rationality and the problems of administrative theory

《rationality and the problems of administrative theory》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

RationalityandtheProblemsofAdministrativeTheoryRICHARDHARTWIG*Mr.HartwigisaninstructorinthePoliticalScienceDepartmentofIowaStateUniversity.‘Ibelievethatthehistoryofphilosophicalethicscanbeviewedasanexplorationofinsolubleproblems,someofwhich,byrephrasing,havebeentransformedintoanswerablequestions.’(WayneLeys,EthicsandAdministrativeDiscretion.)Theconditionofpublicadministrationhasbeendiagnosedasserious,butasyetthereisnocure.Ourinabilitytointegratethemanypartialtheoriesofadministrativebehaviourisonesymptonofthe‘crisis’,andtheambiguityofkeyconceptssuchas‘organizationaleffectiveness’and‘administrativeresponsibility’isanother.Publicadministrationtheoryhastraditionallyattemptedtoexplainandevaluatethebehaviourofcompleteorganizations.Theproblemisthatthecompositebehaviourofcompleteorganizationsinvolvesmultipletypesofrationality.Conceptsusedinthecontextsofmultipletypesofrationalityarenecessarilyambiguous,andtheoriesignoringthefullrangeoftypesofrationalityarenecessarilypartial.Mycentralthesisisthatthefivetypes(andtwoaspectshevels)ofrationalitydescribedbythepoliticalphilosopher,PaulDiesing,maybeusedasacomprehensiveframeworkofidealtypesaroundwhichtoorganizeadministrativetheory.AccordingtoWaldo,theconditionofpublicadministrationisconfusionastothenature,boundaries,methodologyandsubjectmatterofthe‘field’,‘discipline’,or‘profession’,andanycurewillhavetocomefromoutsidethetraditionalframeofreferenceofpublicadministrationtheory(Waldo,1968).Thewriteracceptswithoutfurtherargumenttheexistenceofanidentitycrisisandwhatmightbecalledaparadigmcrisisofthefield-matterswhichhavebeenextensivelydiscussedelsewhere(seeDahl,1947;Waldo,1964,1968,and1975;Marini,1971;Self,1973;Ostrom,1973;andMouzelis,1967).’Theproblemsexist.Thequestionis:‘Whatistobedone?’*IwouldliketothankEdPortis,JohnDorsey,CharlesHartwig,GaryWamsley,FredericThayerandtheeditorsofPublicAdminisfrutionfortheirhelpfulcommentsonearlierdraftsofthispaper.I59 PUBLICADMINISTRATIONTHEPROBLEMOFMULTIPLEPARADIGMSTostartatthebeginning,aparadigmcrisisisaboutorder.Itiseithertheperceivedabsenceofanadequateexplanatoryorder,oritistheabsenceofanacceptabletheoryofhowadesiredordermightbeachieved.Aparadigmproblemmaythushavebothscientificandnormativeaspects.EltonMayo’sfamousWesternElectricstudies(1927-1932)provideagoodexampleofascientificparadigmcrisisinpublicadministration-andonewithnormativeimplications.Asiswellknown,theinvestigatorsorginallyintendedonlytocorrelateworkers’performancewithobjectivefactorssuchasplantlightingandfatigue.Theirfinding,thatperformanceincreasedinbothtestandcontrolgroups,couldnotbeexplainedintermsoftheclassicmodelofthescientificadministrationschool.Theultimateoutcomeofthesestudieswasthecreationofthehumanrelationsmovement,whichemphasizedthepsychologicalandultimatelysociologicaldeterminantsofjobperformance(Mouzelis,1967,pp.97-119).Humanrelationstheoryrepresentedashiftfrommechanisticparadigmtoonewhichemphasizedtheimpactofinformalgroupnormsandvaluesuponindividualbehaviour.Buttheparadigmproblemdidnotendthere.Itisimportanttonotethatclassicorganisationtheorywasneverentire&discredited;humanrelationstheorywasneverentirelysuccessful;andthetwoapproacheshaveneverbeenadequatelyintegrated.Intheliteratureoforganizationalstudiesnewapproachesandtheorieshaveaddedfundamentallynewparadigmswithoutreplacingordestroyingthecredibilityofolderones.HerbertSimon’sAdministrativeBehaviour(1949,forexample,demonstratedtheinherentcontradictionsoftheclassicprinciplesofadministration,yetSimondidnotclaimtohavedestroyedthetraditionaltheories.Theproblem,hewrote,wassimplythat‘hierarchy’,‘spanofcontrol’,andsoforth,werereallyonlycriteriafordescribinganddiagnosingadministrativesituations.‘Nosingleoneofthesecriteriaisofsufficientimportancetosufficeasaguidingprinciplefortheadministrativeanalyst.Inthedesignofadminis-trativeorganizations,asintheiroperation,over-allefficiencymustbetheguidingcriterion.Mutuallyincompatibleadvantagesmustbebalancedagainsteachother,justasanarchitectweighstheadvantagesofadditionalclosetspaceagainsttheadvantagesofalargerlivingroom.’(Simon,1945,p.36.)Simon’sframeworkofanalysiswasclearlynovel,yetnotacompletebreakwithtradition.Atthesametime,ithadlittleincommonwithhumanrelationstheory.AccordingtoPeterSelf,Simon’sparadigmisessentiallythatoftheeconomist/planner(Self,1973,pp.29-37).Fromthisperspective-some-timescalled‘rationalmanadministrativetheory’-thecentralproblemistoincrease(economic)eficiencyindecisionmaking.Simonseesthepsycho-logical/sociologicalfactorswhicharecentraltohumanrelationstheoryessentiallyasconstraintsupon,individualrationality.Bythemselves,individualsarelimited,butwithanadequateorganizationaltheory,theycan160 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYbemotivated,controlledanddirectedtowardorganizationalends.Bywayofcontrast,humanrelations/organizationdevelopmenttheoristssuchasDouglasMcGregor(1960,1970)believethatpeoplearemotivated.Employeesorcivilservantswilltakeinitiativesandinvolvethemselvesintheirworktotheextentthatsuchworkcontributestotheirpersonaldevelopment.Thedifferencesbetweentheschoolsofthoughtwhichthesetwomenrepresentarethusprofound-asisindicatedbytheacerbicexchangebetweenSimonandChrisArgyrisinthePublicAdministrationReview(Argyris,1973;Simon,1973).TheincrementalisttheoriesofLindblomandWildavskyprovideanotherexampleofparadigmconflictinorganizationalanalysis.Itisanarticleoffaithforincrementalists(amongothers)thatthetraditionaldichotomybetweenpolitics/policymakingandadministrationisuntenable.Theyarguethatadministratorsmakepolicy,thatpolicydecisionshavepoliticalimplications,andthatthepoliticalimplicationsofadministrativestructurecannotbeignored.Centralized,hierarchicaldecisionstructures,farfrombeingneutralinstrumentsofchoice,mayfunction(orbedesigned)tocentralizepower.Fromthisperspective,ecomomicefficiencymaybeahighlymisleadingcriterionbywhichtoevaluateorganizationalbehaviour.(SeeBraybrookeandLindblom,1963;Lindbolm,1968;Wildavsky,1964,1966;andCaidenandWildavsky,1974.)Acomprehensivelistingofschoolsandimplicitparadigmswouldextendfarbeyondwhathasbeenmentionedhere.NodiscussionofadministrativetheorywouldbecompletewithoutadiscussionofMaxWeber’swork,forexample.Ostrom(1973,p.29)haswrittenthatWeber’sworkonbureaucracycomplementsandsupportstheclassicparadigmofadministrativethought.Yet,Weberisalsodifferent,andheclearlyovershadowsTaylor,Fayol,andtheotherfallenidolsoftheclassictradition.TheNormativeAspectoftheParadigmProblemItwasstaunchlybelievedintheearlypartofthiscenturythatascienceofpublicadministrationdidexist,andthatitcouldserveasabasisforvaluejudgements.Itisapeculiarcharacteristicofthefieldthattheprofessionhadalreadyassumeditsplaceintheeverydayworldbeforethelimitationsofitstheoreticalbasebecameapparent.Someoftheselimitationswerepointedoutbythepioneersofthehumanrelationsschool,butanotherkeyissuewashighlightedbyCarlFriedrichin1940.Hearguedthat‘publicpolicy...isacontinuousprocess,theformationofwhichisinseparablefromitsexecution’(Friedrichp.416).Butifthisistrue,itfollowsthatadministratorsmustexerciseconsiderableindependentdecision-makingpower.Andhowisthistobejustified?‘Whenceshouldanadministratorgetnormsforofficialdecision-makingandofficial-relatedpersonalconduct,sofarastheseareneithergiventohimbyconstitutionallyderivedsources(orthesesourcesconflict!)norflow“inexorably”fromtechnical-scientificsources?Regardlessofproven-ience,howcansuchnormsbejustified?’(Waldo,1975,p.216.)161 PUBLICADMINISTRATIONWaldowritesthatintheAmericancontextthisproblemisprobablyinsoluble,andinanycase,itinvolvesquestionsofpoliticaltheoryandconstitutionalengineeringwhichextendfarbeyondthecustomarybound-ariesofpublicadministration(Waldo,1975,p.216).Nevertheless,suchissueswillnotdisappear.Administratorsandadministrativeagenciesaretoldtoactresponsibly,whichistosay,inthepublicinterest.Itissmallcomfortforthemtobetoldthatneither‘responsibility’nor‘thepublicinterest’hasbeenadequatelydefined.lHowthendoweevaluateorganizationalbehaviour?Empirically-mindedstudentsofpublicadministrationattempttomeasureorganizationaleffective-ness,whilethosewithmorenormativeinterestsattempttoevaluateadministrativeresponsibility.Afterdecadesofefforttowardclarificationofthoughtandlanguage,however,both‘organizationaleffectiveness’and‘administrativeresponsibility’remainhighlyambiguousconcepts.RichardSteersrecentlyreviewedseventeenmultivariatemodelsoforganizationaleffectiveness,findingthattheyemployedoverfourteendifferentmeasuresofeffectiveness.Heconcededthepossibilityofsayingthatattemptstomeasureorganizationaleffectivenessarefutile,butpreferredtoconcludethattheeffectivenessconstructissocomplexastodefysimpleattemptsatmodeldevelopment(Steers,1975,pp.546-49).Thedflicultiesencounteredindefining‘administrativeresponsibility’areperhapsevenbetterknown.Ina1969JournalofPoliticsarticle,CharlesGilbertfoundthat‘responsibility’hadbeenusedintheliteratureofpublicadministrationtorefertoatleasttwelvedifferentvalues:‘responsiveness’,‘flexibility’,‘consistency’,‘stability’,‘leader-ship’,‘probity’,‘candor’,‘competence’,‘efficacy’,‘prudence’,‘dueprocess’,and‘accountability’.Norare‘effectiveness’and‘responsibility’theonlyambiguousconceptsinthevocabularyofadministration.‘Legitimacy’-atermcrucialtoanyevaluationofadministrativeresponsibility-providesafurtherexample.Websterdefines‘legitimate’as‘conformingtorecognizedprinciplesoracceptedrulesandstandards’.Ontheotherhand,SeymourMartinLipsetwritesthat‘legitimacyinvolvesthecapacityofthesystemtoengenderandmaintainthebeliefthattheexistingpoliticalinstitutionsarethemostappropriateoneforthesociety’(Lipset,1960p.64).Butconformingtoacceptedrulesisnotthesamethingasabelief,andthetwomeaningsarenotalwayscongruent.TYPESOFRATIONALITYANDTHEPARADIGMPROBLEMTheambiguitiesdescribedabovedonotresultfromsimplecarelessness.Apartialexplanationwouldbethatthedifferentschoolsofadministrativethoughthaveevolveddistinctmeaningsforkeyconcepts.Thus,‘humanrelations’theoristsoftenuse‘responsibility’inasubjectivesensewhichisab-sentintheworkofmoretraditionalanalysts-reflectingdifferencesintheunderlyingparadigmsofthetwoschool^.^Onasecondlevel,however,theambiguitiesandmultiplemeaningsofkeyconceptsarerelatedtothenature162 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYofpublicadministrationasafield.FrederickMosher(1975p.6)haswrittenthatpublicadministrationisaprofessionofgeneralists.Studentsofadminis-tration,likepriests,aretrainedtodealwiththewholeman,andpublicadministration,likeanthropology,aimstounderstandwholeorganizations.ItisforthisreasonthatDwightWaldocallspublicadministrationaprofessionratherthanadiscipline-moreanalogousto‘medicine’thanto‘economics’(Waldo,1975,pp.220-26).Thisisinparttosaythatpublicadministrationisanappliedscience.Buthistorically,publicadministrationtheoryhasattemptedtodescribeandevaluatethecompositeactivitiesofcompleteorganizations.Andthekeypoint,asPaulDiesinghaswritten,isthattheactivitiesofcompleteorganizationsnecessarilyinvolvemultipletypesofrationality(Diesing,1962,p.239).Thisexplainsmuchoftheambiguitywehaveencountered,becausethesameconcepts,usedinthecontextsofdifferenttypesofrationality,willhaveirreduciblydifferentmeanings.Take,forexample,‘responsibility’.Anemployeewhofollowstheordersofhisorganizationalsuperiorisnormallyconsideredtobeacting‘respons-ibly’,butwhatdoesthismean?Onepossibilityisthattheemployeeobeyshissuperiorbecauseofacalculationthatitisinhisownbestinteresttosticktothebargainhemadeintakingthejobinthefirstplace.Ifthisisnotthecase,heisattemptingtomaximizeutility,andisactingaccordingtothedictatesof‘economic’rationality.Ontheotherhand,theemployeemightfollowordersbecausehefeelsloyaltohissuperiorandidentifiesemotionallywithhisboss,and/orwiththelargerorganization.Thiswouldnormallybeconsideredanirrationalmotive,butinDiesing’susage(tobedefinedbelow),itwouldin-volve‘social’rationality.‘Responsibility’hasadistinctmeaningineachofthetwosituations,asthesameactionresultsfromdifferentmotivations-inthefirstcase,conscious,andinthesecond,largelyunconsciousorunwilled.AndasisnotedintheRikerandZavoinaarticlefromwhichthisexampleisadapted(1970,p.48),theimplicationsofthetwoexplanationsforpredictingfuturebehaviourarelikewisequitedistinct.Amanwhofeelsresponsiblemayremainloyaltoanorganizationlongafterloyaltyhasbecomeeconomicallyorobjectivelyirrational.Ineffect,ourexplicitorimplicitunderstandingofthenatureofrationalityaffectsthemeaningsofthewordsweuse.Andiftheaboveexampleisvalid,animprovedunderstandingofthedifferentmodesofrationalitymayhelpacademicsandpractitionersaliketoclarifytheirlanguageandtheirtheories.In1960,JohnPfiffnerpublishedanaward-winningarticleinthePublicAdministrationReviewentitled‘AdministrativeRationality’.Pfiffner’sthesiswasthatthereareseveraltypesofrationalitytobefoundinadministrativebehaviour.Rationaldecisionmaking,hewrote,isgenerallythoughtofas:‘thegatheringofallpertinentfacts,canvassingthealternativesforaction,andselectingtheonewhichwillproducemaximumresultsthroughathoughtprocesswhichphysiologyassociateswiththecerebrum.Itisseenasacoldlycalculatingprocesswhichexcludeshumansympathyandsentiment.’(Pfiffner,1960,p.126.)163 PUBLICADMINISTRATIONPfiffnerarguedthatadministrativerationalitydiffersfromorthodoxconceptsofmeans-endrationalityintakingintoaccountanadditionalspectrumoffactsrelativetoemotions,politics,power,groupdynamics,personality,andmentalhealth.(Seealso,Gore,1972.)Pfiffnerwasclearlyright.Multipletypesofrationalityaremanifestedinadministrativebehaviour.Byitself,however,Pfiffner’sarticlewasnotsufficienttosortouttheconfusionsurroundingadministrativetheory.Fewscholarsheededhisadvice,andmostimportantly,theindividualtypesofrationalityhadnotbeenfullyoradequatelydescribedasof1960.Thisintellectualgapwasfilledonlytwoyearslater,however,withthepublicationofPaulDiesing’sbook,ReasoninSociety.ThecentralthesisofthepresentpaperisthatthefivemodesofrationalitydescribedbyDiesingmaybeconsidered-todate-thefundamentalpara-digmsofsocialscience,andthatpublicadministrationtheoryandorganiza-tionaltheoryingeneralcanbeclarifiedandorganizedwithreferencetothesebasicparadigms.Thisargumentcanonlybepersuasivetotheextentthatthereaderisfamiliarwith,andagreeswith,Diesing’sdescriptionsofthedifferenttypesofrationality.ItisthusnecessarytosummarizebrieflyDiesing’sanalysisof‘rationality’.LEVELSANDTYPESOFRATIONALITYPaulDiesing’sanalysis,inReasoninSociety,isdirectedagainsttheview,commontomanysocialtheories,thatrationalityisidenticalwithefficiency.Technicalrationalityistheefficientachievementofasinglegoal;economicrationalityisthemaximumachievementofapluralityofgoals,andnoothertypesofrationalityareadmitted.AnexampleofthisviewisprovidedbyvonMises,whowrites:‘Theeconomicprincipleisthefundamentalprincipleofallrationalaction,andnotjustaparticularfeatureofacertainkindofrationalaction...Allrationalactionisthereforeanactofeconomizing.’(vonMises,1960,p.148.)Similarpositions,Diesingnotes,areoftenassumedbyorganizationtheorists,whentheytake‘rationalizing’ofanorganisationtomeananincreaseinitsproductiveefficiency.‘Arationalorganizationisanefficientone,andotherprinciplesormodesoforganizationarethoughttobenon-rationalorirrational.’(Diesing,1962,p.1.)Suchaconceptionofrationalityisseverelylimited.Thecriterionofeffici-encyisapplicableonlytomeans,notends-unlesstheseinturnaremeanstofurtherends-andthusthemostbasicdecisionsarenotsubjecttoselectionorchoicebyrationalprocedures.Furthermore,therearemanycommonactivitieswhichcanbeevaluatedbystandardsofefficiencyonlywithgreatdifficulty.‘Suchlimitationsmaybeavoidedbydefiningrationalityintermsof164 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYeffectiveness.Effectivenessisdefinedasawiderconceptthanefficiency,andrefersinReasoninSocietytothesuccessfulproductionofanykindofvalue,leavingopenthequestionofthekindsofvaluetheremaybe.Theefficientachievementofpre-determinedgoalsisbutaspecialkindofeffectiveness;ifthereareotherkindsofvaluebesidesgoalvalues,therearepresumablyotherkindsofeffectivenessorrationalityaswell(Diesing,1962,p.3).Thisisnottosaythatthemeaningof‘efficiency’maynotbeexpandedtoencompassagreatdealmoreofhumanactivity.Wordsarenotoriouslyflexible,andasWildavskycomplains,themeaningof‘efficiency’hasbeengreatlyexpandedasitisusedinsystemsanalysisandprogrammebudgeting(1966).Diesingiscarefultoexplainhimselfonthispoint,writingthat:‘Mypurposeisnottoargueaboutterminologyortoattacksacredsymbols;consequentlyifanyonewishestoarguethatsocial,legal,andpoliticalrationalityareallultimatelyspecialvariantsoftechnicaloreconomicrationalityIshallnotobject,solongasallthedifferencesIpointoutarerecognizedandpreserved.Idoobjecttothepropositionthattech-nicaloreconomicreasoningcanbeappliedtosocial,legal,andpoliticalproblemswithoutveryextensivemodification.NordoIthinkthattheuniquecharacteristicsofsocial,legal,andpoliticalreasoningcanbebroughtoutbyderivingorreducingthemtoeconomicreasoning;quitethecontrary,theyarebroughtoutmostclearlybycontrastwitheconomicreasoning.’(Diesing,1962,pp.2-3.)Diesingdistinguishesbetweenthree‘aspects’or(asIcallthem)‘levels’ofrationality.HefollowsMannheimindistinguishingbetween‘substantial’and‘functional’rationality,andnotesthatathirdaspectofrationalitycanbededucedbyabstractionfromtheothertwo.Thatis,sincedecisionsaremadeaccordingtoprinciples,andorganizedstructuresembodyprinciplesoforder,principlescanalsobethoughtofasrational(Diesing,1962,p.4).Forourpurposes,thedistinctionbetweensubstantialrationalityandfunctionalrationalityisofparticularimportance.‘Adecisionoractionissubstantiallyrational(myemphasis)whenittakesaccountofthepossibilitiesandlimitationsofagivensituationandreorganizesitsoastoproduce,orincrease,orpreserve,somegood.Thisdefinitionincludestwopoints:thedecisionmustbeaneffectiveresponsetothesituationinthatitproducessomepossiblegood,andtheeffectivenessmustbebasedonintelligentinsightratherthanonluck.Anorganizationisfunctionallyrational(myemphasis)...whenitissostructuredastoproduce,orincrease,orpreserve,somegoodinaconsistent,dependablefashion.Theconsistentlygoodresultsmustbebasedprimarilyonaninternalstructurewhichisabletocontinueeffectiveoperationthroughvariationsofpersonnelandthroughchangesofenvironment.’(Diesing,1962,pp.3-4.)ThewayinwhichthefivetypesofrationalitywereidentifiediscentraltothevalidityofDiesing’swork.Theunusualapproachoftheauthorallowsfortheclaimthattheanalysismaybecomprehensive,thatallthebasicvarieties165 PUBLICADMINISTRATIONofrationalityhavebeen,ormaybe,incorporated.Beforedescribingtheindividualtypesofrationality,itthusseemsprudenttodescribethewayinwhichtheywereindentified.Thefivetypesofrationalitywerediscoveredneitherthroughsystemizationofcommonusage,norprimarilythrougha-priorilogicalanalysis.TheywereidentifiedasbeingtheoutcomesofwhatS.C.Peppercalls‘naturalselectivesystems’(Pepper,1958,inDiesing,1962,p.6).Thatis,throughthelargelyunconsciousdecisionsofmillionsofpeople,choiceisexercisedonculturetraits-techniques,rules,beliefs,values.‘Effective’culturaltraitsaresuccess-ful,aretransmitted,andreappear,muchinthewaythatbiologicalevolutionissaidtooccur.Theselectionofeffectiveculturaltraitsresultsfromautomaticrewardsandpunishmentswhichfollowtheuseofatrait.Thenatureoftherewardorpunishmentwillvaryaccordingtothenatureoftheselectivesystem:onetraitmightfunctiontoincreasepower,andanothertodecreaseanxiety.InDiesing’sscheme,selectivesystemsdonotoperatethroughthesurvivalornon-survivalofindividualsorsocieties.Whatsurvivesistheculturaltraititself.Ultimately,theoperationofaselectivesystemismanifestedasatrendofdevelopment.Atrendofdevelopmentisagradualandcumulativeprocessofchangethroughwhichcertaindefinitecharacter-isticsappearinaculture.Aprocessofthisnatureisautomaticinthatitoccursapartfrom,orevencontraryto,conscioushumanintentions.Technologicalprogressissuchatrend,andgivesrisetotechnicalmodesofthinkingandtechnicallyrationalformsoforganization.TheothermajortrendswhichDiesingidentifies-andwhichcorrespondrespectivelytoeconomic,social,andlegalrationality-areeconomicprogress,integration,andlegalism.Eachtrendmaybeintellectualizedintheformofprinciplesfordecisionmaking,andeachisproductiveofitsdistinctivekindofvalue.Toqualifyasatypeofrationality,atrendofdevelopmentmustbevirtuallyuniversal,althoughsincethevariousformsofrationalityareinpartopposedtoeachother,thecorrespondingtrendscouldnotalloccuratthesametime.However,asDiesingwrites,‘onlywhenbothempiricalevidenceofnearuniversalityandanalyticalevidenceoftheexistenceoflogicalconditionsofeffectivenesscoincidecanoneclaimtohavediscoveredatypeofrationality’(1962,pp.7-8).Havingidentifiedthevarious‘formsofpracticalreason’,hecontinuesineachcasetodescribethehistoricaltrendofdevelopmentandthetypeoforganizationitproduces;themethodofmakingdecisions;theconditionsunderwhichthismethodisappropriate;andthekindofgoodproduced.TYPESOFRATIONALITYTechnicalrationalityappearsinactionsundertakeninordertoachieveagivenend.Themeans-endconstructisfundamentaltotechnicalrationality,andalsotothecloselyrelatedformofeconomicrationality.Actionswhicharerepeatedoverandoverbecomestandardizedintotechniques,whichare166 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYalsogoaloriented,andthusexemplifytechnicalrationality.Technologicalprogressistheassociatedtrendofdevelopment,someformsofwhichinvolveincreasedhumanskill,otherstheincreasinguseofmachineryandnewformsofenergy,andstillothersspecializationoflabour.Successfultechnologiesortechniquestendtoperpetuatethemselves(Diesing,1962,pp.9-11).Technologicalprogresscanbeseeninvirtuallyallsocieties,butismostobviousinWesternculture,whereithasbeenreinforcedbythemechanismsandassociatedvaluesofeconomicprogress.Organizationsthatachievetheirendsefficientlyarefunctionallyrational,whiledecisionsleadingtoefficientgoalachievementaresubstantiallyrational.Atechnicallyrationalorgan-izationisoneinwhicheachsteporactivitymakesthemaximum(feasibleorreasonable)contributiontoaproductivesequence,resultingintheachieve-mentofagivengoal.However,technicaldecisionsarenotpossibleuntilaftertheeconomicquestionsofcomparativecostshavebeenanswered.Thebestmeanstoanendisthatwhichmakesthegreatestcontributionattheleastcost,butthiscannotbedetermineduntilcomparativecosts(prices)havebeenset.Forexample,inanunderdevelopedcountry,thecivilengineermustbeabletocomparethecostsofmachineryandmanhoursinordertodeterminewhetherlabour-intensiveorcapital-intensivemethodsofconstructionaremostefficient.Thisisbynomeansasimplerequirement,however,asitimpliestheexistenceofamarketeconomyandallthechangeswhichaccompanyeconomicprogress,forinstance,makinglabourintoacom-modity.5Itfollowsinsimilarfashionthatanorganizationcanbetechnicallyrationalonlyinthecontextofasystemofprices.Thebasicprincipleoftechnicalrationalityis‘adaptmeanstoends’.Thisprinciplealsoappliestoeconomicrationality,butismoreusefullyputintermsoftheprincipleofeconomizing:‘Economizingisanevaluationandselectionofends,anditoccurswhentwoormoreendsareincompetitionwitheachother.Itoccurs,forexample,whenoneisinapositiontoask,“ShouldIdothisorthattoday?”...or“Whatorderofpriorityshallweassignthesetasks,incaseweareunabletodoallofthem?”’(Diesing,1962,p.43.)Economizingisnecessarywhentheachievementofoneendimpliesthesacrificeofanother,andeconomizingispossibleonlytotheextentthatendsarecomparableonsomescale.Economicprogressfacilitateseconomizingbyincreasingthealternativenessofendsthroughtheremovalofmorallimitationsonendsandontheuseofmeans,andbythedevelopmentofmediaofmeasurementandcomparison(Diesing,1962,pp.44-47).Puresocialrationalityistherationalityofsocialsystems.Asocialsystemisanorganizationofculturalroles,includingexpectations,obligations,andideals.Onlybysharingacommonsocialsystemwhich,likeadrama,providesindividualswithcomplementaryrolescanpeopleunderstandandrelatetooneanother.Allaction,thought,emotion(andsymbolism)hasabasicsocialaspect,sincepersonalityiscomposedvirtuallyentirelyofsocialrolesandfeelings.Socialrationalityismosteasilyseeninthewritingsof161 PUBLICADMINISTRATIONanthropologistsdescribingtheunderlyingsocialstructureandbehaviourpatternsofisolatedpeopleswhosecultureshavebeenundisturbedbyeconomicprogress.Thebasictrendofisolatedsocialsystems,Diesingwrites,istowardgreaterintegration.(Socialsystemsareintegratedwhentherolesofwhichtheyarecomposedareinternallyconsistent.)Theuniversalityoftheintegrativetrendindicatesthepresenceofsomeprincipleofeffectiveness(i.e.rationality).Thiseffectivenessmaybeseenasthatofpromotingaction:‘Integrationisalogicalpreconditionforthesuccessfulcompletionofanysocialaction.Itmakesactionpossibleby(1)channellingthenecessaryemotionalenergyandpreventingitfrombeingdiffusedandlost;(2)elim-inatingconflicts,whichwouldblockaction;(3)providingsupportingfactorswhichstrengthenactionandcarryittocompletion.Also,(4)itmakesactionmeaningful...byrelatingittopastactionswhichitfulfillsandtofutureactionswhichpreserveandcontinueitsachievement.Anisolatedaction,withnohistoryandnoconsequences,isinsignificant;itdisappearsandisforgotten.’(Diesing,1962,pp.84-85.)Anintegratedsocialsystemisthusarationalone,asitiseffectiveinmakingactionpossibleandmeaningful.Anintegrativetrendinasocialsystemisatrendtowardrational,socialorganization.(Atrendtowardequilibriumisasecond-bestsolution.)Morespecifically,arationalsocialorganizationmaybesaidtoconsistofstructuralelementsandvalueelements.Structurally,rolesmustbeinternallyconsistent;theymustfittogetherwithoutconflict;thereshouldbenosharpdiscontinuitiesinrolesequences;andthesocialsystemshouldbecompatiblewiththenon-socialenvironment.Invalueterms,thecompletelyintegratedorsociallyrationalorganizationwillbecharacterizedbyparticularism,loyalty,andascription-valuesusuallyconsideredanti-theticaltomeans-endrationality.Alloftheseelementsareconducivetothestabilityofthesystem.Sociallyorienteddecision-makingattemptstochangepersonalitiesandsocialrelationsinthedirectionofgreaterfundamentalharmonyandstability.Themeans-endschemeisessentiallyirrelevant,becauseneithermeansnorendscanbetreatedasfixedorindependent-letalonemeasurable.Incon-trasttoamaximisingdecision,whichbeginswithgivenends,theintegrativeapproachtreatsendsassymbolicofhiddenvalues,fears,strains...(Diesing,1958,p.2).Thesubjectofanalysisisaninterrelated,largelyunconscious,processorsystem-meaningthattheterm‘function’willusuallybemoreappropriatethan‘end’orspecificpurpose.Meansandendsmaychangeun-predictablyinthecourseofproblem-solving,andonlyverygeneralgoalssuchasintegrationortension-reductionareatallrelevant.Socialandeconomicrationalityarethusclearlyopposedinmanyrespects,yeteachformofreasonpresupposesandisentirelydependentupontheother.Forexample,theintegrationmadepossibleinsocialorganizationsandinpersonalitiesprovidesthestabilitythatenablescalculationandtechnicallyrationalactiontooccur(Diesing,1962,p.95).Legalrationalityistherationalityoffundamentalrules.Thefundamental168 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYrulesinasocietymayconsistoftheconstitutionandthelaws,themoralorder,andperhapsthestatussystem.Thetendencyofsocietiestodeveloprules,Diesingbelieves,isaconsequenceoftheneedtoregulateandstructureconflictsamongthenormativeelementsofsociallife(Diesing,1962,p.124).Rulesservetoproducepredictabilityandformalorder.Onaninstitutionallevel,legalsystemsmaybethoughtofasthepublicsettingofsocialandeconomiclife,whichistosay,‘themeansthroughwhichthewholeofasocietyparticipatesineachindividualaction,offeringitsacquiescenceorsupport’(Diesing,1962,p.135).Ifintegrationisthefundamentalsocialtrend,legalismmaybeseenasthefundamentallegaltrend.Legalismcanbedefinedas:‘(1)atrendtowardcomplexityofdistinctionsandclarityofdetail,suchashighlytechnicalterms,(2)atrendtowardclearanddistincthierarchicaldifferentiation,forexample,jobspecification,(3)atrendtowarduni-formity,equality,anduniversalizationwheredifferentialsarenotinvolved,and(4)moregenerally,atrendtowardrigidity,unchangeability,actionaccordingtorule.Oncerulesaremade,theymaybeclarified,mademoreprecise,extended,butnotchanged.’(Diesing,1962,p.140.)Legalismoccursonlyinthepresenceofclearandpermanentdifferencesofinterest.Intightlyintegratedsocialsystems,conflictisavoidedwithouttheneedforintricate,formalrules.Indeed,fromthestandpointoftheintegrativeideal,lawisadistinctlyinferiorsolution,becauseitstabilizesconflictswithoutresolvingthem.Nevertheless,alegalordercanprovidesolutionsandpreventdisputeswhenothermethodsfail.Thisisitsclaimtorationality(Diesing,1962,pp.143-46).Politicalrafionalityistherationalityofdecision-makingstructures.Politicaldecisionsareconcernedwiththepreservationandimprovementofdecisionstructures.Alldecisions,tobesure,occurwithinsomesortofdecisionstructure,buttheidentifyingcharacteristicofpoliticaldecisionsisthattheyhavedecisionstructuresastheirspecialsubjectmatter.6Itistruethatineverydaylanguage,‘politics’isusedinamoregeneralsense,andthattheconceptoftenreferstotheimplementationofpublicorgroupdecisions.However,intermsofDiesing’sframework,suchactivitylieslargelyintherealmoftechnicalandeconomicrationalityandisnotofdirectconcernhere(Diesing,1962,p.170).Forexample,apoliticallyrationaladministratorintheU.S.ArmyCorpsofEngineerswillbeconcernednotonlywiththecontentofawater-useplan,butalsowiththestateoftheon-goingdecision-makingprocess.Hemaydecidetoincludealow-priorityprojectinhisplaninordertoplacateregionalinterests,andifthisisdonepurelytosecurehisownposition,itrepresentsnomorethanself-servingtechnicalrationality:fittingameanstoanend.However,suchaconcessiontopotentiallydamagingpressuremightbemadeintheinterestofpreservingaviableworkingrelationshipbetweentheCorpsofEngineersandtheCongress,servingtodefusetensionsandfacilitatesubstantivedecisionmaking.Ifsointended,andcarefullyexecuted,thismanoeuvrewouldbeanexampleofpoliticalrationality.’169 PUBLICADMINISTRATIONDiesingconcludeshisbookwithanoveralldefinitionofreason(1.962,pp.235-47).Suchadefinition,henotes,musthavetwoparts:onetocorrespondtotherationalityoforganizations(functionalrationality)andtheothercorrespondingtotherationalityofdecisions(substantialrationality).Functionalrationalityisseensimplyasorder,andrationalnormsarethere-foreprinciplesoforder.Technicalrationalityisanorderofproduction.Rawmaterialsenterasystem,areprocessed,andbecomeproducts.Economicorderisanorderofmeasurementandcomparisonofvalues.Commoditiesaremeasuredandexchangedinordertomaximizevalue.Socialrationalityisanorderofinterdependenceorsolidarity,whichexistswhenpeopleunder-standoneanother,acttogether,andsharecommonexperiences.Legalorderisanorderofavailability.Itspecifieswhichresourcesareavailabletoeachlegalentity,andwhichpersonscorrespondtowhichactionsandroles.Legalorderexistswheneachpersonknowswhathecandoandmustdo.Finally,thereisthe‘political’orderofdiscussionanddecision.Structuresofthistypearerationaltotheextentthattheyadequatelyprovideforthegatheringandprocessingofinformation,fortakingdecisions,andforcheck-ingtheeffectivenessofsuchdecisions.Nowiffunctionalrationalityisorder,Diesingwrites,subslantialrationalitycanbestbeseenasthemakingoforder,orcreativity(p.240).Substantiallyrationaldecisionsofvarioustypesultimatelycreatethedifferentordersoffunctionalrationality.Thatis,underfavourableconditions,economicallyrationaldecisionswill(indirectly)createaviableeconomicorder;legaldecisionswill(indirectly)createlegalorder;andsoforth.Thus,economicorderarisesoutoftheconstantmeasurementandcomparisonofvalueswhichoccursasindividualsseekthemostdesirablealternatives;legalorderarises,givenpropitiouscircumstances,asaresultofthecontinualapplicationofrulestocases;andsoon.TYPESOFRATIONALITYANDTHEPARADIGMPROBLEMAllfivetypesofrationalityindecisionmaking,onboththefunctionalandthesubstantial“levels”,aretobefoundinadministrativebehaviour.Yet,althougheachmodeofpracticalreasonisrelatedto,andisdependentupon,theothers,eachstillreflectsadistinctandindependentsourceofevaluativecriteria,andtherearemanysituations-torepeat-inwhichthecriteriaoftechnical,economic,social,legalandpoliticalrationalityconflict.This,Ibelieve,largelyexplainsthefundamentalcontradictionsandambiguitiesoftheliteratureofpublicadministration.Thestudyofcompleteorganizationsnecessarilyinvolvesmultipleaspectsandtypesofrationality,whichmeansthatanentirelysuccessfulanalysisandevaluationoftotalrationalitywouldrequiretheexistenceofanintegratedtheoryofthesocialsciences.Inthemostgeneralsensethen,theparadigmcrisesofpublicadministrationandorgani-zationaltheoryreflectthoseofsocialscienceandpoliticalphilosophy.Thisistosaytheobvious,thatthereisnosingle,integratedtheoryofthesocialsciences,andlikewise,thatthereisnoone,agreeduponorsatisfactorilyI70 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYdevelopedphilosophyofpoliticsandgovernmentwhichseemsadequateinrelationtotheproblemsofthemodernworld.Publicadministrationcannotfairlybeblamedfortheconditionofthesocialsciencesandphilosphy,butneitherdosuchextenuatingcircumstancesnegatetheexistenceofapara-digmcrisisinpublicadministration.Thestateofpublicadministrationtheoryisstillconfusion,andtheproblem,inPeterSelf‘swords,isthe‘greatdiversityofpartialexplanationsoforganizationalbehaviour,withnorealacademicwayofsynthesisingthemorchoosingbetweenthemwhentheyconflict’(1973,p.49).Muchprogresswillhavebeenmadeifwecanclearourheadsandourvocabulariesoftheconfusioninwhichwehaveworkedforsolong.This,Ibelieve,canbethecontributionofDiesing’sbook.Again,thethesisofthispaperisthatpublicadministrationtheorycanbeusefullyorganizedandbroughtintofocusthroughthelensof‘rationality’.Thus,ifthereaderwillacceptthevalidityofDiesing’sanalysis,theoverallevolutionofpublicadministrationtheorycanbesummarized,andhopefullyclarified,asfollows:Theearlystudentsofindustrialmanagementandpublicadministration,manyofthemengineers,believedthatthefundamentalproblemsoforganizationswerethoseoffechniculrationality,whichistosayofincreasingnarrowly-definedefficiency.Theirimplicitmodelofbureaucracywasthatofamachine,forwhichtheyprescribedprinciplesoforganizationandco-ordination.Thehumanrelationstheoristsdiscoveredtherelevanceofsocialrationalitytotheoperationofbusinessorganizations.Theycametoappreciatetheroleofemotionsandmoraleonthejob,andthesignificanceofinformalorganizationingeneratingsupportfor,oroppositionto,formalorganizationgoals.HerbertSimoninsistedthatorganizationsshouldbeanalysedprimarilyintermsofdecisionmaking.Sincegovernmentagenciesandbusinessesareessentiallypurposiveentities,heargued,theanalystshoulddirecthisattentiontothethenneglectedproblemofgauginglevelsof‘rationality’-i.e.‘efficiency’-inorganizationaldecisionmaking.InwritingAdminisfrufiveBehavior,Simonemphasizedtheobjectiveofincreasingtheeconomicrationalityoforganizations.Hedefined‘rationality’intermsof‘efficiency’,butdefined‘efficiency’verybroadly(p.75).Heisopentocriticismhereonlybecausehemissedseeingtheinherentconflictsbetweeneconomicandnon-economicvarietiesofrationality(seeSelf,1973,pp.35-36).Insum,Simonextendedtheargumentofthetraditionaltheoriststoamoresophisticatedlevel,fromaconcernwithtechnicalefficiencytoaconcernwitheconomicefficiency-whileatthesametimerecognizingthepsychologicalandotherlimitationsonthe‘rationality’ofisolatedindividuals.Weber,althoughwritinghalfacenturybeforeSimon,wassomewhatunfairlybroughtintothediscussionatthispoint,whenhisworkbecameavailableinEnglishtranslation.Like(early)Simonandthetraditionaltheorists,hepositedadistinctionbetweenpoliticsandadministrationandheneglectedtheroleofinformalorganization-whichthehumanrelationstheoristsandsociologistslikeRobertMerton,havesubsequentlystressed.Weberwasfundamentallyinterestedintheprocessesofsocietalrationalization,and171 PUBLICADMINISTRATIONindeedourunderstandingof‘rationality’isfundamentallybaseduponhiswork.Diesingtakesthedistinctionbetween‘substantive’and‘functional’rationalityfromKarlMannheim,whointurnderivedhisconceptionofthetermfromWeber,althoughinmodifiedform.Weberthusconceivedofanumberofmeaningsof‘rationality’,buthedescribedbureaucracyessentiallyintermsoftechnicalandlegalrationality.Bureaucracyisameanstoanend,butthisendisachievednotsimplythroughtechnicalknowledge,butalsobymeansofrulefollowingapointwhichisperhapsexcessivelyobviousinaGermanicculture.IfWeber’sworkistobecriticized,itshouldbeforinadequateattentiontothepotentialdecreaseinorganizationalefficiencywhichmayresultfrom‘legalism’(seeMouzelis,1967,p.47).Suchanoversight,tobesure,isquiteunderstandableinGermany,wheresocietalnorms-i.e.sociallyrationalnorms-areinvisiblysupportiveofbureau-craticefficiency.Finally,theimportanceandinevitabilityofpoliticalcriteriainadministrativedecisionmakinghasbeenrecognizedbymanywriters.Thepolitics/administrationdichotomyhasbrokendown,andpoliticalrationalityhasbeenaccordeditsuncertainplaceinthepuzzleofadministrativetheory.*Policyanalysis,whichistodaysopopular,normallyassumestheper-spectiveof‘substantial’rationality.Thepurposeofpolicyanalysisistodefine,andsubsequentlyachievecertainobjectiveswithoutcausingunaccept-ablechangesinbackgroundvariables.Abureaucracy,however,issimplyameanstoanend,andoneorganizationisasgoodasanother.AsOstromwrites:‘Whenthecentralprobleminpublicadministrationisviewedastheprovisionofpublicgoodsandservices,alternativeformsoforganizationmaybeavailablefortheperformanceofthosefunctionsapartfromanextensionandperfectionofbureaucraticstructures.’(1973,p.19.)Moststudentsofbureaucracy,however,taketheorganizationasgivenandattempttoimproveitsperformance.Thesameistrueformostpractitioners,whomustnecessarilyworkwithinthesystem,sotospeak.Theythustendtoadopttheperspectiveof‘functional’rationality.Thisisimportant,becauseitcomplementsthenecessarilypartialperspectiveofpolicyanalysis.FrankSherwoodwritesthat:‘Thereisaparticulardangerwhentheorganizationisregardedonlyasavehiclefortheachievementofadesiredprogramorproject.Theinterestsoftheorganizationandtheachievementofaspecifiedprojectmaycollideinanumberofways...Projectstendtoinvolvetargetsandtermination.Institutionsarecharacterizedbyshiftinggoalsandcontinuity.’(1973.p.200.)Itisnotuncommon,however,foranalyststoconsideronlythe‘substantial’oronlythe‘functional’levelsofrationality,ortoconfusethetwo.Thus,asMouzeliswrites,thesupposedlyintegrativecharacteroforganizationtheorydisappearswhenonemovesfromtheindividualleveltoconsiderationoftheorganizationasawhole:‘Indeed...whenoneconsidersthewholeorganisationasanetworkof172 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYdecisioncentresorasaninformationprocessingsystem,itisverydifficulttoaccountforitsculture,foritsstatussystem,andforallthoseorganiza-tionalfeaturestraditionallytreatedundertheinformalorganizationlabel.’(1967,p.162.)TheMeaningojConceptsTheDiesingframeworkalsoseemshelpfulinclarifyingthevocabularyofadministration.Thisisobviouswithrespectto‘organizationaleffectiveness’,sinceDiesingdefines‘rationality’intermsofeffectiveness.Hedefines‘effectiveness’initially--itwillberecalled-asthesuccessfulproductionofanykindofvalue,leavingopenthequestionofthekindsofvaluetheremaybe(1962,p.3).Heconcludes,asdescribedabove,thattherearefivefundamentaltypesofeffectiveness(rationality),whichoperateonthelevelsofindividualandorganizationaldecisionmaking.(LawrenceMohr’sdistinctionbetween‘transitive’and‘reflexive’effectivenessisanalogous,ifnotidentical,toDiesing’sdistinctionbetween‘substantial’and‘functional’rationality/effectiveness(1973.p.475).)Alltypesandlevelsofeffectiveness/rationalityaretobefoundinformalorganizations.Mostformalorganizations,tobesure,areprimarilyproductivesystems,meaningthatdecision-makingisdominatedbythecriteriaoftechnicalandeconomicrationality.However,allhumanorganizationsalsodevelop(sociallyrational)varietiesofinformalorganizationaswell.Iftheorganizationislarge,itwillalsodevelopastatussystemandasetofrulesbywhichtomediatedisputes-i.e.alegalorder.Finally.everyformalorganizationwillneedtodevelopadecisionstructureorcontrolsystem(manifestingpoliticalrationality),whichmayormaynotcorrespondtotheformalorganizationchart(Diesing,1962,p.239).Mostevaluationsoforganizationalperformanceconcentrateonlyuponcertaintypesofeffectiveness,oftenspecifiedintermsofconcretegoals.However,thisneednotalwaysbethecase.Therelativeimportanceofthedifferenttypesandlevelsofrationality/effectivenessinparticularorganizationsundervary.ingcircumstancesisnotself-e~ident.~Adecisiontostudyonlyacertaintypeofeffectivenesshasnormativeimplications,andmustbejustified.Or,ifmultipletypesofeffectivenessaretobestudiedsimultaneously,thereremainsthedifficultproblemofhowtointegratethepotentiallyconflictingstandards.Next,considertheconceptof’legitimacy’.‘Legitimacy’canbeseentohavediscretemeaningsinthecontextsofdifferenttypesofrationality.Thedictionarydefinitionofthetermas‘conformingtorecognizedprinciplesoracceptedrulesandstandards’clearlybelongstotheparadigmoflegalrationality.Ontheotherhand,Lipset’sstatementthat‘legitimacyinvolvesthecapacityofthesystemtoengenderandmaintainthebeliefthattheexistingpoliticalinstitutionsarethemostappropriateonesforthesociety’(1960,p.64)alsoincorporatessocialcriteriaofrationality.Socialrationalityisrelevantherebecauseabeliefistheproductnotsimplyofrulesorinform-ation,butofaffectorfeelingsaswell.FeelingsoflegitimacyarelargelyI73 PUBLICADMINISTRATIONproductsofthemeaning-generatingandconsensualmechanismsofsocialsystems.Mythirdexampleofconceptualambiguitywasthatof‘administrativeresponsibility’.Gilbert(1959)foundthatthetermhasbeenusedinprofes-sionaldiscoursetorefertoatleasttwelvedifferentvalues,anddifferentschoolsofadministrativethoughtusetheconceptinradicallydifferentways.Traditionally,studentsofadministrativebehaviourhavedefined‘respon-sibility’as‘accountabilitytoanoutsideentity,forsomekindofper-formance’.Fromthehumanrelationsororganizationdevelopmentperspective,however,thecentralconnotationof‘responsibility’issubjectiveorpsychological:theresponsibleemployeeistheindividualwhobehaves‘responsibly’becausehefeelsresponsible.InFrederickMosher’swords,‘thismeaningismorenearlysynonymouswithidentification,loyalty,andconsciencethanitiswithaccountabilityandanswerability’(1968,p.8).Thelikelihoodofanemployeeperformingaccordingtohighobjectivestandardswilldependnotsimplyuponhisintelligence,thelevelofextrinsicrewardsprovided,andthepossibilityofsanctionsfornon-performance,butalsouponprocessesofsocializationandassociationwithin(andwithout)theorganization.Thisimpliesthatdirectivesfromhierarchicalsuperiorsmaynotbeobeyediftheyruncountertodeeplyheldloyaltiesofsubordinates.Stillotherbasicmeaningsof‘responsibility’areillustratedinMaxWeber’sfamousessay,‘PoliticsasaVocation’(GerthandMills,1946,pp.77-128).Weberarguesthatthepoliticalleader’sconductissubjecttotheoppositeprincipleofresponsibilityfromthatofthecivilservant:‘Thehonorofthecivilservantisvestedinhisabilitytoexecutecon-scientiouslytheorderofthesuperiorauthorities,exactlyasiftheorderagreedwithhisownconviction.Thehonorofthepoliticalleader,oftheleadingstatesman,however,liespreciselyinanexclusivepersonalrespon-sibilityforwhathedoes,aresponsibilityhecannotandmustnotrejectortransfer.Itisinthenatureofofficialsofhighmoralstandingtobepoorpoliticians,andaboveall,inthepoliticalsenseoftheword,tobeirresponsiblepoliticians.’(GerthandMills,1946,p.95.)Weber’scivilservantdoesnotobeyhisorganizationalsuperiorsimplybecausehierarchicalorganizationseemstobethemost(technically)efiicientwayofstructuringapublicentity.Weber’sbureaucratobeysbecauseitis-orseems-righttofolloworders,andhethushasadutytodoso.Thisistheessenceoflegalrationalityorlegalism(Shklar,1964,p.1).Thestatesmanorpoliticalleader,ontheotherhand,hasadutynottofollowrules,buttomakepolicydecisions.Thepoliticalleaderisprimarilyresponsibleformakinggooddecisions,andthushemustconcernhimselforherselfwiththestateofthedecisionmakingandinformationgatheringprocess,whichistheessenceofDiesing’sconceptionofpoliticalrationality.However,sinceWeberwrote,theabsolutedistinctionbetweenpoliticsandadministrationhasbrokendown,whichmeansthattopadministratorsarenowexpected(withinlimits)tomakepolicydecisions.Consequently,it174 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYseemsclearthatthehigheradministratorsandcivilservantsmustbepreparedtoassumeadditionalpoliticalresponsibilityfortheiractions.Anadequateexaminationoftheproblemof‘administrativeresponsibility’isnotpossiblewithinthescopeofthispaper.Nevertheless,ageneral-andbynowpredictable-hypothesismaybeadvancedatthispoint.Inthelowercase,Isuggestthatthemeaningoftherelationalconcept,‘responsibility’,shouldbederivedfromthemeaningof‘rationality’,asdefinedhere.Iftherearefivetypesandtwolevelsofrationality,itthusfollowsthattherewillbeanequalnumberoftypesandlevelsofresponsibility.Thus,abureaucratwhofollowsdirectivesfromhierarchicalsuperiorsisactinginalegallyresponsiblemanner;anorganizationwhichspecifiesandreasonablyattemptstoachieveits‘substantial’goalsattheleastpossiblecostinscarceresourcesiseconomicallyresponsible;anorgan:ationwhichmanifestsasenseofcommunityamongitsemployeesissocdllyresponsible;andorganizationalleaderswhoassumeobligationstomakepolicydecisions,anddosowithintheconstraintsofthepowersituation,areactinginapoliticallyresponsiblemannervis-a-vistheirorganizations.Intheuppercase,however,‘Respon-sibility’maybeseentohaveanimportantmeaningofitsown.Itake‘AdministrativeResponsibility’torefertoevaluationsoforganizationalactivitywhichinvolvethecriteriaofmultipletypesofrationality.Decisionmakingofthissortinvolveschoosingbetween,orcombining,thecriteriaofthedifferenttypesand/orlevelsofrationality.Itisnecessaryinordertoanswerthefollowingtypeofquestion:Howshouldonedealwithasituationwheresocialconflictpartiallyblockstheachievementofbasicorganizationalgoals?Shouldabusinessfirmvoluntarilyforgoprofitsintheinterestofprotectingtheenvironment?Towhatextentshouldagovernmentagencyassumepolicy-makingresponsibilitynotspecificallygranteditintheagency’scharter?Suchquestions,Isubmit,cannotbeansweredfromtheperspectiveofanysingletypeofrationality/responsibility.loCONCLUSIONThereisstillnocure,butperhapstheconditionofpublicadministrationcannowbemoreaccuratelydiagnosedandtreated.Thiscondition,Ibelieve,ischaracterizedbyblurringofvisionandaconfusionofimages.Fivenewsetsofbifocalshavebeenmetaphoricallyprescribedforthoseamictedbytheparadigmproblem.Inordertoseeandunderstandthetotalityofpublicorganizations,itmaybehelpfultolookthroughallfivesetsofglasses.Eachbifocalrepresentsatypeofrationality,withtwolevels.Eachisabasicpara-digm,whichenablesustomakesenseofpartofthesocial-economicuniverse.Theultimateproblemofsocialscienceisthatwehaveasyetnowayofresolvingthefiveimagesoftheworldintooneclearpicture.Thecentralproblemofpublicadministration,however,hasbeenthatofnotcorrectlyidentifyingthedifferentparadigms,andofnotrecognizingthelegitimacyofeachperspectiveontheuniverseofpublicorganizations.Therearewell-developedtraditionsofscholarshipinpublicadministration175 PUBLICADMINISTRATIONandmanagementwhichcorrespondroughlytotheperspectivesofwhatDiesingcallstechnical,social,andlegalrationality-orsomecombinationofthethree.Somestudentsofpublicsectorproblemshaveadoptedexplicitlyeconomicandpoliticalperspectives,andyetotherwritershavebeguntoanalyseadministrativebehavioursimultaneouslyfromboth‘technical’and‘social’pointsofview(seeArgyrisandSchon,1975).Yetironically,theover-alloutcomeoftenseemsconfusionratherthanenlightenment,parochialism,ratherthantheemergenceofacomprehensivestructurefororganizationalanalysis.Aparticularlyinsidiousformofparochialismisthatwhichinsiststhat(economic)rationalityshouldbethecentralorganizingfocusforthestudyofpublicadministration.‘Rationality’willinevitablybemisleadingasanorganizingconcept,totheextentthatone’sunderstandingofthetermisoverlynarroworpartial.If‘rationality’isdefinedas(economic)efficiency,itbecomesvirtuallyimpossibletogivesufficientconsiderationtowhatDiesingcallsthesocial,legal,andpoliticalmodesofpracticalreason.‘Humanrelations’ororganizationdevelopmenttheoriesofadministrationarethusseenasunreasonableorirrational,anditisdifficulttoaccepttheviewpointsofthosewhoemployothermodesofreason-forexample,thepointofviewofthosewhoadvocateincrementalbudgeting.”MyproposalistouseDiesing’stypesandlevelsofrationalityasacom-prehensiveframeworkofidealtypes.I2The‘scientific’aspectoftheparadigmcrisisofpublicadministrationcanbeexplained-ifnotentirelyresolved-intheseterms.Thesolution,asIseeit,isnottostrain(andthusdistort)one’svisionbyattemptingtounderstandallproblemsfromasingleperspective.Itisrathertoemployeachtypeofanalysiswithinitscarefullydefinedrangeofapplicability,withtheknowledgethateachapproachwillbecomplementedorlimitedbytherelevanceofothertypesofrationalityinagivensituation.The‘normative’aspectoftheparadigmcrisishastodowiththeproblemofcombiningthecriteriaofdifferenttypesandlevelsofrationality,wherethisisnecessary.Insomecases,itisnecessarytochoosebetweenthecriteriaofdifferenttypesofrationality-forexample,whenthepoliticallyprudentsolutiontoa-problematicsituationisnotefficientorevenquitelegal.Inothercircumstances,theresolutionofaproblemmaydemandthecombinedmethodologiesoftheefficiencyexpertandthesocial-psychologist.Diesinghimself(1958)andmanyothershaveaddressedsuchissues,butmuchmoreneedstobedonebeforewecanreasonablyprescribeorevaluatesolutionstotheproblemof‘AdministrativeResponsibility’.Publicadministration,likeanthropology,hassetforitselftheambitiousobjectiveofstudyingthebehaviourofcompleteorganizations,andthebehaviourofcompleteorganizationsnecessarilyinvolvesmultipletypesofrationality.Theissuesarecomplex,andtheproblemsdifficult,butapublicadministrationexplicitlyorientedtowardtheinstitutionalstudyandevalu-ationofpublicorganizationsthroughtheuseofthefullrangeofrationalthoughtandbehaviourwouldbeamongtheleastparochialofprofessions.176 RATIONALITYANDTHFPROBLEMSOFADMINISTRATIVETHEORYNOTESI.Fordiscussionsofparadigmconflicts,seeKuhn,1970:Stephens,1973;Beardsley.1974,andRicci.1977.2.Hart,1967,andSpiro.1969.areparticularlyusefulinunderstandinghowtheconcept‘responsibility’isused.For‘administrativeresponsibility’,see,forexample,Maass,195I.Pennock.1952;Friedrich.andFiner,inAltshuler.1968;Gilbert,1959;Mosher,1968;Dorsey.1971;andKernaghan.1973.On‘thepublicinterest’,thefollowingsourcesillustratetherangeofdisagreementwhichexists:Schubert,1957;Redford,inO’Donnell,1966;Harmon,1969;Cochran,1974;andMitnick.1976.3.Thehumanrelationsofsocial-psychologicalapproachtoorganizationsisnowoftencalledthefieldoforganizationdevelopment.4.Diesing.pp.1-2.notesthat‘thevariousformsofpsychotherapy...canhardlybeconceivedastechniquesfortheprocessingofsomekindofhumanproduct....Themediationoflabourdisputes.ofpoliticaldisputes.andofinternationaldisputesissimilarlyresistanttoatechnicaloreconomictreatment’.5.Diesingwritesthatinthecourseofeconomicprogress:‘Oneculturalelementafteranotherhasbeenabsorbedintotheever-wideningeconomy.subjectedtothetestofeconomicrationality.rationalized.andturnedintoacommodityorfactorofproduction.Sopervasivehasthisprocessbeenthatitnowseemsthatanythingcanbethoughtofasacommodityanditsvaluemeasuredbyaprice,andthatallvaluescanbethoughtofasutilities’(1962.p.24).6.Diesingusestheconcept.‘decision-makingstructure’verybroadlytorefertothesystemofcommunicationwithinwhichhabitsofthoughtareappliedtomaterialstoresultindecisions.Inotherwords.adecision-makingstructureis:‘thesetofsocioculturaldeterminantsofpracticalthought,sincedecisionsareitsproduct.Notallthedeterminantsofthoughtareincluded,butonlythosewhichcanbecontrolledandstructured.Biologicaldeterminants...arenotincluded’(1962.p.170).7.Itmightbepossible,Diesingnotes,toconsiderpoliticaldecisionsasaspecialclassoftechniquesappropriateformaintainingdecisionstructures:‘Ifapersongetsintoacentralpositioninadecisionstructureandwishestodoagoodjobthere.hemustusethesetechniques.Buthereagain.asinthecaseoflaw.wearedealingwithpublic‘techniques’.Theyarepublicbecausetheyserveapublicgoal.Thisgoalispublicinthesensethatitbelongstoastructureratherthantoanindividual.andisrationallypriortotheaimsofIfanincreaseofcentralauthorityiscalledforandachieved.theresultsaccruetoarole.nottoaprivateperson.andcanbeusedforpersonalendstoaverylimitedextent.’(1962.p.217).8.Thissynopsisofthedevelopmentofpublicadministrationtheoryisnotmeantasintellectualhistory,andobviouslydoesnotincludeallthesignificantcurrentsoforganizationalthinking.Ithopefullydoesincorporateallthebasictypesofrationalitywhicharerelevanttoadministrativetheory.9.Aformalorganizationisnormallycreatedtoperformaparticularfunction.Whenthisfunctionisnolongerrequired.however,theorganizationwilloftenremainintact.havingfoundnewlegitimatinggoals.Thisistheprocessofgoaldisplacement.whichineffectrepresentsthetriumphoffunctionaloversubstantialrationality(seeHuntington.1968:pp.17-19).10.SeeDiesing’sarticle‘SocioeconomicDecisions’,1958.foragooddiscussionofhowthecriteriaoftwotypesofrationaldecision-makingmaybecombinedunderdifferentconditions.II.Thedoctrineofincrementalbudgeting,aspreachedbyAaronWildavsky(amongothers)canbeconsideredanattempttocombinethecriteriaoftechnical,economic.andpoliticalrationality.12.SeeMouzelis.1967.Ch.2.esp.pp.45-46.foradiscussionoftheproperusageoftheconcept.‘idealtype‘.I77 PUBLICADMINISTRATIONREFERENCESC.Argyris,‘SomeLimitsofRationalManOrganizationalTheory.’PublicAdministrationReview,33,May/June,1973pp.253-267,andD.A.Schon,TheoryinPractice.SanFrancisco:Jossey-Bass,1975.P.Beardsley,‘PoliticalScience:TheCaseoftheMissingParadigm.’PoliticalTheory,February1974,pp.46-61.D.BraybrookeandC.Lindblom,AStrategyofDecision:PolicyEvaluationasaSocialProcess.NewYork:FreePress,1963.N.Caiden,andA.Wildavsky,1974,PlanningandBudgetinginPoorCountries,NewYork:JohnWiley,1974.C.Cochran,‘PoliticalScienceand“ThePublicInterest”’JournalofPolitics,36,May1974,pp.327-355.R.Dahl,‘TheScienceofPublicAdministration:ThreeProblems.’PublicAdministrutionReview,7,Winter1947,pp.1-11.P.Diesing,ReasoninSociety:FiveTypesofDecisionsandTheirSocialConditions,Urbana:UniversityofIllinoisPress,1962,and‘SocioeconomicDecisions’,Ethics,69,October1962,pp.1-18.J.T.Dorsey,‘Administrators:BureaucratsandPolicy-Makers’,inJ.C.WahlkeandA.N.Dragnich(eds.)GovernmentandPolitics,NewYork:RandomHouse,1971,2nded.,pp.361-395.C.J.Friedrich‘PublicPolicyandtheNatureofAdministrativeResponsibility,’1940,inA.A.Altshuler(ed.)ThePoliticsoftheFederalBureaucracy,NewYork:Dodd,Mead,1968,pp.414-425.H.Finer,‘AdministrativeResponsibilityinDemocraticGovernment’,1941,inA.A.Altshuler,(ed.)ThePoliticsoftheFederalBureacracy,pp.425-32.NewYork:Dodd,Mead.H.H.Gerth,andC.W.Mills,1946,FromMaxWeber:EssaysinSociology,NewYork:OxfordUniversityPress.C.Gilbert,‘TheFrameworkofAdministrativeResponsibility,’JournalofPolitics,21August1959,pp.373-407.W.J.Gore‘JohnPfiffner’sJourneytoAmbiguity,’PublicAdministrationReview,32,July/August,1972,pp.318-327.M.M.Harmon,‘AdministrativePolicyFormulationandthePublicInterest,’PublicAdministrationReview,69,September/October1969,pp.483-491.H.L.A.Hart,PunishmentandResponsibility,NewYork:OxfordUniversityPress,1969.S.P.HuntingtonPoliticalOrderinChangingSocieties,NewHaven,Connecticut:YaleUniversityPress,1968.K.Kernaghan,‘ResponsiblePublicBureacracy:ARationaleandFrameworkforAnalysis,’CanadianPublicAdministration,16,Winter1973,pp.572-603.T.Kuhn,TheStructureofScientflcRevolutions,2nded.Chicago:UniversityofChicago:UniversityofChicagoPress,1970.C.E.Lindblom,ThePolicy-MakingProcess,EnglewoodCliffs,NewJeney:Prentice-Hall,1968.S.M.Lipset,PoliticalMan,GardenCity,NewYork:Doubleday,1960.A.Maass,MuddyWafers,Cambridge,Massachusetts:HarvardUniversityPress,1951.D.McGregorTheHumanSideofEnterprise.NewYork:McGraw-Hill,1960,andTheProfessionalManager.St.Louis:McGraw-Hill,1970.F.Marini,(ed.)TowardaNewPublicAdministration:TheMinnowbrookPerspective,Scranton:Chandler,1971.B.Mitnick,‘ATypologyofConceptionsofthePublicInterest’,AdministrationandSociety,8,May1976,pp.5-28.L.Mohr,‘TheConceptofOrganizationalGoal’,AmericanPoliticalScienceReview,67,June1973,pp.470-479.F.C.Mosher,(ed.)AmericanPublicAdministration:Past,Present,Future,University,Alabama:UniversityofAlabamaPress,1975,andDemocracyandthePublicService,NewYork:OxfordUniversityPress,1968.178 RATIONALITYANDTHEPROBLEMSOFADMINISTRATIVETHEORYN.P.Mouzelis,OrganisationandBureaucracy,Chicago:Aldine,1967.V.Ostrom,TheIntellectualCrisisinAmericanPublicAdministration,University,Alabama:UniversityofAlabamaPress,1973.J.R.Pennock,‘Responsiveness,Responsibility,andMajorityRule’,AmericanPoliticalScienceReview,46,September1952,pp.790-807.S.C.Pepper,TheSourcesofValue,Berkeley:UniversityofCaliforniaPress,1958.J.Pfiffner,‘AdministrativeRationality’,PublicAdministrationReview,20,Summer1960,pp.125-132.E.S.Redford,‘TheNever-EndingSearchforthePublicInterest’.1966,inM.E.O’Donnell(ed.),ReadingsinPublicAdministration,Boston:HoughtonMimin.pp.295-304.D.Ricci,‘ReadingThomasKuhninthePost-BehaviouralEra’WesternPoliticalQuarterly,30,March1977,pp.7-34.W.H.RikerandW.J.Zavoina,‘RationalBehaviourinPolitics:EvidencefromaThree-PersonGame’.AmericanPoliticalScienceReview,64,March1970,pp.48-60.G.A.SchubertJr.,“‘ThePublicInterest”inAdministrativeDecision-Making:Theorem,Theosophy,orTheory?’AmericanPoliticalScienceReview,64,March1957,pp.48-60.P.Self,AdministrativeTheoriesandPolitics,Toronto:UniversityofTorontoPress,1973.F.Sherwood,‘TechnicalAssistanceforEducationinPublicAdministration:LessonsofExperience’,inC.ThurberandL.S.Graham,(4s.)DevelopmentAdministrationinLatinAmerica,Durham,N.Carolina:DukeUniversityPress1973,pp.195-227.J.Shklar,Legalism,Cambridge,Massachusetts:HarvardUniversityPress,1964.H.A.Simon,‘OrganizationalMan:RationalorSelf-Actualizing?’,PublicAdministrationReview,33,July/August1973,pp.346-353,andAdminisIrariveBehavior,NewYork:FreePress,1945.H.J.Spiro,ResponsibilityinGovernment,NewYork:VanNostrandReinhold,1969.R.M.Steers,‘ProblemsintheMeasurementofOrganizationalEffectiveness’,AdministrativeScienceQuarterly,20,December1975,pp.546.557.J.Stephens,‘TheKuhnianParadigmandPoliticalInquiry:AnAppraisal.’AmericanJournalofPoliticalScience,August1973,pp.467-488.G.Strauss,‘SomeNotesonPower-Equalization’,1963inH.Leavitt(ed.)TheSocialScienceofOrganizations,EnglewoodCliffs,NewJersey:Prentice-Hall,pp.41-84.L.vonMises,EpistemologicalProblemsofEconomics.Princeton,NewJersey:VanNostrand,1960.D.Waldo,‘EducationforPublicAdministrationintheSeventies’,inF.C.Mosher,(ed.),AmericanPublicAdministration:Past,Present,Future.University,Alabama:UniversityofAlabamaPress,1975;‘ScopeoftheTheoryofPublicAdministration’.1968,inJ.C.Charlesworth,(ed.)TheoryandPracticeofPublicAdministration,Mongraphno.8inaseriessponsoredbytheAmericanAcademyofPoliticalandSocialScience,Philadelphia,pp.1-26;ComparativePublicAdministration:Prologue.Problems,andPromise,PapersinComparativeAdministration,SpecialSeries:No.2,ComparativeAdministrationGroup,February1964.A.Wildavsky,‘ThePoliticalEconomyofEfficiency:Cost-BenefitAnalysis,SystemsAnalysis,andProgramBudgeting’,PublicAdministrationReview,26,December1966,pp.292-310,andThePoliticsoftheBudgetaryProcess,Boston:LittleBrown,1964.M.N.Zald,andP.Denton,‘FromEvangelismtoGeneralService:TheTransformationoftheYMCA.’AdministrativeScienceQuarterly,8,September1963,pp.214-234.179 PUBLICADMINISTRATIONSocialandbconomicAdministrationEditedbyR.A.B.LeaperVolume12Number1Spring1978ALANMAYNARDTheMedicalProfessionandtheEfficiencyandEquityofHealthServicesA.J.CULYERWhatAccountsfortheHigherCostsofTeachingHospitals?PHILIPR.JONESTheGrowthoftheBritishMedicalAssociation:ACaseStudyinCollectiveBehaviourReviewArticle:F.M.LEWESMeasuringSocialWellBeingReviewsThejournalispublishedthreetimesayear,inSpring,SummerandAutumn,byBASILBLACKWELLinassociationwiththeUNI-VERSITYOFEXETER.Theannualsubscriptionisf8.00(inland),f9.60(overseas),$20.00(USAandCanada).Ordersandremittancesshouldbesentto:JournalsDepartment,BasilBlackwellBrMottLtd.,108CowleyRoad,OxfordOX4lJF,England.Iao