高速铁路外文资料37

www.elsevier.com/locate/tra

Evaluationofmanagementstrategiesfortheoperationof

high-speedrailwaysinChina

W.G.Wong

aa,*,B.M.Hana,L.Ferreirab,X.N.Zhuc,Q.X.Sun

ccDepartmentofCivil&StructuralEngineering,TheHongKongPolytechnicUniversity,HungHom,

Kowloon,HongKong

bSchoolofCivilEngineering,QueenslandUniversityofTechnology,Australia

SchoolofTra󰀢candTransportation,NorthernJiaotongUniversity,Beijing,People'sRepublicofChina

Received24May1999;receivedinrevisedform8July2000;accepted17July2000

Abstract

High-speedtrain(HST)operationshaverecentlybeenintroducedinrailpassengertransportationmarketsworldwide.Althoughthetechnologiesforsuchoperationshavelevelledatspeedsofaround300km/h,theoperatingparameterstobeadoptedineachapplicationwilldi󰀹erfromcountrytocountry.TheoperatingenvironmentwillbeoneofthecrucialsuccessfactorsfortheimplementationofHSToperationsinChina.Thispapercomparesthreedi󰀹erentmanagement/ownershipmodelswhichmightbeusedinChina.Thepaperanalyzesthecharacteristicsofeachmodelandproposesanoptimalplanofanopera-tionalsystemtodevelopHSToperationsinChinabyusingahierarchygoalachievementmatrixap-proach.Ó2001ElsevierScienceLtd.Allrightsreserved.

Keywords:High-speedrailway;Railwaymanagement;Hierarchymethod

1.Introduction

ThecurrenteconomicenvironmentinChinaisconducivetotheconstructionofhigh-speedrailways(HSRs).Since1993,theChinesegovernmenthasadoptedaseriesofe󰀹ectivemacro-economicmeasuresaimedatreducingthein¯ationrateandsmootheningthepaceofeconomicdevelopment.In1997,therateofincreaseofGDPremainedat8.8%,whilethein¯ationratewaslessthan3%.However,theincreaseintheGDPhassloweddownsincethebeginningof1998duetothein¯uenceoftheAsianeconomiccrisis.Thegovernmentexpectstostimulateconsumption

*Correspondingauthor.Tel.:+852-2766-6067;fax:+852-2334-6389.E-mailaddress:cewgwong@polyu.edu.hk(W.G.Wong).

0965-8564/01/$-seefrontmatterÓ2001ElsevierScienceLtd.Allrightsreserved.PII:S0965-8564(01)00003-9

278W.G.Wongetal./TransportationResearchPartA36(2002)277±289

byincreasinginvestmentespeciallyininfrastructure.Thismacro-economicsettingprovidesaverygoodopportunityforthedevelopmentofHSR.Accordingtothecentralgovernment,atotalinvestmentofnearly30billionUSDwillbespentonrailwaydevelopmentfrom1998to2002,anannualrateof6billionUSD(RailwayMinistryofChina,1998).Theinvestmentinroadprojectsisgreaterthan20billionUSDannuallyfrom1998to2000,mainlyduetodi󰀹erencesininvestmentprioritiesatthedi󰀹erentlevelsofgovernment.Localgovernmentshavemoreenthusiasmforroaddevelopment,whilstrailinfrastructureprojectsaremainlysupportedbycentralgovernment.Investmentinairtransportwas1.15billionUSDin1998(Wang,1999).

Fromthepointviewofenvironmentalprotectionandimprovementintheservicelevelofpassengertransport,high-speedtrain(HST)operationsarethoughttobethemostappropriateoption(Han,1997;Qian,1995).Especiallyformiddledistancejourneys(200±700km),HSToperationscomparewellintermsoftraveltimesavings.Chinaisavastcountrywithapopulationof1.2billionandwithlimitedcultivableland.Thus,itisanessentialpolicyobjectivetodeveloppublictransportation,whichoccupiestheleastland,possesseslargecapacityandconsumeslessenergy,comparedtoprivatetransportation.TheHSRhasobviousadvantagescomparedwithhighwaysandcivilaviation.Railways,asanimportantnationalinfrastructure,arenotasad-vancedinChinacomparedwiththedevelopmentofthenationaleconomy.TheGDPhasin-creasedby4.45timesfrom1978to1996,whiletheoperatinglengthoftherailwayshasbeenextendedbyonly33%duringthesameperiod(NationalStatisticalYearbook,1997).Chineserailwaysranklowerthan100intheworld,intermsofthedensityoftherailwaynetwork(trackkilometerspermillionsquarekilometers).

Thedevelopmentofrailwaysonalarge-scalehasbeenincludedinthe``9thFive-YearPlan(1996±2000)''bytheChinesegovernment.TheconstructionofaHSRbetweenBeijingandShanghai(BSHSR)haspriority.ThisprojectisaimedatpromotingtheeconomicdevelopmentofacorridoralongtheBSHSR,themostdevelopedbeltintheMainland,includingthebiggestcitiessuchasBeijing,Tianjin,Jinan,Xuzhou,Nanjing,andShanghai.

HSRsisahightechnologydomain,whichincludesnewmaterials,newtechniquesandinfor-mationtechnology.TheconstructionofHSRswillnotonlyimprovethetechnologyofrailwaydevelopment,butwillalsostimulatethedevelopmentofhightechnologyindustries.Inaddition,itcanincreasethedemandformaterialproductionandprovideadditionalemploymentopportu-nities.

TheBSHSRwillbethelargestHSRprojectintheworldwithatotallengthofmorethan1300km.Thus,choosinganoptimalmodeofconstructionandoperationhasbecomeakeyfactorforthesuccessofthislargeinfrastructureproject.

Since1970s,railwayshaveenteredaperiodofdeclineinindustrialcountriescausedmainlybythechallengeofroadandair.Inordertoremedythissituation,newtechnologiessuchasHSTs,heavy-haultransportandinformationtechniqueshavebeendeveloped.Reformsofoperationalsystemshasbeenputforwardandarebeginningtoplayanimportantroleinrailtransportation(Ferreira,1997a).

InEngland,railwayreformbeganin1983.Japan,whichprivatisedthestaterailwayandes-tablishedsevenrailwaycorporations,hasachievedimportantgrowthbothinpassengerandfreight¯ows.AlthoughUSrailwaysareprivatelyownedandoperatedbymorethan500railwaycompanies,theirgrowthwasstimulatedbythelooseningofcontrolofrailwayfacilities,operationand®nancepoliciesin1976(Xiao,1994).

W.G.Wongetal./TransportationResearchPartA36(2002)277±289279

Thefocusofrailwayreformindi󰀹erentcountrieshasbeenputtingrailwaysintotransportmarketasindependententities.Therelationshipbetweenthegovernmentandthetransporten-terprisesisclearlyde®ned.Therearetwoprincipalmodesfortheseindustrialisedcountries,thesplitmodeandtheaggregativemode.

BasedontheexperienceofindustrialisedcountriesandtheconditionsofChineserail-ways,weproposedthreealternativesforBSHSR(twoprincipalmodesforrailwayopera-tionsandthetraditionalmode):(a)usingtheexistingthreerailwayadministrativebureaux(traditionalmode);(b)separatingoperationsfrominfrastructure(splitmode);(c)combiningrailoperationswithinfrastructure(aggregativemode)andcreatingasinglecorporation.

Itshouldbenotedthatthetraditionalmodeofconstructionandoperationhasmanydisad-vantagesinmeetingtheneedsofanewmarketeconomyenvironment.Thismode,whichisbasedontheplannedeconomysystem,isthemainreasonforreformsintherailwaypolicyarea.Thefocusofthisreformistodeterminehowtoestablishorreorganizetheincorporatedenterprisesinaccordancewithmoderneconomicregulations.

2.Alternatives

Twomainownershipmodelsareemerginginpractice,namely,theverticallyintegratedrailwaywithorwithoutseparateinternalbusinessunits,andtheverticallyseparatedrailwaywithtrackinfrastructuremanagedandownedseparatelybymultipleoperators.Theseparationmodelhasbeenadoptedorproposedinsomecountries,notablyinGreatBritain,Germany,theNetherlandsandSweden(NashandPreston,1994;JanssonandCardebring,1989).TheEuropeanUnionhasapolicyofmovingtowardstheseparationmodel(NashandPreston,1994).Asimilarapproachhasbeenfollowedforinter-statefreighttransportinAustralia,followingthecompetitionrelatedproposalsadoptedbyFederalandStategovernments(Hilmeretal.,1993).

2.1.Traditionalmode(executedbythreerailwayadministrativebureaux)

AccordingtothepresentrailwayadministrativesysteminChina,theBeijing±Shanghaiclassicrailwaysbelongstothreerailwayadministrativebureaux(RAB),RABofBeijing,RABofJinanandRABofShanghai,respectively(seeFig.1).TheBSHSRcouldbedividedintothreesectionswitheachbeingadministeredbyoneofthethreebureaux.

Advantages:

·Thiswouldbethepreferredoptionoftherailwayadministrativebureaux.Currently,themajorrevenuesoftheRABscomefrompassengertransportation.TheconventionalrailwayofBeij-ing±Shanghaiplaysanimportantroleinpassengertransportation(morethan200passengertrainsareoperatedperday).IftheBSHSRwereimplemented,thetra󰁏cvolumeforbothpas-sengersandfreightwouldincreasesigni®cantly.

·EaseofcoordinationoftheclassicrailwayandtheHSR.Becausebothrailwayswouldbeoperatedbythesameowner,itwouldbeeasiertocoordinatethefreightandpassengertrain

280W.G.Wongetal./TransportationResearchPartA36(2002)277±289

Fig.1.PositionofHSRbetweenBeijing±Shanghai.operations.TherewouldbetwotypesofpassengertrainsontheHSRbetweenBeijingandShanghai,namely:themediumspeedtrain(MST)andtheHST(HanandSun,1993;Hu,1998).TheMSTs(withmaximumspeedsof160km/h)wouldoperatepartiallyonconventionalrailwaysandpartiallyonHSRs.ThenecessityofMSTonBSHSRisdueto:(a)thefreighttransportvolumeonBeijing±Shanghaicorridorisquitehighandthedemandissteadilyin-creasing.InordertobalanceBSHSRandtheparallelconventionalrailway,mostpassengertrainsshouldbeoperatedontheformer;(b)thestructuralspeedofpassengertrainsonthecon-ventionalrailnetworkis160km/h;and(c)ifallpassengertrainsusedHSTsets,alargecapitalinvestmentwouldberequired.56

Disadvantages:

·SincetheRABisnotresponsiblefornewinfrastructurefundingaccordingtocurrentpolicy,fundingfortheprojectwouldhavetocomefromothersources.

·ItwouldnotfavourtheestablishmentofamodernizedenterprisesystemandmoveHSRsintoamarketeconomyenvironment.

W.G.Wongetal./TransportationResearchPartA36(2002)277±289281

2.2.Splitmode(separatingoperationsfrominfrastructure)

Undertheverticallyintegratedmodel,operatorsandtrackownerstendtohaveacustomer±serviceproviderrelationship.Theinfrastructureproviderexiststoservicetheneedsofitsclient(s).Thelattermayconsistofseveralbusinessunitssuchaspassengerservicesandvarioustypesoffreightservices.Insomecases,eachbusinessgroup`owns'itsowntracksegments,whicharedividedamongstoperatorsonthebasisofmajoruser.Userchargesmaybeleviedtonon-mainusersusinganinternalcosttransfersystemdesignedtoachieveaccountabilityand`valueformoney'outcomes.Itisarguedthatoneofthedrawbacksoftheverticallyintegratedmodel,isitsinabilitytoreadilyandfairlyaccommodatenewentrantsintheformofoperatingcompetitors,sharingacommontrackinfrastructure.Ifexistingrailwaysystemsarepubliclyowned,itispossibletoopenuptracktonewentrantsthroughdirectinterventionbygovernments.However,thequestionoffairnessindealingwithpotentialcompetitors,wouldrequirestrictcontractualarrangementsrelatedtocostsandservicequality.Thetermsandoperatingconditionsoftrackaccessneedtoextendtotraindispatchingrules.

Incontrast,theverticallyseparatedmodelhasbeenputforwardasawayofincreasingcom-petitionintherailsector,aswellasplacingrailandroadinfrastructureinvestmentandoperationsonanequalfooting.ThemainstatedaimoftheseparationoftrackfromoperationsinGreatBritainwastoensurecompetitioninserviceprovisionandhenceimprovedcustomerserviceatlowercosts.Ascompetitionhasnotyetmaterialisedinpractice,thebene®tsofseparationmayturnouttobesmallrelativetothecostsoflossofco-ordinationandtransactioncosts,suchascontractspeci®cationandenforcement(Dodgson,1995).AccordingtoBruzeliusetal.(1994),theverticalseparationofrailwayfunctionsinSwedenappearstohaveresultedinaloweringofthequalityofserviceprovidedbythetrackowner.Thismodelhasseriousimplicationsfortheoverallproductivityofrailoperations,giventhenatureoftherailwaybusiness,withitscloseindepen-denceofinvestments.Inaddition,thebargainingpowerofnewentrantstonegotiatecontractswithamonopolytrackowneractingtoachievecommercialobjectives,needstobeadequatelysafeguarded.Thecompetitivepressuresontrainoperators,whicharebeingsoughtthroughthismodel,areindangerofbeingabsenttotheinfrastructureprovider.

Themoste󰀹ectiveorganisationalmodeltobeadoptedneedstotakeintoaccountthespeci®caimsoftherailwayorganisation(s),aswellastheexistinglevelsofe󰁏ciency,pricesandcustomerservice,asdiscussedinFerreira(1997b).

Undertheverticalseparationmodel,infrastructureandtrainoperationswouldbemanagedindependently.Becauseofthelargeinvestmentneededandthefactthatthesocialbene®tisconsiderable,thegovernmentshouldusuallyprovidesupportfortheconstructionofHSR.Meanwhile,apassengertransportationcorporation(PTC)wouldmanagetheoperationofHSR.ThestructureofthisapproachisshowninFig.2.

Advantages:

·Thesplitmodemodelenablesthereleaseofrailwaytransportationenterprisesfromtheheavyburdenofowning®xedinfrastructureandallowsthemtocompetefairlyinthemarketplace.·Thismodelwouldprovideforaclearlyde®nedrelationshipbetweenthecompaniesandthegovernment.

·Facilitatescostcontrolandhenceincreasesthepotentialforcommercialpro®ts.

282W.G.Wongetal./TransportationResearchPartA36(2002)277±289

Fig.2.Structureofsplitmodealternative.Disadvantages:

·Thegovernmentwouldneedtoinvestsu󰁏cientfundstobearthecostsoftheproject.

·TheneedtocoordinatetherelationshipbetweenthePTCandIC,aswellastherelationshipbetweenthePTCandtherelativeRABs.

·Prolongedpayo󰀹periodofinvestmentininfrastructure.

ThePTCcouldtaketwopossibleforms:(a)Adirectlyjointstockcorporation.Underthismodel,thecorrespondingrailwayadministrativebureauxbecomethemainpartnersandpartic-ipateinthedailybusiness.ThiswouldfacilitatethecoordinationoftrainoperationbothonHSRandonconventionaltrack.(b)Initially,alimitedHSRcompanywouldbeformed.ThiscouldbeconvertedtoajointstockcorporationofHSRatalaterstage.

EconomicrelationshipbetweenPTC,ICandRAB.ThePTCwouldtakechargeoftheoperationoftheHSR,includingmarketing,trainoperationsmanagement,ticketingandothertechnicalmanagement.TheICwouldtakechargeofthemaintenanceofinfrastructure,communicationandsignallingequipment,powersupplyequipment,aswellaspropertydevelopment.ItwouldimposeafeeonthePTCtopartiallypayfortheoccupationofHSR.CertainfeeswouldalsobeimposedontheRAB,asthemediumspeedtrainsoftheRABwouldusetheHSR.TheirrelationshipscanbedescribedbythediagramshowninFig.3.

2.3.Aggregativemode(combiningtrainoperationswithinfrastructuremanagement)

Thismodeisstillverypopularinrailwaytransportationaroundtheworld,astheinfrastructureandtrainoperationsaremanagedandownedasasingleentity.Allfunctions,suchascon-struction,maintenanceandoperationofHSR,wouldbeundertakenbyahigh-speedrailwaygroup(HSRG).ThestructureoftheHSRGisshowninFig.4.

W.G.Wongetal./TransportationResearchPartA36(2002)277±289283

Fare for usage of equipment on classic railwaySurrogate fee for ticketsFig.3.EconomicrelationshipsamongIC,PTCandRABs.Fig.4.Structureofaggregativemode.Advantages:

·FacilitatestheestablishmentofanincorporatedcompanyandallowstheHSRGtobecomeaself-operatingandself-pro®tingentity.

·FacilitatestheintegrationofallfunctionsoftheHSR.·Reducesinternalexchangecosts.

Disadvantages:

·Itdoesnotfavourtheestablishmentofcompetitionmechanismsbecauseofthelargecapitalinvestmentrequired.

·Doesnotallowtheanalysisofcostcontrolascomparedwiththesplitmode.

TheformsoftheHSRgroupcorporation.TherearetwowaystoestablishanHSRG.Onewouldbetodirectlycreateajointstockcorporation.Thiswouldrequireanextendedperiodoftimetoevaluatetheassetsofrelevantpartners.ThesecondwayistocreatealimitedcompanyofHSRandthentoconvertittoajointstockcorporationofHSRatalaterdate.

EconomicrelationshipbetweenHSRGandRABs.TheHSRGwouldtakechargeofallmattersconcerningtheHSR.Byusingthesalesnetworkoftheconventionalrailways,theHSRGcouldsaveoninvestmentandmaybemoree󰀹ective.TheonlydisadvantageisthesurrogatefeepayedtotheRAB.TheHSRGwouldpaythefeeforusageofequipmentoftheconventionalrailway,suchasstations,motorcardepots,etc.Ontheotherhand,themediumspeedtrainsoftheRABwouldborrowfromtheHSR,sotheyinturnpayafeetotheHSRG.

284W.G.Wongetal./TransportationResearchPartA36(2002)277±289

3.Evaluationoftheoperatingsystem3.1.General

ThealternativemanagementmodelsproposedinSection2havebeenassessedintermsofthemostlikelynetbene®ttobederivedfromtheirimplementation.Asystembasedonaseriesofappraisingindiceswasdevelopedtoassesseachmodel.Themulti-layerevaluationmethodproposedbySaaty(1978)wasusedatthisstage.Thismethodconsistsofthefollowingsteps:

1.de®nethegeneralgoalandthesub-goalstobeachieved(G);2.de®nethesetofappraisingindicesrelevanttoeachsub-goal(I);

3.foreachofthealternativesbeingtested,assignrelativeweightsforeachsub-goal(B1);4.foreachsub-goal,assigntheimportanceofeachindex(B2);

5.Assignvaluesforeachindextorepresentthedegreetowhichitimpactsoneachalternative(B3);

6.EstimateanoverallrankingforeachalternativebasedonthethreematricesB1;B2andB3.Thismethod,whichhasbeenwidelyappliedinsolvingcomplexproblemsofdecisionmaking,isparticularlywellsuitedtocaseswheretherearecon¯ictinggoalsandsub-goalstobemet.Wheneachofthosesub-goalshavedi󰀹erentlevelsofimportance,itisnecessarytouseanevaluationframeworkwhichproducesasinglerankingforthealternativesbeinginvestigated.

Oneofthemainadvantagesofthemethodologyproposedhereisitstransparencyandeaseofapplication.Therelativerankingscanbechangedtore¯ectdi󰀹erentweightsgiventoeachoftheobjectivesbythemainstakeholders.Inthisway,thesensitivityofthe®nalresultstochangesintheinputassumptionscanbeeasilytested.

Establishmentofahierarchymodel.Inthecurrentapplication,ahierarchymodelwasused,consistingof4layersand11indices.Thelatterwereselectedspeci®callytocovertherangeofindicatorswhichdecisionmakersarelikelytobemostconcernedwithwhenassessingeachofthethreealternativemodelsproposedhere.Indiceswhichhavelessin¯uenceontheselectionoftheoptimalalternativewereexcludedatthisstage.However,themethodologycaneasilybeadaptedtoincludeotherindicessuchaseconomicimpactfactors,inadditiontothe®nancialfactorswhichhavebeenincluded.

The®rstlayerconsistsofthegeneralgoalwhichistoselectapreferredmodelfromthethreealternatives.Thesecondlayerincludesthreesub-goals:®nancialbene®ts,sociale󰀹ectsanden-terprisemanagement.Inthethirdlayer,11indiceshavebeenputforwardthatcorrespondtothethreesub-goals.Thelastlayercomprisesthethreealternatives(seeFig.5).

Determinationoftheindexvalue.Mostindicesusedarequalitativeinnature.Thevaluesusedforeachindexwerearrivedatusingapanelofexpertsineach®eld.Thebasicrelationshipsbe-tweeneachindexandthethreealternativesarediscussedbelow.

Fundraising.TheBSHSRisalarge-scaleprojectandtheraisingoffundsisakeyfactortoitssuccess.Theexpertsproposedseveralmethodsoffundraisingsuchas:(a)Governmentinfrastructureinvestmentfunding.

(b)AdditionalchargesonfreightandpassengerscarriedontheexistingrailwaybetweenBeijingandShanghai.

(c)Stocksorbondsissuedondomesticmarketorinternationalmarketandbankloans.

W.G.Wongetal./TransportationResearchPartA36(2002)277±289285

Fig.5.Structureofthemulti-layerevaluationsystem.Alternative1hasless¯exibilityinfundraisingthantheothertwo.Alternative3maybemorefavourableatfundraisingthan2,asitintegratesthewholeHSRsystem.Thepreferredorderforraisingfundsisalternatives3,followedby2and1.

Costcontrol.Alternatives2and3aremoree󰀹ectivethan1becausetheywillresultininde-pendententitiesbeingformed.Alternative3reducesinternalexchangecostscomparedto2.However,itiseasiertocontrolcostsinthesplitmodemodelthanintheaggregativemodel,becausethePTCofHSRdoesnotneedtoconsidertheburdenof®xedassets.Thepreferredorderforcostcontrolisalternatives2,3and1.

Productivity.ThePTCinthesplitmodecanenterthetransportmarketwithouttheheavyburdenof®xedcapitalrequirements.Alternative2hasthepotentialtoresultinhigherproduc-tivitythantheothertwoalternatives.Alternative3willbemoreproductivethan1becauseofitsmodernizedoperatingsystem.

Propertydevelopment.Alternative3hasaslightadvantageasitcanbettertailorlanduseandpropertyneedstothelevelofpassengerdemandbeingforecast.

Pro®tability.Thisisaveryimportantindexforanenterprise.Thesplitmodemodelwouldhavepotentiallythehighestpro®tability,asthePTChastheleastcapitalburden.Themodernincor-poratedmode(alternative3)couldhaveamoree󰁏cientsystemanditspro®tabilityislikelytobehigherthaninalternative1.

FurtherDevelopment.TheestablishmentofamodernenterprisesystemwillensurethefurtherdevelopmentofHSRinChina.Alternatives2and3willbemorefavorableinthisrespect.

286W.G.Wongetal./TransportationResearchPartA36(2002)277±289

Consideringcurrenttrendsinworldrailwayoperationsandmanagement,alternative2wouldhaveahighervaluethanalternative3.

Improvementinthelevelofservice.Inalternative2,thePTCcancompetefairlywithothertransportationmodesandgraduallyimproveitsservicelevelcomparedwiththeothertwoal-ternatives.Asthetraditionalmodelacks¯exibility,itislikelytobeinimprovingservicelevels.Thepreferredorderforlevelofserviceisalternatives2,3and1.

EnthusiasmoflocalgovernmentandRABs.Alternative1isthemostfavorableforRABsbe-causetheycangetthelargestpro®tsfromtheprojectwithoutlargeinvestment.Localgovern-mentsmay®ndeasiertobecomeashareholderinalternative2comparedto3.ThepreferredorderforenthusiasmoflocalgovernmentandRABsisalternative1,2and3.

Coordination.Sincealternative2callsforcoordinationbetweenthePTC,theICandRABs,thismodelhasthelowestvalue.ItwouldbeeasiertocoordinatetheHSRGwithRABsthantoco-ordinatebetweenthethreeRABs.Thepreferredorderforcoordinationisalternative3,1and2.ManagementofMST.Becauseofthefeaturesofpassenger¯owandthecurrentsituationoftheChineserailways,itisindispensabletooperatemediumspeedtrainsontheBSHSRforalongperiodaftertheBSHSRisconstructed.Alternative1hasanadvantagebecausethesameownerwouldoperateboththeHSRandtheexistingrailnetwork.

Establishmentofmodernenterprisesystem.Underalternatives2and3thereisaneedtoes-tablishamodernizedenterprisesystem.Alternative1needsalongerperiodtoconverttoamodernsystem.3.2.Results

Inordertoarriveatanoverallrankingforthethreealternatives,asurveyofexpertswasundertaken.Theresultsofthatsurveyenabledcomparisonmatricestobebuiltforeachofthe11indicesselected.Theprecedencematrix,showingthevalueoftherelativeweightofeachalter-nativecorrespondingtoeachindex,isgivenby

010:2040:2630:270:2820:2520:2440:2380:3930:3490:3930:208

B3󰀁@0:3060:3950:4050:3520:3790:390:4290:3270:2330:280:396A:

0:4900:3420:3240:3660:3690:3660:3330:280:4190:3270:396InmatrixB3,thesumofeachcolumnrepresentingeachindexis1.0.Thehighereachindividualvalueis,thehighertherelativescoreforeachalternative.Forexample,0.204,0.306and0.49representtheweightsforthethreealternativesintheindexrelatedtofundsraising.Inthisex-ample,alternative3hasanobviousadvantageinfundsraising.Whilstalternative1istheweakestinrelationtothesameindex.

Thedecisionontheweightstobeassignedtoeachfactoriscrucialinthedecision-makingprocess.Theresultsofthesurveyofexperts,expressedbyaprecedencematrixwhichre¯ectsthethreesub-goals,aregivenby

0:380:170:160:110:18

0000B2󰀁@000000

0

0

0:380

00000:430:1900000:280:3

1T

0

0A:0:42

W.G.Wongetal./TransportationResearchPartA36(2002)277±289287

InmatrixB2,the11indicesaredividedintothreegroupscorrespondingtothreesub-goals.Forexample,thevalueof0.38,whichistherelativeweightoffundraisinginthe®rstgroup,isthemostimportantfactorinthisgroup.Inotherwords,the®rstrowrepresentstheweightsofthe11indicestosub-goalone,namely,®nancialbene®t.

Forthe®rstlayer,thepriormatrixisshownbelow,inwhichtheweightofthethreesub-goalsisgiven.Thesub-goalofsocialbene®tshasthehighestvalue,astherailwayisapublicinfrastructuregood.

010:355

B1󰀁@0:395A:

0:250Thus,accordingtotheprincipleofthehierarchymethod,theprecedencematrixforthethreealternativesintermsofthethreesub-goalsisgivenby

010:2410:2700:303

W3󰀁󰀂B3󰀃3Â11󰀂B2󰀃11Â3󰀁@0:3550:3950:316A:

0:4040:3350:381Fromthismatrix,itcanbefoundthatalternative2isthebestaccordingtosocialbene®t,whilealternative3wouldbeoptimalintermsoftheothertwosub-goals.

The®nalmatrixforthethreealternatives,intermsofthegeneralgoal,canbeestimatedby

W4󰀁W3B1󰀁󰀂0:268;0:361;0:371󰀃T:

Fromtheaboveresults,itwasfoundthatalternatives2and3arevaluedmuchhigherthanal-ternative1.

3.3.Limitationsofthemethodology

Themethodusedtorankalternativerailwaymanagementmodelsisabletoprovideaquan-ti®ed®nalrankingwhichcanbeusedtoselectthosealternativeswithhighrelativescores.Lowerrankedalternativescanbediscardedfromfurtheranalysisatthisstage.Therefore,themethodisappropriatefora®rstscreeningofalternativestobefurtherassessedduringamoredetailedinvestigationofimpacts.

Theuseofrelativeweightsatseveralstagesintheprocedureinordertoranksub-goals,indicesandalternatives,reliesonalargedegreeofsubjectivityonthepartofthepanelofexperts.Thefactthattheprocesscallsforsinglevaluestobeassignedtoeachweightmeansthatadi󰀹erentsetofexpertscouldyielddi󰀹erentvaluesforthesameweights.Althoughthisinherentsubjectivitycanbeminimizedbytheappropriateuseofexperts,theoverallresultsobtainedshouldbeseenascentralvalueswithuncertaintydistributionsattachedtoeachrankingvalue.Asaresult,alter-nativeswhichhavecloseoverallperformancerankingsarelikelytoneedtobefurtherinvestigatedbeforea®nalresultcanbeobtained.Thelevelofuncertaintyintheestimationoftherelativeweightsforeachindexislikelytobehigherifthealternativesbeingassessedarenewanddi󰀹ersigni®cantlyfromestablishedmanagementmodels.ThisisthecaseofnewrailwaymanagementandoperatingmodelsappliedinChina.Thepanelofexpertsusedtoarriveattheweightsforeach

288W.G.Wongetal./TransportationResearchPartA36(2002)277±289

indexneedstoundertakeimpactassessmentswhichmaybedi󰁏culttoquantify,intheabsenceofpastexperienceintheChineserailwaycontext.

Anotherlimitationofthemethodisthepotentialforbiasonthepartoftheanalystthroughthede®nitionofappropriateindicesforeachsub-goal,aswellastherelativeimportanceofthoseindices.Theexclusionofsomeindices,orthedownplayingoftheirimportance,couldin¯uencetheoverallresultssigni®cantly.

Theimpactoftheshortcomingsofthemethodologyissigni®cantlyreducedbythewayinwhichtheentireprocessismadetransparenttothedecision-maker.Sincetheweightsusedateachstagearesubjecttoscrutinyandreview,itispossibletodetectandcorrectforanybiaswhichmaybeevident.4.Conclusions

TheresultsimplythattheChineserailwayshavetoembarkonsigni®cantreformintheop-erationalsysteminordertoadapttoamarketeconomy.ThecentralgovernmentofChinahasdecidedtocreateaHSRbetweenBeijingandShanghai,andwehaveproposedtheestablishmentofamodernentitybyapplyingthepriorexperiencesofindustrialisedcountries.

Comparingthesplitmodewiththeaggregativemode,itseemsthatthelatterisslightlymoreattractive(0.371)thantheformer(0.361).Thismaybebecausethe®rstindex(compensationandraisingofconstructionfunds)hasaveryimportantin¯uence.Withthereinforcementofanin-tegratednationalpowerandthepushforrailwayreform,thesplitmodemaybecometheoptimalalternative,althoughtheaggregativemodeliscurrentlyrecommended.

Therecentreformpolicyofnationalenterprisesshowsthetrendinestablishingamoderncorporationsystem.TheconstructionandoperationofBSHSRwillbemanagedbyacorporationbody,thoughthemaininvestmentwillcomefromthecentralgovernment.Itisproposedtocreateanewindependententerpriseoutsidetheexistingrailsystem.Therefore,theHSRcorporationneedstocoordinatewiththethreerailwayadministrativebureaux(Beijing,JinanandShanghai).TheHSRwillplayaveryimportantroleinimprovingthetransportationlevelofservicebetweenlargeurbancentres.5.ListofacronymsBSHSRHSRHSRGHSTICMORMSTPTCRAB

High-speedrailwaybetweenBeijingandShanghaiHigh-speedrailway

High-speedrailwaygroupHigh-speedtrain

InfrastructurecompanyMinistryofrailwayMediumspeedtrain

PassengertransportationcorporationRailwayadministrativebureau

W.G.Wongetal./TransportationResearchPartA36(2002)277±289289

Acknowledgements

TheworkdescribedinthepaperwasfullysupportedbyagrantfromtheResearchGrantCounciloftheHongKongSpecialAdministrativeRegion,China(ProjectNo.PolyU5043/98E).

References

Bruzelius,N.,Jensen,A.,Sjostedt,L.,1994.Swedishrailwaypolicy:acriticalstudy.DepartmentofTransportationandLogistics,ChalmersUniversityofTechnology,Goteborg.

Dodgson,J.,1995.Separatingrailwayinfrastructureandoperations:theBritishexperience.In:FourthInternationalConferenceonCompetitionandOwnershipinLandPassengerTransport.Rotorua,NewZealand.

Ferreira,L.,1997a.Planningsingletrackrailoperations:anoverview.TransportationResearchPartA31A(4),335±348.

Ferreira,L.,1997b.Railtrackinfrastructureownership:investmentandoperationalissues.Transportation24(2),183±200.

Han,B.M.,1997.L'optimisationdelaplani®cationetd'exploitationd'unr󰀁eseauferroviaire󰀂agrandevitesse,Ph.D.thesis,LiegeUniversity,Belgium.

Han,B.M.,Sun,Q.X.,1993.Thestudyofhigh-speedtrainoperationplanbetweenBeijingandShanghaiCollectionofhigh-speedtrainbetweenBeijingandShanghai.NorthernJiaotongUniversity,China.

Hu,S.J.,1998.Studyontheovertakingorganizationofmediumspeedtrainbyhigh-speedtraininasection.JournaloftheChinaRailwaySociety20.

Hilmer,F.G.,Rayner,M.R.,Taperall,G.,1993.Nationalcompetitionpolicy.ReportbytheIndependentCommitteeofInquiry.AGPS,Canberra.

Jansson,J.O.,Cardebring,P.,1989.Swedishrailwaypolicy:1979±88.JournalofTransportEconomics&Policy23(3),329±337.

Nash,C.A.,PrestonJ.M.,1994.Competitioninrailtransport:anewopportunityforrailways.WorkingPaper397,InstituteforTransportStudies,UniversityofLeeds.NationalStatisticalYearbook,1997.China.

Qian,Z.H.(Ed.),1995.IntroductionofHigh-SpeedRailway.RailwayPublishHouseofChina,Beijing,China,pp.20±31.

Saaty,T.L.,1978.TheAnalyticHierarchyProcess.McGrawHill,NewYork.

Wang,Q.Y.,1999.Reviewoftransportationin1998andprospectoftransportationin1999.IntegratedTransport1,1±3.

Xiao,X.,1994.Actualityanddevelopmenthistoryofrailwayoperationsysteminabroad.Studyofrailwaydevelopment.China.