张五常《蜜蜂的神话》英文版 24页

TheFableoftheBees:AnEconomicInvestigationAuthor(s):StevenN.S.CheungSource:JournalofLawandEconomics,Vol.16,No.1(Apr.,1973),pp.11-33Publishedby:TheUniversityofChicagoPressStableURL:http://www.jstor.org/stable/724823Accessed:01/11/200807:20YouruseoftheJSTORarchiveindicatesyouracceptanceofJSTOR"sTermsandConditionsofUse,availableathttp://www.jstor.org/page/info/about/policies/terms.jsp.JSTOR"sTermsandConditionsofUseprovides,inpart,thatunlessyouhaveobtainedpriorpermission,youmaynotdownloadanentireissueofajournalormultiplecopiesofarticles,andyoumayusecontentintheJSTORarchiveonlyforyourpersonal,non-commercialuse.Pleasecontactthepublisherregardinganyfurtheruseofthiswork.Publishercontactinformationmaybeobtainedathttp://www.jstor.org/action/showPublisher?publisherCode=ucpress.EachcopyofanypartofaJSTORtransmissionmustcontainthesamecopyrightnoticethatappearsonthescreenorprintedpageofsuchtransmission.JSTORisanot-for-profitorganizationfoundedin1995tobuildtrusteddigitalarchivesforscholarship.Weworkwiththescholarlycommunitytopreservetheirworkandthematerialstheyrelyupon,andtobuildacommonresearchplatformthatpromotesthediscoveryanduseoftheseresources.FormoreinformationaboutJSTOR,pleasecontactsupport@jstor.org.TheUniversityofChicagoPressiscollaboratingwithJSTORtodigitize,preserveandextendaccesstoJournalofLawandEconomics.http://www.jstor.org THEFABLEOFTHEBEES:ANECONOMICINVESTIGATION*STEVENN.S.CHEUNGUniversityofWashingtonEconomistspossesstheirfullshareofthecommonabilitytoinventandcommiterrors....Perhapstheirmostcommonerroristobelieveothereconomists.GEORGEJ.STIGLEREVERsinceA.C.Pigouwrotehisbookson"welfare,"1adivergencebetweenprivateandsocialcostshasprovidedthemainargumentforinstitutinggovernmentactiontocorrectallegedlyinefficientmarketactivities.Theanalysisinsuchcaseshasbeendesignedlesstoaidourunderstandingofhowtheeconomicsystemoperatesthantofindflawsinittojustifypolicyrecom-mendations.Bothtoillustratetheargumentandtodemonstratethenatureoftheactualsituation,thequesthasbeenforreal-worldexamplesofsuchdefects.Surprisinglyenough,asidefromPigou"spollutingfactoryandSidgwick"slighthouse,convincingexampleswerehardtocomeby.2Itwasnotuntil1952,morethanthirtyyearsafterPigou"sinitialanalysis,thatJ.E.Meadepro-posedfurtherexamplesandrevitalizedtheargumentforcorrectivegovern-*Facts,likejade,arenotonlycostlytoobtainbutalsodifficulttoauthenticate.Iamthereforemostgratefultothefollowingbeekeepersandfarmers:LeonardAlmquist,NatGiacomini,AncelGoolsbey,L.W.Groves,RexHaueter,HaroldLange,LavarPeterson,ElwoodSires,ClarenceSmith,KenSmith,JohnSteg,P.F.Thurber,andMrs.GeraldWeddle.Allofthemprovidedmewithvaluableinformation;someofthemmadeavailabletometheiraccountingrecordsandcontracts.R.H.Coaseinspiredtheinvestiga-tion,YoramBarzelsawthatitwasconductedthoroughly,andMrs.LinaTongrenderedherassistance.Theinvestigationispartofaproposedresearchinthegeneralareaofcontracts,financiallysupportedbytheNationalScienceFoundation.1A.C.Pigou,WealthandWelfare(1912);andTheEconomicsofWelfare(1920).2Pigouhadofferedotherexamples.TheexampleoftworoadswasdeletedfromlatereditionsofTheEconomicsofWelfare,presumablyinanattempttoavoidthecriticismbyF.H.KnightinSomeFallaciesintheInterpretationofSocialCost,38Q.J.Econ.582(1924).Therailroadexamplehasnotenjoyedpopularity.MostofPigou"sexamples,how-ever,weredrawnfromlandtenurearrangementsinagriculture,butanexhaustivecheckofhissourcereferenceshasrevealednohardevidenceatalltosupporthisclaimofin-efficienttenurearrangements.11 12THEJOURNALOFLAWANDECONOMICSmentactions.3Meade"sprimeexample,whichsoonbecameclassic,concernedthecaseoftheapplefarmerandthebeekeeper.Inhisownwords:Supposethatinagivenregionthereisacertainamountofapple-growingandacertainamountofbee-keepingandthatthebeesfeedontheappleblossom.Iftheapple-farmersapply10%morelabour,landandcapitaltoapple-farmingtheywillincreasetheoutputofapplesby10%;buttheywillalsoprovidemorefoodforthebees.Ontheotherhand,thebee-keeperswillnotincreasetheoutputofhoneyby10%byincreasingtheamountofland,labourandcapitaltobee-keepingby10%unlessatthesametimetheapple-farmersalsoincreasetheiroutputandsothefoodofthebeesby10%....Wecallthisacaseofanunpaidfactor,becausethesituationisduesimplyandsolelytothefactthattheapple-farmercannotchargethebeekeeperforthebees"food....4Meadeappliedasimilarargumenttoareciprocalsituation:AndWhiletheapplesmayprovidethefoodofthebees,thebeesmayfertilizetheapples....Byaprocesssimilartothatadoptedinthepreviouscasewecanobtainformulaetoshowwhatsubsidiesandtaxesmustbeimposed....5anotherwell-knownwork,FrancisM.BatorusedMeade"sInexampletoinfer"marketfailure":Itiseasytoshowthatifappleblossomshaveapositiveeffectonhoneyproduction...anyPareto-efficientsolution...willassociatewithappleblossomsapositiveLagrangeanshadow-price.If,then,appleproducersareunabletoprotecttheirequityinapple-nectarandmarketsdonotimputetoappleblossomstheircorrectshadowvalue,profit-maximizingdecisionswillfailcorrectlytoallocateresources...atthemargin.Therewillbefailure"byenforcement."ThisiswhatIwouldcallanownershipexternality.6Itiseasytounderstandwhythe"applesandbees"examplehasenjoyedwidespreadpopularity.Ithasfreshnessandcharm:thepastoralscene,withitselfinimageofbeescollectingnectarfromappleblossoms,hascapturedtheimaginationofeconomistsandstudentsalike.However,theuniversalcredencetothelightheartedfableissurprising;forintheUnitedgivenStates,atleast,contractualarrangementsbetweenfarmersandbeekeepershavelongbeenroutine.ThispaperinvestigatesthepricingandcontractualarrangementsofthebeekeepingindustryinthestateofWashington,thelocationhavingbeen3SeeJ.E.Meade,ExternalEconomiesandDiseconomiesinaCompetitiveSituation,52Econ.J.54(1952).4Id.at56-57.5Id.at58.6FrancisM.Bator,TheAnatomyofMarketFailure,72Q.J.Econ.351,364(1958). THEFABLEOFTHEBEES13selectedbecausethePacificNorthwestisoneofthelargestapple-growingareasintheworld.Contrarytowhatmostofushavethought,appleblossomsyieldlittleornohoney.7Butitistruethatbeesprovidevaluablepollinationservicesforapplesandotherplants,andthatmanyotherplantsdoyieldlucrativehoneycrops.Inanyevent,itwillbeshownthattheobservedpricingandcon-tractualarrangementsgoverningnectarandpollinationservicesareconsistentwithefficientallocationofresources.I.SOMERELEVANTFACTSOFBEEKEEPINGAlthoughvarioustypesofbeespollinateplants,beekeepingisconfinedalmostexclusivelytohoneybees.8ThehiveusedbybeekeepersinthestateofWashingtonisoftheLangstrothdesignwhichconsistsofoneortwobroodchambers,aqueenexcluder,andfromzerotosixsupers.Abroodchamberisawoodenboxlargeenoughtocontaineightortenmovableframes,eachmeasuring9-~by17-5/by1-3%inches.Withineachframeisawaxhoneycombbuiltbythebees.Inthehexagonalcellsofthiscombthequeenlayshereggsandtheyoungbees,or"brood,"areraised.Itisherealsothatthebeesstorethenectarandpollenwhichtheyuseforfood.Honeyisnotusuallyextractedfromthischamberbutfromtheframesofashallowerbox,calledasuper,placedabovethebroodchamber.Thequeenexcluder,placedbetweenthesuperandthebroodchamber,preventsthelayingofeggsintheuppersection.9Thebees,andconsequentlythebeekeepers,workaccordingtoayearlycycle.AroundthebeginningofMarch,aWashingtonbeekeeperwilldecidewhetherhewantstoprepareforthepollinationseasonbyorderingboosterpackagesofbeesfromCaliforniatostrengthenhiscolonies,depletedand7Thepresenceofapplehoneyinthemarketisthereforesomewhatmysterious.WhileoccasionallyappleorchardsintheNorthwestdoyieldnegligibleamountsofnectar,bee-keepersarefranktopointoutthatthedandelionandotherwildplantsintheorchardareoftenthesourcesof"apple"honey,socalled.Elsewhere,asinNewYork,itwasre-portedthatappleorchardsyieldedslightlymorenectar.See,forexample,A.I.&E.R.Root,TheABCandXYZofBeeCulture386(1923).Theexplanationforthisdivergenceoffacts,tomymind,liesinthedifferentlengthsoftimeinwhichthehivesareplacedintheappleorchards:inRoot"sdaythehiveswereprobablyleftintheorchardsforlongerperiodsthantoday.8SeeGeorgeE.Bohart,ManagementofWildBees,inU.S.Dep"tofAgriculture,Bee-keepingintheUnitedStates109(Ag.HandbookNo.335,1971).[HereinaftercitedasBeekeeping...].Leafcutters,forexample,haverecentlybeenintroducedforthepollina-tionofalfalfaandcloverseeds.Butthesebeesyieldnohoneycropandareseldomkept.9ForfurtherdetailsseeSpencerM.Riedel,Jr.,DevelopmentofAmericanBeehive,inBeekeeping...8-9;A.I.&E.R.Root,supranote7,at440-58;CarlJohansen,Bee-keeping(PNWBulletinNo.79,rev.ed.March1970). 14THEJOURNALOFLAWANDECONOMICSweakenedduringthewinterandearlyspring.Alternatively,hemaydecidetobuildupthecolonybytransportingthehivestofarmsorpasturesinwarmerareas,suchasOregonandCalifornia.Thecolonyhatchescontinuouslyfromspringtofall,andthegrowthrateisrapid.Rearedonpollen,theinfantbeesremaininthebroodstageforaboutthreeweeksbeforeenteringtheproduc-tivelifeofthecolonyforfiveorsixweeks.Activeworkersspendthreeweekscleaningandrepairingthebroodcellsandnursingtheyoung,thenliveouttheremainderoftheirshortlivesforagingforpollenandnectar.x?Becauseofthebees"quickgrowth,theworking"strength"ofacolonyincludesbothbroodandworkers,andincreasesfromaboutfiveframesinearlyspringtoabouttwelvebylatesummer.Springistheprimaryseasonforfruitpollination,andbeekeepersusuallymarketastandardcolonystrengthofroughlyfourframesofbeesandtwotothreeframesofbroodforpollina-tionservices.Butsinceemptyframesareneededtoaccommodatetheexpand-ingcolony,two-storyhives,with16or20frames,areused.Theswarmingperiod,beginninginmid-summerandlastinguntilearlyfall,isthepeakhoneyseason,andtheyieldperhivewillvarypositivelywiththecolonystrength.Becausethemaximizationofhoneyyieldrequiresthatthecoloniesbeofequalstrength,theyareusuallyreassortedinpreparationforthemajorhoneyseason,sothatthenumberofcoloniesatthe"peak"isgenerallylargerthanthenumberinspring.n1Whenpollenfailsinlatefall,thehivesbecomebroodlessandthebeepopulationbeginstodecline.Duringtheidlewintermonthsadultbeesliveconsiderablylongerthanintheactiveseason,andtheycansurvivethewinterifabout60poundsofnectarareleftinthehive.ButinthenorthernpartofthestateandinCanada,wherecoldweathermakestheoverwinteringofbeesmorecostly,thecommonpracticeistoeliminatethebeesandextracttheremaininghoney.Itshouldbenotedherethatbeescanbecaptured,andthattheycanbeeasilyeliminatedbyanyofalargenumberofpesticidesprays.J2Thecostofenforcingpropertyrightsinnectaristhereforemuchlowerthaneconomistshavebeenledtobelieve.10ForfurtherdetailsseeCarlJohansen,supranote9;F.E.Moeller,ManagingColoniesforHighHoneyYields,inBeekeeping...23;E.Oertel,NectarandPollenPlants,inBeekeeping...10.11AccordingtoasurveyconductedbyRobertK.Lesserin1968,basedonasampleof30outof60commercialbeekeepersinthestateofWashington,thetotalnumberofpeakcoloniesis14.6%higherthanthatofspringcolonies.SeeRobertK.Lesser,AnInvesti-gationoftheElementsofIncomefromBeekeepingintheStateofWashington74(un-publishedthesis,Sch.ofBus.Admin.,GonzagaUniv.,1969).12See,forexample,A.I.&E.R.Root,supranote9,at97-103;EugeneKeyarts,BeeHunting,GleaningsinBeeCulture329-33(June1960);U.S.Dep"tofAgriculture,ProtectingHoneyBeesfromPesticides(Leaflet544,1972);CarlA.Johansen,HowtoReducePoisoningofBeesfromPesticides(PamphletEM3473,Wash.St.Univ.,Col-legeofAg.,May1971);PhilipF.Torchio,Pesticides,inBeekeeping...97. THEFABLEOFTHEBEES15Fewagriculturalcrops,tomyknowledge,exhibitahigheryear-to-yearvarianceofyieldthandoesthehoneycrop.Severalnaturalfactorscontribute.Coldweatherandraindiscouragethebeesfromworking,andwindsaltertheirdirectionofflight.Also,thenectarflowsofplantsaresusceptibletoshocksofheatandcold.?3Theplantsyieldingmosthoneyaremint,fire-weed,andthelegumes,suchasalfalfaandtheclovers.Fruittreesusuallyhavelownectarflows,althoughorangeblossoms(inCalifornia)areexcel-lent.Indeed,thepollinationoffruits,especiallythecherryinearlyspring,mayactuallydetractfromtheyieldofhoney:lesshoneymaybeinthehiveafterpollinationthanwasthereinitially,owingtothebees"ownconsumption.Anotherreasonforthelowhoneyyieldfromfruittreesistherelativelyshorttimethatthehivesareleftintheorchards.Cross-pollinationisaccidentallyeffectedasthebeesforagefornectarandpollen.PollinationserviceswerenotmarketedbeforeWorldWarI,primarilybecausesmallfarmshadenoughfloweringplantsandtreestoattractwildinsects.Itwasnotuntil1910andtheadventofmodernorcharding,withitslargeacreageandorderlyplanting,thatmarketsforpollinationservicesbe-gantogrowrapidly.24Today,theservicesaredemandednotonlyforpro-ductionoffruitsbutalsoforthesetting(fertilizing)ofseedsforlegumesandvegetables.Evidenceisincontrovertiblethatthesettingoffruitsandseedsincreaseswiththenumberofhivesperacre,thatthepollinationpro-ductivityofbeesissubjecttodiminishingreturns,and,despitesomebee-keepers"claimstothecontrary,beyondsomepointthemarginalproductivitymayevenbenegative.?5Thereisalsostrongevidencethatpollinationyieldwillimproveifthehivesareplacedstrategicallythroughoutthefarmrather13SeeE.Oertel,supranote10;C.R.Ribbands,TheBehaviourandSocialLifeofHoneybees69-75(1953);RogerA.Morse,PlacingBeesinAppleOrchards,GleaningsinBeeCulture230-33(April1960).Owingtoitsweather,WashingtonisnotoneofthebetterhoneyyieldingstatesintheUnion.DatamadeavailabletomebytheU.S.Dep"tofAgricultureindicatesthatovertheyears(1955-1971)Washingtonranks24thamong48statesinyieldpercolonyand20thinthetotalnumberofcolonies.TheU.S.Dep"tofAgriculturedata,likethoseobtainedbyLesser,providenoinformationonthedif-ferenthoneyyieldsandpollinationrequirementsofvariousplantsandarethereforeoflittleuseforourpresentpurpose.ItshouldbenotedthattheU.S.Dep"tofAgricultureoverallyielddataaresignificantlylowerthanthoseobtainedbyLesserandbyme.SeeRobertK.Lesser,supranote11.14SeeM.D.Levin,Pollination,inBeekeeping...77.15Id.;9thPollinationConference,Report,TheIndispensablePollinators(Ag.Exten-sionServ.,HotSprings,Ark.,October12-15,1970);G.E.Bohart,InsectPollinationofForageLegumes,41BeeWorld57-64,85-97(1960);J.B.Free,PollinationofFruitTrees,41BeeWorld141-51,169-86(1960);U.S.Dep"tofAgriculture,UsingHoneyBeestoPollinateCrops(Leaflet549,1968);GetMoreFruitwithHoneyBeePollinators(PamphletEM2922,Wash.St.Univ.,March1968);ProtectBerryPollinatingBees(PamphletEM3341,Wash.St.Univ.,February1970);IncreaseCloverSeedYieldswithAdequatePollination(PamphletEM3444,Wash.St.Univ.,April1971);HoneyBeesIncreaseCranberryProduction(PamphletEM3468,Wash.St.Univ.,April1971). 16THEJOURNALOFLAWANDECONOMICSthansetinonespot.16Thecloseraparticularareaistoahive,themoreeffectivewillbethepollinationwithinthatarea.Althougheachindividualbeewillforageonlyfewsquareyards,thebeesfromaonehivewillcollectivelypollinatealargecirculararea,17andthisgivesrisetoaproblem:givenahighcosttocontrolfullytheforagingbehaviorofbees,ifsimilarorchardsarelocatedcloseoneanother,onewhohiresbeestopollinatehiswilltoownorchardinsomedegreebenefithisneighbors.Thiscomplicationwillbefurtherdiscussedinthenextsection.InthestateofWashington,about60beekeeperseachorown100coloniesmore;attheseasonthestate"sgrandtotalofpeakcoloniesisabout90,000.Myinvestigation,conductedinthespringof1972,coveredasampleofninebeekeepersandatotalofapproximately10,000springcolonies.(Oneofthesebeekeepersspecializedincut-combhoneyandhewillbetreatedaseparatelyinfootnote.)Table1liststhebee-relatedplantscoveredbymyinvestigation.AsseenfromColumns(3)and(4),someplants(suchascherrytrees)re-quirepollinationservicesforfruitsettingbutyieldnosome(suchasmint)yieldhoneywhilerequiringnohoney;pollinationservice;andsome(suchasalfalfa)areofareciprocalnature.Notethatwhenalfalfaandthecloversaregrownonlyforhay,pollinationservicesarenotrequired,althoughtheseplantsyieldhoney.Thepracticeofrelocatinghivesfromfarmtofarm,bytruck,enablesthebeekeepertoobtainmultiplecropsayear,eitherinrenderingpollinationserviceorinextractinghoney.whilethemaximumobservednumberofcropsperhiveperyearisfourHowever,andtheminimumisthatatwo,myestimateishiveaveragesonly2.2cropsayear.Morefrequentrotationnotonlyinvolvesgreatercostsofmovingandofstandardizinghives,thebutabbreviateshoneyyieldpercrop.thesouthernpartoftheInstate,wheretherela-tivelywarmclimatepermitsanearlyworkingseason,beekeepersusuallybeginbypollinatingeithercherryoralmond(inCalifornia)inearlyspring.Thehivesmayormaythenbemovednorthwardinlateappleandsoftfruitsnotspring,when(andsomelatecherry)begintobloom.18Theleaseperiodforeffectivepollinationduringspringbloomisnomorethanaweek.Butthen,foramonthortwobetweentheendoffruitpollina-TheIndispensable16See,forexample,DouglasOldershaw,ThePollinationofHighBushBlueberries,inPollinators,supranote15,at171-76;RogerA.Morse,supranote13.17Thereis,however,littleagreementasfromonetothreemiles.Forgeneralforagingtohowbehavior,farabeeseecouldfly:estimatedrangeisD.Levin,supranoteO.W.Park,ActivitiesofM.14,at79;Honeybees,inTheHive18andtheHoneybee125,149-206(RoyA.Grouted.,1946);C.R.Ribbands,supranote13.Followingthepracticeoflocalbeekeepers,weusetheterm"softfruit"toreferpeaches,andtopears,apricots,generallygrowninthesamearea,andofteninthesameorchard,asapples.(Bystandardusage,thetermrefersonlythevariousberrytoplants.) TABLE1BEE-RELATEDPLANTSINVESTIGATED(StateofWashington,1971)(2)(3)(4)NumberPollinationSurplus(5)(1)ofServicesHoneyApproximatePlantsBeekeepersRenderedExpectedSeasonFruits&NutsApple&Soft7YesNoMid-April-Mid-MayFruitsaBlueberry1YesYesMay(withmaple)Cherry(early)1YesNoMarch-EarlyCherry2YesNoAprilCranberry2YesNegligibleJuneAlmond(Calif.)2YesNoFebruary-MarchLegumesAlfalfa5YesandNoeYesJune-SeptemberRedClover4YesandNoYesJune-SeptemberSweetClover1NodYesJune-SeptemberPastureb4NoYesLateMay-SeptemberOtherPlantsCabbage1YesYesEarlyApril-MayFireweed2NoYesJuly-SeptemberMint3NoYesJuly-SeptemberaSoftfruitsincludepears,apricots,andpeaches.bPastureincludesamixtureofplants,notablythelegumesandotherwildflowerssuchasdandelions.ePollinationservicesarerenderedforalfalfaandthecloversiftheirseedsareintendedtobeharvested;whentheyareemployedfornectarextraction.dSweetclovermayalsorequirepollinationservices,butsuchacaseisnotcoveredbythisinvestigation. JOURNALECONOMICS18THEOFLAWANDtionandthebeginningofsummernectarflow,thehiveshavelittlealterna-tiveusage.Sincethisperiodissubstantiallylongerthanthetimeneededforthebeekeepertocheckandstandardizehishivesforthehoneycrops,hewillgenerallybeinnohurrytomovethemandwillprefertoleavethemintheorchardswithnoextracharge,unlessthefarmerisplanningtospraywithinsecticide.TheappropriateseasonsforthevariousplantslistedinColumn(5)ofTable1,maynot,therefore,matchthelengthsofhiveleases.Leaseperiodsaregenerallylongerforhoneycrops,forthecollectionofnectartakesmoretime.ThesixthcolumninTableThe1indicatesthevarioushive-densitiesemployed.numberofhivesperacredependsuponthesizeoftheareatobeser-viced,thedensityofplanting,and,inthecaseoffruitoftheorchards.thepollinationoffruits,thepollination,theageForhivesarescatteredthrough-outthefarm,usuallywithhigherdensitiesemployedinolderorchardsbecausethetreesarenotstrategicallyplacedtofacilitatethecrossingofpollen.Themostpopularchoicesareonehiveperacreandonehivepertwoacres.Itisinteresting,andeasilyunderstood,thatfarmersdemandsignificantlyfewerforpollinationthanthenumberhivesrecommendedbyentomologists:T9bothareinterestedinthemaximizationofyield,butforthefarmersuchmaximiza-tionissubjecttotheconstraintofhiverentals.Whenbeesareemployedtoproducehoneyonly,thehivesareplacedtogetherinonelocation,calledanapiary,forgreatereaseofhandling.20Therelativelylargevariationinhivedensitiesrequirediflegumesare,orarenot,tobepollinatedisdiscussedinthenextsection.Beforeweturntoananalysisofthepricingandcontractualbehaviorofbeekeepersandfarmers,mustIpointthatthetwogovernmentoutprogramswhichsupportthebeekeepingindustrydidnotconstituterelevantconstraintsfortheperiodunderinvestigation.Thehoneyinprice-supportprogram,initiated1949,involvespurchaseofhoneyatsupportedpricesbytheCommodityCreditFortheperiodunderCorporation?21investigation,however,thesup-portedpricewasabout20perlowerthanthemarketoftheAgriculturalActofcentprice?22Section804effectuatedin1971anddesignedto1970,reimbursebeekeepersforanylossduetopesticidesprays,hasbeenlargelyignoredby19Seenotesupra.152oSee,forexample,W.P.Nye,BeekeepingRegionsintheUnitedStates,inBee-keeping...17.21SeeHarryA.Sullivan,HoneyPriceSupportProgram,inBeekeeping...136.22From1970to1972thesupportedpriceswerenearthemarketwholesalepricewasaboveperpound.11.5Betweencentsperpound,whereassevenyearsinwhichthe14cents1950and1965wereCCCpurchasednohoney,andtwoyearsofnegligibleamounts.SeeHarryA.Sullivan,supranote21,at137. THEFABLEOFTHEBEES19beekeepersbecauseofthedifficultyoffilingeffectiveclaimswiththefederalgovernment.23II.THEOBSERVEDPRICINGANDCONTRACTUALBEHAVIORItiseasytofindconclusiveevidenceshowingthatbothnectarandpollina-tionservicesaretransactedinthemarketplace:insomecitiesoneneedlooknofurtherthantheyellowpagesoftheTelephoneDirectory.Buttheexis-tenceofpricesdoesnotinitselfimplyanefficientallocationofresources.Itis,therefore,necessarytodemonstratetheeffectivenessofthemarketindic-tatingtheuseevenofthoseresources-bees,nectar,andpollen-which,ad-mittedly,areelusiveincharacterandrelativelyinsignificantinvalue.Indoingso,Ishallnotattempttoestimatethestandardsetsofmarginalvalueswhichanefficientmarketissaidtoequate:theburdenofsuchataskmustrestuponthosewhobelievethegovernmentcancostlesslyandaccuratelymaketheseestimatesfortheimpositionofthe"ideal"tax-subsidyschemes.Rather,Iofferbelowananalysisbasedontheequimarginalprinciple.Totheextentthattheobservedpricingandcontractualbehaviorfailstofalsifytheim-plicationsderivedfromthisanalysisweconcludethat(1)theobservedbehaviorisexplained,and(2)theobservationsareconsistentwithefficientallocationofresources.A.TheAnalysisThereciprocalsituationinwhichabeekeeperisabletoextracthoneyfromthesamefarmtowhichherenderspollinationservicesposesaninterestingtheoreticriddle.Thetraditionalanalysisofsuchaconditionreliesonsomeinterdependentproductionfunctions,andis,Ithink,unnecessarilycomplex.24Themethodemployedheresimplytreatspollinationservicesandhoneyyieldascomponentsofajointproductgeneratedbythehive.Thatis,therentalpriceperhivereceivedbyabeekeeperforplacinghishivesonafarmmaybepaidintermsofhoney,ofamoneyfee,orofacombinationofboth.Themoneyfeeorthehoneyyieldmaybeeitherpositiveornegative,buttheirtotalmeasurestherentalvalueofthehive.23See7U.S.C.?135b,note(1970);Pub.L.No.91-524?804.Myjudgmentisbasedbothonthebehaviorofbeekeepers(seenextsection)aftertheinitiationoftheActandonthecomplexityofrelevantclaimformswhichIhaveathand.InApril1972beekeepersassociationswerestilllobbyingforeasierclaimingconditions.24InJ.E.Meade,supranote3,at58,thisproblemissetupintermsoftheinter-dependentfunctionsx1=H1(ll,cl,x2)andx2=-H2(12,c2,X1).IfindMeade"sanalysisdifficulttofollow.Elsewhere,OttoA.DavisandAndrewWhinstonemploythefunctionsC1=C1(ql,q2)andC2=C2(ql,q2)intheirtreatmentofcertain"externalities."Itisnotclear,however,thattheauthorshadthebeeexampleinmind.SeeOttoA.Davis&AndrewWhinston,Externalities,Welfare,andtheTheoryofGames,70J.Pol.Econ.241(1962). 20THEJOURNALOFLAWANDECONOMICS(/E>~I1-80 THEFABLEOFTHEBEES21ThesolutionisillustratedinFigureI.Weassumethatthehivesareal-waysstrategicallyplaced.InFigureIathecurve(0N/0h)adepictsthevalueofthemarginalnectarproductofafarminwhichbeehivesareusedonlymint,oralfalfagrownonlyfortheextractionofnectar(aswithfireweed,forhay),withthefarmingassetsheldconstant.Giventhemarket-determinedwealthmaximizationimpliesthatrentalpriceofOAperhive,constrainedOQ"ofhiveswillbeemployed.Inthiscase,thebeekeeperwillberemuneratedonlyinhoney,andwillpayanapiaryrentequaltoareaABC(orDBperontheotherhand,depictsthehive)tothefarmer.Thecurve(OP/Oh)b,forafarmwhichemployshivesvalueofthemarginalpollinationproductorchards).Herethenumberofforpollinationonly(aswithcherryorapplehivesemployedwillbeOQ,whichagainistheresultofwealthmaximization.feeperhiveisagainOA,andWithzerohoneyyield,themoneypollinationtheorchardrentisrepresentedbytheareaAGH.WenowturntothejointproductcaseinFigureIb,wherehivesareusedofnectar(asinthesettingofbothforpollinationandfortheextractionand(ON/Oh)crespectivelyalfalfaandcloverseeds).Thecurves(OP/Oh)carethevaluesofmarginalpollinationandofmarginalnectarproducts.Theiristhetotalmarginalvalue.verticalsummation,thesolidline(0V/Oh)c,OQ"ofhives,thepointWealthmaximizationimpliestheemploymentofwheretherentalpriceperhiveequalstheaggregatemarginalvalue.AsThisimpliesthatthevalueofdrawn,areaHIJissmallerthanareaJKM.belowpointK,asitdoesheretheaveragenectarproduct,(N/h)c,mustpassatL.Inthiscasetherentalpriceperhive,KQ",willconsistofLQ"inhoneyproductsituation,ofcourse,yieldandKLinpollinationfee.Forthisjointitispossibletoconstructacaseinwhich(N/h)cpassesabovepointK,thusyieldinganapiaryrent.Itisalsopossibletoconstructcaseswherethenumberofhivesemployedyieldszeroornegativemarginalproductivity,ineithernectarorpollination.Inotherwords,zeroornegativemarginalproductisco.nsistentwithefficientproductivityinonecomponentofthejointallocationofresources.numbersofpotentialparticipantsUnderopencompetition,therearelargetotalmarginalvaluecurvefortheineachofthecasesabove.Theaggregatethehorizontalsumma-market,orthemarketdemandforhives,isthereforetionofalargenumberofthesolidcurvesinFiguresIaandIb.Similarly,summationofthemarginalcoststhemarketsupplyofhivesisthehorizontalofproducingandkeepinghivesofallactualandpotentialbeekeepers.BothmarketcurvesareshowninFigureIc.25Assumingnocostsforcollatingbidsthesecurves,butforourpresent25Morevariablesareusuallyusedinthederivationofthem.purposelittleisgainedbyincorporating 22THEJOURNALOFLAWANDECONOMICSandasksorforformingrentalcontractsamongallactualandpotentialparti-cipants,thepriceperhive,OA,isdeterminedinthemarket.TheParetocon-ditionissatisfied:thevalueofthemarginalproductofahiveisthesameoneveryfarm,andinturnequalstherentalpriceandthemarginaloppor-tunitycostofproducingthehive.B.TestsofImplicationsBeforewederiveandtestsomeimplicationsoftheaboveanalysis,itisnecessarytopointoutthelimitationsoftheinformationathand.Sincenoattemptismadetoestimatethemarginalvaluesortheelasticitiesofthemarginalproducts,wewillseektoconfirmthemarginalequalitieswithsomeobservedaveragevalues.Theseincludeapiaryrent,pollinationfees,honeyyieldsperhive,andthewholesalepriceofhoney.Wealsohaveinformationonthenumberofhivesemployedondifferentfarms,andsomeothernumer-icaldata.Mychoiceofdataforthehoneyyieldperhive,however,mustbequalified.Thelargefluctuationsinyieldfromyeartoyearandevenfromfarmtofarmcausedbyuncontrollablenaturalphenomenamakestheuseoftheactualobservedyieldsofaparticularyear,orevenofafewyears,irrelevantforourpurposes.Take,forexample,theexceptionallypooryearof1971when,inmanycases,theyieldperhivewasjustone-thirdofthatinanormalyear.Thiswindfalllossisirrelevantfordecision-making(althoughtheexpectedvarianceisrelevant),anditcannotbeattributedtomarket"failure."Lackingsufficientdatatocomputethehoneyyieldperhiveextractedfromvariousplantsovertime,Iresorttotheexpectedyieldsasreportedbybeekeepers.Fortunately,theirestimatesforyieldsun-dercomparableconditionsexhibitremarkableconsistency.AnoverallviewofthepricingstructureisshowninTable2.Sinceahivehasdifferentrentalvaluesfordifferentseasons,wedividethetimeperiodintothreeproductiveseasons:earlyspring,latespring,andthehoneyseason(summertofall).Surplushoneyisnotexpectedintheearlyspringseason,althoughnectarmayaccumulateinthebroodchamberandtheremaybeagaininbroodstrength.Mostbeekeepersinthestateareidleduringthisseason,andpollinationisconfinedtoalmondinCaliforniaorcherryinthesouthernpartofWashington.Therentalvalueofhivesisthehighestinthemajorpollinationseasonoflatespring(ApriltoJune),secondhighestinthemajorhoneyseason,andlowestintheearlyspring(March).ThepollinationfeeslistedinTable2arebasedon1971data,buttheyhaveremainedroughlyconstantfrom1970to1972.Thewholesalehoneyprices,however,arebasedon1970andearly1971data,astheunexpectedlylowhoneyyieldthroughoutthecountryin1971generatedaasharpriseinprices(from14centsapoundinApril1971to32centsapoundinMarch TABLE2PRICINGSCHEMESANDEXPECTEDHONEYYIELDSOFBEE-RELATEDPLANTS(StateofWashington,1970-1971)SurplusHoneyHoneyPricesPerPollinationExpected(poundsPound(whole-FeesSeasonsPlantsperhive)sale,1970)1971)EarlyAlmond(Calif.)0-$5-$8SpringCherry0-$6-$8Apple&SoftFruits0-$9-10LateBlueberry(withSpringmaple)4014?$5pmatorCabbage15130$8_$9-$10pollinationCherry0season)Cranberry513?$9Alfalfa6014.5?0Alfalfa(with25-3514.5?$3-$5Summerpollination)andFireweed6014.5?0E:arlyMint70-7511?0FallPasture6014?0(ma)orRedClover601400honeyRedClover(with0-3514?$3-$6season)pollination)SweetClover6014?0 24THEJOURNALOFLAWANDECONOMICS1972).Theapiaryrentsarepaidmostlyinrefinedandbottledhoney,andarethereforeconvertedintomoneyvaluesaccordingto1970retailhoneyprices.Tomaintainconsistencywithpollinationfees,theapiaryrentsarecomputedperhive,althoughinthelattercontractsthenumberofhivesisnotstipulated.Thefollowingtestimplicationsarederivedfromouranalysis:(1)Ourfirstimplicationisthat,atthesameseasonandwithcoloniesofthesamestrength,therentalpriceperhiveobtainedfromdifferentfarmsorbydifferentbeekeeperswillberoughlythesamewhetherthehiveisem-ployedforpollination,forhoneyproduction,orforacombinationofboth.By"roughlythesame"Idonotmeanthathiverentalsareinvariableamongdifferentbeekeepers.Rather,Imeanthat(a)anydifferenceswhichdooccurarestatisticallynomoresignificantthanthoseformostothercommoditiesinthemarket,andthat(b)thereisastrongnegativecorrelationbetweenthepollinationfee(hiverentalinmoney)andtheexpectedhoneyyield(hiverentalinkind).Datafromtheearlyspringseasonarenotsuitabletotestthisimplicationbecauseduringthisperiodtherearegreatvariationsincolonystrength,inthegainsinbroodandunextractednectar,andindistancestravelledbybee-keeperstodeliverthehives.26Lackingsufficientinformationtomakeap-propriateadjustmentsforthesevariationsincalculatingtherentalpriceperhive,weconcentrateondatafromthelatespringandsummerseasons.Incontractingforpollinationservices,beekeepersofferdiscountsforlargernumbersofhivesandforlesselaboratehivedispersals.Ofthefourbee-keepersfromwhomdetailedrecordsareavailable,forexample,eachservedfrom10to14farmsofapplesandsoftfruits;theirmeanhiverentalsinthemajorpollinationseasonrangedfrom$9.20to$9.68andtheircoefficientsofvariationfrom0.025to0.053.27Toreducetheeffectsonpricegeneratedbydiscounts,weusethemeanrentalsfortheabovefourbeekeepersandthereportedmeansfrombeekeeperswhodidnotmaintainrecords.Ourdatathuscompriseseparateobservationsofthemeanhiverentalofeachbee-keeper,ofeachdifferentplant,and(forthesummerseason)ofeachdifferent26Inthepollinationofalmond,forexample,$5.00ischargedforaone-storyhiveand$6.00to$8.00foratwo-storyhive.Ontheonehand,WashingtonbeekeepershavetotraveltoCaliforniatoobtainthisamountwhentheycouldhaveearnedthesamefeelocallyinthepollinationofearlycherry.Ontheotherhand,however,thebroodgainisgreaterwithalmondthanwithcherry;also,unextractednectarinthebroodchambergainssignificantlyinthecaseofalmondbutislikelytosufferanetlosswithearlycherry.27Ananalysisofvarianceperformedforthesefourbeekeepersshowsnosignificantdifferenceintheirmeanrentalsinthepollinationofappleandsoftfruits.However,thecoefficientofvariationoftheirmeans,0.018,islowerthanthosecomputedfromalargerbodyofdata.Thissimplyindicatesaverylowvariationamongthefourwhoprovideddetailedrecords. THEFABLEOFTHEBEES25expectedhoneyyieldforthesameplant.Thelatterseparationisrequisitebecausetheexpectationofhoneyyieldvariesgreatlydependingonwhetherpollinationis,orisnot,requiredinthecaseofsuchplantsasalfalfa.Thecoefficientofvariationofthemeanhiverentalsamongbeekeeperswhoengagedinthepollinationofapples(includingsoftfruits)andcherries(9observationsintotal)is0.035.Theexpectedhoneyyieldfortheseobserva-tionsiszero.Whenweextendthecomputationtoincludecranberry,blue-berryandcabbagepollination(13observationsintotal),withexpectedhoneyyieldsconvertedintomonetarytermsandaddedtothepollinationfees,thecoefficientofvariationis0.042.Wemaymeaningfullycompareourcoeffi-cientsofvariationswiththosecitedbyGeorgeStigler:28automobileprices(0.017)andanthracitecoalprices(0.068).Another,andmoreilluminating,wayoftestingourimplicationisthroughtherelationshipXo-Xl+X2,(1)whereXoisthetotalrentperhive,Xlistherentpaidinmoney,andx2istheexpectedrentpaidinnectar.Duringthemajorpollinationseason,x1isposi-tiveforallourobservations,butduringthesummerhoneyseasonnegativevaluesforx1(thatis,paymentsinapiaryrents)arecommon.Asnotedearlier,x2mayalsobepositiveornegative,butitisgenerallyeitherzeroorpositiveforthelatespringandsummerseasons.Inthemajorpollinationseason,themeanvaluesofequation(1)are$9.65-$9.02+$0.64.ThevarianceofXocanbebrokendownto=2Covo2o62X1+2x2+(xl,x2).(2)Withatotalof13observationsinlatespring,thecorrespondingvaluesare0.166=-1.620+2.317--3.771.Thevariabilityinx1isalmostentirelyaccountedforbythevariabilityinx2,asreflectedbythelargenegativecovarianceterm.Thecoefficientofcorrela-tionbetweenx1andx2is-0.973.Turningtothesummerhoneyseason,wehaveatotalof23observations,coveringmint(3),fireweed(2),pasture(4),sweetclover(1),redclover(6),andalfalfa(7).Themeanvaluesofequation(1)are$8.07-$1.30+$6.77.Thevaluescorrespondingtoequation(2)are0.806=5.414+6.182-10.791.Again,mostofthevariabilityinxlisstronglyandnegativelycorrelatedwiththatofx2.TheremainingvarianceforXo(withacoefficientofvariationof28GeorgeJ.Stigler,TheEconomicsofInformation,69J.Pol.Econ.213(1961). 26THEJOURNALOFLAWANDECONOMICS0.111)islargerherethaninthemajorpollinationseason.Thiscanbeex-plainedasfollows.First,highrisksareassociatedwiththeexpectedhoneyyields,andbeekeepersseemwillingtosettleforlower,butmorecertain,in-comes.Sincexlismorecertainthanx2,beekeepersseemwillingtoacceptalowerXowithahigherratioofxltox2,29andthevariabilityinthisratioislargerinsummerthaninspring.Similarly,theywillacceptalowerex-pectedmeanofx2formintthanforotherhoneycrops,sincemintisgenerallyknowntohavethesmallestvarianceinexpectedhoneyyieldofanycropinthestate.30Asecond,andmoreimportant,factorcontributingtothelargervarianceofx0isthepremiumpaidtobeekeeperstoassumetheriskofpollinatingcrops(notablyredclover)wheretheuseofpesticidespraysonneighboringfarmsposesthedangeroflossofbees.Sinceourinformationisinadequatetosupportadjustmentsforthesefactors,theresultantdistortionsmustremain.Evenso,thecoefficientofcorrelationbetweenxlandx2com-putedfromthedatais--0.933.(2)Theprecedingevidenceconfirmsthattherentalpricesofhivesem-ployedindifferentusesbydifferentbeekeeperslieonaroughlyhorizontalline.However,itdoesnotconfirmthatthesepricesareequatedtothemarginalproductivities.RefertoFigureI,forexample:theemploymentofhivesmightbeatapointsuchasEratherthanatG,B,orK.Wenowturntosometestableimplicationsregardingthetendencytowardtheequalizationofpriceandmarginalproductivity.Oneobviousimplicationisthat,iftheemploymentofhivesrendersnovaluablepollinationservices,thenanapiaryrentwillalwaysbeobserved.Intheentirebodyofevidenceavailabletome,thereisnotasingleobservationtothecontrary,3aandthismeans,referringtoFigureIa,thattheemployment29Thisstatementisdrawnonlyfromcasualconversationswithbeekeepers;noattemptwasmadetoseekrefutingevidence.30Inconclusiveevidenceindicatesthathiverentals(paidinhoney)obtainedfrommintisabout40centslessthanthoseobtainedfromotherhoney-yieldingplants.Althoughavailableinformationisinsufficientforustocomputetheyear-to-yearvari-ancesofthehoneyyieldsofdifferentplants,rangesofyieldsasrecalledbybeekeepersarelargerthanmostagriculturalcrops.Becausehoneyfromminthasanundesirablystrongflavorthatexcludesitfromtheretailmarket,itiseithersoldtobakeriesorusedtofeedbeesduringthewinter.Quiteunderstandably,onionhoneysharesthedistinctionofbeingmuchcheaperthananyother.Generallyratedasthebestisorangehoney,whichcommandsawholesalepremiumofabout1to2centsapound.Betweentheextremes,differentvarietiesofhoneyhaveroughlythesamevalueandaregradedmorebyclaritythanbytaste.31Onebeekeeperspecializingincut-combhoneyreportedthathepaysapiaryrentseventhoughnosurplushoneyisexpected,providedthatgainsinbroodstrengthandinunextractednectarareexpectedtobesubstantial,aswhenthehivesareplacedinafarmwithmaples.Thisbeekeeperisexcludedfromourfirsttestofimplicationbecausehedidnotengageinpollinationandhiscolonieswereofgreaterstrengths. THEFABLEOFTHEBEES27ofhivesistotheleftofpointE.Itshouldbenotedherethatevenintheab-senceofdemandforpollinationsomeiseffectedwhenbeesforagefornectarfromalfalfaandtheclovers,butthisisnottobetreatedasaserviceunlesstheseedsareharvested.Lessobviousimplicationscanbeobtainedfromthecaseofafarmwherehivesmaybeemployedfornectarextractiononlyorjointlywithpollinationservices.Whenwediscussedthereciprocalcase,asdepictedinFigureIb,itwasnotedthateitheranapiaryrentorapollinationfeemaybepaid.Withsimplemanipulation,thefollowingimplicationsareevident:(a)Ifanapiaryrentispaidinthecaseofajointproduct,andifthemarginalpollinationproductispositive,thenumberofhivesemployedperacreisnecessarilygreaterthanwherebeesareusedonlyfornectarex-tractiononthesameorasimilarfarm.(b)Ifapollinationfeeispaidinthecaseofajointproduct,thenumberofhivesemployedperacreisnecessarilygreaterthanwherebeesareusedonlyfornectarextractiononthesameorasimilarfarm.WhilebothimplicationsindicateatendencytowardpointK(inFigureIb),welacksufficientinformationregardingthemarginalpollinationproducttotest(a)above.Butsinceineveryavailableobservationinvolvingpollinationandnectarextractionapollinationfeeispaid,onlyimplication(b)isrelevantforourpurposes.Theevidence,obtainedfromredcloverandalfalfafarms,stronglyconfirmstheimplication.Thedensityofhivesemployedisatleasttwiceasgreatwhenthebeesareusedforbothpollinationserviceandnectarextractionaswhenusedfornectarextractiononly.Asarule,thisincreaseinhivedensityleadstoasharpdecreaseintheexpectedhoneyyieldperhive.Inthetypicalcase,thedensityofhivesinalfalfaandcloverfarmsforpollinationservicesisabout2.5timeswhatwouldbeemployedfornectarextractiononly,andtheexpectedhoneyyieldperhiveisreducedby50percent.Thisindicatesthemarginalnectarproductofahiveisclosetozeroandpossiblynegative.InCut-combhoneyismoreexpensivethanordinaryhoneybecausethecombwax,whichgoeswiththehoney,isaboutthreetimesthepriceofhoneyperpound.Onlyhoneyoftopgrades(veryclear)willbeextracted.Thisobservationisimpliedbythelawofdemand,sincewiththecombtop-gradehoneybecomesrelativelycheap.Impliedbythesamelawalsoisthatthisbeekeeperchoosestoforgopollinationcontractssothatahigherhoneyyieldcanbeobtained(seeevidenceinimplicationtest2).Evenduringthemajorpollinationseason,whenlittlehoneycanbeexpected,hepreferstoplacehishivesinfarmswherethecolonieswillgaingreaterstrengththanwouldoccuriftheywereusedforpollination.Forarelateddiscussiononsimilarimplicationsofthelawofde-mand,seeArmenA.Alchian&WilliamR.Allen,ExchangeandProduction:TheoryinUse78-79(1969).TheseimplicationsareacceptedhereinspiteofthecriticismsinJohnP.Gould&JoelSegall,TheSubstitutionEffectsofTransportationCosts,77J.Pol.Econ.130(1969). 28THEJOURNALOFLAWANDECONOMICSoneextremecase,inaredcloverfarmthehivedensitywithpollinationser-vicesisreportedataboutsevenoreighttimesthatfornectarextractiononly;sincetheexpectedhoneyyieldisthenreducedtozero,themarginalnectarproductofthehiveisclearlynegative!But,asnotedearlier,zeroornegativemarginalproductinonecomponentofajointproductisconsistentwithefficientallocationofresources.(3)Itremainsforustoshowthattherentalpriceofahiveisroughlyequaltothemarginalcostofkeepingit.Lackingdataonmarginalcost,wewillshowthatthepriceapproximatestheaveragecost,asimpliedbycompeti-tion.Wewillmakethecomparisonintermsofsomegeneralconsiderations.Theexpectedannualincomeofaspringcolonyunderanormalrateofutiliza-tion,asof1970-1971,isabout$19.00.Thisincludesrentalsfromapollina-tioncrop,ahoneycrop,anoccasionalextracrop(forsomehives),andasmallamountfromthesaleofbeeswax.32Thecostsofdeliveringormovingahiveandoffindingandcontractingthefarmersforitsuseareestimatedtototalabout$9.00peryear.33Thisfigureisobtainedasfollows.Somebee-keepersleasesomeoftheirhivestootherbeekeepersonasharecontractbasis;thelessorreceives50to55percentofwhateverincomeinmoneyandinkindthelesseeobtainsfromthefarmers.Sincethelessorcouldhavecon-tractedtoservethefarmershimselfandobtainedtheentireincomeofthe$19.00,thefactthathehaschosentotake45to50percentlessindicatesthat$9.00mustapproximatesuchcosts.Theinterestforgoneinkeepingahiveisabout$3.00peryear.34Thecostofrenewingthecolonystrengthinearlyspringisabout$4.50,thepriceofastandardboosterpackageofbees.35Thisleavesabouttocoverthecostsofdepreciationofthehivevalue,$2.50thelaborinvolvedincheckingandstandardizinghives,spaceforkeepinghivesinthewinter,andtheequipmentusedforhoneyextraction.32InLesser"sinvestigation(supranote11)theactualmeanannualincomeofaspringcolonyfortheyear1967wasestimatedtobe$14.71,andtheactualhoneyyieldsofthatyearwereslightlylargerthanourexpectedhoneyyields.Butin1967thepriceofhoneywasabout16%lowerthanthatin1970;andLesser"sestimateofpollinationincomeperhiveisabout37%lowerthanmine,owingbothtoariseinpollinationfeesinrecentyearsandtodifferentsamplingsofbeekeepers.AccordingtoLesser"sestimate,beeswaxconstitutes4.4%ofthebeekeeper"stotalincome.33Themovingcostscoverlabor,truck,andotherhive-handlingequipment.Dependingonthetimeoftheyear,acompletehive(withsupers)weighssomewherebetween80and250pounds.34Acompletehive,usedbutingoodcondition,sellsforabout$35.00.Theborrowingrateofinterestforthebeekeepersisaround8%.35Thenectarleftunextractedinthebroodchamber,whichconstitutesthemajorcostofoverwintering,isnotcountedaspartofincomeandthereforeisnotcountedaspartofthecost. THEFABLEOFTHEBEES29C.CharacteristicsoftheContractualArrangementsContractsbetweenbeekeepersandfarmersmaybeoralorwritten.Ihaveathandtwotypesofwrittencontracts.Oneisformallyprintedbyanassocia-tionofbeekeepers;anotherisdesignedforspecificbeekeepers,withafewprintedheadingsandspaceforstipulationstobefilledinbyhand.36Asidefromsituationswhereathirdpartydemandsdocumentedproofofthecon-tract(aswhenabeekeeperseeksabusinessloan),writtencontractsareusedprimarilyfortheinitialarrangementbetweenparties;otherwiseoralagreementsaremade.Althoughawrittencontractismoreeasilyenforceableinacourtoflaw,extra-legalconstraintsarepresent:informationtravelsquicklythroughthecloselyknitsocietyofbeekeepersandfarmers,37andthemarketwillpenalizeanypartywhodoesnothonorhiscontracts.Oralcontractsarerarelybroken.Pollinationcontractsusuallyincludestipulationsregardingthenumberandstrengthofthecolonies,therentalfeeperhive,thetimeofdeliveryandre-movalofhives,theprotectionofbeesfrompesticidesprays,andthestrategicplacingofhives.Apiaryleasecontractsdifferfrompollinationcontractsintwoessentialaspects.Oneis,predictably,thattheamountofapiaryrentseldomdependsonthenumberofcolonies,sincethefarmerisinterestedonlyinobtainingtherentperapiaryofferedbythehighestbidder.Second,theamountofapiaryrentisnotnecessarilyfixed.Paidmostlyinhoney,itmayvaryaccordingtoeitherthecurrenthoneyyieldorthehoneyyieldofthepre-cedingyear.38Ingeneral,contractualarrangementsbetweenbeekeepersandfarmersdonotmateriallydifferfromotherleasecontracts.However,somepeculiarar-rangementsresultingfromcertaincomplicationsareworthnoting.First,be-causeoftheforagingbehaviorofthebeesafarmerwhohiresbeesmaybene-fithisneighbors.Second,theuseofpesticidespraysbyonefarmermaycause36Somebeekeepersusejustpostalcards.ThegeneralcontractualdetailsreportedbelowaresimilartothosebrieflymentionedinGrantD.Morse,HowAboutPollination,GleaningsinBeeCulture73-78(February1970).37Duringmyconversationswithbeekeepers,Iwasimpressedbytheirpersonalknowl-edgeofoneanother,includingdetailssuchasthenumberofhivesowned,thekindsoffarmsserved,andtherentsreceived.38Whilewemayattributethisbehaviortotheaversionofrisks,theapiarycontractsarenotthesameassharecontracts.Rather,theyresemblefixed-rentcontractswithwhatIhavecalled"escapeclauses."Fordiscussionofthe"escapeclause"andthestipulationsofthesharecontract,seeStevenN.S.Cheung,TheTheoryofShareTenancy,ch.2&4(1969).OneimpressionIobtainisthatapiaryrentsgenerallyinvolvesuchlowvaluesinWashingtonthatelaborateformationsandenforcementsofapiarycontractsarenotworthwhile.Infurtherinvestigationsofthesecontracts,stateswithhigherhoneyyieldsarerecommended. 30THEJOURNALOFLAWANDECONOMICSdamagetothebeesonanadjacentfarm.Andthird,fireweed,whichyieldsgoodhoney,growswildinforests.Letusdiscusseachinturn.TheCustomoftheOrchards.Asnotedearlier,ifanumberofsimilarorchardsarelocatedclosetooneanother,onewhohiresbeestopollinatehisownorchardwillinsomedegreebenefithisneighbors.Ofcourse,thestrategicplacingofthehiveswillreducethespilloverofbees.Butintheabsenceofanysocialconstraintonbehavior,eachfarmerwilltendtotakeadvantageofwhatspilloverdoesoccurandtoemployfewerhiveshimself.Ofcourse,contractualarrangementscouldbemadeamongallfarmersinanareatodeterminecollectivelythenumberofhivestobeemployedbyeach,butnosucheffortisobserved.Acknowledgingthecomplication,beekeepersandfarmersarequicktopointoutthatasocialrule,orcustomoftheorchards,takestheplaceofexplicitcontracting:duringthepollinationperiodtheownerofanorchardeitherkeepsbeeshimselforhiresasmanyhivesperareaasareemployedinneighboringorchardsofthesametype.Onefailingtocomplywouldberatedasa"badneighbor,"itissaid,andcouldexpectanumberofincon-veniencesimposedonhimbyotherorchardowners.39Thiscustomarymatchingofhivedensitiesinvolvestheexchangeofgiftsofthesamekind,whichapparentlyentailslowertransactioncoststhanwouldbeincurredunderexplicitcontracting,wherefarmerswouldhavetonegotiateandmakemoneypaymentstooneanotherforthebeespillover.40TheCaseofPesticideSprays.Attheoutset,wemustrememberthattominimizethelossofbeesfrominsecticideusageisnotnecessarilycon-sistentwithefficientallocationofresources.Therelevantconsiderationiswhetherthegainfromusingthepesticideisgreaterthantheassociatedlossofbees,intotalandatthemargin.Providedthatthecostsofformingcon-39Thedistinctionbetweenanoraloranimplicitcontractandacustomisnotalwaysclear.Acommonpracticeinsomeareasisthateachfarmerletshisneighborsknowhowmanyhivesheemploys.Perhapstheabsenceofacourtoflawtoenforcewhatcouldinfactbeahighlyinformalagreementisthereasonwhyfarmersdenytheexistenceofanycontractamongthemgoverningtheemploymentofhives.40Sincewithasufficientlyhighrewardthenotorietyofbeinga"badneighbor"willbetolerated,thelikelihoodofexplicitcontractingriseswithincreasingrentalvaluesofhives.Alternativelyandconcurrently,withahighenoughrentalpriceofhivestheaveragesizeoforchardsmayincreasethroughoutrightpurchases,ortheshapesoftheorchardsmaybesotailoredastomatchtheforagingbehaviorofthebees.Bydefinition,giventhegainstheleastcostlyarrangementwillbechosen.Somebeekeepersreportedthattherearepeculiarsituationswheretheforagingbehaviorofthebeesforcesaone-waygift,butthesesituationsarenotcoveredbythepresentinvestigation.Evenundertheseraresituations,theabsenceofbothcontractualandcustomaryrestraintsmaynotresultinadifferentallocationofresources.SeeStevenN.S.Cheung,TheTheoryofInter-individualEffectsandTheDemandforContracts(Univ.ofWashington,Inst.ofEcon.Res.). THEFABLEOFTHEBEES31tractspermits,beekeepersandfarmerswillseekcooperativearrangementssuchthattheexpectedmarginalgainfromusingthepesticideisequaltothevalueoftheexpectedmarginalbeeloss.Intheabsenceofthearrangements,however,thetotalgainfromusingthepesticidemaystillbegreaterthantheassociatedloss;thegreatertheexpecteddamagedonetobees,thegreaterwillbethegainfromthecooperativearrangements.4xWhenapollinationcontractisformed,thefarmerusuallyagreestoinformthebeekeeperbeforesprayinghiscrop,butthisassurancewillnotprotectthebeesfrompesticideusedonneighboringfarms.Inareasdominatedbyorchardswhichrequirepollinationatroughlythesametime,suchastheapple-growingdistricts,thisagreementwillsuffice,fornofarmerwillapplythesprayduringthepollinationperiod.Butinregionswhereadjacentfarmsrequirebeepollinationatdifferenttimes,ordonotrequireitatall,afarmerwithnopresentobligationtoanybeekeepermaysprayhisfieldsandinflictdamagestothebeesrentedbyotherfarms.Inthissituation,onlycooperationoveralargegeographicareacanavoidbeeloss,andwefindjustsucharrange-mentsinthepollinationofcranberriesbutnotofredclover.CranberryfarmsnearSeattleareusuallyfoundinclusters,andsprayingisconductedshortlyafterthebloom,whichmayvarybyasmuchasaweekortwoamongneighboringfarms.Althougheachcranberrygroweragreesnottosprayuntilthecontractedbeekeeperremovesthebeesfromhisfarm,thisdoesnotprotectbeeswhichmaystillremainonadjacentfarms.Thereforethebeekeepersmakeafurtherarrangementamongthemselvestoremoveallhivesonthesamedate,thusinsuringthatallthebeesareprotected.Redcloverpresentsadifferentsituation.Sincetheplantisoftengrowninareaswhereneighboringfarmsrequirenobeepollination,thepesticidedangerisreportedlyhighandbeekeepersdemandanadditional$1.00to$2.00perhivetoassumetherisk.Butjustasthebeekeeperscooperatewithoneanotherduringcranberrypollination,acloverfarmercouldmakearrangementswithhisneighbors.Giventhatneighboringfarmershavethelegalrighttousepesticide,thecloverfarmerwouldbewillingtopaythemanamountnotex-ceedingthebeekeeper"sriskpremiumiftheywouldrefrainfromsprayingduringthepollinationperiod.Althoughnosucharrangementsareobserved,itwouldseemthatthecostsofreachinganagreementwouldbenohigherthanthoseencounteredinthecaseofthecranberries,andwemustinfer,pendingempiricalconfirmation,thatthegainfromusingthespraysisgreaterthantheassociatedloss.Thiswouldparticularlyapplywhenasinglefarmrequiringpollinationislocatedamidstalargenumberoffarmswhichrequiresprayingduringthatsameperiod.41Forafullerdiscussion,seeStevenN.S.Cheung,supranote40. 32THEJOURNALOFLAWANDECONOMICSTheCaseofFireweed.Ihaveathandtwotypesofapiarycontractper-tainingtofireweed,ahoneyplantwhichgrowswildintheforest.ThefirstisbetweenabeekeeperandtheWeyerhaeuserCompany,ownerofprivatetimberland;thesecondisbetweenabeekeeperandtheWaterDepartmentoftheCityofSeattle.Twodistinctionsbetweenthemareworthnoting.First,whilebothcontractsstipulate25centsperhive,Weyerhaeuserasksamini-mumchargeof$100,andtheWaterDepartmentaminimumof$25.Intheapiaryforfireweedhoney,thenumberofhivesusedbyabeekeeperismorethan100butlessthan400.ThusithappensthatinthecaseofWeyerhauser,theapiaryrentisindependentofthenumberofhives,whereaswiththeWaterDepartmentitisdependent.The"underpriced"rentleviedbytheWaterDepartmentwouldhaveimpliedsomesortofqueuingexceptthataseconduniquefeatureisincorporatedinitsapiarycontracts:nobeekeeperisgrantedtheexclusiverighttothefireweednectarinaparticulararea.Theimplicationisthatcompetitionamongbeekeeperswillreducethehoneyyieldperhiveuntilitsapiaryrentisnomorethan25cents;whilenobeekeeperattemptstoexcludeentrants,thepartiesdoseekamutualdivisionofthetotalareatoavoidchaotichiveplacement.Finally,fireweedalsogrowswildinthenationalforestsandforthiscaseIhavenocontractathand.Myinformationisthatapiaryrentismeasuredbythehive,issubjecttocompetitivebiddingamongbeekeepers,andhasareportedrangeof25to63centswiththewinnerbeinggrantedexclusiverighttoaparticulararea.III.CONCLUSIONSWhetherornotKeyneswascorrectinhisclaimthatpolicymakersare"distillingtheirfrenzy"fromeconomists,itappearsevidentthatsomeeconomistshavebeendistillingtheirpolicyimplicationsfromfables.Inadesiretopromotegovernmentintervention,theyhavebeenpronetoadvance,withoutthesupportofcarefulinvestigation,thenotionof"marketfailure."Somehavedismissedincavalierfashionthepossibilityofmarketoperationsinmattersofenvironmentaldegradation}aswitnessestheassertionofE.J.Mishan:Withrespecttobodiesoflandandwater,extensionofpropertyrightsmayeffectivelyinternalizewhatwouldotherwiseremainexternalities.Butthepossi-bilitiesofprotectingthecitizenagainstsuchcommonenvironmentalblightsasfilth,fume,stench,noise,visualdistractions,etc.byamarketinpropertyrightsaretooremotetobetakenseriously.4242E.J.Mishan:AReplytoProfessorWorcester,10J.Econ.Lit.59,62(1972).AsimmediaterefutationofProfessorMishan"sclaim,Ireferthereadertoafactualexample:ProfessorJohnMcGeehasjustpurchasedahouse,separatedfromthatof THEFABLEOFTHEBEES33Similarly,ithasbeenassumedthatprivatepropertyrightscannotbeen-forcedinthecaseoffisheries,wildlife,andwhateverotherresourceseconomistshavechosentocall"natural."Landtenurecontractsareroutinelytakenasinefficient,andtosomethemarketwillfailintheareasofeduca-tion,medicalcare,andthelike.Then,ofcourse,thereisthefableofthebees.Ineachcase,itistruethatcostsinvolvedinenforcementofpropertyrightsandintheformationofcontractswillcausethemarkettofunctiondifferentlythanitwouldwithoutsuchcosts.Andfewwilldenythatgovern-mentdoesaffordeconomicadvantages.Butitisequallytruethatanygov-ernmentactioncanbejustifiedonefficiencygroundsbythesimpleexpedientofhypothesizinghighenoughtransactioncostsinthemarketplaceandlowenoughcostsforgovernmentcontrol.Thustoassumethestateoftheworldtobeasoneseesfitisnoteventocomparetheidealwiththeactualbut,rather,tocomparetheidealwithafable.IhavenogroundsforcriticizingMeadeandothereconomistswhofollowthePigoviantraditionfortheiruseofthebeeexampletoillustrateatheo-reticalpoint:certainly,resourceallocationwouldingeneraldifferfromwhatisobservedifthefactorswere"unpaid."Mymaincriticism,rather,concernstheirapproachtoeconomicinquiryinfailingtoinvestigatethereal-worldsituationandinarrivingatpolicyimplicationsoutofsheerimagination.Asaresult,theirworkcontributeslittletoourunderstandingoftheactualeconomicsystem.hisneighborbyavacantlot.Thatthespacewouldremainvacanthadbeenassuredbythepreviousownerwho(uponlearningthatathirdpartywasplanningtobuythelotandconstructahousethere)hadnegotiatedwiththeneighbortomakeajointpurchaseoftheground,thusprotectingtheirtwohouseholdsfromthe"filth,fumes,stench,noise,visualdistractions,etc."whichwouldbegeneratedbyanewneighbor.