Resourceswillbedirectedtotenantfarmsfromallotheruses.Asmorefarmingresourcesaredirectedtotenantfarms,themarginalproductoflandinthesefarmswillbehigher-andthemarginalproductsoftenantinputswillbelower—thanthoseofsimilarresourcesemployedelsewhere.Thisimpliesthatthemarginalproductsoftenantinputswillbelowerthantheassociatedmarginalopportunitycosts;similarly,resourcesinvestedbytenantswillyieldreturnsataratelowerthantheinterestrate.
Thereareseveralcircumstantialandtechnicaldifficultiesinconfirmingtheseimplicationsempirically.First,owingtothewaryearspreceding1947andthemultiplegovernmentalreformprogramsconductedinTaiwanagriculturesince1953,thedataIuseareprimarilyfrom1948to1951,thoughonoccasionIusedataof1947and1952also.[1]Cross-sectiondata,whenavailable,willbeemphasized.Second,inflationandnewcurrencyissuesduringtheperiodunderstudyruleouttheuseofpricedata.Withdifferentcropsharvestedatdifferentdates,forexample,inflationwouldrendertherelativepricesofcropsuseless.Third,forthegreaterpartoftheavailabledata,wecannotseparateownerfarmsfromtenantfarms.Thus,inthenextchapterIshallattempttoranktheprefecturesinTaiwanaccordingtotherespectivemagnitudesofresourcesaffectedbythesharerestriction.Finally,fortechnicalreasonsIshallnotattempttomeasurenumericallythechangesinmarginalproductivity.Onlyimplicationsofthesechangeswillbetested.
ObservationsfromTaiwanwillbepresentedintwochapters.Inthischapter,Ifocusonthereallocationofinputresources.Inthenextchapter,withmoredataavailableoncropyields,Ishallundertakemoredelicatetreatmentofoutputresponses.Asmuchaspossible,Ichoosedatafromprimarysources.
AftercontractswererenegotiatedinApril1949,underwhichamaximumflatrentalrateof37.5percentoftheactualannualyieldwasputintoeffect,adjustmentsininputintensityoccurredinavarietyofways.
A.AdjustmentsintheIntensityofLaborInput
AlthoughthepopulationinTaiwanincreasedduringtheperiodunderstudy,thecivilianlaborforcedeclined.Thepopulationincreasewaslargelyjuvenile,andtherewasanintensificationofnationaldefense.[2]
Intable2,column1containsindexnumbersforthefarmingpopulationduringtheperiodofsharerestriction.Thisrevealsthereallocationoflaborresourcesfromothersectorstotheagriculturalindustryasawhole.Withnegligiblechangesintheareaofcultivatedlandinthesameperiod,[3]theresultwasdecreasinglandholdingperfarmer(column6).Columns2and3revealthatthepopulationofbothownerandtenantfarmersincreased,withtheformerrisingatafasterratethanthelatterafter1950.Thisiswhytheratiooftenantfarmerstoownerfarmers,putedfromtheactualpopulation,declinesin1950and1951(column4).
However,theratiooflandundertenancytoownerlanddeclinedevenmoresignificantly.Beforethesharerestriction,56.3percentoftheprivatelandhadbeenundertenancy.InDecember1949—somesixmonthsafterthesharecontrol—only44.5percentwasundertenantcultivation.[4]Thisindicatesthat20.9percentoflandundertenantcultivationwasconvertedtoownercultivation.Ofthisportionoflandconverted,onlyabout2to3percentcouldbeaccountedforbyleasedismissalsandlandpurchasesbytenants.[5]Themajorfactor,therefore,waspartialrepossessionoflandfromthetenantsbythelandowners,whichinvolvednoleasedismissal.
Thereductionoflandsizeperleasecontractthroughpartiallandrepossessionwasnotrestrictedbylaw.Thelandownerscouldthemselvescultivatethelandrepossessedinthismanner,hirefarmhandstodothetilling,orsellittopartiesotherthanthetenants.This,togetherwithlandpurchasedbytenants,accountedfortheincreaseinownerfarmers.Ontheotherhand,theownerscouldleasetherepossessedportionstonewtenants,resultingintheriseintenantfarmersandleasecontractsshownincolumn3.
Allthisleadstothevaluesincolumns7and8,whichmayberegardedasresultsof"criticalexperiments"suggestingconfirma-tionofthehypothesisofincreasedfarmingintensity.Measuredinhectaresoflandperfarmer,theholdingperownerfarmerincreased,andtheholdingpertenantfarmerdecreasedunderthesharerestriction.Thedataconfirmnotonlythatresourcereallocationtookplacebut,morespecifically,thatunderthesharerestrictiontheintensityoflaborinputtolandwassignificantlyhigherintenantfarmsthaninownerfarms.
Severalremarksneedtobemadehere.First,thevaluesgivenincolumns7and8indicatethatbeforethesharerestrictionthelandholdingpertenantfarmerwasslightlylargerthanperownerfarmer.Thiswassolargelybecauseasmallportionofthecultivatedlandwasrentedtotenantsunderstateownership,andthelaborintensityforthisportionoflandwasreportedlylow.[6]Butwelackinformationseparatingthistypeoftenantfromthosecultivatingprivateland.Thus,insteadoffurnishingroughestimates,weletthisbiasremainthroughout.Thecorrectionofthiswouldleadtoslightdecreasesforallvaluesincolumn8.Second,sincetherentlimitationwasaflatrentalrateof37.5percentforalltenantcontracts,sometenantfarmswereunaffectedbythiscontrol(seechapter8,table4).Iftheunaffectedtenantfarmersweretransferredfromcolumn8tocolumn7,whichistheproperprocedureforourpresentpurpose,thedifferencebetweenthesecolumnsafter1949wouldbeevengreater.[7]Finally,itisclearthatthemajoradjustmentsinlaborintensityoccurredquickly,andfurtherchangesafter1949wereminor.Thisisalsoseenincolumns9and10,wheretheaveragelandsizesforeachleaseandtractdeclinedonlyslightlyfromDecember1949toJune1951.