choice valuation of traffic restrictions：noise pollution and congestion preferences

TransportationResearchPartDjournalhomepage:www.elsevier.com/locate/trdNotesandcomments

Choicevaluationoftrafficrestrictions:Noise,pollution,andcongestionpreferences.Anote

CarlosPestanaBarrosa,PeterU.C.Diekeb,*abInstitutoSuperiordeEconomiaeGestão,TechnicalUniversityofLisbon,RuaMiguelLupi,20,1249-078Lisbon,PortugalSchoolofRecreation,Health,andTourism,GeorgeMasonUniversityManassas,Virginia20110,USAarticle

Keywords:PollutionLisbonMixedlogitmodelPublicpolicyinfoabstract

ThisNotelooksatthechoicevaluationrestrictionsontrafficenteringthecityofLisbon,basedonindividualpreferencesinrelationtonoise,pollutionandcongestion.Theanalysisemploysaquestionnairedistributedin2007toascertainthesignificantcharacteristicsoftravelingtoLisbon,withtheaimofcurbingthenumberofcarsthatareassociatedwiththeprobabilityofindividualssupportingachargeonmotorvehiclesenteringthecity.Themodelalsotakesintoaccounttheuncontrolledheterogeneityofthedata.PublishedbyElsevierLtd.1.IntroductionVariouscitieshaveintroducedacongestionchargeinanattempttoreducethenumberofmotorvehiclesenteringcity-centers;Singapore(1976),Rome(1998andextendedin2001),Durham(2002),London(2003andextendedin2007),Stock-holm(2007),andValletta(mid-1990s).ItisalsounderconsiderationatmanymorecitiesincludingLisbon.HerewelookatthebehavioralprocessthatleadsGreaterLisboncitizenstoevaluatethenegativeexternalitiesoftrafficandsupporttheimpositionofachargeontrafficenteringthecity.2.ContextLisbonisanoldcity,withmanybuildingsinapoorconditionbecauseofarentpolicythatformorethan50yearslimitedrentincreasesleadingtoownersneglectingmaintenanceandrefurbishment.Asaresultmanyfamiliesrelocatedtothesub-urbsandcommuteintothecityforwork.Theresulthasbeenhighlevelsofnoise,pollutionandcongestion,inadditiontosevereparkingproblemsincentralLisbon.Whilenopolicytorestrictthetraffichasformerlybeenannounced,in2007,themunicipalcouncilimplementedamorerigorousenforcementofcity-centerspeedlimitsusingradarandclosedcircuittele-visiononprincipalarteries.Thisdoeslittle,however,toreducecongestion,pollutionandnoise,becauseitfailstoaddressthenumberofvehicles.Themediahasassumedtheroleofraisingawarenessoftheproblemfocusingonpoliciestotransfercommutingbycarwithbusortrainjourneys.WhereasLisbonisreasonablywellservedbyitspublicly-ownedbusandmetronetworks,congestionisfuelledbythedailyinfluxfromtheoutersuburbs,principallyalongfourmajorhighways,onesecondaryhighwayandonecoastaldual-car-riageway.AlthoughthelocalrailwaynetworksofferservicesfromfourdirectionsintoLisbon,threeofthesepre-datingtherapiddevelopmentofthesatellitesuburbanareas.*Correspondingauthor.Tel.:+17039934260;fax:+17039932025.E-mailaddress:pdieke@gmu.edu(P.U.C.Dieke).1361-9209/$-seefrontmatterPublishedbyElsevierLtd.doi:10.1016/j.trd.2008.03.006348C.P.Barros,P.U.C.Dieke/TransportationResearchPartD13(2008)347–350Moreover,theriverTagusisamajortransportbarrier.Lisbonstandsonitsnorthside,whilemanythousandsofworkerslivetothesouthinfarcheaperdistrictsornewtowns.Untiltheinaugurationin1998ofasecondmajorbridgesomedistanceupriverandtheopeningin1999ofaprivately-ownedraillinkbeneaththeonlyroadbridgeintocentralLisbon,workersfromawideradiustothesouthwereobligedtocrosstheriverbyroadand/orferry.InPortugal,chargesareleviedonlyoninter-urbanhighways.Thus,thecountryhasnoexperiencewithurbancongestioncharging.Manycitizens,however,areawareofLondon’sschemeandvariouspoliticianshaveexpressingsupportofasimilarpolicyforLisbonandthesecondcity,Porto.3.ResearchdesignWeconsiderindividualswhotraveltoLisbonbycaranddeclarethatsupportforrestrictionsonprivatecarsenteringthecity.Thereareanumberofwaystolookatthefactorsinfluencingthissupport:󰀂Individualcommuterswhochoosetosupportthechargearethosewhoareattractedbychoiceattributesrelatedtopol-lution,noiseandcongestion.Thisisatraditionalhypothesisintransportationdemandmodels(HensherandPuckett,2007;Safirovaetal.,2007).󰀂Acommuterwhochoosestosupportthechargeischaracterizedbysocio-economiccharacteristicsregarding,age,income,levelofeducation,socialclass,marriedandwithchildren(SrinivasanandAthuru,2005).󰀂Travelerswhochoosetosupportthechargearethosewhohavebeenadequatelyinformed.Perceptionformationderivesfrominformationthathelpstravelersclarifyandevaluatetraveldecisions.󰀂Travelerswhochoosetosupportthechargearethosewhohavetemporalconstraintsandthuslesstimetowanttospendontravel(Espinoetal.,2006).Toexaminethesealternatives,amixedlogitrepresentationisusedthatassumestheprobabilityofsupportingthecharge,canbedescribedbyalogit-probabilityfunctionoftheexogenousvariables(Train,2003).Theprobabilityforchoiceisas:Zþ1PrðChoiceijviÞ¼Pðb;viÞNðb5jl5;r5Þdb5ð1ÞÀ1whereN(󰀂)isthenormaldistribution,andvi¼b0þb1Pollutionþb2Noiseþb3Congestionþb4Incomeþb5Civilstateþb6Classþþb7Educationþb8Groupþb9Newspaperþb10TVþb11TreserveþeiVariablesaredeterminedusingaquestionnaire.Variableviismeasuredbytheprobabilitythatanindividualdeclaresthathesupportsthechargetakingavalueofunityforapositiveresponse,andmeasureXiasobservedcharacteristics.4.EmpiricalanalysisTogatherinformation,thequestionnairewasadministeredtoastratified,randomsampletravelersenteringLisbondur-ingthemorningrush-hour.1Thesamplewasstratifiedbytransportmode(car,bus,trainandmetro),usingthetravelerdata-baseavailableintheNationalTrafficInstitute.Threethousandquestionnaireswereadministered.Theinterviewersapproachedthepassengeronthechosenmodes,usingarandomprocedure,inwhichthequestionnairewasdistributedbyhand,togetherwithastampedenvelopetoreturnthecompletedquestionnaire.Thisprocedureresultedinthereturnbymailof1652com-pletedquestionnaires(Dillman,1978).

Therateresponsedoesnotdiffersignificantlyamongthepopulationbyagevariable(v2=8.53,p=0.05)orgender(v2=7.55,p=0.05)indicationtherespondentsarerepresentativeofLisbontravelers.Thegeneralcharacteristicsofrespon-dentsweremale(52%),withanaverageageof45andeducatedtocollegedegreelevel;formoredetailsseeTable1.Thesurveycomprisedthreesections:thefirst,characterizesthecommutersinrelationtotheirsocio-economicprofileandincludedvariablessuchassocialclass,age,gender,familyandeducation;thesecondincludesthetripographicvariablesandconstraintssuchasbudget,travelfrequency,andtemporalconstraints;andthethird,theirpreferenceswithregardtonoise,pollutionandcongestion.Aseven-pointLikert-typescaleassessesthepolicyattributesrangingfrom‘‘withoutimpor-tance”(1)to‘‘veryimportant”(7).Anopen-endedquestionwasaskedonthewillingness-to-payacongestionchargethatrangedfromzeroto€10,withanaveragevalueof€5andastandarddeviationof€4.8.Table1showsthesignificantvariables.Themixedlogitresultsdisplayevidenceofunobservedheterogeneity,dependingonincomeandeducation,commontothethreesegmentsanalyzed.Thesurveyalsoshowsheterogeneityintheeffectofnoiseonvehicledrivers,inducingtheesti-mationofamixedlogitmodelwithrandomcoefficientsforthesevariables.ThefinalresultsareseeninTable2.Theoverallfitofthismodelisreasonablygood,withaChi-squarestatisticsquarevalueof19.5with41degreesoffreedomandlevelofsignificanceof0.0067.Theothermodelsdisplaysimilarvalues.Thevariablesare,forthemostpart,sta-Thequestionnairewaspre-testedonstudentsofeconomicsattheTechnicalUniversityofLisbonandfollowingtheadministrationofthefinalsurvey,astratifiedrandomsub-setof50respondentswerecontactedbyphoneseeifanyproblemshadarisenincompletingthequestionnaire;noneweredetected.

1C.P.Barros,P.U.C.Dieke/TransportationResearchPartD13(2008)347–350Table1

CharacterizationofthevariablesVariableChargePollutionNoiseCongestionDescriptionDoyoufavorachargetorestrictcarsfromenteringthecity:yes=1,No=0DestinationattributeshypothesisWhatwastheimportanceofpollutioninyourdecision?(1-withoutimportance;7-extremelyimportant)Whatwastheimportanceofnoiseinyourdecision?(1-withoutimportance;7-extremelyimportant)Whatwastheimportanceofcongestioninyourdecision?(1-withoutimportance;7-extremelyimportant)Socio-demographiccharacteristicshypothesisCommutermonthlyincome(1-lowerthan1000€;5-equaltoorhigherthan5000€)Maritalstatus(1-single;2-married;3-withchildren)Socialclass(1-lower;2-middle;3-upper-middle)Education(numberofyearsofeducation)(4-primaryschool,24-Doctorate)TravelingroupsInformationhypothesisWhatwastheimportanceofinformationobtainedinnewspapersinyourdecision?(1-withoutimportance;7-extremelyimportant)WhatwastheimportanceofTVinformationinyourdecision?(1-withoutimportance;7-extremelyimportant)TemporalconstraintshypothesisHowlongisyourjourney?(1-lessthan15min;2–15minormore,butlessthan30min;3-morethan30min;4-Morethan45min;5-morethananhour)Mina0111Maxb1777Mean0.4905.4545.9235.470349Std.Dev.–1.4651.3681.532IncomeCivilstateClassEducationGroupNewspaperTV1114011533241771.9812.2602.12414.0130.2524.7664.0381.1401.4510.9113.461?1.5241.483Longab154.0361.018Min–Minimum.Max–Maximum.Table2

Mixedlogitparameterestimatesandt-statistics(dependentvariable:DoyoufavorachargeoncarsenteringLisbon?:Yes=1,No=0)VariablesMixedlogitsegment:DriversCoefficients(t-ratio)InterceptPollutionNoiseCongestionIncomeCivilstateClassEducationGroupNewspaperTVLongRandomeffectsNoiseIncomeEducationObservationsLogLikelihood*Mixedlogitsegment:traincommutersCoefficients(t-ratio)3.212(21.089)*0.132(3.930)*0.051(2.836)*0.213(2.934)*–0.317(2.955)**0.512(3.658)*–0.052(2.712)*0.512(0.743)0.012(3.201)À0.135(À2.249)*–0.315(2.030)0.051(2.273)568À295.316Mixedlogitsegment:buscommutersCoefficients(t-ratio)4.135(11.473)*0.218(4.498)*0.412(3.844)*0.351(3.209)*–0.128(2.928)**0.612(2.407)**–0.051(1.804)0.213(1.862)***0.123(0.820)À0.035(À2.212)**–0.591(3.969)**0.125(3.227)442À299.9672.498(1.882)***0.163(3.281)*0.174(3.229)*–0.204(2.188)**0.427(2.679)*–À0.083(À2.001)*0.086(0.862)0.152(0.985)À0.196(À2.397)*0.379(2.605)*0.255(2.036)**0.032(3.113)642À294.75Dependentvariable:choosingtorestrictcarsenteringthecity.Means10%.**Meansstatisticallysignificantat5%.***Meansstatisticallysignificantat1%.350C.P.Barros,P.U.C.Dieke/TransportationResearchPartD13(2008)347–350tisticallysignificantandthreeheterogeneousvariablesareidentified.Twoarecommontoalltravelmodetypes,educationandincome,whilenoiseisspecifictodrivers.Thefindingspointedtoasignificantcorrelationbetweentheprobabilityofsupportingachargeoncarsandtheexogenousvariables.Forthemixedlogitframework,theprobabilityofsupportingachargeincreaseswithmostofthevariablesbutsup-portdecreaseswithgroupstravelingincarsandwithlong-distancetravelinallmodesoftransport.Sincethethreeenvironmentalvariablesarepositivelyrelatedwiththeprobabilityofsupportingthechargeandaresta-tisticallysignificantthereissomesupportforthenotionthatacceptanceofachargeisrelatedtoenvironmentalconcerns.However,noiseisrandomandthereforereflectsheterogeneousopinionamongthecardrivers.Equally,thesocio-economicvariablesare,forthemostpart,positive,althoughincomeandeducation,classandcivilstatusarealsopositiveandstatis-ticallysignificantandhomogenous.However,thevariablegroupisnegativeforcars.Thismixtureofresultssuggeststhatsocio-economicsarenotconsistentforcesaffectingsupportforcongestioncharging.Regardingknowledgeandinformation,TVandNewspaperarebothpositiveandstatisticallysignificantvariablesindicatingthisisanimportantaffectbuttheneg-ativesignsforlong-distancecommutingindicatethattemporalconstraintsarenot.5.ConclusionsTheseresultssuggestthatahighlevelofsupportfortrafficrestrictionsexistsamongaffluenttravelersandthatcommut-ersarehighlysensitivetotheeffectsofpollution,noiseandcongestion,althoughlesssotonoise.Environmentalawarenessissupportedbytheinformationobtainedfromnewspapersandtelevision.Thelengthoftravelnegativelyaffectssupportforacharge.Thusonaverage,commuterssupportachargeforentryintoLisbon,butcardriverswhotravelingroupsareopposedtoit,asarepersonswithlongdailyjourneysinallmodesoftransport.ReferencesDillman,D.(Ed.),1978.MailandTelephoneSurveys:TheTotalDesignMethod.Wiley,NewYork.Espino,R.,Román,C.,Ortuzar,JdeD.,2006.Analysingdemandforsuburbantrips:AmixedRP/SPmodelwithlatentvariablesandinteractioneffects.Transportation33,241–261.Hensher,D.A.,Puckett,S.M.,2007.Congestionandvariableuserchargingasaneffectivetraveldemandmanagementinstrument.TransportationResearchPartA41,615–626.Safirova,E.,Gillingham,K.,Houde,S.,2007.Measuringmarginalcongestioncostsofurbantransportation:donetworksmatter?TransportationResearchPartA41,734–749.Srinivasan,K.K.,Athuru,S.R.,2005.Analysisofwithin-householdeffectsandbetween-householddifferencesinmaintenanceactivityallocation.Transportation32,495–521.Train,K.,2003.DiscreteChoiceMethodswithSimulation.CambridgeUniversityPress,Cambridge.