BridlingDictators
RulesandAuthoritarianPolitics
GRAEMEGILL
GreatClarendonStreet,Oxford,OX26DP, UnitedKingdom
OxfordUniversityPressisadepartmentoftheUniversityofOxford. ItfurtherstheUniversity’sobjectiveofexcellenceinresearch,scholarship, andeducationbypublishingworldwide.Oxfordisaregisteredtrademarkof OxfordUniversityPressintheUKandincertainothercountries ©GraemeGill2021
Themoralrightsoftheauthorhavebeenasserted Allrightsreserved.Nopartofthispublicationmaybereproduced,storedin aretrievalsystem,ortransmitted,inanyformorbyanymeans,withoutthe priorpermissioninwritingofOxfordUniversityPress,orasexpresslypermitted bylaw,bylicenceorundertermsagreedwiththeappropriatereprographics rightsorganization.Enquiriesconcerningreproductionoutsidethescopeofthe aboveshouldbesenttotheRightsDepartment,OxfordUniversityPress,atthe addressabove
Youmustnotcirculatethisworkinanyotherform andyoumustimposethissameconditiononanyacquirer
PublishedintheUnitedStatesofAmericabyOxfordUniversityPress 198MadisonAvenue,NewYork,NY10016,UnitedStatesofAmerica
BritishLibraryCataloguinginPublicationData Dataavailable
LibraryofCongressControlNumber:2021937199
ISBN978–0–19–284968–7
DOI:10.1093/oso/9780192849687.001.0001
Printedandboundby CPIGroup(UK)Ltd,Croydon,CR04YY
LinkstothirdpartywebsitesareprovidedbyOxfordingoodfaithand forinformationonly.Oxforddisclaimsanyresponsibilityforthematerials containedinanythirdpartywebsitereferencedinthiswork.
Preface
Thisbookisaboutthewaythatpoliticsattheapexofauthoritarianpoliticalsystemsisstructured.Thishasforlongbeenanareathatscholarshavebeenreluctant toexplorebecauseoftheobviousdifficultiesofdeterminingwhatwentonwithin theblackboxatthetopofthesepoliticalsystems.Thisisdifficultenoughtopenetrateindemocraticsystems,wheretransparencyissupposedtobeaqualityof thesystem,letaloneinauthoritarianpolitieswheremeasuresareoftentakento hidetheactivitiesoftherulers.Personalrelationships,whichconstitutealarge partoftheessenceofleadershippolitics,arealsonotoriouslydifficulttounderstand.Facedwithsuchdifficulties,mostscholarshaveeschewedthissortoffocus, insteadconcentratingonotheraspectsofauthoritarianrule.
Whilenotgainsayingthewidelyacknowledgeddifficultiesofthissortof analysis,thisbookrestsontheassumptionthatinmanyregimes,theexisting sourcesmakeitpossibletoidentifyrecurringpatternsofactionwithinthe leadershipthatreflectimportantaspectsofthedynamicsofpoliticsatthetopof thoseregimes,andthatthosepatternsarebestunderstoodintermsofrules.This clearlyrunscountertothemainthrustofmostoftheliteraturethatdoesexiston authoritarianleadershippoliticsandisthereforeunusualinthisfield.Butwhat alsomakesitunusualisthatitisinformedbyanalysisofcasesofthecommunist single-partysystem,atypeofregimethathasoftenbeenignoredinthegeneral literatureonauthoritarianpolitics.Thisissomewhatcuriousbecausecommunist regimeswereclearlyundemocraticandthereforeonthefaceofitqualifiedas authoritarianinnature.Furthermore,twocommunistsingle-partyregimes,the USSRandChina,wereamongthelongestlastingauthoritarianregimesinthe twentiethcentury(andinthelattercaseextendingintothetwenty-firstcentury), andprovidedbothinspirationandamodelformanyotherautocratstofollow. Accordingly,thisbookincludesanalysisofthesetwocommunistsingle-party states,alongwitharangeofotherregimetypes,toproduceatypologyofrules thathaveservedtostructurepoliticsintheseregimes.Whethersuchrulesapply toindividualregimesotherthanthosediscussedhereisamatterforempirical investigation,butthefactthattheycanbeseentooperateacrossthewhole spectrumofregimetypessuggeststhattheydo.Attheveryleast,itmeansthat theacceptedparadigmsofauthoritarianelitepoliticsdonotapplytoasignificant numberofregimes,andthereforerequiremajorrethinking.
Onecaveatisinorder.Itisnotunusualforthissortofleadership-focusedanalysisofauthoritarianregimestobecriticizedfornotemphasizingtheterrorand repressionthatundoubtedlyoccurredineachcountry,orfor‘normalizing’these
regimesandseemingtosuggestthattheywereactuallylikemanyothercontemporaryregimes.Suchcriticismisill-judged.Ifonewereseekingtoexplainan authoritariansystemasawhole,theabsenceofreferencetothescaleofsufferingexperiencedundereachregimewouldbeaseriousomission.Butitisbyno meansclearthatanalysisofthatsufferingwouldcontributesignificantlytoour understandingofthepoliticsofleadership.Thus,thefactthatthatsufferingisnot analysedhereisnottodenytheenormityofit;rather,itisrecognitionthatthesufferinganditsscaleisnotdirectlyrelevanttoourunderstandingofauthoritarian leadershipasdefinedinthefocusofthisbook.
Thisbookhasbeenalongtimeinthegestation.Ifirstbecameinterestedin comparativeleadershipincommunistsystemswhenIcompletedaMaster’sthesis onthecultsofpersonalityofJosephStalinandMaoZedongintheDepartment ofPoliticsatMonashUniversityintheearly1970s.IreturnedtoitwhenIbegan toteachacourseinauthoritarianpoliticsintheearly1980sintheDepartmentof GovernmentandPublicAdministrationattheUniversityofSydney.Thiswasa timewhenthequestionofleadershipwashighontheagendainbothcommunist powersand,beingintheearlystagesofwhatSamuelHuntingtoncalledthe‘third wave’ofdemocratization,wasalsoaperiodwhenthefailuresofauthoritarianleadershipwereseentobeanimportantfactorinbringingaboutregimechange.Butit wasnotuntil2014whenIreceivedanAustralianResearchCouncilgranttopursuethisproject(forwhichIthanktheARC)thatmythoughtswerecrystallized intotheformthatisnowreflectedinthisbook.Alongtheway,theDepartment ofGovernmentatSydney,formanyyearsmyprofessionalhome,hasbeenacontinuingsourceofstimulusandencouragement.Overthelasttenyearsofmytime there,theDepartmentbuiltintooneoftheleadingpoliticalsciencedepartmentsin theregionandboththenandsinceIretiredhasbeenaconstantencouragement andsourceofadvice.ParticularlyimportanthavebeenRodTiffen,whosesober advicehasalwaysbeenspoton(ifnotalwaysappreciatedatthetime!)andFred TeiweswhohasforlongbeenthepremierinternationalscholarofChineseelite politicsandatrenchantadvisorwhohassetmestraightonnumerousoccasions. ThelateT.H.RigbyattheAustralianNationalUniversityandArchieBrownat OxfordUniversityhavealsobeenimportantguidesinmyunderstandingofleadership,inboththeSovietandmoregeneralcontexts.MyunderstandingofLatin AmericanmilitaryregimeswasgreatlyenhancedbythestimulatingsemesterI spentintheKelloggInstituteforInternationalStudiesatNotreDameUniversity. ForthatIthankJimMacAdamswhomadeitpossible,theDirectorPaoloCarozza, thestaff,inparticularSamuelValenzuelaandAnibalPerez-Linanfortheiradvice, andthefellowswhosharedtheadventureinSouthBend.Iwasabletotestoutsome ofmyideasattheHigherSchoolofEconomicsinMoscow,forwhichIthankAndreiMelville.Themanuscripthasalsobeenimmenselyimprovedbythecomments ofanonymousreviewers.Myfailuretoacceptalloftheirpointsdoesnotdetract fromthevalueoftheirinput.
Mostimportantly,therockuponwhichitallhasrestedforalmostfivedecades hasbeenHeather,whoseloveandsupporthasbeenunstintingandwithoutwhich whateverIhaveachievedwouldhavebeenmuchthepoorer.Andfinally,ithas beenBethanyandAudreywhohaveshownmeagainwhatismostimportantin life,anditisnotauthoritarianoligarchs!
3.1.Politburomeetingdata
8.1.Rulecompatibilitywithregimetype
9.1.Oligarchstrategiesandtactics
9.2.Strategiesandrules
9.3.Oligarchtacticsandrules
10.1.Regimelongevity(inyears)
OnAuthoritarianLeadership
Dictatorshipshavedominatedhumanhistory.Themostcommonformofpolitical regimehistoricallyhasbeensometypeofauthoritarianrule.Therecentdominance ofthedemocraticparadigmisactuallyquiteunusualinhistoricalterms,butit isalsoonlyrelativelyrecentlythattherehasbeensustainedscholarlyanalysisof non-democraticregimes.Therewascertainlysomeattentiontonon-democratic regimesearlier,butmuchofthistooktheformofthestudyofparticularregimes, beittheRomanEmpire,Europeanabsolutism,NaziGermanyortheSovietUnion. Therewaslittleattemptatunderstandingnon-democraticregimesasatypebefore thepost-wardevelopmentofstudiesoftotalitarianism.1 Afewcomparativestudiesdidappeararoundthistime,2 especiallyrelatedtomilitaryregimes.3 Butitwas notuntiltheendoftheColdWarandtheturnofthemillenniumthatthestudyof non-democraciesexpanded,perhapsreflectingwidespreadrecognitionthatFrancisFukuyama’sclaimaboutthe‘endofhistory’markedbythetriumphofliberal capitalistdemocracywasutterlywrong.4
Thefocusofmuchofthisrecentscholarshiphasbeenontherelationship betweenrulersandruled,reflectingrecognitionthatnotonlyisthisthechiefdimensionuponwhichnon-democraticregimesdifferfromdemocratic,butalso thatthisismuchmorecomplexthantheearliertotalitarianschool’srelianceupon coercionastheprincipalbasisforachievingpublicpassivitysuggests.Muchof thisscholarshiphaslaidbarethewayinwhichsomeoftheseregimesworkin termsofhowtheymaintaintheirruleovertheircitizens/subjects.Themechanismsofincorporationandofbuyingoffhavebeenshowntobeasimportant asthefearofcoercionasmeansforsuchregimestoretaincontrol.Thisfocus ontheregime–peoplerelationshipisveryimportantandwarrantsthesortofattentionithasreceived.However,therehasbeenmuchlessofafocusuponthe leadershipsofsuchregimesandhowtheywork.5 Thisispartlybecauseofthe difficultyofpenetratingtheupperreachesofsuchregimestoseethedynamic wherebytheleadingdecision-makersgoabouttheirtasks.Thisisdifficultenough indemocraticsystemswheretransparencyissupposedtobeacentralprinciple,butinnon-democraticsystemswherethereisusuallynotevenlipserviceto transparency,thisisallbutimpossible.
Inpartthisrelativeabsenceofattentiontoauthoritarianleadershipisalsoexplainedbythecontinuingechoofthe‘greatman’theoryofhistory.Ithasbeen alltoocommonforobserverstoconceiveoftheseregimesintermsoftheperson
BridlingDictators.GraemeGill,OxfordUniversityPress. ©GraemeGill(2021).DOI:10.1093/oso/9780192849687.003.0001
oftheleader:Stalin’sUSSR,Hitler’sGermany,Mussolini’sItaly,Pinochet’sChile, Gaddhafi’sLibya,Mubarak’sEgypt,Khomeini’sIran,Saddam’sIraq,Mugabe’s ZimbabweandNiyazov’sTurkmenistanareafewexamplesofthewayauthoritarianregimeshavebeenpersonalized.Althoughthisnotionoftheembodiment oftheregimeinthepersonoftheleadercanoccurindemocraticsystems(e.g. DeGaulle,Thatcher,Trump),itoccursmostcommonlywithregardtoauthoritarianregimes.Thereisnoproblemwiththissortofshorthandwayofreferringto regimesaslongasitdoesnotproducewhatintoomanycasesithas:theassumptionthattounderstandhowtheregimeworks,yousimplyhavetounderstandthe leader.Clearly,thepersonalityoftheleaderisanimportantfactorinshapinghow aregimeworks,butitisnottheonlyfactor.Biographyisnotthesameasanalysis ofthepoliticalsystem.Centraltothecurrentstudyissomethingthatisalltoooftenforgotteninthefocuson the leader:he6 doesnotrulealone.Eveninthemost dictatorialofregimeswheretheleaderwieldsimmensepowerandhiswordislaw, thedictatorwillbesurroundedbysomewithwhomhediscussesissues.Theymay becronies,orfamilymembers,orpeopleinhisthrall,andtheymayonlyproffer opinionsratherthanparticipateinactualdecision-making,ortheymaysimplybe ‘yesmen’,butthroughtheirinteractionwiththeleadertheycaninfluencedecisionaloutcomes.Furthermore,asinglepersoncannotdecideeverything,withthe resultthatineffectsomedecisionswillbedelegatedtoothers.Inthissense,every authoritarianleadershipwillhaveacollectivedimension.
Somerecentscholarshiphasacknowledgedthis,7 butthereremainsanabsence ofempiricalanalysisofhowleadershipsoperateincontemporarynon-democratic systems.However,thishasnotstoppedscholarsfromtheorizingabouthowsuch leadershipsfunction.Twoofthemostimportantrecentworkslookingatthisquestion8 areinbasicagreementaboutthedynamicsofauthoritarianleadership.Inthe initialperiodofanauthoritarianregime,theleaderisnotinapositiontodominatehiscolleagues,andaprocessofnegotiationoccurswherebyapower-sharing agreementisworkedout.9 Oncethishasbeendone,thedictatorwillinevitably seektoacquiremorepower(therebybreakingthedealthathadbeenarrivedat) attheexpenseofhiscolleagues,andtheywillseektoresistthis.Theseimperativesofseekingenhancedpoweraresystemicratherthanafunctionofpersonal choicebecauseofthebasicuncertaintyoftheauthoritariansystem.Thisuncertaintystemsfromtwo‘intrinsicfeatures’10 ofauthoritarianpolitics:theabsenceof anindependentauthoritytoenforceagreementsandresolvedisputes,andtheeverpresenceofviolenceastheultimatearbiterofconflict.Itisthedictator’sdrivefor increasedpowerthatdestabilizesthepower-sharingarrangementinitiallyagreed, andgiventheabsenceofanexternalarbiter,theonlywayhiscolleaguescanblunt thisisthroughacrediblethreattoremovehimfromoffice.BarbaraGeddesetal. alsoarguethattheothermembersoftheleadershipnotonlywanttorestrictthe dictator,butalsotoincreasetheirownpowerattheexpenseoftheircolleagues. Thismeansthatwidespreaddistrustand‘intensecompetition’isthenormalstate.
BothMilanSvolikandGeddesetal.acknowledgethepossibilitythat‘procedures maybecomeinstitutionalizedovertime’11 orthat‘institutionalized“rulesofthe game”’12 maydevelop,butthisisnotexploredsystematicallyoriscursorilydismissedasbeingpossibleonlywhenthoserulesrestonmutualadvantageand acknowledgementofthepowerofkeyparticipantsorthefearofviolencebehind them.13 Thisreflectstheassumptionthatauthoritarianleadershipisalwayscharacterizedbyadifferencebetweentheleaderontheonehandandhis‘allies’or ‘innercircle’(orleadershipcolleagues,whomayalsobedividedamongthemselves)ontheother,witheachhavingdivergentinterests:thedictatortomaximize hispowertomakehimselfinvulnerabletochallenge;the‘allies’toincreasetheir powerinordertoenhancetheirpersonalpositionsandtoconstrainthedictator. Svolikenvisagestwostates.Onehecalls‘establishedautocrats’wherethedictator issopowerfulthatheisinvulnerabletosuccessfulchallenge,thesecondis‘contestedautocracy’where‘thedictator’salliesaregenuinepoliticalplayerswhoshare powerwithhimandconstrainhischoices’14 andwheresuccessfulchallengeispossible.Svolikalsorecognizesthatinstitutionscanhelptostabilizetheleader–allies relationship,butonlywhenthereisanevenbalanceofpowerbetweenleaderand alliesinhis‘contestedautocracy’.15 Geddesetal.acknowledgethatalongwiththis conflictoverdivergentinterests,thereisalsocooperationinordertopreventsuch conflictfromdestroyingtheregime.However,inthismodelthebasicdynamicof authoritarianleadershipremainscompetitionandconflictwhichisconstrained onlybytherelativepowerofthedifferentactors.
Thisquestionofcompetitionandconflictandhowitcanbemanagedhasbeen thefocusofaliteratureemphasizingco-optationasthemeansofadictatoravoidingconflictandconsolidatinghisposition.Centraltothisexplanationisusually saidtobepoliticalinstitutions—thepoliticalpartyorlegislature16—whichcanact asarenaswithinwhichnegotiationscanoccurbetweenthedictatorandpotential oppositiongroups.Intheseinstitutionscontrolledbargainingcantakeplace,the distributionofrentscanbeworkedoutandpolicycompromisesorganized.This isseenasameansofguaranteeingpotentialoppositioncontinuingaccesstoresourcesandtherebyallayingtheirsuspicionthatadictator’spromisescouldinthe futurebebreached.Inexchangeforsuchconcessionsbytheleader,thepotential oppositionprovidessupportforthedictatorandisco-optedintotheregime.That oppositiontherebygainsastakeinthesystembecauseitismoreadvantageousfor themtosupportthantoopposetheregime.17 Itisarguedthatbyenablinggreater transparencyandinformationflow,andbygivingpotentialoppositionguaranteedaccesstosomestateresources,institutionscanalleviatethecommitmentand monitoringproblemsthataccompanyamoresecretiveprocessandtherebydefuse conflict.18
AmorerecentstudybyAnneMengalsohasatitshearttheideathatthe dictatormakessomeconcessionswhichlimithispowerbutinsodoingconsolidatehisrule.19 Inthisexplanation,personalruleisconstrainedbyaprocessof
institutionalization,definedascreating‘rulesandproceduresthattietheleader’s handsbyempoweringotherelites’.20 Suchrulesandproceduresareseenas‘executiveconstraints’andconsistofconstitutionalrulesspecifyingtheleadership successionprocessincludingtheidentityofthesuccessor,termlimitsforthe leadershippost,andtheappointmentofelitestohigh-rankingcabinetposts(vicepresident/primeministerandtheministerofdefence).Specifyingthesuccession processincludingwhowastosucceedwassaidtoprovidesomecertaintyand therebydepressconflictatthesetimesofpotentiallyhighlevelsofconflict,while appointmenttohighpositions(termlimitswereseentobelesseffective)isseen tobeeffectivebecausetheleader’sdelegationofauthoritytoelitesguarantees theircontinuingaccesstofuturerents.Elitesaregivenastakeinthesystemand therebydissuadedfromchallengingthedictator.However,Mengargues,notall dictatorsgodownthispath.Individualdictatorsadoptsuchconstraintswhen theyareweakrelativetotheelitesandthereforevulnerabletobeingdeposed, especiallyattheoutsetoftheirrule;strongleadersarelesslikelytoadoptsuch constraints.
Thereisclearlyoverlapbetweenthemoregeneralco-optationargumentand Meng’sargumentaboutrulesandprocedures:bothhaveattheirheartthedictator’sdelegationofsomepowerandaccesstoresourcestoothersinexchangefor theirsupport.Thereisalsosomeconvergenceontheimportanceofpower:for Svolik,institutionscanbeeffectiveonlywhenthereisanevenbalanceofpower betweendictatorandothers,whileforMenginstitutionscanonlyconstrainleaders‘whentheychangetheunderlyingdistributionofpowerbetweenleadersand elites’.21 Moreambiguousistheidentityofthatgrouptowhichthedictatormakes concessions.MostofthetimeMengseemstobetalkingaboutthedictator’sintimates,thosereferredtointhisbookasthe‘oligarchs’,whilemuchofthediscussion intheotherliteraturereferstostrategicallyplacedelitesinthesocietymorebroadly whichcanactthroughsuchinstitutionsasalegislatureoraparty,22 similartothat groupreferredtointhisbookasthe‘elite’(definitionsarediscussedinthefollowingpages).Totheextentthattheco-optationliteratureisreferringtothiswider group,itsfocusisdifferentfromthatofthisbook,althoughthebasicstrategythey discussofrewardingco-operationmayberelevanttothenarrowergroup,asshown inChapter 9.Furthermore,neithertheco-optationliteraturenorMengseeksto explainthedynamicsofhowthenarrowleadershipgroup(theoligarchs)operates onaday-to-daybasis.
While,initsfocusonrulesandproceduresandthewaytheystructureaction, Meng’sanalysisisclosertothatofthisbookthanthemoregeneralco-optation literature,ithasanumberofproblematicaspects.Somearesharedwiththewider literatureandarediscussedbelow,somearemorespecificandarediscussedhere. Therearethreespecificaspectsofherargumentthatarerelevant,allrelatingtoher argumentabouttheappointmentofelitestoresponsiblepositionsandtheresultant delegationofpowerandauthority.23
a) Mengarguesthatinstitutionsareonlyeffectivewhentheychangethe ‘underlyingdistributionofpower’.However,inthecaseoftheexecutive constraintsshediscusses,itisnotclearthatthesealtertheunderlying distribution ofpowerasopposedtotheconditionsunderwhichthatpoweris exercised.Furthermoreshearguesthatonlyaweakleaderadoptsthesemeasures,butiftheyactuallymakehimstrongerbyconsolidatinghisposition (andtherebydoaffecttheunderlyingpowerdistribution),inwhatsensecan thesebeconsideredtobeconstraints?
b) Mengarguesthatstrongleaderschoosenottoadopttheseexecutiveconstraintswhileweakleadersmaketheoppositechoice.However,itisnot immediatelyclearthataleaderattheoutsetofhisruleactuallyhasmuch ofachoiceaboutwhethertodelegateornot.Noleadercandoeverything alone,sosomeformofdelegationisinevitable.Thequestionmosttimesis notwhetherdelegationoccurs,buttheconditionsunderwhichthisisdone andtheformittakes.
c) Meng’sexplanationexplicitlyappliestotheearlystagesofaregimeandit isnotclearthatitcanalsoapplytochangesofleaderduringaregime’slife. Whiletheremayhavebeenablanksheetofpaperatthetimetheregime wasfounded(althougheventhisisunlikely),subsequentleadershipchanges takeplacewithincontextsshapedbytheregime’spreviousexperience.Meng acceptsthis,butthismustclearlyhaveimplicationsforthewayinwhichthe sortofnegotiationssheenvisagesasbeingconductedarecarriedoutwhen achangeofleadershipoccurs.
Theemphasisupondelegation,co-optationandco-operationmayappeartobea sharpcontrastwiththeliteratureemphasizingconflict,butthisismoreappearancethanreality.Inbothliteratures,powerdispositionwithintherulinggroup remainsthekeydeterminantofleadershippolitics.Mengarguesthatonlyweak dictatorsmakethesortsofconcessionscentraltoherargument,andalthoughthis isnotalwaysexplicitlydiscussedintheco-optationliterature,itisimplicit;itis therelativestrengthofoppositionforcesthatencouragesthedictatortonegotiate andreachanagreement.Ifastrongdictatordoesnotengageinsuchnegotiation,theco-optationliteratureisnotrelevanttohisregime.Furthermore,most oftheco-optationliteraturefocusespurelyupontheactofarrivingatapact betweenleaderandoppositionratherthanathowthatarrangementworksout overthelifeoftheregime.Thosecontributionstotheliteraturethatdotakea longer-termviewofhowthepactactuallyworksinpractice24 tendtoconcentrateontherelationshipbetweentherulinggroupandbroadersocialforces.This meansthattheco-optationliteraturedoesnotreallyexplainhowpoliticswithin thatrulinggroupworksonaday-to-daybasis.Andthereforeneitherdoesitgainsaytheessenceoftheimageofauthoritarianleadershippoliticsstemmingfrom theconflictliterature.
Thisimageiswellillustratedbythefollowingquotationsfromtwoleading representativesofthisliterature,GeddesandSvolik:
• ‘Lifeinadictatorialeliteisthusinsecure,dangerousandfrightening.’
• ‘Membersofdictatorialelitesliveingrim,dog-eat-dogworlds.’
• ‘Theabsenceofbindinglimitsandinstitutionalchecksonthedictator,however,meanthatonlycrediblethreatstooustthedictatordeterhimfrom renegingonagreementsandabusinghissupporters.’
• ‘Dictatorsmustfeartheirclosestallies.’
• ‘Conflictoverthedistributionofpowerbetweendictatorsandtheirsupportersafflictsallnewdictatorships…earlyperiodsindictatorshipstendtobe unstable,conflictual,andsometimesbloody.’
• ‘Thecentralpoliticaldynamicofauthoritarianpolitics[is]conflictoverthe distributionofpowerwithintheregime’sleadershipgroup.’25
• ‘Politicalconflictsmaybe,andindeedfrequentlyare,resolvedbybruteforce.’
• Adictatorhasthe‘desireandopportunitytoacquiremorepowerattheexpenseofhisallies.Underdictatorship,aneffectivedeterrentagainstsuch opportunismmustbebackedbyacrediblethreatofthedictator’sremoval byhisinnercircle.’26
Thisisastarkandclearpictureoftheinherentlyviolentanduncertainnatureof lifeinauthoritarianpoliticalleaderships,withtherelativepowerdispositionofthe leadercomparedwithothermembersoftheleadershipthedefiningcurrencyof leadershippolitics;thebasicdynamicisoneofcontinualcompetitionandconflict constrainedonlybytherelativepowerofthedifferentactors.Butisitaccurate? Fourpointscanbemadeattheoutsetwithregardtothis.
a) Thewaythecentraldynamicofauthoritarianpoliticsissetupismisleading. Itisassumedthatwhiletheremaybemanyinternaldivisionswithinthe leadership,thereisabasicbinarydividebetweentheleaderontheonehand andhisleadershipcolleaguesontheother.Inpracticeitisrarelythissimple. Withintheleadershiptherewillalwaysbesomewhoarecloselyalliedtoand supporttheleadereveniftheremainderoftheleadershipismoreequivocal orafraidofandsetonrestraininghim.Envisagingtheleadershipinterms ofabinarydividemayassistaninterpretationbasedonthecontinuingclash oftwoopposedinterests,butitdoessobyoversimplifyingwhatisusuallya structurally(andoftentactically)morecomplexandoftenshiftingsituation.
b) Whileitiscertainlythecasethatinsomeregimessomeofthetimethequestionofpowerandwhowieldsitisparamount,isthisreallythedefiningor centralissueallofthetime?Inmostregimes,somesortofpowerequilibriumisusuallyreached,andalthoughthismaynotbeembeddedinstone (justasitisnotindemocraticsystems),thenatureofthepowerrelationship
isnotalwaysunderquestionornegotiation.Likeleadershipseverywhere, authoritarianleadershipsmustcometogripswithpolicyquestions,and whilemanyofthesewilldoubtlesshaveapowerdimension,thiswillnot inevitablybethedefiningfactor;theleader’s(norforthatmatterindividual members’oftheleadership)positionisnotunderquestiononallissuesand atalltimes.
c) Thefocusoncontinuingconflictobscuresordownplaysthelevelofconsensusandagreementthatmustexistforanyleadershiptofunctionforalength oftime.Itmaybethatboththerealityandpromiseofmaterialgainsuch asthatimpliedintheco-optationliteraturecanbeinstrumentalinmaintainingtheunityoftheleadership,27 butitisdebateablewhetherthisalone couldsustainaunitedleadershipoveranextendedperiodwithoutsome otherbasisofunitybeingpresent.Thenatureofsuchagreementandunity isdiscussedbelow(Rule8),butsimplythroughitspresence,suchagreement mustmodifytheHobbesianstatesketchedinmuchoftheliterature.
d) Thereseemstobeanassumptionthatdelegationnecessarilyweakensand constrainsaleader.Itcertainlycoulddothisinsomecases,butitisalso possiblethatthroughtyingotherstohimandgivingthemareasontosupporthim,delegationcouldconsolidatehisposition,enhancehispowerand increasehisreach,makinghimstrongerratherthanweaker.
Clearly,someregimessometimesmaybelikethemodelsuggestedbythequotationsonpage6,butnotallregimesarelikethis,norisitclearthatanyregimeis likethisallofthetime.
Itisdifficulttoseehowaregimewhoseleadershipoperatedinsuchafashion couldmaintainitselfinpowerforanylengthoftime,andwhileitistruethatsome authoritarianregimesareofrelativelyshortduration,somedemonstrateconsiderabledurability:of171‘countriesunderdictatorship’identifiedbyJenniferGandhi between1946and2002,81.3percentexperiencedauthoritarianruleformorethan 10years,withmostforconsiderablylongerperiods.28 Geddes29 hasalsoproduced somefiguresfortheaveragedurationofdifferenttypesofauthoritarianregime: militaryregimeslastedanaverageof9.5years,personalistregimes15.5years,and single-partyregimes29years.WhileGeddes’analysishascomeundersomecriticism,30 otheranalyseshaveagreedthatanaveragelifespanofover10yearsis commonforauthoritarianregimes.31 Ifregimeswerecontinuallycharacterizedby relationswithintheleadershipofthesortsuggestedabove,isitlikelythatsomany regimescouldsurvivemorethan10years?Aregimewhereadictatorruthlessly usedthecoerciveapparatus(secretpoliceormilitary)tofrequentlypurgetherulinggroupmaybeabletoremaininpowerforsometime,butitwouldbemore likelytocollapseundertheweightofthedistrustandantagonismswithintherulinggroupunlesssomethingotherthancommondistrustheldthemtogether.But inanycase,suchacontinualturnoverofleadingfiguresimpliedinthenotionof
therollingpurge(oratleastfearofit)doesnotseemtohavebeenpresentinmany oftheregimescommonlyadjudgedtobeauthoritarian.Thuswhiletheaccepted viewofauthoritarianleadershipmaybetrueofsomeregimesatsometimes,itis certainlynottrueofallandmaybenotevenofmost.Thekeyquestionisnotone focusedpurelyonpoweranditsdisposition,butonethatlooksmorebroadlyat howauthoritarianregimeleadershipswork.
Theprincipalargumentofthisbookis,inshort,thatratherthanbeingcharacterizedbyaHobbesianstateofnaturewithinleadershipcircles,theleadershipof authoritarianregimesisgenerallystructuredbyaseriesofrulesthatgovernrelationsbetweenindividualsintheleadership.Thesearerulesthataremeantto structuretheactivityofandrelationshipswithintheleadinggroupasawhole,and arethereforebroaderthanprinciplesidentifiedasmeansforthedictatortoconsolidatehispower.32 BothSvolikandGeddesetal.acceptthatsuchrulesmayexist, buttheyseethemasfarlessimportantthanthepowerdispositionoftherespectiveactors,andtheygenerallydonotexaminetheirnature.33 Therelativepowerof differentactorsremainsimportant,butthisalonecannotexplainthedynamicsof authoritarianleadership.Forthiswemustturntotherulesofleadershippolitics. Themostimportantrolethattheserulesplayistoorganize,coordinateandstructuretheinteractionsleadingfigureshavewitheachother.Thefollowingchapters willidentifytheserulesandshowhowandwhytheyareimportant.Butfirstwe needtobeclearwhatwearetalkingabout.
Theauthoritarianregime
Thestudyofnon-democraticpoliticsisbedevilledbyaplethoraoftermsthat areusedalmostinterchangeablyandwithoutcleardefinition:‘authoritarianism’, ‘autocracy’,‘tyranny’,‘dictatorship’and‘despotism’,tonameonlythemostcommon.Thereisnoclearandaccepteddistinctioninmeaningamongtheseterms, althoughthelastthreedoseemtohaveasignificantlymorepejorativeedgetothem thanthefirsttwo.Muchofthediscussionabouttherespectivemeritsoftheseterms amountstomeaninglesshair-splitting.Inthisbook,‘authoritarianregime’willbe usedwhenreferringtotheregime.Whenitcomestotheleader,thetermusually usedwillbe‘leader’or‘dictator’,eventhoughthelatterdoesseemtoimplygreater powerandarbitrarinessthanisoftenthecasewiththosethusnamed.
Thedefinitionofanauthoritarianregimehasalsobeenamatterofsomedebate.JuanLinzdefinedsucharegimeintermsofcomparisonwithatotalitarian regime.34 Mostscholarshavesoughttodefineauthoritarianregimesmoreinterms ofthecomparisonwithdemocraticthantotalitarianregimes,andheretheessentialnubofthedefinitionhasbeenthat,unlikedemocracies,authoritarianregimes arenotresponsibletothepopulaceoverwhomtheyrule.Thisdoesnotnecessarily meanthatsuchregimesareinsensitivetotheirsubjects,thattheydonotcareabout
howthepopulaceseesthem,orthattheydonotseektoprovidepoliciesthatserve someoftheinterestsoftheirsubjects.Itmeansthatthereisnoeffectivemechanismfortheregimetobeheldaccountablebythepopulace:thereisnoinstitutional meansforthepeopletoreplaceitsrulers.35 Inprinciple,thisclearlydistinguishes authoritarianregimesfromdemocracies.Inthelatter,thepeoplecanreplaceits governors;intheformertheycannot.However,inpractice,thedistinctionisnot asclearcut,andthisisshownbytheexistenceofoneparticulartypeofregime,the electoralauthoritarianregime(EA),wherealthoughsuchreplacementisunlikely, itcanhappen.Thisisdiscussedlaterinthechapter.
Thereisalsocontestationaboutthetypesofregimesthatconstitutetheauthoritarianfamilyandhowsuchregimesshouldbedistinguished.Notwithstanding theargumentthatformalinstitutionsareof‘questionablerelevance’inauthoritariansystemsandthereforeshouldnotconstitutethebasisofatypology,36 as subsequentanalysiswillshow,thewayregimesfunctionisshapedinpartbytheir institutionalcontours.Alsoithasbeenonthebasisprincipallyoftheidentificationofsuchinstitutionsthatregimetypeshavebeendistinguished.Oneofthe mostpopularofsuchtypologiesisthatofBarbaraGeddes,whoidentifiedthree basictypes:military,singleparty,andpersonalist.37 Thistypologyisnotperfect; itisnotallinclusive—contemporaryIrandoesnotfitintoanyofthecategories— anditencountersambiguitieswherehybridregimesareconcerned,forexample bothmilitaryandsingle-partyregimescanhaveapersonaldictator(Chileand Pinochet,theUSSRandStalin),andmilitaryregimesmayalsohavemajorparties (SyriaunderHafezel-Assad,EgyptunderNasser).Norisitsensitivetochanges withinaregimeovertime.Nevertheless,anexpandedversionofthistypologywill beusedhere.Therearetworeasonsforthis.First,thiswilllocatethisbookandits argumentsinthevastbulkofliteratureincomparativepoliticsthatusesthissort oftypology.Second,manyoftherulesuponwhichthisbookfocusesarerelated directlytotheinstitutionsandhowtheyfunction.Ratherthanbeingof‘ambiguousrelevance’,institutionsarethereforemuchmoreimportant.However,using suchaninstitution-basedtypologymustbesensitivetotheambiguityinvolved inhybridity,totheimplicationsoftherealsobeingotherinstitutionspotentially significanttotheeponymousinstitutioninaparticularregimeandtopossible changesovertime.Accordingly,althougheachregimetypewillbenamedinaccordancewithitsleadinginstitution,whenitcomestoanalysisofthoseregimes thepresenceofotherinstitutionsandchangeswithintheregimewillbetakeninto accountindiscussionofregimetypesandtheirrules.
Therearefivebasicregimetypes:
1) Single-partyregime.Thisisaregimewhereapoliticalparty,definedasan organizationwhoseprimarypurposeistoseekpowerandwielditthrough theoccupationofofficebyitsmembers,38 isthemostimportantinstitution inthesystem.Membershipofthepartyvestspowerinpeopleatalllevels,
andaccesstopoweratthetopofthepoliticalstructureismonopolizedby thepartyanditsmembers.Onlyonepartyispermittedtoplayameaningful roleinthesystem,althoughsatellitepartiesmaybeallowedtoexist,but withalmostnopower,asforexampleinChinaandtheformercommunist Poland.
2) Electoralauthoritarianordominantpartyregime.Thisisasystemwhere regularelectionsareheld,buttheelectoralarenaisweightedsofarinfavour oftherulingpartythatitisverydifficult,ifnotalmostimpossible,foritto bedefeated.ThisisthemostcommonformofauthoritarianruleinthepostColdWarworld.39 Somescholarshavesoughttodistinguishbetween‘competitiveauthoritarianism’and‘electoralauthoritarianism’or‘competitive authoritarianism’and‘hegemonicauthoritarianism’,40 withtheformerexistingwhenthereisapossibilitythattherulingpartycouldbedefeatedatthe pollsandthelatterwhenitcouldnot.Thisisintheoryausefuldistinction, butinpracticedifficulttooperationalizebeforearegime-changingelection actuallytakesplace.Accordinglyhere,theterm‘electoralauthoritarianism’ (EA)willbeusedforaregimewithapartysystemwithadominantparty andacompetitivebutneitherfairnorequalelectoralsystem.
3) Militaryregime.Thisisasysteminwhichthemainofficesinvestedwith powerareheldbymilitaryofficers,anditusuallycomesintoexistence throughamilitarycoupthatoverthrowsacivilianadministration.The militaryregimealsohastwosub-types:thecorporatemilitaryregimewhere themilitaryisinstitutionallyinvolvedinruling,andamorepersonalist militaryregimewheretheinstitutionasawholeisnotdirectlyinvolvedin politicallifebutmilitaryofficersfillthemainpoliticalposts.Bothsub-types aremarkedbythefactthatwhileexercisingcivilpower,militaryofficers usuallystepbackfromtheirmilitaryresponsibilities.
4) Monarchy.Thisisasystemwhereaccesstopowerisdeterminedbypersonal lineage,withpowerbeingdominated(butnotalwaysmonopolized)by membersofasinglerulingfamily.Monarchieshavetwosub-types:the dynasticmonarchywhereruleissharedamongthemembersoftheruling familyundertheheadshipofthekingorqueen,andanon-dynastic monarchywherepowerisconcentratedinthehandsoftheindividual monarchalone.41
5) Personaldictatorship.Thisiswhereasingleindividualhaspersonaldiscretionoverpoweranditsinstrumentsrelativelyunconstrainedbyothers,and thereforesodominatesthesystemthatheisthemostimportantfeatureof theregime;42 ifapartyormilitaryexists,itisofadistinctlysecond-order rank.Manyscholarshavepreferredtoseepersonalismasaqualitythat cancharacterizeallofthetypesofregimenotedinthischapter.Individual leaderscanbethemostimportantactorsinregimeswhereapoliticalparty orthemilitaryremainthedominantinstitutionalstructure,inwhichcase
theregimeappearsasamixedorhybridtype.Furthermore,regimescan gothroughstageswhere‘institutional’ruleisreplacedby‘personal’rule; thesestages(andthereforepotentiallychangesinregimetype)canonly bedistinguishedthroughcloseempiricalanalysis.Inthisbook,inorder nottoexcludethoseregimeswhereadominantleaderdidnotreallyshare powerwithapartyormilitaryinstitution(suchascontemporaryBelarus), thepersonaldictatorialregimewillbeassumedtobeaseparatetype.The questionofstages,oftherelationshipbetweenapersonaldictatorandother aspectsoftheregime,isdiscussedfurtherbelowintermsofthedistinction betweenadominantandapredominantleader.
Theauthoritarianoligarchy
Whilethedominanceofapersonaldictatorisacentralfocusintheanalysisof authoritarianleadership,scholarsalsorecognizethepresenceofothersthanthe leaderwithintheleadership:witnessreferencestothingslikecliques,factionsand croniesaswellassuchinstitutionsasthepolitburo,juntaandcabinet.Butsuch recognitionrarelyextendstosystematicanalysisoftherolesuchgroupsplayin decision-makingandinthestabilityofthepowerstructuregenerally.Thisisacentralfocusofthisbook,butthismeansthatweneedtobeclearwhomwearetalking about.Therearetwogroupingsofleadersrelevanttothisdiscussion:theoligarchs andtheelite.43 Theoligarchsarethosefewpeoplewhoimmediatelysurroundthe leader.Theyarehisclosestcolleagues,andusuallyfilltheleadingformalpositions inthepoliticalsystem;theyarethemembersofthepolitburo,juntaorcabinet. Thisnotionofoligarchemphasizesthepoliticalpowerandpositionthesepeople enjoy.Itisthereforedifferentfromtheconceptionofoligarchsdefinedinterms oftheirpossessionofmaterialresourcestheymayusetowieldpowerandprotect theirpersonalwealthandsocialposition.44 Itismoreapositionaldefinition,and closertothatofAristotle(althoughhedoesassumetheoligarchywillbecomposed ofthewealthy)butdifferentfromthatusedwithregardto,forexample,Russiaor Indonesiawherethetermisexplicitlyrelatedtocorruptwealthybusinessmen.The basicmeaningofoligarchyinthisbookis,therefore,thattheoligarchsaretheclose associatesofthedictator.45 Comparedtothosebelowthem,andespeciallythegeneralpopulace,theyareverypowerful.Comparedtotheleader,thepowerofthe rank-and-fileoligarchsmayberathermoremodest.Relativetooneanother,individualoligarchsmayhavemoreorlesspower(andthismayshiftovertime)and theymayoccupydifferentranks;thenotionofoligarchydoesnotassumethatall oligarchsareequal.46 Thedictatorishimselfanoligarch;heispartoftheoligarchy andtheleaderofit,notseparatefromit.
Fortheviewoftherank-and-fileoligarchsasbeingvulnerabletotheleader,at hisbeckandcallandalwaysindangerofbeingremovedorevenkilled,scholars
oftenemphasizethe‘crony’aspectoftherelationship.Thisimpliesthattheyare inleadershippostspurelybecauseoftheirrelationshipwiththeleader,andthat werethatrelationshiptosour,theircontinuedpresenceintheleadershipwould beunderthreat.Thereisasenseinwhichthisistrue.Anoligarchwhoseposition restsonlyonhispersonalrelationshipwiththeleaderisvulnerabletochangesin thatrelationship.Butthereversealsoapplies:personalrelationshipspre-datingthe leadercomingtopowercanprovideabasisoftrustandcontinuingmutualsupport thatcouldoffsetthepresumedvulnerability.Establishedrelationships,evenifthey undergochangewhentheleadercomestopower,canshaperelationsbetweenthe leaderandotheroligarchs,includingprovidingabasisforhispreferenceforsome overothers.Forthoseleaderswhoseektocreateafactionalbasisfortheirrule,such relationshipscanactasafoundationuponwhichfactionscanrest.Suchestablished relationships,andthetrustthattheymayinvolve,canalsomeanthattheoligarch willenjoygreaterfreedomofmovementandindividualautonomyfromtheleader thanwouldotherwisehavebeenthecase.Italsomeansthatthesituationinthese leadershipsisrarelyoneoftheleaderversustherest:someoftheoligarchswill inevitablybesupportersoftheleader,almostregardlessofwhathappens.
Personalrelationshipscanhaveexistedamongtherank-and-fileoligarchsas well.Wherethoserelationswereamicableorcooperative,carriedforwardinto powertheycanprovideabasisforsolidaritywhichcanunderpinfaction-creation, orsimplyconstituteanideationalbasisforcooperation.Thiscouldbeimportantshouldtheoligarchs(orasectionofthem)believethattheyneedtodefend themselvesagainstpotentialattackbyadictator.Itcouldalsobeimportantinany conflictthateruptsamongtheoligarchsthemselves.Whenconflictoccursamong therank-and-fileoligarchs,perhapsinthecompetitivequestfortheleader’sbenevolenceorinthestruggleforpowerorresources,organizationamongthemisa logicalresponse,andpriorpersonalrelationscanhelptoeasethecreationofsuch organization.Priorhostilerelationsamongindividualoligarchscanhavetheoppositeeffect,thatofprovidingabasisforthegenerationofoligarchconflict.When suchconflictunrolls,ifthedictatorstaysaboveit,itmayenhancehispowerover hiscolleagues.
Personalrelationsarenottheonly,andperhapsnoteventhemostimportant, factorindefiningthenatureoftheoligarchs.Centraltothisistheinstitutional rolestheoligarchsoccupy.Inmostauthoritariansystems,theoligarchsareformallymembersofaleadingcouncil,likeapolitburo,cabinetorjunta,withthe leaderthechiefofficerofthatbody.Butwhilethismaybethepositionthatsignifiesmembershipoftherulinggroup,itisusuallynottheonlypostanoligarch occupies.Usually,theywillalsoheadoneofthebureaucratichierarchiesofwhich thesystemconsists.Mostcommonhasbeenoligarchswhoarealsogovernment ministers,headinggovernmentdepartmentswhosetaskitistoadministersome sectorofnationallife.Commontoohavebeentheheadsofthesecurityservice andofthemilitary,withthelattersometimesrepresentedbyanumberofpeople
fromthedifferentservicebranchesofthemilitary.Theheadoftherulingpolitical partymayalsoappearinthisgroup.Oftenthisotherjobistheoligarch’smainone, inthesensethatittakesupmoretimethanhisinvolvementintheactivitiesofthe narrowrulingoligarchy,butultimatelyitistherecognitionofhimasanoligarch throughmembershipoftheoligarchythatisofgreaterpoliticalimportance.
Theeliteisamuchbroadergroupingofwhichtheoligarchsarepart,withthe boundarybetweennarroweroligarchyandbroaderelitesometimesindistinctand occasionallycontested.Thenon-oligarchpartoftheelitewillcomprisedifferent sortsofpeopledependinguponthetypeofregime,butitwillusuallyincludehigh andmiddlelevelpoliticaladministrators,themilitaryofficercorps,andwealthy privateeconomicinterests.Thisgroupoftenhasitsownorganizationalvessels; leadingpartyinstitutions,thestatelegislatureandcivilservicehierarchy,theformalmilitarystructures,andprivateandpubliccorporationsoftenfigureinthe operationsoftheelite.Thisbroadgroupisacentralunderpinningoftheoligarchy, andtheneedtokeepitcontentandonsideisamajorcontinuingtaskforthe oligarchs.Theelitealsocomprisesthebasicrecruitinggroundfortheoligarchy. Theprincipalfocusofthisbookisontheoligarchy,nottheelite,althoughthe relationshipbetweenthetwoisanimportantconsideration.
Notwithstandingthiscollectivenatureofauthoritarianleadership,apreeminentindividualleader,commonlycalledthe‘dictator’,usuallyemerges.There maybesomecases,ormaybeindividualinstances,whenthedictatormakesadecisiontotallyonhisowninitiativewithnootherinput,butthisisnotcommon.The dictatorusuallyfunctionsinacollectiveenvironmentinthesensethathewillbe surroundedbyandinteractwithotheroligarchsonaregularbasis.Thenatureof thatinteraction,andtheroleplayedbytheotheroligarchs,willbeinfluencedbythe powerpossessedbythedictator.Threebasicmodelsofauthoritarianleadership exist:
1) Realcollectiveleadershipwherenosinglepersonconsistentlyhasgreater powerthantheothers;allaremoreorlessequalanddecisionsaresubjectto brokerageandcompromise.Suchsituationsare,inpractice,rare.Personal ambitionisusuallysufficienttodriveoneormoreindividualstoattempt togainprimacy,leadingtothereplacementofsuchcollectivismbyapreeminentleader.Therearetwosortsofpre-eminence:predominanceand dominance.
2) Predominantleader,whereonepersonispre-eminentbutdoesnotusually ignoretheviewsandpreferencesofotheroligarchs.Theleadermaynothave sufficientpowertoignoretheoligarchscollectively,butismorepowerful thananyofthemindividuallyandprobablymanyofthemcombined.Predominanceusuallyrestsuponastablealliancewithsomeoftheoligarchs. Thepredominantleadercanbestrongorweak,adistinctionreflectedinhis abilityconsistentlytogethiswayonissues.
