The Effects of the High Speed Railway on Urban Development: International Experience and Potential Implications for China - Effects of the High Speed Railway on Urban Development

(1)

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

The Effects of the High Speed Railway on Urban Development: International

Experience and Potential Implications for China

Yin, M.; Bertolini, L.; Duan, J.

DOI

10.1016/j.progress.2013.11.001

Publication date

2015

Document Version

Final published version

Published in

Progress in Planning

Link to publication

Citation for published version (APA):

Yin, M., Bertolini, L., & Duan, J. (2015). The Effects of the High Speed Railway on Urban

Development: International Experience and Potential Implications for China. Progress in

Planning, 98, 1-52. https://doi.org/10.1016/j.progress.2013.11.001

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

(2)

The

effects

of

the

high-speed

railway

on

urban

development:

International

experience

and

potential

implications

for

China

Ming

Yin

a,

**

,

Luca

Bertolini

b,

*

,

Jin

Duan

a

SoutheastUniversity,China b

UniversiteitvanAmsterdam,TheNetherlands

Abstract

Theunfoldinghigh-speedrailway(HSR)networkisexpectedtohaveagreatimpactonChinesecities.Thispaperdiscussesthe internationalexperienceofthedirectandindirectdevelopmenteffectsoftheHSRnetworkoncitiesattheregional,urbanand station-arealevel.ItthendiscussesthepotentialdevelopmentimplicationsandplanningchallengesforChinabytranslatingthe internationalexperienceintoaChinesecontext.Finally,futuretopicsforresearchareidentified.

#₂₀₁₄_Elsevier_Ltd._All_rights_reserved.

Keywords: Effectsofhigh-speedrail;UrbandevelopmentChina;Planningchallenges;Researchchallenges

Contents

1. Introduction ... 2

2. Directeffects... 4

2.1. Reducedtraveltime ... 4

2.2. Impactonthemodalshareofrailwaysinthetransportmarket ... 4

2.3. Towardsanintegratedtransportsystem... 7

2.3.1. Aircraft:fromcompetitiontocooperation... 7

2.3.2. Integrationwithtraditionalrailwaysystemsandexpressbuses... 7

2.3.3. IntegrationbetweenHSRandurbantransportnetworks ... 8

2.3.4. TheHSRstation:aconnector ... 9

3. Indirecteffects:regionallevel... 10

3.1. Agreaterdevelopmentgapbetweentheconnectedandtheunconnected ... 10

3.2. RedistributionandrelocationalongtheHSRcorridor... 14

3.3. Areconstructionoftheurban-regionalsystem... 17

3.3.1. CommuterHSRandanewmetropolitanprocess... 17

3.3.2. LongdistanceHSRandanintegratedfunctionalregion ... 17

www.elsevier.com/locate/pplann

*Correspondingauthorat:DepartmentofPlanningGeographyandInternationalDevelopmentStudies,UniversityofAmsterdam,Plantage Muidergracht14,1018TVAmsterdam,TheNetherlands.Tel.:+31205254007.

**Correspondingauthorat:SchoolofArchitecture,SoutheastUniversityatNo.2SipailouRoad,Nanjing210096,China.Tel.:+862583793040. E-mailaddresses:l.bertolini@uva.nl(L.Bertolini),ym2010seu@gmail.com(M.Yin).

http://dx.doi.org/10.1016/j.progress.2013.11.001

(3)

4. Indirecteffects:urbanlevel ... 18

4.1. Acatalystfortherestructuringofurbansystems... 18

4.1.1. Acatalystfortherestructuringoftheurbanspatialstructure... 18

4.1.2. Anopportunityfortheurbaneconomy... 19

4.2. AsynergybetweenHSRandurbandynamics... 19

5. Indirecteffects:station-arealevel... 21

5.1. HowtodefineanHSRstationarea... 21

5.2. Thegeneralfunctionofastationarea... 23

5.3. Propertydevelopmentinthestationarea... 24

5.4. Urbanqualityofthestationarea ... 24

5.5. Thedevelopmentofstationareas:acomplexprocess... 25

6. Theinternationalexperienceofeffectsofthehighspeedrailwayonurbandevelopment:summaryoffindings.... 26

7. PotentialimplicationsandplanningchallengesforChina... 27

7.1. ThepotentialimplicationsandplanningchallengesofdirecteffectsofHSR... 27

7.1.1. GenerationofHSRtraffic... 27

7.1.2. Realizationofamodalshift... 30

7.1.3. Developmentofanintegratedtransportsystem ... 31

7.2. ThepotentialimplicationsandplanningchallengesofindirecteffectsofHSRontheregionallevel... 33

7.2.1. Enlargingthegapbetweentheconnectedandnotconnected:populationgrowthandeconomicactivity 33 7.2.2. RedistributionalongtheHSRline:tourism,businessandtheknowledgeeconomy... 33

7.2.3. TheHSRasafactorinthereconstructionoftheChineseurban-regionalsystem... 34

7.3. ThepotentialimplicationsandplanningchallengesofindirecteffectsofHSRontheurbanlevel ... 36

7.3.1. AcatalystforreconstructingtheChineseurbansystem... 36

7.3.2. HSRandtheneedforsynergywiththestrategyofurbantransformation ... 39

7.4. ThepotentialimplicationsandplanningchallengesofindirecteffectsofHSRonthestation-arealevel.... 41

7.4.1. Abalancebetweennodeandplace... 41

7.4.2. Copingwithrisinglandvaluesandrealestaterentsintheareaaroundthestation... 42

7.4.3. Qualityofplaceofthestationarea:urbandesignandbeyond... 42

7.4.4. Acomplexstationareadevelopmentprocess:thelimitsofaphysicalplanningtraditioninChina.... 43

7.5. PotentialimplicationsinChinaofHSR’seffectsonurbandevelopment:summaryoffindings... 43

8. Afinalnoteonresearchchallenges... 47

References ... 49

1. Introduction

High-speed rail (HSR) is expectedto have a great impact on Chinese cities. China’s emerging HSR network, at 11028km, is already the biggest in the world (data as of December 2013: http://www.nra. gov.cn/zggstlzt/).This vastnetworkdrasticallyreduces travel times between major cities. It also binds cities togetherasintegratedurban-regionalareasandactsasa catalyst for reshaping their urban systems. Although China’sconstructionof HSRsharescommon elements withsimilar,longer-standingprojectsinWesternEurope, Japan,KoreaandTaiwan,theuniquemagnitudeandpace ofmodernChineseurbanization,aswellasthestrategies adoptedtomanageit,presentsomenotabledifferences. Improving or at least not further depressing the qualityoflifeincitiesinthefaceofrapid,uncontrolled urbanization hasbecome the keychallenge of spatial development in China. Across the country, the

urbanization level grew from 36.22% in 2000 to 49.68% in 2010 (6th Census of Population inChina, 2010).Meanwhile,theurbanbuilt-upareahasexpanded from21,847.64km2in2000to40,533.8km2in2010, anincreaseof85.5%(Wangetal.,2012).Theresident population living in cities has increased from 455 millionin2000to665millionin2010(6thCensusof PopulationinChina,2010),oranincreaseof46.15%in tenyears.Thisrapid,extensiveurbanization,coinciding withthecountry’seconomicriseoverthepast30years, hascreatedhugechallenges,includingthedepletionof scarce land resources, disruption of natural environ-ments,airpollution,trafficcongestion,poorlivingand poor public healthconditions. However, itis believed thatrapidandextensiveurbanizationissettocontinue. By2015,theurbanizationlevelisexpectedtorise5% with respect to 2010 (12th Five-Year Plan, 2011), meaningthateven morepeoplewould crowdinto the cities andthat the built-up area would be larger still.

(4)

Meanwhile,improvingthequalityofurbanizationinall itsecologicalandsocialdimensionshasbeengivenhigh priority inthe 12thFive-YearPlan,whichisthe most importantcurrentnationalpolicydocument.TheHSRis amajor infrastructurethatisexpected tosignificantly affect urban development. Insight into these impacts, andofwaysofinfluencingthem,iscrucialforanyplan aiming todirect the development of acity or region. Accordingly,thispaperlooksattheroleofHSRinthis specialphaseofChina’srapidandextensive urbaniza-tion, and the opportunities and threats it poses in improving thequalityofurbanization.

Atthenationallevel,eventhoughthereisnoofficial national spatial strategy, some clear directions for the spatial development of China have been set. The development direction has transformed from focusing ontheEastCoastinanunbalancedwayatthebeginning ofthereformingandopening-upinthe1980stoamore balanced development between inlands and the East Coastinordertoreducetheever-growingdevelopment gapbetweenthetwo(Li,2012).Thestrategycanbeseen reflectedinforinstancethepoliciesof‘‘The Develop-mentofChineseWesternRegions’’(CPGPRC,2000),or ‘‘RisingofCentralChina’’(CPGPRC,2006)’’.Theissue israisedofwhetherthehigh-speedrailwaycansupport the national spatial strategy of the coordinated and balanceddevelopmentacrossthecountry.

This largely implicit national spatial strategy has been more explicitly implemented by many spatial plans of megacity regions. The Chinese central governmenthasapprovedmanymegacityregionplans since the year 2000, especially following the 2008 financialcrisis,withtheaimtopushthedevelopmentof differentareasinChina.Theseplanssetoutthespatial strategies for improving the connection between different cities in the megacity regions, coordinating industrialdevelopmentbetween them,andreinforcing theircompetitiveposition.Thispaperdiscussestherole of HSRisthesemegacityregionstrategiesandhow it cansupportthem.

Onanurbanlevel,spatialtransformationhasbecome the most important strategyof many cities.After the rapid andextensivedevelopment of the last30 years, more andmore citiesfacethe challenges of whathas beencalledthe‘‘urbandisease’’,whichincludesurban sprawl, a poorly built physical environment, traffic congestion and the loss of urban and environmental quality, especially for the megacities of Beijing, ShanghaiandGuangzhou.Inordertosolvethis‘‘urban disease’’throughspatialtransformation,manypractices have been implemented such as adopting a so-called polycentricurban system, developing newtowns,and

encouraging the deployment of public transport. The interrelationships between high-speed railway and spatial structure are considered as a critical factor in thedevelopmentofcities.Thispaperwillhighlightthe roleofHSRintheurbanspatialtransformationstrategy andhow torespond the opportunities it presents at a locallevel.

The development of the areas that surround HSR railwaystationshasbecomeaheatedtopicofdiscussion inChina.Thestrategyforthestationareaistocombine thedevelopmentof thestationwiththatof anewcity centre or new town. HSR is considered to be an importantconditioninthedevelopmentofstationareas; this paper considers the role HSR will play in that development.

Alloftheseissuesmakeitimportanttopayattention tothepotentialeffectsofHSRonurbandevelopmentin China.Althoughtherearesomediscussionsaboutthis topicbasedontheinternationalexperienceandthereis a recent publication pointing out the challenges stemmingfromHSRdevelopment(Chen,2012),there isstillnosystematicdiscussionrelatingtoactualurban developmenteffectsabroad, potentialeffectsinChina and ensuing planning and research challenges. This paper aimstofillthisgap.Accordingly,we mustfirst assessthe impactofHSRabroadandthendiscussthe Chinesecontext.

Due to the very recent nature of HSR and urban developmentinChina,littleevidenceexistsofitsactual impactoncities.Thisobservationprovidestherationale for our approach andthe structure of the paper:first looking for evidence of actual impacts in contexts where enough time has passed to observe them, and thenreflectingonpotentialimpactsinChina,basedon thespecificcharacteristicsoftheChinesecontext,and, whenavailable,onemergentevidence.

Thepaperaddressesthefollowingsetsofquestions: 1.What are the urban development effects of HSR?

Whatare theexperiencesof othercountries? 2.Howshouldplannersrespondtotheexpectedeffects

ofHSRinChina,giventhespecificitiesofthelocal context?WhatcanChinalearnfromtheexperiences ofothercountries?

3.What are important areas for further research for China?

The first set of questions is answered through a literature review covering the urban development effectsof HSRinJapan(Shinkansen),France(TGV), Spain(AVE),Germany(ICE)andothercountries.Inthe literature,thereareusuallytwodifferentperspectivesin

(5)

discussingtheurbandevelopmenteffectsofHSR.The first perspectivedistinguishes between direct,indirect andexternal effects.Directeffectsare thedirect costs and benefits of HSR for passengers (travel costs, reducedtravel times)andHSRbuilders andoperators (construction and operating costs, ticket revenues). Indirecteffectsare thecostsandbenefitsfor actorsin theeconomic systemthatdonotuse,buildoroperate the HSR themselves. External effects are costs and benefitsfor which no evidenteconomic value canbe assigned,suchaseffectsonthenaturalenvironmentor onthequalityoflife.Thesecondperspectivediscussed inthe literature classifies HSR’s effectsinto those at national,regional,cityandstationlevels,accordingto their spatial scope. These two perspectives are not mutuallyexclusive.Direct,indirectandexternaleffects canbeshownateachspatiallevelandtheeffectsateach spatial level include both direct and indirect compo-nents.

InordertoshedlightontheeffectsofHSRonurban development,this paper adoptsa combined perspec-tive. It focuses on the direct and indirect effects of HSR on urban development (external effects, while important, are still not reflected in enough consoli-datedfindingsintheliterature).Thedirect(transport) effectsconstitutearelativelywell-defineddomain,but onecuttingacrossdifferentspatiallevels;ittherefore seemsappropriatetodiscuss theseeffectsseparately, andbycombining thedifferentspatial levels.On the contrary, indirect effects are both more difficult to isolateandmorefocusedonaparticularspatiallevel, and so the discussion of indirect effects will be organized alongspatial levels.There arefewstudies focusingonthenationallevel, hencethey have been integratedintotheregionallevel, whichisalsomore directly relevant consideringthe urban development focusofthis paper.

Theliterature review addresses thedirect transport effects (Section 2) and indirect effects of HSR at a regionallevel(Section3);urbanlevel(Section4);and atalocal/stationarealevel(Section5).Itisimportantto noteherethatthisliteraturereviewframeworkischiefly awayof organizingthematerial,anddoesnotaimto provideexplanationsitself,for instancebysuggesting causal relationships. Combining these insights with thosespecifictotheChinesecontext,inSection7,we will answer the second question by discussing the potentialurbandevelopmentimplicationsandplanning challenges of HSR for China. Finally, answering the thirdresearchquestion,inSection8,thispaperpoints out the main HSR-related topics deserving future researchinaChinesecontext.

2. Directeffects 2.1. Reducedtraveltime

Reduced travel time is one of the most important effectsof HSR,whichoftenhelpsitwin politicaland publicsupport.Journey timeson high-speedlines are drastically lowerthanthose ofconventional railways. An illustration of this can be seen in Japan which inaugurateditsShinkansenhigh-speedservicein1964. Overthefirst11.5yearsofoperation,itisestimatedthat 977millionjourneysweremadeonthenetwork,saving passengersatotalof2246millionhoursovertheuseof conventional trains (Sanuki, 1980). Enormous time savings can also be seen in France, Germany, Italy, Sweden,Spain,SouthKoreaandTaiwan(seeTable1). Fast travel time has become an important reason for peopletochooseHSRasatransportmode.InItaly,for instance,71.2%ofHSRusersalongtheRome–Naples corridor declare that their main reason for choosing HSR is the reduced travel time (Cascetta, Papola, Pagliara,&Marzano,2011).

2.2. Impactonthemodal shareofrailways inthe transportmarket

Everynewmodeoftransportimpactsthedistribution of modal shares in the transport market. HSR is no exception. With the advantage of high frequency, shortertravel times, acomfortabletravelling environ-ment and ‘‘centre-to-centre’’ connections, HSR has a tremendousopportunitytowinmarketsharefromother transportmodesincertaincontexts,mainlydepending onthe timetravelgainsrelativetoothermodes.

The gain to passengers, in terms of time savings, varies according tothe distance of the journey(Hall, 2009; Pepy &Leboeuf,2005; Pepy &Perren,2006). For trips less than 150km, HSR has little advantage over conventional trains or cars, which give greater flexibility and easier door-to-door connectivity. For distances between 150km and 400km, passengers reach their end destination faster with HSR than by driving on a highway. Between400km and 800km, HSRisthefastestmodeforpersonaltravelling,beating bothairandroadtravelintermsofdoor-to-doorjourney times. Beyond 800km, air travel is faster (Pepy & Leboeuf, 2005; Pepy & Perren, 2006). Japanese and EuropeanexperiencessuggestthatHSRshouldobtain 80–90%oftrafficinthedistancerangebetween150km and500km,andcapture50%uptoabout800km.The competitiverangecouldpotentiallyextendto1500km

(6)

atthemaximumpracticablepresentspeedof350km/h (Hall,2009).

There are many before-and-after-opening quantita-tivefieldsurveysofshiftsinmodalsharefor different HSR lines (see Table 2 for an overview). Such shifts involve bothtraffic substitution(from other modesto HSR) and new traffic generation (trips that were previously not taken). Looking at the cross-country comparison of modal share of HSR for similar distances,itisclearthatthevolumeoftrafficsubstituted to and generated by HSR is highly country-specific. TakeSpain’sAVEasanexample.The modalshareof railontheMadridtoSevillecorridoraftertwoyearsof operation(includingAVEandtraditionalrail)rosefrom 14% to 51%, while the market share of aviation decreased from 40% to13%; road journeys (busand car)droppedfrom44%to36%(Givoni,2006).Ofthose who chose HSR, some 32% switched from air, 25% fromcar,14%fromexistingrailservicesand26%were passengers whose journeys were triggered by the presence ofthe HSR(Vickerman,1997).Similarly,in FranceandJapan,thevolumeoftrafficgeneratedwas far greater than would be expected in an average context. For example, on France’s Paris–Lyon line, passenger numbersreached19.2 millionin1985– an increase of7 millionrelativeto1980,higher thanthe forecastincreaseof5.9million.Ofthis,49%wasnewly generated traffic. Likewise, in Japan, passenger numbers exceeded forecasts by 20% in the first year of operationoftheSanyo line(Osaka–Hakata).

By contrast, passenger numbers fell short of expectations in Germany, Taiwan, and South Korea (Chang&Lee,2008;Vickerman,1997;Yung-Hsiang,

2010).InthefirstfiveyearsofoperationofGermany’s ICE,accordingtotheestimationofDeutscheBahn,ICE traffic accounted for 28% of long-distance passenger revenues. Some 12% of this HSR traffic came from formerroadandairpassengers.Allfigureswerelower than expected (Vickerman, 1997). In Taiwan, the amount of passengers of HSR in the year 2010 was, afterthreeyearsofoperation,only50.5%oftheamount originally forecast (101,000 trips per day, while the originalforecastwasforadailyridershipof200,000). Similarly,inSouthKorea,theamountofpassengerson theKTX’sSeoul-Busanlineintheyear2004wasonly 46%of theamountoriginally forecast(Chang &Lee, 2008).

According to the authors cited, several country-specific reasons explain the differences in the modal shareofHSRacrossthecountriesstudied.Inthecaseof Germany,thespatial andurbanstructure may explain thelower-than-anticipateduseofHSR.Comparedwith France,theGermanurbansystemdoesnothaveagreat mono-centric focus, and urban centres are relatively closetoeachother.Furthermore,thereisan extensive inter-urbanhighwayinfrastructurewhichisfreetodrive on,whereasmotoristsinFrancemustpaycostlytolls. This may explain why HSR passenger numbers in FrancehavebeenhigherthaninGermany.Inthecaseof Taiwan,businesstraveldemandwaslowerthanforecast mainly due to manufacturing companies moving to Southeast Asia and the Chinese mainland in the intervening period (Yung-Hsiang, 2010). Also, poor integrationbetweendifferenttransportmodesisseenby thisauthorasareasonfortheweakperformanceofHSR inTaiwan. As is the case in many other countries,it

Table1

Traveltimes:HSRversusconventionalrailways.

Country/line Journeytime:

conventional railway Journey time:HSR Change Distance (km)

Yearopened Source

Japan Tokyo–ShinOsaka 6h30

0 ₄_h0 _38.33% ₅₁₅ ₁₉₆₄

Sands(1993)

ShinOsaka–Hakata 8h300 4h400 45.10% 554 1972/1975

France Paris–Lyon – 2h200 – 427 1981/1983 Bonnafous(1987)

Germany Frankfurt–Cologne 2h130 590 55.64% 177 2002 Ahlfeldtand

Feddersen,(2010)

Italy Rome–Naples 1h450 ₁_h₅0 _38.10% ₂₁₃ ₂₀₀₅ _Cascetta_et_al.₍₂₀₁₁₎

Spain Madrid–Seville 5h550 ₂_h₃₀0 _57.45% ₅₃₅ ₁₉₉₂ _De_Rus_and_Inglada,

(1997)

Sweden Stockholm–Eskilstuna 1h400 ₁_h _40.00% ₁₁₅ ₁₉₉₇ _Fro¨idh₍₂₀₀₅₎

South Korea

Seoul–Busan 4h100 ₂_h₄₀0 _36.00% _441.7 ₂₀₀₄

ChangandLee,(2008)

Seoul–Mokpo 4h400 ₂_h₅₈0 _36.43% _414.1 ₂₀₀₄

Taiwan Taipei–Zuoying

(Kaohsiung)

(7)

Table2

ChangeoftransportmodemarketsharesaftertheintroductionofHSR.

HSRsystems(nation/line) ModesplitafterthestartingoftheoperationofHSR Source

SanyoShinkansen line

Bythefirstyearofopening,in1975,passengernumbershadincreasedasmuch as40%comparedwith1971.

-55%transferringfromthenarrow-gaugelines

-30%fromothermodesoftransport(including23%fromairlines) -thestimulusofnewtravelaccountedfor6%

-theremaining9%increasewasofanunknownorigin

Okabe(1980)

TGVSudEstline Growthintrainuserscamefrom:

-33%airtravel -18%automobiletravel -49%newlyinducedtraffic

Bonnafous(1987)

ICEnetwork InthefirstfiveyearsofoperationofICE,passengernumbersmorethandoubled; ICEtrafficaccountedfor28%ofDeutscheBahn’slong-distancepassenger revenues.12%ofthistrafficwasdivertedfromroadandair.

Vickerman(1997)

AVE Madrid-Seville

Aftertwoyearsofoperation,rail’sshareofthemarketgrewfrom14%to51%; airdecreasedfrom40%to13%;andbusandcarsharedecreasedfrom44%to 36%.

Some32%ofHSRtravellerswereformerairpassengers;25%formercar passengers;14%weredivertedfromconventionalrailway;and26%werethe resultofnewlygeneratedtraffic.

Givoni(2006),

Vickerman(1997)

Svealandline Therailservice’smarketshareincreasedfrom6%toabout30%forregional trips.

Newrailpassengergrowthcamefrom: -30%theformerSJbusservice

-25%regionalpublictransportauthority’sformerbusservice-15%fromcar travellers

-30%fromnewtravellers

Fro¨idh(2005)

Rome-Naples Betweentheyears2005to2007theshareoftrainandcarchangedfrom49%and 51%to55%and45%,respectively:

-22.3%ofweekdayHSRtrainusersrepresentanewly-generateddemand (12.5%generatedbynewtrips,9.8%generatedbyanincreaseoftripfrequency)

-7.8%demandwassubtractedfromcartravel -0.6%demandwassubtractedfromairandbustravel

-69.2%demandwassubtractedfromIC(Intercity)andEC(Eurostar) conventionaltrains

Cascettaetal.(2011)

Gyeongbuline BeforetheopeningofKTX,theSouthKoreanHSR,fromApril2003toMarch

2004,themarketshareofautomobile,expressbus,airandconventionalrailwas 12.1%,7.8%,42.2%and38.0%,respectively.AftertheinitiationofKTX,the high-speedservicecaptured50.4%oftheentiretransportmarket,whilethe expressbus,airandconventionalrailhadtheirsharereducedto9.4%,4.7%, 25.0%and10.5%,respectively,intheperiodfromApril2004toMarch2005.

ChangandLee(2008)

Honamline BeforeKTXopened,fromApril2003toMarch2004,themarketshareof

automobile,expressbus,airandconventionalrailwas59.1%,14.3%,3.1%and 23.5%,respectively.Aftertheinitiationofthehigh-speedoperation,KTX captured21.1%oftheentiretransportmarket,whiletheexpressbus,airand conventionalrailhadtheirsharereducedto53.8%,14.9%,1.1%,and9.0%, respectively,intheperiodfromApril2004toMarch2005.

Taipei–Kaohsiung HSRenteredserviceinApril2007andhada25.03%inshareofthemarket Yung-Hsiang(2010)

thatmonth.Oneyearlater,ithada49.64%shareofthemarket.8%ofdemand wasnewafterHSRenteredtheintercitytransportationmarket:

BetweenApril2005andApril2008,marketshareofothermodeschangedonthe Taipei-Kaohsiunglineasfollows:

-Theintercitybusservicedecreasedfrom35.29%to22.28% -Conventionalrailwentdownfrom7.76%to2.50% -Airtraveldecreasedfrom28.73%to4.97% -Cartraveldroppedfrom28.22%to20.61%

(8)

appearsthatdevelopinganintegratedtransportsystem iscrucialtotheperformanceoftheHSR.Thispointwill beexpanded uponnext.

2.3. Towardsanintegratedtransportsystem

Developing an integrated system is not only beneficial to the system itself but also to passengers andthe regionsserved bythe lines.It isimportant to remember that HSR is only one element of a total transport system in an urban area and that it should contributetobringingpassengers‘‘door-to-door’’inthe vastmajorityoftripsreachingbeyondtheaccesspoints ofHSR.Aproperlyintegratedtransportsystem,where HSR is complemented by other modes, can avoid disorderly competition,encourageasensible distribu-tion of occupancy and stimulate cooperation among operators.Itisalsoworthrememberingthattheimpact of HSR may not only be beneficial to the cities it connects,butalsotothecitiesinsurroundingregions.In ordertoachievethis,itisdesirablethatanurbanregion be linked to a network of HSR lines, with regional services operating in harmony with HSR through seamlessconnectionsandtimetable-matching.Donein thisway,anintegratedtransportsystemcanimprovethe spatial penetration of HSR,puttingit within reach of morepassengers.Thefollowingsectionswilldiscussin moredetailtherelationshipswithdifferentmodeswith whichHSRcanintegrate.

2.3.1. Aircraft:fromcompetitiontocooperation HSR competes with air travel in medium to long distances. Hall (1991) has made some assumptions comparing the time differencesbetween HSR andair travelintermsofdoor-to-doorjourneys.Ina city-centre-to-city-centreHSRjourney,travellersareassumedtobe abletoreachthetrainstationin15minandthenthetrain in another five minutes; at the other end, 15min are allowedtoreachthefinaldestination.Forair,themodelis based on a 45min journey tothe airport andanother 60minforcheck-inandwaitingtime;attheotherend, five minutes are calculated to exit the terminal and another45mintoarriveatthefinaldestination.Thetotal accessandwaitingtimeisthusmodelledat35minbyrail and155minbyair.Onthisbasis,thebreak-evendistance between HSR and airis530km fora 200km/h high-speedtrainor960kmfora300km/hhigh-speedservice (Hall,1991).

Evidence also suggests that HSR competes strongly with air travel on many shorter-distance links between major metropolitanareas (see alsoTable 2). In France, between theyears1980 and 1985,the total numberof

journeysonthe450kmParis–Lyonroutewentupby56%. Pertrain,thenumberofpassengersroseby151%whilethe numberperaeroplanedeclinedby46%onaverage(van denBerg&Pol,1998).In Japan,whentheShinkansen Hikari trains began full-scale operation in 1967, the passengervolumeofNagoyaAirportwasreducedfrom fifthto19thpositionamong88nationalairports;passenger volumesshrankby75%andtooksevenyearstoreturnto theirpre-Shinkansenlevel(Sanuki,1980).

EventhoughHSRisacompetitortotheaeroplane,it canalsoprovidefast,comfortableaccesstotheairport and enlarge the service area of airports that are connectedwiththeHSRnetwork.Airportscanbenefit from HSR’s complementary service and develop higher-level services such as intercontinental flights. AtypicalexampleistheairportofLyon-Satolas,which isthethird-largestinFrance.SincetheParis–LyonHSR began in 1981, the volume of passengers using the airportinLyon–Satolasfellatfirst;however,becausea HSR terminus was built there, the airportwent on to benefit from improved accessibility to an enlarged radiusof 400km,which contributed tothe launch of intercontinental flights there. By 1996, international traffichadsurpassedthe50%markatLyon-Satolasfor thefirsttime.ThisexampleclearlyillustrateshowHSR andairtravelcancomplementeachother:betteraccess by HSR increases the development prospects of international flights, and a growing number of such flightsmakeitmoreattractivetolocateHSRstationsin airports(Thompson,1995;vandenBerg&Pol,1998). Therearetwoscenariosinwhichtheintegrationof HSRandairtravelcantakeplace.Onewayistocreatea direct HSR line to an airport (as in Paris, Lyon, FrankfurtorAmsterdam),orevenoffering aseamless rail–air journey possibility as is the case the case in Germany where Deutsche Bahn teams up with Lufthansa to offer code share ticketing and shared baggagehandling(Grimme,2006),orwiththe partner-ship between Thalys and KLM inSchiphol, Amster-dam’sairport.The otherscenarioforintegrating HSR andairtravelisthroughapublictransportlink(metro, regional railway or express bus) to connect the two modes such as between airports and HSR stations in LilleEurope,BrusselsSouth,MadridAtochaorLondon St. Pancras. In Table 3, we can see thatseveral HSR stations in Europe have good connections to airports throughpublictransportandcarlinks

2.3.2. Integrationwithtraditionalrailwaysystems andexpress buses

HSRreliesonintegrationwiththetraditionalrailway system,expressbusesandrelatedmodes.InJapan,the

(9)

greatsuccessofShinkansenislargelyattributabletothe reinforcement and improvement inthe service of the narrow-gauge lines that feed into the high-speed service.Similarly,whentheShinkansenwasextended to the southern city of Hakata, the number of buses operatingbetweenHiroshima andHamada toconnect withtheShinkansenwasincreasedandhigh-speedferry boatsfromMiharatoShikokuIslandwereintroduced. These initiatives reinforced and complemented the high-speedtrainservice(Okabe,1980).It isasimilar picture in the southern Italian region of Campania, where the HSR stations are part of an integrated transport strategy. The new HSR station of Naples connectsseamlesslywithregionalrailservicesoperated bytwodifferentprivatecompanies.TheHSRstationis alsointegratedwiththeregionalmetrosystem(Cascetta &Pagliara,2008).

Taiwan’sHSRstoryprovidesacontrastingexample; herethedecisionwastakentolocatemanyHSRstations on the outskirts of the cities away from traditional railway stations in the citycentres. Only four of the eight HSR stations in operation (Taipei, Banciao, Zuoying and Taichung) share services with conven-tional railway systems. The frequency of traditional railway also does not fulfil passengers’ requirements due to poor timetable matching; this is one of the reasonsforthepoorperformanceofHSRinTaiwanin termsofpassengernumbers(Yung-Hsiang,2010).Due tothelegacy ofpoor integration thatthiscreated, the Taiwanesegovernmentisnowaimingtostrengthenthe linksbetweentheHSRandtraditionaltrains.However, it faces major civil engineering (in centrally located stations)andtechnicalandoperational(inperipherally locatedstations)challengesindoingso.

Integration with traditional railway systems and expressbusesisveryimportant,butnoteasytoachieve. Firstly,plannershavetoconfronttechnicalissues,such as different track gauges, electrificationvoltages and signalling systems (Hall, 2009). Secondly, there are

operationalchallengessuchastimetable-matchingand integratedticketing.Thirdly,thereisacivilengineering andfinancialchallengeincreatingadditionalcapacity inprime citycentrelocations.Underground construc-tionsolutionsareoftenrequired;however,governments maynot bekeen ontheextracost anddisruptionthat tunnellingcanprovoke. On theotherhand,peripheral locationsneedtoconnectnewstationsandoldstations. This poses a different integration problem and the challengeoffindingtheextrafundingtoinvestinnew publictransportinfrastructure.

2.3.3. Integrationbetween HSRand urban transportnetworks

Asatransportmode,HSRrarelyprovidesa door-to-doorservice betweencities.Thus,integrationwiththe urban transport systemincludingpublic transport and car parking at each end of the journey is crucial. Adaptations to the urban transport system are nearly always required andthe level andfeasibility of these adaptations depends, chiefly, on the location of the station,whetheritiscentralorperipheral(vandenBerg &Pol,1998).

For central locations, public transport connections aretypicallybetter;however,caraccessibilityisusually poorer. According to many analysts, no matter how muchplannerswouldliketoseepeopletravelbypublic transport,goodcaraccessibilitytotheHSRstationscan be a decisivefactor, especially for businesstravellers whoplaceapremiumoncomfortandshorttraveltime (de Jong,2007).Withrespecttocaraccessibility, two elements canbe distinguished: roadconnections and, especially,parkingfacilitiesadjacent toHSRstations. Despitetherealitiesofurbantrafficpolicy(encouraging publictransport,limitedcar parkingspaces,effortsto discouragetheuseofcars),itdoesseemnecessaryand desirable to provide selective car access to centrally positioned HSR stations. Selective car access can be achieved by introducing or intensifying the price

Table3

ExamplesofconnectionsbetweenHSRstationsandairports.

HSRstation Airport Transporttime

bytrainfrom station

Connection time(h)

Via Traveltime

bycar (off-peaktimes)

LilleEurope ParisCharlesdeGaulle 0:49 1 Direct 1:50

LilleAirport 0:29 1.67 LilleFlanders 0:16

BrusselsSouth BrusselsNationalAirport 0:21 4 Direct 0:19

ParisCharlesdeGaulle(F) 1:51 2 Direct/ParisNorth 2:35

LondonSt. Pancras

LondonHeathrow 0:56/0:45 8 Direct(4)/Paddington(4) 0:41

(10)

mechanism on access roads and/or parking facilities neartheHSRstation.Anotherwaytoachieveselective car access is to lay out dedicatedaccess roads to so-called ‘‘destinationcarterminals’’suchas multi-story carparks(asattheHSRstationofLie`geinBelgium). Promotion of taxi transport canalso helpto keep the HSR station internally accessible for travellergroups that place a value on speed and direct door-to-door connectivity. Taxisoffer individualaccessibility with-outthe needforadditional(long-term)parkingspaces (van denBerg&Pol,1998).

By contrast, peripheral locations are characterized by easier car accessibility and worsepublic transport accessibility. At these locations, public transport integration is typically focused on the connection betweentheHSRstationandthecitycentre,ratherthan onorbitallinks.Poorpublictransportconnectionsarea reasonwhyHSRdoesnotalwaysliveuptoexpectations intermsofpassengervolumeordevelopmentimpactsat these stations (see also Section 3). Take the Gifu-HashimastationinJapan,forexample.SincethisHSR station opened almost 30 years ago, relatively little development has happened at the station or its immediatesurroundings. Oneof the mainreasons for thisfailureisthattheurbanpublictransportserviceto thecitycentrehasnotbeenimproved(Sands,1993).In France, an example of poor integration between the

HSRstationandthe citycentreisatAvignon,where, duetofinancialproblems,thereisonlyashuttlebusto connect the station with the city centre. The bus is affectedbyroadtraffic,leadingtodelaysorconnection disruptions during rush hours. This lack of intercon-nectivityhasimpactedthedevelopmentpotentialofthe station area (Todorovich, Daniel Schned, and Lane, 2011).

2.3.4. The HSRstation:aconnector

Ideally,differenttransportmodesshouldintegrateat HSR stations. The physical organization of Euralille, the HSR station at the heart of the city of Lille, in France,literally visualizes thisambition (see Fig.1). While in other instances this ambition may be less visible,itisequallyimportant.Therailwaystationisa nodewhichsupportsthetransferbetweenmodes,even ifthisconnection cannotbe completedinthe railway station for reasons such as transport organisation, or capacity. It is a ‘‘connector’’ (Van der Spek, 2003) whichjoinsupvariousformsoftransportandallkinds offunctionsrelatedtotravelling.

The ideal state in HSR stations is for seamless transfer. It should be safe, reliable, fast, hassle-free, comfortable and pleasant, thus meeting travellers’ demands. The design of the connector should reduce ‘‘transfer resistance’’; the station area should be

Fig.1. Euralille,integratingHSR,traditionalrail,localpublictransportandroadinfrastructureattheheartofthecity. Source:Euralille

(11)

spatially defined by all possible transfer routes connecting the different transport modes. The area neednothaveauniformshape–thespacewilldepend onthesituationofthestopsandmayextendtomultiple levelstoinclude indoorandoutdoorspace,as wellas public and private space (Peek & Louw, 2007). To enablethesetransfersandmeettheneedsofpassengers, a railway station must provide the following: secure access,loadingandstandingforvehiclesofallmodesto stop or be parked; clear spatial orientation for passengersmovingbetweentransportservices;waiting areasandinformationthatsupportspassengerswhoare making connections; and related services such as ticketing and lost-and-found offices (Zemp, Stauffa-cher,Lang,&Scholz,2011).

Tapiador, Burckhart, and Martı´-Henneberg (2009)

use a concept of inter-modality as measured by the distance between modal stops (such as bus stops, regional trainstops, car parks, bicycle racksand taxi ranks)andthetimerequired foratravellertotransfer from one mode to another, taking into account all possiblecombinations.Thisisappliedthrough empiri-cal research of 27 HSR stations in Europe. The empiricalworkshowsthattheSwisssystemprovidesan exampleofgoodpractice.Thetransfersystemthereis promotedbythefollowingfeatures:punctualandhigh density trains, bus and coach services; consistent departureandarrivalplatforms;timetablesthatoptimise combinations between different modes of transport; goodaccesstothestationforprivatecarsandtaxis;and safe storage areas for bicycles. Combined, these features helpprovide agood door-to-door servicefor commutersandothertravellers.Attheotherendofthe scale,theToledoHSRstationinSpainisrecognizedasa caseofpoorplanningforinter-modality.Caraccesswas penalizedasthenewsystemincludedanincreaseinthe cost of car parking near the station. The result was increased traffic congestion in neighbouring areas. Furthermore,theregionalbusstationishalfamilefrom the HSR and connecting local bus services are infrequent(Tapiador etal., 2009).

Sofar,wehavediscussedthedirecteffectsofHSR, focusingonthetransporteffects.Inthenextsectionwe willexaminetheindirecteffects,whichareofparticular interest to economists, urban geographers and urban planners.

3. Indirect effects:regionallevel

Ata regionallevel, HSRchanges the absoluteand relativeaccessibilityofdifferentcities.Thiscanincrease mobility opportunities, enabling businesspeople and

touriststotravellongerdistances.Throughchangesinthe relativeaccessibilityofcentres,HSRalsoinfluencesthe choice of location for individuals (housing, place of work)andforfirms(officesites),leadingtoaregional spatialredistribution.

A great deal of research both with empirical and model-based approaches hasbeen done toinvestigate HSR’s spatial development effects at regional level. Moststudiesfocusonthedisparitybetweenregionsthat areconnectedtoHSRandthosethatarenot.Theyalso examine the distribution of cities along HSR lines within theregionandhowthisstructurechangesafter theintroductionofHSR.Thisresearchwillbediscussed inthissection.

3.1. Agreaterdevelopmentgap betweenthe connectedand theunconnected

Researchpointsouttherelationshipbetweenregional economicandsocialdevelopmentandconnectiontothe HSR. In Japan, for example, Okabe (1980), Kamada (1980), Sanuki (1980), Hirota (1984), Nakamura and Ueda (1989), Amano and Nakagawa (1990), Sands (1993) and others have extensively investigated the Shinkansen regional development effects. In France,

Bonnafous (1987), Pieda (1991), Plassard (1989) and

Nyfer(1999),amongothers,focusedon theimpactof TGV. There isalso morerecent evidencefrom Spain, SouthKoreaandtheUK.Thepicture,intermsofcause/ effectrelationshipsis,however,morecomplex.InJapan, forexample,datashowsacloselinkbetweenGDPand ridership of the Shinkansen; the government there believes that transport infrastructure spurs economic growth, butshowingasimplecause/effectrelationship has proven difficult (Smith, 2003). In France’s TGV Atlantique, substantial growth has taken place at Le Mans, Nantesand Vendome afterconnection to HSR. However,thesewerecitieswherelocalconditionswere alsobuoyant(Banister&Berechman,2000).

Several studies show that the disparity between connectedandunconnectedcitiesandregionstendsto become larger as HSR may create new location advantages for the connected cities and regions, but alsodisadvantagesforcitiesandregionsnotservedby the new networks (Vickerman, Spiekermann, & Wegener, 1999).Thesedisparitiescanbe capturedby lookingatthe economicperformanceofbusinessesin information, communications and entertainment (van den Berg & Pol, 1998) that can be expected to be sensitivetoHSRaccess.Theseincludedevelopmentsin businessservices,urbantourism,andcongressandfair activities (see Table 4). Research has compared

(12)

Table4

DevelopmenteffectsofHSRontheregionalscale.

HSRsystems Developmenteffects Source

Population

StationlocationsontheHSRrouteshowedmarginally-higherthanaverage populationgrowth.

AmanoandNakagawa

(1990)

Tokaidoline CitieswithaShinkansenstationexperiencedgrowthratesthatwere22%higherthan citieswithoutanHSRstation.

Hirota(1984)

Threeofsixprefectures(provinces)thathaveoneormoreShinkansenstationgrew morethanthenationalaverage.ForthosewithoutanyHSRstation,somegrewbelow thenationalaverageandsomeevenshrunk.

NakamuraandUeda(1989)

TohokuLine CitiesclosetotheHSRgrewinpopulationby32%.Citiesinthesameregion,but locatedfurtherawayfromtheHSRline,sawnopopulationgrowth.

ObermauerandBlack(2000)

PrestonandWall(2008)

Madrid-Seville Thedisparitiesfrom1991to2001becomelargerthanfrom1981to1991between connectedandnon-connectedcities.AlongtheHSRline,therearetwolargecities, CiudadRealandPuertollano.CiudadRealsuccessfullyattractedpopulationand housinginvestmentfrominsideandoutsidetheprovince,whilePuertollanoshoweda completelydifferentpattern.Itspopulationgrewatarateofonly4.29%from1981to 2007;thisrateinCiudadRealwas38.9%.ThedevelopmentofCiudadRealis consideredtohavehappenedattheexpenseofthelatter.

Garmendiaetal.(2008),

Garmendia,deUren˜a,and

Coronado(2011)

HS1 Ashford,thecityconnectedtotheHSR,thepopulationreceivedanupliftof11%in the1990scomparedwiththeSoutheastofEnglandasawhole.

InterCity125/225 Within1hofLondon,bothHSRandnon-HSRtownsshowedlargeincreasesin population,incontrasttothereductionofpopulationinLondon.Amongthetowns within2hofLondon,allmainnon-HSRtowns,interestingly,experienceda noticeableincreaseexceptforthreetowns.However,thepercentageofpopulation increasewassmallerthanintownswithin1hofLondon.Evenmoresurprisingly, HSRtownsdemonstratedapopulationdeclineoraslowerincreasethanthenational average,exceptforYorkandCardiff.Fortownsover2hawayfromLondon, populationsdecreased,exceptinSwansea.

ChenandHall(2011)

Employmentandlabourforce

Tokaidoline 16–34%higheremploymentgrowthincitieswithHSRstations(intheretail, industrial,constructionandwholesalesectors).

Hirota(1984)

Employmentgrowthatstationlocationsof1.8%versus1.3%innon-stationlocations AmanoandNakagawa (1990)

Joetsuand Tohokuline

Employmentintheretailsectorroseby0.4%andby2.8%duetothecombinedeffects ofexpresswaysandHSR,respectively.EmploymentintheITsectorwasup22%for thecombinedHSR/expresswayeffect.Employmentreducedby3.6%incitieswithout expresswaysorHSR,

NakamuraandUeda

(1989)

Sanyoline Bothterminalstationsandintermediatestationsshowedsignificantgrowthinthefood andaccommodationsectors.

Specificemploymentsectorimpacts,suchasretailing,weresmallanddeclinedwith distancefromthestations.

Brotchie(1991),Hirota

(1984),Okabe(1980)

HS1 Ashford’semploymentreceivedanupliftof6%inthe1990scomparedwiththe

averagegrowthintheSoutheastofEnglandasawhole.

InterCity125/225 AmongHSRtownswithin1-htraveltimeofLondon,financialservices,information technology(IT)andotherbusinessserviceswererepresentedmoreheavilythanother services.Fortownswithin2htraveltime,thepicturewasmoremixed.Bristolhad dominanteconomicactivitiesinfinance,ITandotherbusinessservices,whileLeeds experiencedashiftfrompublicadministration,educationandhealthservicestowards finance,ITandotherbusinessservices.ButYork,Cardiff,NewportandSwanseawere characterizedbyhighemploymentinpublicadministration,educationandhealthsectors. ForthoseHSRtownsover2hawayfromLondon,representationwasstrongestinpublic administration,educationandhealthservicesbetweentheperiodof1995and2006,except forEdinburghandDarlington.Ontheotherhand,theserviceindustryinnon-HSRtowns, ingeneral,showedhigherpercentagesofemploymentinpublicadministration,education andhealthservices.

(13)

Table4 (Continued)

Regardingemploymentintheknowledgeeconomy,HSRtownswithin1hofLondon appearedstrong,whiletownswithina2-hradiusalsoperformedaboveaverage, strongestinknowledge-intensiveservices,exceptforDoncaster.ForHSRtownsover 2hfromLondon,theonlyavailabledata(Newcastle)showedless-than-average performance.

Fornon-HSRtownswithin1hofLondon,onlyCambridgeexceededthenational averagefiguresforbothhighandmediumtechnologyandknowledge-intensive services.Amongthenon-HSRtownswithin2hofLondon,Bournemouth, SouthamptonandNorwichprovedquiteknowledge-intensive,especiallyfor knowledge-intensiveservices;although,thisresultedfromlocalpoliciesaimedat promotingknowledge-intensiveactivities.

Atlanticlines AsaresultoftheTGV,employeeswhowerenotabletoworkinParisbecauseofthe highcostoflivingnowhavethepossibilitytodoso.Parisbenefitsfromtheenlarged (labour)marketresultingfromtheTGVwhileLeMans,acitylocatedalongtheHSR line,experienceddecreasesinemployment.

Nyfer(1999),

Rietveldetal.(2001)

Economicactivities:tourism

Paris–Lyon(1983) Underwenttwocontradictorychanges:therewerefewerovernightstays,butthere werealsonewtravelpackagesforusersoftheTGV.Wintertourismdidnotseemtobe affectedbytheTGV.

Bonnafous(1987)

AtlanticHSR(1989) InLeMans,acityconnectedtotheAtlanticHSR,severaltourismdevelopment projectsoccurred,particularlyinbusinesstourism.Conventionsandtradeshowsdrew innationalandinternationalclients,whereasbefore,theyhadonlyoperatedona regionalscale.Hotelstaysincreased,buttherewasadecreaseintheaveragenumber ofnightsstayed,contributingtothedeclineofthiscity’shotels,whichwerealreadyin difficulttimes.

InTours,acitylocatedintheLoirevalley,tourismactivitiesexperiencedalarge increasewhentheHSRwasimplemented.Theimpactisundoubtedlyvisiblewhen consideringthegrowthofvisitorsfromParis.Howevertheimpactoninternational touristattendancewasweaker,thismaybeduetotheinterestinvisitingthefamous— ChaˆteauxdelaLoire,whicharespreadintheregionandnotdirectlyservedbythe HSR.

MassonandPetiot(2009)

Paris–Marseille (2001)

Thesouthernregionbenefitedfromanincreaseinshort-staytravel(extended weekend-getaways)andfromanincreaseintravelfromspecificmarkets(i.e.young adults,seniors,high-incomeprofessionalsandinternationaltravellers).

MassonandPetiot(2009)

SanyoShinkans enline

VisitornumberstothethreeprefecturesalongtheHSR(Hiroshima,Yamaguchiand Fukuoka)increasedby7.9%,1.3%,and11.8%,respectively,whilefortheouter regions,nearlyhalfhadadecreaseinvisitors(Shimane,6.5%;Oha,0.4%; Miyazaki,2.8%).Moreover,somecitiesalongtheShinkansenlinesawabigdropin thenumberofvisitorsbecauseinter-citytrainsdonotstopthereanymore.One exampleisOnomichi.In1964,thecityhadasmanyas1,764,000visitors;this droppedto1,605,000in1975,theyeartheShinkansenopened.Overnightstaysdid notgoupproportionatelyforintermediatecities.

Okabe(1980)

Economicactivities:businessandknowledge-intensiveeconomy

Tokaidoline TheShinkansenhadamajorimpactonthedistributionofadministrative,financial andothercontrollingfunctions.Asageneraltrend,itcanbesaidthat,intheTokaido region,thecontrollinginfluenceofTokyoandOsakaisbecomingstronger,whilethat oftheNagoya,Kyoto,OsakaandKobeareas,sometimescalledthemiddlecapitals,is weakening.

Kamada(1980),

Sanuki(1980)

Paris–Lyon Businesstravelincreasedby56%,whileleisuretravelwasup112%. Pieda(1991)

EvidencedoesnotsuggestthattheParis-Lyonlineresultedinfirmsmovingfrom LyontoParis;instead,theoppositetrendwasobserved.BranchofficesofParisian firmsweresetupalongtheHSRline,particularlythoseinthehigh-techsectorinthe regionRhoˆne-Alpes.ItseemsthatthepresenceofaHSRstationplayedaroleinthe regionallocationchoiceoffirms,althoughtheresearchalsosuggeststhatthiswasnot decisivefactor.

Bonnafous(1987),

Sands(1993),

Rietveldetal.(2001)

ThelinecausedrelocationeffectswithintheregionsofcitiesthathaveaHSRrailway station.Lyon,forinstance,wasabletoattractmanyfirmsfromcompetingcitieslike GrenobleandGeneva.

Nyfer(1999),Rietveld,

Bruinsma,Delft,and

(14)

empirical evidence of cities’ performances withHSR and without HSR, which had been used as acontrol group,inordertostudytheeffectsofHSRincountries likeJapan,France,SpainandtheUK.Forexample,in 1991, Brotchie researched the social and economic impact of the Japanese Shinkansen by comparing populationgrowth,employmentandeconomicactivity incitiesthathadanHSRstationversusthosethatdidnot (Brotchie, 1991). The results showed that the cities connected with the Shinkansen performed better on average, especially for employment in information-basedindustries.InamorerecentexamplefromtheUK,

Chen and Hall (2011) investigated six London-based longdistancerailwaylines–allofthemover150km,of

whichtwo areHSRandfourare non-HSRinorderto compare performance of the local economy and knowledge-intensive development in cities served by the different lines.Theyclassified the effectsof HSR intothreeinfluentialzones,namelycitieswithin1hof London,within2h,orgreaterthan2h.Citiesconnected toHSRwithin1hof Londonsawastrongimpact on privateandknowledge-intensiveactivitiesandseemed to gain from the spillover effects of value-added activities from London. By contrast, in cities not connected with HSR, urban development remained focusedon localservice activitiesand asolid knowl-edgeeconomydidnotdevelop.Ofthecitieswithin2h travelling time from London and connected to HSR,

Table4 (Continued)

Paris–Lille Lilleattractedseveralinternationalaccountingfirmsandmajorofficesofsome nationalbanks.

ThelargeconferencecentrebuiltshortlyafterthearrivalofHSRisenjoyingstrong business;planstoenlargeitareunderway.

Uren˜a,Menerault,and

Garmendia(2009)

Madrid–Barcelona Zaragoza,amedium-sizedcitybetweenMadridandBarcelona,stagedthe2008 InternationalExpoandhasbeenselectedtostagetheinternationalFloraliashowin 2014.

HS1 TherehadbeengreatergrowthinthenumberofnewbusinessesinAshfordthan

elsewhere.Generally,HSRservicesappeartohavereinforcedthetrendinlocal economiestowardstheknowledgeeconomy.

Within1hofLondon,bothHSRandnon-HSRtownshavedemonstratedstrengthin privateandknowledge-intensiveactivities.However,innon-HSRtowns,theoffice developmentsmightbe,toalargeextent,attributedtoroutinepublicservicesor privateback-officesratherthanknowledge-intensivetypesofactivity.

Within2hofLondon,HSRtownshavereversedtheirpreviousdecline.Somecities weregiveneconomicadvantagesascriticaltransporthubsforHSRservicesfrom London.This,tosomedegree,alleviatedLondon’sexcessivedominationofcertain economicactivities;however,someofthecitiesintheNorthofEnglandcontinuedto bepath-dependentonheavyindustry.Non-HSRcitiesshowedbusinessprofilesthat werehighlypublicsector-oriented.

Beyondthe2hradius,HSReffectsappearweakerandalsocontrasting.

ChenandHall(2011)

InterCity125/225 AsaresultoftheHSR,MontabaurandLimburg,twosmallcitiesservedbytheline, experienceda2.7%upwardshiftineconomicactivitycomparedwiththerestofthe region,andthistrendwasnotreversedduringthesubsequentyears.Theservice helpedthetwocitiesattractnewresidents,whichincreasedthelocalemployment poolsandconsumermarkets,andeventuallyattractednewbusinessesthathelpedto drivethetowns’growth.

AhlfeldtandFeddersen

(2010)

Frankfurt–Cologne AsaresultoftheHSR,MontabaurandLimburg,twosmallcitiesservedbytheline, experienceda2.7%upwardshiftineconomicactivitycomparedwiththerestofthe region,andthistrendwasnotreversedduringthesubsequentyears.Theservice helpedthetwocitiesattractnewresidents,whichincreasedthelocalemployment poolsandconsumermarkets,andeventuallyattractednewbusinesseswhichhelpedto drivethetowns’growth.

AhlfeldtandFeddersen

(2010)

Economicactivities:wholesaleandretail SanyoShinkans

enline

Inretailandwholesaleactivity,theShinkansenseemsnottohaveaffectedoutput greatly.Althoughtherewereincreasesinretailandwholesaleactivityinonlytwo HSRcities(OgoriandTokuyama)therewasariseinretailsalesofmorethan10%in theyearfollowingthe1975openingoftheSanyoline.Inaddition,wholesale activitiesbecameincreasinglyconcentratedintheexistingcentres

Sands(1993)

(15)

someof reversedtheirpastdeclines; thesecitieswere giveneconomicadvantagesastransporthubswithHSR servicesfromLondon.Onthecontrary,citiesnotserved byHSRwerecharacterizedbyhighunemployment,low officerentsandhighlypublic-orientedservices.Beyond the 2-h ring, HSR effects appeared weaker and also contrasting(Chen&Hall,2011).

Nevertheless,thelinkbetweenHSRconnectionand retail and commercial activity is not so clear. For instance,eventhoughtherewasanincreaseinretailand commercialactivitiesincitieswithaShinkansenstation (Preston& Wall, 2008), onlytwo citieswith stations sawariseinretailsalesofmorethan10%intheyear followingthe1975openingoftheSanyoLine(Okabe, 1980; Sands, 1993). The research indicates that the face-to-face demands of knowledge intensive sectors such as businessservices, R&D,education and so on rosesignificantlyintheknowledgebasedeconomy;the HSR,whichisprimarilyapassengertrain,reducedthe cost of face-to-face meetings having a significant impact on these sectors; however few people would travelforretailandwholesalepurposes.

While thesestudies may identify clear patterns of correlation between HSR and development, they are muchlessconvincingwhenitcomestothedirectionof causation.DoesHSRdrivegrowthorisitthatcitiesthat are already growing(or citieswith greaterconditions forgrowth)attractHSR?

The model-based studies based in the fields of transporteconomicsandspatialeconomicsalsosupport thesefindings.Thesemodelsconsiderdifferentkindsof accessibility as akey conditionfor development, and HSRasafactorinchangingaccessibilitylevels.Atthe European level, the effects of HSR on accessibility patterns have been measured with different kinds of indicators. These include weighted average-distance indicators (Gutie´rrez, 1996) (Fig. 2), a gravity-type indicator (Bruinsma & Rietveld, 1993) and a daily accessibility indicator (Spiekermann & Wegener, 1996). The results have shown that as planned a EuropeanHSRnetwork willincrease theaccessibility imbalances between the main cities and their hinter-lands. In particular, the position of the cities in the north-westernpartofEuropewillbereinforced,while thecontinentalcountrysidewillbecomerelativelyless accessible. In Japan, Murayama (1994) studied the transformation of the urban system paying special attention to changes in the railway accessibility of cities.Thisresearchhighlightsthetransformationofthe spatialstructurebroughtaboutbytheemergenceofthe Shinkanseninthe1960s.TheShinkansencitiesgained the highestlocationadvantage, while non-Shinkansen

citiesinremoteregionsbecameevenmoreperipheral. Such studies have also been carried out on specific stretchesofHSRlines.Forexample,Gutie´rrez(2001)

studied the accessibility effectsof the HSR linefrom Madrid to France via Barcelona. The HSR was predictedtoincreaseaccessibilitydifferencesbecause, inSpain,thelargeragglomerationstendtobenefitmost from the infrastructureonanationalscale (Gutie´rrez, 2001).

Whilenon-HSRregionscould thusfacechallenges in their prospects for economic development and transformation, they will not necessarily lose out as longastheycanimprovetheirinter-regionaland intra-regionalaccessibility byothermeansof transport.For example,inFrance,thethreesub-regionsofCambresis, Sambre-Avesnois and Berck-Montreui were not con-nected tothe FrenchHSR network. Of these, Berck-Montreui, showed continuous employment growth abovethenationalaverageandbecameanincreasingly popular destination for migration; the other two performedfarbelowaverage.Akeyfactorbehindthis wasthedevelopmentoftheA16highway,significantly improvingtheconnectionoftheregionwithParis(Chen &Hall,2012).

3.2. Redistributionand relocationalongtheHSR corridor

Among the cities along HSR corridors, there has been a redistribution of social, economic and spatial elements. This has been shown by both empirical indicators such as population growth, employment levels, real estate market values, GVA (Gross Value Added),GDHI(GrossDisposableHouseholdIncome), and theoretical indicators such as different kinds of accessibility indicators (Gutie´rrez, 1996), a supply-driven model (Sasaki, Ohashi, & Ando, 1997), and Krugman’s core-periphery model (Masson & Petiot, 2009). Majorcities along HSR corridorshave gained development opportunities by being connected to the HSR. As well as improving the hinterland’s accessi-bilityforthosewholiveinthecity,theHSRalsomakes the major city more accessible for those who live outside the city. Reducedtravel timedue tothe HSR enables dominant cities to become more competitive and enlarge their area of influence. Consumers and purchasersofbusinessservicesfromotherregionscan reachthecitymoreeasilythanbeforetheopeningofthe HSR.Gutie´rrez(1996)predictedthatmajorcitieslike LondonandPariswouldbecomethechiefbeneficiaries after development of the European HSR network by measuringtheaccessibility change.

(16)

Both empirical and theoretical studies point to a more complex situation for intermediate cities. The adventofHSRhasfilledthehigh-speedconnectiongap incertaindistance rangesthatwascaused byaircraft, traditionallyapoint-to-pointlinkbetweenlargercities. HSR’spotentialtointegratecitieswithinregionsmay be underestimated (Blum, Gercek, & Viegas, 1992; Ross, 1994; Uren˜a, Menerault, & Garmendia, 2009). Indeed,someresearchhasarguedthatincountrieslike France,the impacton secondarycitiesisgreaterthan theimpactonlargercentralcities(Cervero&Bernick, 1996).HSRconnectivityhelpsovercomeintermediate cities’ traditional isolation, improving their location advantages.Thenewopportunityforcommutingtoor conducting business in the larger central cities helps

attractcertaintypesofmetropolitanbusinessactivityto the intermediate cities. These include meetings of metropolitanprofessionals,mid-levelbusiness, techni-calconsultancyfirmsandurbantourism(thestagingof congresses, scientific meetings, seminars and so on) (Uren˜aetal., 2009).

On the other hand, there is a risk that railway operators focus primarily on achieving high speeds, concentratingon profitable trafficbetween the largest urbancentres whilebypassingsmaller andless profit-ableplaces(Hall,2009).Thiscouldcause ‘‘backwash effects’’–aconnectiontotheHSRnetwork caneven resultineconomicactivities beingdrainedawayfrom intermediate citiesandanegativeoverall impact(van den Berg & Pol, 1998). Examples of intermediate

Fig.2. Absoluteandrelativechangesinaccessibilitytothemaineconomicactivitycentres1993–2010. Source:Gutie´rrez(1996)

(17)

‘‘middlecapitals’’thatareweakeningafterconnection toHSRincludeNagoya,alongtheTokyo–Osakalinein Japan(Kamada,1980),andLeCreusot,alongtheParis– LyonrouteinFrance.

Murakami and Cervero (2010) compared the population and labour markets within 5km of the ShinkansenstationsontheTokaidoline.Since1992,the operatingcompanyhasbeenrunningfastertrains that do not stop at five intermediate stations: Odawara, Shizuoka,Hamamatsu,Gifu-HashimaandMaibara.As destinations,theseintermediatestopshavebecomeless attractiveforbusinesspassengersandlessprofitablefor the privatized Shinkansen company. They have gen-erally lost both manufacturing and service-based activitieswithin5kmofthe HSRstations(Murakami & Cervero, 2010). Masson and Petiot (2009) used Krugman’score-peripherymodeltopredicttheimpact ofaproposedHSRlinkbetweenSpainandFranceon tourism activityin Perpignan, France.The prediction wasthatPerpignan’svitalityasatouristcentrewould facetheriskofdeclinebecauseofthethreatofthemajor cityBarcelona.

Inlightoftheambivalentimpactssketchedabove,a keyquestion seemsto bewhich kindof intermediate citiesandsecondarycities(destinationcities,butlower inthe urban hierarchy)can benefit from HSR.Three criteriahavebeenfoundtodetermineanHSRservice’s potentialimpactonintermediateandsecondarycities: citysize, network location and distance from central cities. All three criteria combine to determine the developmentpotentialofacity(Stanke,2009).

The first criterion, the size of a city and its metropolitan area, has been identified as a critical factorinhowHSRservice affectsthe developmentof thatcity.deJong (2007)positsarelationshipbetween the attractiveness of HSR stations (office rentals in station areas are used as indicators) and the factors which could influence the attractiveness of a HSR station.Thisresearch foundthatthe economicsize of the region, such as population and Gross Domestic Product (GDP),is one of the most important factors whichwillinfluencethecity’sattractiveness.Thismay explain why Ashford (population: 110,000), gained relativelylittlenewdevelopmentdespiteitbeingonthe Eurostarline(Preston, Wall,&Larbie, 2006) or why Lyon,whichhasasmallerpopulationandeconomythan otherurbanizedEuropeanregions,cannotcomparewith metropolisessuchasMilanorBarcelonainspiteofits excellentHSRconnections(Thompson,1995).

Thesecondcriterion,networklocation,isidentified asacriticalfactorforintermediatecities.Thislocation does not only rely on the HSR network. This is

important, as demonstrated by the example of Lille, locatedonthecrossroadsofthetriangleofParis,London andBrussels–anetworklocationwhichhasboostedthe city’s economy. However, other transport modes also contribute.ArecentDelphiSurveyamonginternational expertsstatesthat thelocationofanintermediateHSR cityatatransportationnodewithstrongconnectionsto otherregionalandinter-regionalnetworkswaslistedas the most importantprerequisitefor its future develop-ment (Loukaitou-sideris, Cuff, Higgins, & Linovski, 2012).ChenandHall(2011)studiedtheManchesterand Lille sub-regions and showed that the simultaneous developmentoftheintra-regionaltransportnetworkand aHSRhubstrategyarecrucialforspill-overeffectsfrom the arrival of HSR. The massively increased rail frequencies between Manchester and its sub-regions, including Central Lancashire, Merseyside, Greater Manchester North and Pennine Lancashire, reflected the wider attraction of a revived centre. However additional Manchester–London services reduced track capacity available for other operators and, to some degree,limitedotheroperators’ services.Theexpectation isthatthisunsolvedintra-regionalsystemwillnotsustain themomentumoflocaleconomicdevelopment.Onthe contrary in Lille, the regional government further attempted to improve its intra-regional network by connectingperipheralcoastalareastoLillethroughthe regional TGV service (TERGV) and progressively expandedtheservicerangeofTERGV.Thesedifferences arecluestoexplainingthefactthatmostsub-regionsin Manchester have experienced continued decline, whereas the Lille region, albeit slowly has been transformingpositively(Chen&Hall,2012).

Thethirdcriterion,thedistancefromcentralcities,is greatly impacted by the travel time between an intermediate or secondary city and the central city (usuallythecapitalcity).AccordingtoHarmon(2006), threetraveltimebandsexist:commutingmarket(1hor less), primarymarket (1.5–2.5h) and longer-distance market (over 2.5h). Firstly, within the commuting market, high income earners can take advantage of high-speed trains for daily commuting. Secondly, within the primary market, the time band enables travellerstomakeareturntripwithinonedaywithtime for short activities (business meetings, leisure visits). This is particularly important for business travel, increasingthedynamismofbusinessactivity,provided thatthelocaleconomiccontextisfavourable.Thirdly, within the longer-distance market, the high-speed services have expanded the potentialfor short breaks and weekend getaways. Social changes meanthat all groups are taking moreshort trips, but railtravel has

(18)

grown especially for middle-aged couples with a comfortableincome.Ineachtimeband,thecompetitive positionoftherespectivehousing,businessservicesand leisure facilities relative to the central city will determine thedevelopmentcourse.

3.3. Areconstructionoftheurban-regionalsystem When considering the impact on the regions connected or not connected to the HSR and the redistributionofactivitiesalongtheHSRcorridor,itis important toexaminetheregionalstructure. TheHSR plays adifferent role as a commuting mode andas a long-distance transport mode. We will discuss them separatelyinthenextsections.

3.3.1. Commuter HSRand anewmetropolitan process

Atfirst,HSRwasthoughtofasanalternativetoair transport, connecting large cities 400–600km apart. Yet, with many intermediate stations being brought within afeasiblecommuting timeofthe largercities, HSR can also become part of the commuter infra-structure. In the process of a metropolitan sprawl, householdsandeconomicactivitieshavethepotentialto relocateandmoveoutwards.Meanwhile,the introduc-tionofHSRcanimprovetheconnectionbetweensmall citiesintheperipheryandthecitycentre.Thistypeof HSRconnectionfacilitatesorreinforcesthe attractive-nessofthesesmallcitiesorsuburbanareasandbonds them moretightly with the metropolitancore.People can commute to the metropolis from faraway places (Uren˜aetal.,2010).Intheprocess,somemetropolitan activities may relocate and take on a suburban role compared to the traditional role, which is more polarizedtowardsservingitssurrounding region.This leadstoanewmetropolitanprocess.

TypicalexamplesofthisareCiudadReal(Madrid) and Ashford (London) (Garmendia, de Uren˜a, Riba-laygua, Leal, and Coronado, 2008; Preston & Wall, 2008). Ciudad Real (65,703 inhabitants in the year 2003)isasmallcitylocatedontheHSRlineof Madrid-Seville. Before the HSR, the accessibility of Ciudad Realwasverypoorandcommutingandparticipationin themetropolitanprocessesofMadridwasalmost non-existent. Itexperienced littlepopulation growthinthe past, and only 10.9% of the regional population was concentrated there in 1981. After the opening of the AVEin1992,theaccessibilityofCiudadRealimproved immensely.JourneytimestothecentreofMadridwere reduced tojust51minutes.Thisattractedpeoplewho werenotoriginallyfromthecityortheprovincetolive

inCiudad Real,travelling dailyor weekly toMadrid. Onesurveyfoundthat39% and33% ofthe dailyand weeklycommuters,respectively,werebornoutsidethe province of Ciudad Real (Garmendia et al., 2008). Meanwhile, the proportion of the region’s population livinginCiudadRealhadincreasedto13.8%by2004. ThisevidenceshowsthattheHSRcanhelpisolated cities to attract housing investment from within the province(30% inthe caseofCiudad Real)as wellas fromoutside the province(9%).The HSRhas helped the city increase its spatial polarizing capacity. The samestoryhappenedinAshford,88kmfromLondon. After the opening of Ashford International in 1996, journey times to London were reduced from 70 to 37min.This led toan estimated85% increase inthe accessibility of continental destinations. Compared with the Southeast as a whole, the population went up by 11% and employment by 6%. Since 1996, propertypricesinAshfordhaverisen26.5%,compared with23.2%intherestoftheregion.Vacancyratesfor propertiesinAshforddroppedfrom13%in1998/99to 8%in2004/05.Bycontrast,vacancyratesinSoutheast England as a whole, increased from 7% to 9%. The trendisthatAshfordisbecomingapotentialcommuting centreofLondon(Preston&Wall,2008).

EvidenceconfirmsthattheconstructionofHSRcan havetheeffectofincreasingpopulationincitiesserved bythehigh-speedcorridors.Italsoshows,however,that HSRcancontributetotheurbansprawlprocess.Atleast in some cases, the HSR can also facilitate the development of these small cities or suburban areas asspecializedsub-centresofthemetropolitanareawith high-level office developments so that people do not need tocommute tothe metropolis.For instance,the evidenceof theimpactof theUK’sInterCity125/225 rail corridor suggeststhat the cities connected to the HSRinacommuting distance(within 1h ofLondon) show spill-over effects of high-value-added activities fromLondon(Chen&Hall,2011).

3.3.2. LongdistanceHSRand anintegrated functionalregion

MostHSR isbuiltfor inter-metropolitan transport. The HSR connects two or more metropolitan areas, linkedtogetherviaachainofcitiesbymeansofa high-speed train. This may have the effect of creating an extended functional region – an integrated corridor economy.Blum,Haynes,andKarlsson(1997)examine economicintegrationalongacorridorintheshortand mediumterm(based onempiricalevidence) andlong term(based on model studies). In the short term, the HSR will not only integrate the goods and services

(19)

market,butalsolabour,commercial,inter-personaland leisureactivities.Inthemediumterm,aHSRwilllead tosome relocation of households and firms along its corridor. This includes relocation from the capital to medium/small-sized cities and from medium/small-sizedcities to the capital, as hashappened on routes suchas Paris–Lyon.ForfurtherevidenceseeTable4. Lookingatthelong-termperspectivewiththeaidof dynamic model, Blum et al. (1997), conclude that withincertaintransportcorridors,HSRcouldleadtoa completely new location pattern with larger travel volumesandamajorchangeintravelpatterns.Afterthe establishment of a new HSR connection, one could expectanon-linearreaction.Itsfullforcewillnotbefelt until sometime after the new line has opened up. Relatedtothis,Haynes(1997)usedspatialinteraction modelsandempiricalevidencetoexplainthatHSRcan facilitatetheeffectivenessofregionallabourmarkets.It canreduceinformationdecayoverdistance,orthecost of market adjustment,, thus making labour migration easier.

4. Indirect effects:urbanlevel

4.1. Acatalystfortherestructuringofurban systems

At an urban level, HSR is a catalyst for the restructuring and reshaping of urban systems. This section will discuss the extensive research whichhas been carried out on this spatial and economic restructuring.

4.1.1. Acatalystfor therestructuringoftheurban spatialstructure

According to the particular locations of a HSR stationsinacity, Hall(2009)identifiedthreekinds of urbanimpactsofHSRdependingonthelocationofthe station.Thefirstiswhenthestationislocatedbesideor withinthetraditionalCentralBusinessDistrict(CBD). Thisimproves or reinforcesthe CBD’sattractionasa placeforcommercialinvestment.Thisistrueinstations such as Lille Europe, Brussels South, Rotterdam Central, King’s Cross-St Pancras in London and Charmartin (St.Martin) inMadrid. A secondtype of stationisusuallylocatedontheedgeofcities,adjacent to,butseparatedfrom,themajorurbancentres.Thiscan helpdevelopcomplementarysub-centres.Forexample, the new station at Kassel-Wilhelmsho¨he in Germany has created a sub-centre, but has not displaced the original mediaeval location of the city centre. Other typicalexamplesare NewOsaka,LyonPartDieuand Stratford in London. The third type of station is differentinconcept,butequallyambitious.Thisisthe useofanewstationasthebasisforanewcommercial ‘‘edge city’’ on the urban periphery such as Shin-Yokohama in the Kanagawa prefecture in Japan, Ebbsfleet railway station (a park-and-ride station) outside London, the new HSR station in Avignon, FranceandAmsterdamSouth(Hall,2009).Allofthem are showninTable5.

Thelattertwotypesofstationlocationscanalterthe centreofgravityofacity’scoreandspurredevelopment of underutilized areas in the urban periphery ( Todor-ovich etal.,2011).Meanwhile,theyalsopromote the

Table5

ModelsandexamplesofthreekindsofHSRstationlocation(basedonHall,2009). Typeone:locatedbeside

orwithinthetraditionalCBD

Typetwo:locatedontheedgeofcities Typethree:locatedinEdgeCity

Model

Examples LilleEurope,BrusselsSouth; RotterdamCentral;LondonKing’s Cross-StPancras;Madrid Charmartin(St.Martin)

LondonStratford, Kassel-Wilhelmsho¨he newstation;New Osakastation; LyonPartDieustation

Shin-Yokohamastation; LondonEbbsfleetrailway station;AmsterdamZuidas, AvignonnewHSRstation