Policies for synchronization in the transport–land-use system

Bert van Wee a,n, Wendy Bohte a, Eric Molin a, Theo Arentze b, Feixiong Liao b

a Delft University of Technology, Transport and logistics group, PO Box 5015, 2600 Delft, The Netherlandsb Eindhoven University of Technology, Urban Planning Group, Faculty of the Built Environment, The Netherlands

a r t i c l e i n f o

Available online 22 November 2013

Keywords:SynchronizationLand useInfrastructureEvaluationModelCBA

a b s t r a c t

This paper presents an overview of options for synchronization in the transport and land-use system.We distinguish between synchronizing (a) transport networks, (b) activity locations, (c) transportnetworks and activity locations, and (d) ICT-based decoupling of activities from time and/or locations.Synchronizations in both time and space apply to these four forms of synchronizing, resulting in eightsynchronization options. These eight synchronization options were then linked to different categories ofpolicy options: (a) regulation, (b) pricing, (c) land use planning (d) infrastructure planning, (e) specificpublic transport policies, (f) marketing and communication, and (g) time related policies. We explain therelevance of these policy options for synchronization. Next we apply our structured overview to a casestudy, the redevelopment of the Rotterdam Soccer stadion of Feyenoord. Finally we discuss the relevanceof Cost-Benefit Analysis and Multi-Criteria-Analysis for the evaluation of policy options, concluding thatCBA is the preferred method of evaluation in most but not all cases.

& 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The transport system allows people to carry out activities indifferent places. Together with the land use system it is of paramountimportance for accessibility. Following Geurs and Van Wee (2004) asa point of departure we define accessibility in the case of passengertransport as ‘the extent to which land-use and transport systemsenable (groups of) individuals to reach activities or destinations bymeans of a (combination of) transport mode(s)’.

Increasingly researchers and policy makers are aware of thenotion that accessibility can be further improved by synchronizingdistinguished components of the transport system, as well as thetransport and land use system. Limiting ourselves to academicliterature, recent examples of papers include Guihaire and Hao(2008) and Guo and Wilson (2011) on synchronization of publictransport networks, Levine (2005) and Bertolini et al. (2012) onthe integration of land use and transport planning, Mokhtarian(2002) on telecommuting and travel, Farag et al. (2007) one-shopping, and Lyons and Urry (2005) on conducting activitieswhile traveling.

Synchronization between transport networks and/or activitylocations can have important accessibility benefits for individuals.Because the trips and activities that people undertake during a dayare interdependent, the synchronization of activity locations andtransport networks often affects more than just one trip or activity.

Trips may be chained or several activities may be carried out atone location or whilst traveling. Moreover, other travel modes maybecome attractive. For example if childcare becomes available neara railway station, people may choose to cycle to the childcare andtravel to their work by train, instead of using the car for bothactivities. These benefits of synchronization are especially impor-tant as far as travel is a derived demand. Note that this does notapply to all travel; sometimes people travel for the fun of it(Mokhtarian and Salomon, 2001).

To optimally synchronize the transport and land use systemdedicated transport and spatial policies are of great importance.Examples of such policies are decisions on roads and the connec-tions between roads of different types. In addition, public trans-port is subsidized in many countries, and subsidies often dependon the level of service characteristics and/or travel demand, whichboth depend on, for example, connectivity between modes. Inaddition to policies focusing on the transport networks, severalland-use policies can increase synchronization. For example, localmunicipalities play a key role in the activities that can be locatednear railway stations, resulting in not only the accessibility ofthese activities by rail being influenced but also the trip-chainingoptions for rail passengers. Furthermore, time-related policies caninduce the synchronization of transport networks and activitylocations. Examples are policies which determine the openinghours of shops and services and flexible working times. Policiesmay also simultaneously influence temporal and spatial accessi-bility. In particular ICT related policies, such as the stimulation ofteleworking and teleshopping, influence both dimensions ofaccessibility. Several of the above-mentioned policy options have

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/tranpol

Transport Policy

0967-070X/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.tranpol.2013.10.003

n Corresponding author. Tel.: þ31 15 2781144.E-mail address: g.p.vanwee@tudelft.nl (B. van Wee).

Transport Policy 31 (2014) 1–9

an impact on accessibility, but in this paper we focus on synchro-nizing and related policies.

To the best of our knowledge there is no systematic overview inthe literature of the policy options for synchronization. This paperaims to fill this gap. We do this by developing a conceptualizationof synchronization options (Section 2), and linking this conceptuali-zation to distinguished categories of options for public policies(Section 3). Section 4, as an example of current synchronizationpolicies, presents a case study from Rotterdam, The Netherlands.Section 5 discusses how policy options for synchronizing could beevaluated. Finally Section 6 summarizes the main conclusions anddiscusses the implications of our paper.

2. Conceptualization of synchronization in time and space

In this paper, synchronization options refer to spatial and/ortemporal synchronizations between locations of activities and/ortransportation networks. Synchronizations can increase the accessi-bility of individuals by decreasing travel distances and the travel timebetween successive activity locations, or by enabling the performanceof multiple activities at the same location or even at the same time.

This section first discusses the distribution of activity locationsand transport networks over time and space. Next this positioningis related to synchronization options by distinguishing eightspatial and temporal dimensions that can be included in policies.

2.1. Activity locations and transport networks in time and space

The ease with which individuals can perform their daily activitypatterns strongly depends on the distribution of activities andtransport networks over time and space. Firstly, a ‘favorable’spatial and temporal distribution of activity locations and trans-port networks leads to improved accessibility. Travel distances,travel times, travel costs, and effort to reach activity locationsdepend on the locations and availability in time of activitylocations and transport networks. The spatial and temporal con-nections between activity locations, between transport networks,and between activity locations and transport networks have greatinfluence on accessibility, particularly for performing multipleactivities. The synchronization of activity locations and transportnetworks in space and time therefore increases the accessibility ofindividuals.

Secondly, if activities can be allocated more flexibly in time andspace, people can arrange their own activity schedule accordingto their particular preferences and minimize travel disutilities.The spatial and temporal flexibility of activities depends on the levelto which other people, goods, facilities and space are needed to carryout their activities and how unique these requirements are.For example a social meeting with a friend requires a specific person,whereas for buying groceries several supermarkets may providereasonable alternatives. Also some activities have very strict timeframes (e.g. taking an exam) while others can be performed withina larger time slot (e.g. shopping or fitness). Currently the devel-opment of ICT increasingly weakens the need for synchronizationof activities in time and place, because activities become moreand more flexible within or even independent of space and time.For example, talking on the phone is more or less independent ofplace, but depends on time while email is also largely independentof time. When activities can be performed anywhere and anytimethis may increase accessibility and decrease the demand for travel.People can shop on the Internet from their office, or work in thetrain and video conferencing may decrease the need to travel forbusiness meetings. The increased flexibility of activities alsoincreases options to combine activities and travel, such as workingin the train.

2.2. Synchronization options

Most activities are still framed in time and space, albeit to avarying degree. Therefore synchronizing times and locations ofactivities and transport networks would enable individuals toexecute their daily activity programs more easily. Policies whichaim to make activities independent of time and or space alsoenable people to transfer more easily from one activity to another.For activities which can be performed anywhere and anytime thedemand for adapted spaces such as shops and offices decreases(Couclelis, 2009).

Table 1 categorizes the synchronization options. Such a cate-gorization enables researchers to study synchronization in astructured way, and policy makers to systematically design syn-chronization policies. It shows that synchronizations are possiblealong the dimensions space and time. The components that can besynchronized include (a) transport networks, (b) activity locations,(c) activity locations and parts of transport networks, and finally(d) that activities can be decoupled from space and/or time.Table 1 links these dimensions and components.

Below we give some examples of options for synchronizationcategories as presented in Table 1:

1. Spatial synchronization between transport networks, e.g.: parkand ride facilities.

2. Temporal synchronization between transport networks, e.g.:synchronizing the timetables of either different public trans-port modes or within one mode (e.g. different lines of a railwaynetwork).

3. Spatial synchronization of activity locations, e.g.: mixing usecategories such as houses, shops, health services and schools.

4. Temporal synchronization between activity locations, e.g.: theopening of health care services after regular working hours atoffices.

5. Spatial synchronization between transport networks and activ-ity locations, e.g.: building offices or high density residentialareas near railway stations.

6. Temporal synchronization of (parts of) transport networks andactivity locations, e.g.: matching the hours when bus servicesare provided and the temporal component of activities such asconcerts.

7. Decoupling activities from place, e.g.: offering services throughtelecommunication, such as allowing people to telework, offer-ing the possibility to teleshop and having a helpdesk by phone.

8. Decoupling activities from time, e.g.: using email, flexibility ofoffice hours, teleshopping 24/7.

In Section 3 we present policy options for synchronizationfollowing this categorization.

3. An overview of policy options for synchronization

Section 2 discussed options for synchronization (what ispossible?). This section links policy options to these options for

Table 1A categorization of synchronization options.

Dimensions\components

Betweentransportnetworks

Betweenactivitylocations

Between transportnetworks and activitylocations

Decouplingof activities

Space 1 3 5 7Time 2 4 6 8

B. van Wee et al. / Transport Policy 31 (2014) 1–92

synchronization (how can policy influence synchronizationoptions?). Policy options can be categorized distinguishing restric-tions/regulations, pricing policies, land-use planning, infrastruc-ture provision, specific public transport policies, marketing andinformation, and time related policies, such as with respect toopening hours. Table 2 links these policy options to the synchro-nization options as presented in Section 2.

Below we discuss Table 2. It is beyond the scope of this paper topresent all the options for different policy instruments to stimulatesynchronization. We provide some key examples only. Regulations(a) can improve synchronization via all synchronization options.Regulations can force road authorities to improve connectivity ofparts of the road network, and can force synchronization betweendifferent public transport networks (synchronization option 1).Regulations can also be used to synchronize public transporttimetables (2). They can be used for land-use policies, examplesbeing policy plans expressing which activities are allowed atwhich locations, or more specifically regulations with respect tolarge-scale retail facility locations (3). Regulations can determinethe opening hours of potentially related activities, such as child-care facilities and working hours (4). The Flemish government setregulations for the maximum distance between a public transportstop and dwellings (5). Authorities can use regulations to forceorganizers of concerts to schedule the timing of concerts to fitpublic transport timetables, and vice versa (6). Regulations canprovide employees with the right to telework (7) or can providerules for 24 h virtual access to some websites (8).

Pricing (financial incentives) (b) can include both levies andtaxes, as well as subsidies. They can be used for just about allsynchronization options. For example, demand dependent sub-sidies for public transport companies stimulate synchronizationbecause synchronization decreases travel times and so increasestravel demand, and consequently subsidies for public transportcompanies. This first applies to networks (1) and timetables (2).It also applies to the provision of access to high density locations ingeneral (5) or at times when demand is high, such as after aconcert or soccer match (6). Financial incentives can be used topersuade shop owners and service providers to locate at specificlocations (3) and synchronize opening hours (4). Fiscal incentives,such as taxing employer compensations for commuting, canencourage less commuting and more teleworking (7). Peak hourroad pricing may encourage people to avoid rush-hour travelingwhen work can be done by teleworking (8).

Land-use planning (c) can be used to cluster activities spatially(mixing use categories) in high densities, reducing barriers forspatial (3) and temporal (4) synchronization. In addition, land useand transport can be synchronized by building such mixed use andhigh densities areas, or office locations or residential areas near arailway station and other public transport nodal points (5).

Infrastructure planning (d) can be used to synchronize distin-guished components of the transport networks. For example,metro or tram lines can be well connected to railway stations,and park and ride facilities enable car and public transport travelto be easily combined in one trip (1). In addition, planninginfrastructure in a way that areas with many origins and destina-tions are well served leads to synchronizing the transport net-works and land use (5).

Specific public transport policies (e) allow for several types ofsynchronization. Linking bus and tram timetables to those of trainservices leads to synchronization within and between transportnetworks (2), providing bus services to designated areas leads tosynchronizing transport and land use (5), providing bus serviceswhen concerts or soccer games are finished leads to synchroniza-tion of transport and activity times (6).

Marketing and information (f) can be used for just about allsynchronization options. In the area of marketing the positiveTa

ble

2Dom

inan

trelation

shipsbe

twee

npolicycatego

ries

andsynch

ronizationop

tion

s.

Tran

sport/sp

ace(1)

Tran

sport/time(2)

Activitylocation

s/sp

ace(3)

Activitylocation

s/time(4)

Tran

sport-

location

s/sp

ace(5)

Tran

sport-

location

s/time(6)

Decou

plin

gof

place

(7)

Decou

plin

gof

time(8)

Reg

ulation

(a)

nn

nn

nn

nn

Pricing(b)

nn

nn

nn

Land-use

planning(c)

nn

n

Infrastructure

provision

(d)

nn

Public

tran

sport

policies(e)

nn

n

Marke

tingan

dinform

ation(f)

nn

nn

nn

nn

Timepolicies(g)

nn

B. van Wee et al. / Transport Policy 31 (2014) 1–9 3

effects of synchronization can be advertised, for example advertis-ing P&R facilities or bus services for specific activities (such asconcerts or soccer games). Information services, web-based andothers, can provide the traveler with information about synchro-nization options, examples being connections between trains andlocal public transport, walking routes between stations andactivity locations, and maps of areas with mixed use.

Time-related policies (g) can be implemented to increasesynchronization options, for example by synchronizing the open-ing hours of shops, childcare facilities, car parks, and workinghours (4), or by setting the right conditions for teleworking,teleshopping, and providing 24/7 services (8).

Some synchronization practices are quite common, for examplelinking components of public transport networks is practicedfrequently, as is the planning of high concentrations of activitiesnear public transport nodes (such as railway stations). On theother hand, the full potential of synchronization has often notbeen utilized due to several barriers. Barriers can be related to theorganization of public bodies: more or less independent bodiescan be responsible for distinguished parts of the synchronizationpolicies. In addition policies often do not result from ‘granddesigns’ but evolve over time and sequentially, as a result of whichat least temporal frictions may occur, and often ‘system optimalsolutions’ are difficult to obtain. Cultural barriers between dis-ciplines (e.g. between transport engineers and planners) may alsoexist, leading to the non-optimal collaboration of public bodies.Furthermore, for some categories of synchronization public andprivate bodies need to collaborate, as in the case of urbandevelopment. Goals of public and private parties may differ, whichcan result in – from the perspective of synchronization – sub-optimal results. Note that if each party optimizes its own goals,this does not necessarily lead to the system optimum.

4. Synchronizing policies in a structural design plan: the caseof Rotterdam Stadionpark

4.1. Introduction

To illustrate a selection of the policy options of Section 3, thissection describes the synchronization policies that have beenincluded in the structural design plan of Rotterdam Stadionpark(see Fig. 1 for a map of the current situation). In The Netherlandsstructural design plans are not binding but developed by munici-palities to give direction to spatial planning. The Stadionparkdevelopment project aims to contribute to the comprehensiveregional development of South Rotterdam. Currently it is first andforemost the host of Rotterdam's largest soccer club Feyenoord.The ambition is that by 2030 the Stadionpark area will have beenre-developed to be the figurehead of Rotterdam, City of Sports,as well as an important hub for South Rotterdam. The plan todevelop an area for all levels of sport, for beginners as well asinternational top athletes, with sport, education and leisurerelated services, including a new Feyenoord stadium. It will haveexcellent public transportation connections, with an intercity trainstation and new tram connections to the northern part ofRotterdam across the Maas River. Section 4.2 discusses optionsfor synchronization for the Rotterdam case, Section 4.3 illustratesthe added value of synchronization of some specific transport andland use options via a scenario based simulation study.

4.2. Synchronization options

The structural design plan includes several of the synchronizationoptions and policies that are categorized in Table 2. Stadionpark is

Fig. 1. Rotterdam Stadionpark area, current situation.Source: http://www.rotterdam.nl.

B. van Wee et al. / Transport Policy 31 (2014) 1–94

positioned as a sport-related facility area with a high quality publictransport node. The plans also include large PþR developments,making sure the road network is well connected to the publictransport network. Following Table 2 these combinations ofsynchronization options and policies can be derived from theplans:

4.2.1. Spatial synchronization between transport networksInfrastructure (1d): Two major roads will be connected.Infrastructure (1d): The new tram tangent and the upgrading of

the railway station will significantly improve the connection of thelocal public transport of Rotterdam-South to the regional andnational public transport networks.

Infrastructure (1d): By providing good public transport connectionsfrom PþR locations in other areas of Rotterdam to the Stadionpark,people will be encouraged to leave their car in the suburbs ofRotterdam rather than causing congestion in the Stadionpark area.

4.2.2. Spatial synchronization between activity locationsLand-use planning (3c): By combining education and sport

facilities in the same area it will be easier to take part in sportingactivities during or after school.

4.2.3. Spatial synchronization between transport networks andactivity locations

Pricing (5b): Parking prices aim to reallocate car parking awayfrom the Stadionpark area. PþR locations in other areas ofRotterdam will be free if people continue their trip with publictransport.

Land use planning (5c): Restaurants, sport and beauty facilities,and childcare will be clustered in Stadionpark, so different activ-ities can be combined.

Land use planning and infrastructure (5c/d): Because educationis becoming increasingly specialized students have to travelfurther to reach their school and the level of access of schools isbecoming more important. In Stadionpark several schools arelocated near the railway station, including sport-related educationfacilities.

Infrastructure (5d): A large number of new PþR places will bedeveloped in Stadionpark.

Infrastructure (5d): Slow mode infrastructure will connectseveral green areas.

Infrastructure (5d): A new tram tangent will connect theStadionpark area to other important areas of Rotterdam, includinga large shopping mall and the university campus.

Infrastructure (5d): The intercity status of the railway stationenables inhabitants of Rotterdam-south to reach employmentopportunities in other parts of the Randstad more easily.

4.2.4. Temporal synchronization between transport networksTime policies (2g): Trams and trains will ride with high frequen-

cies, therefore transfer-times between PT networks will be short.

4.2.5. Temporal synchronization between transport networks andactivity locations

Land-use planning and time policies (6c/g): By developingfacilities that keep people in the area before and after soccergames, the use of the car and PT infrastructure will be spread.There are already a large cinema and some restaurants in the area.The development will further encourage the new trend of com-bined retail formulas like stores with restaurants.

Public transport policies and time policies (6e/g): The DutchRailway company will increase the train frequencies.

4.2.6. Decoupling of activities from time and placeRegulation, land-use planning and time policies (7a/cþ8a/c/g):

A final aim is to accommodate people who want to be involved insport wherever and whenever they want, including having enoughfacilities for individual sports, extending opening hours andproviding flexible membership of sport facilities.

The above listing of synchronizing policies in the structuraldesign plan shows the importance that the municipality ofRotterdam attaches to the use of synchronization to reach itsaims. It is to be expected that with the furthermore detaileddevelopment of the plans for the Stadionpark area new policies inthe area of regulation, pricing and marketing will also includesynchronization dimensions. Examples are the promotion of theuse of PþR to reach the soccer stadium, variation of parking pricesduring the day, regulations concerning the type of facilities thatcan be developed and conditions for the Dutch railway companyconcerning train frequencies at the new or upgraded railwaystation.

4.3. Simulations of transport and land use scenarios

Next we did a simulation study related to the Rotterdam caseaiming to show the importance of a selection of synchronizationoptions, in particular some options for transport and land usepolicies. More specifically the study aims to answer if positivesynergy effects between policies in the areas of land use and thetransport system can be expected. The simulation is based on theapplication described in Liao et al. (2013b). For a detailed descrip-tion of the study area and model readers are referred to thisprevious study. In this section we briefly describe the databasesand model used. Next we discuss the specific scenarios consideredin the present study and the results of the simulations.

4.3.1. The modelThe simulation study is based on the supernetwork model

developed in Liao et al. (2013a, 2013b). The model can be used as atool to analyze the opportunities a land-use–transportation sys-tem offers for implementing daily activity programs of individuals.The supernetwork model originally proposed in Arentze andTimmermans (2004) is an extended model where not onlytransfers between networks of different modalities but alsoimplementations of activities (at particular locations) are repre-sented as links. A path through the network represents a particularway of implementing an activity program as a result of a full rangeof choices a traveler can make. The choice dimensions include thesequence of activities, locations of flexible activities, transportmodes, possible transfers during trips, parking locations androutes of trips. Extended link costs functions are used to representthe values individuals assign to a wide range of relevant attributesof trips and locations including travel time (waiting and in-vehicle), travel costs (parking, ticket, and fuel), location-basedfacilities for activities (e.g., floor space), transport modes (basepreferences) and transfers (inconveniences). The values incorpo-rated in link costs functions in this application have been esti-mated based on a specifically designed series of stated choiceexperiments that involved a large and representative nation-widesample of individuals (Arentze and Molin, 2013). As a conse-quence, the costs of a path effectively represent an estimate of the(dis)utility an individual would experience when the activityprogram is implemented in that way. Using micro-simulation,least costs paths are computed for a representative sampleof individuals. Since least costs paths represent preferred waysof implementing the activity programs, the associated costs offera measure of accessibility. Synchronization strategies can be

B. van Wee et al. / Transport Policy 31 (2014) 1–9 5

analyzed due to the integrated and high-detail representation ofthe land-use–transportation system the model offers.

4.3.2. Study area delineation and data4.3.2.1. Study area and base year. Synchronization policies canhave effects at the regional (or even national) level. We selectedthe Den Haag–Rotterdam–Dordrecht corridor (LHS of Fig. 2) as thestudy area, which takes up the majority share of population andfacilities of Zuid-Holland (The Netherlands). We expect that effectsoutside this area are limited (but not absent). 2010 is the base year.

4.3.2.2. Population and activity program. Data of activity programs,which are input to the model, were provided by MON (Dutch dailymobility travel dairy survey). For the years 2006 and 2007, data of9184 individuals of at least 12 years old living in the corridor wereavailable. To be able to estimate the behavior of all people living inthe area we weighed the data based on data of the full population.The average numbers of activities and trips per person are 1.5 and2.5 respectively. An activity is classified as fixed or flexibledepending on whether the activity can be conducted only at onefixed location or at one of multiple locations. Fixed activitiesinclude work, education, chauffeur, and use of municipality serviceand flexible activities concern shopping and four categories ofleisure activities (going-out, culture, sports, and recreational-tour).For flexible activities, the location choice sets are defined based onan empirically estimated location choice model.

4.3.2.3. Location, transport and price tariff data. The followingdatasets are used to describe the integrated land-use andtransportation system. For mapping the land use system, floorspace data linked to the different activity categories distinguishedin this study at the level of 4-digit postal code areas are used. Fordescribing the public transport networks detailed data of the routesand timetables of all modes of public transport (PT) including bus,tram, metro and train, are used. The fares for bus/tram, stop trainand intercity train are set to estimated values (0.12 €/min for bus/tram, 0.14 €/min for stop train and 0.15 €/min for intercity train)

which approximate the actual tariff system in the Dutch context.As for private transport networks, a database of the Dutch roadnetwork is used that includes all road categories from nationalhighways to local neighborhood roads. Since this database doesnot include car and bicycle speed data, speed values used in themodel were based on assessments of averages. Assessments of carspeeds are differentiated by time of day (peak hours and non-peakhours) and type of road (urban local, regional and national). On theother hand, assessments of bicycle speed are differentiated by roadtype only. Car fuel costs are also based on assessments of averagevalues differentiated by road type (with higher costs per km forlocal roads in urban area, etc.). With regard to parking locations it isassumed that parking is possible at activity locations, PþRs (park-and-ride facilities) and THs (transport hubs, train stations). Parkingprice is set according to the parking tariff zoning system inRotterdam. Actual parking costs for an activity are calculated as alinear function of parking duration (activity duration). On theother hand, it is assumed that bicycles can be parked everywherefor free. Our network includes the nine PþR facilities in the studyarea. Furthermore, train stations generally offer parking facilitiesand, hence, are identified as additional PþR facilities. For adetailed description of the data and assumptions, readers arereferred to Liao et al. (2013b).

4.3.3. ScenariosThis section describes the scenarios. We define the base Scenario

(labeled as Base) according to the settings as described above. Threepolicy scenarios were defined:

Scenario 1: Transportation changes (labeled as Trans)Increase of train connection frequencyThis increase is part of a program aiming at concentrating

spatial developments around the railway stations and improvingtransfer options in the south-wing of the Randstad.

Upgrade Rotterdam Stadion stationThe Rotterdam Stadion station, originally only used for special

events on weekends, is upgraded to an intercity station for general

Fig. 2. Study area: Den Haag–Rotterdam–Dordrecht corridor.

B. van Wee et al. / Transport Policy 31 (2014) 1–96

usage. At the same time, Rotterdam Blaak station (2.5 km north) isdowngraded from an intercity to a stop train station.

Introduction of a new tram lineLinked to the upgrade of Rotterdam Stadion station, a new

east–west tram line (approximately 14 km) is introduced having astop at this station.

Scenario 2: Land-use changes (labeled as Land)Spatial developments – transport nodes-orientedThis scenario aims to increase the attractiveness of the areas

near transport nodes in the central area of Rotterdam (see map onthe right-hand-side of Fig. 2). In areas around public transportnodes, we assume a 25% increase in floor space in each segmentrelated to shopping and the four leisure activities.

Increase of parking cost at activity locationsTo reduce car traffic in the city center of Rotterdam, the parking

price in each parking tariff zone is doubled compared to the basescenario. The increase in parking cost is supposed to stimulate theuse of PþR and bike-and-ride (BþR) facilities.

Scenario 3: Combine Scenarios 1 and 2 (labeled as Comb)To demonstrate the added-effects of combining transportation

and land-use changes, Scenarios 1 and 2 are combined as Scenario 3.

4.4. Results

Figs. 3 and 4 show the results for the whole corridor area. First,Fig. 3 shows the distribution of trips across the main transportmodes, walking, cycling, public transport (PT) and car. Car pas-senger, PþR and BþR modes are not represented in this figure.Next, Fig. 4 shows the results for PþR and BþR.

Fig. 3 shows, as expected, that improvements in PT (Scenario 1)result in an increase in the use of PT. This increase is mainly due toa switch from walking to traveling by PT. At first sight this seems abit strange, but the results can be explained by the fact that theshare of walking trips drops because people substitute activitylocations at nearby destinations to more attractive activity locations at

a larger distance that become accessible by PT. Mode changes becomemore pronounced under the Land scenario because high parking costrepresses car use, and increasing attractiveness at transport hubsstimulates PT use. The differences between the combined scenario andthe land use scenario are small.

Fig. 4 displays the number of PþR and BþR users. Trans andLand both promote multi-modal trips, reducing the number ofcars, especially those going into city centers, as intended by policy-makers. It shows that synergy effects between the land use andtransport policies occur: the combined effect (Comb) on BþR (39additional trips) exceeds the sum changes due to the transport(12) and the land use (15) policies. For PþR also positive synergyeffects occur, but these are smaller than those for BþR.

We now zoom in on the Rotterdam Stadion station area. Fig. 5shows for the different scenarios the use of Rotterdam Stadionstation. Shown is the percentage of trips across all modes thathave an arrival or departure at the Stadion station for the peopleliving in the municipality of Rotterdam and the whole study arearespectively. As the station is not in use in the scenarios Base andLand, the trip ratio in these scenarios is zero. The station is usedmore often by people living in Rotterdam compared to the totalpopulation in the larger study area. Again we see that synergyeffects occur: in the combined scenario the share of people usingthe station is higher than that in Trans only. The difference islargest for people living in the municipality of Rotterdam.

To conclude; our simulation study shows that combined landuse and public transport policies may lead to positive synergyeffects.

4.5. Concluding remarks

We conclude this case study with some final comments. Thefirst comment relates to assumptions with respect to travelbehavior. Such assumptions are always needed to model travelbehavior. This applies to any modeling of travel behavior, so alsofor modeling the impacts of synchronization policies on travelbehavior. It is most common to assume that the traveler aims tomaximize his or her utility (Random Utility Maximizing), thoughalternatives exist, such as Random Regret Minimization (e.g.Chorus, 2010; Chorus et al., 2008), or premises of Prospect Theory,such as loss aversion (e.g. Van de Kaa, 2010). Note that theconceptualization of policy options is not affected by the beha-vioral assumptions regarding travel behavior.

Secondly, in Section 3 we discussed the importance of barriersfor the implementation of synchronization policies. Because thedecision making process of the Rotterdam case is not finished yet,we cannot tell which barriers played which role in the finaldecisions. Nevertheless some reflections can be made. A firstpotential barrier is that financing the plans is too expensive for themunicipality of Rotterdam only. In The Netherlands municipalities

33.41 33.06 31.92 31.70

15.70 16.05 18.43 18.80

22.30 22.27 20.32 20.31

17.02 17.03 17.11 17.10

102030405060708090

100

Base Trans Land Combwalk PT car bike

% trip ratio

Fig. 3. Mode distribution.

66 77

178 195

289 301 304 328

0

100

200

300

400

Base Trans Land Comb

#person

P+R B+R

Fig. 4. Use of PþR and BþR.

0

0.77

0

0.97

0

0.93

0

1.48

0

1

2

Base Trans Land CombCorridor Rotterdam

% (trip ratio)

Fig. 5. Use of Rotterdam Stadion station.

B. van Wee et al. / Transport Policy 31 (2014) 1–9 7

have limited ‘own’ budgets and rely on co-financing from theMinistry of Infrastructure and the Environment. Diverging inter-ests of the Ministry as opposed to the municipality of Rotterdamcan both support as well as endanger the implementation ofsynchronization policies. Because the Ministry will considerimpacts at the national and regional level, its involvement canincrease the scope of impacts of the policies, and so add to theinterest of the municipality. Especially if synchronization hasbenefits at the regional or national level, the involvement of theMinistry can play a positive role. On the other hand, the Ministry isless interested in local effects, which can endanger positivedecisions leading to local benefits from synchronization. A secondpotential barrier could be that the benefits of synchronization arerelatively abstract, and less clear or even not at all known todecision makers. Our involvement by applying our model in thisreal world case aims to reduce this barrier. Planners and otheractors are now familiar with the options for synchronization andtheir benefits, and can communicate them to decision makers.

5. Evaluation synchronizing policies

An important question from a policy perspective is: how toevaluate candidate policy options for synchronization? In the casestudy presented in Section 4 we only discuss impacts on modechoice. As explained in the introduction synchronization is not afinal policy goal, but it aims to contribute to better accessibility.So evaluations should at least include any indicator for accessibility.But synchronization policies may have multiple other effects. Forexample, the costs of public transport services may change, as wellas the costs of extending the opening hours of childcare facilities.Furthermore, if synchronization results in a shift from car to publictransport, emissions of CO2 and pollutants, as well as noise impacts,may decrease.

We first briefly discuss the evaluation of individual accessi-bility. When evaluating network synchronization the inclusion ofmulti-modal travel and multi-activity trips in the measurement ofaccessibility is crucial. However it is very difficult to include multi-activity and multi-modal travel in individual accessibility mea-sures (Fiorenzo-Catalano, 2007). The early supernetworkapproaches allow the modeling of integrated multi-modal net-works and (single activity) multi-model trips (Sheffi, 1995;Van Nes, 2002; Carlier et al., 2002). Later, inspired by the work onextended supernetwork representations as developed in Nagurney(2003); Arentze and Timmermans (2004) introduced the multi-state Supernetwork, which integrates different networks of pas-senger transport, activity locations and activities of travelers into asingle network. The multi-state supernetworks were enhanced inimportant ways in Liao et al. (2010, 2011, 2012, 2013a, 2013b),the resulting model was already introduced in Section 4.3. Whereconventional Supernetworks are able to identify the shortest pathsfor single trips through a multi-modal network, multi-state Super-networks are able to identify the shortest paths for complete dailyactivity programs. Hence, the model is able to identify which pathproduces the least generalized costs to complete a daily activityprogram in a given urban system. The ease (e.g. highest utility orleast generalized costs) with which this path can be conducted isan indicator of the accessibility of that urban system. Hence, anaggregation across a population of individuals is used as a measureof accessibility of a complete transport–land use system. For athorough discussion of the advantages and potential of activity-based models for analyzing the effect of land-use policies on travelbehavior we refer to Shiftan (2008).

After evaluating the accessibility effects of synchronization poli-cies, the next question is how to evaluate their cons. These mainlyrelate to the additional costs of infrastructure, land development in

areas with higher square meter prices, the additional costs ofchanging opening hours, and the costs of removing barriers asaddressed in Section 3. A simple cost-effectiveness evaluationmethod (comparing costs per change in any accessibility indicator)is therefore not an option. Two important evaluation methods thatin principle are suitable in this case are Cost Benefit Analysis (CBA)and Multi-Criteria-Analysis. Below we briefly present both optionsfollowed by a discussion on the suitability of both options forour case.

In most western countries CBA is the method used for ex anteevaluations of transport infrastructure projects (Hayashi andMorisugi, 2000). A CBA aims to capture the pros and cons of aproject, although it is limited to those pros and cons that can beincluded in a welfare-based economic framework. The valuation ofeffects is as much as possible based on the preferences ofconsumers, mainly on their willingness to pay (WTP) for theeffects. For some effects the WTP of consumers is considered tobe less suitable, CO2 emissions being an example. For those effectsvaluations based on policy preferences are generally used. Costsand benefits occur in different years within the time horizon of aCBA. To deal with this, they are presented as so-called ‘net presentvalues’, implying that taking into account interest and inflation itis better to have 1 euro or dollar nowadays than in, for example2030. The so-called discount rate is used to express this time-dependent valuation. Final results are often presented in summar-izing indicators, main examples being the difference betweencosts and benefits, the benefit-cost ratio and the return oninvestment. Almost every handbook on transport economics paysattention to CBA in transport (see, for example, Blauwens et al.,2008; Button, 2010). However, some effects of policies cannoteasily be quantified or expressed in monetary terms. One can thinkof the impact of a new motorway or rail line on the landscape andnature, or the value of a nice bridge because it is a landmark.If effects are present which are of high importance but difficult toquantify or monetize, a CBA may at best be suitable for partialevaluations of the pros and cons of the project under considera-tion. An MCA is probably then preferable – it can include anyeffect, positive or negative, and by setting weights seeminglyincomparable effects can be included in the equation (such asthe landmark function of a bridge and its travel time gains).

CBA is a popular evaluation method in the area of transportbecause most costs and benefits are quite well known (at leasttheoretically) and can be expressed in monetary terms, becausethe risk of double counting effects is limited, and because it is lesseasy to manipulate results than in case of an MCA. An importantquestion is: does this also apply to the options for synchronizationpolicies? We argue that it does, as discussed above (see the textson the multi-state supernetwork synchronization leading toincreased accessibility and related reductions of travel times, lessinconvenience when people need to switch modes, and thepossibilities of changes in activity locations). Such benefits canbe quantified, either via utility-based accessibility indicators basedon the multi-state Supernetwork approach as described above, orusing the logsum indicator (e.g. De Jong et al., 2007), or viachanges in the different components of this indicator as traveltimes and the marginal value of time. The multi-state supernet-work and logsum approaches include valuations of changes inconvenience as well as attractiveness of locations for activities.An alternative would be more traditional approaches for evaluations.The common approach is to multiply changes in generalizedtransport costs (time, money, effort) by the number of travelersand the value they attach to reduction in travel times (and costs).In addition induced demand (new travel) is valued according tothe so called ‘rule of half’. Using this approach the evaluator canbenefit from the literature providing values for the valuation ofpeople of these changes in generalized transport costs, not only

B. van Wee et al. / Transport Policy 31 (2014) 1–98

travel time savings, but also changes in the convenience (e.g.changing modes), e.g. Wardman (2001). However, some effects ofsynchronization policies may be more difficult to quantify orexpress in monetary terms. For example, if synchronizing landuse and the transport networks led to increased liveability on thestreets (more walking, more cycling) there may be benefits inaddition to transport related benefits, such as increased (percep-tions of) safety, and a greater appreciation of the environment.Such effects are difficult to quantify and express in monetaryterms. If such effects play an important role probably a combinedCBA (for effects that can be monetized easily) and MCA (for othereffects) would be preferable. If there is a strong argument forimplementing such policies, an MCA may be preferable.

6. Conclusions

Synchronization between transport networks and/or activitylocations can have important accessibility benefits for individuals.Synchronization options refer to spatial and/or temporal synchro-nizations between locations of activities and/or transportationnetworks. To optimally synchronize the transport and land usesystem dedicated transport and spatial policies are of greatimportance. This paper provides a systematic overview of policyoptions for synchronization.

The components that can be synchronized include (a) transportnetworks, (b) activity locations, (c) activity locations and parts oftransport networks, and finally (d) activities can be decoupledfrom space and/or time. By linking these components to thedimensions ‘time’ and ‘space’ a 4 by 2 table is obtained categoriz-ing synchronization options. Next we link these options tocategories of policy instruments: restrictions/regulations, pricingpolicies, land-use planning, infrastructure provision, specific publictransport policies, marketing and information, and time relatedpolicies, such as with respect to opening hours. The paper discussesthese links, considering several examples of policy options.

The Rotterdam Stadionpark case study shows that our con-ceptualization of synchronization options and policy instrumentscan easily be applied to a real world case. In addition it shows thatcombining land use and transport policy options can inducepositive synergy effects: the combined effect of both policiesexceed the sum of the land use and transport policies separately.

Finally we discuss evaluations of candidate synchronizationpolicies. We argue that the accessibility impacts of synchroniza-tion policies can be evaluated elegantly using the multi-statesupernetwork approach. To evaluate all the important pros andcons of policy options for synchronization (accessibility andothers) a CBA will often be the method of preference. However,if important pros or cons cannot be quantified or expressed inmonetary terms, an MCA or hybrid methodology (CBA and MCA)may be preferred.

Acknowledgment

This research is made possible by financial support from theSustainable Accessibility of the Randstad (SAR) program, financedby NWO and the Dutch Ministry of Infrastructure and theEnvironment.

References

Arentze, T.A., Molin, E.J.E., 2013. Travelers’ preferences in multimodal networks:design and results of a comprehensive series of choice experiments. Transp.Res. Part A 58, 15–28.

Arentze, T.A., Timmermans, H.J.P., 2004. A multi-state supernetwork approach tomodeling multi-activity, multi-modal trip chains. Int. J. Geogr. Inf. Sci. 18 (7),631–651.

Bertolini, L., Curtis, C., Renne, J., 2012. Station area projects in Europe and beyond:towards transit oriented development? Built. Environ. 38 (1), 31–50.

Blauwens, G., De Baere, P., Van de Voorde, E., 2008. Transport Economics.De BoeckAntwerpen.

Button, K.J., 2010. Transport Economics, 3rd edition Edward Elgar, Cheltenham, UK.Carlier, K., Fiorentzo-Catalano, S., Lindveld, C., Bovy, P., 2002. A supernetwork

approach towards multimodal travel modeling. In: Proceedings of the 82-thAnnual Meeting of the Transportation Research Board, Washington, D.C. (CD-ROM).

Chorus, C.G., 2010. A new model of random regret minimization. Eur. J. Transp.Infrastruct. Res. 10 (2), 181–196.

Chorus, C.G., Arentze, T.A., Timmermans, H.J.P., 2008. A random regret-minimization model of travel choice. Transp. Res. Part B 42 (1), 1–18.

Couclelis, H., 2009. Rethinking time geography in the information age. Environ.Plan. A 41 (7), 1556–1575.

De Jong, G., Daly, A., Pieters, M., van der Hoorn, T., 2007. The logsum as anevaluation measure: review of the literature and new results. Transp. Res. PartA 41 (9), 874–889.

Farag, S., Schwanen, T., Dijst, M., Faber, J., 2007. Shopping online and/or in-store?A structural equation model of the relationships between e-shopping and in-store shopping. Transp. Res. A 41 (2), 125–141.

Fiorenzo-Catalano, S., 2007. Choice Set Generation in Multi-Modal TransportationNetworks (Dissertation). Delft University Press, Delft.

Geurs, K.T., Van Wee, B., 2004. Accessibility evaluation of land-use and transportstrategies: review and research directions. J. Transp. Geogr. 12 (2), 127–140.

Guihaire, V., Hao, J.-K., 2008. Transit network design and scheduling: a globalreview. Transp. Res. Part A 42 (10), 1251–1273.

Guo, Z., Wilson, N.H.M., 2011. Assessing the cost of transfer inconvenience in publictransport systems: a case study of the London Underground. Transp. Res. Part A45 (2), 91–104.

Hayashi, Y., Morisugi, H., 2000. International comparison of background conceptand methodology of transportation project appraisal. Transp. Policy 7 (1),73–88.

Levine, J., 2005. Zoned Out: Regulation, Markets, and Choices in Transportation andMetropolitan Land Use. RFF Press, Washington DC.

Liao, F., Arentze, T.A., Timmermans, H.J.P., 2010. Supernetwork approach for multi-modal and multi-activity travel planning. Transp. Res. Rec, 2175, 38–46.

Liao, F., Arentze, T.A., Timmermans, H.J.P.., 2011. Constructing personalized trans-portation network in multi-state supernetworks: a heuristic approach. Int. J.Geogr. Inf. Sci. 25 (11), 1885–1903.

Liao, F., Arentze, T.A., Timmermans, H.J.P., 2012. A supernetwork approach formodeling traveler response to park-and-ride. J. Transp. Res. Rec. 2323, 10–17.

Liao, F., Arentze, T.A., Timmermans, H.J.P., 2013a. Incorporating space–time con-straints and activity-travel time profiles in a multi-state supernetworkapproach to individual activity-travel scheduling. Transp. Res. Part B 55, 41–58.

Liao, F., Arentze, T.A., Molin, E.J.E, Bothe, W., Timmermans, H.J.P., 2013b. Effects ofintegrated land-use transport scenarios on travel patterns: a multi-statesupernetwork application. In: Proceedings of the 13th WCTR Conference, Rio,Brazil, 22 pp. (CD-Rom).

Lyons, G., Urry, J., 2005. Travel time use in the information age. Transp. Res. Part A39 (2–3), 257–276.

Mokhtarian, P.L., 2002. Telecommunications and travel: the case for complemen-tarity. J. Ind. Ecol. 6 (2), 43–57.

Mokhtarian, P.L., Salomon, I., 2001. How derived is the demand for travel? Someconceptual and measurement considerations 2001. Transp. Res. Part A 35 (8),695–719.

Nagurney, A., 2003. Supernetworks: paradoxes, challenges and new opportunities.In: Paper presented at the 1st International Conference on the Economic andSocial Implications of Information Technology, Washington, D.C, pp. 229–254.

Sheffi, Y., 1995. Urban Transportation Networks: Equilibrium Analysis WithMathematical Programming Methods. Prentice Hall, New Jersey.

Shiftan, Y., 2008. The use of activity-based modeling to analyze the effect of land-use policies on travel behavior. Ann. Reg. Sci. 41 (1), 79–97.

Van de Kaa, E.J., 2010. Prospect theory and choice behaviour strategies: review andsynthesis of concepts from social and transport sciences. Eur. J. Transp. Inf. Res.10 (4), 299–329.

Van Nes, R., 2002. Design of Multimodal Transport Networks: A HierarchicalApproach (Dissertation). Delft University Press, Delft.

Wardman, M., 2001. A review of British evidence on time and service qualityvaluations. Transp. Res. Part E 37 (2–3), 107–128.

B. van Wee et al. / Transport Policy 31 (2014) 1–9 9