Specialization and Concentration from a Twofold Geographical Perspective: Evidence from Europe

Abstract

Cutrini E. Specialization and concentration from a twofold geographical perspective: evidence from Europe, Regional Studies. This paper investigates European location patterns during a period of economic integration, seeking to identify the distinct roles played by the different geographical levels. The evolution of localization in Europe proved much more complicated empirically than the predictions based on Krugman's hypothesis. Using Eurostat regional data for the period 1985–2001, the paper shows that while manufacturing employment trickled down among regions, after the completion of the European Single Market a slight agglomeration occurred, but only across national boundaries. National specialization has emerged particularly in the European Union founding Member States. Moreover, there is evidence of an increasing polarization of the North–South divide closely connected with the growing concentration of high-technology sectors.

Cutrini E. La spécialisation et la concentration d'un point de vue géographique à deux temps: des preuves européennes, Regional Studies. Cet article cherche à examiner la distribution européenne des emplacements pendant une période d'intégration économique et à identifier les rôles différents joués par les divers niveaux géographiques. L'évolution des emplacements en Europe s'est avérée beaucoup plus compliquée du point de vue empirique par rapport aux prévisions fondées sur l'hypothèse de Krugman. A partir des données régionales Eurostat pour la période allant de 1985 jusqu'à 2001, cet article cherche à démontrer que, pendant que l'emploi industriel s'infiltrait dans les régions au compte-gouttes, il y a eu une certaine tendance à l'agglomération suite à l'échéance du marché unique, mais seulement de façon transfrontalière. Une spécialisation nationale a vu le jour, notamment dans les pays fondateurs de l'Ue. Qui plus est, il y a des peuves d'une polarisation croissante du clivage Nord–Sud, ce qui se rapporte étroitement à la concentration croissante des secteurs à la pointe de la technologie.

Emplacements Spécialisation Concentration Intégration économique européenne Analyse géographique à deux temps

Cutrini E. Spezialisierung und Konzentration aus einer zweiteiligen geografischen Perspektive: Belege aus Europa, Regional Studies. In diesem Beitrag untersuche ich die Standortmuster in Europa während einer Periode der wirtschaftlichen Integration, um die charakteristischen Rollen zu identifizieren, die von den verschiedenen geografischen Ebenen wahrgenommen werden. Die Evolution der Lokalisierung in Europa erwies sich in empirischer Hinsicht als weitaus komplizierter als die Prognosen auf der Grundlage der Krugman-Hypothese. Mit Hilfe von Eurostat-Regionaldaten weise ich für den Zeitraum von 1985 bis 2001 nach, dass das Beschäftigungsniveau der produzierenden Industrie innerhalb der Regionen zwar einem Trickle-Down-Effekt unterlag, aber nach Vollendung des Europäischen Binnenmarkts eine leichte Agglomeration auftrat, allerdings nur über nationale Grenzen hinweg. Eine nationale Spezialisierung hat sich insbesondere in den Gründungsmitgliedsstaaten der EU herausgebildet. Darüber hinaus liegen Belege für eine zunehmende Polarisierung des Nord–Süd-Gefälles vor, die eng mit der wachsenden Konzentration von High-Tech-Sektoren verknüpft ist.

Lokalisierung Spezialisierung Konzentration Europäische Wirtschaftsintegration Zweiteilige geografische Analyse

Cutrini E. Especialización y concentración desde una perspectiva geográfica dual: el ejemplo de Europa, Regional Studies. En este artículo examinamos los modelos de ubicación europea durante un periodo de integración económica con el fin de identificar los distintos papeles desempeñados por los diferentes niveles geográficos. La evolución de la localización en Europa resulta ser empíricamente mucho más complicada que las predicciones basadas en la hipótesis de Krugman. Usando datos regionales de Eurostat para el periodo 1985–2001, en este artículo mostramos que mientras el empleo manufacturero sufrió un efecto ‘goteo’ entre las regiones, tras la creación del Mercado Único Europeo ocurrió una ligera aglomeración, pero sólo entre fronteras nacionales. La especialización nacional ha surgido especialmente en los estados miembros fundadores de la UE. Además, hay muestras de una mayor polarización de la división norte/sur estrechamente conectada con la creciente concentración de sectores de alta tecnología.

Localización Especialización Concentración Integración económica europea Análisis geográfico dual

Keywords:

• Localization
• Specialization
• Concentration
• European economic integration
• Twofold geographical analysis

JEL classifications:

• C43
• F15
• N60
• R12

Acknowledgements

The author thanks the participants of the seminars at the University of Insubria, at the Polytechnic University of Marche, and at the Kiel Institute for the World Economy for their suggestions on earlier drafts of this paper. The author also thanks the participants at the Conference on ‘Spatial Econometrics and Statistics’ held in Rome, Italy, May 2006, and at the ERSA Conference held in Volos, Greece, in September 2006. The author acknowledges the financial support provided by the PRIN 2005 – ‘Economic Growth: Institutional and Social Dynamics’. Any errors are the author's responsibility.

Notes

Aiginger and Davies (2004) have already shown that specialization and concentration cannot diverge if relative measures are used.

Decreasing concentration was widespread across Spanish NUTS3 regions during the 1980s (Paluzie et al., 2001), across Italian NUTS2 regions from the early 1970s to the late 1990s (Rombaldoni and Zazzaro, 1997; De Robertis, 2001; Ciciotti and Rizzi, 2003) and, more recently, also in Germany (Suedekum, 2006).

Combes and Overman (2004) pointed out that:

the fact that Spanish regions did not change much with respect to one another does not mean that Spanish regions did not become more specialised relative to the rest of the EU [European Union]. (p. 21)

For a classification of different polarization, concentration, and specialisation measures, see Bickenbach and Bode (2006).

Overall localization is conceptually and analytically composed of two economic phenomena: the specialization of economies and the agglomeration of industries (Cutrini, forthcoming 2008).

The model suggests that decreasing transport costs will lead to an increase in specialization and a decrease in regional concentration.

Today, services make up the largest sector in most European economies and there are services that are of great importance for the distribution of regional income and welfare (for example, financial services and research and development). Any full assessment of concentration and specialization in Europe should include them.

The sectors manufacturing of leather and leather products (DC, division 19) and manufacture of coke, refined petroleum products and nuclear fuel (DF, division 23) have been excluded from the analysis because of the overwhelming amount of missing and confidential data.

This study refers to the distinction between absolute and relative measures drawn by Bickenbach and Bode (2006). Therefore, measures based on the uniform reference are considered as absolute measures, while those based on a non-uniform reference are labelled as relative measures.

These forces may be related to intra-industry input–output linkages, labour-market pooling, and industry-specific knowledge spillovers, but they might also indicate a high dependence on natural resources.

The two geographical components of the concentration index for each industry k can be easily derived by factor decomposition (for details on the formal decomposition of the localisation indices, see Cutrini, 2006).

Brülhart and Traeger (2005) test for the significance of temporal changes of regional localization by relying on a block-bootstrap, that is, resampling observations from different countries separately.

Intermediate spatial aggregation level is defined here as the level at which the within-group component is disentangled from the between-group component. Instead, the highest level of aggregation is the macroeconomic geographical benchmark (the set of European regions).

Throughout the present paper, ‘Europe’ refers to the 145 European regions taken together.

The sector includes chemicals; machinery and equipment n.e.c.; electrical and optical equipment; transport equipment; furniture, recycling; and manufacturing n.e.c.

Audretsch and Feldman (1996) showed the different propensity of manufacturing industries to generate spatial knowledge spillovers.

On this reasoning, the functional specialization of different localities is the aggregate outcome of a microeconomic change – induced by the decreased transportation and communication costs – in the firm's trade-off between the benefits of vertical integration and the advantages of spreading the different functions across space. When spatial transaction costs (that is, the costs of coordination and monitoring across fairly wide distances) decrease substantially, firms that used to perform managerial, research and development, and production tasks under a single roof prefer to become multi-plant organizations.

From 1985 onwards, Greece was characterized by increasing specialization in labour-intensive industries (food, textiles, and wearing apparel) and in non-metallic mineral products, which is a manufacturing industry closely linked to the construction industry (Table A8).

Geographical dispersion is referred to in the latter two cases, since Aiginger and Pfaffermayr (2004) and Aiginger and Davies (2004) used absolute concentration measures and their results are not directly comparable with the present ones. Moreover, nominal value added is the activity indicator.

Congestion costs are a crucial dispersion force in the models of Krugman and Livas (1996) and Puga (1999).

The model of Krugman and Livas (1996), which was inspired by the Mexican liberalization programme, suggests that falling trade barriers might affect a firm's location within each country. The fundamental idea is that, in a restrictive trade policy, forward and backward linkages foster the clustering of economic activity. As soon as protective measures are removed, the central place (usually the capital city) loses the advantage it had in a relatively closed economy, and firms, which now mainly sell to external markets, are more willing to migrate to peripheral regions, especially if relocation means better access to the international market.

In New Economic Geography models (for example, Krugman, 1991; and Puga, 1999), labour mobility has an important role in sustaining agglomerations; in a symmetric way, labour immobility is an important dispersion force.