polycentric urban regions and the quest for synergy: is a network of cities more than the sum of the...

19

Click here to load reader

Upload: evert

Post on 25-Dec-2016

214 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

http://usj.sagepub.com/Urban Studies

http://usj.sagepub.com/content/42/4/765The online version of this article can be found at:

 DOI: 10.1080/00420980500060384

2005 42: 765Urban StudEvert Meijers

than the Sum of the Parts?Polycentric Urban Regions and the Quest for Synergy: Is a Network of Cities More

  

Published by:

http://www.sagepublications.com

On behalf of: 

Urban Studies Journal Foundation

can be found at:Urban StudiesAdditional services and information for    

  http://usj.sagepub.com/cgi/alertsEmail Alerts:

 

http://usj.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://usj.sagepub.com/content/42/4/765.refs.htmlCitations:  

What is This? 

- Apr 1, 2005Version of Record >>

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 2: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

Polycentric Urban Regions and the Quest forSynergy: Is a Network of Cities More thanthe Sum of the Parts?

Evert Meijers

[Paper first received, February 2004; in final form, August 2004]

Summary. Polycentric urban regions, or urban networks, are often associated with the notion ofsynergy, the assumption being that the individual cities in these collections of distinct butproximally located cities relate to each other in a synergetic way, making the whole network ofcities more than the sum of its parts. Drawing on economic network theories, an analysis of thepresence of synergy is carried out for the Randstad region in the Netherlands, which is oftenconsidered a classic example of a polycentric urban region. The analysis focuses on the synergymechanisms of co-operation and in particular complementarity. The results are mixed. In termsof co-operation, the Randstad has become more synergetic. However, the less complementaryeconomic roles of the cities caused a reverse effect.

1. Introduction

Interest in polycentric urban regions hasincreased rapidly during the past decade.This can be partly explained by the emergingbelief that regions in general are becomingthe most important spatial level of inter-national territorial competition. In addition,polycentric urban regions are believed to bethe next stage in the expansion of urbanliving space, particularly in densely populatedcountries or regions. The daily urban space ofactors would not only cover the city, itssuburbs and its surrounding rural area, butwould extend to include other cities as well.In polycentric urban regions, cities seem tohave coalesced in functional and morphologi-cal terms into larger and more dispersedregional urban systems. In the literature, poly-centric urban regions are often defined as

collections of historically distinct and bothadministratively and politically independentcities located in close proximity and well con-nected through infrastructure (see Kloostermanand Lambregts, 2001). However, as the litera-ture on polycentric urban regions is stilllimited and therefore rather unconsolidated(Bailey and Turok, 2001), a diversity ofconcepts is applied, which are largely synon-ymous with the polycentric urban regionconcept used here. Recent examples include‘multicore city-regions’ (Westin and Osthol,1994), ‘network cities’ (Batten, 1995), ‘citynetworks’1 (Camagni and Salone, 1993)and ‘polynucleated metropolitan regions’(Dieleman and Faludi, 1998a). Moreover, interms of ideas on spatial structure andinterurban relationships, the polycentricurban region concept builds on older conceptssuch as the ‘dispersed city’ (Burton, 1963),

Urban Studies, Vol. 42, No. 4, 765–781, April 2005

Evert Meijers is in Delft University of Technology, OTB Research Institute for Housing, Urban and Mobility Studies, PO Box 5030,2600 GA Delft, The Netherlands. Fax: þ31 15 2783450. E-mail: [email protected]. The author would like to thank HugoPriemus, Marjolein Spaans, Henny Coolen and Frans Dieleman for their helpful comments. The research on which this article isbased is funded by the Delft Centre for Sustainable Urban Areas research programme within Delft University of Technology, andby the Habiforum Foundaton’s research programme ‘System innovation, urban and regional land use and area development’ (SRG).

0042-0980 Print=1360-063X Online=05=040765–17 # 2005 The Editors of Urban Studies

DOI: 10.1080=00420980500060384 at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 3: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

‘megalopolis’ (Gottmann, 1961) or the ideas ofStein and Mumford on ‘the regional city’(Stein, 1964).Interest in polycentric urban regions has also

increased amongpoliticians andurbanplanners,as evidenced by their appearance as planningconcepts in planning policies in a wide varietyof European countries including Belgium(‘urban networks’), Denmark (polycentric‘national centres’), Estonia (‘urban networks’),France (‘reseaux des villes’),Germany (‘Metro-polregionen’; ‘Stadtenetze’), Greece (‘twinpoles’), Italy (‘city networks’), Ireland (‘linkedgateways’), Lithuania (‘Metropolis Vilnius–Kaunas’), the Netherlands (‘urban networks’),Poland (‘duopols’) and Switzerland (‘vernetzteStadtesystem’). Interestingly, planning policyconcepts referring to polycentric urban regionsoften make use of the network metaphor. Thisis by no means a coincidence. The metaphorof the network emphasises the complex andstrong relationships between the cities andthus the coherence and unity of the region.Moreover, networks are associated with econ-omies of scale, critical mass and synergy. Notsurprisingly, the network metaphor hasbecome part of the standard vocabulary ofadministrators, planners and other policy-makers promoting or otherwise dealing withpolycentric urban regions. In particular, theidea of synergy, or ‘being more than the sumof the parts’ is a central objective in manypolicies for polycentric urban regions.Examples are manifold, but can for instancebe found in the Flanders Structural Outline(Ministerie van de Vlaamse Gemeenschap,1997) which introduces the concept of theFlemish Diamond (see also Albrechts, 1998)and in the proposals for a new national spatialpolicy in the Netherlands in which the conceptof ‘urban networks’ (for instance, the Randstadregion) plays a prominent role (Ministerie vanVolkshuisvesting, Ruimtelijke Ordening enMilieubeheer, 2004). The perhaps rather loosecollection of cities in both regions wouldsuddenly turn into a metropolis, and plannershave suggested that the Randstad and FlemishDiamond could compete with the highest-levelmetropolitan agglomerations such as Londonand Paris, claiming that they provide

economies of scale without incurring the costsor agglomeration diseconomies that theselarge metropolises entail.This article addresses the more theoretical

underpinning of this quest for synergy bypolicy-makers. Its aim is to explore whetherthere is a theoretical basis for assuming thepresence of synergy in polycentric urbanregions and, building on that, to develop aconceptual model to analyse the presenceof synergy in polycentric urban regions.Subsequently, the presence of synergy in theRandstad region is explored. The structure ofthis article is as follows. First, an overviewof economic network theories in relation tothe principle of synergy is presented. Assum-ing that polycentric urban regions may beconsidered as a network in some sense,this may shed some light on how synergywithin such regions can be achieved as thesame network fundamentals and synergyprinciples are likely to apply. The thirdsection addresses this issue of synergy in poly-centric urban regions, resulting in a concep-tual model to analyse the presence ofsynergy. Before applying this model to theRandstad in sections 5 and 6 in order to estab-lish whether there is synergy developing,the fourth section introduces this ‘prime’ or‘classic’ example of a polycentric urbanregion (Hohenberg and Lees, 1985; Batten,1995).

2. Synergy in Networks

The concept of networks is strongly linked tothe notion of synergy. The aim of this sectionis to present some fundamental knowledge onnetwork synergies, drawing on recent analysisof synergy in economic theory. The networksconsidered in these analyses are mainlynetworks between firms, transport networksand communication networks.The word ‘synergy’ comes from the Greek

(syn þ ergos) and refers to a ‘situation inwhich the effect of two or more co-operatingor combined bodies or functions is largerthan the sum of the effects each body orfunction alone can achieve. Synergy is oftenpopularly formulated as 1þ 1 . 2, which,

766 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 4: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

however, emphasises well that synergy can beexpressed as the rise in performance of anetwork through efficient and effective inter-action. Although widely used in economictheory, synergy is a rather fuzzy concept asit hides different mechanisms from whichsynergy is derived. Capello and Rietveld,analysing the synergy concept in a variety ofeconomic theories at the micro-, meso- andmacro-levels arrive at three distinct meanings

– Synergy means that when two or moreactors co-operate, there is a positive resultfor both of them;

– synergy means that when co-operativebehaviour is implemented in order toexploit complementarity in the productionof a particular good, advantages are obtainedby economic actors taking place in a group;

– synergy means that when individuals orfirms ‘voluntarily and non-voluntarily’ arepart of a group, externalities may bepresent and exploited by these individualsor groups in conducting their economicactivity (Capello and Rietveld, 1998, p. 64).

The presence of one of these three synergy-releasing mechanisms—co-operation, com-plementarity or externality—combined withnetwork behaviour generates synergy andhence economic benefits for the actors.

Next to these different mechanisms, thetype of network is also relevant. Differentsynergy mechanisms play a role in the varioustypes of network. Networks in general aremade up of nodes (cities, households, firms,organisations, individuals), linkages betweenthe nodes (infrastructure, relationships, ties),flows (people, goods, information, capital) andmeshes.

Networks can be labelled as being of the‘club’ type or of the ‘web’ type (Capineriand Kamann, 1998). In club networks, actorsshare a common objective, activity orservice, while also having parallel interestsand transaction chains. A classic but tellingexample is the tennis club. Members of sucha club cannot afford the common objective—facilities for tennis—on their own, butby organising themselves they can. Moremembers lead to positive externalities such as

lower membership fees or extended openinghours. Negative externalities arise, however,when all members want to play tennis at thesame time. Web networks, on the other hand,are characterised by different activities ofthe actors. These are complementary insteadof similar and are linked in a serial way.A typical example is a chain of enterprises orbusiness units each undertaking a certainphase in the production of a product.

Distinguishing between club and web net-works is relevant as synergy in both networksis achieved in different ways. In club networks,so-called ‘horizontal synergy’ can be achieved;in web networks, ‘vertical synergy’. In the caseof horizontal synergy, the synergy derives fromco-operation leading to economies of scaleand so-called positive network externalities.Characteristic of these economies of scale isthat they apply only to participants in thenetwork. Externalities are present when thecosts of participating in the network are lessthan the benefits of the co-operation. Thevalue of co-operation is well articulated byCapineri and Kamann

Actors have the choice between indepen-dent ‘stand alone’ strategies where theyperform all activities themselves usuallyat higher costs resulting in lower perform-ance and strategies of co-operation whichresult in the transfer of activities and/orresources to other actors increasing a largerange of types of dependencies but alsoimproving their performance (Capineriand Kamann, 1998, p. 42).

Vertical synergy is the surplus valuefollowing from agglomeration or specialis-ation effects. Complementarity is the keysynergy mechanism here. The synergyresults from a specialisation process, redistri-buting resources and activities among theparticipating actors according to theircompetence. This means that the individualperformance of actors improves as they canfocus their efforts on their core activities,abandoning non-core activities unnecessarilyabsorbing energy (Capineri and Kamann,1998). Classic network development theoriesalso point to specialisation as the outcome of

SYNERGY IN POLYCENTRIC URBAN REGIONS 767

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 5: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

a process of rationalisation of the networkstructure once it has achieved a considerablelevel of complexity.To summarise, synergy is achieved through

the mechanisms of co-operation, comple-mentarity and externalities linked to them.Co-operation leads to horizontal synergypossibly achieved in club-type networks,complementarity to vertical synergy possiblyachieved within web-type networks. External-ities are present in both. They represent themost important economic advantage of networkbehaviour

It is . . . a matter of exploiting scaleeconomies in complementary relationshipsand synergic effects in co-operative acti-vities, achieved through participation inthe network (Capello, 2000, p. 1927).

3. Synergy in Polycentric Urban Regions

Transferring the concept of synergy to citiesdoes not seem too difficult. In fact, the firstcities emerged because of synergy, develop-ing from the advantages that arose fromagglomeration economies. Living andworking in cities entail advantages such asthe supply of public services, specialisedproducts and services, a large and diversifiedurban market and easy exchange and avail-ability of information. The question, however,is how such agglomeration economies canalso be organised in a network of cities.This brief analysis of the synergy concept

and some network fundamentals sheds somelight on the way synergy can be achieved inpolycentric urban regions. The cities makingup a polycentric urban region can be consideredthe nodes in a network that is further made upby infrastructure, interurban relationships andflows. A city in itself is an accumulation ofmany other nodes such as households, firms,individuals, organisations, each one connectedto other nodes by infrastructure, flows andinterdependencies. So, in a polycentric urbanregion, a multitude of other networks can befound, by no means restricted to the scale ofthe polycentric urban region. However, here

we focus on a macro level, thus on the poly-centric urban region as a network of cities.As it can be assumed that polycentric urban

regions are indeed networks (there are nodes,linkages, flows and meshes), it is likely thatthe same basic knowledge of synergy in net-works applies to these spatial phenomena.2

Depending on whether polycentric urbanregions are networks of the club or the webtype, the same mechanisms will lead tosynergy. This means also that in polycentricurban regions, synergy is established throughthe mechanisms of co-operation and comple-mentarity (and externalities involved in both).Two important questions remain. The first

is whether a polycentric urban region is aclub- or a web-type network. The second ishow these two mechanisms should be givena translation relevant to spatial phenomenasuch as polycentric urban regions. Categoris-ing a network of cities as a club-type orweb-type network seems a fruitless endea-vour, given its complex nature. In fact,Capineri and Kamann (1998) state that, inreal life, networks will have both club-typeaspects and web-type dimensions. This isalso the case with polycentric urban regions,which opens up different ways throughwhich synergy can be established. Polycentricurban regions may be characterised as clubnetworks when cities having similar charac-teristics join forces to achieve some kind ofa common objective or common interests.This co-operation then generates economiesof scale. Examples include co-operationbetween cities performing similar economicroles, such as port cities or tourist cities. Butcities can also co-operate when facingsimilar urban problems or challenges—forinstance, relating to segregation, a weakeconomic base, the need for efficient publictransport or waste disposal. On the otherhand, polycentric urban regions resembleweb networks when the individual citiesperform different economic roles and hostcomplementary urban facilities, activities,residential and working environments.Comparable to the distinction between club-type networks and web-type networks, is theclassification of city networks by Camagni

768 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 6: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

and Salone (1993). They refer to club-typeurban networks as ‘synergy networks’, whileweb-type networks are labelled ‘complemen-tarity networks’. Although both club-typenetworks and web-type networks can bepresent in polycentric urban regions, itseems that web-type networks are of particu-lar relevance for polycentric urban regions.With these, proximity matters in the sensethat market areas overlap. Club-type networkscan play a role, but they are also importantin networks among distant cities (thinkfor instance of club-type networks such asMETREX or Eurocities which build oncommon interests).

The second question to be answered is howco-operation and complementarity (and theexternalities involved) should be interpretedin the context of polycentric urban regions toprovide a meaningful framework for analysis.As mentioned before, our focus is on thenetwork at a macro level in polycentricurban regions—that is, a network betweenthe cities rather than between firms, organis-ations or persons located in these cities.Consequently, co-operation is interpreted asco-operation between cities. Public adminis-tration tends to be organised in a territorialhierarchy. However, many spatial issuesthese days call for an approach that is formu-lated and implemented at multiple scales andacross several administrative tiers. Addition-ally, an increasing number of spatial issuesare, or preferably should be, addressedthrough a governance rather than a govern-mental mode. This requires the involvementof multiple public, private and organisedinterest-groups, thereby taking into accountthat different issues call for different allianceswith different spatial competencies and differ-ent life-spans (see Boelens, 2000). Seen fromthis perspective, a focus on co-operationbetween cities seems too narrow. What isneeded is regional organising capacity—thatis, the ability to co-ordinate developmentsregionally through a more or less institutiona-lised framework of co-operation, debate,negotiation and decision-making in pursuitof regional interests in which a multitude ofpublic and private stakeholders participate

(Meijers and Romein, 2003). The externalitiesthat may arise depend on the utilisation andfunctioning of such frameworks. Synergyrequires a high level of interaction whichwill generate the necessary network cohesionto make up for the increased interdependency(see Capello and Nijkamp, 1993). Moreover,actors must be willing and able to adjusttheir internal profile and external behaviour.Free-rider behaviour is to be avoided. Toestablish whether synergy has developed, weneed to consider the extent to which suchframeworks are present in polycentric urbanregions.

Although some previous work on concep-tualising ‘complementarity’ as an interurbanrelationship has been done (Ullmann, 1956;Lambooy, 1969; Camagni and Salone,1993), the concept has remained rathervague, despite its increasingly frequent, butoften casual, appearance in both academicwritings and policy documents. This lack ofconceptual clarity probably explains why theconcept has so far not been empiricallyanalysed.

In urban regions, complementarity refers tothe specific nature of a relationship betweentwo or more relatively similar activities orplaces. ‘Activities’ include economic activi-ties, such as commercial services, or urbanfacilities such as education, culture andmedical care. ‘Places’ on the other handrefer to business milieus or residentialmilieus. As places make up cities and mostactivities take place within cities, it is alsoexpedient to refer at a macro level to citiescomplementing each other. For activities andplaces (or indirectly cities) to be complemen-tary, they need to satisfy two importantpreconditions relating to supply and demand

(1) There must be differentiation in thesupply of activities and/or places.

(2) The geographical markets of demand forthese activities or places must at leastpartly overlap.

To give some examples, two universities arecomplementary if they offer different acade-mic education, while they are at the sametime recruiting their students from more or

SYNERGY IN POLYCENTRIC URBAN REGIONS 769

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 7: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

less the same region. Similarly, two hospitalsare complementary when they provide fordifferent medical specialisations, or specialisein different kinds of treatment—for instance,standardised routine operations versusspecialised knowledge-intensive care whilstserving more or less the same region. Twoor more residential areas are complementarywhen they offer different residential milieus,thus providing alternatives to match the differ-ent preferences of a regional population (seeMusterd and van Zelm, 2001). At a macrolevel, two cities are complementary whenone specialises in, for instance, financialservices and the other one in transport andlogistical services, each also providing theseservices to businesses or citizens located inthe other city.Complementarity often leads to spatial

interaction. In fact, Ullmann (1956), whodescribes complementarity as differentiation,argues that complementarity is the mainexplanation for the development of spatialinteraction. Similarly, Batten (1995) statesthat links between the cities in a polycentricurban region (or ‘network city’ as he termsit) are forged on the basis of complementaryfunctions rather than on the basis of distanceor demand thresholds. However, merecomplementarity does not suffice for spatialinteraction to occur. Following Stouffer(1940), Ullmann (1956) points to the role ofintervening opportunities (intervening sourcesof supply), as well as the role of transfer-ability (the costs of interaction) in determiningwhether or not spatial interactions arise fromcomplementarity. So, spatial interactions onlypartly reveal the complementarity relationshipspresent.One of the ideas behind the polycentric

urban region concept is that it is not one citythat provides a complete array of economicfunctions, urban facilities or residential andbusiness environments, but rather the wholesystem of cities within a region. Such a situ-ation would provide for externalities. Whentwo cities complement each other, then thecitizens and companies in one place can takeadvantage of the various functions the othercity has to offer. These functions can then be

more specialised, as the demand market onwhich they build is larger given the overlap-ping of hinterlands. In such a way, companies,citizens and tourists can choose from a larger,more specialised and diverse collection ofurban functions (public services, facilities,business services), businesses milieus andresidential milieus. In other words, comple-mentarity is strongly linked to agglomerationeconomies.To sum up, synergy in polycentric urban

regions is generated through

– co-operation (regional organising capacityor frameworks for co-operation and theirfunctioning leading to horizontal synergy);

– complementarity (differentiation in theeconomic roles of cities, in urban facilities,in business and residential milieus coupledwith a regional demand leading to verticalsynergy).

In the next section, this conceptual model willbe applied to the Randstad region in order toestablish whether there is synergy presentand developing.

4. The Randstad: Is It More than the Sumof the Parts?

The Randstad provides an excellent casefor analysing the extent of synergy in apolycentric urban region, not only becausemany authors have claimed that it is aclassic stereotype of such a network of cities(see for instance Stein, 1964; Hohenberg andLees, 1985; Batten, 1995), but also becauseseveral authors have suggested the presenceof synergy. For instance, Camagni andSalone (1993) refer to the Randstad as anexample of a complementary network madeup of specialised and complementarycentres. This is fully endorsed by van derKnaap (1994) who states that in the Dutchcontext there seems to be a dominance ofcomplementary networks, based on verticalintegration, spatial specialisations and comp-lementary urban centres. This would beparticularly exemplified in the Randstad,where it is mainly expressed in the relationsbetween the larger and smaller cities. After

770 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 8: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

briefly introducing the Randstad region, wewill analyse whether such claims can bebased on empirical evidence.

The Randstad is the horseshoe-shapedurban constellation in the western part of theNetherlands (see Figure 1). The Dutch word‘Rand’ means ‘rim’ and refers to the positionof the Randstad encircling a green open areacalled the Green Heart. Seven million peoplelive in the Randstad (44 per cent of the Dutchpopulation) and 45 per cent of the nationalemployment is located in what is less than20 per cent of Dutch territory. The anchors of

the Randstad are formed by the four largestcities of the Netherlands: Amsterdam, Rotter-dam, The Hague and Utrecht. Together withsome dozens of other medium-sized andsmaller cities and the absence of one predomi-nant centre, the Randstad can be characterisedas a stereotypical polycentric urban region.This polycentric pattern is basically inheritedfrom the past, as fragmented political andadministrative structures have prevailed inthis area (in fact, in a large part of north-west Europe) for centuries (Dieleman andFaludi, 1998b). However, the Randstad, as it

Figure 1. The Randstad.

SYNERGY IN POLYCENTRIC URBAN REGIONS 771

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 9: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

is nowadays, is also the result of urban andregional planning, being, with its counterpartthe Green Heart, at the core of Dutch planningpolicies since the 1950s (Zonneveld, 1991;Faludi and van der Valk, 1994). A commondivision of the Randstad is into a north wing(including Amsterdam, Utrecht and surround-ing cities) and a south wing (The Hague,Rotterdam and surrounding cities). Suggestionshave been made that these two wings provide amore accurate definition of coherent urbanregions than does the Randstad (van derLaan, 1998; Kloosterman and Lambregts,2001).The Randstad region still dominates in

Dutch planning. The proposals for a newnational spatial policy introduce the conceptof ‘urban networks’ and the Randstad isthe most important one (Ministerie vanVolkshuisvesting, Ruimtelijke Ordening enMilieubeheer, 2004). In order to increase theRandstad’s competitiveness, emphasis is puton the need for synergy. This pursuit ofsynergy is also expressed by several othernetworks of relevant actors in the Randstad.For instance, the Delta Metropolis Associa-tion (see also next section) states as the firstguiding principle for the development of theRandstad

In a European perspective, the developmentof the Delta Metropolis is aimed at theachievement of synergy in the activities ofthe people living there; more synergy ispossible and needed in the light of social,economic and cultural activities andefforts aimed at a sustainable environment(Delta Metropolis Association, 2000, p. 1).3

And, the ‘Projectgroep RandstadinbrengVijfde Nota’, in which several provinces,regional authorities and the major cities inthe Randstad were represented, has presentedas a first major administrative outline that

The urbanised western part of our countryneeds a quality leap towards a Blue GreenDelta Metropolis. This must be seen in thecontext of the pursuit to let the Randstadbe more than the sum of the parts, inorder to take a competitive position in

Europe. Co-operation and fine-tuningbetween networked cities, each havingtheir own specialities, and the wetlands ofthe Green Heart will give the Randstad asa whole added value and will increase thequality of life. The necessary increase inquality requires an extended support forfacilities and amenities, distinct profilingof areas and an increase in diversity(Bureau Regio Randstad, 2001, p. 6).

The following sections present the results of anexplorative analysis of synergy in theRandstad, addressing co-operation and com-plementarity. Most attention, however, willbe paid to the synergy mechanism of comple-mentarity for two reasons. First, it was arguedthat, compared with co-operation, this mech-anism is most prominent in polycentric urbanregions as proximity matters. Secondly,claims for the Randstad being a complemen-tary network have been made (Camagni andSalone, 1993; van der Knaap, 1994).

5. Co-operation or Horizontal Synergy inthe Randstad

There are three administrative tiers in theNetherlands: national government, provincesand municipalities. The Randstad extendsover four different provinces, which compli-cates the recognition of and dealing with thecomplex, multiscalar interplay of spatialtrends and forces in an integrated way.Although several attempts have been madeto add a formal administrative tier, in particu-lar at the supralocal scale, the existing frame-work has proved to be rather resistant tochanges. It has become slowly apparent thatmultilevel governance requires co-operationacross scales and across actors, includingprivate actors.In the Randstad, it seems that actors are

increasingly aware of this need. A largevariety of co-operation networks hasemerged in this region in an attempt to over-come the lack of administrative layersbetween the municipal and provincial levelsand between the provincial and nationallevels. Around the 4 largest Randstad cities,

772 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 10: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

city-regions have been formed including thelarge cities and 10–20 adjacent municipali-ties. These co-operation platforms addressissues such as transport, traffic, regionalspatial development, housing, employment,economic affairs and youth welfare. Severalmore or less similar co-operation networksexist around smaller cities (Dordrecht,Leiden) or more or less homogeneousregions in terms of economic activities orlocation. At a higher spatial level, we findco-operation networks in the northern andsouthern wings of the Randstad. Comparedwith the north wing, the co-operation in thesouth wing is closer, probably due to agreater urgency for co-operation because ofa lower level of economic development. Theco-operation in the south wing involves themunicipalities of The Hague and Rotterdam,the province of South Holland and variousco-operation platforms at the city-regionlevel. Finally, also at the scale of the entireRandstad, two interesting co-operationnetworks have emerged in recent years. Thefirst is a formal co-operation between thefour provinces, four regional authorities(city-regions) and four major cities in theRandstad, together organising the ‘BureauRegio Randstad’ (Randstad Agency). Theirobjectives are to foster a balanced anddynamic development of the western part ofthe Netherlands, the Randstad in particular,as an entity consisting of high-quality urbanand rural environments and to strengthen theinternational competitiveness of the Randstad,in particular within Europe. Next to thisformal co-operation between public actors,an informal co-operation platform hasemerged: the Delta Metropolis Association.It was established in 2000 by 12 munici-palities and 4 chambers of commerce fromwithin the Randstad. Being an opennetwork, the number of members has gradu-ally grown and now also includes housing cor-porations, organisations of the agriculture andhorticulture branches, an employer’s organi-sation, the transport sector, environmentalorganisations and water boards. Negotiationsare taking place to involve other organi-sations, for instance the universities, as well.

The co-operation platform functions as athink-tank predominantly occupied with thecentral ambition to let the ‘in principle yetpresent metropolis grow to full stature’(article 2 of the statutes of the Association).All in all, it can be typified as a lobby groupfor the interests of the Randstad region. Inter-estingly, all these co-operation networks wereinitiated by actors within the region, ratherthan dictated by a higher level of government.

The patchwork of co-operation networks inthe Randstad provides for a considerableregional organising capacity in the Randstad.All these networks are examples of clubnetworks as they share a common objectiveand have at least partly corresponding inter-ests and together fund the organisation ofthe network. Thus synergy is likely to beachieved, although the development of extern-alities depends on the functioning of theseoften rather new co-operation networks. Inter-estingly, Capello (2000) has demonstrated thedevelopment of externalities for actors whenparticipating in a co-operation network in anactive and serious way. It needs to be saidthat this co-operation does not mean that thecities do not compete with each other. Onthe contrary, they have competed for centuriesand will continue to contend in the future.However, the awareness that many issuescall for a co-ordinated approach coupledwith the idea that regions have become theprincipal geographical platform for compe-tition (see for instance Storper, 1995; Scott,1998) fuelled a reconsideration of the intra-regional interdependencies resulting in moreco-operation. Interesting in this respect isthat, at the level of the Randstad or its twowings, co-operation networks are formedbetween the larger cities, while the maincompetitors of these cities tend to be thenearby smaller (satellite) cities, who haveprofited from the process of selective out-migration of middle-class and higher-incomehouseholds as well as economic activities(van der Wouden and de Bruijne, 2001).Obviously, the co-operation networks in theRandstad still have to prove themselves, butsome successes are already visible—forinstance, the Randstad region office in

SYNERGY IN POLYCENTRIC URBAN REGIONS 773

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 11: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

Brussels which takes care of the region’sinterests in the European context, and thejoint approach towards policy proposals,investment plans, etc. emanating from thenational government. Perhaps the biggestsuccess so far is that the Randstad region fea-tures prominently in spatial as well as in manysectoral national policies. Obviously, there ismore horizontal synergy in the Randstadthan there was before, even though this hori-zontal synergy is hard to quantify at thismoment.

6. Complementarity or Vertical Synergyin the Randstad

The presence of complementarity in theRandstad has already been assumed byseveral authors (Camagni and Salone, 1993;van der Knaap, 1994), although these assump-tions lack strong empirical justification. Wehave seen that complementarity relates todifferentiation in economic roles, urban facili-ties and business and residential environments.Moreover, a geographical overlapping ofdemand markets for these activities and placesis also important. Consequently, analysingsuch a multifaceted concept as complemen-tarity is a complex undertaking. Here, theexploratory analysis of the presence of comple-mentarity in the Randstad is limited to one ofthese facets—namely, the differentiation ineconomic roles, or economic profiles, of themain cities within the Randstad. Differentiationin economic roles also indicates to some extentthe differentiation in business environments,as one city may provide better conditions forcertain firms and this then becomes explicitin its economic profile. Economic profilesalso present an indication of differentiationin urban facilities as these are part of a city’seconomic profile. Another reason for focusingon the macro level of economic roles ofcities is that the extent of complementaritythat is generally assumed to be present in theRandstad is usually linked to the division oflabour between its main cities (see forexample, van der Wouden and de Bruijne,2001). Of course, the need for selectivityhere does not deny the fact that the issue of

complementarity requires extended analysisand, consequently, this limits the significanceof the results as regards the synergy derivedfrom complementarity in the Randstad.The economic profile of a city is made up

by the number of jobs in that city classifiedby economic sector. For our analysis, weused a data-set derived from the NationalInformation System on Employment (LISA)database. This database is a registration ofall the establishments in the Netherlands,including government and non-commercialorganisations. An establishment in this data-base is defined as a location of a firm, organ-isation, institution or independent professionin which or from which an economic activityor independent liberal profession is beingpractised by at least one employed person.Multi-establishment firms have separaterecordings for each establishment. Theeconomic activities are coded according tothe European Union wide NACE Rev. 1classification (Nomenclature statistique desActivites economiques dans la CommunauteEuropeenne), which is, up to two digits,similar to the ISIC Rev. 3.1 (InternationalStandard Industrial Classification of allEconomic Activities). The database has beenupdated annually since 1996. Our data-setcontains the 1996 and 2002 data on thenumber of jobs classified according tothe two-digit NACE Rev. 1 classificationof economic activities for the municipalitiesof Amsterdam, Rotterdam, The Hague andUtrecht. For reasons of comparison, we alsohave data-sets for the main cities in the northwing (municipalities of Amsterdam, Haarlem,Zaanstad, Utrecht, Amersfoort, Almere,Haarlemmermeer and Hilversum) and southwing (Rotterdam, The Hague, Leiden,Dordrecht, Zoetermeer and Delft).In this paper, correspondence analysis is

used to analyse the differentiation in the econ-omic roles of cities. Correspondence analysisis a technique to analyse the associationbetween rows and columns of a table ormatrix by representing the rows and columnsas points in a low-dimensional Euclideanspace (in practice, often a two-dimensionalplot). Categories with similar distributions

774 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 12: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

will be represented as points that are close inspace and categories that have very dissimilardistributions will be positioned far apart. Foran extensive discussion of correspondenceanalysis, see Greenacre (1993) and Clausen(1998).4 Although often used as a tool toenable graphic interpretation of complexdata, correspondence analysis also providesa single statistic that describes the extent ofdifferentiation in the economic profiles of agroup of cities. This statistic is called thetotal inertia. Total inertia is a measure of theextent to which the profile points are spreadaround a centroid, representing the averageprofile. The larger the distance of the categorypoints from the centroid, the higher the inertia.The highest attainable inertia is equal to thedimensionality of the problem (in our case,the number of cities—1). This maximumwould be reached if all the cities have comple-tely different economic profiles, whereas zeroinertia is attained when they all have exactlythe same economic profiles. In reality,values will be far from the maximum, asreaching the maximum value would imply—for example, that all schools are located inone city, all supermarkets in another one andall doctors in yet another one. In otherwords, cities have a large component ofemployment in non-tradeable economicactivities. Correspondence analysis correctsfor such activities in the sense that they donot, or hardly, contribute to the total inertiastatistic. In order to enable a comparison ofthe inertia between polycentric urban regions,we defined a complementarity ratio forwhich we normalised the total inertia by

dividing it by the maximum total inertia poss-ible and multiplying this by 100, resulting in avalue between 0 and 100.

Table 1 presents the results of our analysis,the total inertia and complementarity ratiosfor the Randstad and its south and northwings for 1996 and 2002. The table revealsclearly that the economic profiles of citieswithin the Randstad, or its south and northwings, are becoming less differentiated.This trend towards more homogeneous eco-nomic roles for cities indicates a decreasingcomplementarity in terms of economic roles(on average, a decrease of nearly 10 per centin 6 years). The north wing of the Randstadis the most complementary region, with aratio of 6.3 in 2002. This is 29 per centhigher than in its counterpart, the southwing, and the Randstad as a whole whendefined by the 4 cities. This is 66 per centhigher than if the Randstad is defined usingall the larger cities. However, the ratio in thenorth wing is decreasing faster than in thesouth wing, but not as fast as in the Randstaddefined by the 4 largest cities.

One of the main advantages of correspon-dence analysis is that it graphically displaysassociations, thus enabling an easier interpret-ation of complex contingency tables. Here, weare interested in associating cities witheconomic activities (see Figure 2). The 2dimensions displayed represent a reasonable84.2 per cent of the total inertia. The twoaxes together indicate the origin (0.0), whichresembles the average profile of the 4 cities.This plot requires careful interpretation. If2 cities lie close together, then their economic

Table 1. Total inertia and complementarity ratios in the Randstad and spatial sub-divisions in 1996and 2002

Polycentric urban regionTotal inertia

Complementarityratio

Trend(percentage)

1996 2002 1996 2002

Randstad (N ¼ 4) 0.170 0.148 5.7 4.9 212.9Randstad (N ¼ 14) 0.545 0.493 4.2 3.8 29.5North wing (N ¼ 8) 0.489 0.440 7.0 6.3 210.0South wing (N ¼ 6) 0.263 0.245 5.3 4.9 26.8

N ¼ Number of cities included.

SYNERGY IN POLYCENTRIC URBAN REGIONS 775

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 13: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

Figure 2. Cities associated with economic activities in the Randstad (N ¼ 4), 2002 (percentage of inertiaexplained: 84.2 per cent).

Key

01 Agriculture, hunting and related

service activities

61 Water transport

22 Publishing, printing and reproduction

of recorded media

63 Supporting and auxiliary transport

activities; travel agencies

23 Manufacture of coke, refined

petroleum products and nuclear fuel

64 Post and telecommunications

24 Manufacture of chemicals and

chemical products

65 Financial intermediation,

except insurance and

pension funding

28 Manufacture of fabricated metal

products

66 Insurance and pension funding

35 Manufacture of other transport

equipment

72 Computer and related activities

40 Electricity, gas, steam and hot water

supply

75 Public administration and

defence; compulsory social

security

45 Construction

80 Education

51 Wholesale trade and commission trade

91 Activities of membership

organisations n.e.c.

55 Hotels and restaurants

92 Recreational, cultural and

sporting activities

776 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 14: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

profiles are more or less similar. The samecondition applies to the economic activities.Economic activities lying close together aremore or less similarly distributed betweenthe cities. The distances between cities andeconomic activities are more complicated,since these are not defined as chi-squareddistances. All cities influence the location ofan economic activity and, the other wayaround, all economic activities contribute tothe location of a city. In general, cities andactivities will be close to each other whenthe observed value for this pair of points inthe table is larger than expected, and the dis-tance will be large when the observed valueis less than the expected value. For reasonsof clarity, only those economic activitiesthat contribute to at least 1 per cent of thetotal inertia value are depicted. Economicactivities that do not meet this thresholdvalue are either insignificant in terms of thenumber of jobs or because the distribution ofjobs in this activity over the 4 cities issimilar to the distribution of all jobs overthese cities. This is for instance the casewith retail trade, where the expected numberof jobs in the 4 cities is more or less equalto the observed number.

In 2002, the three largest cities, Amsterdam,Rotterdam and The Hague seem to have rela-tively distinct profiles, while Utrecht has amore general and average profile and is, there-fore, located closer to the origin. Still,Utrecht’s economic profile is more similar tothat of Amsterdam than to the other cities.Out of the origin it is possible to distinguishthree axes in which a number of economicactivities are grouped that are dominated byone of the three main cities. Amsterdam has arelatively dominant position in the commercialservices sector, in particular in financial inter-mediation, Internet and communication tech-nology, publishing and printing, recreation,culture and sports, as well as hotels and restau-rants. Rotterdam, on the other hand, holds astrong position inmanufacturing and transport,undoubtedly related to its large port. Thisincludes heavy industries such as the petro-chemical and chemical industries, the manu-facturing of fabricated metal products and

of transport equipment. Not surprisingly,Rotterdam specialises in water transport andsupport and auxiliary transport activities.Finally, Rotterdam has relatively more jobsin the construction and public utilitiessectors. The Hague, the seat of government,is particularly dominant in public adminis-tration. Moreover, it has a position in agricul-ture, but this seems predominantly due to thepresence of a number of agricultural interest-groups. The Hague shares, with Amsterdamand Utrecht, a strong position in activities ofmembership organisations and post andtelecommunications. Utrecht is relativelymore associated with wholesale trade and edu-cation. In general, it seems that there is a con-siderable division of labour between the threelargest cities in the Randstad, each specialisingin either commercial services (Amsterdam),manufacturing and transport (Rotterdam) orpublic administration (The Hague). Thespread of economic activities over the fourcities is to a large extent similar to the patternobserved for 1996.

Figure 3 presents the developments in theeconomic profile of the four main cities inthe Randstad between 1996 and 2002. Theorigin here is the weighted average of theaverage economic profiles in 1996 and 2002.Figure 3 provides an accurate picture of devel-opments in the differentiation in economicprofiles (roles) of the four main cities in theRandstad. Rotterdam accounts for the largestchange in position as it became considerablyless different in economic profile given itsmove towards the average economic profile.This could be due to the number of jobs inthe manufacturing and transport sectorsdecreasing relative to other sectors. Alterna-tively, other cities may have seen a consider-able increase in the number of jobs ineconomic activities associated with them.This is for instance the case for The Haguewhich, contrary to Rotterdam, moved furtheraway from the origin, indicating increaseddifferentiation. The number of jobs in thepublic administration sector increased con-siderably in the 1996–2002 period. Similarto The Hague, Utrecht’s economic rolebecame more distinct, although trends

SYNERGY IN POLYCENTRIC URBAN REGIONS 777

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 15: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

indicate that it will converge with that ofAmsterdam. Amsterdam’s economic roleremained more or less equally distinct. Theoverall conclusions regarding synergy in theRandstad will be presented in the conclusions.

7. Conclusions

The network metaphor first entered socialsciences several decades ago. Recently, ithas also proved to be a fruitful inspirationbase for exploring the relationships betweenactors, phenomena and elements in thespatial sciences. Thus it may be consideredthe contemporary equivalent to the systemsapproach which, although still relevant, hadits heyday in the 1970s and also directedresearch in a wide variety of fields of study.In urban and regional planning, the network

metaphor has been increasingly applied todescribe interurban relationships. Analysingthe synergy concept in economic networktheory and transferring the findings to spatialphenomena such as polycentric urbanregions reveal that it is exactly these relation-ships between cities that determine whether ornot synergy is present. In particular, it is thenature of these interurban relationships thatis decisive when considering synergy. It wasfound that polycentric urban regions canindeed be more than the sum of their partsby means of co-operative and, in particular,complementary relationships and externalitiesarising from them.The results of our explorative analysis of

synergy in the Randstad are mixed. The firstsynergy mechanism, co-operation, seems tobe increasingly present in the Randstad. The

Figure 3. Development in differentiation of economic roles of Randstad cities between 1996 and 2002(percentage of inertia explained: 79.4 per cent).

778 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 16: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

bottom–up establishment of both formal andinformal frameworks for a regional co-ordina-tion of regional issues and developments hasled to more regional organising capacitythan there was before. The wide variety ofco-operating actors present in the two frame-works operating at the scale of the Randstadare even explicitly aiming for increasedsynergy. Although most partnerships at themoment do not yet have an extensive trackrecord of successes, given their relativelyrecent establishment, it appears that theseclub-type networks provide much better con-ditions for achieving horizontal synergy thanever before. Thus, it is likely that synergyhas increased.

We argued that complementarity, thesecond synergy mechanism, is of particularrelevance for polycentric urban regions. Ouranalysis here was limited to differentiationin the economic role performed by cities,which we inferred from the economic profileof cities within polycentric urban regions.The analysis revealed that the main cities inthe Randstad perform distinct roles, each ofthem specialising in either commercial ser-vices, manufacturing and transport, publicadministration, or trade and education. Atthe same time, the extent of complementarityin economic roles diminished during the1996–2002 period by almost 13 per cent,leading to less synergy at the macro level ofthe cities in the Randstad. Following thecommon divide of the Randstad into a northwing and a south wing reveals that comple-mentarity in the north wing is substantiallyhigher than in the south wing. Whether ornot the balance between both synergy mech-anisms is shifting to more or less synergy inthe Randstad remains a crucial question.Answering it requires further research intoother facets of complementarity at the microlevel of certain activities and particularplaces, the extent to which differentiation inthese activities and places is matched by aregional demand for them, as well as of theexternalities present in the co-operationframeworks.

The decline in complementarity in termsof economic profiles of cities should be

looked at in the proper perspective. Otherresearch on developments in the sectoralcomposition of cities has shown that the sec-toral specialisation of cities of all sizes hasalready been declining considerably for along time (Duranton and Puga, 2001). Fromthis viewpoint, our findings regarding thedevelopment in synergy through the mech-anism of complementarity are perhaps notsurprising, nor as negative as they mayseem. Interestingly, Duranton and Puga(2001) argue that cities are increasingly dis-tinguished by their functional specialisationrather than by their sectoral specialisation.They demonstrate that, in US cities, head-quarters and business services functions ofmanufacturing companies cluster in largercities, whereas production functions clusterin smaller cities. Similar processes arelikely to be found in the Randstad. Althoughthe spread of sectoral economic activitiesover the region is becoming more homo-geneous, it may well be that, within a sectoraleconomic activity, different functions (forinstance, front-offices versus back-offices,headquarter versus production, high-skilltop-level functions versus low-skill routinefunctions, etc.) can be found in differentcities.

Finally, we need to draw a link with thecontemporary and emerging debate on thechanging spatial organisation of urbanisedregions and in particular on the changingrelations between cities. Whereas interurbanrelationships have long been defined interms of hierarchy, it is increasinglyassumed that these vertical Christaller-liketype relationships are being replaced, or atleast supplemented with more horizontalnetwork-like relationships between citiessuch as complementarity and, at the insti-tutional level, co-operation. This new modelof spatial organisation is generally referredto as a network urban structure. From ouranalysis, it follows that the development of anetwork urban structure in urban regionsis beneficial in terms of synergy as the hori-zontal network-like relationships betweencities that build such a structure correspondexactly with the synergy mechanisms.

SYNERGY IN POLYCENTRIC URBAN REGIONS 779

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 17: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

Notes

1. The term ‘city networks’ is generally appliedto: systems of distant cities linked to eachother by some functional characteristic; and,a system of proximal cities more or lesslocated within each others’ functional hinter-lands. Polycentric urban regions are citynetworks in the latter meaning.

2. A comparable analogy between firm networksand urban networks is for instance drawn byDematteis (1991), Emanuel (1990) andCamagni (1993).

3. ‘Delta Metropolis’ is synonymous with‘Randstad’.

4. See Kloosterman and Lambregts (2001) for anearlier example of the application of corre-spondence analysis. Rather than analysinggeneral economic profiles of cities, theyfocus on patterns of convergence in businessstart-up profiles of cities in the Randstad.

References

ALBRECHTS, L. (1998) The Flemish Diamond: pre-cious gem and virgin area, European PlanningStudies, 6, pp. 411–424.

BAILEY, N. and TUROK, I. (2001) Central Scotlandas a polycentric urban region: useful planningconcept or chimera? Urban Studies, 38,pp. 697–715.

BATTEN, D. F. (1995) Network cities: creativeurban agglomerations for the 21st century,Urban Studies, 32, pp. 313–327.

BOELENS, L. (2000) Inleiding, in: L. BOELENS

(Ed.) Nederland Netwerkenland, pp. 5–12.Rotterdam: NAi Uitgevers.

BUREAU REGIO RANDSTAD (2001) Naar eenblauwgroene Deltametropool. Utrecht: BureauRegio Randstad.

BURTON, I. (1963) A restatement of the dispersedcity hypothesis, Annals of the Association ofAmerican Geographers, 63, pp. 285–289.

BUTTON, K., NIJKAMP, P. and PRIEMUS, H. (Eds)(1998) Transport Networks in Europe.Cheltenham: Edward Elgar.

CAMAGNI, R. (1993) From city hierarchy to citynetworks: reflections about an emergingparadigm, in: T. R. LAKSHMANAN andP. NIJKAMP (Eds) Structure and Change in theSpace Economy: Festschrift in honour of MartinBeckmann, pp. 66–87. Berlin: Springer Verlag.

CAMAGNI, R. and SALONE, C. (1993) Networkurban structures in northern Italy: elements fora theoretical framework, Urban Studies, 30,pp. 1053–1064.

CAPELLO, R. (2000) The city network paradigm:measuring urban network externalities, UrbanStudies, 37, pp. 1925–1945.

CAPELLO, R. and NIJKAMP, P. (1993) Measuringnetwork externalities. Paper presented at theNectar Cluster 2 Workshop, Certosa diPontignano, Siena, June.

CAPELLO, R. and RIETVELD, P. (1998) The conceptof network synergies in economic theory:policy implications, in: K. BUTTON, P.NIJKAMP and H. PRIEMUS (Eds) TransportNetworks in Europe, pp. 57–83. Cheltenham:Edward Elgar.

CAPINERI, C. and KAMANN, D. J. F. (1998)Synergy in networks: concepts, in: K. BUTTON,P. NIJKAMP and H. PRIEMUS (Eds) TransportNetworks in Europe, pp. 35–56. Cheltenham:Edward Elgar.

CLAUSEN, S.-E. (1998) Applied correspondenceanalysis: an introduction. Sage UniversityPapers Series on Quantitative Applications inthe Social Sciences, 07-121, Thousand Oaks, CA.

DELTA METROPOLIS ASSOCIATION (2000)Hoofdlijnen ontwikkelingsconcept. Delft:Vereniging Deltametropool.

DEMATTEIS, G. (1991) Sistemi locali nucleari esistemi a rete: un contributo geografico all’inter-pretazione delle dinamiche urbane, in: C. S.BERTUGLIA and A. LA BELLA (Eds) I sistemiurbani, Vol. 1: Le teorie, il sistema e le reti,pp. 417–441. Milan: Franco Angeli.

DIELEMAN, F. M. and FALUDI, A. (1998a) Poly-nucleated metropolitan regions in northwestEurope: theme of the special issue, EuropeanPlanning Studies, 6, pp. 365–377.

DIELEMAN, F. M. and FALUDI, A. (1998b)Randstad, RhineRuhr and Flemish Diamond asone polynucleated macro-region, Tijdschriftvoor Economische en Sociale Geografie, 89,pp. 320–327.

DURANTON, G. and PUGA, D. (2001) From sectoralto functional urban specialisation. DiscussionPaper 2971, Centre for Economic PolicyResearch, London.

EMANUEL, C. (1990) Polimorfismo di imprese e diterritorio: una possibile convergenza discipli-nare nellesame del caso italiano, RivistaGeografica Italiana, 1990(1), pp. 13–37.

FALUDI, A. and VALK, A. G. VAN DER (1994) Ruleand Order: Dutch Planning Doctrine in theTwentieth Century. Dordrecht: Kluwer AcademicPublishers.

GOTTMANN, J. (1961)Megalopolis, The UrbanizedNortheastern Seaboard of the United States.New York: The Twentieth Century Fund.

GREENACRE, M. J. (1993) CorrespondenceAnalysis in Practice. London: Academic PressLtd.

HOHENBERG, P. M. and LEES, L. M. (1985)The Making of Urban Europe: 1000–1950.Cambridge, MA: Harvard University Press.

KLOOSTERMAN, R. C. and LAMBREGTS, B. (2001)Clustering of economic activities in polycentric

780 EVERT MEIJERS

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 18: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

urban regions: the case of the Randstad, UrbanStudies, 38, pp. 717–732.

KNAAP, G. A. VAN DER (1994) Ruimtelijkecomplexen en schaalspanning op de stedenring.EGI-onderzoekspublikatie 19, Erasmus Univer-siteit Rotterdam.

LAAN, L. VAN DER (1998) Changing urbansystems: an empirical analysis at two spatiallevels, Regional Studies, 32, pp. 235–247.

LAMBOOY, J. (1969) City and city region in theperspective of hierarchy and complementarity,Tijdschrift voor Economische en SocialeGeografie, 60, pp. 141–154.

MEIJERS, E. and ROMEIN, A. (2003) Realizingpotential: building regional organizing capacityin polycentric urban regions, European Urbanand Regional Studies, 10, pp. 173–186.

MINISTERIE VAN DE VLAAMSE GEMEENSCHAP

(1997) Ruimtelijk Structuurplan Vlaanderen.Integrale versie. Brussel: Ministerie van deVlaamse Gemeenschap.

MINISTERIE VAN VOLKSHUISVESTING, RUIMTE-

LIJKE ORDENING EN MILIEUBEHEER (2004)Nota Ruimte, Ruimte voor Ontwikkeling, deel3: kabinetsstandpunt. The Hague: Ministerievan VROM.

MUSTERD, S. and ZELM, I. VAN (2001) Polycentri-city, households and the identity of places,Urban Studies, 38, pp. 679–696.

SCOTT, A. J. (1998) Regions and the WorldEconomy, The Coming Shape of Global

Production, Competition and Political Order.Oxford: Oxford University Press.

STEIN, C. S. (1964) A regional pattern fordispersal, Architectural Record, September,pp. 205–206.

STORPER, M. (1995) The resurgence of regionaleconomies, ten years later: the region as anexus of untraded interdependencies, EuropeanUrban and Regional Studies, 2, pp. 191–221.

STOUFFER, S. A. (1940) Intervening opportunities:a theory relating mobility and distance,American Sociological Review, 5, pp. 845–867.

ULLMAN, E. L. (1956) The role of transportationand the bases for interaction, in: W. L.THOMAS (Ed.) Man’s Role in Changing theFace of Earth, pp. 862–880, Chicago, IL: TheUniversity of Chicago Press.

WESTIN, L. and OSTHOL, A. (1994) Functional net-works, infrastructure and regional mobilization,in: L. LUNDQVIST and O. PERSSON (Eds)Northern Perspectives on European Integration,pp. 43–57. Stockholm: nordREFO.

WOUDEN, R. VAN DER and BRUIJNE, E. DE (2001)De stad in de omtrek: problemen en perspectie-ven van de vier grootstedelijke gebieden in deRandstad. The Hague: Sociaal en CultureelPlanbureau.

ZONNEVELD, W. A. M. (1991) Conceptvorming inde ruimtelijke planning: patronen en processen.Amsterdam: Planologisch en DemografischInstituut.

SYNERGY IN POLYCENTRIC URBAN REGIONS 781

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from

Page 19: Polycentric urban regions and the quest for synergy: Is a network of cities more than the sum of the parts?

at Erciyes Universitesi on May 20, 2014usj.sagepub.comDownloaded from