WPC 4^.QBPF.Z/""K:?& AMa>STq5K3ܹU)y&et-k"4 W#y<(ζHtv9Vy[wjv [Z̈́Uhn0+ ) vj~uj fH'W]H?hr3py-}.lZpI_U#(휛:,"vY+LxTP cy#b[w9Ҳ;sj8psǙaû@)#kf/=P(j`jnD܎7QpIx8S xw'M_f,^ixQ@JdZ)ρIOqJYc1)ҡ;əJz.}g"ghR$K(*x13`Ǻ>*G #!UF % 0(UN/w@}4 mHP LaserJet 5/5M - Standard,,,,009Z+.Courier New RegularX($&d9 Z 6Times New Roman Regular E EE EU" 3|xU !  ݛX˛XXXAConsistentMethodforCalibratingContingentValueSurveyData  CarolMansfield,NicholasSchooloftheEnvironmentDukeUniversity#XCXXX˛#X˛XXXCAbstract `    Thispaperoutlinesastatisticalmethodforcalibratingthedatafromcontingentvalue(CV)surveysderivedfromtheassumptionthatindividualsmakeconstrainedutilitymaximizingdecisions.IndividualsreactdifferentlytoCVquestions,andaneffectivecalibrationmethodmustaccountforthis.#XCXXX˛#X˛XXXCԀThemethodproposedinthis#XCXXX˛#X˛XXXCpaper#XCXXX˛#X˛XXXCԀallowsustodeterminetheinfluenceof `  individualcharacteristicsonbias,asdistinctfromtheirinfluenceonthepreferenceparameters.Toillustratethelogicofthisapproachaspecificfunctionalformforindividualpreferenceswasusedtoderiveclosedformanalyticalexpressionsforanindividual'swillingnesstopay(WTP)andwillingnesstoaccept(WTA).Thesefunctionsallowsystematicdeviationsinindividualresponsestobeexplicitlymodeled#XCXXX˛=#X˛XXXCbyprovidingastructuralinterpretationoftheerrorterm#XCXXX˛e#X˛XXXC.The  frameworkisappropriateforbothopenendedanddichotomouschoicedata.WeillustratethisapproachwiththreeseparateCVdatasetsusingbothtypesofresponses.Ineachcase,thecalibratedCVresponsesarecloseinvaluetoresponsesderivedfromsimulatedmarketexperiments.#XCXXX˛#  X˛XXXCԢJEL񢢢񢢢Classicficationnumbers:H4,Q20tionaiton ` 28pages