncontent.Inshort,Ievidencetosupporttheconjecturethattheremaybemultiple pivotalplayersonintheU.S.SupremeCourtwithunusualontheinstitutionaloutput. Whilemedianauthorsarelessapttoalterthelegalrule,thisauthoralsospursmoreseparate opinionwritingrightfromtheoutset.Whentheyauthorthemajorityopinion,medianmembers ofthemajoritycoalitiondepresstheimmediatecirculationofseparatecompetingopinionsand,as bargainers,theseactorsareespeciallypersuasiveduringtheopinionwritingstage. Bydissectingtheauthor-bargainerdyad,explicitlyaccountingfortheauthor'sdecisionto alterthemajorityopinionattherequestoftheothermembersoftheCourt,Chapter4provides alookintowhenandwhythemajoritycoalitionfracturesafteranauthorelectsnottomodify theopinioncontentafterarequestfromanengagedjustice.Ithatthreatsmightbe especiallypersuasive(asdiscussedinChapter2)becausebargainingjusticesaremorelikelyto writeseparatelyifthey'veindicatedawillingnesstodosoiftheauthordoesnotchangethe majorityopinion.Yetthedescriptiveevidencepresentedinthischaptersuggeststhatjusticesonly taketheadditionalsteptopubliclynoteinternaldiscordinaminorityofcaseswhentheauthor refusestoaccommodaterequestsand,mostfrequently,theresultantseparateopinionsdolittleto harmonthemajoritydecision. 5.4Limitations Althoughtheresearchdesignemployedineachchapterattemptstoutilizeappropriatestatistical modelingtechniquestothevariousstagesoftheopinionwritingprocess,theanalysesare notwithoutlimitations.InChapter2,theiterativeaccommodationdecisioniseffectivelyreduced toabinarychoice.Thismodelingdecisionresultedfromtheobservationthatmostchangestothe opiniondraftswerenotsubsequentlyalteredorremovedinfuturedrafts.Thatis,onceanauthor 120 accommodatedarequest,theaccommodationdecisionstoodinallbuttherarestinstances. Inotherwords,accommodatedrequestsforchangewereremovedfromtheopinioninfewerthan onepercentofallinstances.Additionally,Ifoundthat,likegametheoreticbargainingmodels predict,nearly95percentofbargainingendedintheperiod.Thatis,justicesrarelyextended thebargaininggamebeyondtheinitialmemorandumregardlessoftheauthor'saccommodation decision. TheapproachinChapter3toexaminemajoritycoalitionmembers'initialreactiontothe majorityopiniondraftsimilarlyeschewsawayfromtheiterativenaturetheprocess.Again, thiswaspurposefultodeterminethepropensityofjustices'willingnesstoengageinbargaining insteadofsimplyjoiningthemajorityorwritingseparately,butitcertainlymissesnuance.In particular,whileoccasionallyjusticesoptedtojointheopinion,theywerespurredtobargain afterseeingthereactionsofotheractors. BothChapters3and4examineseparateopinionwritingbehavior,yettheresultsarenot yetcomparabletooneanother.Takentogether,Ithatmorethanhalfofallconcurringopinions arewrittenafterajusticeengagesinbargainingwiththemajorityopinionauthor.Thismeans thatasizableminorityofseparateopinionsarewrittenwithoutpriorengagement,asIin Chapter3.WhileIcanpointtosystematicfactorsthatmotivateajusticetowriteseparatelyfrom theoutsetŒe.g.,thecaseraisesnumerouslegalissues,thejusticeisideologicalremovedfromthe restofthewinningcoalition,orthemedianjusticeistaskedwithauthoringtheCourt'sopinion Œtheanalysisistoolimitedtoascertainwhetherthesetypesofconcurrencesdifferincontent fromtheseparateopinionsthatresultfromabreakdownintheexchangebetweenthebargainer- authordyad.Indeed,thecurrentresultsinChapter3providenoinsightintothecontentofthese concurrenceswhatsoever.Conversely,Chapter4thetypeofconcurringopinionthatstems fromnon-accommodation.However,sincethisanalysisislimitedtothreerandomlyselectedterms itprovidesonlylimited,largelyqualitativeinsightintojusticelearningandtheeffectivenessof threatsovertime,forexample. Finally,duetodataavailability,theentireprojectislimitedtocasesdecidedduringthe 121 BurgerCourt,whichhadinfrequentturnoverinpersonnel.Thestabilityofmembershipmayreduce theleverageofcertaincollegialcharacteristicsŒlikethatofthefreshmanjusticeŒanditdoesnot allowforanexaminationofchiefjusticeeffects.Thecomprehensivenessofthedatasettoinclude everycasedecidedbymajorityopinionensuresthatthisparticularCourtheardalmosteveryissue that any Courtmightface,andIamtheresultsaregeneralizablebeyondtheperiodunder consideration,butastheprivatepapersoftheRehnquistCourtarereleased,itwillbeinformative tonotewhat,ifany,differencesexistinstyleandsubstance. 5.5AvenuesforFutureResearch Theprevioussectionsuggeststhatitwouldbevaluabletosimplyextendtheanalysisofconcurring opinionsthatresultattwodisparatestagesduringtheopinionwritingprocess:intheabsenceof bargainingandafterthefailureofabargainertoachievesatisfactorycompromisewiththemajority opinionauthor.Adescriptiveanalysiswouldshedlightonwhetherthetypesofconcurrencesthat resultinthesestagesaresubstantivelydifferentfromoneanother.Thisinandofitselfwouldbe illuminating,buttogethermyresearchsuggeststhatweknowlittleofwhatconcurringopinions actuallyaccomplish.Todate,theoreticalandempiricalstudiesoftheseseparateopinionstreat thebehaviorasmerelyexpressiveaction(e.g.,Carrubbaetal.2012;Corley2010).YetIthat themerethreatofaseparateopinionisoftenenoughtoinduceactiononbehalfoftheopinion authorandextantworksuggeststhatconcurringopinionsmightcomplianceamongthe lowercourts(HansfordandSpriggs2006).Thisevidencepointstotheneedforanewtheoryof concurringopinions:giventhattheyarecostlytowriteandtheiroccurrencedoesnothavelegally bindingimpactontheinstantcase,whyexactlydojusticeselecttowriteseparately? Infutureresearch,Iplantoexaminethe immediate ofpublishingaconcurring opinionandexplainthepotentialforconcurringopinionstooffer future Whenthebar- gainingperiodendsandarespondingjusticehasmadelittleheadwaywiththeauthor,theresults ofmycurrentanalysislendsupporttothenotionthatajusticemightgainimmediateexpressive 122 fromgoingpublicwithherindividualperspective.Again,themodalcategoryindicates thatmanyoftheseconcurrencesareemphaticinnatureanddolittletoaltertheinterpretationof themajority'slegalrule.Ontheotherhand,twothirdsofconcurrences do seektoofferadif- ferentinterpretationofthecaseandmightthereforeprovideanimportantsignaltolowercourts concerningthepreferencesofthejusticesonthecurrentCourtaswellasprovidinganavenuefor non-compliancewhenthereismembershipreplacement. Forexample,currentempiricalevidencesuggeststhatajusticeismorelikelytowritea concurringopinionastheideologicaldistancefromthemajoritycoalitionmedianincreases(Car- rubbaetal.2012).Descriptivedatarevealconcurrenceson both sidesofthecoalitionalmedian, raisinganunexplainedquandaryconcerningtheofconcurrencesasasignaltolower courts:ifthemedianmemberoftheCourtauthorsaconcurrenceinaprecedent-settingcase,it signalsthatmedianisnotentirelywiththelegalruleestablishedinthemajorityopinion. Lowercourts,then,haveachoicetofaithfullyexecutethelegalruleandriskunfavorablereviewby theSupremeCourtortoadoptthelegalrulemostpreferredbythemedianmemberoftheSupreme CourtbyannouncingadecisionthatisatleastpartiallyinconsistentwiththeCourt'smajorityopin- ionbutwouldlikelygarnerthesupportoftheCourt'smedianandallofthejusticesinthedissent. Asaresult,wewouldexpecttoseeincrementalmovementawayfromtheholdinginthemajority decision. Toanimatethispuzzle,consider Roev.Wade (1973),oneofthemostfamousdecisions tocomeoutoftheBurgerCourt.Inthis7-2dispositionaldecision,themajorityoftheCourt endorsedJusticeBlackmun'smajorityopinionthatcenteredontheConstitution'simpliedright toprivacyandestablishedathreetierapproachtobalanceawoman'srighttochoosewiththe state'slegitimateconcernformaternalandfetalprotection.ModerateJusticeStewartoffereda concurringopinioninthiscasetonotehisbeliefthatthelegalruleoversteppedsinceitdealtwith legislationnotyetbeforetheCourt.Hecounteredbypointingoutthatthereisalegitimatestate interestandthattheseobjectivesarefiamplysuftopermitaStatetoregulateabortionsasit doesothersurgicalprocedures,andperhapssuftopermitaStatetoregulateabortionsmore 123 stringentlyoreventoprohibittheminthelatestagesofpregnancy.flWhileBlackmun'sopinion oughttohavehadtheforceoflawandStewart'soughtnot,thelastfortyyearsofabortionrulings havechippedawayatthemajority'sopinionin Roe andhaveinsteadmorefaithfullyadheredto Stewart'sconcurrence,discardingthethreetierapproach,andconsistentlyupholdingregulations andrestrictionstoabortionaccess. Putplainly,concurringopinionsmaygivelowercourtsalternativeinterpretationsofthe legalrulethateventuallyreplaceoratleastsubstantiallyaffecttheimplementationofthemajority opinion.Futureworkshoulddevelopaformaltheoretic-approachtoevaluatetheroleofconcurring opinionswithinthejudicialhierarchy. 124 REFERENCES 125 REFERENCES Carrubba,Cliff,BarryFriedman,AndrewD.MartinandGeorgVanberg.2012.fiWhoControlsthe ContentofSupremeCourtOpinions.fl AmericanJournalofPoliticalScience 56(2):400Œ412. Clark,TomS.andBenjaminLauderdale.2010.fiLocatingSupremeCourtOpinionsinDoctrine Space.fl AmericanJournalofPoliticalScience 54(4):871Œ890. Corley,PamelaC.2010. ConcurringOpinionWritingontheU.S.SupremeCourt .NY:SUNY Press. Corley,PamelaC.,PaulM.Jr.CollinsandBryanCalvin.2011.fiLowerCourtonU.S. SupremeCourtOpinionContent.fl JournalofPolitics 73(1):31Œ44. deTocqueville,Alexis.1835. DemocracyinAmerica .2003ed.London:PenguinBooks. Epstein,Lee,DanielE.Hoe,GaryKingandJefferyA.Segal.2005.fiTheSupremeCourtDuring Crisis:HowWarAffectsOnlyNon-WarCases.fl NewYorkUniversityLawReview 80(1):1Œ116. Epstein,LeeandJackKnight.1998. TheChoicesJusticesMake .Washington,DC:CQPress. Hammond,ThomasH.,ChrisW.BonneauandReginaldS.Sheehan.2005. StrategicBehaviorand PolicyChoiceontheU.S.SupremeCourt .PaloAlto,CA:StanfordUniversityPress. Hansford,ThomasG.andIISpriggs,JamesF.2006. ThePoliticsofPrecedentontheU.S.Supreme Court .Princeton,NJ:PrincetonUniversityPress. Lauderdale,BenjaminandTomS.Clark.2015.fiEstimatingVPreferencesfromRoll- CallDataUsingConditionalAutoregressivePriors.fl. Maltzman,Forrest,JamesF.IISpriggsandPaulJ.Wahlbeck.2000. CraftingLawontheSupreme Court .NewYork:CambridgeUniversityPress. Martin,AndrewD.,KevinM.QuinnandLeeEpstein.2005.fiTheMedianJusticeontheU.S. SupremeCourt.fl NorthCarolinaLawReview 83(5):1275Œ1322. McGuire,KevinT.2012.Preface.In NewDirectionsinJudicialPolitics ,ed.KevinT.McGuire. 1sted.NewYork:Routledge. Rohde,DavidandHaroldJ.Spaeth.1976. SupremeCourtDecisionMaking .SanFrancisco:W.H. Freeman. Schwartz,EdwardP.1992.fiPolicy,Precedent,andPower:APositiveTheoryofSupremeCourt Decision-Making.fl JournalofLaw,Economics,andOrganization 8:219Œ252. Segal,JefferyA.andAlbertD.Cover.1989.fiIdeologicalValuesandtheVotesoftheU.S.Supreme CourtJustices.fl AmericanPoliticalScienceReview 83:557Œ565. 126 Segal,JefferyA.andHaroldJ.Spaeth.1993. TheSupremeCourtandtheAttitudinalModel .New York:CambridgeUniversityPress. Segal,JefferyA.andHaroldJ.Spaeth.2002. TheSupremeCourtandtheAttitudinalModel Revisited .NewYork:CambridgeUniversityPress. Spriggs,JamesF.II,ForrestMaltzmanandPaulJ.Wahlbeck.1999.fiBargainingontheUS SupremeCourt:Justices'ResponsestoMajorityOpinionDrafts.fl JournalofPolitics 61(2):485Œ 506. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.1998.fiMarshallingtheCourt: BargainingandAccommodationgontheUnitedStatesSupremeCourt.fl AmericanJournalof PoliticalScience 42(1):294Œ315. Wahlbeck,PaulJ.,JamesF.IISpriggsandForrestMaltzman.2009. TheBurgerCourtOpinion WritingDatabase .Washington,DC:GeorgeWashingtonUniversity,DeptofPoliticalScience. 127