Whenthecultivatorhastogivetohislandlordhalfofthereturnstoeachdoseofcapitalandlaborthatheappliestotheland,itwillnotbetohisinteresttoapp,then,heisfreetocultivateashechooses,hewillcultivatefarlessintensivelythanontheEnglishplan[fixedrent]；hewillapplyonlysomuchcapitalandlaboraswillgivehimreturnsmorethantwiceenoughtorepayhimself:sothathislandlordwillgetasmallershareevenofthosereturnsthanhewouldhaveontheplanofafixedpayment.[4]

WhatisimportanthereisthatMarshallsawthataccordingtothisanalysisthesharetenantwillbegettingaresidualreturhooseinsteadafixedrentcontractorsellinghislandownershiptothetenantoutright.[5]AndwhatisnaturalisthatMarshallwasreluctanttoletthetenantsresidualearningremainunexplained:

Ifthetenanthasnofixityoftenure,thelandlordcandeliberatelyandfreelyarrangetheamountofcapitalandlaborsuppliedbythetenantandtheamountofcapitalsuppliedbyhimselftosuittheexigenciesofeachspecialcase.[6]

Marshallsviewwasthatthemetayertenant”haspracticalfixityoftenure,”[7]andhereferredtoanarticlebyHenryHiggs.[8]Higgsnotedthattherentalsharesdodiffer,andthat”rigidasmetayermayatfirstsightseemtobe,itissusceptibleofconsiderableelasticity.”[9]Henonethelesssharedthe”custom”,whichunfortunatelyinvolvedasampleofonlyasinglefarm.[10]Itisthenotionof”fixityoftenure,”perhaps,thatledMarshalltomentionthepossibilityofadjustmentsonlyinafootnote:

Ifthelandlordcontrolstheamount[ofcapital]freelyandinhisowninterest,andcanbargainwithhistenantastotheamountoflaborheapplies,itcanbeprovedgeometricallythathewillsoadjustitastoforcethetenanttocultivatethelandjustasintensivelyashewouldundertheEnglishtenure[fixedrent]；andhissharewillthenbethesameasunderit.[11]

Marshallprovidednogeometricproof,anditisaninterectonlyincertainspecialcases,vary.[12],supposethehorizontalaxisoffigure4measuresnotonlylaborbutnonlandinputsingeneral（andassumeawaytheassociatedsubstitutionproblem）.（notany）rentalpercentage,andgiventhatthelandlordhasbeenprovidingasufficientlylargeamountofnonland”capital”input,thenhemayadjustthenonlandinputsprovidedbyeithercontractingpartysothat（a）therentalineundersharerentisthesameasunderfixedrentand（b）thetenantsre,,orthatthelandlordhasbeenprovid-ingapartofthenonlandinputsbuttheamountistoosmall;then,exceptbyaccident,therentalpercentagemustbevaried（inadditiontoadjustingnonlandinputsoverland）:giventheratioofnonlandinputstolandthatisconsistentwithwealthmaximization,andgiventhattherelativesharesofnonlandinputscontributedbythecontractingpartiesarestipulated,thereexistsoneandonlyonerentalpercentagewhichisconsistentwithequilibrium.[13]Thisdifficultparagraphwillbeeclearerwhenweetothenextsection.

ItshouldbenotedthatMarshall,likeSmith,Jones,andMillbeforehim,attemptedto”rank”variouslandtenurearrangementsaccordingtosomenotionofeconomicefficiency.[14]Theydidnottackletheissuebyidentifyingaspecificsetofpropertyrightconstraintssubjecttowhic,thefreelyalienablerightsimplicitintheiranalysessuggestthattheconstraintofprivatepropertyrightswasassumed.[15]ButwhereasSmithandJonesviewedasharelease,thoughwasteful,”custom.”Smithspredictionthatfixedrentswouldreplacesharerentshadfailedtoetrue;and,asnotedearlier,thetermsofasharecontractmightyieldtheimpressionofbeingcustomarilyfixed.

Amongcontemporarywriterswhoperformedsimilar”rank-ings”andalsoreliedonthenotionof”custom”areRainerSchic-kele[16][17]SchickeleandHeadyfurtheredthetaxapproachintheiranalysesofasharelease,,theyconcludedthatequilibriumisatpoints,wherethemarginaltenantcostequalsthemarginaltenantreceipt;thatis,

（1-r）.,iftenantinputisincreasedfromt1tot2,thenareaABCrepresentsa”present”fromthetenanttothelandowner[18]—sincetheaddedtenantreceipt（areat1ACt2）islessthantheaddedtenantcost（areat1ABt2）,however,proceedwiththeparadox:iftenantinputisatt1,thenareaMEA（tenantsresidualearning）representsa”present”weent1andt2,bothpartieswillbeoffering”presents”toeachother!