The world city network: Evaluating top-down versus bottom-up approaches

Stefan Lüthi, Alain Thierstein, Michael Hoyler

Research output: Contribution to journalArticlepeer-review

34 Scopus citations

Abstract

The growth of the knowledge economy has led to new forms of business networks linking cities and towns across different spatial scales. Various attempts have been made to analyse these networks empirically using the interlocking network model of the Globalization and World Cities (GaWC) research network. Two approaches can be distinguished from a spatial perspective: a global ‘top-down’ approach that studies the world city network from the perspective of the largest advanced producer service firms, and a macro-regional ‘bottom-up’ approach that starts with the most important knowledge-intensive firms located within specific territorial boundaries. This paper compares and critically assesses the methodological implications and empirical outcomes of both approaches with reference to case studies of the German space economy. Both approaches pursue similar objectives: to investigate external relations of cities, both transnationally and on the national scale. Differences exist in the theoretical argumentation: the top-down approach is grounded in world city research; the bottom-up approach is anchored in debates in regional science and economic geography. In this paper, we argue for the need of scale-sensitive interpretations of connectivity patterns resulting from different approaches to the interlocking network model and conclude with some tentative recommendations for the methodological direction of future research in world city network studies.

Original languageEnglish
Pages (from-to)287-294
Number of pages8
JournalCities
Volume72
DOIs
StatePublished - Feb 2018

Keywords

  • Germany
  • Interlocking network model
  • Knowledge economy
  • Scale
  • World city network

Fingerprint

Dive into the research topics of 'The world city network: Evaluating top-down versus bottom-up approaches'. Together they form a unique fingerprint.

Cite this