Ben Derudder is Professor of Urban Geography at Ghent University’s Department of Geography and an Associate Director of the Globalization and World Cities (GaWC) research network. He is one of the editors of Regional Studies, and will be a plenary speaker at the 2019 RSA Annual Conference. His primary research focus is on measuring and interpreting patterns in global urban networks, a summary of which can be found in a recent Regional Studies paper and the second edition of World City Network: a Global Urban Analysis (both published with Peter Taylor).

When I was a Geography student at Ghent University (Flanders, Belgium), most of the instruction I received into ‘regional geography’ and ‘regional studies’ was fairly traditional. Largely, it consisted of fact-filled introductions to the world’s, Europe’s and Belgium’s ‘regions’, or used such pre-defined ‘regions’ as the self-evident background for discussing geographical variation. It was only towards the end of the programme that this simplistic approach to ‘the region’ was problematized and things got more interesting and relevant. Looking back now, I recall two major insights which are probably obvious to most regional studies researchers and practitioners.

First, regions do not present themselves naturally. They are not given or self-evident, but constantly (re)produced by a series of actors. These actors may or may not be purposefully bent on devising regional schemes, but they do partake in decision-making through which regional patterns emerge, are reproduced, and disappear. Second, and related to this, looking for clear-cut regional borders is a futile undertaking. The concern for regional delimitation, which largely equates regionalization with classification, tends to produce regional patterns that are neatly demarcated as a simple, mosaic geography. But in reality regions have fuzzy and overlapping borders, and ultimately depend on the actors being studied.

As I pursued a PhD in Geography, my interest in ‘regions’ initially waned. There was a constant awareness that regions quite obviously matter in geography, but for some time I thought that as a concept it was largely inconsequential for what became the focus of my research: inter-city connections at the global scale. Most of this research, carried out under the umbrella of the Globalization and World Cities (GaWC) research network, focused on selected key agents who facilitate the reproduction of globalization: producer service firms (e.g. firms producing financial, accountancy, and consultancy services) creating a ‘world city network’ through their day-to-day work practices. Applying network projection techniques to location matrixes of hundreds of these firms across hundreds of cities, this research inter alia allows identifying some of the geographies of the ‘world city network’. Table 1 provides a glimpse into one of these geographies: it gives an overview of the 15 most connected city-dyads in 2018.

Table 1: The 15 most connected city-dyads in the world city network (based on 2018 data, see Taylor and Derudder, 2016 and Derudder and Taylor, 2018).

1 London New York
2 Hong Kong London
3 Hong Kong New York
4 London Singapore
5 New York Singapore
6 London Paris
7 New York Paris
8 London Shanghai
9 Beijing London
10 New York Shanghai
11 Beijing New York
12 Dubai London
13 London Tokyo
14 London Sydney
15 Chicago New York


Figure 1: most connected cities and city-dyads in the world city network (based on 2013 data, see Hennemann et al., 2015). This circular layout shows network (cities) nodes and edges (city connections) in the world city network concentrically around the connectivity map, with cities roughly and intuitively positioned in line with their world-regional location. Colour codes are used to identify cities in the same world-region, with edges coloured based on the dominant direction of the connection.

As already suggested, this research may seem very far removed from regional studies. In fact, such geographies of global inter-city connectivity may at first sight even seem the antithesis of the traditional regional depictions seeking meaningful organization of world-space: the connections in Table 1 and Figure 1 straddle the global economy, with seemingly very little regional ordering. However, once we start digging deeper, things turn out to be quite more complex than this. In Figure 1, the borders of the circular layout show dense connections which suggests that the bulk of the inter-city connections are actually quite ‘regional’. This turned out to be a recurring finding in our research: we consistently discover there to be much starker regional orderings to inter-city connections than initially anticipated and as suggested in Table 1.

To reveal the key dimensions of this regionality, we applied principal components analysis to location data revealed overlapping groups of cities (Derudder and Taylor, 2018). Alongside two ‘global processes’, which we term intensive and extensive globalization, we find above many regional components. Intensive globalization refers to a component consisting of New York and London as well as a select number of major headquarter cities: a global pattern of decision-making centres. Extensive globalization, in turn, refers to a much larger component of mostly capital cities of medium-sized national economies or the non-capital economic centres of such countries (e.g. Auckland, Dubai, Istanbul and Lagos): a global pattern of regional gateways. However, beyond these two ‘global patterns’ we find above all components representing broad regional orderings of cities. For example, there are Canadian, Australasian, Scandinavian, Indian, Chinese, and range of European groups of cities. Table 2 provides an overview of the 15 most typical cities in 3 of these regional components (i.e. cities with the highest component scores). In each case, it is possible to use a straightforward regional label because of the strong regional ordering. However, at the same time, there are a range of ‘outreaches’ that are sometimes surprising and sometimes less so.

Scandinavian cities for instance, have similar service mixes but the firms producing this pattern also tend to be located in the key world cities of New York and the specialized European financial centres of Luxembourg and Dublin. In the other two components, there is a clear-cut (post)colonial legacy: London is part of the Indian component (New York is not), while Madrid/New York are part of the Latin American component (London is not). In all three components there are trans-regional outreaches, but nonetheless again sometimes with some sort of regional ordering attached to it (e.g. Riyadh and Jakarta for Indian cities; Baltic cities for Scandinavian cities). The overall finding here is that leading producer service firms have reacted to, and contributed to, economic globalization through their location decisions on placing offices to service clients in cities across the world. Spatially they are network-makers rather than region-builders, but their creation of a world city network for servicing global capital inevitably reflects uneven regional markets for their services as well as their regional origins. As a result, they are producing regional patterns in putatively global inter-city connections.

Table 2: 15 most central cities in three of the regional components in the world city network (based on 2018 data; see Derudder and Taylor, 2019)

‘Scandinavia’ ‘India’ ‘Latin America’
1 Copenhagen Mumbai/Bombay Sao Paulo
2 Oslo Dubai Buenos Aires
3 Stockholm Bangalore Rio De Janeiro
4 Helsinki Hyderabad (India) Santiago
5 Luxembourg New Delhi Brasilia
6 New York (New York) Calcutta Mexico City
7 Gothenburg Ahmedabad Madrid
8 Bergen Pune Luxembourg
9 Malmö Chennai/Madras Belo Horizonte
10 London Manama Porto Alegre
11 Vilnius London George Town (Cayman)
12 Riga Cochin/Kochi Curitiba
13 Tallinn Jakarta New York (New York)
14 Arhus Cairo Campinas
15 Dublin Riyadh Tokyo

And so, in recent years the two major ‘regional’ insights from my Geography programme have come back through the backdoor: I have come to re-appreciate that regions and region-making are central to what I am doing. First, I had forgotten that regions are not self-evident, but are constantly (re)produced by a range of different actors. In carrying out their everyday work, many of the producer services firms we are studying have devised location strategies through which new regional formations are being constructed. Obviously the firms we study are not necessarily bent on devising regional schemes, and this is why regionality may not be the most obvious concept to adopt as we make sense of the geographies they produce. However, these firms do partake in location decision-making through which unfolding regional patterns emerge and may subsequently be enacted. The result is a world regional geography as practised by leading producer service firms.

The particular economic globalization process we measure – world city network formation – does therefore not support stories about a blanket-effect of global processes such as Friedman’s (2005) ‘flat world’; but neither does it support Florida’s (2008) ‘spiky world’ riposte to Friedman. Rather, we find that the latter world is far more regionally ordered than Florida and others propose: our results show new world regional geographies being created to satisfy different forms of service provisioning. So although Agnew (2013) rightly identified the curious paradox that contemporary globalization has been accompanied by the declining relevance of regional schemas at that scale, this needn’t be the case. However, and second, although there is a world-regional ordering to putatively global inter-city connections, this does not imply the presence of neat demarcations. As table 2 shows, the regions created by producer services firms are fuzzy, overlapping and porous. Teasing out the meaning of these regions and researching how they emerge, enacted and disappear may further enliven our imagination of what regions can be about.

Are you currently involved with regional research, policy, and development, and want to elaborate your ideas in a different medium? The Regional Studies Association is now accepting articles for their online blog. For more information, contact the Blog Editor at


Agnew, J. (2013) Arguing with regions. Regional Studies, 47(1), 6-17.

Derudder, B. & Taylor, P.J (2018) Central flow theory: Comparative connectivities in the world city network, 2013. Regional Studies, 52(8), 1029-1040.

Derudder, B. & Taylor, P.J. (2019) Three globalizations shaping the twenty first century: understanding the new world geography through its cities. Under review.

Florida, R. (2008) Who’s your city? Basic Books, New York.

Friedman, T.L. (2005) The world is flat. Farrar, Straus and Giroux, New York.

Taylor, P.J. & Derudder, B. (2016) World City Network: a Global Urban Analysis (2nd edition). London and New York: Routledge.