FromtheaboveSchickele（andHeadyafterhim）claimedthatpartofthenonlandinputsmustbesharedbythelandownertoattainefficientresourceuse.Ageneralruleof"efficient"inputsubsidizationbythelandownerwassought.Usinga50percentrentalshare（r=0.5）,andusingthehorizontalaxisoffigure4tomeasurenonlandinputsingeneral,Schickeleconcludedthat"ifthelandlordwouldshareequallyinalltheinputcostsgoingintocropproductiontheintensitywouldbecarriedto[t2],thatis,tothesamedegreewhichcorrespondstomaximumefficiency."[19]Itseconomiccontentaside,whatcouldhavebeenmerelyageometric"accident"waspoundedintoanillusionbyHeady.Headyconverted
and
（1-r）intomarginal（output）costcurves,butwiththecurvesarbitrarilydrawn.[20]Thus,onSchic-kele'sproposal.Headygeneralized:"Thecostofvariablefactors（whereonefactorsuchaslandisfixed）mustbedividedbetweenthelandlordandthetenantintheproportionsthatholdforthedivisionoftheproduct[toattainefficiency]."[21]Letmeclarify.
Returntofigure4,whereweuser=0.4.Supposetherentalpercentage,r,is0.5（thatusedbySchickele）.SchickeleseemstohavethoughtthattxwouldnecessarilydividethedistanceOt2intotwoequalparts.Thus,ifthelandownercontributesnonlandinputsfromt1tot2,hewouldthenbesharingexactlyhalfthenonlandinputcost-orsharingit"intheproportionsthatholdforthedivisionoftheproduct."Butthisiswrongevenifthemarginalproductcurveislinear.Themarginaltenantreceiptcurve,recall,is1—rof
ateverypoint,measuredvertically.Andonlyatonepointwillthesamepercentageholdforahori-zontalmeasure.Whatisinterestinghereisnotthegeometricillusionitself,butaparadoxicalresultoneobtainswithoutit:Givenanyrentalpercentage,andawageratejusthighenoughtoreducetenantinputtozero（accordingtothetax-equivalentanalysis）,thevalueoflandisstillpositive（sincethetotalproductisstillgreaterthanthetotalwage）.Woulditbesensibletosay,then,thatincertaincasesthevalueoflandtothetenantisposi-tivebutheisnotwillingtoofferanysharepaymentforitsuse?
Itisnotclearwhatismeantby"custom"initsusagebySchic-keleandHeadyintheirdiscussionsoflandtenure.WhereasMillhadvisualized"custom"onlyassomethingnonpetitive,SchickeleandHeadyreferredtoboth"custom"and"pe-tition"intheiranalysesofsharetenancy.Oneinterpretationisthatby"custom"theymeantasituationwherethepostulateofwealthorutilitymaximizationdoesnotapply.Yetwithoutanysuchbehavioralpostulate,themeaningofpetitioncannotbedefined.Toplicatethematter,attimesthepostulateofmaximizingbehaviorisimplicitforthetenantbutnotforthelandowner.ThisisexplicitlystatedbyCharlesIssawiiinhisanalysisofsharetenancy.ConcludingalsothatequilibriumisatA（seefigure4）,Issawiiadmittedthat,inhisanalysis:
Ithasbeenimplicitlyassumedthroughout……thatlandlordsdonotrespondreadilytosucheconomicmotivationsasthepossibilityofincreasingtheirinebyinvestment;iftheydid,thedistinctionbetweenfixedrentsandshare-croppingwould,naturally,losemostofitssignificance.Inthepastthisassumptionhas,toalargeextent,heldtrueformostunderdevelopedcountriesand,toaslightlylesserextent,itstillholdstrue.[22]
Thiskindofanalysisisquitepopularin"underdevelopment"literature.AndonewondershowananalystfollowingMarshallwouldfareshouldhediscardtheideaofcustom.WefindthisinastudybyD.GaleJohnson.[23]JohnsonformalizedMarshall'sanalysisingreaterdetail,andhisequationsledhim.toconclude:
Underacrop-sharelease,ifthelandlord'sshareofthecropsishalf,thetenantwillapplyhisresourcesinproductionofcropsuntilthemarginalcostofcropoutputisequaltohalfthevalueofthemarginaloutput.Thesametenant,however,willconducthislivestockoperations,whereimportantcostsarebornebythelandlordandthereceiptsarenotsharedwithhim,intheusualmanner.Thelandlordwillnotinvestinlandassetsunlessthevalueofthemarginalproductistwicethemarginalcost.[24]
Johnsonnoted,however,thathisanalysisisbasedon"circumstancesinwhich……thetenantandthelandlord……eachviewshisinterestseparately,"[25]whichissimilartoMarshall'ssuppositionthatthetenant"isfreetocultivateashechooses."Thissupposition,ofcourse,rendersthemeaningofacontractnebulous.Whatisinterestinghereisthatevenifweacceptthissupposition,Johnson'sconclusionisfoundedonasetofconstraintswhicharedifficulttospecify.Thiswillbeshowninthenextsection.
ButJohnsonwasreluctanttoaccepttheimplicationofinefficientresourceuseunderasharecontract,andhedevotedonesectiontoinvestigatingotherpossibleadjustments.[26]Hefoundthat,"thoughadmittedlyinadequate,theavailableevidenceindicatesthatthecrop-sharecontractyieldsatleastasmuch,ifnotmore,rentperacrethandoesthecashleaseonparablefarms."[27]Inanattempttoreconcilethisapparentconflictbetweentheoryandfact,Johnsonarguedthatwithashort-termleasethetenantisnotreallyfreetocultivateinanywayheseesfit.[28]Thus,theactualintensityoftenantinput"willdependuponwhathethinks'hecangetbywith'."[29]
ItisdifficulttounderstandwhyJohnsondidnotdiscardhistheoreticalanalysisandstartanew-byconsideringthatthecontractingpartiesarefreeonlytoacceptornottoacceptacontract,andthatthey"cangetbywith"onlyasmuchastherestraintsofpetitionallow.Thesechoicesareexactlythesameasfixedrentandwagecontracts,whichareimpliedbytheconstraintsofprivatepropertyrights,andwhichJohnsonhadinmind.Atheoryconstructedontheserightswillrevealthatthetermsofasharecontractareexpressedthroughthemarket-determinedrentalpercentageandratioofnonlandinputstoland.YetonthispointJohnsonwasindoubt: