Cities 28 (2011) 160–170

Contents lists available at ScienceDirect

Cities

journal homepage: www.elsevier .com/locate /c i t ies

Heterogeneous development of metropolitan spatial structure: Evidence fromcommuting patterns in English and Welsh city-regions, 1981–2001

M.J. Burger a,⇑, B. de Goei b, L. van der Laan c, F.J.M. Huisman c

a Department of Applied Economics, Erasmus University Rotterdam and ERIM, P.O. Box 1738, 3000 DR Rotterdam, Netherlandsb Judge Business School and Fitzwilliam College, University of Cambridge, Storeys Way CB3 0DG, Cambridge, United Kingdomc Department of Applied Economics, Erasmus University Rotterdam, Netherlands

a r t i c l e i n f o

Article history:Received 21 September 2009Received in revised form 4 October 2010Accepted 22 November 2010Available online 24 December 2010

Keywords:City-regionsSpatial structureCommutingUrban networksEngland and Wales

0264-2751/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.cities.2010.11.006

⇑ Corresponding author. Tel.: +31 (0)10 4089579; fE-mail addresses: mburger@ese.eur.nl (M.J. Burger

(B. de Goei), lambertvanderlaan@gmail.com (L. van de(F.J.M. Huisman).

URL: http://www.mjburger.net (M.J. Burger).1 For example, city-regions characterized by the his

smaller centers (i.e., conurbations), such as the area ar

a b s t r a c t

In the contemporary literature on urban systems, it is often suggested that the conceptualization of urbansystems as monocentric spatial entities has become increasingly problematic. However, by analyzingemployment and commuting patterns in English and Welsh city-regions between 1981 and 2001, itcan be shown that not all city-regions are experiencing a shift toward a polycentric spatial structure.Although most city-regions in Southeast England and the Midlands are becoming more polycentric,the spatial structures of many city-regions in the North have not shown significant change. In fact, someare becoming more monocentric. In addition, polycentricity takes different forms, which indicates thatthe development of metropolitan spatial structure can be characterized as a heterogeneous spatialprocess.

� 2010 Elsevier Ltd. All rights reserved.

Introduction

The contemporary literature on urban systems often argues thatpolycentricity has become the dominant metropolitan form inWestern Europe. The literature claims that the monocentric model,in which a principal city offers labor demand and the surroundingterritory offers labor supply, is increasingly inaccurate (Meijers,2007). Indeed, studies have hypothesized that firms and house-holds are increasingly located outside the principal city where theycreate new centers while maintaining significant linkages with theoriginal core. In addition, past studies have conjectured that prox-imate but historically autonomous urban centers within the samecity-region are undergoing a spatial and functional integration.1

The outcome of these developments is hypothesized as a spatiallyand functionally integrated city-region with multiple centers at thesupra-local scale.

Therefore, it is argued that modern urban life is taking place in apolycentric city-region, not a monocentric city. The emphasis onpolycentricity reflects the move from territories dominated by a

ll rights reserved.

ax: +31 (0)10 4089141.), bastiaan.degoei@cantab.netr Laan), fhuisman@ese.eur.nl

torical co-location of multipleound Middlesbrough.

principal center to territories in which no center predominates(Kloosterman & Musterd, 2001; Meijers, 2007). The city-regionindicates a shift in emphasis from the principal city and its suburbsas the unit of analysis to the principal city and its wider, surround-ing territory (Davoudi, 2008; Parr, 2005).

Changes in the relationship between cities and surroundingterritories are often seen as consequences of economic and socio-cultural developments. Examples of such developments includechanges in the demography of Western populations, the changeof Western economies from manufacturing economies to serviceeconomies, and an increased demand for customized productsand services (Champion, 2001; De Goei, Burger, Van Oort, & Kitson,2010). At the same time, increasing urban diseconomies, such ascongestion, high land prices and pollution, have made the principalcity increasingly unattractive (Richardson, 1995). These develop-ments have resulted in complex and interlinked processes ofdeconcentration and restructuring (Champion, 2001).

Some scholars have argued that the size of the population andbusinesses residing in the core is dependent on the stage of urbandevelopment (Tosics, 2004; Van den Berg, Drewett, Klaasen, Rossi,& Vijverberg, 1982). Others perceive the development of spatialstructure as a path-dependent process in which the initial shapeof cities and urban systems determines spatial structure (see e.g.,Kloosterman & Lambregts, 2007; Nitsch, 2003). The focal point ofthese discussions provides a clear view of the development of spa-tial structures. Many past empirical studies on urban systems havebeen cross-sectional in assessing spatial structure at only one

3 Yet, the concept of city-region is in fact much older and dates back to the work oLösch (1944). A good overview of the development of the city-region concept is

M.J. Burger et al. / Cities 28 (2011) 160–170 161

specific point in time. More recently, scholars have focused onmodeling the dynamics of metropolitan spatial structures (e.g.,Aguilera & Mignot, 2004; De Goei et al., 2010; Green, 2008; Guth,Holz-Rau, & Maciolek, 2009; Nielsen & Hovgesen, 2005).

There is a large amount of literature on the emergence of thepolycentric city-region (e.g., Kloosterman & Musterd, 2001; Parr,2005), including an increasing number of empirical studies onpolycentric development. However, two elements in this field ofresearch remain unexamined. First, with regard to spatial hetero-geneity in urban development processes, many studies assume, of-ten implicitly, that urban systems evolve from a monocentricspatial structure toward some form of polycentric spatial structure.However, the original state differs between city-regions. Somecity-regions are predominantly monocentric; others are predomi-nantly polycentric. In reality, most city-regions fall somewhere be-tween those poles (De Goei et al., 2010). As such, the spatialstructure of city-regions does not necessarily have to change inthe same direction.2 Secondly, previous studies have often neglectedthat polycentricity may take different forms. Patterns of deconcen-tration can differ across urban systems, ranging from exchangecommuting between the original center and the surrounding terri-tory to a fully networked city-region. In the fully networked city-region, the original center and surrounding territory become indis-tinct (Schwanen, Dieleman, & Dijst, 2003; Van der Laan, 1998).

This article aims to contribute to the existing literature onmetropolitan spatial structure by analyzing the development ofcity-regions at the supra-local scale. Metropolitan spatial structurehas traditionally been assessed by examining the distribution ofinternal characteristics of cities, such as the number of residentsor jobs. However, many of the theoretical foundations for the poly-centric model are based on flows linked to the physical movementof goods, people and services (De Goei et al., 2010; Hall, 2001).Also, the position of centers within a network of cities increasinglydetermines the importance of centers (Batten, 1995; Hall, 2001).Hence, the metropolitan spatial structure is not only assessed byexamining the distribution of population or employment withina city-region but also by assessing the distribution of flows be-tween the different parts of the city-region. As such, a typologyof city-regions has been developed based on the spatial configura-tion of commuting networks (Schwanen et al., 2003; Van der Laan,1998). To accomplish this, a number of related features of urbansystems should be taken into account. Such factors include the net-work density (i.e., the degree of functional integration of the differ-ent parts of the city-region) and the openness of city-regions.

This article uses data on journey-to-work travel in 22 city-regions in England and Wales from 1981–2001. Employment andcommuting patterns are a useful source for investigating the devel-opment of metropolitan structure because labor market interac-tions between centers constitute the majority of all daily tripswithin city-regions (Clark & Kuijpers-Linde, 1994). In addition,commuting data is still the most elaborate, reliable, and relevantinteraction data available. This paper addresses the heterogeneityin metropolitan spatial structure present in the English and Welshurban system by comparing the spatial structures of city-regionsover time.

The remainder of this article is organized as follows. The nexttwo sections contain a discussion of the literature on changing ur-ban systems. In particular, the paper focuses on the level of thecity-region (supra-local scale). However, the current trend towardmega-city-regions (Hall & Pain, 2006; Hoyler, Kloosterman, &Sokol, 2008) at a higher geographical scale is beyond the scope ofthis article. In empirical setting and methodology, the data set

2 In this, it should be acknowledged that Hall and Pain (2006) specifically reportedon differences between European mega-city-regions in relation to their degree opolycentricity at the inter-regional scale.

provided by Parr (2005) and Davoudi (2008).4 For example, the city-region of Newcastle contains the principal city Newcastle

the suburbs Gateshead, North Shields and South Shields, the lower order urbancenters Chester-le-Street, Durham and Sunderland, and rural areas in Tynesdale andNorth Northumberland.

f

and research methods are introduced. Empirical results contains anoverview of the main findings. A discussion and conclusion followsin Concluding remarks and limitations.

From a monocentric city to a polycentric city-region

From cities to city-regions

Among academics and policy analysts, the debate on urbandevelopment has shifted from the city to the city-region as theprimary unit of analysis (Davoudi, 2008; Parr, 2005, 2008). Asindicated by Parr (2008, p. 3017), the increasing interest in thecity-region as a unit of analysis echoes ‘a growing appreciation thatin certain important respects the city is not a satisfactory unit ofanalysis, since many of its external interaction are with adjacent(and sometimes extensive) areas’.3

Although there are multiple spatial imaginations of the city-region, such as specific administrative units or NUTS-2 areas (Parr,2005), this research assumes city-regions to be functional urbanregions. Embedded in urban systems theory, cities and their sur-roundings are herein perceived as functionally interdependententities linked through economic flows, such as inter-firm trade,labor market interactions, and shopping (Berry, 1964; Parr,1987). In this, the dominant flows determine the delineation ofcity-regions (Davoudi, 2008). The city-region is usually perceivedas a principal city and its surrounding territory (Parr, 2005,2008). The surrounding territory consists of suburban areas, ruralvillages and, most importantly, subordinate urban centers. The dif-ferent parts of the city-region together form a functional economicspace (Davoudi, 2008).4

From monocentric to polycentric city-regions

The study of metropolitan spatial structure originates from ur-ban location theory and can be traced back to the theoretical workby Alonso (1964) on the monocentric model, which he adaptedfrom the work of Christaller (1933) on hierarchical urban systems.From a morphological point of view, the monocentric model pre-dicts a strict division of labor between the central business district(CBD) and suburban areas, where the CBD is the source of labor de-mand and the suburbs are the source of labor supply. In otherwords, the size distribution of centers in terms of employment ishighly unbalanced in a monocentric city. From a functional pointof view, the monocentric spatial system is best represented by astar-shaped pattern of interactions, where the journey-to-workflows are directed from the suburbs to the CBD (Haggett andChorley, 1967). As indicated by Parr (2008), the monocentric citymodel can be perceived as a special case of the traditional Löschiancity-region (Lösch, 1944). Like the monocentric city, the traditionalcity-region is perceived as a nodal area in which the principal cityis characterized by a deficit labor supply and the territory sur-rounding the principal city by a deficit labor demand (Parr, 1987,2005). Accordingly, journey-to-work flows are directed from thesurrounding territory to the principal city.

The monocentric model was challenged from two sides (Van derLaan, Vogelzang, & Schalke, 1998). First, the model was unable to ex-plain the emergence of employment locations in the suburbs and,hence, the emergence of sub-centers (Garreau, 1991). Likewise, at

f

,

162 M.J. Burger et al. / Cities 28 (2011) 160–170

the level of the city-region, the monocentric model has been unableto explain a shift in employment from the principal center to thesubordinate centers in surrounding territories and the resulting de-crease of commuting from the surrounding territory to the principalcity. Second, the existence of an enormous amount of ‘excess’ or‘wasteful’ commuting – the difference between the average com-muting distance projected by the monocentric density functionbased on travel-minimizing behavior and the actual required com-muting distance – made the monocentric model unrealistic (Small& Song, 1994). The most obvious reason for the ‘failure’ of the mono-centric model is the increasing percentage of journey-to-work travelfrom the principal city to the surrounding territory and between dif-ferent parts of the surrounding territory.5

Whereas the monocentric concept gave way to that of polycen-tric city-regions, the resulting metropolitan spatial structure oftenremains unclear (Van der Laan, 1998).6 For example, a polycentriccity-region can be characterized by a spatial structure whereby ex-change commuting only occurs between the original center andthe surrounding territory or by a more fragmented pattern in whichthe original principal city and subordinate centers have becomeindistinct. A more elaborate discussion of these issues is providedin the next section.

Metropolitan spatial structure

Conceptualizing the importance of centers

Every assessment of spatial structure starts off by determiningthe importance of centers. As indicated by Preston (1971), onecan make a distinction between the absolute importance of a cen-ter or its nodality and the relative importance of a center or its cen-trality. In the light of employment and journey-to-work flows, thenodality of a center can be expressed by the total number of jobs itprovides, whereas the centrality of a center can be defined by thenumber of jobs in excess of those demanded by the center’s owninhabitants. The distinction between nodality and centrality canbe traced back to the work by Christaller (1933), who argued thatif the importance of a center was determined solely on the basis ofits size, then part of its importance must be ascribed to the settle-ment itself as an agglomeration and another part to the settlementas a provider of jobs, goods and services to the surrounding terri-tory. Accordingly, it is possible to separate the external importancefrom the local importance of a city. The importance a center c in anopen system (i.e., including linkages with other systems) can thenbe decomposed7 as follows:

Cci ¼ Nc � Cco � Lc;

where Cci = the surplus importance of a center based on incomingflows from other places within the same spatial system; its internalcentrality. Nc = the absolute importance of a center; its nodality.Cce = the surplus importance of a center based on incoming flowsfrom other places outside the spatial system; its external centrality.Lc = the local importance of a center based on internal flows (peopleworking and living in the same place).

For example, when examining the importance of a center assource of jobs in a city-region, it can be argued that Nc representstotal employment in part of the city-region c, Cci represents

5 It should be noted that, over time, the monocentric model could also over-predictcommuting distance as total commuting distance could drop when more people workand live in sub-centers (Small & Song, 1994). However, this would also mean that thepredictions from the monocentric model are no longer reliable.

6 The use of polycentricity should not be confused with the concept of multicen-tricity. Multicentricity refers to the existence of multiple centers, while polycentricityemphasizes a certain balance in importance of these centers.

7 See Burger and Meijers (2010) for a more elaborate discussion of these issues.

incoming commuting in part of the city-region c from other partsof the city-region, Cce represents incoming commuting in part ofthe city-region c from places situated outside the city-region, andLc represents the number of employees in the part of the city-re-gion c that also live there.

Polycentricity, network density and openness of city-regions

Analogous to the distinction between nodality and centrality,there are two main approaches for assessing the spatial structureof city-regions (Green, 2007; Meijers, 2008). The morphological ap-proach employs the attributes or internal characteristics of centers,such as the number of jobs (e.g., Meijers, 2008; Spiekermann &Wegener, 2004). The functional approach classifies metropolitanspatial structure based on the structure of flows within spatial sys-tems (e.g., De Goei et al., 2010; Green, 2007). Both approaches lookat the balance in the distribution of importance across centers,where polycentricity tends to be associated more with a balanceddistribution of the importance of these centers (Kloosterman &Lambregts, 2001; Meijers, 2008). Whereas the morphological ap-proach assesses the balance in the size distribution or distributionof absolute importance of centers, the functional approach typi-cally looks at the balance in the distribution of functional linkagesbetween centers (Burger & Meijers, 2010; Fig. 1), where most oftenattention is paid to the distribution of intra-regional flows and,hence, internal centrality scores (e.g., Green, 2007).

For conceptual clarification, in line with graph theoretical con-ceptualizations of spatial structure (e.g., Limtanakool, Schwanen,& Dijst, 2008), we explicitly disentangle the spatial structure ofcity-regions from related aspects of the spatial organization ofcity-regions, such as the degree of network density (i.e., the extentto which the different parts of a city-region are functionally linked)and outward openness (i.e., the extent to which the city-region isconnected to other city-regions). Not including network densityin our measure of functional polycentricity was necessary becauseit is possible to encounter spatial systems that are strongly net-worked in a hierarchical organization and spatial systems that arenot networked at all, where centers are relatively equal in termsof their connectivity to other centers. If both primacy and networkdensity scores are combined, we may find that spatial systems witha highly unbalanced distribution of functional linkages but a highnetwork density would receive a similar score as spatial systemswith a highly balanced distribution of functional linkages but alow degree of network density (Burger & Meijers, 2010). A similarargument can be made with regard to the openness of a city-region.

This does not mean that the degree of network density andopenness are not important aspects of the spatial organization ofa city-region. In fact, it can be expected that the increased mobilityof households and businesses will result in increasing networkdensity and increasing openness within city-regions. Moreover, afunctional polycentric urban region is non-existent without a min-imum degree of interaction between centers (Champion, 2001;Parr, 2004).

A Typology of city-regions based on commuting

Based on the orientation of commuting patterns in city-regions8,it is possible to distinguish between four prototypical city-regions, asshown in Fig. 2 (Schwanen et al., 2003; Van der Laan, 1998). We dis-tinguished three types of journey-to-work patterns between differ-ent parts of a city-region: traditional commuting (commuting fromthe surrounding territory to the principal city), exchange commuting

8 It should again be stressed that we look here only at the directionality of thenkages between the different parts of the city-region and not at the strength of thenkages between the different parts of the city-region.

lili

Morphologically Monocentric Morphologically Polycentric

Functionally Monocentric Functionally Polycentric

Fig. 1. Morphological polycentricity versus functional polycentricity.

Polycentric Criss-Cross Polycentric Decentralized

Monocentric Polycentric Exchange

Fig. 2. Functional typology of the spatial structure of city-regions.

M.J. Burger et al. / Cities 28 (2011) 160–170 163

(commuting from the principal city to the surrounding territory) andcriss-cross commuting (commuting between the different parts ofthe surrounding territory).

The monocentric city-region is characterized by a low degree ofexchange and criss-cross commuting because the majority of com-muters living in the surrounding territory travel to the principalcity, while the surrounding territory attracts very few commuters.In a polycentric exchange city-region, commuting has becomereciprocal; commuting is no longer directed solely from the

surrounding territory to the principal city but also from the princi-pal city to the surrounding territory. In a polycentric criss-cross ur-ban system, the different parts of the surrounding territory havebecome more dominant because they attract commuters fromother parts of the surrounding territory. In this, parts of the sur-rounding territory have become complementary to the principalcity and are increasingly important as centers within the city-re-gion. Commuting flows decentralize as the number of workerscommuting between the different parts of the surrounding terri-tory and bypassing the (former) principal city increases. Yet, thedegree of exchange commuting remains low. Finally, a decentral-ized polycentric city-regionis characterized by a multi-orientedcommuting pattern in which there is no longer a dominant center.In a polycentric decentralized system, there is a large amount ofboth criss-cross and exchange commuting. Two types of city-re-gions fit this classification: (1) a formerly monocentric city-regionin which employment has spread from the urban core to the urbanfringe (i.e., ‘edgeless cities’) (Lang, 2003) and (2) city-regions char-acterized by the historical co-location of multiple smaller centers(i.e., conurbations). In such urban systems, the amount of tradi-tional commuting may become very small.

Empirical setting and methodology

Previous research on commuting and spatial structure in England andWales

Despite the growing amount of theoretical literature on chang-ing urban systems, only a small number of recent empirical studieshave quantitatively assessed the development toward polycentriccity-regions in England and Wales. Studies by Coombes, Charles,Raybould, and Wymer (2006), Nielsen and Hovgesen (2008), Green(2008) and De Goei, Burger, Van Oort, and Kitson (2008, 2010) haveanalyzed the dynamics of spatial structure in England and Walesempirically by using commuting data, whereas most work on

164 M.J. Burger et al. / Cities 28 (2011) 160–170

English and Welsh urban networks has been done at the inter-re-gional scale (i.e., between supra-local territories).

Analyzing commuting data, Green (2008) found that Englandand Wales became more connected at the inter-regional scale be-tween 1981 and 2001. In addition, Green (2008) concluded thatthere is an increasing tendency for people to live and work at theurban periphery, where travel patterns have become increasinglydiffused. On a similar note, Nielsen and Hovgesen (2008) examinedthe development of nationwide commuting flows in England andWales between 1991 and 2001. They found an increase in the aver-age commuting distance during this period, which is exemplified byan increasing connectivity between rural areas and main centers.

Focusing on commuting networks in the Greater Southeast, DeGoei et al. (2008), De Goei et al. (2010) concluded that the mega-city-region still could not be regarded as a functional polycentricurban region. However, they found some evidence for spatial inte-gration at the supra-local scale (i.e., the city-region), as well as adecentralization of the system at the inter-regional scale, becausethe hub function of London is decreasing in significance. On a sim-ilar note, Coombes et al. (2006) found some evidence for increasinglinkage formation between the principal cities of the city-regionsin the East Midlands urban network.

However, the empirical literature on spatial structure in gen-eral, and those studies in particular, pay limited attention to thespatial heterogeneity present between different city-regions. Theremainder of this article contributes to the existing empirical liter-ature on urban systems by jointly addressing the dynamics, spatialheterogeneity and the specific polycentric pattern of city-regionsin England and Wales.

Commuting and city-regions in England and Wales

The present study used journey-to-work data between localauthority districts from 1981 and 2001 to analyze the developmentof metropolitan spatial structures in England and Wales. Thesedata were obtained from the Special Workplace Statistics (Set C)in the British Census.9 The Census Interaction Data Service (CIDS)1991/2001 common geography was used to avoid potential prob-lems with the changes of district-boundaries over the past 20 years(Boyle & Feng, 2002). A subdivision of city-regions is made using thecommon geography. In this, we used the classification created byCoombes (2000), which comprises 43 British city-regions. We de-fined city-regions by functional linkages and areal associations basedon the 2001 Census. In this, ‘the city-regions are delineated on thebasis of commuting data through an algorithm that optimises theboundaries on the basis of a size of employment criterion and a min-imum threshold of self-containment of flows to workplaces’ (Robson,Barr, Lymperopoulo, Ress, & Coombes, 2006, p. 8). Accordingly, theclassified city-regions are to a large extent self-contained becausemost people (>80%) who work in city-regions also live there.

The delineation of Coombes (2000) is non-nodal and exhaustivebecause there may be several employment centers within eachcity-region, and every area in Great Britain is allocated to a city-region. Compared to other classifications of city-regions in Englandand Wales (see Davoudi, 2008; Robson et al., 2006), an obviousadvantage of the Coombes classification is that it does not assumea priori the existence of only one main center within a city-region.Hence, the bottom-up approach of delineating city-regions is use-ful for exploring polycentric patterns in a wider region (Davoudi,2008). Based on the classification of city-regions by Coombes(2000), we selected the 22 largest English and Welsh city-regions,

9 Census output is crown copyright and is reproduced with the permission of thecontroller of HMSO and the Queen’s printer for Scotland. Sources: 1981 census:Special workplace statistics (Set C) and 2001 census: Special workplace statistics(level 1).

whereby the more rural areas are excluded because these areas aretypically not metropolitan (for a similar selection, see Champion &Coombes, 2007). An overview of the city-regions included in theanalysis is presented in Fig. 3.

Quantifying the spatial structure of city-regions

The degree to which incoming commuting and employment iscentralized within city-regions is estimated to assess the develop-ment of metropolitan spatial structure of English and Welshcity-regions between 1981 and 2001. To assess the degree of mor-phological and functional polycentricity in a city-region, we used aprimacy index (Adolphson, 2009; Van der Laan, 1998). The mor-phological primacy index is calculated as the ratio of employmentin the largest center (i.e., the principal city) and the total employ-ment in the city-region and is thus based on the balance in the dis-tribution of nodality scores. The functional primacy index iscalculated as the ratio of incoming commuting into the largest cen-ter originating from the city-region and the total incoming com-muting originating from the city-region and is thus based on thebalance in the distribution of internal centrality scores. A city-re-gion is considered morphological monocentric if its employmentis highly concentrated in one (principal) city. A city-region is con-sidered functional monocentric if most commuting flows originat-ing from other parts of the city-region are directed at the principalcity (and no flows are directed from the principal city to the sur-rounding territory). The larger the degree of morphological andfunctional primacy is in a city-region, the lower the degree of mor-phological and functional polycentricity in a city-region.

In addition to the summary measures of polycentricity, we lookat specific patterns of polycentricity, which correspond to thetypology of polycentric spatial structures introduced in the previ-ous section. In this, we make a distinction between the monocen-tric and three types of polycentric city-region: the exchange, thecriss-cross, and the decentralized city-region. This typology isbased on (1) the degree of exchange commuting, or the degree towhich commuters living in the principal city are oriented towardthe surrounding territory and (2) the degree of criss-cross com-muting, or the degree of commuting between different parts ofthe surrounding territory. We compared the degree of exchangeand criss-cross commuting with the degree of traditional commut-ing, where commuting is defined as journey-to-work travels be-tween the area of residence and another area of employment.Hence, we excluded the journey-to-work travel of people whowork and live in the same area. A more technical description ofthese measures is provided in Table 1.

Building on the discussion in the previous sections, we describetwo additional features of city-regions. The degree of network den-sity reflects the extent to which different parts of the city-regionare networked or functionally interdependent and can be mea-sured as the ratio of the actual connections between the differentparts of the city-region compared to the total potential connectionsbetween the different parts of the city-region (Green, 2007). Here,the total potential connections between the different parts of acity-region are defined as the total number of employees in acity-region. A generally low ratio of the sum of internal centralityscores by different parts of the city-region compared to the num-ber of employees within a city-region indicates a low level of net-work density. Likewise, the openness of a city-region can bedefined as the ratio of the number of employees in a city-regionresiding in other city-regions compared to the total employmentin the city-region (Patuelli, Reggiani, Nijkamp, & Bade, 2009).10

10 Alternatively, this measure can be conceptualized as the ratio of the sum ofexternal centrality scores of the different parts of the city-region divided by the sumof nodality scores of the different parts of the city-region.

Fig. 3. Selected 22 English and Welsh city-regions.

Table 1Orientation of commuting and four types of urban systems.

Exchange commuting

Low High

Crsiss-cross commuting High Criss-cross DecentralizedLow Monocentric Exchange

Notes:Exchange commuting index: IC(PC ? ST)/[ IC(PC ? ST) + IC(ST ? PC)].IC(PC ? ST) = incoming commuting in the surrounding territory emerging from theprincipal city.IC(ST ? PC) = incoming commuting in the principal city emerging from its sur-rounding territory.Criss-cross commuting index: IC(ST ? ST)/[IC(ST ? ST) + IC(ST ? PC)].IC(ST ? ST) = commuting between different parts of the surrounding territory.IC(ST ? PC) = incoming commuting in the principal city emerging from its sur-rounding territory.

11 Central London as principal city is here defined as the Camden, City of LondonCity of Westminster, Islington, Kensington and Chelsea, Lambeth and Southwarkwhich together form the Central London partnership.

12 However, part of the variation between city-regions can be explained by thedifferent areal units used to characterize the different parts of the surroundingterritory.

M.J. Burger et al. / Cities 28 (2011) 160–170 165

Empirical results

Spatial organization of English and Welsh city-regions

Table 2 indicates the spatial organization of the 22 selected city-regions in England and Wales in 2001 on the basis of employmentand commuting trips. Based on the morphological and functionalprimacy indices, the majority of city-regions exhibit some extentof a functional and morphological polycentric pattern. On average,only 36.7% of the incoming commuting flows to the city-region and

28.8% of employees concentrate in the principal city. Middlesb-rough, Preston, Reading and Southampton are the most relativelypolycentric city-regions in England and Wales. Other city-regions,such as Nottingham, Leeds and Leicester, can still be characterizedas relatively monocentric. Overall, there is a strong correlation(0.77) between the degree of functional and morphological poly-centricity (see also Table 3); city-regions that are relatively func-tional polycentric are also relatively morphological polycentric.

Now we turn to related aspects of the spatial organization ofcity-regions: the degree of network density and outward openness.Table 2 shows that on average, 28% of the employees in a city-region commute between different parts of the city-region, whileapproximately 10% of the employees live outside the city-regionin which they work. However, there are considerable differencesbetween city-regions, especially with regard to the degree ofnetwork density. London11 appears to be the most networkedcity-region, with over half of its employees commuting betweenthe different parts of the city-region.12 With respect to the outwardopenness of city-regions, (unsurprisingly) the geographically iso-lated city-regions (Cardiff, Middlesbrough, Newcastle and Plymouth)are the most closed, while smaller city-regions located in between

,

Table 2Spatial organization of English and Welsh city-regions (2001).

City-region Functionalprimacy

Morphologicalprimacy

Networkdensity

Outwardopenness

Southampton 17.8 15.8 31.1 8.4Reading 19.8 11.9 29.0 18.7Preston 20.6 13.0 28.4 7.2Plymouth 24.4 24.1 19.8 2.7Middlesbrough 25.0 16.9 31.6 3.9Brighton 26.7 31.5 22.4 8.0Newcastle 33.3 22.2 37.1 4.2Manchester 34.1 19.2 34.5 9.7Birmingham 35.5 35.2 30.5 11.2Cardiff 35.8 29.8 29.1 4.3London 38.2 23.0 51.4 6.2Bristol 39.5 28.0 24.6 7.2Stoke 39.5 26.9 25.3 12.7Portsmouth 40.4 28.5 22.2 11.2Sheffield 40.8 39.4 20.4 7.9Northampton 44.4 43.7 18.6 18.7Norwich 45.9 22.0 28.0 5.5Liverpool 46.2 31.7 28.3 10.6Coventry 47.0 43.5 21.7 19.9Leeds 49.0 50.8 13.7 14.0Nottingham 49.5 35.5 38.7 11.3Leicester 53.1 41.6 34.0 12.4

Mean 36.7 28.8 28.2 9.8Standard deviation 10.3 10.7 8.2 4.9

Note: sorted on functional primacy.

Table 3Correlation matrix of the different dimensions of the spatial organization of city-regions.

(1) (2) (3) (4)

(1) Functional primacy 1.00(2) Morphological primacy 0.77 1.00(3) Network density �0.07 �0.43 1.00(4) Outward openness 0.34 0.45 �0.31 1.00

166 M.J. Burger et al. / Cities 28 (2011) 160–170

larger city-regions (Coventry, Northampton, Reading) are the mostopen.

Concerning the relationship between the different aspects ofthe spatial organization of city-regions (see Table 3), there is afairly significant negative correlation between the degree of mor-phological primacy and network density. This means that a higherdegree of integration between parts of the city-region character-izes morphological polycentric city-regions. However, relation-ships between the degree of functional primacy and networkdensity are virtually non-existent. Therefore, functional polycen-tric city-regions are not more networked than functional monocen-tric city-regions, which support the decision not to combine thesetwo measures into one index.

Relatively polycentric city-regions are less open than relativelymonocentric city-regions. As indicated by Burger and Meijers(2010), this occurs because monocentric city-regions often have adisproportionately large center with a relative overrepresentationof higher-order functions. Because it is well known that higher-ranked employees (in terms of education and income) are willingto commute longer distances to work (Schwanen & Dijst, 2002),these city-regions are more likely to attract employees residingoutside the city-region.13

13 Of course, the assessment of mono/polycentricity is to some extent dependent onthe territorial delineation used as relations outside the region are not taken intoaccount here. This is in geography better known as the Modifiable Areal Unit Problem(MAUP) (Burger, Van Oort, & Van der Knaap, 2010; Openshaw & Taylor, 1979).However, given that the city-regions in our study are to a large extent self-sufficientin terms of employment, we expect that this problem is limited in our analysis.

Yet, polycentric spatial structures differ across city-regions.Fig. 4 indicates the presence of specific polycentric patterns for dif-ferent English and Welsh city-regions. The demarcation lines rep-resent the point at which the volume of criss-cross commuting isequal to the volume of traditional commuting and the point atwhich the volume of exchange commuting is equal to the volumeof traditional commuting. The values on the horizontal axis indi-cate the ratio of the volume of exchange commuting comparedto the volume of traditional commuting, while the values on thevertical axis indicate the ratio of the volume of criss-cross com-muting compared to the volume of traditional commuting. Accord-ing to this figure, the monocentric model and decentralizedpolycentric model prevail.

With respect to the indicated geographical differences inmetropolitan spatial structure, the city-regions in the Notting-ham–Sheffield–Leicester triangle have a relatively monocentricstructure. Bristol, Coventry and Stoke represent more of an ex-change-commuting model. The city-regions of Brighton, Mid-dlesbrough, Plymouth, Preston, Reading and Southampton fallbetween a criss-cross and decentralized spatial structure. In thelatter three city-regions, the volume of criss-cross commuting isover three times the volume of traditional commuting. The largecity-regions around London, Birmingham, Newcastle, Cardiff, andManchester fall between a monocentric and criss-cross spatialstructure. Most strikingly, many city-regions in Southeast Englandare characterized by a relatively high level of decentralization,while the larger city-regions in North England and the Midlandstend to have the most monocentric spatial structures.

The dynamics of the spatial structure of city-regions, 1981–2001

This section discusses the changes in the spatial organization of22 English and Welsh city-regions over the last two decades of the20th century. Table 4 shows the change in spatial organization be-tween 1981 and 2001 in terms of percentage points (pp). A numberof conclusions can be drawn from the data. First, there is a cleartendency toward polycentric spatial structures, exemplified by ageneral decrease of morphological and functional primacy. In addi-tion, city-regions have become more networked and open.

Second, with respect to geographical differences, the pace ofpolycentric development has generally been faster in city-regionsin the Midlands and South England than in city-regions inNorth England and Wales. Bristol, Coventry, Leicester, London,Portsmouth, and Stoke have shown the largest polycentric devel-opment, closely followed by Birmingham, Norwich and Plymouth.London and Portsmouth show a much greater tendency toward afunctional polycentric structure than toward a morphological poly-centric structure.14 On the contrary, Leeds and Nottingham have be-come more monocentric, while the metropolitan spatial structures ofother city-regions (Cardiff, Middlesbrough, and Sheffield) haveshown only small changes. In general, there is a significant correla-tion of 0.84 between change in functional primacy and the changein morphological primacy (see Table 5).

Yet city-regions that have become relatively more polycentrichave not become relatively more networked; Middlesbrough andCardiff have shown a much higher degree of integration thanLondon or Portsmouth. There is also a significant positive correlationof about 0.55 between an increase in functional or morphologicalprimacy and an increase in network density. Although this might ap-pear counterintuitive, it can be explained that a relative decrease injourney-to-work trips from the surrounding territory to the princi-pal city coincides with a relative increase of journey-to-work trips

4 Yet, overall, there is a strong correlation of 0.84 between the change in functionalrimacy and the change in morphological primacy.

1

p

Fig. 4. Functional spatial structures in England and Wales (2001).

Table 4Changes in spatial organization of English and Welsh city-regions in percentagepoints (1981–2001).

City-region Functionalprimacy, D

Morphologicalprimacy, D

Networkdensity, D

Outwardopenness, D

London �13.7 �2.6 0.7 1.6Bristol �12.9 �6.6 2.6 2.8Portsmouth �12.5 �5.1 0.2 2.7Coventry �11.0 �7.2 3.0 7.9Leicester �11.0 �8.3 2.5 3.8Stoke �10.3 �6.9 2.4 3.5Norwich �9.9 �4.0 3.4 1.7Plymouth �9.4 �6.3 5.3 1.0Birmingham �8.3 �6.1 2.2 4.0Manchester �7.4 �3.3 3.6 4.6Southampton �7.1 �5.0 4.0 2.9Liverpool �5.3 �4.4 �0.9 2.8Reading �5.3 �2.6 2.7 7.8Preston �4.6 �0.3 5.4 3.0Newcastle �4.4 �1.9 5.7 1.8Brighton �4.1 �3.3 2.9 1.0Sheffield �4.0 �1.3 3.9 3.1Cardiff �1.7 2.3 7.1 1.9Northampton �1.7 �0.1 5.3 6.2Middlesbrough �0.6 0.7 7.9 �2.3Nottingham 2.0 0.2 3.1 4.3Leeds 7.1 3.6 4.9 5.4

Mean �6.2 �3.1 3.5 3.2Standard deviation 5.2 3.2 2.1 2.3

Note: sorted on D functional primacy.

Table 5Correlation matrix of changes in the different dimensions of the spatial organizationof city-regions.

(1) (2) (3) (4)

(1) D Functional primacy 1.00(2) D Morphological primacy 0.84 1.00(3) D Network density 0.53 0.56 1.00(4) D Outward openness 0.04 �0.09 �0.26 1.00

M.J. Burger et al. / Cities 28 (2011) 160–170 167

within the different parts of the surrounding territory. The correla-tion between changes in outward openness and changes in primacywas insignificant.

Turning to the development of specific spatial structures, Fig. 5shows that most city-regions are developing a more decentralizedpattern, whereas city-regions in South England and the Midlandstend to do this at a faster pace than city-regions in North Englandand Wales. In Cardiff and Northampton, only some exchange

commuting has developed; in Middlesbrough, only criss-crosscommuting has developed.

These findings are supported by the shift in the relative share ofthe different types of journey-to-work trips in city-regions overtime (see Table 6). According to this table, many city-regions expe-rienced a large decrease in the share of journey-to-work trips with-in the principal city (column 1) and a small decrease in the relativeshare of journey-to-work trips within the different parts of the sur-rounding territory (column 2). Concerning the relationship be-tween the principal city and its surrounding territory, column 3shows a decrease in the relative number of journey-to-work tripsfrom the surrounding territory to the principal city (ST ? PC; col-umn 3; traditional commuting) and an increase of the flows fromthe principal city to the surrounding territory (PC ? ST; column4; exchange commuting). The data lead to the conclusion thatthe surrounding territory has become relatively more important.The latter point is also strengthened by the fact that the share ofcommuting between the different parts of the surrounding terri-tory (ST ? ST; column 5; criss-cross commuting) has increased.Overall, the surrounding territory became a more common termi-nus for journey-to-work trips.

City-regions that have experienced a considerable shift towarda decentralized metropolitan spatial structure are often the city-re-gions that have also experienced a relatively large decrease in theshare of principal city residents employed in the principal city(Table 3; within PC; column 1) or a large decrease in the share of

Fig. 5. Change in functional spatial structure in England and Wales (1981–2001).

Table 6Shift in relative share of the different types of journey-to-work trips in percentagepoints (1981–2001; italicised cells more than 3% points change).

City-region 1 2 3 4 5Within PC Within ST ST ? PC PC ? ST ST ? ST

Birmingham �4.7 2.5 �1.6 1.2 2.5Brighton �2.9 0.0 0.0 0.5 2.4Bristol �5.0 2.4 �1.8 0.7 3.7Cardiff 0.1 �7.2 2.2 1.7 3.2Coventry �6.3 3.3 �0.6 1.9 1.7Leeds �0.5 �4.5 3.0 0.6 1.3Leicester �6.2 3.7 �2.1 0.8 3.8Liverpool �2.6 3.5 �2.0 �0.2 1.2London �0.8 0.1 �6.7 0.4 7.0Manchester �2.4 �1.2 �1.1 �0.1 4.7Middlesbrough �1.5 �6.4 1.8 �0.5 6.6Newcastle �2.2 �3.5 0.5 0.7 4.4Northampton �1.5 �3.9 2.1 1.5 1.7Norwich �3.2 �0.2 �0.9 0.5 3.8Nottingham �2.4 �0.7 2.2 �0.4 1.2Plymouth �6.2 0.9 �0.1 1.4 4.0Portsmouth �2.2 2.1 �2.7 1.3 1.5Preston �0.5 �4.9 0.1 0.6 4.7Reading �1.5 �1.2 �0.9 0.6 3.0Sheffield �1.9 �1.9 0.9 1.1 1.8Southampton �3.8 �0.1 �1.2 1.1 4.1Stoke �5.4 3.0 �1.4 1.7 2.1

Mean �2.9 �0.6 �0.5 0.8 3.2Standard deviation 2.0 3.3 2.1 0.7 1.6

Within PC = within principal city, within ST = within surrounding territory,ST ? PC = from surrounding territory to principal city, PC ? ST = from principal cityto surrounding territory, ST ? ST = between different surrounding territories.

168 M.J. Burger et al. / Cities 28 (2011) 160–170

traditional commuting (Table 5, column 3), as in the case ofLondon. With respect to the surrounding territories, we find thatcity-regions in which the surrounding territory has experiencedan increase or a small decrease in the share of journey-to-worktrips within the surrounding territory (Table 3; within ST; column2) have decentralized the most. This indicates that increasing labor

demand within the surrounding territory coincides with a rise ofcriss-cross and exchange commuting and a decrease of traditionalcommuting.

Yet the amount of decentralization is, on average, still rathermodest, and the spatial structure of city-regions is generallychanging at a low pace. Given the time period of 20 years under re-view in this research, it could be that radical changes in the spatialstructure of English and Welsh city-regions did not occur. How-ever, there are also considerable differences between city-regionsbecause their spatial structures can move in different directionsand at different paces. The gray-marked cells in Table 5 indicatewhich city-regions showed the largest changes. Coventry, Leicesterand Stoke experienced large decreases in the share of the commut-ing within the principal city and large increases in commutingwithin the surrounding territory. In Cardiff, Middlesbrough, Leeds,Newcastle, Northampton and Preston, the share of the commutingwithin the surrounding territory decreased considerably. London,in particular, had a relatively large decrease in traditionalcommuting and a large increase in criss-cross commuting. Criss-cross commuting also significantly increased in the Middlesb-rough, Newcastle and Southampton city-regions.

It is clear that the city-regions that have transitioned to a morefunctional polycentric spatial structure are characterized by a de-crease in traditional commuting and an increase in exchange andcriss-cross commuting. The much greater tendency of Londonand Portsmouth toward a functional polycentric structure than amorphological polycentric structure can easily be explained bythe fact that the decrease in the share of journey-to-work tripswithin the principal city has been limited in these city-regions.

Concluding remarks and limitations

The urban geography and planning literature has often assumedthat all modern metropolitan systems are becoming polycentric. Atthe same time, the heterogeneity of the development of city-regions has not been the main object of various studies. Our anal-ysis of commuting and employment data shows that, although

15 Of course, the delimitation is to some extent also dependent on the delimitationalgorithm and procedure that is used (Cörvers, Hensen, & Bongaerts, 2009).

M.J. Burger et al. / Cities 28 (2011) 160–170 169

many city-regions are polycentric, the spatial structure differs con-siderably across city-regions. A related question is which kind ofspatial structure prevails in specific regions. During the period1981–2001, decentralization in the spatial structure occurredacross city-regions. These results are in line with the findings ofDe Goei et al. (2010) for the Greater Southeast, which also observeddecentralization at the supra-local scale, exemplified by a rela-tive increase in exchange and criss-cross commuting. Similardecentralization of employment and population at the supra-localscale has been observed in Denmark (Nielsen & Hovgesen, 2005),France (Aguilera & Mignot, 2004), and Germany (Guth et al., 2009).

Nevertheless, the extent of change in the overall spatial patternis modest. In addition, the results show a large spatial differentia-tion. It was found that not all city-regions are moving in the samedirection. Whereas the majority of city-regions in SoutheastEngland and the Midlands are becoming more polycentric, thespatial structure of many city-regions in North England is changingslowly or even becoming more monocentric. Having observedthese different trends, the question remains why some city-regionsbecome polycentric in form and other city-regions do not. Furtherempirical research should address this question in detail by quan-titatively linking the spatial heterogeneity in the dynamics ofmetropolitan spatial structure with the heterogeneity in the initialshape, economic and socio-cultural developments and local and re-gional land use policies across city-regions.

Further research can use different approaches (De Goei et al.,2010). First, more detail on the relationship between changingmetropolitan spatial structure and the increasing flexibility andmobility of firms is needed. In particular, the link betweenadvancements in transport, ICT, a developing service economy,and changes in metropolitan spatial structure deserves furtheranalysis. A starting point here would be the research of Iaonnides,Overman, and Schmidheiny (2008), which has shown that ICTweakens agglomeration forces and provides incentives to relocateeconomic activities to smaller centers. Particular attention shouldbe paid to how changes in the balance between agglomerationeconomies and diseconomies (e.g., pollution, crime) triggerchanges in the spatial organization of metropolitan areas andhow a group of functionally linked centers share agglomerationeconomies (Meijers & Burger, 2010).

Secondly, the relationship between the changing metropolitanspatial structure and the increasing flexibility and mobility ofhouseholds can be further examined. Advances in transport andICT make households not only more mobile, but they can alsochange residential preferences. Advances in transport and ICT al-low for the potential of changes in the demography of developedeconomies and the life styles of people and can thus affect metro-politan spatial structure (Champion, 2001). Such demographicdevelopments include the rise of two-earner and single personhouseholds, the increasing number of working women, higher lifeexpectancy, and lower fertility (Hall & White, 1995). These demo-graphic developments have changed the residential preferences oflarge groups of people, causing changes in the spatial organizationof urban systems (Van Ham, 2002). Thirdly, the effect of local andregional policies on metropolitan spatial structure can be analyzedfurther. This includes policies that have the explicit intention tochange metropolitan spatial structure and policies that uninten-tionally accomplish this task. Likely, there are multiple drivers be-hind the changing spatial organization of city-regions. However,these drivers may differ across various locations. In addition, theoriginal shape of city-regions also influences changes in their spa-tial organization. Accordingly, it will be important to distinguishbetween city-regions and different determinants of metropolitanspatial structure.

Although commuting constitutes one of the most importanteconomic interactions within a city-region, a significant limitation

of this study is that it only looked at commuting and employmentdata. Several authors have pointed out that journey-to-work travelis not a perfect indicator for all economic interactions within a ter-ritory and should be used alongside other indicators to gain a real-istic insight into the interdependence of places and structure inurban systems (Lambregts, Kloosterman, & Van der Werff, 2005;Parr & Hewings, 2007). Accordingly, policymakers should be care-ful drawing inferences solely based on this study. Places are notonly functionally connected through labor market relations butalso through trade, capital movements, leisure trips, and shoppingtrips (Lambooy, 1998; Parr, 2005). Admittedly, commuting tripsare often the only type of data available and may provide a surro-gate representation of other types of spatial interaction (Parr,2005). Yet the degree to which commuting relations are a proxyfor other types of spatial interaction is far from clear and shouldbe addressed in future research on metropolitan spatial structure.In particular, it is interesting to examine whether polycentric city-regions can still be regarded as functional polycentric when othertypes of economic interaction are taken into consideration. A sim-ilar point can be made with regard to morphological polycentricityand the internal characteristics of centers taken into consideration.

The evaluation of spatial structure is scale-dependent and con-tingent on the choice of analytical framework (European, national,regional, local) and the measurement of functional spaces that aredependent on the type of economic interaction (Hall & Pain, 2006;Parr, 2008).15 In our analysis, we evaluated the spatial structure ofcity-regions delimited on the basis of commuting and employmentdata. However, some urban functions, such as centers of commandand the supply of specialized goods and services, have a larger geo-graphical scope concentrated in a few major cities (Parr, 2008). Inthis light, the city-regions of Brighton, London, Northampton, Read-ing, Southampton and Portsmouth studied in this article can be con-sidered secondary city-regions within some primary city-region at ahigher geographical scale, with London as the principal center.

Lambregts et al. (2005, p. 32) rightly remarked that ‘polycen-tricity is up to certain extent in the eye of the beholder’. Preferably,indicators suggesting that a territory is polycentric should beweighed against indicators that suggest otherwise, taking into ac-count the spatial scale at which polycentricity is studied. Only inthis fashion will it be possible to completely grasp how territoriesare spatially organized and how this spatial organization changesover time. Thorough understanding of these processes is necessaryto explain the dynamics of spatial structure and to inform policy.

References

Adolphson, M. (2009). Estimating a polycentric urban structure. Case study: Urbanchanges in the Stockholm region. Journal of Urban Planning and Development,135, 19–30.

Aguilera, A., & Mignot, D. (2004). Urban sprawl, polycentrism and commuting. Acomparison of seven French urban areas. Urban Public Economics Review, 1,93–113.

Alonso, W. (1964). Location and land use. Cambridge, MA: Harvard University Press.Batten, D. F. (1995). Network cities: Creative urban agglomeration for the 21st

century. Urban Studies, 32, 313–327.Berry, B. J. L. (1964). Cities as systems within systems of cities. Papers of the Regional

Science Association, 13, 147–163.Boyle, P., & Feng, Z. (2002). A method for integrating the 1981 and 1991 British

census interaction data. Computers, Environment and Urban Systems, 26,241–256.

Burger, M. J., & Meijers, E. J. (2010) Form follows function? Linking morphologicaland functional polycentricity. Working paper, GaWC research bulletin 344.

Burger, M. J., Van Oort, F. G., & Van der Knaap, G. A. (2010). A treatise on the scale-dependence of agglomeration externalities and the MAUP. Scienze Regionali, 9,19–40.

Champion, A. G. (2001). A changing demographic regime and evolving polycentricurban regions – consequences for the size, composition and distribution of citypopulations. Urban Studies, 38, 657–677.

170 M.J. Burger et al. / Cities 28 (2011) 160–170

Champion, T., & Coombes, M. (2007). Using the 2001 census to study humancapital movements affecting Britain’s larger cities: Insights and issues. Journal ofthe Royal Statistical Society A (Statistics in Society), 170, 447–467.

Christaller, W. (1933). Die Zentralen Orte in Süddeutschland. Jena: Gustav Fischer.Clark, W. A. V., & Kuijpers-Linde, M. (1994). Commuting in restructuring urban

regions. Urban Studies, 31, 465–483.Coombes, M. (2000). Defining locality boundaries with synthetic data. Environment

and Planning A, 32, 1499–1518.Coombes, M. G., Charles, D. R., Raybould, S. R., & Wymer, C. (2006). City regions and

polycentricity: The East Midlands urban network, East Midland DevelopmentAgency, Nottingham.

Cörvers, F., Hensen, M., & Bongaerts, D. (2009). The delimitation and coherence offunctional and administrative regions. Regional Studies, 43, 19–31.

Davoudi, S. (2008). Conceptions of the city-region: A critical review. Urban Designand Planning, 16, 51–60.

De Goei, B., Burger, M. J., Van Oort, F. G., & Kitson, M. (2008). Testing the super-region. Town and Country Planning, 77, 458–464.

De Goei, B., Burger, M. J., Van Oort, F. G., & Kitson, M. (2010). Functionalpolycentrism and urban network development in the Greater South East UK:Evidence from commuting patterns, 1981-2001. Regional Studies, 44,1149–1170.

Garreau, J. (1991). Edge city: Life on the new frontier. New York: Doubleday.Green, N. (2007). Functional polycentricity: A formal definition in terms of social

network analysis. Urban Studies, 44, 2077–2103.Green, N. (2008). City-states and the spatial in-between. Town and Country Planning,

223–231.Guth, D., Holz-Rau, C., & Maciolek, M. (2009) Employment suburbanization and

commutertraffic in German city regions. Working paper, TU Dortmund.Haggett, P., & Chorley, R. J. (1967). Network Analysis in Geography. London: Edward

Arnold.Hall, P. (2001). Global city regions in the twenty first century. In A. J. Scott (Ed.),

Global city regions: Trends, theory, policy (pp. 59–77). Oxford, UK: OxfordUniversity Press.

Hall, P., & Pain, K. (2006). The polycentric metropolis: Learning from mega-city regionsin Europe. London: Earthscan.

Hall, R., & White, P. (1995). Europe’s population: Towards the next century. London:University College London Press.

Hoyler, M., Kloosterman, R. C., & Sokol, M. (2008). Polycentric puzzles – Emergingmega-city regions seen through the lens of advanced producer services.Regional Studies, 42, 1055–1064.

Iaonnides, Y. M., Overman, H. G., Rossi-Hansberg, E., Schmidheiny, K. (2008). Theeffect of information and communication technologies on urban structure.Economic Policy, April 2008, 201–242.

Kloosterman, R. C., & Lambregts, B. (2001). Clustering of economic activities inpolycentric urban regions: The case of the Randstad. Urban Studies, 38, 717–732.

Kloosterman, R. C., & Lambregts, B. (2007). Between accumulation andconcentration of capital: Toward a framework for comparing long-termtrajectories of urban systems. Urban Geography, 28, 54–73.

Kloosterman, R. C., & Musterd, S. (2001). The polycentric urban region: Towards aresearch agenda. Urban Studies, 38, 623–633.

Lambooy, J. G. (1998). Polynucleation and urban development: The Randstad.European Planning Studies, 6, 457–467.

Lambregts, B., Kloosterman, R. C., & Van der Werff, M. (2005). Polycentricity and theeye of the beholder: A multi-layered analysis of spatial patterns in the DutchRandstad. Romanian Economic Journal, 8, 19–34.

Lang, R. E. (2003). Edgeless cities: Exploring the elusive metropolis. Washington DC:Brookings Institution Press.

Limtanakool, N., Schwanen, T., & Dijst, M. (2008). Ranking metropolitan areas: Acomparison of interaction and node attribute data. Cities, 27, 26–42.

Lösch, A. (1944). Die Räumliche Ordnung der Wirtschaft. Jena: Gustav Fischer.Meijers, E. (2007). From a central place to a network model: Theory and evidence of

a paradigm change. Tijdschrift voor Economische en Sociale Geografie, 98,245–259.

Meijers, E. (2008). Measuring polycentricity and its promises. European PlanningStudies, 16, 1313–1323.

Meijers, E., & Burger, M. J. (2010). Spatial structure and productivity in USmetropolitan areas. Environment and Planning A, 42, 1383–1402.

Nielsen, T. A. S., & Hovgesen, H. H. (2005). Urban field in the making: New evidencefrom a Danish context. Tijdschrift voor Economische en Sociale Geografie, 96,515–528.

Nielsen, T. A. S., & Hovgesen, H. H. (2008). Exploratory mapping of commuter flowsin England and Wales. Journal of Transport Geography, 16, 90–99.

Nitsch, V. (2003). Does history matter for urban primacy? The case of Vienna.Regional Science and Urban Economics, 33, 401–418.

Openshaw, S., Taylor, P. J. (1979). A million or so correlation coefficients: Threeexperiments on the modifiable areal unit problem. In N. Wrigley (Ed.), Statisticalapplications in the spatial sciences (pp. 127–144). London: Pion.

Parr, J. B. (1987). Interaction in an urban system: Aspects of trade and commuting.Economic Geography, 63, 223–240.

Parr, J. B. (2004). The polycentric urban region: A closer inspection. Regional Studies,38, 231–240.

Parr, J. B. (2005). Perspectives on the city-region. Regional Studies, 39, 555–566.Parr, J. B. (2008). Cities and regions: Problems and potentials. Environment and

Planning A, 40, 3009–3026.Parr, J., & Hewings, G. (2007). Spatial interdependence in a metropolitan setting.

Spatial Economic Analysis, 2, 8–22.Patuelli, R., Reggiani, A., Nijkamp, P., & Bade, F.-J. (2009). Spatial and commuting

networks: A unifying perspective. In A. Reggiani, & P. Nijkamp (Eds.), Complexityand spatial networks: In search of simplicity (pp. 257–271). Berlin: Springer.

Preston, R. E. (1971). The structure of central place systems. Economic Geography, 47,136–155.

Richardson, H. W. (1995). Economies and diseconomies of agglomeration. In H.Giersch (Ed.), Urban agglomeration and economic growth (pp. 123–155). Berlin:Springer.

Robson, B., Barr, R., Lymperopoulo, K., Ress, J., & Coombes, M. G. (2006). Aframework for city regions – Working paper 1: Mapping city-regions. London:Office of the Deputy Prime Minister.

Schwanen, T., Dieleman, F. M., & Dijst, M. J. (2003). Car use in Netherlands dailyurban systems: Does polycentrism result in lower commute times? UrbanGeography, 24, 340–360.

Schwanen, T., & Dijst, M. J. (2002). Travel-time ratios for visits to the workplace: Therelationship between commuting time and work duration. TransportationResearch Part A: Policy and Practice, 36, 573–592.

Small, K., & Song, S. (1994). Population and employment densities: Structure andchange. Journal of Urban Economics, 36, 292–313.

Spiekermann, K., & Wegener, M. (2004). How to measure polycentricity? Paperpresented at ESPON 1.1.3 project meeting, Warsaw, Poland (June).

Tosics, I. (2004). European urban development: Sustainability and therole ofhousing. Journal of Housing and the Built Environment, 19, 67–90.

Van den Berg, L., Drewett, R., Klaasen, L. H., Rossi, A., & Vijverberg, C. H. T. (1982).Urban Europe: A study of growth and decline. Oxford: Pergamon Press.

Van der Laan, L. (1998). Changing urban systems: An empirical analysis at twospatial levels. Regional Studies, 32, 235–249.

Van der Laan, L., Vogelzang, J., & Schalke, R. (1998). Commuting in multimodalurban systems: An empirical comparison of three alternative models. Tijdschriftvoor Economische en Sociale Geografie, 89, 384–400.

Van Ham, M. (2002). Job access, workplace mobility, and occupational achievement.Delft: Eburon Publishers.