Regional Accessibility Trends in Europe: Road Infrastructure, 1957–2012

Abstract

Stelder D. Regional accessibility trends in Europe: road infrastructure, 1957–2012, Regional Studies. This paper describes the construction and analysis of a historical database of European road networks over the period 1957–2012 covering the whole European continent. Road accessibility is analysed over time for cities, regions and countries using a high-resolution population grid. The results show that many peripheral regions have lost accessibility relative to the centre during the first periods, but have been catching up after 1990. The general result of road infrastructure investment policy has been a spreading effect on accessibility since 1970.

Stelder D. 欧洲的区域可及性趋势：道路基础建设，1957 年至 2012 年，区域研究。本文描述 1957 年至 2012 年间，横跨全欧洲大陆的欧洲路网历史数据集的建构及分析。本文运用高解析度人口网格，分析各城市、区域及国家随着时间变化的道路可及性。研究结果显示，在最早几个阶段中，诸多边陲区域相较于核心而言丧失了可及性，但自 1990 年后便开始迎头赶上。道路基础建设投资政策，自 1970 年开始对可及性带来了普及的影响。

Stelder D. Les tendances régionales en matière d'accessibilité en Europe: l'infrastructure routière, de 1957 à 2012, Regional Studies. Cet article présente la création et l'analyse d'une base de données historique sur les réseaux routiers en Europe de 1957 à 2012 et qui couvre l'ensemble de l'Europe continentale. À partir d'une grille de la population à haute résolution, on analyse l'accès routier dans le temps pour les grandes villes, les régions et les pays. Les résultats montrent que, par rapport au centre, beaucoup de régions périphériques ont connu une perte d'accessibilité dans un premier temps. Cependant, depuis 1990, elles sont en phase de rattrapage. Le bilan général montre que la politique d'investissement dans l'infrastructure routière se répand depuis 1970.

Stelder D. Trends der regionalen Erreichbarkeit in Europa: Straßeninfrastruktur, 1957–2012, Regional Studies. In diesem Beitrag wird die Entwicklung und Analyse einer historischen Datenbank von europäischen Straßennetzen im Zeitraum von 1957 bis 2012 für den gesamten europäischen Kontinent beschrieben. Die Erreichbarkeit über Straßen wird für den betreffenden Zeitraum mithilfe eines hoch aufgelösten Bevölkerungsrasters in Bezug auf Städte, Regionen und Länder analysiert. Aus den Ergebnissen geht hervor, dass viele Randgebiete während der ersten Perioden im Vergleich zum Zentrum an Erreichbarkeit verloren haben, was aber ab 1990 wieder wettgemacht wurde. Das generelle Ergebnis der Politik für Investitionen in die Straßeninfrastruktur war ein Ausbreitungseffekt hinsichtlich der Erreichbarkeit seit 1970.

Stelder D. Tendencias de la accesibilidad regional en Europa: infraestructura vial, 1957–2012, Regional Studies. En este artículo se describen la construcción y el análisis de una base de datos históricos de las redes viales europeas durante el periodo entre 1957 y 2012 que abarca todo el continente europeo. Se analiza la accesibilidad por carretera con el paso del tiempo para las ciudades, regiones y países mediante una matriz de población de alta resolución. Los resultados indican que muchas regiones periféricas han perdido accesibilidad relativa al centro durante los primeros periodos, pero se han ido recuperando a partir de 1990. El resultado general de la política de inversión en la infraestructura vial ha sido un efecto de dispersión en la accesibilidad desde 1970.

• Transportation
• Networks

• 交通
• 网络

• Transport
• Réseaux

• Verkehr
• Netze

• Transporte
• Redes

JEL classifications:

• R12
• R42
• R58

Acknowledgements

The author thanks Lewis Dijkstra, Deputy Head of the Analysis Unit in the Directorate General for Regional Policy of the European Commission, for support and intensive discussions had during the whole research project. Dr Arnold Meijster from the Faculty of Mathematics and Natural Sciences of the University of Groningen has provided essential support in parallel programming without which the extensive calculations need for this project would never have been possible. Furthermore, Dr Peter Groote and Dr Marien de Bakker from the Faculty of Spatial Sciences and Professor Jan Oosterhaven from the Faculty of Economics and Business, all University of Groningen, have provided many valuable comments and suggestions. Last, but least, the author thanks Luuk Stelder and Jimme Zoete for their difficult and precise work in the process of digitizing all road maps.

Supplemental data

Supplemental data for this article can be accessed at http://dx.doi.org/10.1080/00343404.2014.952721

Notes

1. The full database discussed in this paper is open source and available at: http://www.regroningen.nl/stelder/.

2. For example Bonnefous and Crozet (1997) focus on the TGV in France, Jensen-Butler and Madsen (1996) on the Great Belt Bridge, and Hensher et al. (2012) on the North West high speed rail link in Australia.

3. The ‘Historical GIS for European Integration Studies’ project has been building up an archive of railways networks (see http://www.europa.udl.cat).

4. European Union project ‘Changes in Road Infrastructure and Accessibility in Europe since 1960’. European Commission; Directorate General Regional Policy, Policy Development; Economic and Quantitative Analysis, No. 2012.CE.16.BAT.040.

5. In (1) each Ai would increase with the same factor λ − β, leaving ai unchanged in (3). This is not the case when a distance-decay like in (2) is used.

6. This assumes regional firm homogeneity. If some regions are dominant in transport cost-intensive activities, they may benefit more from λ than others.

7. See, for instance, McCann (2005) on transport cost functions in New Economic Geography (NEG) models.

8. In the simple example of a two-region model, ignoring the size of the own region in (1) would always give an equal result in (4) for both regions, regardless of their size.

9. Depending on the units in D_ij, a weight of 1 implies an internal travel time of one minute, one hour etc.

10. For various reasons the alternative of coding an existing recent digital network back in time was not chosen. For more details, see Appendix A in the Supplemental data online.

11. Ferry connections were added with a direct travel time code.

12. For any regional aggregate of n cells, the average accessibility is calculated using the population of each cell as weights.

13. An exploratory test for the 30 largest cities showed no significant change in D_ij when class 3 roads were ignored.