[1].Allricecropshavebeenclassifiedasmoncrops,andbecausetheyarepredominantinChineseagriculture,aseparationisusuallymadebetweenriceandothermoncropsinstatisticalsources.Seethefootnoteattachedtotable3.
[2].Oftheforty-eightmontypesofcroprotationlistedinLu,AnalysisofFarmFamilyEconomy,pp.124-34,riceexistsinmost,frequentlywithvariousmoncrops.Specialcropsandvegetablesexistinonlyafew.
[3].Note,however,thatthemarginfortheincreaseintherateofriceplantingmaybefoundnotonlyinfarmswhereonecropofricewasgrownperyear,butalsoinsomefarmswhereonecropofricewasgrownintwoorthreeyears.
[4].SeeTsung-hanShen,AgriculturalDevelopmentonTaiwansinceWorldWarII(NewYork:ComstockPublishingAssociates,1964),pp.174-95.SeealsoD.H.Grist,Rice(London:Longmans,GreenandCo.,1953).
[5].DepartmentofAgricultureandForestry,TaiwanAgriculturalYearbook,1952Edition(Taiwan:ProvincialGovernment,1952),sec.3.1.orgparingthesetwopercentageswiththetotalincreaseof9.9percent,itbeesclearthatpaddyricewasbyfarthepredominantricecrop.
[6].SeeJCRR,TaiwanAgriculturalStatistics,1901-1955(Taipei,1961),p.12.Notallpaddyricehadbeendouble-cropping,ortherewouldhavebeennoincreaseatallincropareaforpaddyrice.
[7].Shen,AgriculturalDevelopmentonTaiwan,pp.174-95.
[8].Ibid.
[9].DAF,TaiwanAgriculturalYearbook,1952,sec.3.1.
[10].SeeDAF,Yearbook1949,1950,1951and1952,sec.4.Inparticular,theareaactuallydamagedwasroughlythesameeveryyearfrom1948to1951.
[11].DAF,Yearbook1952,sec.3.1.
[12].Ibid.,sec.3.1.D.Forpaddyrice,recallthatthedateofcountingmixedupthefirstandthesecondcrops.
[13].Ibid.,sec.3.1.
[14].Ibid.
[15].Suchaseparationisnotpossiblewiththeavailableinformation.Formoredetailedtreatmentsofthemarginalcrops,anattemptwillbemadetorankprefectures.
[16].DAF,Yearbook1952,Sec.3.2.
C.OutputResponsesofMarginalCrops
Beforeweturnourattentiontothebehaviorofthreetypesofmarginalcrops,letussetthestageonwhichtheresponsesofthesecropsmaybeobservedinmoreconvincingdetail.Thelackofoutputdataseparatedfortenantandownerfarmsmakesitdifficulttoconfirmtheimplicationthatoutputresponsesinthetwotypesoffarmsmoveinoppositedirections.OnemethodIemploytomeasuretherelativeresponsesistoranktheprefecturesinTaiwanaccordingtothedegreetowhichtheirfarmingresourceswereaffectedbythesharerestriction,sothataprefectureofthehigherorderisexpectedtorespondwithgreaterproportionalchangesinthemarginalcrops,accordingtomytheoreticalpredictions.
ThePrefectureRanking
Itshouldbeclearthatthedegreetowhichaprefecturewasaffectedbythesharerestrictiondependsontwofactors:(a)theproportionoffarmlandaffected,and(b)theinitialmarketrentalpercentage,sincetheflatratecontrolledtoagreaterrentalsharereductioniftheoriginalpercentagehadbeenhigher.Ourrankingproblemisgreatlysimplifiedbythefactthat(a)and(b)arepositivelyrelated.Intable4,columns4and6revealthattenancyoccurredmorefrequentlyinpaddyfields.Theexception,Hwalien,canbeexplainedbythepredominanceofuplandricecropsinthisprefecture.Columns4and6,togetherwiththeproportionateacreageofpaddyfieldsanddryland(notshown)determinethepercentagevaluesofcolumn2.
Aparisonofcolumns2and3revealsthat,sincethereducedrentalwasamaximumuniform37.5percentoftheannualyieldforalltenantcontracts,notalltenantfarmswereaffected.Thedifferencesinthesetwocolumnsdependprimarilyontheproportionsofpaddyfieldsanddrylandundertenancy.Paddyfieldsaregenerallymorefertileandhadhighermarketrentalpercentagesthandryland.[1]Asisseenincolumns4through7,paddyfieldswerefarmorevulnerabletotheflatratecontrolthanwasdryland.Columns5and7,togetherwiththeproportionateacreageofpaddyfieldsanddrylandaffected(notshown),determinethepercentagevaluesofcolumn3.
Sinceprefectureswithproportionallymorepaddyfieldswereassociatedwithahigherfrequencyoftenancy,andsincepaddyfieldswereassociatedwithhighermarketrentalpercentages,thevaluesincolumn3renderreliablecriteriafortherankingofprefecturesinorderofthedegreetowhichtheywereaffected.Sucharanking,however,isnotwithoutambiguity.Accordingtomostsources,SinchuandTaipeihadthehighestrentalpercentages;accordingtoanother,Taichungwasthemostfertile.[2]Thedisagreementstemsfromthefactthatnosourceisspeakingofaweightedaverageoftherentalpercentages,andheterogeneouslandwithdifferentrentalpercentagesineachprefecturemakesparisonsdifficult.OursolutionhereistoclassifySinchu,Taipei,andTaichunginthesamegroup.Thus,ingroupIwehaveprefectureswherewearecertainthathigherrentalshadprevailed,andthatprivatelandresourceswereaffectedinthegreatestproportions.[3]Insimilarfashion,weplaceKaoshiungandTainaningroupII,andHwalien,Taitung,andPenghuingroupIII.TheFiveCities,whichoccupiedsmallareasinTaiwan,willbeexcludedbecauseofthemassiveputationsrequired.[4]Rankingbythecriteriaofmarketrentalpercentagesandproportionsofprivatelandaffectedisalsoinaccordancewiththeproportionoffarminghouseholdsaffected.Thelatterisshownincolumn8.Wedonot,however,haveinformationonthenumberoffarmersaffected,whichwouldbethemoreappropriatemeasureoflaborresources.[5]
ThreeTypesofMarginalCrops