Assessing the Functional Polycentricity of the Mega-City-Region of Central Belgium Based on Advanced Producer Service Transaction Links

Abstract

Hanssens H., Derudder B., van Aelst S. and Witlox F. Assessing the functional polycentricity of the mega-city-region of Central Belgium based on advanced producer service transaction links, Regional Studies. This paper comments on three aspects of POLYNET's implementation of the Globalization and World Cities (GaWC) model to measure functional polycentricity at the local scale. These aspects are: (1) the use of advanced producer services (APS) office locations as a proxy for inter-city relations, (2) the problematic calculation of the functional polycentricity index, and (3) the narrow conceptualization of functional polycentricity. Alternative methodologies based on transaction link data are introduced to examine whether Central Belgium displays features of a functionally polycentric mega-city-region. The results suggest that this is not the case: Central Belgium displays more features of a hierarchical urban system with Brussels as the dominant service city. However, some indications are found for a preliminary urban network formation.

Hanssens H., Derudder B., van Aelst S. and Witlox F. 根据高端生产性服务交易链评估比利时中部钜型城市区域的功能多中心，区域研究。本文将对全球化与世界城市模型（GaWC）中实施欧洲多中心钜型城市区域可持续发展管理 POLYNET 的三方面进行评论，用以评估地方尺度的功能多中心。此三方面为：(1) 运用高端生产性服务 (APS) 的办室公区位来代理城际关係，(2) 具有疑问的功能多中心索引计算，以及 (3) 对功能多中心的狭隘概念化。本文将引介根据交易链资料的替代研究方法，用以检视比利时中部是否展现做为功能性多中心钜型城市区域的特徵。研究结果则显示并非如此 ：比利时中部展现了更多阶层性的都市系统特徵，并以布鲁塞尔为支配性的服务城市，但研究亦发现初步城市网络形成的部分徵兆。

Hanssens H., Derudder B., van Aelst S. et Witlox F. Une évaluation de la polycentricité fonctionnelle de la mégalopole de la Belgique centrale fondée sur les liens de transaction des services à la production avancés, Regional Studies. Cet article commente trois aspects de la mise en oeuvre par POLYNET du modèle Globalization and World Cities (GaWC; modèle de la Mondialisation et des Villes du monde) qui vise à mesurer la polycentricité fonctionnelle sur le plan local. Ces aspects sont les suivants: (1) l'emploi de l'emplacement des bureaux des services à la production avancés comme substitut aux relations inter-villes, (2) le calcul problématique de l'indice de polycentricité fonctionnelle, et (3) la conceptualisation étroite de la notion de polycentricité fonctionnelle. On présente d'autres méthodologies fondées sur des données relatives aux liens de transaction afin d'examiner si, oui ou non, la Belgique centrale affiche les caractéristiques d'une mégalapole fonctionnellement polycentrique. Les résultats laissent supposer que ce n'est pas le cas. La Belgique centrale affiche plus des caractéristiques d'un système urbain hiérarchique, dont Bruxelles est la ville de services dominante. Cependant, certains indicateurs témoignent de l'établissement d'un réseau urbain préliminaire.

Hanssens H., Derudder B., van Aelst S. und Witlox F. Untersuchung der funktionalen Polyzentrizität der Megastadtregion Mittelbelgien anhand der Transaktionsverknüpfungen von Wirtschaftsdienstleistern, Regional Studies. In diesem Beitrag nehmen wir zu drei Aspekten der Umsetzung des Globalisierungs- und Weltstadtmodells (GaWC) durch POLYNET zur Messung der funktionalen Polyzentrizität auf lokaler Ebene Stellung. Diese Aspekte sind: (1) die Nutzung der Firmensitze von Wirtschaftsdienstleistern als Ersatz für Beziehungen zwischen Städten, (2) die problematische Berechnung des Indexes für funktionale Polyzentrizität und (3) die enge Konzeptualisierung der funktionalen Polyzentrizität. Ebenso stellen wir alternative Methodologien auf der Grundlage von Transaktionsverknüpfungsdaten vor, um zu untersuchen, ob Mittelbelgien Merkmale einer funktional polyzentrischen Megastadtregion aufweist. Die Ergebnisse lassen darauf schließen, dass dies nicht der Fall ist: Mittelbelgien weist eher Merkmale eines hierarchischen urbanen Systems mit Brüssel als dominierender Dienstleistungsstadt auf. Allerdings wurden auch einige Anzeichen für die Bildung eines vorläufigen urbanen Netzwerks festgestellt.

Hanssens H., Derudder B., van Aelst S. y Witlox F. Valoración de la policentralidad funcional de la región mega-ciudad de Bélgica central basada en los enlaces de transacción de los servicios avanzados de productores, Regional Studies. En este artículo analizamos tres aspectos de la implementación del modelo de Globalización y Ciudades Mundiales (GaWC) por parte de POLYNET para medir la policentralidad funcional a escala local. Estos aspectos son: (1) el uso de ubicaciones de oficinas de los servicios avanzados de productores en sustitución de las relaciones interurbanas, (2) el cálculo problemático del índice de la policentralidad funcional, y (3) la estrecha conceptualización de la policentralidad funcional. Introducimos metodologías alternativas basadas en datos de enlaces de transacción para examinar si Bélgica central muestra características de una región mega-ciudad policéntrica funcional. Los resultados indican que este no es el caso: Bélgica central muestra más características de un sistema urbano jerárquico con Bruselas como la ciudad de servicios dominante. Sin embargo, se observan algunas indicaciones para una formación preliminar de una red urbana.

• Central Belgium
• Mega-city-region
• Functional polycentricity
• Transaction links
• Gravity model

• 比利时中部
• 钜型城市区域
• 功能多中心
• 交易链
• 引力模型

• Belgique centrale
• Mégalapole
• Polycentricité fonctionnelle
• Liens de transaction
• Modèle de gravité

• Mittelbelgien
• Megastadtregion
• Funktionale Polyzentrizität
• Transaktionsverknüpfungen
• Schwerkraftmodell

• Bélgica central
• Región mega-ciudad
• Policentralidad funcional
• Enlaces de transacción
• Modelo de gravedad

JEL classifications:

• C21

Acknowledgements

The authors would like to thank the Editor and the anonymous referees for their useful comments on an earlier version of this paper. The usual disclaimers apply. This research work was funded by the Research Foundation-Flanders.

Notes

The remainder of this paper employs the term ‘advanced producer services (APS)’ rather than ‘knowledge-intensive business services (KIBS)’. There are two reasons for this: (1) this paper to some extent conceptually builds on the literature on global cities and the WCN in which APS is more commonly used than KIBS; and (2) apart from the usual KIBS such as management consultancy, law, advertising, etc., the analysis also includes financial services and insurance.

A similar argument has been made in studies that measure intercity relations based on ownership links in multinational companies (cf. Alderson and Beckfield, 2004; Rozenblat and Pumain, 2007; Wall and van der Knaap, 2011).

Belgium's top 300 firms, as listed on the Trends Top website (http://trendstop.rnews.be; accessed on 11 May 2009). This ranking is based on corporate turnover figures for fiscal year 2007.

An FUR comprises an economic core and a ring consisting of those neighbouring municipalities in which at least 10% of the working population commutes to the core on a daily bases. The economic core of an FUR is either a number of neighbouring municipalities with an employment density in excess of 700 jobs per km² and having a total of at least 35 000 jobs, or a municipality with over 35 000 jobs plus its neighbouring municipalities with an employment density exceeding 700 jobs per km². In the Belgian case, eight contiguous FURs make up the MCR. For more details on the geographical definition of the MCR of Central Belgium, see Aujean et al. (2005).

As the questionnaire was written in Dutch and English, but not in French, it has been suggested that the underrepresentation of Walloon firms might be the result of a language bias. There are two reasons why this might be the case but only to a certain extent. First, the bilingual (though, in practice, predominantly French-speaking) Brussels Capital Region is over-represented in the sample. Second, instead of considering the location of the companies to test the assumption of a language bias, it might be more useful to take into account the (native) language of the chief executive officer (CEO)/chief financial officer (CFO) or personal assistant who initially received and dealt with the survey.

On top of the 332 national links, the dataset also included twenty-six international links with service cities abroad.

The reason why the SD of the DIs is calculated instead of the SD of the RCs is that – as will become clear in the third step of the measurement methodology – the SD is needed to become infinite in the limiting case of absolute dominance.

However, although this normalization introduces ‘yardsticks’ that allow for a more refined assessment and comparison of the degree of polycentricity in urban systems, two cautionary remarks need to be made. First, these measures can be interpreted as a linear ratio measure: 0.25 does not represent half the strength of 0.50. Second, this paper is confined to nodal characteristics, while it could be argued that analyses should also focus on flows characteristics in general and their asymmetry in particular.

Although Brussels is situated north of the linguistic border, it officially constitutes a bilingual region. Therefore, the transaction links between Brussels and either Walloon or Flemish FURs are not considered as cross-border linkages.

In a study of the impact of the Belgian linguistic border on the spatial configurations of inter-firm networks of small and medium-sized enterprises located in the Flemish–Walloon border region, Strihan (2008) reported similar observations. Strihan noted that corporate networks do have ties crossing the border, but that these are few and – in her study – predominantly develop along family-based or ethnic axes of solidarity.

The only exception is Model 2, where the Vuong statistic indicates that the zero-inflated Poisson fit is better. However, as the expected number of zeros in both regression models is the same (that is, seventy-five compared with seventy-one observed zeros), this difference in model fit cannot be explained by a better fit for zero inflation. Hence, also for Model 2, a Poisson regression appeared to be the most appropriate model.