Dependingontheirphysicalattributes,somecropsaremoresuitableforinterplantingthanothers.Anycropwhichissuitableforinterplantingmay,ofcourse,beplantedsinglyatalowerplantingcostpercroparea.Meticulouscaremustbeprovidedwhenplantingorharvestingacropbetweenrows.Ontheotherhand,anycropmaybeinterplanted,butweareinterestedonlyinthosewhichhadbeeninterplantedinTaiwan.Notallthe"inter-plantable"cropswereclearlymarginalcropsaddedforfastercroprotation.Forthepresentpurpose,Ieliminatesweetpotatoes,peanuts,barley,andIndiancorn.Thoughthesecropswereinterplantedonsomefarms,theywerelargelyplantedsinglyandweremonlygrown.[9]Next,weeliminatesugarcaneandtobacco.Thesetwocropswerelargelyinterplanted;buttheTaiwanSugarCorporation,agovernmententerprise,grewthegreatestportionofthecanecrop,andtobaccoplantingwasregulatedbythegovernment.[10]Remainingarefiveotherinterplantingcrops—orientalpicklingmelons,cucumbers,watermelons,potatoes,andeggplants.Withahighertenantcostconstraintallowedunderthesharerestriction,thehighertotalyieldobtainablefromthegivenlandthroughinterplantingmadethesecropsan"ideal"choiceasmarginalcropsontenantfarms.
Thefivecropscitedabovealsoservetoillustratetheutilizationofspaceininterplanting.First,theairspaceabovethegroundisutilizedbybiningcucumbersandeggplants:
Cucumbersareusuallyinterplantedwitheggplantsontheridgeofthefield.Betweeneverytworidgesliesaditchfordrainage.Bamboosticksareplacedoneachsideoftheridgeandaretiedtogetheratthetop.Thevinesofcucumbersgrowonthebamboosticks,thusmakinguseofthespacebetweenridges.[11]
Second,theutilizationofgroundspaceisseenintherelayingofwatermelonsandorientalpicklingmelons.Beforetheinitialcropisharvested,thecrawlingvinesofthesemelonsaresuchthatthefruitcanbestrategicallyplaced.Andthird,theunoccupiedsoilspaceundergroundcanalsobeutilized.Potatoes,thoughunmonintheChinesediet,havetheadvantagethattheycanbeplantedsixinchesdeep.
Thesefivecropsarelistedintable6.Inthistable,theprefecturegroupsarelistedincolumn2,withtheirpercentagechangesincropareasincolumn5,andthepercentagechangesinyieldpercrophectareincolumn8.Ingeneral,wefindthat(a)increasesincropareasoccurredmoresignificantlyinprefecturegroupsIandIIthaninIII,(b)theyieldpercrophectaredeclinedingroupsIandII,and(c)theyieldpercrophectareincreasedingroupIII.
Interpretationsfor(a)and(b)abovearethesameasthoseforcitronella.Torepeat,theincreasesincropareaweremoresignificantformarginalthanforintramarginalcrops.Theexhaustionofsuitableinterplantinglandmargins,andperhapsthegenerallypoorplantingorganizationgeneratedbythehighrotationrate,contributedtothedecreaseincrophectareyield.[12]Theimplicationsforthemarginalproductsoffarmingresourcesarethesameasbefore.SinceingroupIIIthereweremainlyfarmsunaffectedbythesharingrestriction,farmingresourcesotherthanlandweredrawnfromthisgrouptogroupsIandII.Thus,acropwhichhadpreviouslybeeninterplantedmightnowbegrownsingly,resultinginariseinyieldpercropareaasdiscernedin(c).Thisalsoconfirmsourpredictionthatoutputresponsesinownerandtenantfarmsmoveinoppositedirectionsunderthesharerestriction.
CropsoftheSeasonalMargin
Perhapsthemostpowerfulconfirmationofmyhypothesisisfoundinathirdtypeofmarginalcrop:vegetables.Therelativelylargecropvarietyonrecordallowsustodetectsomedefinitepatternsofchoiceunderthesharerestriction.OwingtothetropicalclimateinTaiwan,mostvegetablesaregrowninwinter—notthatvegetablescannotbegrowninotherseasons,butinotherseasonsfarmersfindmorelucrativecropstogrow.[13]Underafreemarket,intheshortwinterseason(DecembertoMarch)farmerscultivatesmallplotsoflandasvegetable"gardens."Onlyasmallportionofthebetterland(usuallypaddyfields)ischosenforvegetablesbecausethecostrequiredtocultivatethemisexceptionallyhigh.[14]Thegreaterportionoflandiseitherleftidle,usedtogrowgreenmanureatverylowplantingcost,[15]orusedtogrowwheatandIndiancornintheappropriatefields.
Therelativelyhighcostofplanting,associatedwiththerelativelyhighmarketvalueofvegetables,[16]togetherwiththeidletimemargininwhichtheycouldbegrown,madethemamongthepreferredchoicesasmarginalcropsunderthesharerestriction.Asisshownintable3,thecropareaofvegetablesincreasedthemostofallcropclasses,reachingamaximumincreaseof31.7percentin1951.[17]
Intable7,thenamesofallvegetablesonrecordarelistedincolumn1.Theyarerankedindescendingorderaccordingtothepercentagechangesincroparea(column4).Themajorprefecturescontributingtothepercentagechangesincroparea(increaseordecrease)ofeachvegetablearelistedincolumn8.TheyareallprefecturesbelongingtogroupsIandII,withTaichungandTaipei(groupIprefectures)beingtheoverallmajorcontributors.Withtheexceptionsofvegetableitems1,9,13,and14,forwhichthecrophectareyieldcouldnotbeputed,thepercentagechangesincrophectareyieldarelistedincolumn7.Asisshown,thecrophectareyieldsforallbuttwocrops(garlicandwaxgourd)decreased.[18]Ourinterpretationisthesameasbefore:thegradualuseoflesssuitableland,theurgencyofthetimemarginwhichmustbereleasedformoncropplantations,andperhapsforsometheunfitseason,allcontributedtothediminishingreturnstotenantinputresourcesincultivatingadditionalcrophectares.[19]Overthesameareaof(physical)land,thehighertotalyieldgeneratedbyafasterrateofplantingimpliesthatthemarginalproductoflandundertenantcultivationwashigherthanelsewhere.