REGULATORYIMPACTANALYSISGUIDELINESFORTHETHAILANDGOVERNMENT

Contents1.PurposeoftheGuidelines……………………………………………………………….32.Introduction…………………………………………………………………………………..43.ScopeandApplicationofRIA………………………………………………………….54.TheKeyElementsofRegulatoryImpactAnalysis…………………………….64.1Identificationoftheproblem………………………………………………………..64.2Objectives…………………………………………………………………………………..114.3IdentificationofOptions……………………………………………………………..114.4AssessmentofOptions……………………………………………………………….124.5PreferredOption………………………………………………………………………..134.6Consultation……………………………………………………………………………....134.7ProposedRegulations………………………………………………………………...135.CostBenefitAnalysis……………………………………………………………………146.RegulatoryImpactAnalysisTemplate…………………………………………...317.Casestudy……………………………………………………………………………………34

PurposeoftheGuidelinesRegulatory Impact Assessment (RIA) is a popular tool used by most OECD countries andincreasinglybyAPECdeveloping countries.RIA isused to reviewexisting andnew legislationandregulation.RIAprovidesarobustanalyticalapproachusingcostbenefitanalysistoevaluatethecostsandbenefitstoensurearegulationprovidesanetbenefittosociety.The benefits of RIA are numerous but some of the key benefits include improving businessproductivity, reducing business costs, creating a business environment to invest, grow andcreate jobs, improving consumer choice, protecting the environment and public health andsafety,andprovidingafoundationforimprovingthestandardofliving.ThailandadoptedtheOECDReferencechecklistforregulatorydecision-makingin1995andhasbeencommittedtoundertakingRIAsforlegislationandregulation.However,theOECDChecklistonly provides a list of principles and little guidance on how to undertake RIA. As a result, agovernment study “RIA Situation in Thailand” found a high level of unawareness of therequirementtoconductRIAandveryfewRIAbeingpreparedforGovernment.Mostdevelopedcountries thathave introducedRIAhavedevelopedguidelinesand training tosupport the OECD RIA Checklist. This is necessary in order to provide government agencyofficerswiththenecessaryknowledgeandskillstoprepareRIAs.DevelopedcountriesthathavebeenusingRIAforoverthirtyyearshavebeenableto improvethe quality and efficiency of regulation resulting in significant cost savings to business andcitizensinthehundredsofmillionsofdollars.Thailand’sglobalcompetitivenessrankingin2014was29outof60countriesandscored38onthe corruption perception index. The index indicates the perceived level of public sectorcorruptiononascaleof0(highlycorrupt)to100(clean).The introduction of RIA similar to other developed countries will provide Thailand with thepotentialtoimproveitscompetitivenessrankingandtohelpremovecorruptioninitslegislationand regulations. In doing so, the adoption of RIAwill facilitate reforms that deliver improvedstandardsoflivingandreduceincomeinequality.Thepurposeof theseGuidelines is toprovide a framework for government agencyofficers tobetterunderstandtheRIAprocessandtodeveloptheappropriateskillsintheGovernment’sRIAtrainingprogram.TheGuidelinescoverthefollowingtopics:• Introduction• ScopeandapplicationofRIA• ThekeyelementsofRIA• Publicconsultationprocesses• CorruptionImpactAssessment• RIACaseStudy

1.IntroductionGovernment regulation is sometimes necessary to achieve certain economic, social andenvironmental goals. However, excessive or poorly designed regulation can impose costs onsocietythatoutweighthebenefitsofregulation.Thesecostscanhavenegativeimplicationsforoverall economic performance, including competition, business costs, consumer choice,employmentandinvestmentopportunities.To avoid the problems caused by poorly designed regulation, it is essential that governmentshouldnotresorttoregulationunlessithascompellingevidencethat:• aproblemexists;• governmentactioniswarranted;and• regulation is the best option available to government to deal with the problem in an

efficientandeffectivemanner.Regulatory impact assessment (RIA) is a document that analyses the problem, the need forgovernment intervention into amarket, and the costs and benefits of feasible options to dealwiththeproblem.ThepurposeoftheRIAistoensurethataffectedstakeholdersandthewidercommunityhaveanopportunitytocommentonallaspectsoftheRIAandtheproposedregulation.Ineffect,publicconsultationprovidesgovernmentwiththeopportunitytoseekreassurancefromthosedirectlyaffected that a problem exists, the scope and scale of the problem has been defined and theproposed regulation is the best option to deal with the problem. Accordingly, the RIA helpsgovernmenttomakeaninformeddecisionbeforeitintroducesalaworregulation.Accordingly,bestpracticeRIAcountrieshaveintegratedthekeyfeaturesoftheRIAprocessintopolicydevelopmenttoensurethepreferredpolicyresponsehasbeenrigorouslyassessedandisthebestoption.ThisisasuperiorapproachcomparedtomakingadecisiononaparticularpolicyresponseandsubsequentlyundertakingtheRIAprocess.The government strongly encourages its departments to integrate the key features of theRIAprocess into policy development to facilitate the selection of the best option and to avoidunnecessary delays to the government’s response to economic, social and environmentalproblems.Some OECD countries have been using RIA for over thirty years and have strengthened thestandards in their RIA Guidelines over time to further improve the quality of RIAs. This hasresultedintheadoptionofagreaterrangeofmethodologiesandapproaches.Inthesecountries,governmentdepartmentshavebeenabletolearnandadjusttoincrementalchangestotheRIAGuidelines.Thailand is essentially at thebeginningof theRIA journeyandhasdecided to commence thatjourney learning and applying the fundamentals of the RIA process. The adoption of ‘bestpractice’RIAGuidelinesfromsomeOECDcountriescouldover-whelmgovernmentdepartmentsthatdonothaveexperiencewiththeRIAprocessandjeopardizewholeofgovernmentsupportandadoptionoftheRIAprocess.

The key feature of the RIA process that is critical to the government’s success to deal witheconomic,socialandenvironmentalproblemsisthefirstpartoftheRIA;thenatureandextentoftheproblem.It iscriticalforthefollowingreasons.Firstly,thesolerelianceonhigh-levelaggregateddatatodemonstratethataproblemexistswillinvariablyleadtopoorlydesignedregulationthatfailstodeal with the problem, and in most cases, will unnecessarily regulate some parts of thecommunity; imposing additional costs that makes the targeted sector less efficient andcompetitive.Thefollowingisanexampleoftheuseofhigh-levelaggregateddata. Liquorlicensedpremiseswere the thirdmost frequent typeofpremises (behindresidential andoutdoors) for reportedassaults,withdatasuggesting66.7percentofallassaultswerealcohol-related.Thelimitationswiththishighlevelaggregateddataisthatitsuggeststhatmostliquorlicensedpremisesarelikely,atsomestage,toleadtoalcohol-relatedviolence.Afurtherproblemwiththisrelianceonhigh levelaggregateddata is that thesolutionwillbedevelopedandapplied toallliquorlicensedpremises.All OECD countries struggle to perform high quality analysis of the nature and extent of theproblemand care shouldbe taken inusingRIAs fromother countries that primarily usehighlevelaggregateddata.Secondly, if the nature and extent of the problem is analyzed in appropriate detail withsupportingempiricalevidenceitislikelytoreveal,insomecases,arangeofoptionstodealwiththeproblem,and importantlyavoidapplyinganypolicysolution topartsof thecommunityorindustrysectorsthatarenotresponsiblefortheproblem.Usingthesameexampleaboutliquorlicensedpremises,theuseofdisaggregateddatarevealsadifferentpictureoftheproblem:InSydney,27or12percentofhotelsandnightclubsaccountedfor almost60percentof all assaults athotels andnightclubs.7or3per centof the27hotelsrecorded26percentofallassaults.AstudyinNewcastlefoundthatofthe400ormorelicensedpremisesinthearea,only21or5%hadanaboveaveragenumberofalcohol-relatedincidents,withfourpremisesor1%oflicensedpremisesaccountingforalargemajorityoftheseincidents.AtKingsCross,whichhasthehighestrateofassaults,inexcessof20percentoftheassaultswererecordedat just3 licensedpremises;abar/restaurant,abar/nightclubandabar/strip-club.Asimilar outcomewas found atWynyard/The Rocks areawhere 23.3 percent of assaultswererecordedinornear3licensedpremises.This data analysis changes the size and extent of the problem to a handful of liquor licensedpremisescomparedtothehighleveldatathatsuggestedawidespreadproblemamongstliquorlicensedpremises.Accordingly,whilefurthercausalanalysisisrequiredinrespecttothehandfulofliquorlicensedpremises,itisclearthatapplyingaregulatorysolutionandtheassociatedcoststomostliquorlicensedpremiseswouldadverselyimpactonthesebusinessesanditspatrons.Thirdly,providedthecausesandtheextentoftheproblemcanbeclearlydefinedandthecostsofthe problem can be quantified, the rest of the RIA is relatively straightforward in terms ofundertakingacostbenefitanalysisoftheselectedoptions.Thecostsquantifiedintheproblemsection of the RIA are later treated as the potential benefits in the assessment of the variousoptions.

ManyOECDmembercountryRIAsmakeclaimsofmarketfailureinthenatureandextentoftheproblembutprovideverylittlesupportingempiricalevidence.Inmanycases,regulatoryfailureis far more common and there is a litany of published studies on inefficient and ineffectiveregulation.Accordingly,allgovernmentsandcivilservantsshouldapplytheprincipleofcaveatemptortoallpolicy issues and turn over every stone to ensure that regulation is really needed and willachievethepolicyobjectiveatminimalcosttothecommunity!

2.ScopeandApplicationofRIAThe corruption impact assessment and compliance cost measurement framework should beprepared for new and amending legislation by responsible government agencies prior tosubmittingproposedlegislationtotheCabinetOfficeasshowninthediagrambelow.Regulatory impactanalysis (RIA)willbe initiallypreparedbygovernmentagencies toexistingprimary legislation and other forms of subordinate legislation. Once the RIA system isembeddedwithingovernment,RIAwillneedtobeprepared fornewandamending legislationandsubmittedtotheCabinetOfficetogetherwiththeproposedlegislation.ThailandLegislativeProcess

GovernmentAgencyCorruptionImpactAssessmentComplianceCostMeasurement

OfficeoftheCouncilofState(OCS)

CabinetOffice

Cabinet

Parliament

HM

GovernmentGazette

3.TheKeyElementsofRegulatoryImpactAnalysis(RIA)OverviewARIAiscomprisedofthefollowingsevenparts:

1. Identificationoftheproblem

2. Objectives

3. IdentificationofOptions

4. AssessmentofOptions

5. PreferredOption

6. Consultation

7. ProposedRegulationsParts1and4entailmostofthecontentinaRIA.Thekeyissuesthatneedtobedealtwitharediscussedbelowforeachofthesevenparts.Aswillbeseenthekeyissuesarediscussedatageneralandhighlevel.Thishasoneobviouslimitation.Each reader will interpret differently the expected level of detailed analysis. To obviate this,theseGuidelinesprovideacasestudyofhowtoapply thekeypartsofaRIA toanactualcasestudytodemonstratethelevelofanalysisandcriticalthinkingrequired.3.1. IdentificationoftheproblemA key weakness in RIAs prepared by OECD member countries is the tendency to limit theproblem section to an overviewof theproblemat a high-levelwith aggregateddata. Inmanycases, the actual causes are not analysed. This generally leads to an overstatement of theproblemandresultsinanoverstatementofthebenefitslaterinthecostbenefitanalysissection.Thismakes iteasier to justify the introductionofnewregulations.However, it isalso likely toleadtoover-regulation,increasedregulatoryburdenandnon-deliveryofthepolicyobjective,forexampletosavelives,toprotecttheenvironmentandsoforth.The problem section is the most critical part of the RIA. It provides the opportunity tounderstandtheproblemproperlyandtofindsolutionsthatarecommensuratewiththesizeoftheproblem.Toprepareahighqualityanalysisof theproblem,you require superioranalytical skills andawillingnesstofindanddevelopanalyseswherenodataorlimiteddataexistsabouttheproblem.A key challenge for anyone preparing a RIA is to deal with the differing perceptions andpositions of key stakeholders, including government. It is only natural that everyone in thecommunity has a different understanding of a problem;mostly due to their limited access toknowledgeoftheproblem.ThechallengeisevengreaterwhereaMinisterhasmadeapublicannouncementtointroducealaw or regulation or the government has an election commitment to introduce a law or

regulationtoresolveaparticularproblem.ManycivilservantspreparingaRIAfeelcompelledtojustifytheMinister’spositionorthegovernment’selectioncommitment.Indoingso,theseRIAsmisrepresentthesizeoftheproblembyexcludingcriticaldataandanalyses.It is important forMinisters and Governments to be properly advised. If your analysis of theproblem is different to the Government’s perception of the problem, it needs to be advisedaccordingly.Remember,thepurposeoftheRIAistoenablethedecision-maker(Government)tomake informed decisions. If you do not undertake a detailed analysis of the problem or omitcriticalanalysesthatwouldmateriallychangetheoutcomeofthepreferredregulation,youarenotprovidingtheGovernmentwiththeopportunitytomakeaninformeddecision.YourresponsibilityistogivetheGovernmenttheopportunityandchoicetodecidewhetherthesize of the problem is x or y or z and whether they want to proceed with the preferredregulation.Thatistheirdecision.Notyourdecision.Justmakesurethatyougivethemthebestinformation,notjusttheinformationyouthinkthattheywanttohear!Thebenefitsofrobustanalysisofthenatureandextentoftheproblem:• Confirmorrefutetheperceivedextentoftheproblem;• Enablekeystakeholderstomakeinformedcomments;• Enables some key stakeholders to reconsider their previously held positions where the

extentoftheproblemisgreaterorlesserthantheirperceptionoftheproblem;• Enablegovernmenttomakepolicyadjustmentsiftheanalysisofthenatureandextentof

theproblemismateriallylessthantheperceivedmagnitudeandimpactbelievedpriortotheanalysis.Policyadjustmentscouldincludeamendmentstothedesignoftheproposedregulationsoitiscommensuratewiththesizeoftheproblem,withdrawaloftheproposedregulation,adoptionofotherpolicyalternativesthataremorelikelytobecommensuratewiththesizeoftheproblem,oradecisionthatthereisnoroleforgovernmentgiventhatotherregulatoryschemesaddresstheproblemormarketfailurewasnotdemonstrated.

Itisimperativethattheproposedregulationandfeasiblealternativesmatchthenatureandsizeoftheproblem.Forexample,iftheproposedregulationistobanchildrenlessthansixyearsofage being transported on a motorcycle, the size of the problem must exclude motorcyclefatalities and injuries for children over six years of age aswell as other fatalities and injuriesincurredbyothermodesoftransport:car,bus,truck,minivan,lorry,tractor,pedestrianetc.Thepurposeoftheproblemanalysisistomeasurethenumberofchildrenlessthansixyearsofagethatarekilledorinjuredtravelingonamotorcycle.Arobustanalysisofthenatureandextentoftheproblemiscriticaltoregulatoryreformandthedevelopmentofgoodregulationthatisefficientandeffective.Thefollowingquestionsneedtobeexploredandanswered.• Whoisaffectedbytheproblem?• Whatisthescopeandscaleoftheproblem?

• Istheidentifiedproblemapartofalargerproblem?Ifso,whatisthesizeoftheidentified

problemrelativetothelargerproblem?• Whatisthecauseoftheproblem?

• Istheresufficientempiricalevidencethataproblemexists?• Istheextentoftheproblemidentifiedorisitsidentificationbasedonlyonanecdotal

evidence?• Whataretheeconomic,socialandenvironmentalcostsoftheproblem,andwhobears

thesecosts?• Doestheproblemexistcurrently,orisitmerelyanticipated?• Istheproblemaminorirritantorasignificanthazard?• Arethereanytechnological,economic,political,administrative,socialand/or

environmentalconstraintsthatarerelevanttotheproblem?• Arethereexistingregulationsthatcoulddealwiththeproblem?Ifyes,whyarethese

regulationsinadequate?• Whataretheconsequencesofnottakinganyaction?• Could relying on themarket in conjunctionwith the general application of existing laws

andregulationssolvetheproblem?Ifnot,whynot?• Whatistheexperienceinotherjurisdictionswithdifferentregulatoryframeworks?• Whatisthescopeandscaleoftheprobleminunregulatedjurisdictions?

• Has thescopeandscaleof theproblem increased/decreased inunregulated jurisdictions

duetomarket,technological,regulatoryorenvironmentalchanges?3.2. ObjectivesTo enable proper formulation of policies, it is necessary to give careful consideration to thedesiredoutcomes.Unlessthepolicygoalsareclearlyspecified,theidentificationofappropriatealternativemeansofachievingthemwillbecompromised.Particular care shouldbe taken toensure theobjective isdefinedbroadlyand isnot confusedwith the strategy for its achievement. For example, a reduction in motorcycle fatalities is anobjective whereas compulsory wearing of motorcycle helmets is one strategy or means forachievingthisobjective.Focusingonastrategyratherthanthegeneraloutcomewillhinderafullandproperconsiderationofalternativemeansofachievingthedesiredoutcome.Thatisfindingthesolutiontotheproblem.Theobjectiveshouldidentifytheendstobeachievedorthebroadpolicyoutcomesdesiredratherthanthemeansofitsachievement.3.3 IdentificationofOptionsAfundamentalstageinthepolicydevelopmentprocessistheidentificationandassessmentofallfeasiblealternativestotheproblembeingaddressed.Unlessafullandproperassessmentoffeasiblealternativesisundertaken,theregulatoryproposaladoptedmaynotrepresentthebestsolutiontotheproblem.Thus,itisimportanttoconsiderwhatthemosteffectivetoolwillbetoachievethedesiredoutcome.

Preliminaryconsiderationoftherangeofoptionsavailabletoachievethestatedobjectivemayidentify someoptions thatappear, after closerexamination, inappropriateandunworkable. Inthesecases,theoptionshouldbediscardedandabriefexplanationprovidedwhytheoption(s)isnotfeasible.Theremainingfeasibleoptionsshouldbefurtherexaminedingreaterdetailinordertoprovideasolidbasisforacostbenefitanalysis.Inparticular,greaterdetailonthelikelyoperationoftheoptionswilladdresshoweachoptionwillbeadaptedsoastoprovideasolutiontotheproblemaswellaspotentialimpediments.Dependingontheoption,thefollowingquestionsmayneedtobeconsidered:• Howwouldthealternativework?• Whatroledoesgovernmenthave?• Istheresufficientcommonalityofinterest,withinanindustryorprofessionalassociationto

ensurehighlevelsofvoluntarycompliance?• Howwillconsumerinterestsberepresented?• Doesthealternativediscriminateagainstpersons/groups/industries?• Isthealternativelegallyfeasible?• Doesthealternativerestrictcompetition?• Whatmonitoringwillberequiredandhowwouldmonitoringoccur?• Isthealternativelikelytobeenforceable?• Willnon-compliancebeevident?3.4. AssessmentofOptionsCostbenefitanalysismeasurestheefficiencyorresourceallocationeffectsofaregulatorychangeand alternative options. It calculates the dollar value of the gains and losses for all peopleaffected. If the sum is positive, the benefits exceed the costs. The option that provides thegreatestnetbenefitprovidesthemostefficientresourceallocation.Costbenefitanalysis:• provides decisionmakers with quantitative and qualitative information about the likely

effectsofeachoption• encouragesdecisionmakerstotakeaccountofallthepositiveandnegativeeffectsofeach

option,anddiscouragesthemfrommakingdecisionsbasedonlyontheimpactsonasinglegroupwithinthecommunity

• assesses the impactofeachoption inastandardmanner,whichpromotescomparability,

assistsintheassessmentofrelativeprioritiesandencouragesconsistentdecisionmaking• captures the various linkages between the regulatory proposal and other sectors of the

economy (for example, increased safetymay reduce health care costs), helping decisionmakersmaximizenetbenefitstosociety,and

• helpsidentifycost-effectivesolutionstoproblemsbyidentifyingandmeasuringallcosts.Evenwhenitisdifficulttoestimatesomecostsorbenefitswithprecision,costbenefitanalysismakes clear and transparent the assumptions and judgements made. Further, attempting toquantifycostsandbenefitsencouragesanalyststomorecloselyexaminethesefactors.

Formostregulations,costsarenormallymoreevident,measurable,concentratedononegroupand immediate (in termof time) compared to benefits,which are often less easy tomeasure,morewidespreadandlong-term.Stepsinpreparingafullcostbenefitanalysis1.Specifythesetofoptions2.Decidewhosecostsandbenefitscount3.Identifytheimpactsandselectmeasurementindicators4.Predicttheimpactsoverthelifeoftheregulatoryproposal5.Monetise(attachdollarvaluesto)impacts6.Discountcostsandbenefitstoobtainpresentvalues7.Computethenetpresentvalueofeachoption8.Performsensitivityanalysis9.ConclusionIf you are unfamiliar with cost benefit analysis, you should refer to a reputable economictextbook or governments that have published on this subject. The Australian Government’sHandbookofCostBenefitAnalysis(2006)providesguidanceanddetailacrossawiderangeofpossiblepolicyissues.Thiscanbedownloadedfromtheinternet.Other relevant publications can be found on the OECD’s regulatory reformwebsite andmostOECDmembercountriespublishRIAontheirwebsites.ThisresourceshouldbeusedtoobtainRIAsfromothercountriestoascertainhowtheymeasuredcostsandbenefitsonthepolicyissueyou are working on and also to benchmark the analysis of the problem and regulatoryframeworks.3.5. PreferredOptionAcomparisonofthebenefits,costs,netbenefitandbenefitcostratiosforeachoptionshouldbesummarizedinatabletoenablethereadertoquicklycomparethedifferentoutcomes.Ideallythequantifiablecostsandbenefitswillbediscountedtogenerateanetpresentvaluethatisgreaterthanzero.Some costs andbenefitsmaybedifficult toquantify. Aqualitative accountof these costs andbenefitsmaybeused.Governmentregulatorsandagenciesshouldnotbeconcernediftheproposedregulatoryoptionis not the best option or imposes a net cost. It should be remembered that the cost benefitanalysis framework is not a precise process (given that not all costs and benefits can bequantified) and the purpose of theRIA is to seek further input fromkey stakeholder and thewidercommunitytoenableGovernmentandresponsibleMinisterstomakeinformeddecisions.Theremay be uncertainty and various risks associatedwith an option that is superior to theproposedregulatoryoption,ortheremaybeconsiderabledifficultiesinquantifyingmostofthecostsandbenefitsformostoptions.ItisimportantthatGovernmentandMinistersareprovidedwithanhonestappraisalofthecostsandbenefits

3.6. ConsultationConsultation with affected groups is integral to regulation impact analysis (RIA). EffectiveconsultationiscommencedearlyandpreferablypriortothecommencementoftheRIA.Effectiveconsultationengagesaffectedgroupstocontributetopolicydevelopment.Thisincludesproviding information and data to help the responsible department to define the extent andcauses of the problem, measure the likely compliance costs, analyze the impact of anyrestrictionsoncompetitionandtoidentifyfeasiblealternatives.Document the consultation undertaken by identifying the groups, firms, government agenciesandindividualsthathaveparticipatedintheconsultationprocess.Itisimportanttoacknowledgethecontributionofanystakeholderbydisclosingtheinformationanddatathathelpedtoinformtheanalysisoftheproblem.Similarly, it is important to acknowledge stakeholders that have provided compliance costestimatesthathavebeenusedintheassessmentoftheproposedregulation.Documenttheviewsofgroupsaffectedbytheproposedregulationandanyevidenceprovidedtosupportthoseviews.Aresponseandthereasonsfornotacceptingaparticularviewofakeyaffectedgroupshouldbeprovided.Forexample,theXYZindustryfederationraisedconcernsthattheproposedregulationwouldimposesignificantcompliancecostsonitsmembers.Inresponsetotheseconcerns,thedepartmentheldmeetingswiththeXYZindustryfederationand agreed tomeetwith a representative sample of itsmembers to identify andmeasure thelikelycompliancecosts.Meetingswereheldatthebusinesspremisesof12selectedfirmswhereeachfirmprovidedinformationonthetypeofpersonnelandprocessesrequired,thelikelytimerequired and the associated costs involved with compliance of the proposed regulation. Theaveragecompliancecostwascalculatedat6bahtperunitofproductionandrepresentedabout0.8%ofthecostofproduction.Withthisnewinformation,theXYZindustryfederationreconsidereditssubmissionandagreedthat the compliance costs would not impose a significant burden on its members. The XYZindustry federation also agreed to notifymembers in its next newsletter the outcomes of thecompliancestudy.Thisprovidesacleardemonstration thatmattersraised insubmissionshavebeenconsidered,andcontributes to the transparencyof theregulatoryprocess.Thishelps tobuild trustwithinthe community that government is inclusive and gives consideration to matters raised andmakesappropriateadjustmentstotheanalysisand/orthedesignoftheproposedregulations.3.7. ProposedRegulationsA copy of the proposed regulations should be attached at the end of the RIA to enable keystakeholdersandthewidercommunitytocommentonthescaleandscopeoftheregulations.

5.COSTBENEFITANALYSISRationaleforusingcostbenefitanalysisasthepreferredmethodforregulatoryanalysisRegulationhaspositive (benefits) andnegative (costs) impacts.Usually, the group that incursthecostisdifferenttothegroupthatreceivesthebenefitsoftheregulation.Forexample,motorcartraderregulationimposescostsonmotorcartraderstobelicensedandtoprovidewarrantyanddisclosurerequirementssothatconsumersareprotectedfromfaultyvehiclesandcanmakeinformeddecisions.Providedthebenefitsaregreaterthanthecosts,theregulationisdeemedtohaveprovidedanetbenefittosociety.However, there is an opportunity cost attached with every regulation. The opportunity costimposedonbusinessesandconsumersistheresourcesthatcouldbeallocatedtootherusesintheabsenceofregulation.Usingtheaboveexample,supposethetotalcosttobusinessassociatedwithmotorcartraderregulationis$10million,thentheopportunitycosttomotorcartradersistheforegoneopportunitytohaveallocatedthe$10milliontootheruses.Letssupposethatmostof the $10million cost is attributed tomanagement and staff time complyingwith disclosurerequirements. Lets also suppose thatmotor car traders spend twohourson complianceworkratherthantwohoursonrevenuegeneratingactivities:sellingvehiclesorprovidingafter-salesservices.Thetwohoursexpendedonregulatorycompliancerepresentstheopportunitycosttomotorcartraders:potentialforegonerevenueincomefromtheirbusinessactivity.Inrealsimpleterms, theycouldhavesoldavehicletoacustomer,butwere insteadsittingat thedeskdoingcompliancepaperwork.The opportunity cost of regulation involves society givingup something in order to achieve aregulatoryobjectiveandtheassociatedbenefits.Formostregulations,aspecificbusinessgroupincurs the regulatory costs so that society can benefit from safety, environment, consumerprotectionandsoforth.Iftheopportunitycostacrossthetotalstockofregulationsissignificant,thecosttobusinesscanresult in lower productivity, higher production costs and less competitiveness. This canultimatelyaffectinvestmentandemploymentopportunities.Given that regulation has positive (benefits) and negative (costs) impacts and there is anopportunity cost associatedwith regulation, it is important to evaluate the costs to all partiesandtoensurethetotalbenefitsareinexcessofthetotalcoststhatareimposed.Thisistherationaleforusingcostbenefitanalysisastheprimarytooltoundertakeregulatoryanalysis.Costbenefitanalysiscalculatesthetotalcostsandcomparesthesecostswiththetotalbenefits. A qualitative assessment of benefits (and sometimes costs) is still an importantcomponentofthecostbenefitanalysis. Itallowstheidentificationanddiscussionofdirectandintangible benefits to be considered and possibly weighted in a partial cost benefit analysisassessment. Importantly, the cost benefit analysis facilitates informed decision-making on thebest available data. For example, a partial cost benefit analysis needs to provide the cost andbenefit outcome (net cost or net benefit) for those components that could bemonetized anddiscusswhetherthenon-monetizedbenefitsandcostsarematerialenoughtomakeasignificantdifference on the monetized analysis. Even if this discussion concludes that the proposedregulation is likely to generate a net benefit after consideringmonetized and non-monetizedcostsandbenefits,thereisstillaneedtoconsidertheprobabilityoftheregulationachievingthepolicyobjectiveandwhether thebenefitsof theproposedregulationare significant comparedwith other problemswhere the governmentmay have reason to regulate and deliver greaterbenefitstosociety.

Ananalysisoftheaveragecompliancecostperaffectedbusinessisalsoimportantinformationtothe decision-maker. For example, two different regulatory analyses reveal the followingcompliancecosts.Inoneregulatoryanalysis,thebusinesscompliancecostis$50perannumandthesecondregulatoryanalysis,thebusinesscompliancecostis$10perunitthatrepresentsand8%costincreaseperunit.Evenwherethebenefitscannotbemonetized,areasonabledecision-makerwouldnot consider the $50per annumahuge impost onbusiness. But a $10per unitcompliancecostthatledtoan8%costincreaseislikelytoconcernthedecision-makerintermsofwhether the flow-oneffecton consumerprices and impactsonbusiness competitiveness isworththeintroductionoftheregulation,particularlyifthebenefitstosocietyappearsomewhatsmall relative to other comparable regulatory matters. Even if the decision-maker requestedhis/herdepartment to undertake furtherpolicywork to ascertainwhether therewas a lowercost alternative, theobjectiveof the costbenefit analysishasbeenachieved. Ithas facilitatedinformeddecision-making.Inthiscase,thedecision-makerhasdecidedtheopportunitycosttobusiness and consumers appears, prima facie, too great to achieve the policy objective, andwants a fuller exploration of alternative compliance approaches that can deliver a lowercompliancecost.Notwithstanding the difficulties of quantifying benefits, cost benefit analysis is the onlyanalyticalframeworkthatevaluatesthecostsandbenefitstoallparties.Costbenefitanalysisisnotaprecisetoolbutshouldbeseenasaconceptualframeworktoidentifyingallofthepositiveandnegativeimpactsofaregulationandalternativeapproaches.CostBenefitAnalysisFrameworkCostbenefitanalysisfacilitatesinformeddecision-making.Acost–benefitanalysisshouldassessthe costs and benefits of the regulation and the viable options. In most cases, it providesevidence that the benefits of government intervention outweigh the costs and identifies theoptionthatprovidesthegreatestnetbenefittosociety.Insomecases,thecostbenefitanalysiswill also reveal thatnoneof theoptionsprovideanetbenefit to societyand that governmentinterventionisnotwarranted.Thefullrangeofcostsandbenefitsneedtobeidentifiedandwherepossiblequantified.Therearefourstagestocostbenefitanalysis:• Identifythegroupsaffectedbytheregulation• Identifythetypeofcostsandbenefits• Assessmentofthecostsandbenefits• DecisioncriteriaIdentifythegroupsaffectedDuring the policy development stage and/or the drafting of the proposed regulations, theaffectedgroupswillneedtobeidentifiedforthepurposesofevaluatingthecostsandbenefits.The key affected groups will include those persons, businesses, organizations, groups andindustrysectorsthatwillneedtocomplywiththeregulationandthebeneficiarieswillinvariablybethosepartiesthatreceivethegoodsorservicesfromtheregulatedparty.Thesepartiescouldbeconsumersandotherbusinesses for industry specific regulation, and in thecaseofgenericregulationsuchasenvironmentalregulation,thebeneficiarieswouldbethegeneralpublic.

TypeofcostsandbenefitsTherearevariouscostsandbenefitsthatneedtobeconsideredinaregulatoryanalysis:• directandindirectcostandbenefits• intangiblecostsandbenefitsDirect costsandbenefits are closely related to thepolicyobjectiveof theproposed regulationandtheindirectcostsandbenefitsareby-productsoftheproposedregulation.Direct costs include compliance costs to thoseparties that need to complywith theproposedregulation and the administration costs incurred by Government in enforcing the proposedregulations.Indirect costs comprise social and environmental costs to the community and economy-wideimpactssuchasareductioninemployment.Tangiblecostandbenefitsbydefinitioncanbevaluedandinvolveanexplicitmarketprice.Intangiblecostsandbenefitsdonothaveamarketpriceandamarketvariableneedstobefoundtoapproximatetheirvalue.Commonintangiblecostsandbenefitsincludepositiveandnegativeimpactsontheenvironment.The nature of the proposed regulation will determine which of these costs will need to beevaluated.Ataminimum,thedirectcostsandbenefitswouldneedtobeevaluated.AssessmentoftheCostsandBenefitsDirectCostsGenerally, the direct costs of a regulation can be quantified. Each part of a regulation thatimposesanobligationonaperson,businessororganizationtocomplywithaspecificprovisionimposesadirectcost.Eachregulatoryclauseshouldbeappraisedtodeterminewhetheritislikelytoimposeacostandtoidentifytheperson,business,organization,grouporindustrysectorthatwillincurthedirectcosts.Several pieces of data are required to calculate the direct cost. These normally include thecompliancetimeinvolvedandtheassociated laborcostandthecompliancefrequency(one-offorperiodic).Thiscalculationshouldbeundertakenonatransactionbasisandonanaggregatedbasisfortheexpectedtotaltransactionsacrosstheaffectedgroup.The example below is taken from the RIA for the Motor Car Traders Regulations 2008. Theexample details a regulatory obligation, compliance time involved, labor cost and the totaltransactions.Withthisinformation,calculationsareundertakenforthetransactioncompliancecost,theannualcompliancecosttotheindustryandthepresentvalue(discounted)compliancecosttotheindustry.

Box1:MotorCarTradersRegulations2008RIARegulatoryobligationTheregulationsrequiremotorcartraderstorecordinformationinadealingsbookabouttheacquisitionsanddisposalsofmotorvehicles.Theinformationincludesvehicleidentificationnumber,odometerreading,nameandaddressofownerorbuyer,securityinterest(ifany)androadworthinesscertificate.CompliancetimeinvolvedOn-siteconsultationwithmotorcartradersandobservanceofthecompliancetaskrevealedthatthetimeinvolvedforrecordingdetailsaboutamotorvehicleacquisitionwasabout45minutesandforthedisposalofamotorvehicleabout15minutes. LaborCostTo calculate the labor cost associated with the compliance time involved with a regulatoryobligation,weneedtoestablishthehourlyrate. In theabsenceofany industrydataonhourlyrates paid to personnel involved in this compliance task, the analysis has drawn upon theAustralianBureauofStatistics(ABS)privatesectoraverageweeklyearnings.The ABS average weekly earnings are stated at $1,083.29. This is multiplied by 52weeks toobtainanannualsalaryof$56,331Thetotalnumberofweeksworkedperannumneedstoexcludeannualandpublicholidaysandsick leave entitlements. This equates to 44weeks per annum and ismultiplied by number ofhoursworkedeachweek(41hoursperweek).Thisequatesto1804hoursperannum.The$56,331annualsalaryisdividedby1804hoursperannum.Thisequatestoanhourlyrateof$31.23.The $31.23 hourly rate needs to be augmented with any wage oncosts (payroll tax, workerscompensationpremiums,superannuationcharges)andbusinessoverheadcoststoestablishtheactualhourlyrate.The standard salary oncosts is calculated at 16.5% and 50% for overheads. The $31.23 ismultipliedby1.165and1.5andequatesto$54.57androundedto$55.Theworkingsandformulausedtocalculatethehourlyrateisprovidedbelow.ABSAverageWeeklyEarnings-$1083.29multipliedby52weeks=$56,311perannumNumberofweeksworkedperannum52weeksminus4weeks(annualholidays)minus2weeks(publicholidays)minus2weeks(sickleave)=44weeksAverageweeklyhoursforfull-timeworkers=41hours

Oncostmultiplier(payrolltax,workerscompensation,superannuation)=1.165Overheadcostmultiplier(rent,buildingandlandrates,insuranceandothercorporateoverheads)=1.5=$56,331x1.165x1.544x41=$56,3311,804hoursperannum=$31.23x1.165x1.5=$54.57Thishasbeenroundedupto$55forthepurposesofmakingthecalculationsbelow.Totaltransactions Toascertainthetotalcosttothemotorcartraderindustry,weneedtoknowthetotalnumberofmotor vehicle sales for both new and used vehicles. In this case, the Australian Bureau ofStatisticscollectsdataonannualmotorvehiclesales.There are250,000newcar salesper annumandmotor car traderswouldneed to record theacquisition from the manufacturer or importer/distributor and the disposal to consumers(500,000entriesinthedealingsbook).Thereareabout450,000usedmotorcarsalesperannumandmotorcartraderswouldneedtorecord theacquisition from the sellerand thedisposal to thebuyer.Hence, a totalof900,000entrieswouldneedtobeinthedealingsbookacrosstheindustry.Asaresult,atotalofabout1,400,000entries(500,000newcarsalesand900,000usedcarsales)wouldberecordedinmotorcartraders’dealingsbooksacrosstheindustry.For the purposes of calculating the compliance cost, there are 700,000 acquisitions (250,000newcarsand450,000usedcars)and700,000disposals (250,000newcarsand450,000usedcars).CalculationsTransactioncompliancecostWiththeaboveinformation,wecannowcalculatethecompliancecostpersalestransactionforeachacquisitionanddisposal.PerAcquisitiontransaction:$55perhour(laborcost)dividedby45minutes=$41.25PerDisposaltransaction:$55perhour(laborcost)dividedby15minutes=$13.75AnnualComplianceCost

Withthepertransactioncost,wecannowcalculatethetotalcompliancecostforthe1.4milliontransactionsacrosstheindustry.700,000acquisitiontransactionsmultipliedby$41.25=$28,875,000700,000disposaltransactionsmultipliedby$13.75=$9,625,000Totalannualcompliancecost:$38.5millionPresentValueCostThemotorcartraderregulationshavealifeoftenyearsandthecostshavebeendiscountedby3.5%overthistimeperiod.The$38.5millionannualcompliancecostovertenyearsequatesto$385million(undiscounted)andapresentvalueof$320million(discounted).Theaboveanalysisprovidestwousefuloutcomes.Firstly,itprovidesanestimationofthecostsonanindividualtransactionbasis.Thisenablesconsiderationofwhetherthecomplianceburdenisreasonabletakingintoaccountthepurposeoftheregulationandwhetheritislikelytohaveasignificantimpactonthebusinessorbepassedontothepurchaser.Inthiscase,the$55costisconsideredarelativelysmallcompliancecostrelativetotheretailpriceofmostmotorvehiclesandtothegrossprofitmarginonamotorvehicle.Secondly, the analysis provides the total cost in respect to this specific provision on dealingsbookoverthelifeoftheregulations.This is a basic example of compliance cost calculation. Other regulations may require thecalculation of costs associated with equipment andmaterials used as part of the complianceand/or specialist external assistance (for example, legal or accounting advice). The samemethodology as used in the above examplewould be applied for these extensive compliancerequirements.Themethodologycanalsobeusedtocalculategovernmentcoststoadministerregulationsuchasthetimeinvolvedinprocessingandapprovingapplicationsforalicence,permit,registrationetctherenewaloftheaforesaid,inspectionsandaudits.Theactualsalariespaidtogovernmentagencypersonnelwouldbeusedratherthanaverageweeklyearnings.To recap, the compliance time involved is critical and should be obtained from affectedstakeholders that need to complywith the regulation. Private sector averageweekly earningsshould be used in the absence of reliable industry wages data and industry or governmentstatisticaldatashouldbeusedtodeterminethetotalnumberoftransactions.DirectBenefitsA qualitative assessment of the direct benefits was undertaken as the several governmentagencies thataccessamotor car traders’dealingsbookhadnodata.The followingqualitativeassessmentwasprovidedassshowninBox2.Box2:QualitativeassessmentofthebenefitsTheproposedregulationwillenableConsumerAffairsVictoriatoundertakeroutineinspectionsand investigations into consumer complaints, and the Motor Car Traders Guarantee Fund toassess and pay claims to consumers; both organizations are reliant on the prescribed vehicleidentification informationtomatchthemotorvehiclewiththemotorcartraderandthebuyerduring a contractual dispute. Similarly, Victoria Police and VicRoads rely on the prescribedvehicle identification in the dealings book to trace andmatch stolen vehicles being sold at a

motorcartrader’spremisesandtoensuremotorvehiclesareroadworthyandtransferofcleartitlerespectively.Asmostoftheinformationrecordedinthedealingsbookwouldbeundertakenbyamotorcartraderforstockcontrolpurposes,theonlyfeasiblealternativeconsideredwasavariationtotheprescribedrequirementsinvolvinglessinformationasshowninBox3below.Box3:Alternative-LessInformationPrescribedAnalternative istoprescribe less informationthanintheproposedregulation.However,mostmotor car traders would still keep records regarding vehicle identification for stock recordpurposesandenquireintowhetherthereisanysecurityinterestinatrade-inmotorvehicleandamountstobepaidoutondischargeasthiswouldbeintheirfinancialintereststoensurethattheydidnotincuranypotentialfinancialliabilitypriortothesaleofthetrade-inmotorvehicle.Theregulationandthealternativewerecompared.However,asthequalitativeassessmentofthealternativeand theabsenceofquantifiedbenefits required theassistanceofanotherdecision-making in the form ofmulti-criteria analysis – a balanced score card approach. Box 4 belowshowstheapplicationofmulti-criteriaanalysisforassessingtheregulationandthealternative.Box4:Multi-criteriaanalysisNot all of the benefits can be quantified and a net present value cannot be calculated.Accordingly,themulti-criteriaanalysisapproachhasbeenadoptedtocomparethenetimpactsofthealternatives.Inthisanalysis,thecriteriaare:•Increasedconsumerprotection(75percent);•Reducedcoststobusiness(15percent);and•ReducedcoststoGovernment(10percent).Thesecriteriahavebeenselectedonthebasisthattheyreflectthekeycostsandbenefitsdetailedinthealternatives.Weightings are assigned to each of the criteria reflecting their relative importance to theobjectivesofconsumerprotectionandeconomicefficiency.For each alternative, a qualitative score is assigned to each of the criteria, depending on theimpactofthealternativeonthecriteria.Scoresareassignedrelativetothebasecase–either-5iftheimpactisnegative/undesirable/poorand+5ifthereisapositive/desirable/goodimpact.ThefollowingoptionsareassessedintheTablebelow:BaseCase–informationprescribedbytheActi.emotorcartraderstodeterminethecontentofadealingsbookforacquisitionsanddisposalsofmotorvehicles.Information prescribed in the proposed regulations – same as the Act but prescribed entryrequirementsfortheacquisitionanddisposalofmotorvehicles.Lessinformationprescribed–enablestheremovalofsomeunspecifiedentryrequirementsforthedealingsbook.

Table:AssessmentofnetimpactsofalternativesCriteria BaseCase ProposedRegulation LessInformationCriteria Weighting Score Weighted

ScoreScore Weighted

ScoreScore WeightedScore

Increasedconsumerprotection

0.75 0 0 +5 3.75 +3 2.25

Reducedcost tobusiness

0.15 0 0 -5 -0.75 -3 -0.45

Reducedcosts togovernment

0.10 0 0 +5 0.50 +3 0.50

TotalScore 0 0 3.50 2.10

Under the increased consumer protection criterion, information prescribed in the proposedregulationreceivesascoreof+5asitcoversthecriticalaspectsthatneedtobeincludedinthedealingsbook.Ascoreof+3scoreisallocatedtothelessinformationprescribedoptionasmotorcartraderswouldstillmaintainmostoftheinformationprescribedinthedealingsbookasthedata collected is considered normal business practice but somemotor car traders could omitcritical information such as odometer readings. Accordingly, the less information prescribedoptionwouldstillprovideincreasedconsumerprotectionbutnottothefullextentasscoredintheinformationprescribedintheproposedregulation.Underreducedcoststobusiness,theinformationprescribedintheproposedregulationreceivesa -5 score given the compliance costs incurred (notwithstanding that some costs would beincurredaspartofnormalbusinesspractice)andlessinformationprescribedreceivesa-3scoregiventhatsomebusinessescouldchoosetocollectlessdata.Under the reduced costs to Government criterion, information prescribed in the proposedregulationreceivesa+5scoreasitprovidescertaintyaboutthetypeofinformationcollectedbymotorcartraders.Thelessinformationprescribedoptionreceivesa+3scoregiventhatmotorcartraderswouldaspartoftheirnormalbusinesspracticestillcollectmostoftheinformationprescribedintheproposedregulation.The multi-criteria analysis suggests that the most attractive alternative is the informationprescribedintheproposedregulationasthisgivesassurancethatmotorcartradersdocollectallthenecessaryinformationfortheacquisitionanddisposalofmotorvehicles. HowtoquantifythebenefitswhennodataisavailableontheproblemIftheanalysisofthenatureandextentoftheproblemhasnotprovidingsupportingevidenceonthecostsassociatedwiththeproblem, itwillbedifficulttoquantifythebenefitsofaproposedregulationandthealternatives.In these cases, it isworthwhile undertaking comparative research to ascertainwhether othercountrieshaveconductedempiricalanalyzesthatcanadjustedforthelocalsituationinThailand.For example, the Victorian State Government in Australia, recently reviewed its safe drinkingwater regulations. These regulations prescribe mandatory drinking water standards, waterqualityandthefrequencyofsamplingthatmustbeundertakenbywaterauthorities.Thepolicyobjectiveoftheregulationsistoprotectpublichealth.Theregulationshavebeenhighlyeffective

inpreventingtheoutbreakofwaterbornediseases.Accordingly,theRIAwasunabletoprovidedataonthesizeoftheproblem(casesofdeathsandhospitalizationsandtheassociatedcosts).In this regard, quantifying the benefits in the absence of the regulations is difficult withoutappropriate data. TheDepartment ofHealth searched for empirical data fromother countrieswheretherehadbeenoutbreaksofwater-bornediseases(USA,CanadaandSweden)andappliedthese findings to its analysis to the local situation in Victoria. As shown in Box 5 below, theDepartment of Health was able to estimate the incremental benefits that would arise fromprotectingpublichealthfrompreventingtheseoutbreaksofwater-bornediseases.Box5:SafeDrinkingWaterRegulations2015,Victoria,AustraliaQuantifiableincrementalbenefitsTheestimationofquantifiablebenefits inthisRIAisbasedonincrementalcostsavingsarisingfrom protection of public health in the form of reducing the risk of gastroenteritisoutbreaks/cases.In order to establish the incremental benefits under the options the following health andmortalitycostassumptionshavebeenmade:Thesocietalcostofanepidemicoutbreakwouldbe$163.64perpersonin1995prices,basedonaMonashUniversityandANUreportonanoutbreakonatownof11,000people(DepartmentofEpidemiologyetal1997).Thisisequivalentto$267.84perpersonin2014prices.Thecostofadeathtosocietyisbasedonavalueofastatisticallife(VSL),whichrepresentshowmuchsociety iswilling topay to reduce the riskofdeath.TheVSLestimatedemonstrates thefinancialvaluesocietyplacesonreducingtheaveragenumberofdeathsbyoneandisgivenas$3.5millionin2007(OBPR2008).Thisisequivalentto$4,216,724in2014prices.PotentialhealthcostsofanoutbreakCasestudiesfromsimilar(developed)countrieswithinadequatewaterqualityregulationsandmonitoringhelp toprovide themagnitudeofwhatcouldpotentiallyhappenwithanoutbreak.The following specificoutbreaks inMilwaukee,Östersund andWalkerton areusedasexampleswherepeoplebecameillordied.These casesof specificoutbreaksare summarized in theTablebelow.Theaveragenumberofpeople becoming ill with IID (infections intestinal disease) in an outbreak is around 144,100with around 39.31% of the total population affected on average. The risk of death is moreprevalentinthosewithsuppressedimmunesystems.Country/location Study Year Population

affected byIID

Pathogenbehindoutbreak

Populationaffected as% of totalpopulation

Deaths No ofdeaths asa % ofthoseaffected

USA/Milwaukee Mackenzieetal1994

1993 403,000 Crytosporidium 25.00% 69 0.017%

Canada/Walkerton Salvadori etal2009

2000 2,300 Campylobacterjejuni

47.92% 7 0.304%

Sweden/Östersund Widerstrometal2014

2010 27,000 Crytosporidium 45.00% 0 0.00%

Average 144,100 39.31% 25 0.107%

EstimatedcostsofaMilwaukee-typescenarioinaVictoriacontextIntheMilwaukee1993outbreak,whereCrytosporidiumhadcontaminatedthecity’spublic water supply, approximately 64 people or 93% of the 69 deaths in 1993involved people with AIDS. Another waterborne outbreak in 1994 in Las Vegas,Nevada led to thedeathsof41AIDSpatients (Goldstein et al 1996).This isused toconsidertheimpactsonpeoplewithcompromisedimmunesystems.InordertoestimatethecostofmortalityduringanoutbreakinaVictoriansettingthefollowingassumptionsaremade:• ThepopulationservedbythelargestwatersupplierinVictoriaisestimatedtobe

1.74millionandrepresents30.97%ofthetotalpopulationofanestimated5.62million.

• In2011thenumberofAIDSpatientsinVictoriawas2,282.• AIDS patients who died in Milwaukee in 1993 (64) as a proportion of total

populationofAIDSpatientsin1995(653)isestimatedtobe9.8%.• The total population of AIDS patients in Victoria affected by an outbreak is

estimatedtobe69(2,282x30.97%x9.8%=69).• Theprobabilityofanoutbreakoccurringis1.75%.• Theestimatednumberofmortalities fromawaterborneoutbreakis1.22(69x1.75%)• TheVSLasatJune2014isestimatedtobe$4,216,724.The cost of mortality during an outbreak in Victoria is therefore estimated to be$4,216,724 x 1.22mortalities = $5.12million or 4.21million in 2014 present valuedollars.

25

DecisionCriteriaNetpresentvalueWhere a full cost benefit analysis has been undertaken, the future costs andbenefits need to be discounted to determine the net present value. The netpresentvaluemustbepositivei.eNPV>0inorderfortheproposedregulationtomeettheacceptancecriteria.TheformulaandanexampleareprovidedinBox6below.Box6:NetpresentvalueformulaTodetermine thenet present value (NPV) of an option, the costs andbenefitsneedtobequantifiedfortheexpecteddurationoftheproposal.Thenetpresentvalueiscalculatedas:T

ΣNPV=(Bt-Ct)/(1+r)tt=0

whereBt=thebenefitattimetCt=thecostattimetr=thediscountratet=theyearT=numberofyearsoverwhichthefuturecostsorbenefitsareexpectedtooccur(thecurrentyearbeingyearo)Consideranoption thatwill require industry to installnewequipment to limitairpollution.Theequipmentcosts$5millionto installandwilloperate for thefollowing four years. Ongoing (annual maintenance) costs to business are $1millionayear(inconstantprices).Thebenefitsareestimatedat$3millionayear(inconstantprices).Thediscountratesare3percentand5percent.

Costs Benefits Annual netbenefit

Netpresentvalue

(Ct) (Bt) (Bt-Ct) 3% 5%

$m $m $m $m $mYear0 5 -5 -5.00 -5.00Year1 1 3 2 1.94 1.90Year2 1 3 2 1.89 1.81Year3 1 3 2 1.83 1.73Year4 1 3 2 1.78 1.65Netpresentvalue 2.44 2.09

Source:BestPracticeRegulationHandbook(2010)AustralianGovernment

26

Otherdecision-making tool touse in theabsenceofa full cost-benefitanalysisA full cost-benefit analysis (CBA) represents best practice in evaluating theimpact of viable policy options as it gives decision-makers a strong basis forcomparingpolicyalternativesonthebasisofquantifiable(monetary)costsandbenefits.When the benefits (and in some cases the costs) of the policy options beingconsidered cannot be sufficiently or confidently quantified and monetized, apartial cost benefit analysis should still be undertaken with supplementarydecision-makingtoolstoassistincomparingorrankingoptions.Theseinclude:•_break-evenanalysis;•_cost-effectivenessanalysis;and•_multi-criteriaanalysis.Thesedecision-makingtoolsshouldnotbeusedasasubstitute forcost-benefitanalysisbutasanaidtoimproveapartialcostbenefitanalysis.Break-evenanalysisBreak-evenanalysisisusefulwherethebenefitscanbemonetizedbutthereisadegree of uncertainty of whether the benefits are likely to be accrued. Thisrequires estimating the benefits needed to offset the estimated costs. Box 7belowprovidesanexampleoftheuseofbreak-evenanalysis.Box7:ExampleofBreak-evenanalysisAhypotheticalproposalisexpectedtoimprovesafetybyreducingfatalitiesandpreventing injuries and the cost of the proposal can be estimated withreasonable certainty. While there are widely used estimates of the value of astatistical life (VSL)(assumed here to be $4million) and the value of avoidedinjuries,intermsofhospitalizationcostsandlostproductivity(assumedheretobe $250,000 per injury0, there may be no way of confidently and accuratelyquantifyinghowmanyliveswillbesavedandinjurieswillbeavoidedfromtheproposal.It is possible to use this available information to determine how manyfatalities/injurieswouldneed tobe avoided inorder to justify the costs of theproposal, that is for the proposal to ‘break-even’. Various combinations offatalities and injuries prevented would see the proposal break-even. Forexample, if the total cost of implementing and complyingwith the proposal is$13millionperannum,theproposalwouldneedtopreventthreefatalitiesandfour injurieseachyear tobreak-even,usingaVSLof$4millionandthecostofinjuriesof$250,000.Similarly,preventing2fatalitiesand20injurieswouldalsoallowtheproposaltobreakeven.

27

Judgment needs to be exercised to determinewhether such a proposalwouldachievethemagnitudeofbenefitsrequiredtobreak-evengiventhenatureandsizeof thepolicyproblem(e.gdoes theproposal targeta small elementof theproblem?) and the expected practical effect of the proposal (e.g what is theinterventionlogicandwhatbehaviors/activitiesareexpectedtochange?).In this example, if the current level of fatalities is 2 and the current level ofinjuries is 3, then the break-even won’t be achieved. If the current level offatalitiesisinstead30andinjuriesis100,thenitismorelikelytobeachieved.Inthe latter case, the judgment as to whether the break-even points is feasibleshould be supported by objective data, for example based on historical time-series incident data and the counterfactual/baseline, overseas experience, thesafety-relatedoutcomesexperiencedfromasimilarpolicyproposal,oracademicresearch.Source: Victorian Guide to Regulation (2011), Department of Treasury &Finance,VictorianGovernment.CosteffectivenessanalysisCost-effectiveness analysis is usedwhere the benefits cannot bemonetized. Itcompares alternatives on the basis of the ratio of their costs and a singlequantifiedmeasure such as lives saved. Box 8 provides an example of how toundertakecosteffectivenessanalysis.Itisarelativelysimplecalculation.However,costeffectivenessshouldbeusedprudentlyasitdoesnotaddresstheactual benefits (that is, the costs associatedwith the nature and extent of theproblem).Without this information, it is entirely possible that theOption thatprovidesahigherunitcostmayinfacthaveahigherprobabilityofsavinglives.Thiswouldoccurwheretheoptionthathasthelowestunitcostdoesnotaddresstheprimarycausesoftheactualproblem(roadfatalities)buthasbeenassumedthatthisoptionwilladdressaprimarycauseofroadfatalities.Box8:ExampleofCostEffectivenessAnalysisTwo policy options are aimed at reducing road fatalities. Option A costs $20millionandwouldsave10livesandOptionBcosts$15millionandwouldsave5lives.Thecostforeachlifesavediscalculatedbydividingthecostbythenumberoflivessaved($20milliondividedby10lives=$2million)Option A BCost $20million $15millionLivessaved 10 5Costforeachlifesaved $2million $3millionTheanalysisshowsOptionAhasthehighestcostbuthasthelowestunitcostin

28

savinglives,$2millioncomparedwith$3millioninOptionB.ThiswouldsuggestthatOptionAisthepreferredoption.Multi-criteriaanalysisMulti-criteriaanalysis(MCA)canbeausefultoolwhenitisdifficulttoquantifytheimpacts,particularlythebenefitsofaregulationandalternativeapproaches.MCA is a balanced score card approach and requires judgments about howproposedoptionswillcontributetoaseriesofcriteriathatarechosentoreflectthe costs and benefits associated with the proposals. The criteria should beconsistent with the stated policy objectives for the proposal and weightingaccordingtotheirrelativeimportancetothefinaldecision.Aqualitativescorewouldbeassigned,dependingontheimpactoftheoptiononeach of the criteria measured relative to the base case (i.e in the absence ofregulation).Acriterionratingscalefrom-10to10ispreferredasitiseasiertoinclude more information on the choices made, and this results in a greaterunderstandingoftheproposal.Forexampleascoreof10wouldindicatethattheoptionhas twice the impact of anoptionwith a scoreof 5 (and five times theimpactof anoptionwitha scoreof2etc).Forexample, ifoneoption incurredcosts of $3.5million per year, and another option $7million, then the formeroptionmightreceivearatingof-5,whilethelatterwouldscore-10.Thescoreinthis casewouldbenegative as the costs incurred are relative to thebase casewherenocostsareincurredintheabsenceofregulation.Box 9 below provides an example of how to use multi-criteria analysis. Theweighted scores are calculated by multiplying the score by the criterionweighting.ForexampleinBox9,theweightedscoreforOption1inrespecttoareduction in road-related accidents is +4 and is calculated by multiplying thescoreof+10bythecriterionweighting(40%).Thetotalscoreforeachoptionisthesumoftheweightedscoresforeachcriterion.Box9:ExampleofMulti-CriteriaAnalysis(MCA)Toachieveareductioninroadrelatedaccidents,twooptionsmaybeconsideredandevaluatedbasedonthefollowingsimplifiedmulti-criteriaanalysis,withtheassignmentofscoresrangingfrom-10fornegativeoutcomesto+10forpositiveoutcomes relative to the base case. (Outcomes that maintain the status quowouldreceiveascoreofzero). Basecase Option1 Option2Criteria Weighting Score Weighted

ScoreScore Weighted

ScoreScore Weighted

ScoreReduction inroad –relatedaccidents

40% 0 0 +10 +4 +5 +2

Costs ofcompliance

50% 0 0 -5 -2.5 -3 -1.5

29

andadministrationImprovedtrafficflow

10% 0 0 0 0 -10 -1

Total 0 0 +1.5 -0.5

TheassignedscoresindicatethatOption1isconsideredtoreduceroad-relatedaccidents by twice as much as Option 2. Meanwhile, the compliance andadministrative costsofOption1arehigher than forOption2.Option1hasnoexpectedimpactontrafficflow.Inthisexample,Option1 is thepreferredapproachbecause ityieldsapositivescoreof+1.5.Option2,ontheotherhand,returnsanegativeresultof -0.5andwouldthereforebeconsideredtobeanundesirableproposal.When presenting the results of MCA in a RIA, it is important tot providesufficientcommentarytoexplaintheapproach,particularlyintermsofprovidingjustification for the choice of criteria, the weightings of the criteria, and thescoresassignedtothedifferentoptionsforeachofthecriterion.Source:Source:VictorianGuidetoRegulation(2011),DepartmentofTreasury&Finance,VictorianGovernment.ExpectedQualityofRIACostBenefitAnalysisIn the early years of RIA adoption in Thailand, the skill and experience ofgovernmentagencyofficerswillrestricttheirabilitytoundertakesophisticatedcostbenefitanalysis.Evenwhereofficershavetheskill-set,theabsenceofrobustdatawillpreventtheuseofmoresophisticatedanalyzes.With the improvement of data collection strategies over time, governmentagencieswillbeinapositiontoundertakefullcostbenefitanalysisandemploysophisticatedmethodologiesandanalyzes.Intheinterim,itisexpectedthatataminimum,thecostbenefitanalysisshouldcoverthefollowing:• Where the proposed regulation imposes a direct cost (obligation to

comply)onaperson,business,organization,grouporindustrysector,thedirectcosts(compliancecosts)areassessed.

• Where the costs and benefits of the alternatives cannot be quantified, a

qualitativeassessmentshouldbeundertaken.• The cost to government in administering and enforcing the proposed

regulation should be also assessed using a similar methodology to thedirectcosts.

30

• Where the benefits of the proposed regulation cannot be quantified, aqualitative assessment of thebenefits shouldbeundertaken including ananalysis of the likely size of the benefit with some consideration of theweight of each benefit. It will be particularly useful to draw uponcomparable RIAs from other OECD and APEC countrieswhere they havebeenabletoquantifythebenefitsandtoadjustthesequantifiedbenefitstolocalconditionsinThailand.

• The use of other decision-making tools such as break-even analysis, cost

effectiveness andmulti-criteria analysis should be usedwhere a full costbenefitanalysishasnotbeenabletobeundertaken.

ThisminimumstandardisasignificantimprovementcomparedtowhatexistedpriortotheintroductionoftheseGuidelines.Importantly,aRIAproducedusingtheminimumstandardwillenableinformeddecision-makingbygovernment.Inparticular,itshouldprovideaclearindicationofthecompliancecosttodirectlyaffectedstakeholdersandthecosttogovernmenttoadministerandenforcetheregulations.Wherethecostsassociatedwiththeproblemcannotbequantifiedandhencethepotential benefits cannot be quantified, the RIA also provides importantdecision-making information that thegovernmentdepartmentdoesnothaveagood understanding of the problem, and in some cases, any surety that theproposed regulation or other options are likely to achieve the policy objectiveanddeliveranetbenefittosociety.Inthesecases,thecostbenefitanalysisintheRIAenablesthedecision-makertoerron the sideof cautionand request that further research is requiredon thesize of the problem, the associated costs and the likely benefits thatwould bedelivered before making a commitment to introduce the regulation. Such anoutcome is probable where the cost benefit analysis has revealed significantdirect costs to affected stakeholders that could affect the cost of goods andservices to consumers and/or business competitiveness, investment andemployment opportunities. Another issue of concern may be where the costbenefit analysis reveals significant budgetary costs to Government inadministeringandenforcingtheregulation.Onceagain,thedecision-makermaywant further research and evidence that the policy objective can be achievedwithanetbenefittosociety.Accordingly, the absenceofquantificationof thebenefits should facilitateovertime improvements to the quality of data collection strategies withingovernmentsothatfullcostbenefitanalysiscanbeundertaken.

31

6.RegulatoryImpactAnalysisTemplateREGULATORYIMPACTANALYSISTEMPLATERegulationTitle MinistryorRegulatoryBody ExecutiveSummaryNoticeforSubmissionsSection1:IdentificationoftheProblemWhoisaffectedbytheproblem?Whatisthescopeandscaleoftheproblem?Istheidentifiedproblempartofalargerproblem?Ifso,whatisthesizeoftheidentifiedproblemrelativetothelargerproblem?Istheresufficientempiricalevidencethataproblemexists?Istheextentoftheproblemidentifiedorisitsidentificationbasedonanecdotalevidence?Whatisthecauseoftheproblem?Whataretheeconomic,socialandenvironmentalcostsoftheproblem?Doestheproblemexistcurrently,orisitmerelyanticipated?Istheproblemaminorirritantorasignificanthazard?Arethereanytechnological,economic,political,administrative,socialand/orenvironmentalconstraintsthatarerelevanttotheproblem?Arethereexistingregulationsthatcoulddealwiththeproblem?Ifyes,whyaretheseregulationsinadequate?Whataretheconsequencesofnottakinganyaction?Couldrelyingonthemarketinconjunctionwiththegeneralapplicationofexistinglawsandregulationssolvetheproblem?Ifnot,whynot?Whatistheexperienceinotherjurisdictionswithdifferentregulatoryframeworks?Whatisthescopeandscaleoftheprobleminunregulatedjurisdictions?Hasthescopeandscaleoftheproblemincreased/decreasedinunregulatedjurisdictionsduetomarket,technological,regulatoryorenvironmentalchanges?

32

Section2:ObjectivesWhatarethepolicyobjectives?Section3:OptionsWhatarethepossibleregulatoryandnon-regulatoryoptionsthatmeetthepolicyobjectiveandsolvetheproblem?Dependingontheoption,thefollowingquestionsmayneedtobeconsideredanddiscussed:Howwouldthealternativework?Whatroledoesgovernmenthave?Istheresufficientcommonalityofinterest,withinanindustryorprofessionalassociationtoensurehighlevelsofvoluntarycompliance?Howwillconsumerinterestsberepresented?Doesthealternativediscriminateagainstpersons/groups/industries?Isthealternativelegallyfeasible?Doesthealternativerestrictcompetition?Whatmonitoringwillberequiredandhowwouldmonitoringoccur?Isthealternativelikelytobeenforceable?Willnon-compliancebeevident?

33

Section4:AssessmentofOptionsStepsinpreparingafullcostbenefitanalysis:1.Specifythesetofoptions2.Decidewhosecostsandbenefitscount3.Identifytheimpactsandselectmeasurementindicators4.Predicttheimpactsoverthelifeoftheregulatoryproposal5.Monetize(attachbahtvaluesto)impacts6.Discountcostsandbenefitstoobtainpresentvalues7.Computethenetpresentvalueofeachoption8.Performsensitivityanalysis(differentrangeofdiscountrates)9.Conclusion(comparativeanalysisoftheoptions)Partialcostbenefitanalysis:Whereafullcostbenefitanalysiscannotbeundertaken(mostlyduetosomeorall of the benefits not being able to be monetized), the costs of each optionshouldstillbemonetizedandcomplementedwithotherdecision-makingcriteriasuchascosteffectivenessormulti-criteriaanalysis.Forfullandpartialcostbenefitanalysis,discloseanyassumptionsthathavebeenused for monetizing/quantifying costs and benefits, and the basis for thoseassumptions.Theanalysisshouldmaketransparenttothereaderhowcostandbenefitvalueshavebeencalculated.Forcomplexcalculations,itmaybeusefultoincludethisinformationinanappendix.Section5:ConsultationSection6:ReferencesSection7:AppendicesSection8:ProposedRegulation

34

7.CaseStudyEach reader will interpret the expected level of analysis required for the keyparts of a RIA differently. This will lead to varying levels of quality RIA. Toobviate this, these Guidelines showhow to apply the key parts of a RIA to anactual case study to demonstrate the level of analysis and the critical thinkingrequiredtopreparearobustandhighqualityRIA.The case study relates to the problem of road traffic fatalities and injuries inThailand;specificallythegovernmentdecisiontobanchildrenlessthansixyearsofagefrombeingtransportedonamotorcycle.It should be noted that the use of different types of data sets (fatalities andinjuries,population,costsassociatedwithfatalitiesandinjuries,costinputssuchasaveragemonthlywages,faresforalternativemodesoftransport,etc)thatareusedtogethertomakecalculationsintheassessmentofcostsassociatedwiththeproblemandtheassessmentofthecostsandbenefitsfortheregulatoryproposaland alternatives, should cover the same period of time to ensure accuracy.Otherwise,thecalculationscouldunder-stateoroverstatethecostsandbenefits.Forthepurposesofthiscasestudy,mostofthedifferentdatasetsarefor2010.However,otherdatasetsarefromdifferentyearsandthisaffectstheaccuracyofthecostsandbenefits.ThereadershouldnotbeoverlyconcernedwiththisissuebutfocusonthelevelofanalysisandthecriticalthinkingthathasbeenusedtodeveloptheRIA.KeypointsRoad fatalities and injuries are a significantproblem inThailand.Anumberoforganizations have campaigned to save children frombeing killed and injuredwhilst being transported on a motorcycle. Some of these organizations claimseveralthousandfatalitiesassociatedwiththisactivity.It is importanttoverifytheextentoftheproblem.The case study also demonstrates the need to analyze the size of the problemrelative to the affected population. A risk analysis reveals the probability of achild fatality and injury relative to the size of the child population and alsomotorcycle usage based on vehicle kilometers traveled per annum. ThisinformationisimportantforGovernmentinweighingupwhethertheallocationofscarceresourceswithintheeconomyshouldbeappliedtothisproblemortoanother part of the road fatality and injury problem thatmay provide greaterbenefitstoimprovingroadsafety.Thecostbenefitanalysisrequiresanalysisoftheincrementalcostsandbenefits.That is, theadditional costsandbenefits incurred in theabsenceof regulation.Quantifyingthecostsofbehavioralregulationcanbechallenging.Inthiscase,itis importanttothinkaboutthereasonsanaffectedgroupusesamotorcycle.Inthis case, parents use a motorcycle to transport their child with them to go

35

shopping,work(insomecases),pre-school,healthcentres,visitingrelativesandfriends,recreation,religiousetc.Thecostofabanisnotjusttherestrictiononthefreedomofparentsdependentonmotorcyclestotransporttheiryoungchildrentotheseactivities.Howmanychildrenandparentsorfamilymemberswillneedtousealternativemodesoftransport?While it is likely to be difficult to obtain data on the extent of young childrenbeingtransportedonamotorcycle,considerationneedstobegiventoanactivitywheretheparenthasnochoicebuttouseanalternativemodeoftransport.Forexample,mostchildrenfrom3to5yearsofageattendpre-school.Howmanychildrenattendpre-school?Howmanyparentsaredependentonamotorcycle as their primary mode of transport? How will parents send theirchildren to pre-school if they cannot use amotorcycle? Is their homewithinwalking distance of the pre-school centre? Or do they need to take a bus orminivan,orataxi?Will it take longertowalktoapre-schoolcentrecomparedwithamotorcycle?What is the average time difference between these two modes? What is theopportunity cost (potential income forgone) of the parent or other familymemberthatmayhavetospendmoretimewalkingtoapre-schoolorabusstopcomparedtowhentheytraveledonamotorcycle?Whatistheaverageoperatingcostofamotorcycle?Isthishigherorlowerthanthecostofafareforabusortaxi?Thedifferenceincostsbetweenmotorcyclesand alternative modes of transport is the incremental cost in the absence ofregulation(ban).Howwillabanaffectmotorcycletaxis?Aretherelikelytobeimpactsonrevenueandemployment?Arethereanyunintendedconsequencesofaban?Does walking on sidewalks pose a greater risk than being a passenger on amotorcycle?Can the current pedestrian infrastructure cope with an influx of children andparentswalkingtopre-school?Willitcausetrafficcongestion?Does thecurrentpublic transportsector (busesandtaxis)have thecapacity totransportadditionalchildrenandparentstoandfrompre-school?Arethereareas thathave limitedornopublic transportoptions?Remoteruralareas?Ifso,howmanyparentsmaynotsendtheirchildrentopre-school?Canlow-incomefamiliesaffordtheadditionalcostsofpublictransport?

36

Will theban lead to some familieswithdrawing their children frompre-schoolduetolimitedaccesstopublictransportoptionsand/oraffordabilityissues?Will thereductionofmotorcycleusage leadtoareduction in trafficcongestionandmotorcycleemissions?Howwillthecostsassociatedwithfatalitiesandinjuriesbevalued?Willtheavoidedcostsoffatalitiesandinjuriesbeachievable?Whatiffamiliesrefusetocomplywiththeban?Dothepolicehavethecapacityandresourcestoenforcetheban?These are the type of questions that need to be asked as part of the criticalthinkingbehindthepreparationofaRIA.

37

ObjectivesThe objective of the proposed regulation is to prevent children less than sixyearsofagefrombeingkilledorinjuredasapassengeronamotorcycle.NatureandExtentoftheProblemOverviewoftheRoadSafetyProblemThailand has one of the worst road safety records. Thailand’s total trafficaccidentcostswereestimatedat232.8billionbahtor2.81percentageofGDP.1AsshowninTable1,Thailand’sroadfatalitiesincreasedmarkedlyfrom2,104in1987topeakat16,727in1995anddeclinedto12,858by2005.Table1:TrafficAccidentsinThailandfrom1987to2005

Bangkok(NoofCases) Regional(NoofCases) National(NoofCases)YearAccident Fatality Injury Accident Fatality Injury Accident Fatality Injury

1987 19,745 752 6,333 4,387 1,352 2,256 24,132 2,104 8,5891988 31,175 817 9,565 4,114 1,198 3,939 35,289 2,015 13,5041989 31,709 917 10,005 6,388 4,451 3,076 38,097 5,368 13,0811990 33,064 949 10,701 7,417 4,816 7,551 40,481 5,765 18,2521991 38,355 1,057 10,778 7,946 5,276 8,777 46,301 6,333 19,5551992 46,743 983 11,025 14,586 7,201 9,677 61,329 8,184 20,7021993 64,006 1,011 11,031 20,886 8,485 14,299 84,892 9,496 25,3301994 72,359 1,290 18,849 30,251 13,856 24,692 102,610 15,146 43,5411995 64,469 1,284 21,697 24,898 15,443 29,021 94,362 16,727 50,7181996 60,308 1,069 23,314 28,248 13,336 26,730 88,556 14,405 50,0441997 54,324 903 20,933 28,012 12,933 27,828 82,336 13,836 48,7611998 46,800 732 18,920 26,925 11,502 33,618 73,725 12,234 52,5381999 37,868 594 17,104 29,932 11,446 35,434 67,800 12,040 47,7702000 43,485 1,582 23,368 30,252 10,406 29,743 73,737 11,988 53,1112001 45,711 1,519 22,854 31,905 10,133 31,106 77,616 11,652 53,9602002 48,507 1,734 23,488 43,116 11,382 45,825 91,623 13,116 69,3132003 46,806 1,491 23,597 48,386 11,718 50,555 107,565 14,012 79,6922004 55,381 865 23,597 69,149 12,901 70,297 124,530 13,766 94,1642005 - - - - - - 122,040 12,858 94,364Source:DepartmentofHighways,“TheStudyofTrafficAccidentCostinThailand”,FinalReport,Faculty of Engineering, Prince of Songkla University, September 2007. Note datawas sourcedfromtheRoyalThaiPoliceandBureauofTrafficSafety,DepartmentofHighways.However,theofficialgovernmentdatawouldappeartounderestimatethesizeoftheproblem.TheWorldHealthOrganizationestimatesamuchhighernumberoffatalitiesasshowninTable2.Table2:RoadtrafficdeathsinThailand(2010)Estimatedroadtrafficdeaths Estimatedroadtrafficdeathrate(per100,000

population26,312 38.1Source:WorldHealthOrganisation–GlobalHealthObservatoryDataRepository

1DrPichaiTaneerananon,“TheStudyofTrafficAccidentCostsinThailand”powerpointpresentation,web.worldbank.org

38

The Department of Highways in its “The Study of Traffic Accident Cost inThailand”, (2007) noted under-reporting of traffic accidents, fatalities andinjuries due to police not attending all traffic accidents. Hospital records aremorelikelytoprovideamoreaccuratepictureoftheextentoftheproblem.Asa resultof theunder-reporting, thesizeof theproblemwillbedocumentedrangingfromtheminimumsizeoftheproblem(officialrecords)tothemaximumsizeof theproblem(basedonWHOdata).Thecostsof trafficaccidentswillbecalculatedforthisrangeofdata.Motorcyclistsrepresent74percentofroad fatalities(Table3)andmotorcycles61 percent of registered vehicles (Table 4). While the data highlights thatmotorcycle riders comprise themost road fatalities, thedatadoesnotprovideanyinsightintotheagedistributionofthefatalities.Table3:Deathsbyroadusercategory(2010)Typeofroaduser Percentageofdeaths NumberofdeathsRiders motorized 2 or 3wheelers

74% 10,187

Pedestrians 8% 1,101Passengers 4 wheeled carsandlightvehicles

7% 964

Drivers 4 wheeled cars andlightvehicles

6% 826

Cyclists 3% 413Drivers/passengers heavytrucks

1% 138

Drivers/passengersbuses <1% 100Other 1% 138Total 13,766Source:WorldHealthOrganization–ThailandCountryProfile2013Table4:Totalregisteredvehicles(2010)Cars and 4 wheeled lightvehicles

9,887,706 35%

Motorized2and3wheelers 17,322,538 61%Heavytrucks 816,844 3%Buses 137,943 <1%Other 319,798 1%Totalregisteredvehicles 28,484,829Source:WorldHealthOrganization–ThailandCountryProfile2013

39

Howsignificantistheproblem?Whatisthemagnitudeoftheproblem?ProportionofMotorcycleFatalitiesthatarechildpassengersAfurtherbreakdownof thehigh-leveldata is required to identifyandquantifythe number of children less than six years of age killed and injured as amotorcyclepassenger.TheWorldHealthOrganization(WHO)citesastudybasedondatafromatraumaregistry at the Khon Kaen Regional Hospital in the northeast of Thailand thatshowedchildren0to5yearsand5to9yearsaccountedfor1.8percentand3.9percentrespectivelyofthemotorcycleaccidentpatientstreatedatthehospital.TheWHOalsocitesdatafromtheAsianDevelopmentBank(2004)inrespecttoagedistributionoftrafficfatalitiesinThailand.ThisdataisshowninTable5andthe percentage of child fatalities is similar to the data from the Khon KaenRegionalHospital. Table5:AgedistributionoftrafficfatalitiesinThailandAge(years) Fatalitiespercent<5 1.65-9 1.810-14 2.715-40 60.4>40 33.5Source:WorldHealthOrganization–basedondata fromAsian Development Bank The status of road safety in Thailand. Manila: Asian Development Bank; 2004. Report No.: Country Report: CR 09.Anotherstudywasconductedinto214fatalmotorcycleaccidentsfromautopsyreportsperformedatRamathibodiHospitalinBangkok(responsiblefor9outofthe50metropolitandistricts)from2003to2006.Inthiscase,10or4.7percentof the214 fatalmotorcycle accidentswere to children less than15 years. Thedataanalysisdidnotprovideanyfurtherbreakdownofthisagegroup.OtherjurisdictionswithasimilarprofileBenchmarkingothercountrieswithasimilarprofilewheremotorcyclesarethepredominantmodeoftransportandmotorcyclefatalitiesrepresentmostoftheroad tollmay assist in verifying the above data. AWHO report onmotorcyclesafety for South East Asian countries revealed Indonesia andBangladesh havesimilar profiles to Thailand. In these countries, road traffic injuries ofmotorcyclistscompriseareported25to70percentofthetotalvictims.Ofthesevictims,childrenlessthan10yearsappeartorepresent2to3percent.WhileAustraliadoesnothaveasimilarprofiletoThailand(childrenlessthansixyears of age do not travel onmotorcycles), it is still worth benchmarking thenumber of fatalities for children given that Australia has a strong road safetyrecordtoseewhereThailandstandsincomparison.Australiakeepsrecordsfor

40

children less than 16 years of age. In 2010, children less than 16 years of ageaccounted for 52 passenger (motor vehicles) fatalities or 3.85% of the 1,352roadfatalitiesinAustralia.2TheThailandstudiesaresummarizedinTable6below.Itisreasonablebasedonthisevidenceand theWHOstudyonSouthEastAsiancountrieswitha similarprofiletoThailandtoconcludeatleast2percentofmotorcyclefatalitiesinvolvechildpassengerslessthansixyearsofage(giventhattwoofthestudiesshowedalmost2percentforchildrenlessthanfiveyearsofage).Thenumberofchild fatalities, seriousandslight injurieswillbedetermined inthenext sectionbasedon2percent ofmotorcycle fatalities, serious and slightinjuries.Table6:SummaryofagedistributionoftrafficfatalitiesstudiesAge(years) Khon Kaen

RegionalHospitalADB(2004) Ramathibodi

Hospital<5 1.8 1.6 -5-9 3.9 1.8 -10-14 4.7NumberofchildfatalitiesandinjuriesCalculating2percentofthenumberoffatalitiesfromTables3&4(officialdataandWHOestimationsrespectively),Table7belowshowschildrenlessthansixyears old accounted for an estimated 204 fatalities or 2% of the 10,187motorcyclefatalitiesin2010andanestimated389fatalitiesor2percentofthe19,471motorcyclefatalitiesin2010.Table7:Numberofchildfatalitiesin2010 Totalmotorcyclefatalities Childfatalities-2%oftotal

fatalitiesOfficialData 10,187 204WHOestimations 19,471 389Note:TheWHO estimation of 26,312 fatalities has been adjusted to reflect the 74 percent or19,471ofmotorcyclefatalities.Thenumberofseriousandslightinjurieswascalculatedbasedondetaileddatashown in Appendix 1. This data showed a ratio of one fatality for every 13serious injuriesand39slight injuries. Table8showstheestimatednumberofchildfatalities,seriousandslightinjuries.Table8:EstimatedNumberofChildFatalities,SeriousandSlightInjuries Officialdata WHOestimatesFatalities 204 389SeriousInjuries 2,652 5,057SlightInjuries 7,956 15,171

2 Department of Infrastructure, Transport and Regional Economics, “Road Deaths Australia”2011StatisticalReport,AustralianGovernment.

41

Whatisthenatureoftheproblem–whatistheloss,harmorotheradverseconsequencethatisbeingexperienced,andbywhom?CostofchildfatalitiesandinjuriesThehumancapitalcostmethodologywasusedtocalculatethecostsassociatedwithchildfatalitiesandinjuries.The human capital costmethodology comprises three cost categories: human,propertydamageandgeneralcrash.The human costs category covers loss of productivity, quality of life, medical,EMSandlongtermcare.Thepropertydamagecostscategorycoversvehicleandnon-vehicle damage costs. The general crash costs category covers insuranceadministration,policeadministration,judicialsystem,ERSandtraveldelay.Table9showsthevalueofcostsperfatality,seriousinjuryandslightinjury.ThecostcomponentforeachcostcategoryisprovidedforeachtypeofcrashseverityinAppendix2.Table9:ValueofcostsaccordingtocrashseverityforThailandin2007CrashSeverity Averagevalueofcosts(baht)PerFatality 5,315,556PerSeriousInjury 147,023PerSlightInjury 34,761Source:DepartmentofHighways“TheStudyofTrafficAccidentCostinThailand”,(2007)Table10showsthetotalcostsforchildfatalities,seriousandslightinjuries.ThecostsarecalculatedbymultiplyingthenumberforeachcrashseveritycategoryinTable8bythevalueoftheappropriatecrashseveritycategoryinTable9.Forexample, 204 child fatalitiesby $5,315,556baht= $1,084,373,424baht and soforth.Table 10 also shows the total cost ranges from 1,750,836,936 baht (based onofficialdata)to3,338,605,726baht(basedonWHOestimations).Table10:CostsofChildFatalities,SeriousandSlightInjuriesCrashSeverity Cost(baht)basedonofficial

dataCost (baht) based onWHOestimations

Fatalities 1,084,373,424 2,067,751,284SeriousInjury 389,904,996 743,495,311SlightInjury 276,558,516 527,359,131Total 1,750,836,936 3,338,605,726

42

In thecaseof risk,what is the likelihoodof theadverseeventoccurring?Whatevidencedoyouhavetosupportthisinitialassessment?RiskofchildfatalitiesandinjuriesOECD countries calculate fatalities per 100,000persons, per 10,000 registeredvehiclesandper100millionvehiclekilometrestraveled(VKT).Theseindicatorsmeasure the rate and relative risk of road fatalities taking into accounthuman/vehiclepopulationandtrafficvolumes.The aforementioned fatality rates provide a general indication of risk fornationalandprovincialregions.Thefatalityratesaremoremeaningfulifappliedtothespecificroadlocationswherefatalitiesoccur.NumberofchildfatalitiesandinjuriesrelativetothetotalchildpopulationIt isimportanttomeasurethenumberofchildfatalitiesandinjuriesrelativetothetotalchildpopulationinThailandtoascertaintherelativerisk.Childrenlessthansixyearsofagecomprise6.5millionor10percentofthetotalpopulationof65million.3Withthispopulationdata it ispossibletoestimatethenumberofchildrenthatarelikelytobetransportedonamotorcycle.Giventhat61percentofregisteredmotor vehicles aremotorcycles, it is conceivable that up to 4million children(6.5million*61%)couldbepotentiallytransportedonamotorcycle.Using the official andWHO estimation fatality data and child population data,Table11shows5.1to9.7childfatalitiesper100,000populationofchildrenlessthansixyearsofage.4Theseriousandslightinjuriesper100,000populationisalsoprovidedinTable11below.Table 11: Child Fatalities, Serious and Slight Injuries per 100,000population2010CrashSeverity Officialdata WHOestimationsFatalities per 100,000population

5.1 9.7

Serious Injuries per100,000population

66.3 126.4

Slight Injuries per 100,000population

198.9 379.3

Table12belowshowsthatchildfatalitiesper100,000ofthechildpopulationareconsiderably lower than the fatality rate per 100,000 for the rest of the3NationalStatisticalOffice(web.nso.go.th)2005census.4Calculation:–4million/100,000=40.Hence204fatalities/40=5.1and389fatalities/40=9.7)Similarcalculationswereundertakenforseriousandslightinjuries.Itshouldbenotedthatusing2005populationdatawith2010fatalityandinjurydatahasresultedinaslightover-estimationoftheapplicablerates.

43

population. For example, 5.1 compared with 22.2 for the official data.5 Thisstronglysuggeststhatparentsandothercaregiversaregenerallyrisk-averseandtakeconsiderablecarewhenridingamotorcyclewithayoungchildaboard.TheOECD median fatalities per 100,000 population has been included as abenchmark for the Official and WHO fatality data. However, it is not directlycomparabletothechildfatalityrate.Table12:ComparativeFatalitiesper100,000population2010Official–totalroadfatalities 22.2OfficialChildfatalities 5.10WHO–totalfatalities 42.5WHOChildfatalities 9.70OECDMedianfatalities 6.20Note: OECD median fatalities cited from Department of Infrastructure and Transport,“InternationalRoadSafetyComparisons2010”StatisticalReport,AustralianGovernment.Themaponthenextpageprovidesthefatalityrateper100,000populationforthe76provincesinThailand.Thefatalityrateisthegeneralrateforallroaduserfatalities including child fatalities. Provincial data is based on 2005 from theDepartmentofHighways“TheStudyofTrafficAccidentCostinThailand”(2007).ThedetaileddataisprovidedinAppendix3.Ideally,2010provincialdatashouldbeusedtobeconsistentwiththepreceding2010data.Notwithstandingthis,themaphighlightsthesignificantdifferencesinthefatalityrateper100,000populationacrossprovinces.Thisisnotaperfectindicationofriskasasmallpopulatedregionmayhaveahighfatalityrateduetootherfactors(hightransitoryroadtrafficthroughtheregion).Thereare10provinceswith fatalityratesgreater than30.Theseareshown inred numerals. Several provinces around and includingBangkok have some ofthelowestfatalityrates.Further investigation is required to understand the differences in the fatalityrates betweenprovinces including road design, environment, volume of trafficetc.

5TheofficialandWHOfatalitydataandpopulationdatahasbeenadjustedtoexcludechildfatalityandchildpopulation.Forexample,Officialfatalities13,562(13,766totalfatalities-204childfatalities)/610(61millioni.e65milliontotalpopulation-4millionchildpopulation)=22.2fatalitiesper100,000population.

44

45

RiskAnalysis–FatalitiesandInjuriesper100,000millionVKTThe average number of personal vehicle VKT6 is multiplied by the childpopulation to determine the total number of VKT. Using the number of childfatalities,seriousandslightinjurydatafromTable8,therelevantratesper100millionVKThavebeencalculatedasshowninTable13below.Table13:ChildFatalities, seriousandslight injuriesper100millionVKT(2010)CrashSeverity Officialdata WHOestimationsFatalitiesper100millionVKT

1.8 3.5

Serious Injuries per 100millionVKT

23.4 45

Slight Injuries per 100millionVKT

70.8 135

Table 14 shows a considerable lower fatality rate for children passengers onmotorcycles compared to all other motorcycle fatalities This is similar to theresults inTable12Comparativefatalitiesper100,000populationandprovidesfurtherevidencethatparentsandothercaregiversaregenerallyrisk-averseandtakeconsiderablecarewhenridingamotorcyclewithayoungchildaboard.Table14:Comparativechildandmotorcyclefatalitiesper100millionVKTCrashSeverity Officialdata WHOestimationsChild Fatalities per 100millionVKT

1.8 3.5

Motorcycle Fatalities per100millionVKT

26.7 51

In 2010, OECDmedian fatalities were 0.54 per 100million VKT traveled andapplied to fatalities for all ages groups. 7 Obviously, the median fatality ratewouldbeevenlowerthan0.54forchildrenlessthansixyearsofage.

6AnAnalysisofVKTofMajorCitiesinThailand(2010)measured2810VKTforpersonalvehiclesfortheNakhonRatchasimaprovince.IthasbeenassumedthisisrepresentativeofallprovincesotherthanBangkokthathasahigherVKT.7OECDmedianfatalitiescitedfromDepartmentofInfrastructureandTransport,“InternationalRoadSafetyComparisons2010”StatisticalReport,AustralianGovernment.

46

Whataretheprimarycausesoftheproblem?PrimaryCausesforChildFatalitiesandInjuriesRoad safety literature has demonstrated that there are many differentcontributingfactorsinvolvedincrashes.Thesearecategorizedasenvironmental,humanandvehiclefactors.Road safety empirical studies analyze these factors and seek to determine thekey contributing factors that cause specific type of crashes and recommendappropriatecountermeasurestopreventthesecrashes.Data was recorded for 214 fatal motorcycle accidents from autopsy reportsperformedatRamathibodiHospitalinBangkok(responsiblefor9outofthe50metropolitandistricts)from2003to2006.Thedatacomprised:• age,• gender,• ridingposition,• timeofaccidents,• typeofcrash-singlevehiclecrash(SVC)andmultiplevehiclecrashes

(MVC)• crashobjects• alcoholconsumptionlevels• causesofdeathTable15:PersonalCharacteristics Number(%)

Male 188(87.9)GenderFemale 26(12.1)Rider 183(85.5)RidingPositionPassenger 31(14.5)<15 10(4.7)15-24 96(44.9)25-34 65(30.4)35-44 24(11.2)

Age,years

>45 19(8.9)Age(years);mean±SD(range):27.4±10.76(3-69)Thisstudyfoundmostmotorcyclefatalitiesweremaleriders,15to34yearsofage,alcohol-relatedandoccurredfrom9pmto6.00am.Thestudydidnotseektofindthecausesforchildfatalitiesthatoccurredwhilston amotorcycle.However, the time formost of the high risk accidents occurswhenmostchildrenlessthansixyearsofagewouldbehomeandasleep.

47

Youngmalemotorcyclistsarethehighestriskgroupinmostcountriesincludingin Australia as demonstrated in the study, “Analysis of High Risk and HighSeverity Groups among Motorcyclists”, Monash University Accident ResearchCentre-Report#77-1995MotorcycleaccidentsinotherjurisdictionsThe other vehicle is commonly at fault in multi-vehicle crashes involvingmotorcycles. In an analysis of 900 motorcycle accidents in Los Angeles Hurt,Oullet and Thom (1981) found that the most common motorcycle accidentinvolved another vehicle (75%) causing the collision by violating the right-of-wayof themotorcycleatan intersection,usuallyby turning left in frontof theoncomingmotorcycle.InVictoria,motorcyclistsarecommonlythevehiclegoingstraightaheadinright-turncrashes,beingintherearinrear-endcrashesandintheongoinglaneinsideswipes.8TheThailandAccidentResearchCenter(TARC) isundertakingduring2014/15an in-depth study of the main types of motorcycle accidents to determineappropriatecountervailingmeasures.Detailed analysis of the causes of accidents is lacking in Thailand due toinadequatedata.Detaileddataonchildfatalitiesandinjurieswouldrequirethecollationofcrashcharacteristicssuchas:CrashVictimsAge–segmentedagegroups;Gender–MaleorFemale;X–driver,passengerandpedestrian;Crashseverity–fatality,seriousinjury,minorinjuryandpropertydamage;Location-capitalcity,otherurban,ruraltownshipsandotherruralcategories;DateandTimeofDay;Weather Conditions - dry, wet, slippery from rain precipitation, and frozencategories;Distance from crash site to the home address of victims (only for nationalcitizens);CausalfactorsHuman causal factors - impairment from alcohol/drug use, driver fatigue andunlicensedcategories;

8Haworth.N,Symmons.M&Kowadlo.N,“HazardPerceptionbyInexperiencedMotorcyclists”,MonashUniversityAccidentResearchCentre,ReportNo.179,Dec2000.

48

Vehiclecausalfactors–vehicleageandvehicledefects(tyres,brakes,suspensionandotherdefectscategories)categories;CrashTypeCrashType–crashesthatinvolvedapedestrian,crashesthatoccurredbetweenvehiclesapproachingfromadjacentdirections(intersectionsonly),crashesthatoccurred between vehicles traveling in opposing directions, crashes thatoccurredbetweenvehiclestravelinginthesamedirection,crashesthatoccurredwhile a vehicle was manoeuvring, crashes that occurred while a vehicle wasovertaking, crashes between a vehicle and an obstacle in the path of travel,crashes that occurred when a vehicle left a straight roadway, crashes thatoccurredwhenavehicleleftacurvedroadwayandmiscellaneouscrashes;RoadCharacteristicsRoadCharacteristics-intersectionwithouttrafficlights,intersectionwithtrafficlights, midblock (section of road between intersections) and roundaboutcategories;RoadType1-dividedroadandundividedroadcategories;RoadType2-sealedroadandunsealedroadcategories;RoadType3–straightroad,curvedroadandslopingroadcategories;RoadCondition-goodanddamaged(potholes)categories;Road Infrastructure - no pedestrian pavement, pedestrian pavement withbuildings abutting pavement (no escape area), pavement with roadside area,clear roadsidewith run-off area, roadsideareawith fixedobjects (trees, poles,bridges,fencesetc)categories;SpeedZone–speedlimitcategoriesWith the collation and analysis of the range of variables used, it is likely thatpatternswillemergeandtheremaybeaneedtoidentifysegmentsofthecrashpopulationwhereasubsetofcrashdatamaybemoreappropriatetoconsider.

49

How is the problem currently regulated? Are there deficiencies in theexistingregulatorysystemthatmightfixtheproblemifcorrected?CurrentregulationoftheproblemThailand law requires motorcycle riders and passengers to wear motorcyclehelmets.Motorcyclehelmetsareahighlyeffectiveroadsafetyinterventionthatreduces the frequency and severity of head injuries resulting from trafficcrashes. TheWorldHealth Organisation cites the Cochrane review that claimshelmet use reduces the risk of motorcycle injuries by 69% and motorcyclefatalitiesby42%.9It is estimated thatwhilemostmotorcycle riderswear a helmet only about 9percent of passengers wear a helmet. The government has delivered a publiceducationprogramtoencouragemotorcyclepassengerstowearhelmetsbutthisappearstohavefailedtoreducethehighlevelofnon-compliance.The extent of the problem in regards to all motorcycle fatalities and seriousinjuries could be substantially reduced if Police enforcement together withsubstantialfinesfornotwearingahelmetwereimplemented.Forthistoleadtobroadchangedcommunitybehaviour,thePolicewouldneedtoallocateappropriateresourcesforstoppingmotorcyclistsandtoissuefines.Inparticular,mostpeoplemustfeelthatthereisareasonableprobabilityofbeingapprehendedby aPoliceofficer and issueda finewhile riding amotorcycle. Ifthisisnotthecase,changebehaviouracrossthecommunityislesslikely.Inthisregard, it should be noted that about 80 percent of the Thai population ridemotorcyclesandthathismaycreateasignificantresourcechallengetodealwithsomanyridersandpassengers

9AaronPervin,JonathonPassmore,MirjamSidik,TylerMcKinley,NguyenThiHongTuc&NguyenPhuongNam,“VietNam’smandatorymotorcyclehelmetlawanditsimpactonchildren”,BulletinoftheWorldHealthOrganization2009;87:369-373.

50

AssesstheconsequencesofnoactionWhataretheconsequencesofnottakinganyaction?Couldrelyingonthemarketinconjunctionwiththegeneralapplicationofexistinglawsandregulationssolvetheproblem?Ifnot,whynot?ConsequencesofnogovernmentactionIt is useful to compare the experience of other countries that mandate thewearing of motorcycle helmets where the motorcycle is the main mode oftransport.Thailandhas33.5millionregisteredmotorvehiclesand20million(2013)or60percentare registeredmotorcycles.Vietnamwouldappear tobea comparablecountrygiventhatithas26millionregisteredmotorvehiclesand95percentaremotorized twowheelers.Similarly,Vietnamhasahighroad toll; in2007 therewere12,800 fatalitiesor15 fatalitiesper100,000population.Anestimated60percentofallroadfatalitiesoccuramongmotorcycledriversandpassengers.A study was conducted for all road traffic injury patients with head injuriesadmitted to 20 provincial and central hospitals 3months before and after thenew lawcame intoeffecton15December2007.Thestudy founda16percentreductionintheriskofroadtrafficheadinjuriesandan18percentreductionintheriskofroadtrafficdeath.10Itwouldappearthepublicperceptionthatmotorcyclehelmetswornbychildren,particularly young children, may cause neck injuries has underminedcompliance. Conflicting views expressed by themedical profession in Vietnamhas divided the Vietnamese community and they have erred on the side ofcautionandmostlydecidedtonotlettheirchildrenwearamotorcyclehelmet.Apubliceducationcampaigntocountertheperceptionthatmotorcyclehelmetsdonotcauseneckinjurieswouldmorethanlikelyneedtobelengthycampaignand costly to gain the confidence of the community and to persuade mostparentstoensurethattheirchildrenwearmotorcyclehelmets.Theeffectivenessof such a public education campaign would be dependent on the degree ofcontinued divisive views publicly expressed by some within the medicalprofession. Hence, there is a risk that such a campaignmay fail to deliver anadequate increase intheproportionofchildrenwearingmotorcyclehelmetstojustify such an investment by government where these funds may be more

10 Passmore J,TuNT,LuongMA,ChinhND,NamNP, “ImpactofmandatorymotorcyclehelmetwearinglegislationonheadinjuriesinVietNam:resultsofapreliminaryanalysis”,TrafficInjuryPrevention,2010Apr;11(2):202-6.

51

effectivelyused forothercountermeasures thataremore likely todeliverroadsafetybenefits.Willtheproblemself-correctwithinareasonabletimeframe?Therearemanyfactorsthatcontributetoroadfatalitiesandinjuries.Similarly,governmentsuseawiderangeofcountermeasurestoaddressthesefactorswiththeaimofimprovingroadsafety.In this respect, it isuseful tounderstand the impactof these countermeasuresand whether these have been, or are likely to be adopted by Thailand in thecomingyears.Thailandhasexperiencedhighpopulationgrowthsince1950sandahighgrowthof vehicle ownership (particularly motorcycles) since the 1970s. Anygovernmentstrugglestoexpandroadcapacitytoaccommodaterapidpopulationandmotorvehiclegrowth.Newroadinfrastructuretakesmanyyearstobuild.California,TexasandFloridaexperiencedsimilarpopulationandmotorvehiclegrowth from1950 to the early2000s. While the road fatality toll in theU.S.Apeakedin1976,theroadfatalitytollpeakedinCaliforniain1984,Texasin1986andFloridain2003.Akeyfactorforthedelayinthereductionoftheroadtollinthese states was due to the higher population growth compared with otherstates.From1950to2003thepopulationinCalifornia’sdoubledfrom20to40million,Texas from20 to50millionandFlorida tripled from15 to45million.Oncethepopulationgrowthcurveflattened,thesestatesexperiencedabouta5percentageannualreductionintheirroadtoll.ThepopulationgrowthoverthepastdecadeinCalifornia,TexasandFloridahasdeclined and all three states have experienced significant reductions in theirroadtolls.Bycontrast,theUnitedKingdomhashadminimalpopulationgrowth(51to57millionfrom1950to2003)andhasinvestedheavilyinroadinfrastructureandothersafetycountermeasures.Thishas resulted inadramatic reduction in theroadtollbutwouldhavebeenunlikelyintheeventthatithadpopulationgrowthlikeCalifornia,TexasandFlorida.TheWorldBank(UnitedNations)hasforecastthatThailand’spopulationgrowthwillbegintodeclinefrom2015.BasedontheexperiencesofCaliforniaetal,itislikely that as Thailand expands its road infrastructure, its road toll will alsodeclineoverthenextdecade.ItisdifficulttodetermineThailand’sannualrateofreductiononceitspopulationgrowth curve flattens. It is problematic that Thailandwould achieve a similarannual rate of reduction given that California et al had well established roadnetworks in the1950sandmostofThailand’s roadsarenotdivided toensuremotorvehicles,motorcyclesandpedestriansareseparatedfromeachother.

52

Another factor that needs tobe taken into account is the growth in registeredpassengervehicles.In2004,therewereabout6.5millionpassengervehiclesor33 percent of the total number of registered motor vehicles. By 2013, thenumber of registered passenger vehicles had increased to 13 million or 39percentofthetotalnumberofregisteredmotorvehicles.It is noteworthy that in 2003, Thailand introduced a requirement thatexpressways in Bangkok must exclude motorcycles. This was instigatedprimarily to reduce traffic congestion but also would provide road safetybenefitstomotorcyclists.JustificationforGovernmentInterventionTheanalysisofthenatureandsizeoftheproblemhasrevealedthefollowing:Ataminimum, children less thansixyearsofagecomprised204or2%of the10,187motorcyclefatalities.BasedonWHOestimations,childrenlessthansixyearsofagecomprised389or2%ofthe19,187motorcyclefatalities.The cost to the community from child fatalities, serious and slight injuries isestimatedfrom1.750billionbahtto3.338billionbahtperannum.Therewere5.1or9.7childfatalitiesper100,000populationofchildrenlessthansix years of age. This is lower than the 22.2 to 42.5 fatalities per 100,000populationforallotherroadusers.Theriskofafatalityforachildlessthansixyearsofagebeingtransportedonamotorcycle is 1.8 to 3.5 fatalities per 100 million vehicle-kilometers-travelled(VKT). This is lower than the 26.7 to 51 fatalities per 100,000 population formotorcyclists.Mostmotorcycle accidents occur amongstmale riders, 15 to 29 years,mostlyintoxicatedandfrom9.00pmto6.00am.Thelawrequiresmotorcycleridersandpassengerstowearmotorcyclehelmets.Thelackofcomplianceandenforcementofthecurrentlawwouldappeartonotaddresstheproblemofchildfatalitiesandinjuries.Publiceducationtosupportthecurrentlawregardingthemandatorywearingofmotorcycle helmets would appear to be problematic given the Vietnamexperience where many communities refused to make their children wear amotorcyclehelmetinfearthattheymayincurneckinjuries.Thailand, likeCalifornia,TexasandFlorida,haveexperienced rapidpopulationandmotor vehicle growth.With an expected decline in population growth, anincrease in the proportion of passenger vehicles relative to motorcycles, and

53

improved road networks (road engineering strategies), it is likely that thenumberofoverallfatalitiesandinjurieswilldecline(includingchildren)evenifthegovernmentdoesnotinterveneandintroduceanynewmeasures.TheimprovementsarelikelytobevariedwithgreaterdeclinesinBangkokandothermunicipal areas due to the higher ownership rate of passenger vehiclesrelative to motorcycles. Accordingly, there will be a lag in road safetyimprovement in rural areas, particularly low socio-economic areas with highmotorcycle dependency and low levels of road infrastructure investment(includingroadsafetyengineeringstrategies).The problem analysis has revealed a significant cost associated with childfatalities,seriousandslightinjuriesthatareincurredtravelingonamotorcycle.However,theanalysisalsorevealedthatchildrenlessthansixyearsofageareatlessriskonamotorcyclecomparedtootherroadusers.ThisisnotdissimilartootherOECDcountries.Thesignificanceofthecosttothecommunityisworthyoffurtherconsiderationintermsofwhethera countermeasurecanbedevelopedto specifically address child fatalities and injuries incurred traveling on amotorcycle.

54

OptionsThe following options listed below are examined to assess whether they arelikelytoaddresstheproblem:

• Option1TotalBan• Option2SelectiveBantargetedathighriskareas• Option3WarningSignsforhighriskareas

• Option4PubliceducationAdescriptionofeachoptionisprovidedbelow.Option1TotalBanThisoptionwouldimposeatotalbanonchildrenlessthansixyearsofagefrombeing transported on amotorcycle. A total banwould apply at all times in allareasthroughoutthecountry.Thetotalbanwouldaffectfamiliesthatuseamotorcycleastheirprimarymeansoftransportandwhohaveanestimated970,941childrenlessthansixyearsofage.11Atotalbanwouldrestrictcompetitionanddirectlyaffectmotorcycletaxisfromproviding transport services to children less than six years of age. A total banwouldprovideadvantagestobusandothertaxitransportproviders.However,itisnotclearwhethertheseotherformsoftransporthavethecapacitytomeetthedemandifatotalbanwasintroduced.Compliance and enforcement is problematic given the experience withcomplianceandenforcementofmandatorywearingofamotorcyclehelmet.Option2SelectiveBantargetedathighriskareasThisoptionwouldimposeabanonchildrenlessthansixyearsofagefrombeingtransportedonamotorcycleinselectedareasthatareconsideredhighriskareas(blackspots)andhaveahistoryofmultipleaccidents,fatalitiesandinjuries..High risk areas have not been identified and research would need to beundertaken to identify appropriate areas. The Thailand Accident ResearchCenter(TARC)iscurrentlyundertakingastudytoidentifyblackspotsinseveral

11The970,941childrenisbasedon61%(percentageofmotorcycleuse)of1,591,706childrenenrolled in kindergartens (2007). Source of kindergarten enrolments: Australian EducationInternational “Thailand Regulatory Factsheet 2013” cites Basic Statistics of the Ministry ofEducation2007.

55

provinces. The results of this study could help to inform the likely number ofhighrisksareasinThailand.Thegoalofthisprojectistoimprovetheroadsafetybyimplementingengineeringmeasures,toevaluateperformanceofengineeringmeasures by conducting before-after analysis, and to present the benefits ofengineering measures to policy makers and provide data for other similarprojects. A total of 10 black spot locations will be selected from differentprovinces in Thailand. Then, the process of studying sites, data collection,conceptual and detailed design, and implementation of appropriateimprovementwillbeconducted.Similartooption1,aselectivebanwouldrestrictcompetitionanddirectlyaffectmotorcycletaxisfromprovidingtransportservicestochildrenlessthansixyearsofage inhighriskareas.Aselectivebanwouldprovideadvantages tobusandother taxi transport providers. However, it is not clear whether these otherformsoftransporthavethecapacitytomeetthedemandifaselectivebanwasintroduced.Similar to option 1, compliance and enforcementwould be an issue.However,enforcement would be more manageable for the Royal Thai police to enforcegiventhesmallerareasinvolvedcomparedtooption1.Option3WarningSignsathighriskareasSimilartooption2,highriskareashavenotbeenidentifiedandresearchwouldneedtobeundertakentoidentifyappropriateareas.Appropriate design of warning signs would need to be installed at high riskareas.Thedesignofthewarningsignwouldneedtoclearlycommunicatetothemotorcycle rider that they were entering an area that has a high number ofaccidentsandfatalities.Thiswouldbesimilartothe‘blackspot’signsinstalledathighfatalityintersectionsinVictoria,Australia.Warning signs rely onmotorcyclists andother roadusers to take greater caredriving through these high risk areas. This option is effectively a form of self-regulationand requiresvoluntary compliancebyall roadusers to takeamoreriskaverseapproachwhendrivingthroughhighriskareas.Option4PubliceducationA public education campaign could be undertaken targeted at families withyoung children highlighting the number of child fatalities and injuries and theappropriatemeasures thatcanbe taken to reduce theriskof fatalityor injury.Thiscouldincluderevisitingthemandatorywearingofmotorcyclehelmetsandprovidingmedicallysoundadviceaboutthemeritsofyoungchildrenwearingamotorcyclehelmet.Apubliceducationcampaigntocountertheperceptionthatmotorcyclehelmetsdonotcauseneckinjurieswouldmorethanlikelyneedtobelengthycampaignand costly to gain the confidence of the community and to persuade most

56

parentstoensurethattheirchildrenwearmotorcyclehelmets.Theeffectivenessof such a public education campaign would be dependent on the degree ofcontinued divisive views publicly expressed by some within the medicalprofession. Hence, there is a risk that such a campaignmay fail to deliver anadequate increase intheproportionofchildrenwearingmotorcyclehelmetstojustify such an investment by government where these funds may be moreeffectivelyused forothercountermeasures thataremore likely todeliverroadsafetybenefits.Thereisnoemphaticempiricalevidencethatpubliccampaignsusedsolelyastheprimarycountermeasuredeliverroadsafetybenefits intermsofareductioninfatalitiesandinjuries.Publiccampaignstendtobecomplementarytoinformthepublic about the introduction or changes to countermeasures to deal with aspecificroadsafetyproblem.Theeffectivenessofpubliceducationisdependentonthepublicperceivedriskofchildfatalitiesandinjuriesaswellastheenforcementofregulationsuchasthemandatorywearingofmotorcyclehelmetsforridersandpassengers.GiventhatthelevelofcomplianceiscurrentlylowthroughoutmostpartsofThailand,itisunlikely that current enforcement practices are likely to improve compliancelevels.For these reasons, option 4 is not considered a feasible alternative to solelyaddress child fatalities and injuries, and will not be considered further orassessedasaviablealternative.

57

AssessmentofOptionsOption1:TotalBanCostsDirectCostsAtotalbanwouldimposedirectcostsonfamiliesdependentonamotorcycleastheirprimarymodeoftransport.Thiswouldaffectfamiliesof4millionchildrenlessthansixyearsofage.Thedirectcosttofamiliesinvolvestwotypesofcosts;the opportunity cost to parents or other family members to accompany theirchildrenonalternativemodesof transportandthe incrementalcostassociatedwithalternativemodesoftransport.Theopportunitycostentailsthetimedifferencebetweenafamilymembertakingachildonamotorcycleandalternativemodesoftransport.Thetimedifferenceis the opportunity cost of potential foregone income that could have beenderivedhadthefamilymembernotbeenrequiredtospendadditionaltimeusinganalternativemodeoftransport.Theincrementalcostofalternativemodesoftransportisthecostdifferenceonaperkmbasisforusingamotorcycleandalternativemodesoftransport.The ban would also impose direct costs on motorcycle taxis and restrictcompetitioninthepublictransportsector.Thesedirectcostsarediscussedandquantifiedbelow.OpportunitycosttofamiliesIt is common for families dependent on amotorcycle, for either the father ormothertotransporttheiryoungchildrentopre-schoolinstitutions,shopsaswellastomakesocial,culturalandreligiousvisits.Themotorcycleisalsocriticalfortransporting a sick child to the local doctor, health centre or the hospital inemergencysituations.Under thisoption,parentswouldneed tomakearrangements for leaving theirchildrenthatarelessthansixyearsofageathomewithanothercare-giverwhilethey went to work, shopping, transported older children to school, or madesocial visits to friends or family relatives, or to attend cultural and religiousceremonies.However, not all families would need to take their young children on amotorcyclewhentransportingolderchildrentoschool.MostThaifamilieshavelargeextendedfamilieswheretheycanrelyongrandparentsoroldersiblingstocareforyoungerchildrenwhileaparentistransportingolderchildrentoschool.

58

ItisestimatedthataboutXpercentoffamilieshavelargeextendedfamiliesthatcanassistwithcaringforyoungerchildrenwhentheparentneedstogotowork,shoportransportoldersiblingstoschool.OpportunitycostsforfamilymembersusingalternativemodesoftransportItisdifficultduetotheabsenceofdatatoestimatethenumberoftripsperdayachildistransportedonamotorcycle.However,youngchildrenfromthreetofiveyearsof age attendpre-school classes, kindergartens and/or childcare centres.Invariably, parents or older siblings that are dependent on a motorcycle,transporttheiryoungchildrenonamotorcyclefromhometotheseinstitutions.Atotalbanwouldhaveitsgreatestimpactinrespecttosendingyoungchildrento these institutions and would require parents to consider other modes oftransport.Dataonpre-schoolattendanceisavailableandcostsestimatescanbeundertaken. Up to 970,941 children are transported to and from pre-schoolwhose families are dependent on a motorcycle as their primary mode oftransport. It has been assumed for the purposes of calculating the costs thatthese families would transport their children by motorcycle notwithstandingthatsomeofthesefamiliesmaycurrentlyusealternativemodesoftransport.An alternative mode of transport requires six person trips per day. A familymember escorting a young child to pre-school (2 person trips), the familymember returning home (1 person trip), the family member leaving home topick–upchild(1persontrip)andthefamilymemberescortingtheyoungchildhome (2 person trips). Using the 970,941 affected children andmultiplying 6person trips equates to5,825,646person tripsperdayof pre-school. This hasbeenroundedto6milliontripspersontrips.Familieslivingwithinwalkingdistanceofapre-schoolcentrecouldopttowalk.Given the nature of narrow lanes,manywithout sidewalks, this could actuallyposeagreaterrisktoyoungchildrenthanridingonamotorcycle,particularlyifaccompaniedbyanelderlygrandparent.Bothyoungchildrenandtheelderlyareconsideredvulnerablepedestrians.Alternatively, some parents could decide to use buses to undertake some oftheseactivities.However,eveninthesecircumstances,thefamilywouldneedtobewithineasywalkingdistanceofabusstop.Inothercases, some familieswith the financialmeanscouldusemotorvehicletaxis.Parentswouldneedtomakeappropriatetransportationarrangementsbasedontheir proximity to a pre-school institution and the choice of transport modesavailableinthearea.Walkingdirectlytoapre-schoolinstitutionorwalkingtoabusstopinvolvesacosttofamilymembers;thatisthetimeinvolvedthatcouldhavebeenusedproductivelyforotherpursuitsincludinggeneratinganincome.

59

Parents, older siblings and grandparents already incur this cost when theytransport a young child on a motorcycle. Accordingly, the cost of a total banwould be the additional cost imposed from using an alternative mode oftransport.Theadditionalcostisknownastheincrementalcost.Walkingfromhometoapre-schoolinstitutionandwalkingtoabusstopwouldbe a less efficient mode of transport compared with a motorcycle given thelongertimeinvolvedinthispedestrianactivity. Itisestimated(basedonxxxx)thata round trip for thismodeof transportwould involveaboutonehourperday,Xhoursperweek(basedonthenumberofdaysperweekachildattendsapre-schoolinstitution)andXhoursperannum.AnestimatedX%offamiliescouldwalkdirectlytoapre-schoolinstitutionandanestimatedX%offamiliescouldwalktoabusstop(basedon………).Averagemonthlyearningsareusedtodeterminethehourlyrateforpersonsinvolvedinwalkingwithayoungchild.Accordingly,theestimatedincrementalcostofXbahtper annum is based on the number of families that nominatewalking as theirpreferredmodeoftransportmultipliedbythedailyincrementalcost50baht.A pre-booked passenger vehicle taxi that picked up the child and family carerfromthehomewouldhavethesamelevelofefficiencyasamotorcycle,andmayhave superior efficiency, as vehicles tend to travel at higher speeds thanmotorcyclescarryingyoungchildren.IncrementalcostofusingalternativemodesoftransportIn addition to the incremental cost incurred by family members to usealternative modes of transport to attend a pre-school institution, the familywould incur the incremental cost for paying to use these other modes oftransport.Thatisthedifferencebetweentheoperatingcostsofamotorcycleandtheothermodesoftransport.In regards towalkingdirectly from thehome to apre-school institution, therewouldbean incrementalbenefitas thereareno faresassociatedwithwalking.TheoperatingcostofanaveragemotorcycleisXbaht(basedontheaveragetimeforaroundtriptoapre-schoolinstitution).Theestimatedpercentageoffamiliesthat could walk directly to a pre-school institution is X % (based on ………).Accordingly,theestimatedincrementalbenefitisXbahtperdayperfamilyandXbaht per annum (number of families that nominatewalking as their preferredmodeoftransportmultipliedbythedailyincrementalbenefitXbaht).TheestimatedincrementalcostwithusingabusasthemainmeansoftransportisXbahtperday.Thecostofusingabus involves the fare forthe familycarer(fourtrips@Xbaht)andthechild(twotrips@xbaht.=Xbaht).TheoperatingcostofanaveragemotorcycleisXbahtasdiscussedabove.Thedifferenceinthecost between a bus and motorcycle is X baht. The estimated percentage offamilies that would need to use a bus to transport their child to a pre-schoolinstitutionisX%(basedon………).Accordingly,theestimatedincrementalcostis X baht per day per family and X baht per annum (number of families that

60

nominate a bus as their preferred mode of transport multiplied by the dailyincrementalcostXbaht).Theestimatedincrementalcostwithusingapassengervehicletaxiasthemainmeansoftransport isXbahtperday.Thiscost isbasedonfourfares@Xbahtperday(fourtripsforthefamilycarerandtwotripstrips@xbaht.=Xbaht).Theoperating cost of an average motorcycle is X baht as discussed above. Thedifference in the cost between a taxi andmotorcycle is X baht. The estimatedpercentage of families that would use a taxi to transport their child to a pre-school institution is X % (based on ………). Accordingly, the estimatedincrementalcostisXbahtperdayperfamilyandXbahtperannum(numberoffamiliesthatnominateabusastheirpreferredmodeoftransportmultipliedbythedailyincrementalcostXbaht).The average incremental cost for the various alternative modes of transport(otherthanpedestrian)isestimatedat50bahtandtheaverageincrementalcostto family members accompanying their children on alternative modes oftransport isestimatedat50baht. This imposes total incremental costsof100baht per day per family or 97,094,100 baht per day for the families of theaffected970,941pre-schoolchildren.On an annual basis, this translates into a total incremental cost of almost 20billionbaht (on theassumption thatpre-schooloperates5 timesperweek,40weeksperannum).Asummaryoftheannualcostsisprovidedbelow.Incrementalcosttocarerwalkingtopre-schoolinstitution XmillionbahtIncrementalcosttocarerwalkingtobusstop XmillionbahtIncrementalcostwithwalking (Xmillionbaht)Incrementalcostwithusingabus XmillionbahtIncrementalcostwithusingataxi XmillionbahtTotalincrementalcost: XbillionbahtInaddition,therewouldbeincrementalcostsassociatedwithfindingalternativetransport for taking trips for health, cultural, religious, recreation and othersocialactivities.Asdiscussedbefore,nodataexiststoquantifythesecosts.RestrictiononCompetitionThetotalbanwouldalsorestrictcompetitioninthepassengertransportsector.Motorcycletaxisplayalargeroleinprovidingtransportservicesandwouldnotbepermittedtocarrychildrenlessthansixyearsofage.Thereareanestimated80,000motorcycle taxis thatwouldbeaffectedbythetotalban. It isestimated(based on survey from major motorcycle taxi firms) that young children lessthansixyearsof age compriseX%of taxi tripsand the lossof revenue to themotorcycletaxisectorisestimatedataboutXmillionbahtperannum.

61

Further consultation is requiredwith themotorcycle taxi industry toascertainwhether the loss of pre-school children and their family member would besignificantenoughtoimpactonemploymentopportunitiesintheindustry.IndirectCostsImpactsonearlychilddevelopmentThebanmayhaveunintendedconsequencesforearlychilddevelopmentwheresome familieshave limitedpublic transportchoicesorareunable toafford theadditionalcostsofpublictransport.Theimpactonremoteruralareasmayactuallydetersomefamiliesfromsendingtheirchildrentopre-schoolinstitutionswheretheyhavelimitedornoaccesstopublictransport.Similarly, low socio-economic groups across the country may also withdrawtheirchildrenfrompre-schoolinstitutionsincaseswheretheyaredependentonpublic transport but due to financial hardship cannot afford the incrementalcostsassociatedwithpublictransport.ImpactsonthecapacityofalternativemodesoftransportThebanwouldrequireashiftfrommotorcycletransporttoalternativemodesoftransport.Other than for those families that canwalk fromhome to theirpre-schoolcentre,itisnotclearwhetherthecurrentpublic/privatebusnetworkandmotorvehicletaxishavethecapacitytoabsorbupto6millionpersontripsperday.Impactsonthecapacityofpedestrianinfrastructure,congestionandroadsafetyPedestrianfacilities(sidewalksandbridgesacrossbusyroads)arefairlylimitedthroughout Thailand. It is not clear whether the current road infrastructurecouldcopewithincreasedpedestrians.Ifasignificantproportionofthe6millionpersontripsundertakenonadailybasistopre-schoolsinvolvedwalkingasthealternative mode of transport, the lack of appropriate and safe pedestrianinfrastructure could lead to unintended consequences of increased congestion(pedestriansspillingontotheroadandstoppingtraffic)andtheassociatedsafetyriskstoyoungchildrenandfamilymembers.

62

BenefitsDirectBenefitsTheaffectedfamilieswoulddirectlybenefit fromthebanintermsofno lossoflifeorinjurytotheirchildren.Familiesandthewidercommunitywoulddirectlybenefitfromtheavoidedcostsassociatedwithfatalitiesandinjuries.Atotalbanwouldpreventfrom204to389childrenbeingkilledonamotorcycle.In addition, a total ban would prevent a considerable number of serious andslightinjuresasshowninTable16.Table16:EstimatedNumberofChildFatalities,SeriousandSlightInjuries Officialdata WHOestimatesFatalities 204 389SeriousInjuries 2,652 5,057SlightInjuries 7,956 15,171Thebanonchildrenlessthansixyearsofagebeingtransportedonmotorcycleswouldavoidthecostsassociatedwithchildfatalities,seriousandslightinjuriesand generate annual savings from 1.750 billion baht to 3.338 billion baht asshowninTable17below.Table17:AvoidedCostsofChildFatalities,SeriousandSlightInjuriesCrashSeverity Cost(baht)basedonofficial

dataCost (baht) based onWHOestimations

Fatalities 1,084,373,424 2,067,751,284SeriousInjury 389,904,996 743,495,311SlightInjury 276,558,516 527,359,131Total 1,750,836,936 3,338,605,726However,asdiscussedinthenatureandextentoftheproblemsection,thelevelofcomplianceisanissuewithroadtrafficlaws.Inparticularwithmotorcycles,itisarequirementfortheriderandpassengerstowearamotorcyclehelmet.Yetthe level of compliance varies across Thailand and it is common for themotorcyclerider towearahelmetwhileotherchildren(including thosebelowsixyearsofage)tonotwearahelmet.Intermsofenforcement,themodusoperandiofpoliceistoestablishdesignatedpolice checkspointsonmajor roads to stopandcheck the licence, registrationand third party insurance papers of the driver and to also conduct vehicleroadworthinesswhere appropriate. Given thatmost familieswould undertakesmall trips fromtheirhometoapre-school institutionmostlyalongresidentialstreetsand lanes, it isunlikelythatthecurrent locationofpolicecheckswoulddetectnon-compliancewithatotalbanunlessthepolicespreaditsresourcestoestablishpolicecheckpointsatpre-schoolinstitutions.In view of a possible low compliance with a total ban and the resourceconstraints of police providing police check points at X number of pre-school

63

institutionsacrossthecountry, it isunlikelythat the fullbenefitsofa totalbanwouldeventuateinthefirstfewyearsofitsoperationandmayalwaysstruggletoachievethedesiredbenefitsunlesssupportedbyapubliceducationcampaignandthewillingnessofpeopletocomplywiththelaw.IndirectbenefitsImpactontrafficcongestionandenvironmentTheabsenceofabout970,941motorcyclesduringtheperiodwhenchildrenaredroppedoffandpickedupfrompre-schoolislikelytoreducetrafficcongestionandtheassociatedtraveldelaycosts toothercommuters.Parents transportingtheirchildrenonamotorcyclerepresentabout5percentofthetotalnumberofregisteredmotorcycles.Thereisinsufficientdataontheproportionofregisteredmotorcycles that are likely to be on the road at the same time asmotorcycleswith children traveling to and from pre-school to determine the currentcongestion levels and associated travel costs to predict possible cost savingsunderthisoption.Whileitisdifficulttoestimatethereductioninmotorcycleusagebyfamilieswithpre-schoolchildren,thebanmayalsoprovidesomeenvironmentalbenefitswithreductionsinmotorcycleemissions.ImpactonpatronagelevelsforalternativemodesoftransportAlternativemodes of transport such as public andprivate bus companies, andtaxis are likely to experience increased patronage and revenue from theproposedban.Itisdifficulttoestimatetheexpectedindirectbenefitstoeachofthedifferentalternativemodesoftransport.SummaryofCostsandBenefitsThe incrementalcosts to families tousealternativemodesof transport to taketheir children topre-schoolwasestimatedat about20billionbahtper annumandthebenefitsoftheavoidedcostsassociatedwithfatalities,seriousandslightinjurieswasestimatedtorangefrom1.750billionbahtto3.338billionbahtperannum; leaving a net cost of 18.250 billion baht to 16.662 billion baht perannum.Thebanwouldalsohave indirectcostsandunintendedconsequences forearlychildhood development for families with limited access to public transport orlow-income families that could not afford the additional costs associatedwithpublictransport.Thebanimposesarestrictiononcompetitionthatremovestheoptionofamainpublic transport provider (motorcycle taxis) providing transport options tochildren and families. This calls into question whether the capacity of otherpublictransportoperatorsandthepedestrianinfrastructurecancopewithuptoan additional 6million person trips on a daily basis. In the case of pedestrian

64

infrastructure, additional pedestrian traffic may lead to increased trafficcongestion and road safety for children and family members. The increasedpatronage of public transport may in the medium term result in increasedinvestmenttoimprovecapacityandthereducedmotorcycleusagemayoffsettosomeextenttrafficcongestion.Option2SelectiveBantargetedathighriskareasCostsThe direct costs would be the same as in option 1. However, they would belimitedtospecificareasthatweredeemedhighriskandhavemultiplenumbersof fatalities, serious and slight injuries. These high risk areas are commonlyreferredtoas‘blackspots’wheretheroaddesignandtopographyareinherentlydangerous.ItisdifficulttoquantifythedirectcostsasthenumberandlocationofhighriskareashavenotbeenidentifiedinThailand.Insomecases,ahighriskareamaybespecificroadsthathaveahistoryofmultipleaccidents.The government would incur administrative costs undertaking research intoidentifying high risk areas and installing appropriate road signage advisingfamilies tonot transportyoungchildrenonamotorcycle.ThesecostshavenotbeenabletobequantifiedandconsultationisrequiredwiththeDepartmentofHighways and TARC to ascertain the cost per high risk area and the likelynumberofhighriskareasacrossthecountry.Enforcement would bemoremanageable for the Royal Thai police to enforcegiventhesmallerareasinvolvedcomparedtooption1.BenefitsThere isno guarantee thatmost child fatalities and injuries are locatedwithinhighriskareas.Itisassumedthataselectivebanwouldreduceanundeterminednumberoffatalitiesandinjurieswithoutsignificantlyimpactingonthefreedomof families dependent on a motorcycle as their primary mode of transport totransporttheiryoungchildren.Inthesecircumstances,affectedfamiliesmaybeabletostillusetheirmotorcycletotransporttheirchildtopre-schoolbynavigatingtheirwayaroundthesehighriskroads.SummaryofCostsandBenefitsThekeydirectcostsofaselectivebanonhighriskareaswouldbe incurredbygovernmentinidentifyinghighriskareas,installingappropriatesignageandthecostofenforcement.However,dependingonthenumberandthesizeofthehigh

65

risk areas, the Royal Thai Police may have the capacity and resources toeffectivelymonitorandenforceselectivebansofhighriskareas.However,untilfurtherresearchhasbeenundertakentoascertainwhetherthereis a positive correlation between high risk areas andmost child fatalities andinjuries, it is difficult to assume that a selective ban would be effective inachievingasignificantreductioninchildfatalitiesandinjuriestowarrantpublicsector investment in identifying high risk areas and committing the requiredresourcestoestablishhighriskareasandtoenforcetheselectiveban.Option3WarningSignathighriskareasCostsSimilartooption2,thegovernmentwouldneedtoinvestpublicsectorresourcesinidentifyinghighriskareasbyanalyzingaccidentdataandconsultingwiththelocalcommunity.Ontheassumptionthatthistaskmaytake1,000workinghours@500baht labour cost per hour, the total cost to complete onehigh risk areawouldcostabout500,000baht.If500highriskareaswereundertaken,thetotalcostwouldbeabout250millionbaht.A further 50 million baht may be required for design of the warning sign,managementandadministrationoftheproject.The elements of sign cost include: materials, fabrication, inventory control,maintenanceandinstallationcosts(laborandtransport).Asignmaycost5,000bahteachandabout1,000bahtforlabourandequipmentfortheinstallationofeachsign(needstobeconfirmedfromtheDepartmentofHighway).Anaverageof10signsmaybeneededforeachhighriskareaatacostof$60,000baht.If500highriskareaswereidentifiedacrossthecountry,thetotalcostofsignagewouldbeabout30millionbaht.In summary, the cost to the government could be in the order of 330millionbaht.Thiswouldbeaone-off costwithminimalongoingmaintenance costs toreplacewornandbrokensigns.BenefitsTrafficcontrol(signals,signs,geometry,markings)werefoundinanAustralianstudy to be definitely relevant in about 20 percent of accidents and possiblyrelevant ina further17percentofaccidents. It isnotclearwhetherthiswouldtranslatetoThailand.Notwithstandingthis,ithasbeenassumedthatwarningssignscouldpotentiallyreduce20percentofaccidents.Similartooption2,thereisnoguaranteethatthereisapositivecorrelationbetweenhighriskareasandthelocationofchildfatalitiesandinjuries.

66

Onthebasisthata20percentreductioncouldbeachieved,thiswouldresultinthefollowingestimatednumberofavoidedfatalities,seriousandslightinjuriesasshowninTable18.Table18:EstimatedNumberofAvoidedChildFatalities,SeriousandSlightInjuries Officialdata WHOestimatesFatalities 41 78SeriousInjuries 525 1,011SlightInjuries 1,591 3,034This would potentially avoid the costs associated with child fatalities, seriousand slight injuries and generate annual savings from 350 billion baht to 669billionbahtasshowninTable19below.Table19:AvoidedCostsofChildFatalities,SeriousandSlightInjuriesCrashSeverity Cost(baht)basedonofficial

dataCost (baht) based onWHOestimations

Fatalities 217,937,796 414,613,368SeriousInjury 77,187,075 148,640,253SlightInjury 55,304,751 105,464,874Total 350,429,622 668,718,495SummaryofCostsandBenefitsThecostof identifyinghighriskareasandthedesign,manufacture, installationofwarningsignsisestimatedatabout330millionbaht.Thiswouldbeaone-offcostwithminimalongoingmaintenancecoststoreplacewornandbrokensigns.There is a degree of uncertainty as towhether thewarning signs in high riskareaswoulddelivera20percentreductioninchildfatalitiesandinjuriesgiventhattheactuallocationofchildfatalitiesandinjuriesisunknown.Given that thepotentialbenefitsare in theorder from350millionbaht to669million baht per annum, this option has a greater probability to deliver a netbenefitevenifa10percentreductionwasonlyachieved.ComparisonofOptionsOption1generatesanet cost to the communitybasedon thedata that canbequantified.Itislikelythatthecostswouldbeevengreaterifallofthecostscouldbequantified.Option1also identified severalunintended impacts,particularlyonearly childdevelopment for familieswith limitedaccess topublic transportandlowincomefamiliesthatmightnotbeabletoaffordpublictransport.Option2wasnotmeasuredduetotheinadequatedataonthelikelynumberofchildfatalitiesandinjuriesinhighriskareas.

67

Option 3 generates a net benefit but only achieves a potential 20 percentreductioninthenumberofchildfatalitiesandinjuries.None of the options solely address the problem completely and deliver a netbenefit.Furtherstudiesshouldbeundertakenparticularlyintohighriskareastoascertain whether there is a positive correlation between these areas with ahistoryofmultipleaccidentsandmostchildfatalitiesandinjuriesincurredonamotorcycle.ConsultationAsiaInjuryPreventionFoundationOfficeoftheConsumerProtectionBoardSavetheChildrenThailandRoyalAutomobileAssociationofThailandRoyalThaiPoliceThailandAccidentResearchCenter(TARC)UniversitiesSavetheChildrenThailandInresponsetotherecentproposedbanonyoungchildrenridingonmotorcycles,SavetheChildrencallsforthegovernment'sattentiontochildrenbelowtheageoftwotonotbeallowedonmotorcycles.Inaddition,SavetheChildrenurgesthegovernment,andthepolice,toenforcetheexistinghelmetlawforallpassengers,particularlyallchildren,2yearsoldandup.Withanestimated1.3millionchildreninThailandtravelingonmotorcycles,theThai government has legislated that all people – including children – arerequired towear a safety helmet at all times. Still,many child passengers areoften seenwithout helmets – only 7% of children in Thailand currently wearhelmets while riding motorcycles. This leads to devastating results -approximately2600childrenarekilled,andmorethan72,000areinjured,everyyearinroadcrashes.Save the Children recommends that children under two should not ridemotorcycles because they are at high risk of long-lasting injuries since theycannotwearhelmets safely. Save theChildrenalso recommends that childrenunderfiveonlyridemotorcyclesunderclosesupervisionofanadult.Currently,therearenoregulationsbytheThaigovernmentonthismatter.Thereisalaw,however,thatrequiresallpassengersanddriverstowearhelmetsatalltimes.AllisonZelkowitz,SavetheChildren inThailandCountryDirector,explains, "Atsuchayoungageoftwoandbelow,childrenhaveinsufficientmusclestrengthtosupporttheweightofahelmet–theyaremorepronetoinjuriesandshouldnotbeallowedonmotorcyclesatall.Childrenbetween2to5yearsshouldonlyrideinfrontofanadult,andmustalwayswearachild-sizedhelmetthatfitsproperly,

68

andisfastenedtightly."Save the Children seeks to minimize road crash risks to children throughawareness building, education, enforcement and preventative tools andequipmenttoensuresafetyandnegateunnecessaryexposuretoharm."InThailand,motorcyclesareimportantinprovidingchildrenaccesstoschoolsandhealthfacilities,butsafetymustalwaysbethefirstconcern.”saysAllison.SavetheChildrenalsorecommendstheuseofalternateformsoftransportationfor young children. These include public buses, subways, tricycles (tuk-tuks),taxisandpublicmini-vans.Save the Children in Thailand is currently partnering with the Asia InjuryPreventionFoundation in a collaborative effort called “The7%Project,”whichaimstodecreasemotorcycledeathandinjuryamongThaichildrenbyincreasinghelmetusefrom7%to60%by2017.

69

AppendicesAppendix1:StudyArea Fatalities Serious

InjuriesSlightInjuries

PropertyDamageOnly

Total

Bangkok 715 8,144 42,707 85,414 136,980AmnatCharoen 44 757 2,105 4,210 7,116AngThong 67 876 2,460 4,920 8,323BuriRam 222 3,021 8,466 16,932 28,641Chachoengsao 306 2,128 6,187 12,374 20,995ChaiNat 55 874 2,438 4,876 8,243Chaiyaphum 95 2,583 7,102 14,204 23,984Chanthaburi 136 1,723 4,841 9,682 16,382ChiangMai 396 5,014 14,094 28,188 47,692ChiangRai 287 3,376 9,519 19,038 32,220ChonBuri 527 8,172 22,797 45,594 77,090Chumphon 173 1,748 4,965 9,930 16,816Kalasin 67 1,983 5,441 10,882 18,373KamphaengPhet 117 1,717 4,800 9,600 16,234Kanchanaburi 192 2,301 6,482 12,964 21,939KhonKaen 221 3,699 10,293 20,586 34,799Krabi 140 1,577 4,458 8,916 15,091Lampang 155 2,016 5,660 11,320 19,151Lamphun 52 1,386 3,812 7,624 12,874Loei 115 1,750 4,884 9,768 16,517LopBuri 205 1,800 5,152 10,304 17,461MaeHongSon 20 627 1,720 3,440 5,807MahaSarakham 51 1,784 4,883 9,766 16,484Mukdahan 45 790 2,194 4,388 7,417NakhonNayok 50 937 2,600 5,200 8,787NakhonPathom 254 2,929 8,266 16,532 27,981NakhonPhanom 54 1,332 3,668 7,336 12,390NakhonRatchasima

516 5,544 15,698 31,396 53,154

NakhonSawan 322 2,930 8,369 16,738 28,359Nakhon SiThammarat

218 2,238 6,324 12,648 21,418

Nan 75 1,728 4,766 9,532 16,101Narathiwat 41 1,479 4,047 8,094 13,661Nong BuaLamphu

45 854 2,365 4,730 7,994

NongKhai 106 1,693 4,719 9,438 15,956Nonthaburi 21 1,892 5,129 10,258 17,300PathumThani 115 1,523 4,273 8,546 14,457Pattani 34 1,339 3,657 7,314 12,344PhangNga 46 1,082 2,985 5,970 10,083Phatthalung 92 1,426 3,979 7,958 13,455Phayao 29 828 2,275 4,550 7,682Phetchabun 198 2,208 6,240 12,480 21,126Phetchaburi 66 1,756 4,827 9,654 16,303Phichit 120 1,166 3,317 6,634 11,237Phitsanulok 227 1,567 4,558 9,116 15,468Phra Nakhon Si 220 2,699 7,598 15,196 25,713

70

AyutthayaPhrae 71 1,077 3,006 6,012 10,166Phuket 152 1,993 5,595 11,190 18,930PrachinBuri 263 1,687 4,934 9,868 16,752Prachuap KhiriKhan

205 1,768 5,067 10,134 17,174

Ranong 26 590 1,629 3,258 5,503Ratchaburi 143 2,979 8,241 16,482 27,845Rayong 122 2,847 7,851 15,702 26,522RoiEt 117 2,516 6,950 13,900 23,483SaKaeo 125 1,285 3,648 7,296 12,354SakonNakhon 145 2,115 5,914 11,828 20002SamutPrakan 113 1,210 3,428 6,856 11,607SamutSakhon 205 1,540 4,450 8,900 15,095SamutSongkhram

15 443 1,217 2,434 4,109

Saraburi 347 3,375 9,603 19,206 32,531Satun 25 509 1,409 2,818 4,761SiSaKet 69 1,964 5,394 10,788 18,215SingBuri 54 766 2,144 4,288 7,252Songkhla 271 3,677 10,307 20,614 34,869Sukhothai 100 1,410 3,948 7,896 13,354SuphanBuri 201 2,499 7,031 14,062 23,793SuratThani 285 3,383 9,538 19,076 32,282Surin 244 2,123 6,079 12,158 20,604Tak 67 1,372 3,797 7,594 12,830Trang 108 2,381 6,575 13,150 22,214Trat 44 845 2,342 4,684 7,915UbonRatchathani

384 4,408 12,444 24,888 42,124

UdonThani 304 3,127 8,872 17,744 30,047UthaiThani 60 955 2,664 5,328 9,007Uttaradit 65 977 2,731 5,462 9,235Yala 56 1,251 3,454 6,908 11,669Yasothon 83 999 2,815 5,630 9,527Total 11,721 157,057 460,197 920,394 1,549,369Appendix2

71

Appendix3StudyArea Population Fatalities Per

100,000Population

SeriousInjuries

Per100,000population

Bangkok 8,500,000 715 8.4 8,144 96AmnatCharoen 375,000 44 11.7 757 202AngThong 283,000 67 23.9 876 313BuriRam 1,580,000 222 14 3,021 191Chachoengsao 695,000 306 44 2,128 306ChaiNat 332,000 55 16.6 874 265Chaiyaphum 1,140,000 95 8.3 2,583 227Chanthaburi 527,000 136 25.6 1,723 325ChiangMai 1,700,000 396 23.3 5,014 295ChiangRai 1,200,000 287 23.9 3,376 281ChonBuri 1,400,000 527 37.6 8,172 584Chumphon 498,000 173 34.6 1,748 350Kalasin 985,000 67 6.7 1,983 198KamphaengPhet 729,000 117 16 1,717 235Kanchanaburi 848,000 192 22.6 2,301 271KhonKaen 1,800,000 221 12.3 3,699 205Krabi 457,000 140 30.4 1,577 343Lampang 753,000 155 20.6 2,016 269Lamphun 405,000 52 13 1,386 346Loei 634,000 115 18 1,750 278LopBuri 758,000 205 27 1,800 237MaeHongSon 248,000 20 8 627 251MahaSarakham 960,000 51 5.3 1,784 186Mukdahan 346,000 45 13 790 226NakhonNayok 257,000 50 19.2 937 360NakhonPathom 891,000 254 28.5 2,929 329NakhonPhanom 713,000 54 7.6 1,332 188NakhonRatchasima

2,620,000 516 19.7 5,544 212

NakhonSawan 1,073,000 322 30 2,930 274Nakhon SiThammarat

1,5000,000 218 14.5 2,238 149

Nan 478,000 75 15.6 1,728 360Narathiwat 775,000 41 5.3 1,479 192Nong BuaLamphu

509,000 45 9 854 171

NongKhai 517,000 106 20.4 1,693 326Nonthaburi 1,174,000 21 1.8* 1,892 162PathumThani 1,074,000 115 9.8 1,523 130Pattani 686,000 34 5 1,339 197PhangNga 261,000 46 17.7 1,082 416Phatthalung 520,000 92 17.7 1,426 274Phayao 484,000 29 6 828 172Phetchabun 995,000 198 19.8 2,208 221Phetchaburi 474,000 66 14 1,756 374Phichit 547,000 120 21.8 1,166 212Phitsanulok 851,000 227 26.7 1,567 184Phra Nakhon SiAyutthaya

803,000 220 27.5 2,699 337

Phrae 454,000 71 15.8 1,077 239Phuket 378,000 152 40 1,993 524

72

PrachinBuri 479,000 263 54.8 1,687 351Prachuap KhiriKhan

525,000 205 39.4 1,768 340

Ranong 177,000 26 14.4 590 328Ratchaburi 842,000 143 17 2,979 355Rayong 674,000 122 18.2 2,847 425RoiEt 1,3000,000 117 9 2,516 194SaKaeo 552,000 125 22.7 1,285 234SakonNakhon 1,140,000 145 12.7 2,115 185SamutPrakan 1,262,000 113 9 1,210 96SamutSakhon 532,000 205 38.7 1,540 291SamutSongkhram

194,000 15 7.9 443 233

Saraburi 633,000 347 55 3,375 536Satun 313,000 25 8 509 164SiSaKet 1,465,000 69 4.7 1,964 134SingBuri 212,000 54 25.7 766 365Songkhla 1,400,000 271 19.3 3,677 263Sukhothai 602,000 100 16.6 1,410 235SuphanBuri 849,000 201 23.6 2,499 294SuratThani 1,000,000 285 28.5 3,383 338Surin 1,400,000 244 17.4 2,123 152Tak 539,000 67 12.4 1,372 254Trang 638,000 108 16.9 2,381 372Trat 225,000 44 20 845 384UbonRatchathani

1,845,000 384 20.9 4,408 240

UdonThani 1,570,000 304 19.3 3,127 199UthaiThani 330,000 60 18.2 955 289Uttaradit 460,000 65 14 977 212Yala 512,000 56 11 1,251 245Yasothon 540,000 83 15.4 999 185Total 67,397,000 11,721 157,057