fRANS?OR”ia/,URQADRESEARCHLABORAT~T LIBRARY ILP)&u9P~ I J.*-..,-–—-– Roadaccidentsindevelopingcountries–urban problemsandremedialmeasures by G.D.Jacobsand1.A.Sayer TRANSPORTmdROAD ~SEARCHLABORATORY DepartmentofTransport TRRLSUPPLEMENTARYREPORT839 ROADACCIDENTSINDEVELOPINGCO~WES– URBANPROBWMSANDMWDIALMEASURES by GDJacobsandIASayer 1.- (ThetextofapaperpresentedtotheSeminar‘TransportSafetyinDevelopingCountries’ organisedbytheInternationalCenterforTransportationStudiesinItaly,November1984.) TheworkdescribedinthisReportformspartoftheprogrammecarriedout fortheOverseasDevelopmentAdministration,butanyviewsexpressed arenotnecessarilythoseoftheAdministrationortheDepartmentofTransport OverseasUnit TransportandRoadResearchLaboratory Crowthorne,Berkshire 1984 ISSN0305–1315 CON~NTS Abstract 1. 2. 3. 4. 5. 6.Introduction Accidentpatterns 2.1Casualtiesbyclassofroaduser 2.2Typesofvehiclesinvolvedinaccidents 2.3Thedistributionofaccidentsbyclassofroad Studiesofroad-userbehaviourinThirdWorldcities Remedidmeasures Actiowledgements ReferenwsPage 1 1 2 2 4 5 6 8 10 10 @CROWNCOPYRIGHT1984 Extractsfromthetextmaybereproduced,exceptfor commercialpurposes,providedthesourceisacknowledged ROADACC~ENTSINDEWLOP~GCOUNTWES– URBANPROBLEMSANDREMEDIALMEASURES ABSTRACT Thisreportpresentsareviewoftheroadaccidentprobleminselectedcitiesin developingcountries.Itshows,foranumberofcities,thedistributionof accidentsbytypeofroad,classofroaduserandvehicleinvolved. Itwasfoundthatinthecitiesstudied,theroadaccidentfatahtyrates @erticencedvehicle)wereuptothirtytimesgreaterthanincitiesinthedeveloped world. Studiesofroad-userbehaviourcarriedoutinanumberofThirdWorldcities aredescribed.Resultsshowdriverbehaviourtobemarkedlypooreratpedestrian crossingsandsignrdcontro~edjunctionsintheThirdWorldcitiesthanatsimilar sitesintwocitiesinGreatBritain. Arangeofremedirdmeasuresthatcanbeintroducedatrelativelylowcost inThirdWorldcitiesaredescribed. 1.INTRODUC~ON Studiesofroadaccidents1‘2haveindicatedthataccidentrates(perlicensedvehicle)indevelopingcountriesare highincomparisonwiththoseofWesternEuropeandNorthAmerica.kmostEuropeancountriesthemajority ofreportedinjuryaccidentsoccuronroadsinurbanareas,theproportioninGreatBritainbeingparticularly high,atover75percentofthetotal.AsstatedbySabey3:– “Whfiethedramaofthehighspeedcrashormultiplepfle-upattractswidespreadpubhcityandmakes thegreatestimpactonthepubfic,intermsofhardfacts,thelargestproblemassociatedwithroad accidentstiesinurbanareaswheretrafficspeedsarerelativelylow.Three-quartersofdlinjury accidentsreportedinGreatBritainoccuronroadsrestrictedto30or40milesperhour.Itisvital thereforetoexamineunderwhatcircumstancestheseaccidentsoccurandhowtheproblemmay bealleviated....“ Itisinterestingtonotethatinthe~wherethree-quartersofaflaccidentsoccurinurbanareas,asimflar proportionofthepopulationdsoliveinurbanareas.Ifthisisrepresentativeofothercountries,thenwherethe percentageofthepopulationhvinginurbanareasislow,theproportionofaccidentstakingplaceinurbanareas mightakobeexpectedtobelow.Populationstatisticsfrom20developingcountriesshowedthatindcasesbut one,themajorityofthepopdationhvedinruralareasandforauthesecountriescombined,thepopdation distributionwas28percenturban,72percentrural.InthesecircumstancesthemajorityofaccidentsinThird” Worldcountriesmightbeexpectedtooccurinruralareas;fromTable1itcanbeseenthatthisisnotthecase. TAB~1 Distributionofaccidentsbytypeofarea Country Chile Ettiopia Jamaica Jordan Korea Malaysia Patistan (SindProvince) Swaziland Turkey ZambiaYear 1979 1982 1977 1980 1982 1982 1981 1971 1980 1977UrbanPercentage 85 61 70 80 75 50 62 37 73 56RuralPercentage 15 39 30 20 25 50 38 63 27 a Inonlyoneofthecountriesforwhichdatawereavdableweretheremoreaccidentsinruralthaninurban areas.Itmaywe~bethataccidentstakingplaceinruralareasarelesswellreportedthaninurbanareas.Conversely thehighproportionofaccidentsinurbanareasmaybeduetothefactthatthelargemajorityofvehiclesin developingcountriesarebasedintheurbanareas. AnotherwayofexaminingtheroadaccidentsinThirdWorldcitiesistocomparefatalityrates(ieroad accidentdeathsper10,000vehiclesregisteredineachcity)withthosederivedforurbanareasinthedeveloped world.ResultsaregiveninTable2. Fatalityratesinthecitiesofdevelopedcountriesrangefrom1.4inTokyo,Japanto2.8inGreater Manchester,UK.RatesintheThirdWorldcitiesaremuchhigherandalsoshowmuchgreatervariationwhich may,inpart,reflecttheunreliabilityofthebasicdata.IntheThirdWorldcities,ratesrangefrom10fatalities per10,000registeredvehiclesinBangkok,ThailandandHongKongupto43inSeoul,Koreaand45inAmman, Jordan.Thefactthatavehicleisregisteredinacityisnoguaranteethatitoperatesonlyinthatcityandresults shouldbetreatedwithadegreeofcaution.Nevertheless,figuresinTable2suggestthatincomparisonwith citiesindevelopedcountries,thoseinthedevelopingworldhaveconsiderablyhigherroadaccidentdeathrates. 2.ACC~ENTPA~ERNS InordertoobtainsomeinsightintothenatureofroadaccidentstakingplaceinThirdWorldcities(ietypesof vehiclesinvolved,casualtiesbyclassofroaduseretc)datawereobtainedfromalimitednumberofcitiesand resultsarepresentedbelow. 2.1 &sualtiesbychssofroaduser Table3showstheproportionofcasualtiesamongstthevariousclassesofroaduser.Infewofthecitiesin developingcountriesweredataco~ectedsuchthatadetailedbreakdownofclassofroad-usercouldbeobtained. Thusclassofroad-userwasoftensubdividedsimplyinto‘pedestrian’and‘vehicleoccupant’.Neverthelessttisstifl 2 providessomeinsightintothefactthattheproportionofpedestriancasualtiesishigherinthesecitiesthanin urbanareasinGreatBritain.Forexample,inBombayandAddisAbabatheproportionofpedestriancasualties isthreetofourtimeshigher.DatafromDelhialsosu~eststk~toccupantsofcommercialvehiclesformamuch greaterproportionofcasualtiesthaninGreatBritain. TABM2 Fatahtyratesindifferentcities Registered vehicles (thousands) 33 40 822 239 258 136 328 250 192 254 2449 779 2860 646 ~Fatalities/ 10,000veticles CityYearFatalities AddisAbaba Amman Bangkok Bombay Deh Fortaleza HongKong Karachi Salvador Seoul Greaterbndon GreaterManchester Tokyo Detroit NewYork1982 1980 1983 1980 1980 1982 1983 1980 1982 1982135 180 849 733 663 314 322 665 494 109141 45 10 31 26 23 10 27 26 43 1982 1982 1980 1982 1982553 226 392 97 5552.2 2.8 1.4 1.5 2.2 TABE3 Casualtiesbyclassofroaduser(expressedaspercentageoftotal) 1 1I ChssofRoadUser II Usersof 2-wheeled Motor vehicles City YearPedestrianPedal CycleCar/TaxiUsersof Commercial Publicvehicle occupants TransportOccupantsTotal Other AddisAbaba1982 Amman1981 Bombay1979 Colombo1980 De~i 1983 HongKong 1983 Karachi 1981 UrbanAreas1982 GreatBritain80 66 68 58 36 36 44 < < —20 > 34 > 27 > <16 3 922 <51 > 48 >100 100 100 100 100 100 5 < 11 3 < < —26+ 16 10 —8+3 24 1023 37 1412 100 I 3 Anearlierstudy4ofpedestrianriskratesintwoThirdWorldcitiesillustratedthegreaterdegreeofriskthat pedestriansincurincrossingbusyshoppingstreetsthaninadevelopedcountrysuchasGreatBritain,where studies5’6hadbeenmadeoftheriskthatpedestriansincurredincrossingbusyroads.Bydividingthenumberof personal-injuryaccidentsinvolvingpedestriansonagivensectionofroadina2yeartimeperiodbytheaverage, hourlypedestrianflowameasureofrelativeriskwasobtained. Theabsolutelevelofriskcannot,ofcourse,bemeasuredasthiswouldrequireaknowledgeofthetotal pedestrianflowacrosstheroadinayear.Theriskratesobtainedwererelatedtothevehicleflowmeasuredon samesectionsofroad.the InordertoobtainsimdarrelationshipsinNairobi,KenyaandSurabaya,Indonesia,(workbeingcarriedout in1976)thenumberofpedestrianscrossingbusyshoppingstreetsduringtheperiod1000–1200hoursand 1400–1600hourswasobtainedtogetherwiththevehicleflowsonthesestreetsoverthesametimeperiod.Counts weremadeduringoff-peakperiodssothatvtidcomparisonscouldbemadewithresultsofanumberofpedestrian studiesmadeinGreatBritainwherecountsweremadeduringmid-morningandmid-afternoon.Therelativerisk rateswerethenregressedagainsttheaveragevehicleflowandequationsderivedwhichwerestatisticallysignificant atthe5percentlevel.TheresultsaregiveninFigure1togetherwiththerelationshipobtainedforshoppingstreets inGreatBritain. ItcanbeseenthatthelevelsofriskinNairobiandSurabayaweremuchhigherthaninGreatBritainfor similarlevelsofvehicleflow;theslopeoftheregressiontinecalculatedfromthedatainGreatBritainwasless steepthanforcitiesintheTtirdWorld.Ataflowof1500vehiclesperhour,forexample,theriskratesare86and 172percentgreaterinNairobiandSurabaya,respectively,thaninGreatBritain;ataflowof3000vehiclesper hour,theratesare180and260percentgreater.Thustakingpedestrianandvehicleflowsintoaccount,pedestrian accidentrateswerefoundtobeconsiderablygreaterthaninGreatBritain. 2.2 Typesofvehiclesinvolvedinacctients Table4showstheproportionofvehiclesinvolvedinroadaccidentsinthecitiesforwhichdatawere available. Itcanbeseenthatthereareconsiderabledifferencesbetweenthevariouscities.Asmightbeexpected,with therelativelylowlevelofcarownershipexistingintheThirdWorld,theproportionofcarsandtaxisinvolvedin accidentswasconsiderablylessthaninGreatBritain.Theproportionoftwo-wheeledmotorvehiclesinvolvedin accidentsvariedconsiderablyinthevariouscities,beingaslowasonepercentinAddisAbabaandashighas38 percentinSurabaya. Inthecitieswherecyclerickshawsoperate,suchasBangalore,JakartaandSurabaya,theproportionof thesevehiclesinvolvedinaccidentsisconsiderable.Observationssuggestthereisatendencyforthedriversof theseveticlestodisobeytrafficregulations(trafficsignals,one-waystreetsetc)andthevehiclesthemselves providepassengerswithlittleprotection. IntheThirdWorldcitiesbothcommercialandpubticservicevehiclesform“ahighproportionofthetotal numberofvehiclesinvolvedinaccidents.StudiesinKenya7andelsewherehaveshownthatcommercialvehicles arecommonlyusedtotransportpeople(particularlytoplacesofwork)asweflasgoodsandthenumbersof peopleinjuredinanaccidentinvolvingatruckcarryingpeopleinthiswayisoftenhigh.Seeplate1. 4 City AddisAbaba Bangalore Bangkok Bombay Colombo Delhi Jakarta Salvador Seoul Surabaya BufltupAreas GreatBritain TAB~4 Vehiclesinvolvedinurbanroadaccidents(expressedaspercentageofthetotal) Year 1982 1977 1983 1977 1980 1983 1980 1982 1983 1980 1982Pedal cycles — — 4 12 2 — 8 82-wheeld motor vehicles 1 32 22 12 11 28 3 5 6 38 18Car5– Taxis 58 19 53 58 40 17 49 50 46 18 68VehicleType Commercial vehicles 27 16 12 16 25 18 34 28 19 16 2Public service vehicles 8 10 12 13 18 25 5 7 29 8 4Rickshaws — 23 1 — — 10 — 12Other 6 — 1 — 2 — — 8 — — Total 100 100 100 100 100 100 100 100 100 100 100 Astudy8ofaccidentrecordsofanumberofpubtictransportundertakingsinIndiashowedthatbuseswere involvedinaboutfivetimesmoreaccidentsthanmightbeexpectedfromtheirnumbersontheroadorfromtheir annualvehiclekilometrage.FatatityratespermiMonbuskilometrestravelledwereaboutsixtimesgreaterthanfor pubhctransportinbndonandovertentimesgreaterthanforothercitiesinGreatBritain. Thisstudyincorporatedanexaminationofonecompleteyear’saccidentrecordsoftheDeltiTransport CorporationOTC).Inaddition,about10percentofaubusdrivers(580)wereinterviewedtoobtaininformation ontheirbackground,workingconditions,experienceandknowledgeofthehighwaycode.Thisshowedthat ahostaureceivedsomeformoftrainingfromtheDTCbutrelativelylittletimewasspentactua~ybehindthe wheelofavehicle.QuestionsontheMghwaycodereverdedthattheunderstandingofemergencystoppingdistances wasparticularlypoor. Asurveyoftheconditionof234buses(over10percentofthetotal)infivedepotsoperatedbytheDTC showedthatover90percenthadnorearfights,stoptightsorsidetightsinworkingorderandover47percenthad nofrontorrearindicatorsinworkingorder.Overhalfthebusesexaminedhadeitherextremeoruneventyre pressures.ClearlytheaccidentrecordoftheDTCcouldbeimprovedbyimprovingvehicleconditionanddriver training. 2.3 Thedistributionofaccidentsbyckssofroad hinvestigationsofthepatternofaccidents,urbanareascanbeconvenientlyseparatedintodistinctlocations withdifferentcharacteristics.Inanearfieranalysis4ofthedistributionofaccidentsoccurringinfourThirdWorld cities,numbersofaccidentsinthecentralareas,onAroads(urbanarterialroutes),Broads@rimarydistributor routes),residentialroadsandotherroadswereobtained.TheresultsfromthesecitiesaregiveninTable5,together withthosefromtownsinGreatBritain.FromTable5itanbeseenthatthecitiesoftheThirdWorldstudied,the percentageofaccidentstakingpkceinthecentralareastendedtobemorethaninGreatBritain;thenumber occurringinthecentralareaof~gston,Jamaicabeingparticularl~high. 5 Thislargerproportionofaccidentsinthecentralareasofthesecitiesisprobablyafunctionofland-useand socialactivity,developingcountriestendingtohavethelargemajorityoftheshops,businessesandevenindustry concentratedinthecentresoftheirmajorcitieswithlessactivityontheperipherythanisthecaseindeveloped countries. TABLE5 Accidentdistributioninthecitiesstudied Centralarea‘A’Roads‘B’RoadsOtherTOTAL City No %No%No I% No% K~GSTON901 59.0620. 1521100 (1962) 41.0% MOMBASSA (average1972/73)188 33.310017.711219.8166 29.3566100 NMROBI (1972)540 26.139218.9 66832.247222.82072100 SUUBAYA (1974)969*23.863715.64059.9206750.74078100 Averageof3towns inGreatBritain 39820.3 66134.528815.0579 30.21917100 (1971) *Includingshoppingandbusinessdistricts 3.ST~IESOFRO~USERBEHAVIOURINTN~WORDCITIES Inresearchingroad“accidentsindevelopingcountriesithas,inthepast,proveddifficulttocarryoutdetafled‘on- the-spot’accidentinvestigationsthatenabletherelativecontributionsmadebytheroad,vehicle,humanfactors andenvironmentinroadaccidentstobeassessed.However,insomeThirdWorldcountriesthepolicehave attemptedtoattributeamaincausetoeachaccidentreported(itisWelythatalargeproportionofauroad accidentswerenotreported)andintheanalysis9ofresultsforfivecountries,roadusererrorwasidentifiedas themaincauseinatleast70percentoftheroadaccidents.Thepoticedefinitionsofmainmusesmayvary considerablyfromonecountrytoanotheranditistikelythatinmanyoftheaccidentstherewasmorethanone cause.Thusthesefiguresmustbetreatedwithcaution;itisprobable,foremmple,thattheyunderestimatethe contributionoftheroadenvironment.Neverthelesstheydosupportthefindingof‘on-the-spot’accident investigations10 carriedoutinGreatBritainthatroad-userbehaviouristhemajorcontributoryfactorinroad accidents. Bemuseoftheimportanceofroad-userbehaviourasafactorinaccidentsandbecauseofrequestsmadeby roadsafetyauthorities,preliminarystudies9weremadebytheOverseasUnit,TRRLofroad-userbehaviour, 6 maidyatjunctionsandpedestriancrossings,inselectedurbanareasinanumberofdevelopingcountries.Wherever possibleresultswerecomparedwiththoseofsimilarstudiescarriedoutinGreatBritain. Driverbehaviourwasobservedat‘zebra-type’(ieuncontrolled)pedestriancrossingsinfiveThirdWorldcities andcomparisonsmadewithresultsfromstudiesofbehaviouratsimflarcrossingsinbndonandReading,W.It wasmandatoryfordriverstostopforpedestriansonthecrossingsinaucitiesstudied.Itwasfoundthatthe averageproportionofdriversstoppinginfouroftheThirdWorldcitiesrangedfrom10to17percentwtistin Surabayathepercentagewaswellunderonepercent.TheequivalentvaluesinReadingandbndonwere72and 40percentrespectively. Observationsweredsomadeatsignd-contro~edjunctionsinthesamecitiesandtheproportionsofdrivers @resentedwithafreechoice)stoppingattheredsignalwererecorded.ItcanbeseenfromTable6thattheper- centageofdriverschoosingnottostopattheredsignalintheThirdWorldcitieswasgreaterthaninReadingand hndon.StudieswerecarriedoutinNairobiin1971and1977;theresultsfor1977showedamarkedimprove- mentoverthe1975value.Thissurprisingresultmaybeduetothefactthatthenumberofsignrdsinthecity increasedfromthreetoovertwentybetween1975and1977.&vingmoresignals(andhavingthemsetcorrectly) mayhavebroughtabouttheobservedimprovementindriverbehaviour.Itshouldbenoted,however,thatin Bangkok,AnkaraandSurabayamanyjunctionsweresignal-controlled,butbehaviourwassti~poor. TAB~6 Non-observanceoftheredsignalinselectedThirdWorldcities City Ankara(2sites)1974 Bangkok(9sites) 1975 Nairobi(2sites) 1975 Nairobi(1Osites) 1977 Surabaya(6sites)1975 Surabaya(6sites)1976 Centralbndon(11sites)1977 ReadingArea(19sites) 1977Numberofdriverswhohad afreechoiceofstoppingor notstoppingatredsignal 101 754 203 3045 253 396 364 726Numberofdrivers choosingnotto stopatredsignrd 36 391 101 210 92 130 22 30Percentageofdrivers choosingnottostop atredsignal 35.6 52.0 50.0 7.0 36.0 48.8 6.0 4.1 Recent(unpublished)researchcarriedoutbytheOverseasUnit,TRRLinIslamabadandRawalpindi, Pakistaninvestigateddriverbehaviouratsiteswhereimprovementsinroadmarkingswereintroduced.Atmajor– minorjunctionswherestoplineswereinsta~ed,andatsiteswheredoublewhitefineswerepaintedonthecentre ofthecarriageway,driverbehaviourwaslittlechanged.ItmayweflbethatinThirdWorldcitieslowcostmeasures suchasimprovedroadmarkingsmaybemuchlesseffectiveforcontrolhngdriverbehaviourthaninthedeveloped world. 4.REMEDIALMEASURES Preliminarystudiesofroad-userbehaviourinselectedThirdWorldcitieshaveshownthatbehaviourattraffic signals,pedestriancrossingsanddsoatpriorityjunctionsismarkedlypoorerthanincitiesinGreatBritain. LimitedstudieshavealsosuggestedthatthelevelofknowledgeonroadsafetymattersisnotM@.Whetherthis poorknowledgeandbehaviourhasaneffectonthenumberofaccidentstakingplaceindevelopingcountriesis notyetfullyestabhshedandmoreresearchisneededtoeitabtishwhethersuchfinksetist.However,itispossible thatremedialmeasuressuchasroad-usereducationorpublicity,whosebenefitshavebeenfounddifficultto identifyandquantifyinEuropeandthe~,couldbefarmoreeffectiveindevelopingcountrieswhereetisting standardsofknowledgeandbehaviouraremuchlower. Becauseoftheirrelativelygreaternumbersindevelopingcountries,childrenrepresentamuchgreater proportionofroadaccidentcasualtiesthantheydoindevelopedcountries.Thusacomparisonoftheage distributionofpersontiledindevelopedanddevelopingcountries11showedthat20percentofthosekflledin developingcountrieswereagedunder15whilsttheequivalentfigureindevelopedcountrieswas10percent. ThispointstothevaluesofdevelopingteachingaidsforuseinThirdWorldcountries,simflartothosedeveloped foruseinschoolsinGreatBritainandotherdevelopedcountries. Withthegenerallylowstandardofroad-userbehaviourthatefistsinmanyThirdWorldcountries,which mayinturnbedueeithertoalackofawarenessoftrafficregulationsortoageneral“attitude”towardsroad safety,itisimportantthatadequatetrafficlawenforcementisprovidedbythepolice.Becausefittleresearchhad beencarriedoutinthisfielditisdifficulttoassessthepotentialofpoliceenforcementforaccidentreductionin developingcountries.Thereishoweverlikelytobeconsiderablepotentialinthesecountries,forinmanyofthem thetrafficpolicearenotsowelltrainedorequippedastheyareindevelopedcountries.Further,inmanyThird Worldcitiesthepoliceareobtigedtospendmuchoftheirtimecontrollingtraffic,withlittletimeavailablefor trafficlawenforcement. Themostpromisingevidencefortheroadsafetybenefitsofenforcementindevelopingcountriescomes fromSingaporeandEgypt.InSingapore,acombinedpublicityandenforcementcampaignappearstohaveleadto dropsof19percentinfatahtiesand50percentinseriousinjuries,althoughtherewasariseof20percentin slightinjuries.InEgyptapackageofpoliceenforcementmeasuresincludingradar,increasedpatrolsandheavier penaltiesfortrafficoffenceshashadasignificanteffectonaccidentsontwomajorinter-urbanroads.Ononeof theseroadstherehasbeenanoveraflreductioninthenumberofaccidentsofover50percent. Perhapsthetwomostimportantmeasuresthatcanbeadaptedtoprotecttheroaduserduringthecourseof anaccidentaretheuseofseatbeltsforvehicleoccupants,andcrashhelmetsformotorcyclists.Therehasbeen growingevidencefromthedevelopedworldthatthecompulsorywearingofseatbeltsresultsinasignifi~nt reductionininjuries,particularlythoseofamoreseverenature.Thebenefitsofwearingaseatbeltinanyparticular accidentsituationshouldbesimilarinbotidevelopedanddevlopingcountries.Inviewoftheoftenpoorermedical facilities,thebenefitscouldinfactbegreaterinThirdWorldcitiesinthecaseofthemoreseriousinjuries. Regrettably,fewThirdWorldcountrieshave,asyet,introducedcompulsorywearingofseatbelts. Motorcyclinghasbeenshowntobeaparticularlydangerousactivity.InGreatBritainfore~mplethe fatahtyrate@ermfitionvehiclemilestrave~ed)ofmotorcyclistsin1977wasahnost30timesgreaterthanthat forcardrivers,despitetheintroductionofthecompulsorywearingofhelmetsin1973.Asshowneartier,inmany ThirdWorldcities,motorcyclistsareamajorcasualtygroupparticularlyinthecountriesofSouthEastAsiawhere motorcyclesformahighproportionofthetrafficmti. 8 Widersocialandbehavioraldifferencesbetweendevelopedanddevelopingcountriesmustalsobeconsidered whenassessingthepotentialforroadaccidentreductioninThirdWorldcities.ThusinmanyAsiancitiesthreeor morepeoplecanfrequentlybeobservedridingamotorcycleorscooter.Seeplate2.Therelativeriskofsuchover- loadinghasnotbeeninvestigatedbutnewlegislationand/orstrongerenforcementofexistinglawswouldberequired ifthisproblemistobedealtwitheffectively.ThusinDelhilegislationexistsforthecompulsorywearingofcrash helmets(withmembersoftheSikhreligionbeingexempt)butthelawappliesonlytothedriverofthevehicle; consequentlypassengersrarelywearcrashhelmets.AsitiscommoninDelhitofinduptofivepeople(usuallythe father,motherandchildrenofonefamily)usingamotorcycleorscooteratonetime,thelawisfarlesseffective thaninasimilarcityinadevelopedcountry. AsstatedinSection2,commercialandpubticservicevehiclesareinvolvedinproportionatelymore accidentsinThirdWorldcitiesthanisthecaseinEuropeandNorthAmerica.Thewayinwhichthesevehicles areusedleadstopotentiaHydangeroussituationswithopenlorriesoftencarryinglargenumbersofworkers,and busescarryingpeoplehangingontheoutsideofthevehicle.Seephte3.Rratransitformsofpubtictransport, cyclerickshaws,sharedtaxisetcdsohaveareputationofbeingdangerousvehiclesinwhichtotravel.Seeplate4. Theaccidentrecordofthesevehiclescouldbeconsiderablyimprovedbylegislationprohibitinglorries,busesand minibusesfromcarryingpassengersinadangerousmanner.ResultsfromDetialsosuggestthatmuchcouldbe donetoimprovethesafetyofpassengersinpubticservicevehiclesthroughimprovedvehiclemaintenanceand alsothroughimprovedtrainingofbusdrivers. Therehasbeenincreasingevidencefromthe~andUSAthatrelativelydetafledlocalaccidentinvestigation, combinedwithlow-costengineeringremedidmeasures,canbehi~ycost-effective.Theexperiencebeinggained fromfo~owingthisapproachinthesetwocountriesisofparticularrelevancetotheThirdWorld.Indeveloped countries,agrowingemphasishasbeenplacedinrecentyearsonobtainingvalueformoneyfrommoneyspenton roadsafety.Withtheirlowergrossnationalproducts,thismustalsobeanimportantconsiderationfordeveloping countries. WorkbyJorgensenandWestat12 intheUSindicatedclearlythehighbenefit-costratiosthatcouldbe obtainedfrom“spot”improvementsascomparedwithcontinuouswidentigorovera~modernisationprojects.The tiiteddataavailabletothemstron@ysuggestedthatlow-costprojectsyieldthegreatestsafetybenefitperdo~ar 13showedasimdarresultandhisanalysisofschemesconfirmedthatsmallinexpen- expended.hbndon,bridles siveschemescouldhaveaverymarkedeffectonroadsafety.Inhndontheseincludedparkingrestriction,traffic signals,roadmarkingetc.However,indevelopingcountriessuchcountermeasuresmaybemuchlesseffective because,asstatedearher,thereisevidencetoshowthattheyarelesseffectiveinThirdWorldcitiesowingtopoor road-userbehaviour.FromSection2,theanalysisof~sudtiesbyclassofroad-usersuggestthatpedestriansare particularlyvukerableinThirdWorldcitiesandmuchgreatereffortsshouldbedevotedtoimprovingtheirsafety perhapsbygreateruseofguardrails,footbridgesandtight-contro~edcrossingfacfities.Evenmoreadvantageous wouldbethecompletesegregationofpedestriansandvehiculartrafficbytheestabhshmentofpedestrianprecincts. InNottingham14~forexample,thepedestrianisationofthecentralareasledtoa60percentreductioninM roadaccidentsinatwoyearperiodfo~owingimplementationofthescheme.InthebndonBoroughofHackney, theuseofbarriersonresidentialroadspreventingthemfrombeingusedasthrough-routesresultedina53percent reductioninpedestrianaccidentsanda44percentreductioninotheraccidents.Regrettablytheauthorsknowof noinstanceswheresimdartechniqueshavebeenusedinThirdWorldcities. . Outsidecentralareasofcitiesonlyarelativelysmallproportionofaccidentsmayoccurinclusterssufficiently largetojustifytheuseof‘site-specific’engineeringremedies.Thereisatendencyinresidentialareaandoftenon arterialroutesforaccidenttobescattereddiffuselyoverthestreetsystem.AsseenfromTable5aconsiderable proportionofaccidentsinthoseThirdWorldcitieswheredatawereavailableoccurredonarterialroutes(Aand Broads)andinareasdescribedas‘other’whichwerepredominantlyresidential. WorkattheTRRLiscurrentlydirectedtowardsdeahgwiththisproblemandresearchisbeingcarriedout 15showedthatbytheuseofimprovedjunctiondesign, infivecitiesinEn#and.ApreliminarystudyinSwindon neworredesignedsignal-contro~edpedestriancrossingsandthebanningofrightturns,accidentsinthestudyarea werereducedby10percent.On‘contro~ed’sectionsofarterialroadswherenoimprovementsweremade, accidentsoverthesameperiodincreasedby29percent,whiktonnon-arterialroutes,accidentsremained virtuallythesame. Itisimportantthatsafetyfeatures,suchasthoseinvolvinggeometry,signinganddelineation,beintroduced atthedesignstageratherthanbeaddedlater,ahostasan‘afterthought’.Forexample,itcanoftenbegeatly moreexpensivetowidenthemainroadsataT-junctionafteranaccidentproblemhasbufltupthantoincorporate itattheconstructionstage;thisisbecauseafterconstructionitisoftenfoundthatutihtyserviceshavetobe movedforanylocalwideningscheme. Ithasbecomeappreciatedindevelopedcountriesthatplanningcanhaveaprofoundeffectuponthelevelof roadsafetyinacity.Thelayoutofroadsinresidentialareas,forexample,hasbeenfoundtohaveamajorimpact 16Itisnowgenerallyacknowledgedthatthe‘gridlayout’is uponthelevelofpedestrianaccidentsinparticular. notconclusivetoroadsafety,particularlybecauseofthelargenumberofcross-roadsandtheavailabilityof‘rat runs’throughresidentialstreets,usedbymotoristseitherasshortcutsortobypassmainroadcongestion.Thusat theplanningstageofresidentialareasinThirdWorldcities,considerationshouldbegiventoroadsafety. Finally,whenintroducingroadsafetyremedialmeasuresinThirdWorldcities,itisessentialthatscarce resourcesarenotwastedandthatanymeasuresintroducedarecarefu~yappraisedandanassessmentmadeof theirrelativeeffectiveness. 5.AC~OWLEDGEMENTS TheworkdescribedinthisReportformspartoftheresearchprogrammeoftheOverseasUnit@nitHead: MrJSYerreH)oftheTRRL. 1. 2. 3.6.REFERENCES JACOBS,GDandPHUTC~SON.AstudyofaccidentratesinDevelopkgCountries. Departmentof theEnvironmentDepartmentofTransport,TRRLLaboratoryReportLR546. Crowthorne,1973 (TransportandRoadResearchhboratory). JACOBS,GDandISAYER.Roadaccidentsindevelopingcountries.AccidentAnalysisandPrevention Vol15,pp337–353(1983). SABEY,BE.AccidentsinUrbanAreas.ReprintedfromReportoftheConferenceonTrafficEngineering andRoadSafety.BrightonOtt2–5,1973. 10 4. JACOBS,GDandIASAYER.Astudyofroadaccidentsinselectedurbanareasindevelopingcountries. DepartmentoftheEnvironmentDepartmentofTransport,TRRLReport LR775.Crowthorne,1977 (TransportandRoadResearchLaboratory). 5.MACKIE,AMandSJOLDER.StudyofpedestrianriskincrossingbusyroadsinLondoninnersuburbs. TrafficEngineeringandControlJournal,1965,7(6),376–80~ 6.JACOBS,GDandGDWILSON.Astudyofpedestrianriskincrossingbusyroadsinfourtowns. Minist~ofTransport,RRLReport LR106.Crowthorne,1967(RoadResearchLaboratory). 7.JACOBS,GDandIASAYER.AnanalysisofroadaccidentstiKenyain1972. Departmentofthe EnvironmentDepartmentofTransport,T~LReport SR227,Crowthorne,1976(Transportand.Road Researchhboratory). 8.JACOBS,GDandAJDOWNING.AstudyofbussafetyinDelhi. DepartmentoftheEnvironment DepartmentofTransport,TULReport SR 758. Crowthorne,1982(TransportandRoadResearch Laboratory). 9.JACOBS,GD,IASAYERandAJDOWN~G.Apreliminarystudyofroaduserbehaviourindeveloping countries. DepartmentoftheEnvironmentDepartmentofTransport,T~LReportSR646. Crowthorne, 1981~ransportandRoadResearchtiboratory). 10.SABEY,BEandGGSTAUGHTON.h~eractingrolesofroadenvironment,vehicleandroaduserhr accidents.5thInt.Conf.ofInt.Ass.forAccidentandTrafficMedicine.bndon,September1975. 11.DOWN~G,AJandIASAYER.Aprehinarystudyofchildren’sroadcrossingknowledgeinthree developtigcountries. DepartmentoftheEnvironmentDepartmentof~ansport,TRRLReport SR777. Crowthorne,1982(TransportandRoadResearchhboratory). 12.JORGENSEN,RandAssociates.Costandsafetyeffectivenessofhighwaydesignelements.National CooperativeHighwayResearchProgram.Report197.TransportationResearchBoard,Washington1978. 13.LAND~S,JR.TheworkoftheGLCBlack-spotteam.GreaterbndonCouncdRoadSafetySection, hndon,1979. 14.DALBY,E.Area-widemeasuresinurbanroadsafety. DepartmentoftheEnvironmentDepartmentof Transport,TRRLReportSR517. Crowthorne,1979(TransportandRoadResearchLaboratory). 15.WARD,Hand%DABY.Applicationoflow+ostroadaccidentcountermeasuresaccordingtoanarea- widestrategy.TrafficEngineeringandControlJournal22(ii),November1981. 16.OECD,SpecialResearchGrouponPedestrianSafety.Chairman’sReportandReportofSub-GroupI– ThePedestriansRoadEnvironment.PubhshedbyTransportandRoadResearchhboratory,Crowthorne, 1977. 11 0.032 0.028 0.024 0.020 0.016 0.012 0.008 0.004 00 n l Surabaya l l O Nairobi %6 ~Q Q. l l l o l 0 0400 80012001600 2000 2400 28003200 3600 Averagevehicleflowlh Fig.1Centralareas–Surabaya(1974)andNairobi(1972/73) –pedestrianrelativeriskrates Neg.no. R37518413a Plate1Overloadedcommercialvehicles Neg. no. B1019/82 Plate2MotorscooterridersinDelhi Neg. no. R76181/13 Plate3Overloadedbuses Neg.no.B995/80 Plate4Motorcyclerickshaw (3122)Dd80413701,40012/84HPLtdSo’tonG3371 PRINTED IN ENGLAND ABSTRACT ROADACCIDENTS~DE~DPMENTCO~MES–URBANPROBLEMSANDRE~DIAL~ASURES: G DJacobsandZASayer: DepartmentofTransport,TRRLSupplementaryReport839:Crowthorne,1984 (TransportandRoadResearchbboratory).Thisreportpresentsareviewoftheroadaccidentproblemin selectedcitiesindevelopingcountries.Itshows,foranumberofcities,thedistributionofaccidentsbytypeof road,classofroaduserandvehicleinvolved. Itwasfoundthatinthecitiesstudied,theroadaccidentfatahtyrates(perficencedvehicle)wereupto ttirtytimesgreaterthanincitiesinthedevelopedworld. Studiesofroad-userbehaviourcarriedoutinanumberofThirdWorldcitiesaredescribed.Resultsshow driverbehaviourtobemarkedlypooreratpedestriancrossingsandsignalcontro~edjunctionsintheThirdWorld citiesthanatsimdarsitesintwocitiesinGreatBritain. ArangeofremedidmeasuresthatcanbeintroducedatrelativelylowcostinThirdWorldcitiesare described. ISSN0305–1315 ABSTRACT ROADACCIDENTSINDE~LOPMENTCOUNTMES–URBANPROBLEMSANDREMEDIALMEASURES: G DJacobsandZASayer: DepartmentofTransport,TRRLSupplementaryReport839:Crowthorne,1984 (TransportandRoadResearchhboratory).Thisreportpresentsareviewoftheroadaccidentproblemin selectedcitiesindevelopingcountries.Itshows,foranumberofcities,thedistributionofaccidentsbytypeof road,classofroaduserandvehicleinvolved. Itwasfoundthatinthecitiesstudied,theroadaccidentfatahtyrates@erhcencedvehicle)wereupto thirtytimesgreaterthanincitiesinthedevelopedworld. Studiesofroad-userbehaviourcarriedoutinanumberofThirdWorldcitiesaredescribed.Resultsshow driverbehaviourtobemarkedlypooreratpedestriancrossingsandsignalcontro~edjunctionsintheThirdWorld citiesthanatsimilarsitesintwocitiesinGreatBritain. ArangeofremedialmeasuresthatcanbeintroducedatrelativelylowcostinThirdWorldcitiesare described. ISSN0305–1315