EarlO.HeadyandEarlW.Kehrbergconductedasurvey（Iowa,1949）withseventy-fourselectedpairsofcash-leaseandshare-leasefarms.Theyfound"nosignificantdifferencesbetweenshareandcashleases"（p.661）fortheinputintensitiesoffarming.Theauthors,however,refusedtoaccepttheconflictbetweentheirtheoryandfindings,andresortedtootherfactorswhich"mightaccountforthislackofdifference"（pp.661-62）.Withthetaxapproach,theyhadconcludedthat,intheory,farmingintensityundersharerentisnecessarilylowerthanundercashrent（atpp.658-60）.SeeHeadyandKehrberg,"RelationshipofCropShareandCashLeasingSystemstoFarmingEfficiency,"ResearchBulletin386（IowaStateCollegeAgriculturalExperimentStation,May,1952）,pp.635-83.
[16].Buck,ChineseFarmEconomy,p.156.Nonumericaldataareprovided.
[17].SeealsoLegislativeYuan,StatisticalMonthly（1930）,2,5.
[18].Thisisshownearlierinthissection.ForsomeindirectevidencefromTaiwan,seechapters7and8.
[19].orgputedfromdatainSino-AmericanJointCommissiononRuralRe-construction"JCRRAnnualReportsonLandReformintheRepublicofChina,"mimeographed（1965）,p.35.
[20].Onlyasmallamountofnumericaldatahasbeenprovided,andexamplesoflandownersholdingoveronethousandleasesareperhapscasesselectedtoemphasizetheconcentrationoflandownership.AccordingtoinformationprovidedinJ.P.Gittinger,"VietnameseLandTransferProgram,"LandEconomics（May,1957）,"beforethelandtransferprograminVietnam,2,170persons……havedeclaredtotalholdingsamountingto976,602hectares."Giventhat"theaveragetenantholdingisapproximately2hectaresorperhapsslightlymore,"theaveragenumberofleasesforeachoftheselandownerswasovertwohundred.
4.TransactionCosts,RiskAversion,andtheChoiceofContractualArrangements
Theanalysisthusfarhasbeenprimarilybasedontheconditionthattransactioncosts,andinparticularthecostsofcontractualnegotiationandenforcement,arezero.Thetheoryofsharetenancythereuponderivedshowsthat,subjecttotheconstraintofprivatepropertyrights,economicefficiencyisthesameundervariouslandtenurearrangements.Althoughtransactioncostsexistintherealworld,thetheorysucceedsinexplainingaclassofobservations.Butthepresenceofavarietyofcontractualarrangementsunderthesameconstraintofpetitionposesthequestion:Whyaredifferentcontractualarrangementschosenunderthesamesystemofprivatepropertyrights?Thepurposeofthischapteristoadvance,inaratherinformalmanner,achoice-theoreticapproachbasedonnonzerotransactioncostandriskaversiontoexplaintheobservedcontractualbehaviorinagriculture.TheobservationsusedarelargelydrawnfromtheChineseexperience.
Ifafirmcanincreaseefficiencyinproductionbyemployingproductiveresourcesofmorethanoneresourceowner,acontracttobinetheresourceswillobtain.Theformationofthecontractinvolvespartialtransfersofpropertyrightsinoneformoranother,suchasleasing,hiring,ormortgaging.[1]Thesetransfers,andtheassociatedcoordinationofinputsofvariousfactorsinproduction,arecostlyevents.[2]Therearecostsofnegotiatingandofenforcingthestipulationsofthecontract.
Giventhestateofpersonalwealthdistributionandtheportfoliosofassetsheldasprivatepropertybyresourceowners,someownerswillseekcontractualarrangementswithothersinbiningresourcesforproduction.[3]Thereisavarietyofarrangementsunderwhichthiscanbedone.Atleasttworeasonsmaybeofferedfortheexistenceofdifferenttypesofcontractualarrangements.Firstistheexistenceofnaturalrisk,definedhereasthecontributionbynatureorthestateoftheworldtothevariance（orstandarddeviation）oftheproductvalue.[4]Givenanonzerovariancefortheexpectedoutputyield（thetotalineforthecontractingparties）,differentcontractualarrangementsallowdifferentdistributionsofinevariancesamongthecontractingparties.Underthepostulateofriskaversion,anindividualwillseektoavoidriskifthecostofdoingsoislessthanthegainfromtheriskaverted.Hemayavertriskbysearchingforinformationaboutthefuture（whichmaynotbeattainableevenatinfinitelyhighcost）,bychoosinglessriskyoptionswheninvesting（whichoptionsincludeportfoliodiversification）,orbychoosingamongarrangementswithwhichhisburdenofriskcanbedispersedtootherindividuals-suchasinsuranceandvariouscontractualarrangements.Thelastisoneofourconcernsinthischapter.Asecondreasonfortheexistenceofdifferentcontractualarrangementsisthedifferenttransactioncoststhatareassociatedwitheach.Transactioncostsdifferbecausethephysicalattributesofinputandoutputdiffer,becauseinstitutionalarrangementsdiffer,andbecausedifferentsetsofstipulationsrequirevaryingeffortsinenforcementandnegotiation.[5]
Letmeadvancethefollowinghypothesis:thechoiceofcontractualarrangementismadesoastomaximizethegainfromriskdispersionsubjecttotheconstraintoftransactioncosts.Ishalldevelopthishypothesisandapplyittosomeobservationslater.