第33章 《佃農理論》英語原著 (27)

Anycontractbiningresourcesfromdifferentownersforproductioninvolves,inadditiontonegotiationcosts,theenforcementcostsofcontrollinginputsanddistributingoutput,accordingtothetermsofthecontract.Contractingonasharebasisappearstoinvolvehighertransactioncostsasawhole（thesumofnegotiationandenforcementcosts）thanafixed-rentorawagecontract.Thetermsinasharecontract,amongotherthings,includetherentalpercentage,theratioofnonlandinputtoland,andthetypesofcropstobegrown.[11]Thesearemutuallydecidedbythelandownerandthetenant.Forfixed-rentandwagecontracts,however,giventhemarketprices,onepartyalonecandecidehowmuchoftheotherparty'sresourcesheshallemployandwhatcropsshallbegrown.Andsinceinasharecontractthesharingofoutputisbasedontheactualyield,effortsmustbemadebythelandownertoascertaintheharvestyield.Thusnegotiationandenforcementaremoreplexforasharecontractthanforafixed-rentorawagecontract.

Therankingoftransactioncostsoffixed-rentandwagecontractsappearsuncertain.Thephysicalattributesoflandaresuchthatthecostofenforcingthecontractedamountofinputislowerthanwithlabor.Thatis,the"shirking"oflaborinput,whichmayexistinawagecontract（alsoinasharecontract）withouteitherenforcingtheinputorcheckingtheoutput,iscostlytoprevent.[12]Butwhilethis"shirking"problemdoesnotappearsignificantforlandinputinafixed-rentcontract,policing（orenforcing）themaintenanceofsoilandotherassetsownedbythelandlordismorecostlyforafixed-rentorasharecontractthanforawagecontract.[13]Ifweaccepttheabovereasoning,pendingempiricalconfirmation,andiftransactioncostistheonlyconsideration,thentheminimizationoftransactioncostimpliesthatsharecontractswillneverbechosen.Why,then,aresharecontractschosen?

Supposethetransactioncostiszeroorthesameforallformsofcontract.Letusemployabehavioralpostulateofriskaversion,definedheretomeanthatanindividual,giventhesameexpectedaverageine,prefersalowertoahighervariance.Inagriculture,variablesexogenoustotheproductionfunction,suchasweatherconditionsandpests,areriskfactorswhicharedifficulttoforecastandwhichmaysignificantlyaffectthevarianceofthevalueofoutput.Underafixed-rentcontract,thetenantbearsmost,ifnotall,oftherisk;underawagecontract,thelandownerbearsmost,ifnotall,oftherisk.Sharetenancymaythenberegardedasadeviceforrisksharing（orriskdispersion）；thatis,thevarianceoftheoutputyieldisdistributedamongthecontractingparties.Giventhepostulateofriskaversion,asharecontractwillbemutuallypreferredbythelandownerandthetenant.[14]Risk,however,existsinvaryingdegreesinanytenancy.Why,then,arefixed-rentandwagecontractschosenatall?

Isuggestthatthechoiceofcontractbeanalyzedbyemployingboththedilferencesintransactioncostsandthepostulateofriskaversion.Giventhestateofriskassociatedwithaparticularoutput,ahighertransactioncostwillleadtolowerreturnstotheproductiveassets.Ontheotherhand,giventhetransactioncost,riskaversionimpliesthatassetvaluesandthevariancesofinearenegativelyrelated.[15]Whileinitselfthedispersionofriskunderasharecontractwillleadtohighervaluesforthecon-tractedresources,thehigherassociatedtransactioncostwillleadtolowerassetvalues.Wealthmaximization（orutilitymaximization,dependingontherelevantmeasurementproblem）im-pliesthatthecontractualarrangementchosenwillbetheonewhichyieldsthehighestvaluesforthecontractedresources.

Giventhevarianceofoutputvalueandtherentalpercentage,asharecontractprescribesaspecificdistributionofinevariancesforthecontractingparties.Theassociatedstateofriskdis-persionmaynotconformtothemostpreferablestateaccordingtotheparties'preferencefunctions.Sincesomedispersionofriskispreferredtonodispersionatall,however,asharecontractwillbechosenratherthanafixedrentorawagecontractifthehighertransactioncostisatleastpensatedforbythegainfromriskdispersion.Thereexist,ofcourse,stillotherarrangementsunderwhichthedispersionofriskcanbetailoredtofiteachcase.ButasIshalldiscussinthenextsection,thetransactioncostofanarrangementforriskdispersionmoreflexiblethanasharecontractmaybesohighastomakeitundesirable.

Evidenceisavailabletosupporttheapplicabilityofthiskindofanalysis:

1.Sincetransactioncostisassertedtobehigherforsharethanforfixedrents,therewouldberoomforathirdpartytoinsuretheamountofcropyield.Thatis,ifathirdparty（aninsurancepany）weretoinsuretheexpectedmeanyield,thecontractingpartieswouldchooseafixed-rentcontractandwouldbewillingtopaytheinsuringpartyanamountnohigherthanthesavingfromalowertransactioncostplusapremiumforthealmostcertaininenowobtainedasparedwiththevariableineinasharecontract.Yetweseldomfindsuchacropinsurancewithoutgovernment'stakinganactiverole.Thereason,perhaps,isthatthecostofhandlinginsurancetransactionsmaybesohighastobeprohibitive:theinsuringagentwouldhavetochecknotonlytheactualcropyieldsbutalsotheamountofnonlandinputs.FortheFrenchmetayage（sharecropping）,however,ConstantiaMaxwellobserved:

TheusualprocedureforFrenchseigneurswas,whileretainingthechateauanditsimmediateneighborhoodfortheirownuse,toletouttheirlandsingrostomiddlemenorfermiers（tobedistinguishedfromfermiersexploitants）,whopaidafixedsumtotheproprietorandgatheredtherentsfrommetayageorcensitairesattheirownriskforapersonalprofit.Someofthesemiddlemen,likethelandlords,wereabsenteesandworkedtheestatethroughsub-agents.[16]