企业的性质科斯(英文版),企业的性质科斯英文版pdf百度云

('科斯：企业的性质（英文版）TheNatureoftheFirm(1937)R.H.COASEEconomictheoryhassufferedinthepastfromafailuretostateclearlyitsassumption.Economistsinbuildingupatheoryhaveoftenomittedtoexaminethefoundationsonwhichitwaserected.Thisexaminationis,however,essentialnotonlytopreventthemisunderstandingandneedlescontroversywhicharisefromalackofknowledgeoftheassumptionsonwhichatheoryisbased,butalsobecauseoftheextremeimportanceforeconomicsofgoodjudgmentinchoosingbetweenrivalsetsofassumptions.Forinstance,itissuggestedthattheuseoftheword“firm”ineconomicsmaybedifferentfromtheuseofthetermbythe“plainman.”\'Sincethereisapparentlyatrendineconomictheorytowardsstartinganalysiswiththeindividualfirmandnotwiththeindustry,2itisailthemorenecessarynotonlythatacleardefinitionoftheword"firm"shouldbegivenbutthatitsdifferencefromafirminthe"realworld,"ifitaists,shouldbemadeclear.Mrs.Robinsonhassaidthat"thetwoquestionstobeaskedofasetofassumptionsineconomicsare:Aretheytractable?and:Dotheycorrespondwiththerealworld?"3Though,asMrs.Robinsonpointsout,"Moreoftenonesetwillbemanageableandtheotherrealistic,"yettheremaywellbebranchesoftheorywhereassumptionsmaybebothmanageableandrealistic.Itishopedtoshowinthefollowingpaperthatadefinitionofafirmmaybeobtainedwhichisnotonlyrealisticinthatitcorrespondstowhatismeantbyafirmintherealworld,butistractablebytwoofthemostpowerfulinstrumentsofeconomicanalysisdevelopedbyMarshall,theideaofthemarginandthatofsubstitution,togethergivingtheideaofsubstitutionatthemargin.4Ourdefinitionmust,ofcourse,"relatetoformalrelationswhicharecapableofbeingconceivedexactly."5IItisconvenientif,insearchingforadefinitionofafirm,wefirstconsidertheeconomicsystemasitisnormallytreatedbytheeconomist.LetusconsiderthedescriptionoftheeconomicsystemgivenbySirArthurSalter6.“Thenormaleconomicsystemworksitself.Foritscurrentoperationitisundernocentralcontrol,itneedsnocentralsurvey.Overthewholerangeofhumanactivityandhumanneed,supplyisadjustedtodemand,andproductiontoconsumption,byaprocessthatisautomatic,elasticandresponsive.”Aneconomistthinksoftheeconomicsystemasbeingco-ordinatedbythepricemechanismandsocietybecomesnotanorganizationbutanorganism.7Theeconomicsystem“worksitself.Thisdoesnotmeanthatthereisnoplanningbyindividuals.Theseexerciseforesightandchoosebetweenalternatives.ThisisnecessarilysoifthereistobeorderinthesystemButthistheoryassumesthatthedirectionofresourcesisdependentdirectlyonthepriceTheNatureoftheFirm(1937)R.H.COASE2mechanism.Indeed,itisoftenconsideredtobeanobjectiontoeconomicplanningthatitmerelytriestodowhatisalreadydonebythepricemechanism.8SirArthurSalter\'sdescription,however,givesaveryincompletepictureofoureconomicsystem.Withinafirm,thedescriptiondoesnotfitatall.Forinstance,ineconomictheorywefindthattheallocationoffactorsofproductionbetweendifferentusesisdeterminedbythepricemechanism.ThepriceoffactorAbecomeshigherinXthaninY.Asaresult,AmovesfromYtoXuntilthedifferencebetweenthepricesinXandY,exceptif50farasitcompensatesforotherdifferentialadvantages,disappears.Yetintherealworld,wefindthattherearemanyareaswherethisdoesnotapply.IfaworkmanmovesfromdepartmentYtodepartmentX,hedoesnotgobecauseofachangeinrelativeprices,butbecauseheisorderedtodo50.Thosewhoobjecttoeconomicplanningonthegroundsthattheproblemissolvedbypricemovementscanbeansweredbypointingoutthatthereisplanningwithinoureconomicsystemwhichisquitedifferentfromtheindividualplanningmentionedaboveandwhichisakintowhatisnormallycalledeconomicplanning.Theexamplegivenaboveistypicalofalargesphereinourmodemeconomicsystem.0fcourse,thisfacthasnotbeenignoredbyeconomists.Marshallintroducesorganizationasafourthfactorofproduction;J.B.Clarkgivestheco-ordinatingfunctiontotheentrepreneur;ProfessorKnightintroducesmanagerswhoco-ordinate.AsD.H.Robertsonpointsout,wefind"islandsofconsciouspowerinthisoceanofunconsciousco-operationlikelumpsofbuttercoagulatinginapailofbuttermilk.”9Butinviewofthefactthatitisusuallyarguedthatco-ordinationwillbedonebythepricemechanism,whyissuchorganizationnecessary?Whyaretherethese"islandsofconsciouspower"?Outsidethefirm,pricemovementsdirectproduction,whichiscoordinatedthroughaseriesofexchangetransactionsonthemarket.Withinafirm,thesemarketstransactionsareeliminatedandinplaceofthecomplicatedmarketstructurewithexchangetransactionsissubstitutedtheentrepreneurco-ordinator,whodirectsproduction.10Itisclearthatthesearealternativemethodsofco-ordinatingproduction.Yet,havingregardtothefactthatifproductionisregulatedbypricemovements,productioncouldbecarriedonwithoutanyorganizationatall,wellmightweask,whyisthereanyorganization?0fcourse,thedegreetowhichthepricemechanismissupersededvariesgreatly.Inadepartmentstore,theallocationofthedifferentsectionstothevariouslocationsinthebuildingmaybedonebythecontrollingauthorityoritmaybetheresultofcompetitivepricebiddingforspace.IntheLancashirecottonindustry,aweavercanrentpowerandshop-roomandcanobtainloomsandyarnoncredit.11Thisco-ordinationofthevariousfactorsofproductionis,however,normallycarriedoutwithouttheinterventionofthepricemechanism.Asisevident,theamountof“vertical”integration,involvingasitdoesthesupersessionofthepricemechanism,variesgreatlyTheNatureoftheFirm(1937)R.H.COASE3fromindustrytoindustryandfromfirmtofirm.Itcan,Ithink,beassumedthatthedistinguishingmarkofthefirmisthesupersessionofthepricemechanism.Itis,ofcourse,asProfessorRobbinspointsout,“relatedtoanoutsidenetworkofrelativepricesandcosts,”12butitisimportanttodiscovertheexactnatureofthisrelationship.ThisdistinctionbetweentheallocationofresourcesinafirmandtheallocationintheeconomicsystemhasbeenveryvividlydescribedbyMr.MauriceDobbwhendiscussingAdamSmith\'sconceptionofthecapitalist:“Itbegantobeseenthattherewassomethingmoreimportantthantherelationsinsideeachfactoryorunitcaptainedbyanundertaker;thereweretherelationsoftheundertakerwiththerestoftheeconomicworldoutsidehisimmediatesphere...theundertakerbusieshimselfwiththedivisionoflabourinsideeachfirmandheplansandorganisesconsciously,”but“heisrelatedtothemuchlargereconomicspecialisation,ofwhichhehimselfismerelyonespecialisedunit.Here,heplayshispartasasingleceIlinalargerorganism,mainlyunconsciousofthewiderrôlehefills.”13Inviewofthefactthatwhileeconomiststreatthepricemechanismasacoordinatinginstrument,the?alsoadmittheco-ordinatingfunctionofthe“entrepreneur,”itissurelyimportanttoenquirewhyco-ordinationistheworkofthepricemechanisminonecaseandoftheentrepreneurinanother.Thepurposeofthispaperistobridgewhatappearstobeagapineconomictheorybetweentheassumption(madeforsomepurposes)thatresourcesareallocatedbymeansofthepricemechanismandtheassumption(madeforotherpurposes)thatthisallocationisdependentontheentrepreneur-co-ordinator.Wehavetoexplainthebasisonwhich,inpractice,thischoicebetweenalternativesiseffected.14IIOurtaskistoattempttodiscoverwhyafirmemergesatahinaspecializedexchangeeconomy.Thepricemechanism(consideredpurelyfromthesideofthedirectionofresources)mightbesupersedediftherelationshipwhichreplaceditwasdesiredforitsownsake.Thiswouldbethecase,forexample,ifsomepeoplepreferredtoworkunderthedirectionofsomeotherperson.Suchindividualswouldacceptlessinordertoworkundersomeone,andfirmswouldarisenaturallyfromthis.Butitwouldappearthatthiscannotbeaveryimportantreason,foritwouldratherseemthattheoppositetendencyisoperatingifonejudgesfromthestressnormallylaidontheadvantageof“beingone\'sownmaster;”150fcourse,ifthedesirewasnottobecontrolledbuttocontrol,toexercisepoweroverothers,thenpeoplemightbewillingtogiveUpsomethinginordertodirectothers;thatis,theywouldbewillingtopayothersmorethantheycouldgetunderthepricemechanisminorderTheNatureoftheFirm(1937)R.H.COASE4tobeabletodirectthem.Butthisimpliesthatthosewhodirectpayinordertobeabletodothisandarenotpaidtodirect,whichisclearlynottrueinthemajorityofcases.16Firmsmightalsoexistifpurchaserspreferredcommoditieswhichareproducedbyfirmstothosenot50produced;buteveninsphereswhereonewouldexpectsuchpreferences(iftheyexist)tobeofnegligibleimportance,firmsaretobefoundintherealworld.17Thereforetheremustbeotherelementsinvolved.Themainreasonwhyitisprofitabletoestablishafirmwouldseemtobethatthereisacostofusingthepricemechanism.Themostobviouscostof“organizing”productionthroughthepricemechanismisthatofdiscoveringwhattherelevantpricesare.18Thiscostmaybereducedbutitwillnotbeeliminatedbytheemergenceofspecialistswhowillsellthisinformation.Thecostsofnegotiatingandconcludingaseparatecontractforeachexchangetransactionwhichtakesplaceonamarketmustalsobetakenintoaccount.19Again,incertainmarkets,e.g.,produceex-changes,atechniqueisdevisedforminimizingthesecontractcosts;buttheyarenoteliminated.Itistruethatcontractsarenoteliminatedwhenthereisafirmbuttheyaregreatlyreduced.Afactorofproduction(ortheownerthereof)doesnothavetomakeaseriesofcontractswiththefactorswithwhomheisco-operatingwithinthefirm,aswouldbenecessary,ofcourse,ifthisco-operationwereasadirectresultoftheworkingofthepricemechanism.Forthisseriesofcontractsissubstitutedone.Atthisstage,itisimportanttonotethecharacterofthecontractintowhichafactorentersthatisemployedwithinafirm.Thecontractisonewherebythefactor,foracertainremuneration(whichmaybefixedorfluctuating),agreestoobeythedirectionsofanentrepreneurwithincertainlimits.20Theessenceofthecontractisthatitshouldonlystatethelimitstothepowersoftheentrepreneur;Withintheselimits,hecanthereforedirecttheotherfactorsofproduction.Thereare,however,otherdisadvantages-orcosts-ofusingthepricemechanism.Itmaybedesiredtomakealong-termcontractforthesupplyofsomearticleorservice.Thismaybeduetothefactthatifonecontractismadeforalongerperiod,insteadofseveralshorterones,thencertaincostsofmakingeachcontractwillbeavoided.Or,owingtotheriskattitudeofthepeopleconcerned,theymayprefertomakealongratherthanashort-termcontract.Now,owingtothedifficultyofforecasting,thelongertheperiodofthecontractisforthesupplyofthecommodityorservice,thelesspossible,andindeed,thelessdesirableitisforthepersonpurchasingtospecifywhattheothercontractingpartyisexpectedtodo.Itmaywellbeamatterofindifferencetothepersonsupplyingtheserviceorcommoditywhichofseveralcoursesofactionistaken,butnottothepurchaserofthatserviceorcommodity.Butthepurchaserwillnotknowwhichoftheseseveralcourseshewillwantthesuppliertotake.Therefore,theservicewhichisbeingprovidedisexpressedingeneralterms,theexactdetailsbeingleftuntilalaterdate.AllthatisstatedinthecontractistheTheNatureoftheFirm(1937)R.H.COASE5limitstowhatthepersonssupplyingthecommodityorserviceisexpectedtodo.Thedetailsofwhatthesupplierisexpectedtodoisnotstatedinthecontractbutisdecidedlaterbythepurchaser.Whenthedirectionofresources(withinthelimitsofthecontract)becomesdependentonthebuyerinthisway,thatrelationshipwhichIterma"firm"maybeobtained.21Afirmislikelythereforetoemergeinthosecaseswhereaveryshort-termcontractwouldbeunsatisfactory.Itisobviouslyofmoreimportanceinthecaseofservices-labor-thanitisinthecaseofthebuyingofcommodities.Inthecaseofcommodities,themainitemscanbestatedinadvanceandthedetailswhichwillbedecidedlaterwillbeofminorsignificance.WemaysumUpthissectionoftheargumentbysayingthattheoperationofamarketcostssomethingandbyforminganorganizationandallowingsomeauthority(an"entrepreneur")todirecttheresources,certainmarketingcostsaresaved.Theentrepreneurhastocarryouthisfunctionatlesscost,takingintoaccountthefactthathemaygetfactorsofproductionatalowerpricethanthemarkettransactionswhichhesupersedes,becauseitisalwayspossibletoreverttotheopenmarketifhefailstodothis.Thequestionofuncertaintyisonewhichisoftenconsideredtobeveryrelevanttothestudyoftheequilibriumofthefirm.Itseemsimprobablethatafirmwouldemergewithouttheexistenceofuncertainty.Butthose,forinstance,ProfessorKnight,whomakethemodeofpaymentthedistinguishingmarkofthefirm-fixedincomesbeingguaranteedtosomeofthoseengagedinproductionbyapersonwhotakestheresidual,andfluctuating,income-wouldappeartobeintroducingapointwhichisirrelevanttotheproblemweareconsidering.Oneentrepreneurmaysellhisservicestoanotherforacertainsumofmoney,whilethepaymenttohisemployeesmaybemainlyorwhollyashareinprofits.22Thesignificantquestionwouldappeartobewhytheallocationofresourcesisnotdonedirectlybythepricemechanism.AnotherfactorthatshouldbenotedisthatexchangetransactionsonamarketandthesametransactionsorganizedwithinafirmareoftentreateddifferentlybyGovernmentsorotherbodieswithregulatorypowers.Ifweconsidertheoperationofasalestax,itisclearthatitisataxonmarkettransactionsandnotonthesametransactionsorganizedwithinthefirm.Nowsincethesearealternativemethodsoforganization"-bythepricemechanismorbytheentrepreneur-sucharegulationwouldbringintoexistencefirmswhichotherwisewouldhavenoraisond\'être.Itwouldfurnishareasonfortheemergenceofafirminaspecializedexchangeeconomy.0fcourse,totheextentthatfirmsalreadyexist,suchameasureasasalestaxwouldmerelytendtomakethemlargerthantheywouldotherwisebe.Similarly,quotaschemes,andmethodsofpricecontrolwhichimplythatthereisrationing,andwhichdonotapplytofirmsproducingsuchproductsforthemselves,byallowingadvantagestothosewhoorganizewithinthefirmandflotthroughthemarket,necessarilyencouragetheTheNatureoftheFirm(1937)R.H.COASE6growthoffirms.Butitisdifficulttobelievethatitismeasuressuchashavebeenmentionedinthisparagraphwhichhavebroughtfirmsintoexistence.Suchmeasureswould,however,tendtohavethisresultiftheydidnotexistforotherreasons.These,then,arethereasonswhyorganizationssuchasfirmsexistinaspecializedexchangeeconomyinwhichitisgenerallyassumedthatthedistributionofresourcesis"organized"bythepricemechanism.Afirm,therefore,consistsofthesystemofrelationshipswhichcomesintoexistencewhenthedirectionofresourcesisdependentonanentrepreneur;TheapproachwhichhasjustbeensketchedwouldappeartoofferanadvantageinthatitispossibletogiveascientificmeaningtowhatismeantbysayingthatafirmgetslargerorsmallerAfirmbecomeslargerasadditionaltransactions(whichcouldbeexchangetransactionsco-ordinatedthroughthepricemechanism)areorganizedbytheentrepreneurandbecomessmallerasheabandonstheorganizationofsuchtransactions.Thequestionwhicharisesiswhetheritispossibletostudytheforceswhichdeterminethesizeofthefirm.Whydoestheentrepreneurnotorganizeonelesstransactionoronemore?ItisinterestingtonotethatProfessorKnightconsidersthat:therelationbetweenefficiencyandsizeisoneofthemostseriousproblemsoftheory,being,incontrastwiththerelationforaplant,largelyamatterofpersonalityandhistoricalaccidentratherthanofintelligiblegeneralprinciples.Butthequestionispeculiarlyvitalbecausethepossibilityofmonopolygainoffersapowerfulincentivetocontinuousandunlimitedexpansionofthefirm,whichforcemustbeoffsetbysomedecreasedefficiency(intheproductionofmoneyincome)withgrowthinsize,ifevenboundarycompetitionistoexist.23equallypowerfulonemakingforProfessorKnightwouldappeartoconsiderthatitisimpossibletotreatscientificallythedeterminantsofthesizeofthefirm.Onthebasisoftheconceptofthefirmdevelopedabove,thistaskwillnowbeattempted.Itwassuggestedthattheintroductionofthefirmwasdueprimarilytotheexistenceofmarketingcosts.Apertinentquestiontoaskwouldappeartobe(quiteapartfromthemonopolyconsiderationsraisedbyProfessorKnight),why,ifbyorganizingonecaneliminatecertaincostsandinfactreducethecostofproduction,arethereanymarkettransactionsatall?24Whyisnotahproductioncarriedonbyonebigfirm?Therewouldappeartobecertainpossibleexplanations.First,asafirmgetslarger,theremaybedecreasingreturnstotheentrepreneurfunction,thatis,thecostsoforganizingadditionaltransactionswithinthefirmmayrise.25Naturally,apointmustbereachedwherethecostsoforganizinganextratransactionwithinthefirmareequaltothecostsinvolvedincarryingoutthetransactionintheopenmarket,TheNatureoftheFirm(1937)R.H.COASE7or;tothecostsoforganizingbyanotherentrepreneur.Secondly,itmaybethatasthetransactionswhichareorganizedincrease,theentrepreneurfailstoplacethefactorsofproductionintheuseswheretheirvalueisgreatest,thatis,failstomakethebestuseofthefactorsofproduction.Again,apointmustbereachedwherethelossthroughthewasteofresourcesisequaltothemarketingcostsoftheexchangetransactionintheopenmarketortothelossifthetransactionwasorganizedbyanotherentrepreneur.Finally,thesupplypriceofoneormoreofthefactorsofproductionmayrise,becausethe"otheradvantages"ofasmallfirmaregreaterthanthoseofalargefirm.260fcourse,theactualpointwheretheexpansionofthefirmceasesmightbedeterminedbyacombinationofthefactorsmentionedabove.Thefirsttworeasonsgivenmostprobablycorrespondtotheeconomists\'phraseof"diminishingreturnstomanagement."27Thepointhasbeenmadeinthepreviousparagraphthatafirmwilltendtoexpanduntilthecostsoforganizinganextratransactionwithinthefirmbecomeequaltothecostsofcarryingoutthesametransactionbymeansofanexchangeontheopenmarketorthecostsoforganizinginanotherfirm.Butifthefirmstopsitsexpansionatapointbelowthecostsofmarketingintheopenmarketandatapointequaltothecostsoforganizinginanotherfirm,inmostcases(excludingthecaseof"combination"28),thiswillimplythatthereisamarkettransactionbetweenthesetwoprocedures,eachofwhomcouldorganizeitatlessthantheactualmarketingcosts.Howistheparadoxtoberesolved?Ifweconsideranexamplethereasonforthiswillbecomeclear.SupposeAisbuyingaproductfromBandthatbothAandBcouldorganizethismarketingtransactionatlessthanitspresentcost.B,wecanassume,isnotorganizingoneprocessorstageofproduction,butseveral.IfAthereforewishestoavoidamarkettransaction,hewillhavetotakeoveralltheprocessesofproductioncontrolledbyB.UnlessAtakesoverailtheprocessesofproduction,amarkettransactionwillstillremain,althoughitisadifferentproductthatisbought.Butwehavepreviouslyassumedthataseachproducerexpandshebecomeslessefficient;theadditionalcostsoforganizingextratransactionsincrease.ItisprobablethatA\'scostoforganizingthetransactionspreviouslyorganizedbyBwillbegreaterthanB\'scostsofdoingthesamething.AthereforewilltakeoverthewholeofB\'sorganizationonlyifhiscostoforganizingB\'sworkisnotgreaterthanB\'scostbyanamountequaltothecostsofcarryingoutanexchangetransactionontheopenmarket.Butonceitbecomeseconomicaltohaveamarkettransaction,italsopaystodivideproductioninsuchawaythatthecostoforganizinganextratransactionineachfirmisthesame.Uptonowithasbeenassumedthattheexchangetransactionswhichtakeplacethroughthepricemechanismarehomogeneous.Infact,nothingcouldbemorediversethantheactualtransactionswhichtakeplaceinourmodemworld.ThiswouldseemtoimplythatthecostsofcarryingoutexchangetransactionsthroughthepricemechanismwillvaryconsiderablyTheNatureoftheFirm(1937)R.H.COASE8aswillalsothecostsoforganizingthesetransactionswithinthefirm.Itseemsthereforepossiblethatquiteapartfromthequestionofdiminishingreturnsthecostsoforganizingcertaintransactionswithinthefirmmaybegreaterthanthecostsofcarryingouttheexchangetransactionsintheopenmarket.Thiswouldnecessarilyimplythattherewereexchangetransactionscarriedoutthroughthepricemechanism,butwoulditmeanthattherewouldhavetobemorethanonefirm?Clearlynot,forallthoseareasintheeconomicsystemwherethedirectionofresourceswasnotdependentdirectlyonthepricemechanismcouldbeorganizedwithinonefirm.Thefactorswhichwerediscussedearlierwouldseemtobetheimportantones,thoughitisdifficulttosaywhether"diminishingreturnstomanagement"ortherisingsupplypriceoffactorsislikelytobethemoreimportant.Otherthingsbeingequal,therefore,afirmwilltendtobelarger:a.thelessthecostsoforganizingandtheslowerthesecostsrisewithanincreaseinthetransactionsorganized.b.thelesslikelytheentrepreneuristomakemistakesandthesmallertheincreaseinmistakeswithanincreaseinthetransactionsorganized.c.thegreaterthelowering(orthelesstherise)inthesupplypriceoffactorsofproductiontofirmsoflargersize.Apartfromvariationsinthesupplypriceoffactorsofproductiontofirmsofdifferentsizes,itwouldappearthatthecostsoforganizingandthelossesthroughmistakeswillincreasewithanincreaseinthespatialdistributionofthetransactionsorganized,inthedissimilarityofthetransactions,andintheprobabilityofchangesintherelevantprices.29Asmoretransactionsareorganizedbyanentrepreneur,itwouldappçarthatthetransactionswouldtendtobeeitherdifferentinkindorindifferentplaces.Thisfurnishesanadditionalreasonwhyefficiencywilltendtodecreaseasthefirmgetslarger.Inventionswhichtendtobringfactorsofproductionnearertogether,bylesseningspatialdistribution,tendtoincreasethesizeofthefirm.30Changeslikethetelephoneandthetelegraphwhichtendtoreducethecostoforganizingspatiallywilltendtoincreasethesizeofthefirm.Allchangeswhichimprovemanagerialtechniquewilltendtoincreasethesizeofthefirm.31/32Itshouldbenotedthatthedefinitionofafirmwhichwasgivenabovecanbeusedtogivemoreprecisemeaningstotheterms"combination"and"integration."33Thereisacombinationwhentransactionswhichwerepreviouslyorganizedbytwoormoreentrepreneursbecomeorganizedbyone.Thisbecomesintegrationwhenitinvolvestheorganizationoftransactionswhichwerepreviouslycarriedoutbetweentheentrepreneursonamarket.Afirmcanexpandineitherorbothofthesetwoways.Thewholeofthe"structureofcompetitiveindustry"becomestractablebytheordinarytechniqueofeconomicanalysis.TheNatureoftheFirm(1937)R.H.COASE9IIITheproblemwhichhasbeeninvestigatedintheprevioussectionhasnotbeenentirelyneglectedbyeconomistsanditisnownecessarytoconsiderwhythereasonsgivenabovefortheemergenceofafirminaspecializedexchangeeconomyaretobepreferredtotheotherexplanationswhichhavebeenoffered.ItissometimessaidthatthereasonfortheexistenceofafirmistobefoundinthedivisionoflaborThisistheviewofProfessorUsher,aviewwhichhasbeenadoptedandexpandedbyMr.MauriceDobb.Thefirmbecomes"theresultofanincreasingcomplexityofthedivisionoflabour…Thegrowthofthiseconomicdifferentiationcreatestheneedforsomeintegratingforcewithoutwhichdifferentiationwouldcollapseintochaos;anditisastheintegratingforceinadifferentiatedeconomythatindustrialformsarechieflysignificant."34Theanswertothisargumentisanobviousone.The"integratingforceinadifferentiatedeconomy"alreadyexistsintheformofthepricemechanism.Itisperhapsthemainachievementofeconomicsciencethatithasshownthatthereisnoreasontosupposethatspecializationmustleadtochaos.35ThereasongivenbyMr.MauriceDobbisthereforeinadmissible.Whathastobeexplainediswhyoneintegratingforce(theentrepreneur)shouldbesubstitutedforanotherintegratingforce(thepricemechanism).Themostinterestingreasons(andprobablythemostwidelyaccepted)whichhavebeengiventoexplainthisfactarethosetobefoundinProfessorKnight\'sRisk,UncertaintyandProfit.Hisviewswillbeexaminedinsomedetail.TheNatureoftheFirm(1937)R.H.COASE10ProfessorKnightstartswithasysteminwhichthereisnouncertainty:actingasindividualsunderabsolutefreedombutwithoutcollusionmenaresupposedtohaveorganisedeconomiclifewiththeprimaryandsecondarydivisionoflabour,theuseofcapital,etc.,developedtothepointfamiliarinpresent-dayAmerica.Theprincipalfactwhichcallsfortheexerciseoftheimaginationistheinternalorganisationoftheproductivegroupsorestablishments.Withuncertaintyentirelyabsent,everyindividualbeinginpossessionofperfectknowledgeofthesituation,therewouldbenooccasionforanythingofthenatureofresponsiblemanagementorcontrolofproductiveactivity.Evenmarketingtransactionsinanyrealisticsensewouldnotbefound.Theflowofrawmaterialsandproductiveservicestotheconsumerwouldbeentirelyautomatic.36ProfessorKnightsaysthatwecanimaginethisadjustmentasbeing"theresultofalongprocessofexperimentationworkedoutbytrial-and-errormethodsalone,"whileitisnotnecessary"toimagineeveryworkerdoingexactlytherightthingattherighttimeinasortof\'pre-establishedharmony\'withtheworkofothers.Theremightbemanagers,superintendents,etc.,forthepurposeofco-ordinatingtheactivitiesofindividuals,"thoughthesemanagerswouldbeperformingapurelyroutinefunction,"withoutresponsibilityofanysort."37ProfessorKnightthencontinues:Withtheintroductionofuncertainty-thefactofignoranceandthenecessityofactinguponopinionratherthanknowledge-intothisEden-likesituation,itscharacterisentirelychanged....Withuncertaintypresentdoingthings,theactualexecutionofactivity,becomesinarealsenseasecondarypartoflife;theprimaryproblemorfunctionisdecidingwhattodoandhowtodoThisfactofuncertaintybringsaboutthetwomostimportantcharacteristicsofsocialorganization.Inthefirstplace,goodsareproducedforamarket,onthebasisofentirelyimpersonalpredictionofwants,notforthesatisfactionofthewantsoftheproducersthemselves.Theproducertakestheresponsibilityofforecastingtheconsumers\'wants.Inthesecondplace,theworkofforecastingandatthesametimealargepartofthetechnologicaldirectionandcontrolofproductionarestillfurtherconcentrateduponaverynarrowclassoftheproducers,andwemeetwithaneweconomicfunctionary,theentrepreneur....WhenuncertaintyispresentandthetaskofdecidingwhattodoandTheNatureoftheFirm(1937)R.H.COASE11howtodoittakestheascendancyoverthatofexecutiontheinternalorganisationoftheproductivegroupsisnolongeramatterofindifferenceoramechanicaldetail.Centralisationofthisdecidingandcontrollingfunctionisimperative,aprocessof"cephalisation"isinevitable.39Themostfundamentalchangeis:thesystemunderwhichtheconfidentandventuresomeassumetheriskorinsurethedoubtfulandtimidbyguaranteeingtothelatteraspecifiedincomeinreturnforanassignmentoftheactualresults…Withhumannatureasweknowititwouldbeimpracticableorveryunusualforonemantoguaranteetoanotheradefiniteresultofthelatter\'sactionswithoutbeinggivenpowertodirecthiswork.Andontheotherhandthesecondpartywouldnotplacehimselfunderthedirectionofthefirstwithoutsuchaguarantee…Theresultofthismanifoldspecialisationoffunctionistheenterpriseandwagesystemofindustry.Itsexistenceintheworldisthedirectresultofthefactofuncertainty.40ThesequotationsgivetheessenceofProfessorKnight\'stheory.Thefactofuncertaintymeansthatpeoplehavetoforecastfuturewants.Therefore,yougetaspecialclassspringingUpwhodirecttheactivitiesofotherstowhomtheygiveguaranteedwages.Itactsbecausegoodjudgmentisgenerallyassociatedwithconfidenceinone\'sjudgment.41ProfessorKnightwouldappeartoleavehimselfopentocriticismonseveralgrounds.Firstofall,ashehimselfpointsout,thefactthatcertainpeoplehavebetterjudgmentorbetterknowledgedoesnotmeanthattheycanonlygetanincomefromitbythemselvesactivelytakingpartinproduction.Theycanselladviceorknowledge.Everybusinessbuystheservicesofahostofadvisers.Wecanimagineasystemwhereahadviceorknowledgewasboughtasrequired.Again,itispossibletogetarewardfrombetterknowledgeorjudgmentnotbyactivelytakingpartinproductionbutbymakingcontractswithpeoplewhoareproducing.Amerchantbuyingforfuturedeliveryrepresentsanexampleofthis.Butthismerelyillustratesthepointthatitisquitepossibletogiveaguaranteedrewardprovidingthatcertainactsareperformedwithoutdirectingtheperformanceofthoseacts.ProfessorKnightsaysthat"withhumannatureasweknowit,itwouldbeimpracticableorveryunusualforonemantoguaranteetoanotheradefiniteresultofthelatter\'sactionswithoutbeinggivenpowertodirecthiswork."Thisissurelyincorrect.Alargeproportionofjobsaredonetocontract,thatis,thecontractorisguaranteedacertainsumprovidingheperformscertainacts.Butthisdoesnotinvolveanydirection.Itdoesmean,however,thatthesystemofrelativepriceshasbeenchangedandthattherewillbeanewarrangementofthefactorsofproduction.42ThefactthatProfessorKnightmentionsthatthe"secondTheNatureoftheFirm(1937)R.H.COASE12partywouldnotplacehimselfunderthedirectionofthefirstwithoutsuchaguarantee"isirrelevanttotheproblemweareconsidering.Finally,itseemsimportanttonoticethatevenin.thecaseofaneconomicsystemwherethereisnouncertaintyProfessorKnightconsidersthattherewouldbeco-ordinators,thoughtheywouldperformonlyaroutinefunction.Heimmediatelyaddsthattheywouldbe"withoutresponsibilityofanysort,"whichraisesthequestionbywhomaretheypaidandwhy?ItseemsthatnowheredoesProfessorKnightgiveareasonwhythepricemechanismshouldbesuperseded.IVItwouldseemimportanttoexamineonefurtherpointandthatistoconsidertherelevanceofthisdiscussiontothegeneralquestionofthe"cost-curveofthefirm."Ithassometimesbeenassumedthatafirmislimitedinsizeunderperfectcompetitionifitscostcurveslopesupward,43whileunderimperfectcompetition,itislimitedinsizebecauseitwillnotpaytoproducemorethantheoutputatwhichmarginalcostisequaltomarginalrevenue.«Butitisclearthatafirmmayproducemorethanoneproductand,therefore,thereappearstobenoprimafaciereasonwhythisupwardslopeofthecostcurveinthecaseofperfectcompetitionorthefactthatmarginalcostwillnotalwaysbebelowmarginalrevenueinthecaseofimperfectcompetitionshouldlimitthesizeofthefirm.45Mrs.Robinson46makesthesimplifyingassumptionthatonlyoneproductisbeingproduced.Butitisclearlyimportanttoinvestigatehowthenumberofproductsproducedbyafirmisdetermined,whilenotheorywhichassumesthatonlyoneproductisinfactproducedcanhaveverygreatpracticalsignificance.Itmightberepliedthatunderperfectcompetition,sinceeverythingthatisproducedcanbesoldattheprevailingprice,thenthereisnoneedforanyotherproducttobeproduced.Butthisargumentignoresthefactthattheremaybeapointwhereitislesscostlytoorganizetheexchangetransactionsofanewproductthantoorganizefurtherexchangetransactionsoftheoldproduct.Thispointcanbeillustratedinthefollowingway.Imagine,followingvonThunen,thatthereisatown,theconsumingcenter,andthatindustriesarelocatedaroundthiscentralpointinrings.TheseconditionsareillustratedinthefollowingdiagraminwhichA,B,andCrepresentdifferentindustries.TheNatureoftheFirm(1937)R.H.COASE13ACBXImagineanentrepreneurwhostartscontrollingexchangetransactionsfromx.Nowasheextendshisactivitiesinthesameproduct(B),thecostoforganizingincreasesuntilatsomepointitbecomesequaltothatofadissimilarproductwhichisnearer.Asthefirmexpands,itwillthereforefromthispointincludemorethanoneproduct(AandC).Thistreatmentoftheproblemisobviouslyincomplete,47butitisnecessarytoshowthatmerelyprovingthatthecostcurveturnsupwardsdoesnotgivealimitationtothesizeofthefirm.Sofarwehaveonlyconsideredthecaseofperfectcompetition;thecaseofimperfectcompetitionwouldappeartobeobvious.Todeterminethesizeofthefirm,wehavetoconsiderthemarketingcosts(thatis,thecostsofusingthepricemechanism),andthecostsoforganizingthedifferententrepreneursandthenwecandeterminehowmanyproductswillbeproducedbyeachfirmandhowmuchofeachitwillproduce.Itwould,therefore,appearthatMr.Shove48inhisarticleon"ImperfectCompetition"wasaskingquestionswhichMrs.Robinson\'scostcurveapparatuscannotanswer;Thefactorsmentionedabovewouldseemtobetherelevantones.VOnlyonetasknowremains;andthatis,toseewhethertheconceptofafirmwhichhasbeendevelopedfitsinwiththatexistingintherealworld.Wecanbestapproachthequestionofwhatconstitutesafirminpracticebyconsideringthelegalrelationshipnormallycalledthatof"masterandservant"or"employerandemployee."49Theessentialsofthisrelationshiphavebeengivenasfollows:(1)theservantmustbeunderthedutyofrenderingpersonalservicestothemasterortoothersonbehalfofthemaster,otherwisethecontractisacon-tractforsaleofTheNatureoftheFirm(1937)R.H.COASE14goodsorthelike.(2)Themastermusthavetherighttocontroltheservant\'swork,eitherpersonallyorbyanotherservantoragent.Itisthisrightofcontrolorinterference,ofbeingentitledtotelltheservantwhentowork(withinthehoursofservice)andwhennottowork,andwhatworktodoandhowtodoit(withinthetermsofsuchservice)whichisthedominantcharacteristicinthisrelationandmarksofftheservantfromanindependentcontractor,orfromoneemployedmerelytogivetohisemployerthefruitsofhislabour.Inthelattercase,thecontractororperformerisnotundertheemployer\'scontrolindoingtheworkoreffectingtheservice;hehastoshapeandmanagehiswork50astogivetheresulthehascontractedtoeffect.50Wethusseethatitisthefactofdirectionwhichistheessenceofthelegalconceptof"employerandemployee,"justasitwasintheeconomicconceptwhichwasdevelopedabove.ItisinterestingtonotethatProfessorBattsaysfurther:Thatwhichdistinguishesanagentfromaservantisnottheabsenceorpresenceofafixedwageorthepaymentonlyofcommissiononbusinessdone,butratherthefreedomwithwhichanagentmaycarryouthisemployment.51Wecanthereforeconcludethatthedefinitionwehavegivenisonewhichapproximatescloselytothefirmasitisconsideredintherealworld.Ourdefinitionis,therefore,realistic.Isitmanageable?Thisoughttobeclear;Whenweareconsideringhowlargeafirmwillbetheprincipleofmarginalismworkssmoothly.Thequestionalwaysis,willitpaytobringanextraexchangetransactionundertheorganizingauthority?Atthemargin,thecostsoforganizingwithinthefirmwillbeequaleithertothecostsoforganizinginanotherfirmortothecostsinvolvedinleavingthetransactiontobe"organized"bythepricemechanism.Businessmenwillbeconstantlyexperimenting,controllingmoreorless,andinthisway,equilibriumwillbemaintained.Thisgivesthepositionofequilibriumforstaticanalysis.Butitisclearthatthedynamicfactorsarealsoofconsiderableimportance,andaninvestigationoftheeffectchangeshaveonthecostoforganizingwithinthefirmandonmarketingcostsgenerallywillenableonetoexplainwhyfirmsgetlargerandsmaller;Wethushaveatheoryofmovingequilibrium.Theaboveanalysiswouldalsoappeartohaveclarifiedtherelationshipbetweeninitiativeorenterpriseandmanagement.Initiativemeansforecastingandoperatesthroughthepricemechanismbythemakingofnewcontracts.Managementpropermerelyreactstopricechanges,rearrangingthefactorsofproductionunderitscontrol.ThatthebusinessmannormallycombinesbothfunctionsisanobviousresultofthemarketingcostswhichwerediscussedTheNatureoftheFirm(1937)R.H.COASE15above.Finally,thisanalysisenablesustostatemoreexactlywhatismeantbythe"marginalproduct"oftheentrepreneur.Butanelaborationofthispointwouldtakeusfarfromourcomparativelysimpletaskofdefinitionandclarification.NOTES1.JoanRobinson,EconomicsIsaSeriousSubject(1932),12.2.SeeN.Kaldor,"TheEquilibriumoftheFirm,"44TheEconomicJournal(1934),60-76.3.Op.cit.,6.4.JM.Keynes,EssaysinBiography(1933),223-24.5.L.Robbins,NatureandSignificanceofEconomicScience(1935),63.6.ThisdescriptionisquotedwithapprovalbyD.H.Robenson,Control0fIndustry(1923),85,andbyProfessorArnoldPlant,"TrendsinBusinessAdministration,"12Economica(1932)45-62.ItappearsinAlliedShippingControl,pp.16-17.7.SeeF.A.Hayek,"TheTrendofEconomicThinking,"13Economica(1933)121-37.8.SeeRA.Hayek,op.cit.9.Op.cit.,85.10.IntherestofthispaperIshallusethetermentrepreneurtorefertothepersonorpersonswho,inacompetitivesystem,taketheplaceofthepricemechanisminthedirectionofresources.11.SurveyofTextileIndustries,26.12.Op.cit.,71.13.CapitalistEnterpriseandSocialProgress(1925),20.Cf.,also,Henderson,SupplyandDemand(1932),3-5.14.ItiseasytoseewhentheStatetakesoverthedirectionofanindustrythat,inplanningit,itisdoingsomethingwhichwaspreviouslydonebythepricemechanism.Whatisusuallynotrealizedisthatanybusinessmaninorganizingtherelationsbetweenhisdepartmentsisalsodoingsomethingwhichcouldbeorganizedthroughthepricemechanism.ThereisthereforepointinMr.Durbin\'sanswertothosewhoemphasizetheproblemsinvolvedineconomicplanningthatthesameproblemshavetobesolvedbybusinessmeninthecompetitivesystem.(Sec"EconomicCalculusinaPlannedEconomy,"46TheEconomicJournal[1936]676-90.>Theimportantdifferencebetweenthesetwocasesisthateconomicplanningisimposedonindustrywhilefirmsarisevoluntarilybecausetheyrepresentamoreefficientmethodoforganizingproduction.Inacompetitivesystem,thereisan"optimum"amountofplanning!15.Cf.HarryDawes,"LabourMobilityintheSteelIndustry,"44TheEconomicJournal(1934)84-94,whoinstances"thetrektoretailshopkeepingandinsuranceworkbythebetterpaidofskilledmenduetothedesire(oftenthemainaiminlifeofaworker>tobeindependent"(86).16.Nonetheless,thisisnotaltogetherfanciful.Somesmallshopkeepersaresaidtoearnlessthantheirassistants.17.G.F.Shove,,"TheImperfectionoftheMarket:aFurtherNote,"44TheEconomicJournal(1933>113-24,n.1,pointsoutthatsuchpreferencesmayexist,althoughtheexamplehegivesisalmostthereverseoftheinstancegiveninthetext.18.AccordingtoN.Kaldor,"AClassificatoryNoteofthcDeterminantenessofEquilibrium,"ITheReview0fEconomicStudies(1934)122-36,hisoneoftheassumptionsofstatictheorythat"Ailtherelevantpricesareknowntoailindividuals."Butthisisclearlynottrueoftherealworld.19.ThisinfluencewasnotedbyProfessorUsherwhendiscussingthedevelopmentofcapitalism.Hesays:"Thesuccessivebuyingandsellingofpartlyfinishedproductsweresheerwasteofenergy."(IntroductiontotheIndustrialHistory0fEngland(1920),13.)Buthedoesnotdeveloptheideanorconsiderwhyitisthatbuyingandsellingoperationsstillexist.20.Itwouldbepossiblefornolimitstothepowersofthcentrepreneurtobefixed.ThisTheNatureoftheFirm(1937)R.H.COASE16wouldbevoluntaryslavery.AccordingtoProfessorBatt,TheLawo!MasterandServant(1933),18,suchacontractwouldbevoidandunenforceable.21.0fcourse,itisnotpossibletodrawahardandfastlinewhichdetermineswhetherthereisafirmornot.Theremaybemoreorlessdirection.Itissimilartothelegalquestionofwhetherthereistherelationshipofmasterandservantorprincipalandagent.Secthediscussionofthisproblembelow.22.TheviewsofProfessorKnightareexaminedbelowinmoredetail.23.Risk,UncertaintyandProfit,PrefacetotheRe-issue,LondonSchoolofEconomicsSeriesofReprints,No.16(1933).24.Therearecertainmarketingcostswhichcouldonlybeeliminatedbytheabolitionof"consumers\'choice"andthesearcthecostsofretailing.Itisconceivablethatthesecostsmightbesohighthatpeoplewouldbcwillingtoacceptrationsbecausetheextraproductobtainedwasworththelossoftheirchoice.25.Thisargumentassumesthatexchangetransactionsonamarketcanbeconsideredashomogeneous;whichisclearlyuntrueinfact.Thiscomplicationistakenintoaccountbelow.26.Foradiscussionofthevariationofthesupplypriceoffactorsandproductiontofirmsofvaryingsize,secE.A.G.Robinson,TheStructureofCompetitiveIndustry(1932).Itissometimessaidthatthesupplypriceoforganizingabilityincreasesasthesizeofthefirmincreasesbecausemenprefertobetheheadsofsmallindependentbusinessesratherthantheheadsofdepartmentsinalargebusiness.SecJones,TheTrustProblem(1921),531,andMacgregor,IndustrialCombination(1935),63.ThisisacommonargumentofthosewhoadvocateRationalization.Itissaidthatlargerunitswouldbemoreefficient,butowingtotheindividualisticspiritofthesmallerentrepreneurs,theyprefertoremainindependent,apparentlyinspiteofthehigherincomewhichtheirincreasedefficiencyunderRationalizationmakespossible.27.Thisdiscussionis,ofcourse,briefandincomplete.Foramorethoroughdiscussionofthisparticularproblem,seeN.Kaldor,"TheEquilibriumoftheFirm,"44TheEconomicJournal(1934)60-76,andE.A.G.Robinson,"TheProblemofManagementandtheSizeoftheFirm,"44TheEconomicJournal(1934)242-57.28.Adefinitionofthistermisgivenbelow.29.ThisaspectoftheproblemisemphasizedbyN.Kaldor,op.cit.ItsimportanceinthisconnectionhadbeenpreviouslynotedbyE.A.G.Robinson,TheStructure0fCompetitiveIndustry(1932),83-106.Thisassumesthatanincreaseintheprobabilityofpricemovementsincreasesthecostsoforganizingwithinafirmmorethanitincreasesthecostofcarryingoutanexchangetransactiononthemarket-whichisprobable.30.ThiswouldappeartobetheimportanceofthetreatmentofthetechnicalunitbyE.A.G.Robinson,op.cit.,27-33.Thelargerthetechnicalunit,thegreatertheconcentrationoffactorsandthereforethefirmislikelytobelarger.31.Itshouldbenotedthatmostinventionswillchangeboththecostsoforganizingandthecostsofusingthepricemechanism.Insuchcases,whethertheinventiontendstomakefirmslargerorsmallerwilldependontherelativeeffectonthesetwosetsofcosts.Forinstance,ifthetelephonereducesthecostsofusingthepricemechanismmorethanitreducesthecostsoforganizing,thenhwillhavetheeffectofreducingthesizeofthefirm.32.AnillustrationofthesedynamicforcesisfurnishedbyMauriceDobb,RussianEconomicDevelopment(1928),68."Withthepassingofbondedlabourthefactory,asanestablishmentwhereworkwasorganisedunderthewhipoftheoverseer,lostitsraisond\'êtreuntilthiswasrestoredtoitwiththeintroductionofpowermachineryafter1846."Itseemsimportanttorealizethatthepassagefromthedomesticsystemtothefactorysystemisnotamerehistoricalaccident,butisconditionedbyeconomicforces.Thisisshownbythefactthatitispossibletomovefromthefactorysystemtothedomesticsystem,asintheRussianexample,aswellasviceversa.Itistheessenceofserfdomthatthepricemechanismisnotallowedtooperate.Therefore,therehastobedirectionfromsomeorganizer.When,however,serfdompassed,thepricemechanismwasallowedtooperate.Itwasnotuntilmachinerydrewworkersintoonelocalitythatitpaidtosupersedethepricemechanismandthefirmagainemerged.TheNatureoftheFirm(1937)R.H.COASE1733.Thisisoftencalled"verticalintegration,"combinationbeingtermed"lateralintegration."34.Op.cit.,10.ProfessorUsher\'sviewsaretobefoundinhisIntroductiontotheIndustrialHistoryofEngland(1920),1-18.35.Cf.J.B.Clark,DistributionofWealth(1899),19,whospeaksofthetheoryofexchangeasbeingthe"theoryoftheorganisationofindustrialsociety."36.Risk,UncertaintyandProfit,267.37.Op.cit.,267-68.38.Op.cit.,268.39.Op.cit.,268-95.40.Op.cit.,269-70.41.Op.cit.,270.42.Thisshowsthatitispossibletohaveaprivateenterprisesystemwithouttheexistenceoffirms.Though,inpractice,thetwofunctionsofenterprise,whichactuallyinfluencesthesystemofrelativepricesbyforecastingwantsandactinginaccordancewithsuchforecasts,andmanagement,whichacceptsthesystemofrelativepricesasbeinggiven,arenormallycarriedoutbythesamepersons,yetitseemsimportanttokeepthemseparateintheory.Thispointisfurtherdiscussedbelow.43.SeeKaidor,op.cit.,andRobinson,TheProblemofManagementandtheSizeoftheFirm.44.Mr.RobinsoncallsthistheImperfectCompetitionsolutionforthesurvivalofthesmallfirm.45.Mr.Robinson\'sconclusion,op.cit.,249,n.1,wouldappeartobedefinitelywrong.HeisfollowedbyHoraceJ.White,Jr,"MonopolisticandPerfectCompetition,"26TheAmericanEconomicReview(1936)645,n.27.Mr.Whitestates"Itisobviousthatthesizeofthefirmislimitedinconditionsofmonopolisticcompetition."46.EconomicsImperfectCompetition(1934).47.Ashasbeenshownabove,locationisonlyoneofthefactorsinfluencingthecostoforganizing.48.G.F.Shove,"TheImperfectionoftheMarket,"43TheEconomicJournal(1933).115.Inconnectionwithanincreaseindemandinthesuburbsandtheeffectonthepricechargedbysuppliers,Mr.Shoveasks“…whydonottheoldfirmsopenbranchesinthesuburbs?"Iftheargumentinthetextiscorrect,thisisaquestionwhichMrs.Robinson\'sapparatuscannotanswer.49.Thelegalconceptof"employerandemployee"andtheeconomicconceptofafirmarenotidentical,inthatthefirmmayimplycontroloveranotherperson\'spropertyaswellasovertheirlaborButtheidentityofthesetwoconceptsissufficientlycloseforanexaminationofthelegalconcepttobeofvalueinappraisingtheworthoftheeconomicconcept.50.Batt,TheLawofMasterandServant,6.51.Op.cit.,7.',)