ExperimentalAuctions Economists, psychologists, and marketers are interested in determining the monetaryvaluepeopleplaceonnon-marketgoodsforavarietyofreasons: tocarryoutcost-benefitanalysis,todeterminethewelfareeffectsoftechno- logicalinnovationorpublicpolicy,toforecastnewproductsuccess,andto understandindividualandconsumerbehavior.Unfortunately,manycurrently available techniques for eliciting individuals’ values suffer from a serious probleminthattheyinvolveaskingindividualshypotheticalquestionsabout intended behavior. Experimental auctions circumvent this problem because they involve individuals exchanging real money for real goods in an active market.Thisrepresentsapromisingmeansforelicitingnon-marketvalues. LuskandShogrenprovideacomprehensiveguidetothetheoryandpractice ofexperimentalauctions.Itwillbeavaluableresourceforgraduatestudents, practitioners, and researchers concerned with the design and utilization of experimentalauctionsinappliedeconomicandmarketingresearch. JaysonL.LuskisProfessorandWillardSparksEndowedChairintheDepart- mentofAgriculturalEconomics,OklahomaStateUniversity. Jason F. Shogren is Stroock Distinguished Professor of Natural Resource Conservation and Management, and Professor of Economics and Finance, UniversityofWyoming. QuantitativeMethodsforAppliedEconomicsandBusinessResearch SeriesEditor PROFESSORPHILIPHANSFRANSES ErasmusUniversity,Rotterdam Researchers and practitioners in applied economics and business now have access to a much richer and more varied choice of data than earlier generations. Quantitative MethodsforAppliedEconomicsandBusinessResearchisanewseriesaimedatmeeting theneedsofgraduatestudents,researchersandpractitionerswhohaveabasicgrounding instatisticalanalysisandwhowishtotakeadvantageofmoresophisticatedmethodology intheirwork. Forthcomingtitles StewartJonesandDavidHensher(eds.),CreditRiskModelling:APrimer Experimental Auctions Methods and Applications in Economic and Marketing Research JaysonL.LuskandJasonF.Shogren cambridge university press Cambridge,NewYork,Melbourne,Madrid,CapeTown,Singapore,Sa˜oPaulo CambridgeUniversityPress TheEdinburghBuilding,CambridgeCB28RU,UK PublishedintheUnitedStatesofAmericabyCambridgeUniversityPress,NewYork www.cambridge.org Informationonthistitle:www.cambridge.org/9780521671248 (cid:1)C JaysonL.LuskandJasonF.Shogren2007 Thispublicationisincopyright.Subjecttostatutoryexception andtotheprovisionsofrelevantcollectivelicensingagreements, noreproductionofanypartmaytakeplacewithout thewrittenpermissionofCambridgeUniversityPress. Firstpublished2007 PrintedintheUnitedKingdomattheUniversityPress,Cambridge AcataloguerecordforthispublicationisavailablefromtheBritishLibrary LibraryofCongressCataloguinginPublicationdata Lusk,Jayson. Experimentalauctions:methodsandapplicationsineconomicandmarketing research/JaysonL.LuskandJasonF.Shogren. p.cm.–(Quantitativemethodsforappliedeconomicsandbusinessresearch) Includesbibliographicalreferencesandindex. ISBN–13: 978–0–521–85516–7(hbk.) ISBN–10: 0–521–85516–0(hbk.) ISBN–13: 978–0–521–67124–8(pbk.) ISBN–10: 0–521–67124–8(pbk.) 1.Auctions. 2.Consumers’preferences–Mathematicalmodels. I.Shogren,Jason F. II.Title. III.Series. HF5476.L87 2007 330.072(cid:2)4–dc22 2007002291 ISBN 978-0-521-85516-7hardback ISBN 978-0-521-67124-8paperback CambridgeUniversityPresshasnoresponsibilityforthepersistenceor accuracyofURLsforexternalorthird-partyinternetwebsitesreferredto inthispublication,anddoesnotguaranteethatanycontentonsuch websitesis,orwillremain,accurateorappropriate. Acknowledgments Participating in an auction is a collective process and so is conducting one. We have had the good fortune of learning about experimental auctions with a number of colleagues and students including: Todd Cherry, Keith Coble, TomCrocker,ScottDaniel,TyFeldkamp,SeanFox,SaraGunnersson,Dermot Hayes,LisaHouse,DarrenHudson,WallyHuffman,SaraJaeger,JimKlieben- stein, Cannon Koo, Muhammad Koohmaraie, John List, Christy Lusk, Dana Marcellino,MichaelMargolis,DarrellMark,ChrisMcIntosh,BryanMelton, JimMintert,MelissaMoore,BurtMorrow,BaileyNorwood,GregParkhurst, MattRousu,TedSchroeder,S.Y.Shin,LucineTadevosyan,AbeTegene,Bruce Traill, Carlotta Valli, Bob Wilhelm, and Christine Wilson. Sara Jaeger pro- videdhelpfulcommentsonanumberofthechapters.Seminarparticipantsat theJourne´esd’E´conomieExpe´rimentaleatBureaud’EconomieThe´oriqueet Applique´e, Universite´ Louis Pasteur, Strasbourg provided helpful comments onthematerialinchapters9and10. Thisbookisdedicatedtoourfamiliestheirsupportandpatienceduringthe writingofthisbook. v Contents Listoffigures page ix Listoftables x 1 Introduction 1 1.1 Introduction 1 1.2 Whyexperimentalauctions? 3 1.3 Whatisanexperimentalauction? 5 1.4 Purposeofthisbookandboundariesofcoverage 17 2 Incentivecompatibleauctions:theoryandevidence 19 2.1 Introduction 19 2.2 Theoryofincentivecompatibleauctions 20 2.3 Evidencefrominducedvalueauctions 27 3 Valuetheory 34 3.1 Introduction 34 3.2 Valuationundercertainty 34 3.3 Valuationunderuncertainty 37 3.4 Valuationinadynamicenvironmentwithuncertainty,limitedinformation, andirreversibility 43 3.5 Summary 44 4 Conductingexperimentalauctions:somepreliminaries 46 4.1 Introduction 46 4.2 Experimentaldesign 47 4.3 Samplesizedetermination 55 4.4 Experimentsettingandcontext:fieldversuslaboratory 57 4.5 Conclusions 61 5 Conductingexperimentalauctions 62 5.1 Introduction 62 5.2 Trainingandpractice 62 5.3 Endowmentversusfullbidding 65 5.4 Choosinganauctionmechanism 69 5.5 Multiplegoodvaluation,demandreduction,andfieldsubstitutes 76 5.6 Learningandaffiliationinrepeatedbiddingrounds 80 5.7 Negativevalues 92 5.8 Conclusions 94 vii viii Contents 6 Dataanalysis 95 6.1 Introduction 95 6.2 Censoredregressionswithauctionbids 95 6.3 Quantileregressionwithauctionbids 100 6.4 Paneldataregressionwithauctionbids 103 6.5 Othertypesofdataanalysiswithauctionbids 106 6.6 Conclusions 112 7 Valuationcasestudies 113 7.1 Introduction 113 7.2 InformingPolicyI:beeftendernessgradingsystem 113 7.3 InformingPolicyII:valuingsaferfood 121 7.4 InformingPolicyIII:toleranceforgeneticallymodifiedfood 129 7.5 MarketingI:forecastingmarketshareofanewproduct 137 7.6 MarketingII:preferencesforfreshfoodwithmultiplequalityattributes 141 7.7 MarketingIII:thevalueoffarmfinancialrecords 149 7.8 ControversialgoodsI:demandforgeneticallymodifiedfoodinthree countries 154 7.9 ControversialgoodsII:irradiation 163 7.10 ControversialgoodsIII:foodfromanimalstreatedwithgrowth hormones 169 7.11 Concludingcomments 174 Appendices 175 8 Auctiondesign:casestudies 196 8.1 Introduction 196 8.2 Preferencelearning 196 8.3 Willingnesstopay,willingnesstoaccept,andtheauctionmechanism 199 8.4 Secondpriceauctiontournaments 209 8.5 Preferences:fixedorfungible? 217 8.6 Giftexchange 225 8.7 Calibrationofrealandhypotheticalauctionbids 229 8.8 Hybridauctionsandconsequentialbidding 239 8.9 Concludingremarks 245 9 Validityofexperimentalauctions 247 9.1 Introduction 247 9.2 Auctionbidsandeconomictheory 248 9.3 Reliability 252 9.4 Convergentvalidity 255 9.5 Anomalies 261 9.6 Summary 267 10 Thefutureofexperimentalauctions 269 10.1 Introduction 269 10.2 Tenquestionsworthyoffutureresearch 270 10.3 Concludingremarks 278 References 279 Index 297 Figures 1.1 Experimentalauctionsasabalanceofcontrolandcontext page15 5.1 Expectedcostofsub-optimalbiddingforv =3,N=8, i andvalues/pricesaredrawnfromauniformdistribution on[0,10] 75 5.2 Expectedcostofsub-optimalbiddingforv =7,N=8, i andvalues/pricesaredrawnfromauniformdistributionon [0,10] 76 5.3 Secondpriceauctionbidsforfivebeefsteaksacrossfive biddingrounds 81 5.4 Bidfunctionsandthedeterminationofwinner,loser,and marketprice 90 6.1 Distributionoffifthpriceauctionbidsforanon-genetically modifiedcookieinFrance 101 6.2 Hypotheticalpathmodelwheresixauctionbidsare representedbytwolatentfactors 108 7.1 Thethreetypesoflabelsusedforthevegetableoil 133 7.2 Distributionofroundfiveauctionbidsbylocation 159 7.3 Effectofinformationonbidsforirradiatedpork 166 7.4 Averagewillingnesstopayfor“non-BST”milkinIowa, Arkansas,Massachusetts,California(rural)andCalifornia (urban) 172 B7.1 Financialrecordsinventorysheet 190 B7.2 Financialrecordsbidsheet 190 ix Tables 1.1 Examplesofexperimentalauctionsinaction page7 2.1 Payoffsfrombiddingstrategies 22 2.2 Payofffrombiddingtruevalueinsteadofunder-or over-bidding 23 2.3 Resultsofstudiestestingincentivecompatiblemechanisms ininducedvalueexperiments 31 4.1 Fourexperimentaltreatments 48 4.2 Treatmentsina23design 49 4.3 Higherordereffectsina23design 50 4.4 Acomparisonoftwofractionalfactorialdesigns 51 4.5 Samplesizecorrectiontablefor95%levelofconfidence 57 5.1 Someincentivecompatibleauctions 69 5.2 SummaryofpaneldatacategoriesinListandShogren(1999) 83 5.3 Two-wayfixedeffectsestimationresultsforbidequation 84 5.4 Secondpriceauctionbidsforbeefsteakacrossfivebidding rounds 87 5.5 Aggregateandindividualmodelsoftheeffectofposted pricesonbiddingbehavior 88 6.1 Comparisonoftobittodoublehurdlemodel 99 6.2 Conditionalmeanandquantileregressions 103 7.1 Summarystatisticsofauctionbidsandthevalueof tenderness(n=116) 119 7.2 Priceandprobabilityassumptionsusedinwelfare calculations 120 7.3 Consumerwelfarechangesfromatendernessgrading system(allunitsin$perchoiceoccasion;n=116) 120 7.4 Subjectiveandobjectiveriskandacomparisonofna¨ıveand informedoptionprice(Ra)offivepathogens 127 7.5 Summarystatisticsoftestswithineachadditional salmonellatreatment 128 7.6 Bidsonnon-geneticallymodifiedfoodwithdiffering tolerancelevels 135 x