Explaining)the)Path)and)Pace)of)Nuclear)Weapons)Programs) ) by) ) Lisa)Langdon)Koch) ) ) ) A)dissertation)submitted)in)partial)fulfillment)

Transcription

1 Explaining)the)Path)and)Pace)of)Nuclear)Weapons)Programs) ) by) ) Lisa)Langdon)Koch) ) ) ) A)dissertation)submitted)in)partial)fulfillment) of)the)requirements)for)the)degree)of) Doctor)of)Philosophy) (Political)Science)) in)the)university)of)michigan) 2014) ) ) ) ) ) ) ) ) ) ) ) Doctoral)Committee:) ) ) Professor)Allan)C.)Stam,)Chair) ) Professor)Robert)J.)Franzese,)Jr.) Professor)James)D.)Morrow) ) Assistant)Professor)Philip)Potter)

2 ACKNOWLEDGEMENTS) ) Ithankmycommitteechair,AllanC.Stam.IbenefittedtremendouslyfromAl s expertiseinthesubjectofnuclearweaponsprograms,insightsintotheoryandcases,and enthusiasmformyresearchquestions.ialwaysenjoyedourdiscussionsandal ssenseof humor,andiamgratefulforhisprofessionaladviceandmentorship.ithankmycommittee members,jamesd.morrow,robertfranzese,jr.,andphilipb.k.potter.theirguidance, questions,critiques,andsupporthavebeeninvaluable.ialsothankrobforhismethods workshopfordissertationwriters,andmyfellowparticipantsinthatworkshopfortheir comments.igratefullyacknowledgefinancialsupportformywork:thealfredmeyer Award,endowedbyDr.Meyer sdaughterverameyer,themargaretdowtowsley Fellowship,andtheRackhamOneOTermDissertationFellowship. IthankChuckShipanfordiscussinginstitutionalautonomywithmeandforhiring meashisresearchassistantforasemester,whichwaswhenistartedworkingwithcox proportionalhazardsmodels.ihavelongbeengratefulforchuck sadvicetofollowmydual interestsofworldpoliticsandamericaninstitutions,andfortheeducationinamerican politicalinstitutionsireceivedinhisclassroomthatinformedmyworkonnuclear agencies.iamindebtedtorobertaxelrodforencouragingmyresearchonterrorismwhen Iwasamastersdegreestudentinpublicpolicy,forsupportingmyapplicationtothe politicalscienceph.d.program,andforadvisingmetojettisonearlierresearchquestions. ii

3 PeterFeaver,T.V.Paul,andWillTobeyhelpedmethinkthroughearlyversionsof mytheoryandaskedmequestionsthatshowedmewhereineededtodothemostwork. JanetBoxOSteffensmeierandBenjaminJonesprovidedindispensablemethodological advice.vipinnaranggenerouslysharedanearlyversionofhisrecentbookwithme.for permittingmetoaccesstheirdataandfortheirtimespentansweringmyfollowoup questions,ithankmatthewfuhrmann,jessical.weeks,williamr.thompson,andphilipp C.Bleek.MyfellowgraduatestudentsMatthewWells,GaryUzonyi,andJessicaSteinberg contributedthoughtfulandthoroughcritiquesofsomeofmyearlyworkonthistopic,and MollyReynoldsgenerouslyofferedStataassistanceandsuggestedthefinaltitle.Ithank Mattinparticularformanyhelpfulconversations.Ireliedontheprofessionalknowledge andexperienceofpoliticalsciencelibrariancatherinemorsecountlesstimes.sofia RosenbergspenthoursofherpersonaltimetranslatingSwedishwritingsintoEnglishso thaticouldpuzzleoutthecharacteristicsofswedishnuclearinstitutions. IamindebtedtoDonaldL.Hafner,whotaughtmeaboutnuclearweaponsstrategy andmissiledefensetechnologywheniwasanundergraduatestudentatbostoncollege. Histeachingandhisidealscontinuetoinspireme.Ithankmyfather,JohnW.Langdon, professorofhistoryatlemoynecollege,forhisinsightfulcommentsonmydraftsandfor thedozensoffascinatingconversationswehavehadaboutnuclearweaponsprograms.his vastknowledgeofpoliticalhistoryisunmatchedbyanyoneiknow.finally,ithankmy wonderfulchildren,audrey,paul,andtimothy,whooccasionallyenjoyedanimpromptu moviesothaticouldfinishupapageortwo,andmylovinghusband,matt,forhispatience duringmyresearchandwriting,forhissupportofmywork,andforhisfriendship. ) iii

4 TABLE)OF)CONTENTS) ACKNOWLEDGEMENTS)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).))ii) ) LIST)OF)FIGURES)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).))v) ) LIST)OF)TABLES)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)vi)) ) LIST)OF)APPENDICES)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)))vii) ) ABSTRACT)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)))viii) ) CHAPTER) ) 1. Introduction)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)1) ) 2. Why)States)Start)Nuclear)Weapons)Programs)).)).)).)).)).)).)).)).)).)).)).)).)).))15) ) 3. The)Path)and)Pace)of)Nuclear)Weapons)Programs:)) CivilZMilitary)Relations)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).))57))) ) 4. The)Path)and)Pace)of)Nuclear)Weapons)Programs:)) Nuclear)Bureaucracy)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)93)) ) 5. Case)Studies)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)132) ) 6. Conclusion)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).))).)).)).)).))180) ) APPENDICES)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)))190) ) BIBLIOGRAPHY)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)).)))211 iv

5 LIST)OF)FIGURES) Figure 2.1 Hazardrateofdecisiontostartnuclearweaponsprogram Hazardrateofdecisiontoacceleratenuclearweaponsprogrambylevelof militarizationofthegovernment Hazardrateofdecisiontoacceleratenuclearweaponsprogram: Independence Hazardrateofdecisiontoacceleratenuclearweaponsprogram: Origin v

6 LIST)OF)TABLES) Table 1.1 Pakistaninuclearweaponsprogrambypaceandstage Decisiontostartnuclearweaponsprogram...27 ) 2.2 Eventhistorymodelofnuclearweaponsprogramdecisionmaking Originsofnuclearprogram Civiliannuclearexperience NuclearStrategicRivalries1938O vi

7 LIST)OF)APPENDICES) ) Appendix 1 YearsofDecisionstoStartNuclearWeaponsProgram DecisionstoChangethePaceofNuclearWeaponsProgram CoxProportionalHazardsModelforStartDecisions ListandDescriptionofCovariatesforEventHistoryModelofNuclear WeaponsProgramStartDecisions ResultsTableforEventHistoryModelofNuclearWeaponsProgramStart Decisions CoxProportionalHazardsModelforPaceDecisions ListandDescriptionofCovariatesforEventHistoryModelofProgram AccelerationDecisions ResultsTableforEventHistoryModelofNuclearWeaponsProgram AccelerationDecisions vii

8 ABSTRACT) ) ) Thisdissertationexplorestwokeyquestionsrelatedtonuclearweaponsprograms: First,underwhatconditionsdostatesdecidetostartnuclearweaponsprograms?And second,oncestatesbeginsuchprograms,whenandwhydotheyvarythepathandpaceof theirnucleardevelopment? Astate swealthandresources,andwillingnesstomakethepoliticaldecisionto beginanuclearweaponsprogram,determinewhetheragovernmentwilldecidetostart downthepathtonuclearweapons.pastscholarshiphasidentifiedseveraldifferent possiblefactorsthatincreaseastate sriskofmakingthedecisiontostartsuchaprogram. Usingeventhistoryanalysis,andincludingeverycountryintheworldstartingin1939,I demonstratethatelementsofthesecurityenvironmentoparticularlywhetherastatehasa nuclearstrategicrivaloandpriornuclearreactorexperiencehavethegreatesteffectona state sdecisiontostartaprogram. Onceastatebeginsanuclearweaponsprogram,whataffectsthepathandpaceof thatprogram?ioffertwomaintheoriesregardingthisquestionandmakethefirst scholarlyattempttomodelthepathsofallnuclearweaponsprogramsthathaveever existed.iuseeventhistorymodelstoconductthestatisticalanalysis. Themodelsconfirmboththeories:First,theweakerastate sciviliancontrolover themilitaryis,thelesslikelyastatewillbetoacceleratethepaceofitsprogram.second, themoreindependentastate snuclearbureaucracyis,themorelikelyastatewillbeto viii

9 accelerateitsnuclearweaponsprogram.finally,iexplorethemechanismsunderlying thesetheoriesinthreehistoricalcasestudies. ix

10 CHAPTER)1) Introduction) ) Thepathandpaceofnuclearweaponsprogramsvarywidelyacrosscountries.Not allgovernmentshavethecapabilityandthepoliticalwilltostartanuclearweapons program.ofthosethatdo,thereisnosetpathtotakeonceaprogramhasbegun.nuclear weaponsprogramsexperienceperiodsofaccelerationordeceleration,andsomeareshut downentirely.whydosomestatesstartnuclearweaponsprograms?oncetheystart,why dotheyvarythepathandpaceoftheirnucleardevelopment?whataccountsforthe differences? Astate sexternalsecurityenvironmentcanaffectthatstate sdecisiontopursuea nuclearweaponsprogram.facingamajorsecuritythreatincreasesthestate slikelihoodof decidingtopursuesuchaprogram.however,oncethatprogramhasbegun,thestatemay makemanydifferenttypesofdecisionsregardingthenatureanduseoftheprogram.the pathastatetakesasitpursuesanddevelopsnuclearweaponsislinedwithwhatiterm pacedecisions:decisionstoaccelerate,ordecelerate,theprogram.thesedecisionscan redefineanuclearweaponsprogram.inthisdissertation,ianalyzeaccelerationdecisions. When,andwhy,agovernmentdecidestoaccelerateaprogramismuchless dependentonthethreatenvironment.instead,theproliferationpathwayisheavily 1

11 moderatedbydomesticinstitutionalfactors.astate scivilomilitaryrelations,andthe independenceofitsnuclearbureaucracy,affectthelikelihoodthatthegovernmentwill decidetopushaprogramforward.moremilitarizedgovernmentsarelesslikelyto accelerateanuclearweaponsprogram,whilestateswithmoreindependentnuclear agenciesaremorelikelytomakeanaccelerationdecision. Itestthesetheorieswithstatisticalmodels,usingoriginaldata.Ithenusecase studiestoillustratethevariousmechanismsthatunderliethetheories. Nuclear)Weapons)Programs)and)Decision)Making) ) Agovernmentmusthaveboththenecessarymotivationandopportunitytodecide tostartanuclearweaponsprogram.theleadershipmusthavethepoliticalwillneededto makethedecisiontoembarkuponacourseofactionthatiscostly,willrequireanunknown periodofyearstoproducethedesiredoutcome,andwillaltertheidentityofthestate.and thestatemustalsohavethecapacitytodeveloporobtaintheneededtechnology,fissile material(nuclearfuel),andexpertise.statesthathavethecapabilitytodevelopnuclear weapons,butlackthemotivationtodoso,willnotstartdownthepathtoanucleararsenal. Onceagovernmenthasmadethedecisiontostartanuclearweaponsprogram,how thatprogramisexecuteddependsonsubsequentdecisions.thestatemaychangethepace oftheprogrambydecidingtoaccelerateordeceleratenuclearweaponsdevelopment.pace decisionshaveneverbeensystematicallyanalyzedacrosscases.inthisdissertation,i identifyaccelerationdecisionsineverynuclearweaponsprogramanddeterminewhat conditionsputastateatgreaterorlesserriskofmakingadecisiontoaccelerate. 2

12 ) Thepaceoftheprogramisadifferentconceptthanthatofthestageaprogramisin. Adecisiontoaccelerateanuclearweaponsprogramis,atthetimeitismade,unobservable. Thesedecisionsaretypicallymadeinsecretbyalimitednumberofpeople,andarenot revealedtothoseoutsideeitherthenuclearprogramortherelatedgovernmentagencies.a governmentthatdecidestoaccelerateaprogram perhapsbydecidingtoinitiateacrash programtoacquireabombinashorttimeframe,ortoincreasetherateofweapons production willhavetowaitmonthsoryearsfortheprogramtoachievethedesired outcomeoftheacceleration.thatoutcome say,atestablenuclearweapon willbethe observableresultofthedecision. Thestageofaprogramismeasuredbythelevelofnuclearweaponsdevelopment thestatehasattained.themostsimplemeasureofstagewouldbetoassignaprogramina givenyeartooneofthreeascendingcategories:whethertheprogramhasacquiredfissile material(uraniumorplutonium),whethertheprogramhasweaponizedthatmaterial(i.e. constructedanuclearbomb),andwhetherthestatehastheabilitytodeliverabombtoa target.inreality,thepathofdevelopmentismuchmorecomplex. Therearemanymilestonesalongthepathtodevelopingnuclearweapons.States establisharesearchanddevelopmentprogram,obtainfissilematerial,buildanuclear reactor,operatethereactor,enrichfissilematerialsoitcanbeusedinabomb,testabomb, anddevelopdeliverysystems.however,itisnotalwaysobviouswhenastatemovesfrom onemilestonetothenext,inpartbecausetheseaccomplishmentsareoftencloselyguarded secrets.further,theprogressionalongthepathofnuclearweaponsdevelopmentisnot alwayslinear.statesmayskipoversomemilestones,perhapsaccumulatingastockpileof nuclearweaponsanddevelopingdeliverysystemsbeforeittestsaprototypebomb. 3

13 ConsiderthecaseofPakistan,whichstarteditsnuclearweaponsprogramin1972.A decisiontostartamilitarizednuclearprogramoccurredafteraciviliannuclearprogram hadbeeninplacefornearlyadecade.twodecisionstoacceleratetheprogramweremade inthe1970s,butitwouldnotbeuntil1983thatpakistanreachedthemilestoneof producingweaponsogradeuranium.thetwounobservabledecisionsaffectedtheprogram, butthedesiredoutcomeswerenotrealizeduntilyearsafterward. Table)1.1PakistaniNuclearWeaponsProgrambyPaceandStage Year Type:Pace Event orstage (1965) ) 1977) ) 1998 ) 2007) Pace) Pace) Stage Stage Stage Stage Pace) Stage Stage Pace) (Nuclearreactorgoescritical) 1 Decisiontostartnuclearweaponsprogram 2 Decision)to)speed)up)the)program 3 ) Decision)to)speed)up)the)program 4 ) Produceshighlyenricheduranium 5 Cold testofnucleardevice 67 Nuclearweaponsdeliverablebyplane 8 Nuclearweaponsassumedtobedeliverablebymissile 9 Decision)to)slow)down)the)program 10 ) Plutoniumproduction 11 Fissiontests 12 Decision)to)speed)up)the)program 13 ) 1Chakma2008,13. 2Spector1990,Kapur1987,Khan Perkovich1999,194. 4Kapur1987,103. 5ReportedbyA.Q.KhaninThe$News(Islamabad)on30May1998,thisdateiswidelyaccepted.Thefirst enrichmentofuraniumatlowlevelsoccurredin Khan2013,185. 7Acoldtestisanuclearweaponstestthat,insteadofusingweaponsgradeuraniumorplutonium,usesanonO weaponsogradematerial(suchasdepleteduranium)atthecore. 8McDonnell Spector1990, Spector1990,104O105;Cheema2010, AlbrightandBrannan Cheema2010, ThomShankerandDavidSangerreportedinaNewYorkTimeson17May2009( PakistanisRapidly AddingNuclearArms,U.S.Says )evidenceofacceleration.theinstituteforscienceandinternational Security,amongothers,reportedinMay2009thatPakistanwasnearcompletionoftwonewplutoniumO producingreactorsatitskhushabfacility.indicationsarethatthedecisiontoacceleratewasmadetwoto threeyearsearlier.thepakistannewsservicereportedon18august2007that,accordingtomilitary 4

14 Giventheseriesofmilestoneslistedonthepreviouspage,thePakistanitimeline appearstoskipoversome,andevenreturnstoonelater.forexample,deliverysystems weredevelopedbeforethenuclearweaponstobedeliveredwerefullytested.fissiontests werenotconducteduntilmorethantenyearsafterthebombswereabletobebroughtto theirtargets.pakistandidnotfollowalinearprogressionofstepstowardnuclearweapons. Pakistandidnotprogresssteadilytowardnuclearweapons,either.Inthe1970s, Pakistanaccelerateditsprogramtwice.Itsloweditspacein1989whenbothcivilianand militaryleadersagreedthatpakistanhadasufficientnuclearcapabilityatthatpointto crediblydeterforeignaggressors,andpossiblyalsotoprovidereassurancetotheunited States,whichwasapplyingdiplomaticpressuretoPakistantostoptheirnuclearweapons program. 14 Mostrecently,in2007,Pakistandecidedtoaccelerateitsprogrambyexpanding itsplutoniumproduction,andbuiltnewfacilitiesforthatpurpose. The)Difference)Between)Nuclear)Pace)and)Nuclear)Posture) Theresultsofthe2007PakistanaccelerationwerenotobservedbyAmerican newspapersuntil2009.intelligenceservicesaresometimesabletoobserveeffectsearlier, butnotalways;inthiscase,satelliteimageryrevealedtheincreasedpaceofworkon facilities,buttheobservationsweremadewellafterthedecisionhadbeenmade.decisions aboutthepaceofanuclearweaponsprogramareusuallynotimmediatelyobservable. Anuclearweaponsprogramconsistsoftheresearchanddevelopmentofweaponry andweaponssystems,andoftheproductionoffissilematerialandtheweapons sources,presidentmusharrafhadissueddirectivestoaccelerateitsprogram,includingacceleratingworkthat wasalreadyunderwayonaheavywaterreactoratkhushab. 14Cheema2010,199;Spector104O105. 5

15 themselves.nuclearweaponsprogramsmaynotevenresultintheacquisitionofnuclear weapons.butifastateacquiresnuclearweapons,militaryandcivilianleadersmaydiscuss whetherandhowtheweaponsshouldbedeployed,andinwhatcircumstancestheymaybe used.tacticaldecisionsregardingdeployment,andstrategicdecisionsregardingnuclear doctrine,aredecisionsabouttheuseoftheweapons,notthedevelopment$orproduction$of theweapons. WhenthedriverofatwoOseatersportscardecidestomakeaturn,thecarwill completetheturninalmostthesamemomentthedriverturnsthewheel.butwhenthe driverofasemitruckdecidestomakeaturn,theentirebodyofthetruckwillnotcomplete theturnforseveralseconds,longafterthedriverhasturnedthewheel.anuclearweapons programismuchmorelikethesemitruckthenlikethesportscar:massive,slowto respond,andwithresultsthatrequirepatiencetorealize. Anuclearposturedecisionismorelikethesportscar.Nuclearpostureistheterm forthewayastatearrangesitsexistingnuclearweaponsthathavebeenproducedbythe nuclearweaponsprogram.thewaytheweaponsaremanaged,thenumberandtypesof weaponsthestatepossesses,andthewaythestatehasarrangedthoseweaponsfor possibleuse,arepartofthestate snuclearposture.ifastatefeelsthreatened,itcan respondtothethreatquicklybychangingitsnuclearposture.forexample,astatecould makeadecisiontofuelitsnuclearbombersandplacepilotsonstandoby,andthatdecision couldbeexecuted,andpotentiallyobserved,withinamatterofhours. Respondingtoathreatbymakingadecisiontoaccelerateanuclearweapons program,ontheotherhand,wouldnotbringaboutimmediate,usefulresults.these decisionsmayincurveryhighcostsandareimplementedwithinbureaucraticstructures 6

16 thatrequiretimetoadapt.aprogramdecisionmayrequirebuildingofnewfacilities,hiring newpersonnel,expandingapartnershipwithaforeigngovernment,orfindingnewsources ofmaterialorequipment.decidingtoincreaseworkonanexistingprojecttomoveupthe timeline,forexample,ortoexpandtheprogramfromuraniumbombstoplutoniumbombs aswell,wouldbelongoterm,highlysecretinitiativesthatwouldtakemonthsoryearsto cometofruition. The)Role)of)the)Executive) ) ThetwokindsofdecisionsIanalyzeinthisdissertationareprogramstartdecisions andprogrampacedecisions.howarethesedecisionsmade?whohastheauthorityto makeadecisiontostartanuclearweaponsprogram?andwhomakespacedecisionsonce anuclearweaponsprogramisunderway? Thedecisiontostartanuclearweaponsprogramisapoliticalone,ultimatelymade bytheheadofstate.inmostcases,theheadofstatehasmadethedecisionalongwitha smallnumberofcloseadvisors.subsequentdecisionsregardingthepaceoftheongoing programtypicallyalsorequireauthorizationbytheheadofstate.becausepoweris concentratedintheexecutive,decisionmakingisoftenattributedtothepersonaltraitsof individualleaders.whena strong presidentisobservedadvancinghisnuclearweapons program,ora hawkish primeministerdecidestoexpandtheproductionoffissile material,theassumptionmaybemadethatnuclearpolicymakingdependsonpersonality andleadershipstyle. However,leadersdonotmakedecisionsinavacuum.Domesticactors,likethe military,thenuclearagency,oralegislaturethatcontrolsthepursestrings,mayinfluence 7

17 theheadofstate.internationalactors,suchasastate srivals,orbenefactorswhoimpose pressureonthestatetoceasenucleardevelopment,mayalsoexertinfluenceonthe executive.attributingdecisionmakingtothepersonalcharacteristicsuniquetoeachleader isnotonlytheoreticallyunsatisfying,italsoignorestheinstitutionalstructuresand externalenvironmentthatshapealeader sperceptionsandbeliefs. Inordertomakeadecision,theheadofstatemustobtaininformationanddevelopa setofpossiblepolicyoptionsfromwhichtochoose.stateorganizationsandagencieshave opportunitiestoshapetheheadofstate sunderstandingofthestrategicenvironment,the utilityofnuclearweapons,thefinancialcosttothestate,thepotentialeffectonthestate s internationalreputation,andamultitudeofotherfactorsthataffectdecisionmaking. Militaryleadershipmayhaveadifferentviewpointthanscientificleadership.Agroupthat hasgreaterinfluenceovertheexecutivehasgreateropportunitytoinformandpersuade. Sometimes,anuclearweaponsprogramdecisionismadewhenanewpresidentor primeministercomestopower.someheadsofstatehavestrongfavorableorunfavorable attitudesregardingnuclearweapons,andmayseizetheopportunitytochangestatepolicy upontakingoffice.butalthoughachangeinleadershiphasattimesprecededastart decisionorpacedecision,itisnotacommonorwidespreadcausalfactorleadingtoeither typeofdecision.casesinwhichaheadofstatehadastrongpersonalstakeinthenuclear weaponsprogramareparticularlyinterestingandthuswellknown,butitismoretypical fortheexecutivetorelyuponadvisors,agencies,andotherorganizationstosuggestasetof policychoicesandthentoadvocatefortheirpreferredpolicieswithinthatset.indeed,even thoseexecutiveswhohadexceptionallyhighlevelsofpersonalinvolvementinnuclear affairsstilldidnotoperatealone. 8

18 Becauseleadersareinfluencedbytheirpoliticalandglobalenvironments,andby thepeopleandinstitutionstheyinteractwith,wecanfindpatternsinthecircumstances thataffectwhetheradecisionismadetostartanuclearweaponsprogram,andinwhat happensafterward.eventhoughindividualheadsofstateareunique,andhaveunique personalities,lifehistories,andleadershipstyles,therearecommonsituationsand conditionsthattendtomakeleadersmoreorlesslikelytomakecertainkindsofnuclear weaponsprogramdecisions. Modeling)Decision)Making) Inthisdissertation,Ideveloptheoriesofdecisionmakingspecifictoeachoftwo typesofnuclearprogramdecisions:startdecisionsandpacedecisions.iusestatistical modelsandcasestudiestotestthesetheories.first,iuseeventhistorymodelstotest whethercertainfactorsarelikelytoincreasetheriskofstartdecisions,andwhichfactors arelikelytoincreasetheriskofpacedecisions.ifindthatthecircumstancesthataffectthe riskofadecisiontostartaredifferentfromthosethataffecttheriskofadecisionto accelerate.later,iexplorethesedynamicsinthreecasestudies:pakistan,brazil,andthe UnitedStates. Becauseadecisiontostartoraccelerateanuclearweaponsprogramdoesnot immediatelyproducevisibleresults,andbecausethephysicaloutcomeiseasiertoseethan thedecision,knowledgeofthedecisionusuallycomesonlyafterward.determiningwhen thedecisionwasmadecanbedifficult,asthesedecisionsareoftenmadeinsecret. Likewise,determiningtheprecisecausalinfluencesthatunderliedecisionmakingisa 9

19 difficulttask.hansborn,batesgill,andheinerhänggiwrite: Despiteitsimportance,domesticnuclearweapongovernance issparselyresearched.thisislargelybecauseresearchinthis highlysensitivepolicyareaishamperedbysecrecyinall possessorstates Becauseofthisdifficulty,thereisdisagreementovertheprecisedatesofsomeofthe moresecretivestartandpacedecisions.insomecases,thedateofthedecisionwas documenteddirectlyorbykeywitnesses.butinothers,theformaldecisionislessclear. Throughextensivecaseresearch,includingbothgovernmentdocumentsand secondarysources,andusingnewinformationthathasbecomeavailableonlywithinthe lastfewyears,ihavedevelopedanupdatedlistofalldecisionstostartnuclearweapons programs.thislistofcountriesanddatesimprovesuponpreviouswork. 16 Idonotrelyon publicstatementsmadebygovernments,asastatemayclaimtoadheretonuclearnono proliferationidealswhenitisactuallypursuingaclandestinenuclearweaponsprogram. 17 Usinganeventhistorymodel,Iidentifytheconditionsunderwhichstatesareputat higherorlowerriskofstartinganuclearweaponsprogram.thistypeofstatisticalmodelis usedtoestimatetheriskofsomeeventoccurring.inthismodel,theeventinquestionisthe decisiontostart. Acomprehensivelistofpacedecisionshasneverbeforebeencompiled.Iused primarysourcedocumentswhenpossible,alongwithsecondarysources,toidentifypace decisionsforeachofthetwentyosevenstatesthathavestartedanuclearweaponsprogram 15Bornetal2010,4. 16AtablecomparingthecountriesanddatesIhaveidentifiedwiththosethatotherscholarshaveidentified appearsinappendix1. 17Thereareseveralinstancesofgovernmentsmisleadingdomesticandinternationalaudiencesinthisway. OneexampleisPakistan sproposalforasouthasiannuclearweaponsfreezoneinnovember1972,just monthsafterhavingmadethedecisiontopursueanuclearweaponsprogram. 10

20 orpossessednuclearweapons. 18 Toqualifyasapacedecision,adecisionhadtobemadeto accelerateordecelerateanuclearweaponsprogram,andthus,posturedecisionsarenot included. 19 Iwasabletofindinformationabouteverystate,evenininstanceswhere thoroughcasehistorieshavenotyetbeenwritten. Usingthisnewdatasetofdecisions,Ionceagainemployaneventhistorymodel.In thiscase,thedecisiontoaccelerateanuclearweaponsprogramistheeventinquestion.i analyzetheconditionsunderwhichgovernmentsareatgreaterorlesserriskofmaking accelerationdecisions. StatisticalmodelingandcasestudiesprovideempiricalsupportforthetheoriesI outlinebelow.onecouldarguethatcertaindecisionsseemtohavebeenbroughtaboutby veryspecificcauses,orthateachdecisionmaybesituatedinauniquehistoricaland culturalcontext.idonotclaimthatthetheoriesiproposeexplaineverynuclearweapons programdecision.whileitistruethatsomedecisionsmaybeexceptional,developinga theoryofdecisionmakingisanexerciseinseekingoutcommonfactorsthatsystematically affectthelikelihoodofadecisionbeingmade. Structure)of)the)Dissertation) Why$States$Start$Nuclear$Weapons$Programs$ ) Inorderforastatetodecidetostartanuclearweaponsprogram,twoconditionsare required:motivationandopportunity.theleadershipmusthavepoliticalwilltomakethe 18Ukraine,Belarus,andKazakhstaninheritedtheirnuclearweaponsintheaftermathofthecollapseofthe SovietUnion.Uponbecomingindependentstates,eachofthethreegovernmentshadtodecidehowto managethesovietnuclearweaponsandfacilitiesleftontheirsoilthattheynowsuddenlypossessed. 19ThefulllistappearsinAppendix2. 11

21 decision,andthebeliefthatthestatehas,orwillbeabletobuild,thecapacitytodevelop andproducenuclearweapons. ) Statecapacitytopursueanuclearweaponprogramiscomprisedofwealth,energy, materials,andknowledge.manystateshavethecapabilityneededtopursuenuclear weapons,butchoosenotto.astatemustalsobewillingtoproceeddownthecostlyand controversialpathtothebomb.threekeycategoriesofmotivatorsmaydrivestatesto startanuclearweaponsprogram:security,domesticpoliticalinterests,andnorms. 20 Asthe casehistoriesofnuclearweaponsoseekingstatesindicatethateachofthesemotivatorshas ledvariousstatestomakethedecisiontostartaprogram,recentscholarshiphasused quantitativemethodstosystematicallyanalyzewhichmotivatorsaffectdecisionmaking. 21 InChapter2,Iexplainwhichoftheseconditionsaffectastate sriskofmakingadecisionto startanuclearweaponsprogram. Program$Pace$and$Civil=Military$Relations$ ) Onceanuclearweaponsprogramhasbeeninitiated,thepaththatprogramtakesis dottedwithdecisionsregardinghowtoproceed.withsomeexceptions,securityconcerns arenolongertheprimarymotivatorforhowtoconductaprogram.instead,domestic institutionalcharacteristicsdeterminethelikelihoodofprogramexpansion. TheliteraturehasofferednotheoryofcivilOmilitaryrelationsandthepaceof nuclearweaponsprograms.strongorweakciviliancontrolofthemilitaryhasbeenstudied withregardtocommandandcontrolsystems,thedelegationofauthoritytousenuclear 20Sagan SeeinparticularSinghandWay2004,JoandGartzke2007,Fuhrmann2009,Kroenig2009,andBleek

22 weapons,andthesecurityandpotentialforaccidentallaunchofnuclearweapons. 22 But therehasbeennosystematicexaminationofciviliancontrolofthemilitarywithregardtoa program spath. InChapter3,Iarguethatstateswithweakerciviliancontrolofthemilitarymake differentpacedecisionsthanstateswithstrongerciviliancontroldo.themoremilitarized thegovernmentis,thelowertheriskthatthegovernmentwilldecidetoacceleratea nuclearweaponsprogram.iproposesixreasonswhythisistrue:militarieshave traditionalinterests,worktobuildarsenals,areinwardlyfocused,strivetominimize civilianinterferenceinmilitaryaffairs,desireprestige,andbelievethatnuclearweapons areunusable.ingeneral,militariesaremorelikelytopursuetheirownorganizational interestsratherthantheinterestsofthestateasawhole. Program$Pace$and$Agency$Independence$and$Origin$ ) Thenuclearbureaucracymanagesandoperatesanuclearweaponsprogram,and providesinformationontheprogramtostatedecisionmakers.despitetheimportanceof thenuclearagency,notheoryofbureaucraticpowerandnuclearweaponsprogramshas previouslybeendeveloped.astherehadbeenno compellingmeasuresofbureaucratic powerandautonomy, quantitativeanalysisofthecharacteristicsofnuclearagencieshad neverbeendone. 23 InChapter4,Iproposeatheoryofnuclearbureaucraticinfluenceonthe executive.usingoriginalmeasuresofbureaucraticpowerandautonomy,idemonstratethe effectsofagencyindependenceandagencyoriginontheriskofastatechoosingto accelerateitsnuclearweaponsprogram. 22SeeFeaver1993andSagan Sagan

23 ) Nuclearweaponsprogramsarehighlysecret,extremelytechnical,andcloakedin themysticismoftheirextraordinarydestructivepower,limitingtheunderstandingof programoutsiders.thedearthofcomprehensibleinformationavailabletodecisionmakers heightensthepoweroftheexpertswithinthenuclearbureaucracy.domesticnuclear agencieswithahighlevelofindependencewillbeabletoinfluencepolicymakingtoa greaterdegreethandomesticnuclearagencieswithlessindependence,becauseheadsof stateselectfrompolicyoptionsprovidedtothemfrombelow.themoreindependenta nuclearagencyis,themorelikelyitistobeabletoinfluencedecisionmakerstoadoptits preferences.thekeycharacteristicsdetermininghighinfluenceincludehavingtheearof theheadofstate,havingindependencefromthecabinetorotherexecutivebureaucratic apparatus,experiencinglimitedstateoversight,andreceivinglegalcodificationofsomeor alloftheseinstitutionalfeatures. Case$Studies$ InChapter5,Iexaminenuclearweaponsprogramdecisionsinthreecountries: Pakistan,Brazil,andtheUnitedStates.Statisticalmodelingallowedmetotestbroad theoriesofstatedecisionmakingbehaviorbyanalyzingallstatesthathaveeversought nuclearweapons,orforstartdecisions,byanalyzingallstatesinexistenceduringthetime periodinquestion.casestudiesprovidethebenefitofaddingcontextanddepthtobroad theorizing.ineachofthethreecases,ishowhowdomesticinstitutionalfactorsandthe externalsecurityenvironmentplayedapart ordidn t inkeynuclearweaponsprogram decisionsandoverthelifeoftheprogram. IconcludethedissertationinChapter6withareviewofthekeyfindings,a discussionoftheimplicationsofthisresearch,andalookaheadatfutureresearch. 14

24 CHAPTER)2) Why)States)Start)Nuclear)Weapons)Programs) Whydoonlysomestateswiththecapacitytopursuenuclearweaponsactually embarkuponaweaponsprogram?isthedecisiondrivenbythedesireforthekindof internationalstatusthatcomeswithmembershipinthe nuclearclub?dodomestic politicalinterestsleadastatetopursuenuclearweapons?ordoastate ssecurityconcerns ultimatelyprovokethedecisiontostartaprogram? Thedecisiontostartanuclearweaponsprogramisapoliticalone.Typically,the headofstateauthorizesamilitarizednuclearprogramthatutilizesnucleartechnologyto producefissilematerialanddevelopbombs.rarely,thedecisiontostartaprogramhas amountedtoanordertopurchasean offtheshelf capabilityfromotherstatesoronthe blackmarket.thatapproachhasnotyetprovensuccessfulforanystatethathasattempted it. 1 Regardlessofthedetailsinvolvedinthescaleortypeofthenuclearweaponsprogram, itistheexecutivewhoultimatelyhasmadethedecisiontoweaponize. Pastpersonalexperiences,andindividualpersonalityandcharacteristics,shapethe headofstate sworldview.somepresidentsorprimeministersaremorehawkishand inclinedtowardaggression;othersfavordiplomacyandthepursuitofpeaceover 1LibyaandAustraliabothinitiallyshowedinterestinsimplypurchasingnuclearweapon.Neithereffortwas successful.australiaendedtheirpursuitofnuclearweapons,andlibyaturnedtowardpurchasingthe technologyandmaterialsneededtoconstructtheirownproductionfacility. 15

25 militarizedaction.buttheexecutivedoesnotmakedecisionsinavacuum noteventhe oftensecretandmoreisolateddecisiontopursuenuclearweapons.heorshemaybe influencedbymanydifferentpeople,institutions,andideas.theheadofstatemightbe concernedabouttheriseofarivalstate,orcouldbethinkingstrategicallyaboutpotential futureconflicts.someleadersholdlengthydebatesinparliament,orwithcabinetmembers orsmallergroupsofadvisors.historianshavedetailedthetrustedfriendshipsthathave existedbetweensomeoftheseheadsofstateandthementheyhavenamedtoheadtheir nuclearprograms. 1 Becauseleadersareinfluencedbytheirpoliticalandglobalenvironments,andby thepeopleandinstitutionstheyinteractwith,wecanfindpatternsinthecircumstances thataffectwhetheradecisionismadetostartanuclearweaponsprogram.eventhough individualheadsofstateareunique,andhaveuniquepersonalities,lifehistories,and leadershipstyles,therearecommonsituationsandconditionsthattendtomakeleaders moreorlesslikelytomakethiskindofdecision. Theseconditionscanbecategorizedintermsofmotivationandopportunity.Both mustbepresentforthedecisiontobemade.thestatemusthavethecapacitytodevelopor purchaseanuclearweaponsprogram,oratleastmustholdthebeliefthatitwillsoonbe abletodevelopthenecessarycapacity.andtheremustexistpoliticalwilltostartdownthe pathtonuclearweapons.notallstatesthathavethecapacitytostartaprogramwilldoso, becausenotallcapablestatesalsopossessthemotivationtoproceed. ) 1Ascientist sintenttodevelopweapons,whilepotentiallyinfluential,isnotapoliticaldecision.although somenuclearscientistshavehadagreatdealoffreedomtopursuetheirownagendas,itisthepolitical leadershipofacountrythathastheauthoritytodecidetodevelopnuclearweapons. 16

26 Capacity) Inordertodevelopanuclearweaponsprogram,astateneedswealth,energy, materials,andknowledge.thethresholdsfortheknowledgeandfundingneededtodesign asingleworkingnuclearweaponarerelativelylow.astate scapacityforstartinganuclear weaponsprogramiscenterednotinthefinishedweapon,butinthebuildingandoperation ofthefacilitiesandtechnologiesneededtoproducefissilematerialand,later,delivery systemsfortheweapons. Becauseofthis,nuclearweaponsprogramsareextraordinarilyexpensive. ProducingthefissilematerialneededforatomicbombsisoneofthehighestOcostelements ofanuclearweaponsprogram.internationalrestrictionsontheproductionandtradeof fissilematerialhavemadeitdifficulttoobtainthecomponentsneededtobuildabomb. Whetherastatedecidestoproduceuraniumorplutoniumweapons andinmanycases, bothoptionsareexplored thecostofenrichingorprocessingthematerialisenormous. ConsidertheManhattanProject:approximately80%ofthetotalfundingfortheManhattan ProjectwasallocatedtoOakRidge,the secretcity establishedintennesseethatproduced thefissilematerialfortheu.s.nuclearweaponseffort. 2 In2014dollars,thattotalsmore than$16billion.evenwithadvancesinuraniumenrichmentandplutoniumreprocessing technology,thecostofproducingweaponouseableuraniumorplutoniumstilldominatesa nuclearprogram sbudget. Operatingoneofthesefacilitiesrequiresanenormousamountofelectricity. Whetherastateisoperatingnuclearreactorsandreprocessingplantstoproduce plutonium,orisusinggaseousdiffusionorcentrifugecascadestoenrichuranium,ittakesa 2Schwartz

27 greatdealofenergytooperatetheplantsthatproducefissilematerials.astatemusthave enoughelectricitytodothis. Astatecandevelopitsownplants,facilities,andreactors.Alternatively,itcanbuy themfromotherstatesandcompanies,oritcanobtainassistanceandbuildthemtogether withanotherstate.whileastatedoesn thavetohavetheabilitytobuildallofthis infrastructureitself,itdoeshastoatleasthaveenoughmoneytopurchasethose technologies,andtheinternationalrestrictionsondoingsohavecontinuedtogrowover thelastfivedecades.furthermore,thenonproliferationregimehasevolvedovertimeto moreeffectivelypreventthespreadofthekindoftechnologydeemed dualuse, oruseable forbothcivilianandmilitarypurposes. 3 Ifagovernmentdesiresanuclearreactoror reprocessingplanttoproduceweaponsogradematerial,thebestwaytoensuresuccessisto buildcovertindigenousfacilitiesinordertoavoidscrutinybyinternationalinspectors. 4 Knowledgeisthelastcomponent.Itiscertainlytruethatlessknowledgeable programsmayproceedmoreslowlyandsuffermoresetbacksandbreakdowns,andthat theirweaponrywill,atleastatfirst,belesssophisticated.butstateswithlessexperiencein nuclearscienceandengineeringhavehistoricallysenttheirscientistsabroadtolearn,or haveestablishedpartnershipswithforeigninstitutesanduniversitiesthatallowtheir nuclearexpertstovisitthelessexperiencedcountryinordertoeducatethosescientists. Overall,arelativelackofscientificknowledgeregardingnuclearweaponswillnotbemuch ofahindrancewhenleadersweighthedecisiontostartaprogram. 3Montgomery Afewexamples:NorthKoreadidthisformorethanadecadebeforewithdrawingfromtheNPTin2003.See Hecker2010.SyriabuiltacovertnuclearreactorwithNorthKoreanassistance.SeePollack2010.Taiwan beganbuildingsmalloscalesecretfacilitiesforplutoniumextraction.seespector1990.argentinabegan secretlyconstructingauraniumenrichmentfacilitytoavoidiaeasafeguards.seereiss1995.brazilbuilta secretnuclearreactordesignedtoproduceplutonium.seereiss

28 Itisimportanttonotethatstatesdevelopevenindigenousnuclearprogramswitha greatdealofassistancefromothernuclearpowers.inbuildingnuclearprograms,states havealwaysreceivedmaterialandscientificaidfromothergovernments,andfrom corporations,whicharesometimesprivatebutmostoftenenjoystatesupport.nostatehas pursuedanuclearprogramentirelyalone,andthusnostatehastopossessthecomplete capacitytodevelopsuchaprogramwithoutassistance. 5 Motivation) Manystateshavethecapabilityneededtopursuenuclearweapons,butchoosenot to.astatemustalsobewillingtoproceeddownthecostlyandcontroversialpathtothe bomb.sagan s1996article ThreeModelsinSearchofaBomb identifiesthreedifferent theoreticalmotivatorsthatmaydrivestatestostartanuclearweaponsprogram:security, domesticpoliticalinterests,andnorms.asthecasehistoriesofnuclearweaponsoseeking statesindicatethateachofthesemotivatorshasledvariousstatestomakethedecisionto startaprogram,usingquantitativemethodstosystematicallyanalyzetheconditionsunder whichstatesmakethesedecisionscanhelpdeterminewhatmotivatorsarethemost important. Security$ Statesmaybemotivatedbynationalsecurityfactorstobeginanuclearweapons program.thelogicofbalancing theideathatastateinananarchicinternationalsystem 5Thevastmajorityofnuclearassistanceprovidedbyonestatetoanotheriscivilianassistance,whichmay consistoftransfersoftechnology,material,orknowledge.inafewcases,stateshavereceivedsensitive nuclearassistance,whichisassistanceinthedesignandconstructionofnuclearweapons,thetransferof weaponsogradefissilematerial,orassistanceinbuildingfacilitiesthatcanproduceweaponsogradefissile material.formoreonthistopic,seefuhrmann2009andkroenig

29 willseeknottomaximizepower,buttobalanceagainstastrongerpower hasledscholars totheorizethatastatefacinganopponentpossessingastrongerconventionalthreatmay seektochangetheasymmetryofpowerbypursuingnuclearweapons. 6 Astatewith nuclearrivalsmaydecidetoproliferatetoeventheplayingfield 7 ;Sagan(1996)offersthe term strategicchainreaction, whereeachstatethatdevelopsnuclearweaponsinspires otherstatestoproliferatetofendoffthenewthreatofnuclearattack.alternatively,arival ofanuclearstatemayshyawayfromproliferationoutoffearofsparkinganarmsrace. 8 Notallsecurityconcernsareequal.Anemergingthreatfromaregionalrivalmay motivateastatetomakepolicydecisionsthatwouldnotbeconsideredifamilitarybuildup wereobservedinanally,orinarelativelyweakneighbor.mostrecentscholarshiponthe threatenvironmenthasattemptedtoseparateoutwhichelementsofthesecurity environmentmaypromptstatestodecidetopursuenuclearweapons.amongthese elementsaremeasuresoftheintensityofthesecuritythreat,usuallycalculatedbasedon themilitarizeddisputesastateisinvolvedin,andwhetherastateispartofalongoterm rivalry. 9 WhetherathreatemergesfromanuclearpowerornonOnuclearpowershouldalso mattertoastatethinkingaboutwhethertostartitsownnuclearprogram. Securityguaranteesfrompowerfulalliesmaymitigateastate ssecurityconcerns. UntilthecollapseoftheSovietUnionin1991,thepostOWorldWarIInuclearweaponsera coincidedwiththecoldwar,aperioddominatedbythe superpowers oftheunitedstates andsovietunion.manystatesalignedmilitarilywithoneofthetwopatronstates,whether forprotection,forassistanceinimprovingthesmallerstate smilitaryandweaponry,or 6Potter1982,Kapur2001,Quester1973, ForanandSpector BuenodeMesquitaandRiker1982,Mandelbaum SinghandWay2004,JoandGartzke2007,Kroenig2009,Fuhrmann2009,Bleek

30 both.mostrelevanttothequestionofwhystateschoosetostartnuclearweapons programsisthepresenceorabsenceofasecurityguaranteefromtheunitedstatesor Russia.Ifeithernuclearweaponspowermaintainsaformaldefensepactwithaprotégé, thatprotectionmayobviatetheneedfortheprotectedstatetoacquireitsownnuclear weapons. Domestic$Institutions$ Domesticinstitutionsmaydampenorheightenastate smotivationtopursuea nuclearprogram.similartypesofgovernments,whichsharecertaininstitutional characteristics,mayrespondtothequestionofnuclearweaponsinsimilarways. Thedistributionofpowerandauthoritybetweenacountry smilitaryandcivilian institutionscanimpactnationalsecuritypolicymaking.docivilomilitaryrelationsaffecta governmentaldecisiontobeginorforgoanuclearweaponsprogram?militarized governments thoseinwhichcivilianshaveweakercontrolofthemilitary maybemore likelytochoosetopursuenuclearweapons.therearemanyreasonswhythismaybeso. Forexample,possessingnuclearweaponsallowsamilitarytowieldtremendous destructivepowerthatcoulddeterpotentialenemies.anucleararsenalthatservesasa credibledeterrentcouldfreethemilitarytopursueexpansionistgoalsconventionally, believingthatnoadversarywilldarerespondtoanattack. 10 Andmilitariesmaydesirethe prestigethatnuclearweaponsconfer.ifthemilitary sinfluenceonpolicymakingis 10SeeChapter5foradiscussionofhowPakistan smilitaryseemstohaveusedjustthislogicindecidingto enterkargilin

31 relativelystrong,astatemaybemorelikelytopursuemilitaryinterestsandthusmore likelytomakethedecisiontoaddnuclearweaponstoitsarsenal. 11 Moredemocraticgovernmentsmaybemoreresponsivetopublicopinionandmay shyawayfromnuclearweaponswhenthepopulationopposesthem.autocraciesmaybe morelikelytoembracenuclearweaponsprogramsbecausetheconcernsoftheircitizens cannotbeexpressedinfreeelections. 12 Publicopinionopposingnuclearweaponsseemsto havebeensalientenoughtoswedishpoliticiansforswedentoscuttletheirplanstostudy andultimatelyevendevelopmilitaryusesfornuclearpower;inothercountries,public oppositionhasnotbeenlargeenoughorimportantenoughtobedecisive. 13 Asthereis currentlynouniversalmeasureofpublicfeelingregardingnuclearweaponsacross countries,thisdynamiccannotbedirectlytested. Itisalsopossiblethatdemocraticregimeswouldbemorelikelytouseanational nuclearprogramtoappeasenationalistpopulations. 14 Nuclearweaponsaresymbolicof nationalstrength,ability,andpower:qualitiesthatdemonstratesuperiority.perkovich (1999)arguesthatIndia sdemocraticregimesawthenuclearprogramasawaytoappeal toandmanagenationalistsentiment.yetstatesthatfallonthemoreautocraticendofthe spectrumofregimetypemayalsohaveaninterestinusingnuclearweaponsthatstems fromnationalism anautocraticgovernmentmightuseanuclearprogrameithertoquiet nationalistforcesortowhipupnationalistsentimenttofurtherbolstertheregime. 11ThisisonlyaverybriefoutlineofhowcivilOmilitaryrelationsmayimpacttheinitialdecisiontostarta nuclearweaponsprogram.seechapter3foradiscussionofmilitaryinterests/prioritiesandwhether militariesaremoreorlessaggressivethanciviliangovernments. 12Chubim Jonter2010,Cole1997.TheSwedishelectoratewasgenerallyinfavorofnuclearpursuitsintheearlyand mido1950s,asdidtheprimeministeratthetimeandleadingsocialdemocrats;publicsentimentchangedto oppositionintheearly1960s. 14Perkovich1999,Snyder

32 PerkovichalsopointsouttwoothercharacteristicsthatIndiaheldbecauseofits democraticcharacterthatcouldhaveloweredthechancesthatindiawouldpursuenuclear weapons,oratleastslowedindia sprogress.nuclearweapons,someoftheelitebelieved, wereatoddswithindia shistory:nonviolencehadbeenakeyfactorinachievingindia s hardowonindependence.inaddition,india,asademocracy,couldnotdevoteresourcestoa nuclearprogramattheexpenseofthewellobeingofitscitizenry. 15 Democraciesmaybe morehesitanttoembarkuponthecostlypathtonuclearweapons,knowingthatthe sacrificesnecessarywillnotbetoleratedbytheelectorate. Asmeasuredbyregimetype,then,similardomesticinstitutionsareunlikelytohave acommoninfluenceonthedecisiontostartanuclearweaponsprogram.thekindsof domesticinstitutionsthataremorelikelytoimpactthatdecisioninpredictablewaysare difficulttomeasure.weknowthatthepersonalrelationshipsbetweenheadsofstateand scientistsmatter,butfriendshipscannotbequantified,compared,andassessedacrossa populationofcases.thepowerofthenuclearbureaucracyshouldinfluencetheheadof stateincaseswhereaciviliannuclearprogramprecededamilitarynuclearprogram,but manynuclearbureacraciesdidnotdevelopuntilafterthedecisiontostartanuclear programhadalreadybeenmade.domesticcoalitionsandinterests,eitherprooorantio nuclear,shapethepoliticalenvironmentinwhichaheadofstatemustdecidewhetherto pursuenuclearweapons.measuringthosecoalitionsandtheirimportwouldbea fascinatingbutenormousundertaking.withouttheabilitytomeasurethosecomponentsof thedomesticenvironment,aquantitativeanalysisofthedomesticinstitutionalfactors 15Perkovich1999,47. 23

33 affectingagovernment sdecisiontopursuenuclearweaponsisincomplete.further researchisneededtoaddressthosetopicsmorefully. Prestige Thecachetofnuclearweaponshaschangedovertime.The nuclearclub andthe veryphrasenuclear$clubitselfindicatesthatnuclearweaponsconferexclusivestatusupon thosewhopossessthem wasprestigiousinthemidotwentiethcenturybecauseonlyan advancedsocietycouldmasterthephysics,chemistry,metallurgy,andindustrialscience necessarytobecomeanuclearpower.theawesomeandterribledestructivepowerof nuclearweapons,andthecatastrophicconsequencesofnuclearaccidents,werenotas widelyunderstoodastheywouldcometobeyearslater.now,thosewishingtojointhis clubaremostoftenbrandedasrogues;irresponsiblestateswillingtoriskgreatharmatthe expenseofnationalambition. 16 Astheinternationalnonproliferationregimehasgrownin sizeandscope,thepursuitofnuclearweaponryhasexperiencedadeclineinlegitimacy. 17 Yetnuclearweaponsremaina normativesymbolofmodernity. 18 Asasymbol,they invoketechnologicalability,nationalachievement,militarymight.allfivepermament membersofthesecuritycouncilarenuclearweaponsstates.ofthegroupofeight(g8), fourpossessnuclearweapons,andtherestareciviliannuclearpowers.francesought nuclearweapons,inpart,torestoreprideinfranceandreturnthecountrytoaplaceof globalpowerafterthecountry sembarrassinglossatdienbienphuin Chinalikely 16Sagan Thereisabroadliteratureontheissueofwhodetermineswhennuclearweaponsarelegitimateandwhen theyarenot.nuclearracism,nuclearorientalism,andnuclearapartheidaretermsthatrefertotheviewpoint thatadoublestandardexistsinthenonproliferationregimewhenstatesaverthatonlytheoriginalnuclear powershavearighttopossessnuclearweapons.seehughgusterton, NuclearWeaponsandtheOtherinthe WesternImagination, Cultural$Anthropology14(1),Feb.1999,foranintroductiontothissubject. 18SinghandWay Norrisetal

34 hadprestigeinmindalongwithsecuritywhenitsoughtanendto nuclearbullying by beginningitsownprogram. 20 CountrieslikeLibya,Brazil,orSouthAfrica,whichfacedonly limitedsecuritythreatsatthetimeofthedecisiontostartaprogram,mayprimarilyhave hopedtoattaingreaterglobalstatusthroughtheirendeavors. ) A)Statistical)Model)of)Decision)Making) Whichofthesefactorsarethemostrelevanttothequestionsofwhyandwhenthe politicaldecisionismadetostartanuclearweaponsprogram?statisticalmodelscanhelp identifythemostlikelycausesofadecision.modelsthatallowustoaccountfortheamount oftimethatelapsesbeforeadecisionismadecanalsohelpusthinkaboutthetimingof thesedecisions.inordertoanalyzetheconditionsunderwhichstatesareatriskofstarting anuclearweaponsprogram,wemustexaminetheseconditionsacrosstimeandspace.i beginthisanalysisfrom1939,whenscientistswerefirstbeginningtomakediscoveries aboutnuclearfission,andiincludeeverycountryintheworld. 21 Eventhistorymodels,alsoknownassurvivalmodelsandasdurationmodels,allow theresearchertoexaminetheriskthatsomeeventwilloccur.here,theeventinquestionis thepoliticaldecisiontostartanuclearweaponsprogram.therearetwogroupsofstates: thosethatmakethedecisiontostart,andthosethatdonot.attheoutset,whichgroupa givenstatebelongstoisunknown.byusinganeventhistorymodel,icanidentifythe variablesthatbestpredictwhichstateswillstartprograms. 20Feigenbaum OttoHan,LiseMeitner,andFritzStrassmannfirstsplittheatomin1938.Followingthisdiscovery,1939is generallyconsideredtobetheyearwhennuclearphysicistsbegan,asacommunity,todedicategreateffortto studyingnuclearfission. 25

35 Inanygivencalendaryear,acountryisatriskofmakingthetransitionfromnot pursuingnuclearweapons,whichisstage1,topursuingnuclearweapons,whichisstage2. Eventhistoryanalysismodelsboththelengthoftimespentinstage1andthedecisionto movetostage2.asthereisnoreasontoexpectstage1durationtimestofollowany particulardistributionalform,iuseacoxproportionalhazardsmodelforthisanalysis. FurtherdetailsregardingthemodelcanbefoundinAppendices3O5. 22 The$Decision$to$Start$a$Program Makingthedecisiontostartanuclearweaponsprogramisapoliticalaction.Based onevidencethatincludesdeclassifiedgovernmentdocuments,interviews,andexpert knowledge,wecanpresumethatastate sdecisiontopursuenuclearweaponsismadeby thatstate spoliticalleadership.theheadofstateultimatelydecideswhetherornottostart downtheproliferationpathway.politicaldecisionsareanalyzedinthemodelspresented here.althoughthereareseveralcasesinwhichascientistorgroupofscientistsintendto useanuclearprogramtodevelopweapons,theplansofnuclearscientistsarenotpolitical decisions.forexample,dr.homibhabha,thefatheroftheindiannuclearprogram,aimed tosteerindia sprogramtowardweaponizationlongbeforethepoliticaldecisionwasmade byprimeministerjawaharlalnehruin1964.thisdistinctionisillustratedbythe conversationau.s.majorgeneralobservednehruholdwithbhabha:nehruaskedbhabha ifhecoulddevelopanuclearbomb,andbhabhaansweredthathecould,giventime.nehru thendirectedbhabhanottoproceeduntilnehrugavehimpermission IamindebetedtoRobertFranzese,JamesMorrow,JanetBoxOSteffensmeier,andBenjaminJonesfortheir insightsinto,andadviceon,eventhistorymodeling. 23InPerkovich1999,36. 26

36 27 Table)2.1Decisiontostartnuclearweaponsprogram Country Yearofdecision Argentina Australia Brazil China France Germany India Iran Iraq Israel Japan Libya NorthKorea Pakistan Russia SouthAfrica SouthKorea Sweden Switzerland Syria Taiwan UnitedKingdom UnitedStates Yugoslavia Thereissomedisagreementovertheprecisedatesofsomeofthesedecisions,as theyweremadeinsecret.insomecases,documentationoragreementamongkey witnessesmakesthedateofthedecisionclear.inothers,wecanobservethestartofa program forinstance,whentheargentinemilitaryregimereplacedthecivilianheadof Argentina snuclearprogramwithamilitaryofficial butmustdeducewhentheformal decisiontostartwasmade.throughextensivecaseresearch,includingbothprimaryand secondarysources,andusingnewinformationthatmorerecentlybecameavailable,ihave compiledalistofcountriesanddatesthatismoreaccuratethanthevariouscountriesand