The effects of behavior and attitudes on drop-off recycling activities

ContentslistsavailableatScienceDirect

Resources,ConservationandRecycling

journalhomepage:www.elsevier.com/locate/resconrec

Theeffectsofbehaviorandattitudesondrop-offrecyclingactivities

ShaufiqueF.Sidiquea,∗,FrankLupib,SatishV.Joshib

ab

FacultyofEconomicsandManagement,UniversitiPutraMalaysia,43400UPMSerdang,Selangor,Malaysia

DepartmentofAgriculture,FoodandResourceEconomics,MichiganStateUniversity,AgricultureHall,EastLansing,MI48824,USA

articleinfoabstract

Toreducetheamountofwasteenteringlandfills,policymakersandgovernmentshaveimplementedvariousrecyclingandwastereductionprogramssuchassourcereduction,curbsiderecyclinganddrop-offrecyclingprograms.Thesuccessofarecyclingprogramlargelydependsonhouseholdparticipationandsortingactivities.Abetterunderstandingofrecyclingbehaviorwillhelpusaidthedesignandimprovetheeffectivenessofrecyclingpolicies.Thispaperstudiestheprofileofpeoplewhoutilizedrop-offrecyclingsitesandanalyzesthefactorsinfluencingtheirsiteusage.Theresultsshowthattheusageofdrop-offrecyclingsitesisinfluencedbydemographicfactorssuchasage,education,incomeandhouseholdsize.Attitudinalfactorsarealsofoundtoaffectsiteusage.Recyclerstendtousethedrop-offsitesmorewhentheyfeelthatrecyclingisaconvenientactivityandwhentheyaremorefamiliarwiththesites.

© 2009 Elsevier B.V. All rights reserved.

Articlehistory:

Received27April2008

Receivedinrevisedform19January2009Accepted22July2009

Available online 25 August 2009Keywords:Recycling

Drop-offprogramsDrop-offsites

CurbsideprogramsWastemanagementRecyclingbehavior

1.Introduction

In2007,UnitedStatesresidents,businesses,andinstitutionsproducedapproximately2milliontonsofmunicipalsolidwasteandmorethan50%ofthesolidwastegeneratedwaslandfilled(USEPA,2007).Toreducetheamountofwasteenteringland-fills,localgovernmentshaveimplementednumerousrecyclingandwastereductionprograms,includingsourcereduction,variablegarbagepricing,curbsiderecyclinganddrop-offrecycling.

Drop-offrecyclingisarecyclingprogramwheredesignatedsitesareestablishedtocollectarangeofrecyclablesandusuallyrecyclersthemselvesarerequiredtodepositthesortedrecyclablesinspe-ciallymarkedcontainers.Drop-offrecyclingcentersarelesscostlytooperatecomparedtocurbsideprograms,andtheyarealsofastertoimplementthantake-backanddepositrefundprogramsinvolv-ingmanufacturers(Saphoresetal.,2006).Drop-offcenteroperatorsareabletosaveonlaborandtransportationcostsbecausethesecostsaretransferredtotherecyclers.Drop-offrecyclingisalsocon-sideredtobeafinanciallyviablerecyclingoptioninruralareaswithlowpopulationdensity(Tilleretal.,1997).Asaresult,drop-offrecy-clingisawidelyadoptedrecyclingprogrambylocalgovernments.Asof1998,over12,000recyclabledrop-offsiteswereoperatingintheUnitedStates(USEPA,2000).

Despitewideimplementation,relativelylittleliteratureana-lyzesdrop-offrecycling.Researchoncurbsiderecyclingand

∗Correspondingauthor.

E-mailaddress:shaufique@econ.upm.edu.my(S.F.Sidique).

0921-3449/$–seefrontmatter© 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.resconrec.2009.07.012

variablegarbagepricingismorepopularinthefieldofrecyclingandwastemanagement.Forexample,FullertonandKinnaman(1996),HongandAdams(1999),VanHoutvenandMorris(1999),KinnamanandFullerton(2000)andJenkinsetal.(2003)analyzetheeffectofcurbsiderecycling,togetherwithvariablegarbagepricing,ontheamountofwastegenerationandrecycling.HongandAdams(1999)studytheeffectofdisposalfeeandhouseholdcharacteristicsonrecyclingratesandwastegenerationusinghouseholddatafromPortland,Oregonandfindthatanincreaseinthepriceofsolidwastecollectionincreasesboththedemandforrecyclingandrecyclingrates.Hong(1999)reportssimilarresultsforahouseholdsamplefromKorea.KinnamanandFullerton(2000)findthatanunitfeehasanegativeeffectongarbageweight,andcurbsiderecyclingprogramshaveapositiveeffectonrecycling.Otherstudieshaveestimatedtheconsumerwillingnesstopayforcurbsiderecyclingservices(Lakeetal.,1996;AadlandandCaplan,1999,2003;Blaineetal.,2005).

Arareexampleofrecyclingresearchthatisrelatedtodrop-offrecyclingisthestatedpreferencestudyofadrop-offprogramcon-ductedbyTilleretal.(1997).Theirstudyanalyzedtheeconomicfeasibilityofestablishingadrop-offrecyclingprograminaruralandsuburbanareaofTennesseebyutilizingthecontingentval-uationmethodtocalculatehouseholdwillingnesstopay(WTP)fortheprogram.TheyfoundthattheestimatedWTPdependedonrespondents’income,educationlevel,ageandattitudestowardstheimportanceofrecycling.However,thestudywasbasedonstatedpreferencesofhouseholdsandnotonobservedbehaviororrevealedpreference.SpeirsandTucker(2001)studiedthepro-fileofrecyclersutilizingdrop-offrecyclingsitesinGlasgowand

1S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170

acrossAyrshireinsouth-westScotland.Theyreportrecyclers’traveldistances,theweightsandtypesofrecyclablesanddemographiccharacteristics.However,thestudyismainlydescriptiveanddoesnotanalyzetherelationshipbetweendrop-offsiteutilizationandtheprofileofrecyclers.Similarly,theOhioEPAconductedastudyofparticipationratesandusagepatternsofrecyclersatdrop-offsitesinOhioin2004(Snyderetal.,2004).Theprimarypurposeofthestudywastoassessthepercentageofpopulationinawasteman-agementdistrictwithaccesstorecyclingfacilitiestoseewhetheritmetthesettargetsandthereportedresultsaremainlydescriptivestatistics.Saphoresetal.(2006)analysisofhouseholds’willingnesstorecycleatdrop-offcentersislimitedtoelectronicwaste.

Weaddressthisgapintheliteraturebystudyingindividu-alsutilizingdrop-offrecycling.Weconductedonsiteinterviewsofrecyclersusingeightdrop-offrecyclingcentersaroundthecityofLansinginMichigan.Westatisticallyanalyzetherelationshipbetweenthenumberofvisitsmadetodrop-offrecyclingsitesandrecyclercharacteristicsandtheirattitudestowardsrecycling.Thefindingsprovideinsightsintorecyclingbehaviorthatcanbeusedtoimprovethedesignandeffectivenessofdrop-offrecyclingpolicies.2.Literaturereview

Thesuccessofadrop-offrecyclingprogramislargelydepen-dentonhouseholdparticipationandsortingactivitieswhicharetheessentialbehaviors.Priorresearchindicatesthatrecyclingbehav-iorisinfluencedbycostofrecycling,convenienceofavailablerecyclinginfrastructureandprograms,theextentofenvironmentrelatedawarenessandknowledge,attitudestowardsrecycling,socialnormsandexternalpressures,andhouseholdsocioeconomicstatus.Theseresultsaretypicallybasedonsurveysofbothrecyclersandnon-recyclersandthereporteddifferencesbetweenthesetwogroups(ViningandEbreo,1990;Oskampetal.,1991;EbreoandVining,2001).

Asrecyclingrequiresinvestmentoftime,space,moneyandeffort,makingrecyclingconvenientshouldincreasehouseholdparticipation.DominaandKoch(2002)intheirstudyoftextilerecyclingbehaviorreportthatconvenienceisanimportantdriverofrecyclingbehavior.ViningandEbreo(1990)examinethedif-ferencesbetweenrecyclersandnon-recyclersandconcludethatnon-recyclersweredeterredbytheinconvenienceandthecostsassociatedwithrecycling.Saphoresetal.(2006)studyhouseholds’willingnesstorecycleelectronicwasteatdrop-offcentersandfindthatconveniencefactorssuchasproximitytothedrop-offcenterincreasedrecycling.Horniketal.(1995),basedonameta-analysis,concludethatfrequencyofrecyclablescollectionwasastrongpredictorofrecyclingbehavior.Gonzalez-Torreetal.(2003)exam-ineselectivewastecollectionsystemsthatarefrequentlyusedinEuropeandAmericaandconcludethatasystemthatrequireslesstimeandefforttodisposeandseparatewastewillresultinahigherrecyclingrate.

Concernfortheenvironmentisperceivedtobeimportantinencouragingrecyclingparticipation,butempiricalstudieshaveshownmixedresults.DominaandKoch(2002)findthatpeoplewhohavegreatconcernfortheenvironmentaremorelikelytorecycle.MenesesandPalacio(2005)studythedistributionofrecyclingtaskswithinthehousehold,andreportthathouseholdmemberswithpositiveattitudestowardsecologyandwhoaremotivatedtoprotecttheenvironmentsharedagreaterburdenoftherecycling.However,ViningandEbreo(1990)findthatconcernfortheenvironmentwasindiscriminatelyexpressedbybothrecyclersandnon-recyclers.Similarly,Oskampetal.(1991)didnotfindsignificantdifferencesbetweenrecyclersandnon-recyclersintheirgeneralpro-ecologicalattitudesandbeliefsintheseriousnessofenvironmentalproblems.

Knowledgeabouttheavailabilityofrecyclingprogramsandfacilitiesisnecessaryforeffectiveparticipationinrecycling.Stud-

ieshavefoundthatknowledgeaboutrecyclingprogramsisastrongpredictorofrecyclinginvolvement(GambaandOskamp,1994;Horniketal.,1995).ViningandEbreo(1990)findthatrecyclersweremoreawareofthepublicityassociatedwithrecyclingandmoreknowledgeableabouttherecyclingfacilitiesinthelocalarea.Otherstudieshavetriedtoanalyzetheroleofknowledgeabouttheenvi-ronmentinrecyclingbehavior.Oskampetal.(1991)reportthatthelevelofknowledgeaboutconservationisagoodpredictorofpartic-ipationinrecycling.Studieshavealsoinvestigatedtheeffectsocialinfluencehasonrecyclingbehavior.Socialinfluenceinthiscontextisdefinedasanindividual’sconcernabouttheperceptionofoth-ers,suchasfamilyandneighborsiftheydonotrecycle(ViningandEbreo,1990).Oskampetal.(1991)andDoValleetal.(2004)reportthatsocialinfluenceisanimportantdriverofrecyclingbehavior,butViningandEbreo(1990)donotfindsocialinfluencetobesignificantinexplainingrecyclingbehavior.

Apartfrombehavioralaspects,numerousstudieshavealsolookedattherelationshipbetweendemographicandsocioeco-nomicvariablesandrecyclinginvolvement.Themostcommonlyexaminedvariablesaregender,age,educationandincome(Saphoresetal.,2006).MenesesandPalacio(2005)arguethatwomenboreagreaterburdenofrecyclingtasksthanmeninahousehold,whileArcuryetal.(1987)suggestthatwomenareusu-allyassociatedwithrecyclingtasksbecausetheytraditionallyhaveplayedagreaterroleindomestictasks.Saphoresetal.(2006)findthatwomenaremorewillingtorecycleelectronicwasteatdrop-offcenters.However,otherstudiesfindnolinkbetweengenderandrecycling(GambaandOskamp,1994;WernerandMakela,1998).

Somestudiesfindagetobeasignificantfactorinfluencingrecy-clinginvolvement(ViningandEbreo,1990;GambaandOskamp,1994;Margai,1997;Scott,1999;Saphoresetal.,2006),whilesomeotherstudiesdonot(WernerandMakela,1998;MenesesandPalacio,2005).Contrarytocommonexpectationthatyoungerpeo-plearelikelytobemoreinvolvedinrecycling,someresearchersconcludethatmiddleagedandolderpeoplearemorelikelytorecy-cle(ViningandEbreo,1990;MenesesandPalacio,2005;Saphoresetal.,2006).

Therelationshipbetweeneducationandrecyclingisambiguous.Saphoresetal.(2006)findthathighereducationincreasesthewill-ingnesstorecycle,butseveralotherstudiesreportthateducationhasnosignificanteffectonrecyclingbehavior(ViningandEbreo,1990;Oskampetal.,1991;GambaandOskamp,1994;MenesesandPalacio,2005).Somestudiesfindapositiverelationshipbetweenincomelevelandrecyclinginvolvement(ViningandEbreo,1990;Oskampetal.,1991;GambaandOskamp,1994),butastudybyScott(1999)findsnostatisticallysignificantrelationship.3.Researchobjectivesandhypotheses

Theabovereviewsuggeststhattherearemixedfindingsaboutthedriversofrecyclingbehavior.Onelikelyreasonforsuchmixedresultsisthelackofcorrespondencebetweentheattitudinalandbehavioralentities.AjzenandFishbein(1977)con-cludefromareviewofanumberofattitude–behaviorstudiesthatwhenelements(i.e.target,action,contextandtime)oftheattitudinalentitycorrespondedtoelementsofthebehavioralentity,attitude–behaviorcorrelationswerequitehighandsig-nificant,andwhencorrespondencewaseitherpartialorlacking,attitude–behaviorcorrelationswereinsignificant.GeneralizingonAjzenandFishbein(1977)findings,weconjecturethatthesemixedfindingsaboutdriversofrecyclingbehaviorarebecauseoflackofcorrespondencebetweenthespecificelementsofrecyclingbehav-iorandpresumeddriversofsuchbehavior.Forexample,inmanyofthesestudies,measuresofattitudestowardstheenvironmentandenvironmentalknowledgewereverybroadandgeneral,whilethebehavioralentity(i.e.recyclingbehavior)beingmeasuredwas

S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170165

veryspecific.Inotherwordstherewasnocorrespondencebetweenthetargetandactionelementsoftheentities.Similarly,inhouse-holdsurveysofrecyclingbehavior,thetimeandcontextelementsoftherecyclingbehaviorarenotspecific;therespondentmaynotbetheactualrecycler,andreportedbehaviormaybedifferentfromactualbehaviordependingonthecontext.Furtherbecauseofthedifferentaction,contextandtimingelementsofdrop-offrecyclingcomparedtootherformsofrecycling,thefindingsfromotherrecy-clingbehaviorstudiesmaynotbedirectlyapplicabletodrop-offrecycling.

Themainobjectiveofthisstudyistoanalyzetheinfluenceofsocioeconomic,demographicandbehavioralfactorsondrop-offsitevisits.Thebehavioralaspectsexaminedareenvironmentalaffiliation,perceptionandattitudestowardsdrop-offrecyclingandimplicationsofsuchrecyclingontheenvironment.Thisstudyalsoanalyzestheeffectofdrop-offsitedistancefromhomeonsitevis-its.Toaddressthelimitationsofotherstudies,wespecificallytargetrecyclersrecyclingatdrop-offcenters.Ourmeasureofrecyclingbehavioristhenumberofsuchvisitstothedrop-offcenters,andourinterviewswereaimedatmeasuringattitudesspecificallytowardsrecyclingactivitiesandknowledgeaboutdrop-offrecycling,alongwithotherdemographicandsocioeconomicvariables.Byestablish-ingsuchacorrespondence,weexpectourresultstobemorerobustcomparedtoearlierfindings.

Drawingonthefindingsinearlierliterature,weproposethefollowinghypothesesandthentestthemempirically:

H1.Longerdistancetorecyclingsitesfromhomereducesthenumberofrecyclingvisits.

H2.Increaseinthenumberofdifferenttypesofrecyclablesbroughttoasiteincreasesthenumberofsitevisits.

H3.Highertimerequiredtosortrecyclablesreducesthenumberofsitevisits.

H4.Accesstocurbsiderecyclingreducesthenumberofsitevisits.H5.Demographicfactorssuchasage,gender,maritalstatus,edu-cationandemploymentstatusinfluencesthenumberofsitevisits.H6.Affiliationwithanenvironmentalorganizationincreasesthenumberofsitevisits.

Sinceattitudestowardsrecyclingandawarenessaboutrecy-clingoftherespondentscannotbedirectlyobserved,ascalewasdevelopedtomeasureattitudestowardsrecyclingconsistingof18Likert-scaleitemscoveringvariousaspectsofrecycling,suchasconvenience,attitude,socialpressureandawarenessorfamiliar-ity.TheassumptionisthatsuchaspecificrecyclingattitudescaleprovidesabettercorrespondencewithrecyclingbehaviorbeingmeasuredthancommonlyusedgeneralenvironmentalattitudemeasuressuchastheNewEcologicalParadigm(NEP)scale(Dunlapetal.,2000).Wehypothesizethatincreasedconvenience,posi-tiveattitudetowardsrecycling,positivesocialpressureandhigherfamiliaritywillbepositivelyassociatedwiththenumberofvisitstotherecyclingcenter.4.Methods

Sincewesoughttoanalyzetheeffectsofrecyclercharacteris-ticsonthenumberofdrop-offsitevisits,weconductedasurveyofdrop-offsitevisitors.Thissectiondescribesthesurveydesignanddatacollectionprocess.Thissectionalsoreportsthedescrip-tivestatisticsofthevariablesofinterest.Wealsoconductedfactoranalysistoreducethenumberofourattitudinalvariablesintoafewinterpretablefactorsthatwerelateroperationalizedasexplanatoryvariablesinourstatisticalmodelofdrop-offsitevisits.

4.1.Questionnairedesignanddatacollection

Thedataforthisstudywascollectedthroughin-personinter-viewsconductedateightdrop-offrecyclingsitesaroundtheLansingareainMichigan.Theinterviewsincludedquestionsonthefre-quencyofvisitstodrop-offsitesinthepastthreemonthsandoneyear.Respondents’homeaddresswaselicitedtoenablecalculationofrespondents’traveldistancestotherecyclingsite.Thesurveyalsocontainedquestionssolicitingdemographicinformationoftherespondentsuchasgender,education,employmentstatusincomeandmaritalstatus.Questionswereaskedabouttherespondents’otherrecyclingoptionsassuchcurbsiderecyclingattheirresidence,thetypesofrecyclablestheybroughtonsite,andthetimetheytooktosorttherecyclablestheyhadbroughtduringthevisit.Aquestionaskediftherespondentswereaffiliatedwithanyenvironmen-talorganization(governmentornon-governmentalenvironmentalorganizationsincluded).Thesurveyalsoincludedasetofquestionsassessingtherespondent’sattitudestowardsrecycling.Inanswer-ingthesequestions,respondentswerereadstatementsandaskedtoindicatetheextenttowhichtheyagreeordisagreewiththestate-mentsonafive-pointLikert-scalerangingfromstronglyagreetostronglydisagree.

Thequestionnairewaspretestedandimprovedbeforeconduct-ingtheactualsurvey.Thequestionnairepretestwasconductedbyinterviewingseveralrecyclersatoneofthedrop-offsites.Thepretestresultedinsomewordingrefinementsandrearrangementofquestionsintheinstrument.Thefinalsurveywasconductedforfourweeks,fromthelastweekofOctober2006tothelastweekofNovember2006.Interviewswereconductedateachsitefourtimes,during3hintervalseachtimethroughoutthefour-weekperiod.Thesurveydateschosenforallthesiteswererandomlyselectedtoavoidanypotentialbias.Duringthesurvey,recyclersvisitingthesiteswereapproachedforinterviews.Attheendofthesurvey,weapproached527recyclersandmanagedtocomplete356interviewsfora68%responserate.4.2.Variablesdescription

Table1listsanddefinesthedemographicandotherrelatedvariablesthatwereutilizedinouranalysis.ThevariablesTHREE-MTHSandONEYEARarethenumberofvisitstothedrop-offsitewheretherespondentwasinterviewedinthepastthreemonthsandoneyear.ThevariableDISTANCErepresentstheroundtripdis-tancefromtherespondent’shometotherecyclingsitewheretherespondentwasinterviewed.TheroundtripdistancewascomputedusingMapQuest(www.mapquest.com).ThevariableCURBSIDEisadummyvariableindicatingiftherespondentshadaccesstocurb-siderecyclingpickupattheirhome.Therecyclerscamefromseveraltownshipsandcurbsiderecyclingservicewasofferedonlyinsome

Table1

Definitionofvariables.VariableDefinition

THREEMTHSTotalnumberofsitevisitsinthelastthreemonthsONEYEARTotalnumberofsitevisitsinthelastoneyearDISTANCETotalround-tripdistancefromhometosite

NUMRECNumberofdifferenttypesofrecyclablesbroughtonsiteSORTIMETimetaken(inminutes)tosortrecyclablesbroughtCURBSIDEAccesstocurbsiderecycling(yes=1,no=0)

CDEGREEEducatedwithabachelor’sdegreeorhigher(yes=1,no=0)INCOMEAnnualhouseholdincome($1000’s)HSIZEHouseholdsizeAGEAge(years)

MALEMale(yes=1,no=0)MARRIEDMarried(yes=1,no=0)

FULLEMPEmployedfull-time(yes=1,no=0)

ENVAFF

Affiliatedwithanenvironmentalorganization(yes=1,no=0)

166

Table2

Summarystatisticsofvariables.VariableTHREEMTHSONEYEARDISTANCENUMRECSORTIMECURBSIDECDEGREEINCOMEHSIZEAGEMALEMARRIEDFULLEMPENVAFF

Obs.348348333348344345348348346345347348348346

S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170

Mean4.33014.65219.7126.32216.1660.2520.71877,9352.52048.20.5560.7040.10.263

SD3.45513.80410.2873.47427.3370.4350.45052.7911.26515.1810.4980.4570.4800.441

areasandforselectedmaterials.Similarly,somelargeapartmentbuildingsprovidedonsiterecyclingfacilities.Weuseaccesstocurb-siderecyclingasasummaryindicatorofavarietyofsuchservices.Ourinterviewssuggestedthatdrop-offcenterswereusuallyusedforrecyclingmaterialsthatwerenotcollectedatthecurbside.Hence,wehypothesizethatrespondentswithoutcurbsiderecy-clingserviceattheirresidenceswouldtendtorecyclemoreatdrop-offcentersthanrespondentswithcurbsiderecycling.

Thesummarystatisticsofthevariables(Table2)indicatethattheaveragenumberofvisitstoadrop-offsiteinthepastthreemonthsandoneyearwereapproximately4and15timesrespectively.Theaverageroundtripdistancetraveledbytherespondentstoadrop-Table3

Definition,distributionanddescriptivestatisticsofLikert-scalevariables.Variable

Surveystatement

offsitewasaround19miles.Therespondentsrecycled,onaverage,6differentmaterialseachtimetheyvisitedadrop-offrecyclingsite,andtheyspentapproximately16minsortingouttherecyclablesthattheybrought.Twenty-fivepercentoftherespondentsreportedthattheyhadcurbsiderecyclingserviceattheirresidence.

Themajorityoftherespondents(72%)hadatleastfouryearsofcollegeeducation.Sixty-fourpercentofourrespondentswerefullyemployedandthemeanhouseholdincomewas$77,935.Oursam-plecomprisedof56%malerespondentsindicatingroughgenderbalanceinrecyclingparticipation.Seventypercentoftherespon-dentsweremarried,andtheaveragehouseholdsizewas2.5.Only26%oftherespondentsindicatedthattheywereaffiliatedwithoneormoreenvironmentalorganizations.

Table3showsthestatementsthatwereusedinoursurveytoelicittherespondentsexperience,knowledgeandattitudetowardsrecyclingalongwiththerespectivedistributionofLikert-scaleresponsesanddescriptivestatistics.Thescalewasdefinedas(1)stronglyagree,(2)agree,(3)neitheragreenordisagree,(4)dis-agreeand(5)stronglydisagree.Basedonthemeanscorewecanseethatdrop-offrecyclersdisagreethatrecyclingisadifficulttask(M=4.174,SD=0.825).Theyalsodisagreetoboththestatementsofnothavingenoughsortingtime(M=4.285,SD=0.711)andstoragespace(M=3.797,SD=1.038).Therecyclersalsodisagreethatrecy-clablesstoredmayattractpests(M=4.026,SD=0.825).Mostoftherecyclersagreethattheyarefamiliarwiththerecyclingfacilities(M=1.947,SD=0.847)andthematerialsacceptedforrecyclinginthefacilitiesintheirarea(M=1.724,SD=0.595).

Therecyclersalsoagreethattheirfamilyexpectsthemtorecycle(M=2.312,M=1.012).However,therecyclersarequiteindiffer-

PercentagedistributionofresponseStronglyagree

Agree2.9%3.5%14.7%10.0%57.1%63.2%

Neitheragreenordisagree5.6%0.9%5.9%2.6%5.0%1.5%

Disagree.4%56.8%53.8%62.1%8.8%2.1%

Stronglydisagree35.0%38.2%22.9%25.3%0.3%0.0%

DescriptivestatisticsMean4.1744.2853.7974.0261.9471.724

SD0.8250.7111.0380.8250.8470.595

DIFFICTIMESPACEPESTFACILIMATERI

NEIGHBFRIENDFAMILYGOODREDPOLREDLANDNATRESENVQBREDPOLBREDLAND

BNATRESBENVQ

Forme,householdrecyclingisadifficulttask

Idonothaveenoughtimetosortthematerialsforrecycling

Idonothaveenoughspacetostorethematerialsforrecycling

TherecyclablesthatIstoreattractpestsIamfamiliarwiththerecyclingfacilitiesinmyarea

Iamfamiliarwiththematerials

acceptedforrecyclingintherecyclingfacilitiesinmyarea

Myneighborsexpectmetorecyclehouseholdmaterials

Myfriendsexpectmetorecyclehouseholdmaterials

Myfamilyexpectsmetorecyclehouseholdmaterials

IfeelgoodaboutmyselfwhenIrecycleRecyclingisamajorwaytoreducepollution

Recyclingisamajorwaytoreducewastefuluseoflandfills

Recyclingisamajorwaytoconservenaturalresources

Recyclingwillimproveenvironmentalquality

Ibelievethatmyrecyclingactivitieswillhelpreducepollution

Ibelievethatmyrecyclingactivitieswillhelpreducewastefuluseoflandfills.

IbelievethatmyrecyclingactivitieswillhelpconservenaturalresourcesIbelievethatmyrecyclingactivitieswillhelpimproveenvironmentalquality

2.1%0.6%2.6%0.0%28.8%33.2%

2.9%9.1%21.2%48.5%45.7%48.4%49.0%46.3%44.7%47.9%

7.4%23.5%46.2%47.1%48.7%49.9%47.5%51.3%49.7%50.0%

55.3%39.4%16.5%4.4%4.1%0.6%2.1%1.2%3.5%1.2%

25.3%22.9%12.6%0.0%1.2%0.9%1.2%0.9%1.5%0.6%

9.1%5.0%3.5%0.0%0.3%0.3%0.3%0.3%0.6%0.3%

3.3032.9122.3121.5591.6171.91.5631.5751.6351.553

0.8481.0121.0520.5790.6530.5910.6280.5980.6800.585

48.5%46.8%

47.9%50.9%

1.8%0.9%

1.5%0.9%

0.3%0.6%

1.5711.576

0.6360.622

S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170167

entonthestatementsonwhethertheirneighbors(M=3.303,SD=0.848)andfriends(M=2.912,SD=1.012)expectthemtorecy-cle.Nevertheless,mostoftherecyclersfeelgoodaboutthemselveswhentheyrecycle(M=1.559,SD=0.579).Themeanscoresalsoshowthattherecyclersstronglyfeelthatrecyclingisgener-allybeneficialtotheenvironment.Therecyclersstronglyagreethatrecyclingisamajorwaytoreducepollution(M=1.617,SD=0.653),toreducelandfilluse(M=1.9,SD=0.591),tocon-servenaturalresources(M=1.563,SD=0.628)andtoimproveenvironmentalquality(M=1.575,SD=0.598).Additionally,thesegeneralperceptionsonthebenefitsofrecyclingarestrength-enedbywhattherecyclersbelieveaboutthecontributionsoftheiractivities.Therecyclersstronglybelievethattheirrecyclingactivitieswillactuallycontributetoreducingpollution(M=1.635,SD=0.680),reducinglandfilluse(M=1.553,SD=0.585),conservenaturalresources(M=1.571,SD=0.636)andimproveenvironmen-talquality(M=1.576,SD=0.622).4.3.Factoranalysis

WeusefactoranalysiswithprincipalcomponentanalysistogrouptheLikert-scalevariablesintoasmallnumberofinter-pretableunderlyingfactors.Factoranalysiswillgroupthevariablesthataremeasuringthesameconstruct.Thismethodiscommonlyusedinsocialscienceresearch.WeusetheKaisereigenvaluecri-terionandthescreetest,assuggestedbyNunnallyandBernstein(1997),todecideonhowmanyfactorstoretainbeforeproceedingwithfurtheranalysis.Accordingtotheeigenvaluecriterion,factorswitheigenvaluesgreaterthanoneareretainedandfactorswitheigenvalueslessthanoneareconsideredinsignificantandthere-foreexcluded.Table4reportstheinitialfactorextractionwiththeeigenvaluesandpercentageofvariancesforeachsuccessivefactor.Usingtheeigenvaluecriterionmethod,fourfactorswereretainedforfurtheranalysis.

Thescreetest,ontheotherhand,isagraphicalmethodofdeter-miningthenumberofappropriatefactorstoretain.Thescreetestinvolvesplottingtheeigenvaluemagnitudesontheverticalaxisagainstthecomponentnumbersonthehorizontalaxisandnotingthepointatwhichtheplotbecomesfairlyhorizontal.Thenumberoffactorscorrespondingtothefairlyhorizontalpointindicatestheappropriatenumbertoretain.InFig.1,thepointwherethelinebecomesfairlyhorizontalstartsataboutfactor4.Thus,thescreetestindicatesthatweshouldalsoretainfourfactors,similartotheresultoftheeigenvaluecriterionmethod.

WealsoassessthesuitabilityofourdataforfactoranalysisusingtheKaiser–Mayer–Olkin’s(KMO)measureofsamplingadequacyandBartlett’stestofsphericity(Hairetal.,1998).Thesamplingadequacytestpredictsifdataarelikelytofactorwell,basedoncor-relationandpartialcorrelation,andaKMOmeasurevaluegreaterthan0.6isconsideredacceptable.Bartlett’stestofsphericityisusedtotestthenullhypothesisthatthevariablesinthepopulationcor-relationmatrixareuncorrelated.TheresultsarereportedinTable5,whichshowthatthedatameettherequirementsforfactoranalysis.

WeusetheVarimaxrotationmethod(Kaiser,1958)torotatethefourretainedfactorsinoursolution.TherotatedfactormatrixwithitsfactorloadingsispresentedinTable6.

Table4

Totalvarianceexplained.Factor

InitialeigenvaluesTotal

%ofvarianceCumulative%17.6662.59%42.59%22.0911.66%.25%31.67719.32%63.57%4

1.1681

6.49%

70.06%

Fig.1.Screeplotofeigenvaluesandfactors.

Table5

KMOandBartlett’stest.

Kaiser–Meyer–Olkinmeasureofsamplingadequacy0.841Bartlett’stestofsphericity

Approx.󰀃25075Df

153

Significance

p<0.001

Weconsidervariableswithloadingsgreaterthan0.4tobe‘highlyloaded’andsalienttotheinterpretationofafactor.Usingthiscriterion,thevariablesaregroupedtogetherintheappropriatefactorcategories(refertohighlightedcellsinTable6).Eachfactorisdescribedbasedonthesevariablesandassigneddescriptivenames.WealsocomputetheCronbach’scoefficientalphaforeachfactortotestthereliabilityofscalesoftheitemvariables.Whilethereisnostandardcut-offpointforthealphacoefficient,agenerallyaccept-ablelowerlimitis0.7(Hairetal.,1998)althoughvalueshigherthan0.5areacceptableinexploratoryresearch(Nunnally,1978).Wethenusethefactorloadingstocomputethefactorscores.

Thevariablesthatloadhighlyonfactor1are,GOOD,REDPOL,REDLAND,NATRES,ENVQ,BREDLAND,BNATRESandBENVQ.Thisfac-torislabeledas“Attitude”andcanbebestdescribedasattitudeandbeliefthatrecyclingactivitiesbenefittheenvironment.Alowscoreforthisfactorindicatesthattherespondentshavepositiveattitudeandbeliefthattheirrecyclingactivitiesleadtoenvironmentalben-efitssuchasreducedpollutionandlandfilluse,conservingnatural

Table6

Rotatedfactormatrix.VariableFactor1Factor2Factor3Factor4DIFFIC−0.1090.811−0.138−0.115TIME−0.2050.842−0.002−0.036SPACE−0.1450.679−0.118−0.015PEST−0.2310.4100.104−0.286FACILI0.168−0.0030.0270.853MATERI0.224−0.1980.1290.747NEIGHB0.042−0.0050.8200.086FRIEND0.126−0.0920.7990.020FAMILY0.112−0.1220.7200.042GOOD0.482−0.351−0.0410.190REDPOL0.873−0.0500.0670.105REDLAND0.3−0.1060.0150.101NATRES0.927−0.1120.0070.0ENVQ0.877−0.1130.0480.116BREDPOL0.877−0.0650.1030.057BREDLAND0.917−0.1300.0740.108BNATRES0.924−0.0750.0160.079BENVQ

0.5

−0.144

0.106

0.0

168S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170

resourcesandimprovingenvironmentalquality.TheCronbachs’salphafortheseitemsis0.96.

Theitemvariableswithhighloadingsonfactor2areDIFFIC,TIME,SPACEandPEST.Welabelthisfactoras‘Convenience’asitrelatestorecyclingbeingaconvenientactivitytoundertake.Ahighscoreforthisfactorsignifiesthattherespondentsregardrecyclingassomethingthatisconvenienttothemastheyhavenoissuewithitbeingdifficult,timeconsuming,spaceconsuming,orattractingpests.TheCronbach’salphaforthesevariableitemsis0.69.

Welabelfactor3as‘SocialPressure’becausethevariablesthatloadhighlyonfactor3concernsocialpressuresontherecycler.ThevariablesinthisfactorareNEIGHB,FRIENDandFAMILY.Alowscoreforthisfactorindicatesthattherespondentfeelsthatneigh-bors,friendsandfamilyexpectationsareimportantelementsinencouraginghim/hertorecycle.TheCronbach’salphaforfactor3is0.7015.

Thevariablesthatloadhighlyonfactor4areFACILIandMATERI.Welabelthisfactoras‘Familiar’asitrelatestothefamiliarityofrecyclingfacilities.Alowscoreforthisfactordemonstratesthatrespondentsarehighlyfamiliarwiththerecyclingfacilitiesandthematerialsacceptedintherecyclingfacilitiesintheirarea.TheCronbachs’salphaforfactor4is0.579.

Table7outlinesthefactors,theirrespectivevariablesasextractedbythefactoranalysis,andtheirCronbach’salphacoeffi-cient.Thefactorscondensetheexperience,knowledgeandattitudetowardsrecyclingofoursurveyrespondentsintofournewinter-pretablevariables;namelyAttitude,Convenience,SocialPressureandFamiliarity.Weusethefactorscoresofthesefournewvari-ablesandanalyzetheeffectsoftheseonthenumberofrecyclingvisitstodrop-offcenters.

5.Estimationmodelandresults

Themainpurposeofthearticleistoanalyzethevariablesthatinfluencevisitstodrop-offrecyclingsites.Thissectiondevelopsavisitationmodeltoanalyzetheeffectsofdemographics,environ-mentalaffiliation,andtheattitudeandknowledgevariablesderivedfromthefactoranalysis,onthenumberoftripstakentoadrop-offsite.ThevisitationmodelisdevelopedusingthePoissonregres-sionmethod.Poissonregressionisutilizedbecausethedataforourdependentvariable,thetripsanindividualtakestoarecyclingsiteyi,isclassifiedasacountvariablewhereyicanonlytakediscretevalues(yi=1,2,3,...).Morespecifically,wewillusetheendoge-nousstratifiedandtruncatedPoissonregressionsincewedonotobservezerotripsforanyofthesamplemembersasoursampleisobtainedviatheonsitesamplingmethod.FollowingHaabandMcConnell(2002)thePoissonprobabilitywithonsiteendogeneityandtruncationisexpressedasfollows:

Pr(y󰀂e−󰀂i󰀂yi󰀂yi>0)=

yii−1i−1!(1)

where󰀂iisboththemeanandthevarianceofthedistribution.Sinceitisnecessaryfor󰀂i>0,itiscommonlyspecifiedasanexponentialfunction:󰀂i=exp(xiˇ)

(2)

Table7

Factors,itemvariablesandCronbach’salpha.FactorItemvariables

Cronbach’s␣(1)AttitudeGOOD,REDPOL,REDLAND,NATRES,0.960ENVQ,BREDLAND,BNATRES,BENVQ(2)ConvenienceDIFFIC,TIME,SPACE,PEST0.696(3)SocialpressureNEIGHB,FRIEND,FAMILY0.702(4)Familiarity

FACILI,MATERI

0.579

wherexiisavectorofexplanatoryvariables.Eq.(1)canbesimplified

byre-writingitas:

y󰀂

Pr(y󰀂0)=e−󰀂i󰀂i󰀂y󰀂iiyi>(3)

i!

wherey󰀂i=yi

−1.UsingEq.(3),weobtainthelog-likelihoodofaPoissonfunction

ln󰀄ˇ󰀂󰀂X,y󰀁T

=

󰀃

[−eXiˇ+Xiˇyi−ln(y󰀂i

!)](4)

i=1

andthus,theonsiteendogenousandtruncatedPossioncanbeesti-matedbysimplyrunningastandardPoissonregressionofyi−1on

allXi’s.

Thetimeinvestedinsortingtherecyclablespriortovisitingadrop-offcenterisahouseholddecisionwhichmayaffectthenumberofvisitsandinturnmaybeaffectedbythenumberofvisits,whichgivesrisetoapotentialendogenietyprobleminourestimationmodel.Weconducttheomitted-variableversionoftheHausmantesttoexaminewhetherthereisanendogeneitybiasinourestimatesusingatwo-stageinstrumentalvariablemethod.Ourinstrumentalvariablewasthefittedvaluesobtainedbyregress-ingthelikelyendogenousvariable,SORTTIME,asafunctionofalltheotherindependentvariablesintheregressionsandadditionaldummyvariablesformaterialsrecycled.Ourtestsfailedtorejectthehypothesisofnoendogenietybias.HencewereporttheresultscorrectedforendoegenietywiththeinstrumentalvariableforSORT-TIMEinTable8.

Table8presentstheresultsofthePoissonregressionmodelspredictingthenumberoftripstakentoarecyclingdrop-offsiteinthelastoneyear.Therearetwomodelsinthisanalysis.Model1isthebasicmodelthatusesdistance,numberofrecyclables,sort-ingtime,accesstocurbsiderecyclinganddemographicvariablesasdependentvariables.Model2istheextendedmodelthatincludesallfourLikert-scalevariablesderivedfromthepreviousfactoranal-ysisalongwithallofthebasicvariablesinModel1.Thepurposeoftwomodelsistoseetheincrementaleffectsofattitudeandfamiliar-Table8

Poissonregression.

Dependentvariable:ONEYEARi.e.numberofvisitsinthepastyearVariable

Model1Model2

Coeff.

Std.errorCoeff.Std.errorDISTANCE−0.010.001***−0.010.002***NUMREC0.0350.007***0.0360.007***SORTIME−0.0390.011***−0.0370.012***CURBSIDE0.0130.046−0.0130.047CDEGREE−0.0250.045−0.0040.047INCOME0.0020.001***0.0020.001***HSIZE0.0310.018*0.0480.019***AGE0.010.001***0.0080.001***MALE0.030.0320.0590.033*MARRIED0.0860.042**0.0180.044FULLEMP−0.3030.038***−0.3310.04***ENVAFF

0.114

0.038***

0.0360.041CONVENIENCE0.0460.016***FAMILIAR−0.1480.017***SOCIAL−0.0840.016***ATTITUDE0.0210.017CONSTANT2.4830.146***

2.5610.152***

Observations329

322

Log-likelihood−2343.06−2195.48−2ln(LR/LU)424.33516.83PseudoR2

0.08

0.11

*Statisticallysignificantatthe10%level.**Statisticallysignificantatthe5%level.***

Statisticallysignificantatthe1%level.

S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170169

ityvariablesoverdemographicanddistancevariables.TheresultsshowthatbothModels1and2arestatisticallysignificantwithlikelihoodratioteststatisticsof424.33and516.83respectively.

AllthecoefficientsinModel1arestatisticallysignificantat5%exceptforCURBSIDE,CDGREEandMALE,indicatingthataccesstocurbsiderecyclingservices,genderandeducationleveldonothavestatisticallysignificanteffectsontheexpectednumberofsitevisits.ThecoefficientsonNUMREC,INCOME,HSIZE,AGEandMAR-RIEDinModel1arepositiveindicatingthatincreasednumberofrecyclables,householdsizeandincomeincreasethenumberofvis-its.ThepositivecoefficientforENVAFFindicatesthataffiliationtoanenvironmentalorganizationincreasestheexpectednumberofsitevisits.ThecoefficientsonDISTANCE,SORTIME,andFULLEMParenegative.ThesignificancelevelofthecoefficientsinModel2afteraddingthefourattitudinalvariablesdonotchangeexceptforENVAFF,whichisnolongersignificant,indicatingthatenvi-ronmentalaffiliationwascapturingtheeffectsoftheattitudinalvariables.ThecoefficientsonNUMREC,INCOME,HSIZEandAGEintheextendedmodelremainpositive.ThecoefficientsonDISTANCE,SORTIME,CDEGREEandFULLEMPremainnegativeandstatisticallysignificant.Threeofthefourattitudinalvariables:CONVENIENCE,FAMILIARandSOCIALaresignificantatthe1%level.FAMILIARandSOCIALhavenegativesignsandCONVENIENCEhasapositivesign.

ThecoefficientsonDISTANCEimplythattheexpectednum-berofvisitsreducesby1%asroundtripdistancefromhometositeincreasesbyamile.Thisresultconfirmsearlierfindings(e.g.Saphoresetal.,2006)thatproximitytorecyclingsitesencouragesrecyclingbehavior.ThecoefficientonNUMRECindicatesthatthenumberofsitevisitsisexpectedtoincreasewhenarecyclerrecyclesalargervarietyofrecyclables.Thetimetakentosorttherecyclablesathomewasfoundtoreducetheexpectednumberofsitevisits.TheSORTTIMEcoefficientinModel2indicatesthata1minincreaseinsortingtimereducestheexpectednumberofsitevisitsby3.7%.Sur-prisingly,thecoefficientforCURBSIDEinModel2suggeststhattheavailabilityofcurbsiderecyclingdoesnotsignificantlyaffectdrop-offrecycling.Ourinterviewssuggestedthatthiswasmainlybecauseofthelimitednumberofmaterialsacceptedincurbsideprograms,e.g.mostofthecurbsideprogramsdidnotacceptcoloredglassorcardboard,whichwereacceptedatthedrop-offcenters.

ThenegativecoefficientsonFULLEMPindicatethatpeoplewhoareemployedfull-timearelikelytospendlesstimeonrecyclingactivitieswhencomparedtopeoplewhoareemployedparttimeorunemployed.Anincreaseinannualhouseholdincomeandhouse-holdsizeincreasethenumberofsitevisits.Thepositiverelationshipisasanticipatedbecauselargerandricherhouseholdstendtocon-sumemoregoods.ThisresultalsoconfirmsthefindingsbyViningandEbreo(1990),Oskampetal.(1991),andGambaandOskamp(1994).Thepositiverelationshipbetweenageandnumberofsitevisitsisalsoconsistentwithpreviousfindingsthatolderpeoplehaveahighertendencytorecycle(ViningandEbreo,1990;MenesesandPalacio,2005;Saphoresetal.,2006).

ThepositivecoefficientofthevariableCONVENIENCEinModel2indicatesthatthenumberofexpectedsitevisitsincreaseswhenrecyclingisregardedasaconvenientactivity.Thisresultconfirmsthepreviousfindingsthatconvenienceisanimportantfactorthatencouragesrecyclingbehavior(ViningandEbreo,1990;Horniketal.,1995;DominaandKoch,2002;Gonzalez-Torreetal.,2003;Saphoresetal.,2006).ThecoefficientforFAMILIARsuggeststhatpeoplewhoaremorefamiliarwithlocationsandmaterialsacceptedatthedrop-offcenterinhisorhervicinityareexpectedtomakemorevisitstothecentersthanthelessfamiliarpeople.Thecoeffi-cientforSOCIALimpliesthatpressurefrompeersandfamilyhasapositiveeffectondrop-offsitevisits.Thisresultconformstothefindingsofpreviousstudiesthatindicatesocialpressureisanimportantfactormotivatingrecyclingbehavior(Oskampetal.,1991;DoValleetal.,2004).

Wealsoransimilarregressionswithnumberofvisitsduringthepreviousthreemonthsasthedependentvariable.Thefitoftheover-allregressionswasweaker,mainlyonaccountoflowervariationinthedependentvariablewhileallexplanatoryvariablesremainedthesameastheone-yearvisitmodel.Allthecoefficientestimateshadsimilarsignsandsimilarvaluesastheone-yearmodel.How-ever,onlycoefficientsonDISTANCE,NUMREC,FULLEMP,FAMILIARandSOCIALcontinuedtobestatisticallysignificant.6.Conclusions

Despitetherelativepopularityofdrop-offrecyclingamonglocalgovernmentsandpolicymakers,littleresearchhasempiricallyexamineddriversofdrop-offrecycling.Thisstudyhelpstounder-standtheprofileofpeoplewhoutilizedrop-offrecyclingsitesaswelltheunderlyingfactorsthatinfluencetheirfrequencyofuse.Anotherstrengthofthestudyisthatitconsiderseconomic(e.g.traveldistance,sortingtime,income),demographic(e.g.age,gender)andpsychological(e.g.attitude,knowledge)driversofrecyclingbehaviorinthespecificcontextofdrop-offrecycling.Incomparison,asdiscussed,manyoftheearlierstudieslackclosecor-respondencetodrop-offrecycling.Hencethefindingsofthisstudyarelikelytobemorereliableforunderstandingdrop-offrecycling.

Thestudyresultssuggestthatlocationplaysacrucialroleininfluencingtheusagepatternofdrop-offsites.Recyclersarelikelytouseadrop-offsitemorefrequentlyifthetraveldistancefromhometositeisshorter.Thus,thedecisiontoestablishadrop-offrecyclingprogramshouldfactorinlocationtoencourageitsuse.Ourresultssuggestthatsocioeconomicvariablessuchashouse-holdsizeandincome,whicharelikelyhighlycorrelatedwithhouseholdconsumption(andhencewastegeneration),aregoodpredictorsofrecyclingbehavior,comparedtogenderandmari-talstatus.Therefore,locatingdrop-offrecyclingcentersconvenienttohigherincome,olderneighborhoodsislikelytoleadtohigherutilization.

Theresultsindicatethatbeliefsaboutrecyclingconvenience,familiaritywithrecyclinginfrastructureandsocialpressurearesignificantdriversofrecyclingbehavior.Recyclerstendtousethedrop-offsitesmorewhentheyfeelthatrecyclingisaconve-nientactivityandarefamiliarwiththeavailablerecyclingfacilities.Hencecommunicationandeducationeffortsaimedatimprovingawarenessofrecyclingfacilitiesandrecyclingconveniencecanbeeffectiveinpromotingvisitstorecyclingcenters.Ourresultsforsocialpressureareconsistentwiththegrowingbodyofresearchthatshowshowleveragingsocialnormscanenhanceconservation(Schultz,1999;Chenetal.,2009).Interestingly,beliefsabouttheenvironmentaleffectsofrecyclingwerenotsignificantlyrelatedtothenumberofdrop-offtrips.Assuch,ourfindingsareconsistentwithresearchshowingthatcommunicationappealsbasedonenvi-ronmentalprotectionarelesseffectivethatappealsthatleveragesocialnorms(Goldsteinetal.,2008).Thus,theresultssuggestthatsuchpromotioneffortsaimedatchildrenandthecommunityingeneralcanalsoindirectlyincreaserecyclingbyincreasingsocialpressure.References

AadlandD,CaplanAJ.Householdvaluationofcurbsiderecycling.JournalofEnviron-mentalPlanningandManagement1999;42:781–99.

AadlandD,CaplanAJ.Willingnesstopayforcurbsiderecyclingwithdetectionand

mitigationofhypotheticalbias.AmericanJournalofAgriculturalEconomics2003;85:492–502.

AjzenI,FishbeinM.Attitude–behaviorrelations:atheoreticalanalysisandreview

ofempiricalresearch.PsychologicalBulletin1977;84:888–918.

ArcuryT,ScollayS,JohnsonT.Sexdifferencesinenvironmentalconcernandknowl-edge:thecaseofacidrain.SexRoles1987;16:463–73.

BlaineTW,LichtkopplerFR,JonesKR,ZondagRH.Anassessmentofhouse-holdwillingnesstopayforcurbsiderecycling:acomparisonofpayment

170S.F.Sidiqueetal./Resources,ConservationandRecycling (2010) 163–170

cardandreferendumapproaches.JournalofEnvironmentalManagement2005;76(1):15–22.

ChenX,LupiF,HeG,LiuJ.Linkingsocialnormstoefficientconservationinvestment

inpaymentsforecosystemservices.ProceedingsoftheNationalAcademyofSciences2009;106:11812–7.

DominaT,KochK.Convenienceandfrequencyofrecycling:implicationsfor

includingtextilesincurbsiderecyclingprograms.EnvironmentandBehavior2002;34:216–38.

DoVallePO,ElizabethR,MenezesJ,RebeloE.Behavioraldeterminantsofhouse-holdrecyclingparticipation:thePortuguesecase.EnvironmentandBehavior2004;36:505–40.

DunlapRE,VanLiereKD,MertigAG,JonesRE.Measuringendorsementof

thenewecologicalparadigm:arevisedNEPscale.JournalofSocialIssues2000;56(3):425–42.

EbreoA,ViningJ.Howsimilararerecyclingandwastereduction?Futureorien-tationandreasonsforreducingwasteaspredictorsofself-reportedbehavior.EnvironmentandBehavior2001;33:424–48.

FullertonD,KinnamanTC.Householdresponsestopricinggarbagebythebag.The

AmericanEconomicReview1996;86:971–84.

GambaRJ,OskampS.Factorsinfluencingcommunityresidents’participation

incommingledcurbsiderecyclingprograms.EnvironmentandBehavior1994;26:587–612.

GoldsteinNJ,CialdiniRB,GriskeviciusV.Aroomwithaviewpoint:usingnormative

appealstomotivateenvironmentalconservationinahotelsetting.JournalofConsumerResearch2008;35:472–82.

Gonzalez-TorrePL,Adenso-DıazB,Ruiz-TorresA.Somecomparativefactorsregard-ingrecyclingcollectionsystemsinregionsoftheUSAandEurope.JournalofEnvironmentalManagement2003;69:129–38.

HaabTC,McConnellKE.Valuingenvironmentalandnaturalresources:theecono-metricsofnon-marketvaluation.EdwardElgarPublishing;2002.

HairJF,AndersonRE,TathamRL,BlackWC.Multivariatedataanalysis.5thed.Upper

SaddleRiver,NJ:PrenticeHall;1998.

HongS.Theeffectsofunitpricingsystemuponhouseholdsolidwastemanagement:

theKoreanexperience.JournalofEnvironmentalManagement1999;57:1–10.HongS,AdamsRM.Householdresponsestopriceincentivesforrecycling:some

furtherevidence.LandEconomics1999;75:505–14.

HornikJ,CherianJ,MadanskyM,NarayanaC.Determinantsofrecyclingbehavior:a

synthesisofresearchresults.JournalofSocio-Economics1995;24:105–27.

JenkinsRR,MartinezSA,PalmerK,PoldolskyMJ.Thedeterminantsofhouse-holdrecycling:amaterial-specificanalysisofrecyclingprogramfeaturesandunitpricing.JournalofEnvironmentalEconomicsandManagement2003;45:294–318.

KaiserHF.Thevarimaxcriterionforanalyticrotationinfactoranalysis.Psychome-trika1958;23:187–200.

KinnamanTC,FullertonD.Garbageandrecyclingwithendogenouslocalpolicy.

JournalofUrbanEconomics2000;48:419–42.

LakeIR,BatemanIJ,ParfittJP.Assessingakerbsiderecyclingscheme:aquantita-tiveandwillingnesstopaycasestudy.JournalofEnvironmentalManagement1996;46:239–.

MargaiF.Analyzingchangesinwastereductionbehaviorinalow-incomeurban

communityfollowingapublicoutreachprogram.EnvironmentandBehavior1997;29:769–92.

MenesesGD,PalacioAB.Recyclingbehavior:amultidimensionalapproach.Environ-mentandBehavior2005;37:837–60.

NunnallyJC.Psychometricmethods.3rded.NewYork,NY:McGraw-Hill;1978.

NunnallyJC,BernsteinIH.Psychometrictheory.3rded.NewYork,NY:McGraw-Hill;

1997.

OskampS,HarringtonMJ,EdwardsTC,SherwoodDL,OkudaSM,SwansonDC.

Factorsinfluencinghouseholdrecyclingbehavior.EnvironmentandBehavior1991;23:494–519.

SaphoresJM,NixonH,OgunseitanOA,ShapiroAA.Householdwillingnesstorecy-cleelectronicwaste:anapplicationtoCalifornia.EnvironmentandBehavior2006;38:183–208.

ScottD.Equalopportunity,unequalresults:determinantsofhouseholdrecycling

intensity.EnvironmentandBehavior1999;31:267–90.

SchultzPW.Changingbehaviorwithnormativefeedbackinterventions:afield

experimentoncurbsiderecycling.BasicandAppliedSocialPsychology1999;21:25–36.

SnyderKC,KristelOV,DhmammarungruangB,SangS.ReporttotheOhioEnviron-mentalProtectionAgency:drop-offrecycling—understandingparticipationanddetermininganempiricallybasedaccesscreditmodel.July14,2004.

SpeirsD,TuckerP.Aprofileofrecyclersmakingspecialtripstorecycle.Journalof

EnvironmentalManagement2001;62:201–20.

TillerKH,JakusPM,ParkWM.Householdwillingnesstopayfordrop-offrecycling.

JournalofAgriculturalandResourceEconomics1997;22:310–20.

USEPA,Municipalsolidwastegeneration,recycling,anddisposalintheUnited

States:factsandfiguresfor2000.

USEPA.Municipalsolidwastegeneration,recycling,anddisposalintheUnited

States:factsandfiguresfor2007.

VanHoutvenGL,MorrisGE.Householdbehaviorunderalternativepay-as-you-throwsystemsforsolidwastedisposal.LandEconomics1999;75:515–37.

ViningJ,EbreoA.Whatmakesarecycler?Acomparisonofrecyclersandnonrecy-clers.EnvironmentandBehavior1990;22:55–73.

WernerC,MakelaE.Motivationsandbehaviorsthatsupportrecycling.Journalof

EnvironmentalPsychology1998;18:373–86.