World city network analysis is an exciting field of research. However it does have its depressing side. There is a relatively large constituency of interested parties who consider the results as just another ranking exercise – where is my city in the world pecking order and why isn't it higher? These city partisans, largely in the blogosphere, tend to evaluate research in terms of how it fits with their limited horizons: our world city network results are deemed ‘controversial', meaning suspect, because we tend to rank US cities lower than bloggers would wish. This is all very depressing on several levels. The results are treated as just ‘out there', available for discussion involving minimal effort: there appears to be little or no interest in how results are produced as if this had no relevance. It is part of a contemporary ‘list-mania' where the ranking is all that matters. Ranking brings out the competitive spirit in commentators to replace reasoned thinking. And in the particular case of world city network analysis, the overall outcome is to spectacularly misunderstand what the research, and therefore the results, are all about. Networks are essentially about mutuality, analysing them reveals cooperative relations not competitive ones. No wonder city partisans find our results unappealing; actually it would be even more depressing if they did embrace them!
I have begun with this little rant because it addresses challenges to world city network analysis at their most overt level. Of course, ill-informed discussion in the blogosphere is an easy target but such academic smugness is not my intention here. Quite the opposite in fact: my purpose is to show that peer-reviewed knowledge is by no means immune from the basic assumptions underpinning list-mania. I have personally been caught up in this process through an early research project on how London related economically with other world cities. The latter had to be identified and a ‘roster of cities' was devised by counting numbers of selected firms in a range of cities (Beaverstock et al 1999). The variation in numbers was simplified by dividing cities into strata labelled alpha, beta and gamma. This essentially petty exercise, just a first step in investigating London's external links, has had immense influence: personally it is my most cited article and, with hundreds of citations, it is the most cited article ever published in the journal Cities. I interpret this astonishing success as my ‘alpha-beta-gamma misgiving'.
Why a misgiving? This most successful paper is a very simple taxonomic exercise in which relational thinking is conspicuous by its absence. Hence we have made a major contribution to cities considered as merely separate entities to be ranked and compared. As all my subsequent writings show (deriving from Taylor 2001), such modes of thinking eliminate a crucial part of the complexity of cities and thereby misunderstand them. Hence the alpha-beta-gamma misgiving is just about as large a research embarrassment as can be imagined. What can we learn from it? Of course, there is nothing inherently wrong with simplification, in fact it is a necessity: simplifying complex reality is precisely the task of social science research as it tries to make sense of society continuously changing in myriad ways. However, in this basic pursuit, it is often relations between entities that are over-overlooked or largely ignored. To understand how such unhelpful simplification comes about requires thinking about the way research begins – what presumptions are brought to a subject?
Where to start is a generic conflict between activating social study through investigating attributes in situ, or investigating relations beyond the immediate situation. The latter is not a matter of adding ‘context' to a study, rather it is a holistic argument: external relations are implicated in the very constitution of an entity. This argument occurs in many guises where study of individual units is confronted by systems, ecological or network thinking. Even with the most rigorous research on a unit or units, critical features will be banished from consideration leading, for instance, to subsequent identification of ‘unintended consequences' in modern urban planning, or ‘side effects' in modern pharmaceutical medicine. However well intentioned the motives, such practice based upon limited knowledge can be counter-productive: just so with cities when treating them as units rather than networks. This is the challenge to world city network analysis that is the subject matter of this chapter.
The City Hierarchy Presumption
Cities come in many sizes and central place theory was one very successful means of explaining why this should be so. First presented in 1933, Christaller (1966) modelled urban places as a hierarchy of central places so that size was a function of hinterland market. With the latter arranged in hierarchical order, it followed that cities servicing hinterlands were equally hierarchical, as represented by their different sizes. This idea of inter-city relations as hierarchical has become embedded in thinking about cities and is often generalised as city rankings. The link between hierarchy and ranking is that both imply city competition – a zero-sum game where hierarchies and ranking lists are there to be climbed.
I have documented the workings of this hierarchical presumption in urban economics (Taylor 2009) and it has been equally influential in the study of cities in globalization. The resulting emphases on competition in inter-city relations has meant that with the coming of globalization, cities (city governments) were set a new task to climb Friedmann's (1986) ‘world city hierarchy' so as to reach Sassen's (1991) ‘global city' status at the top. In fact, when developing world city network analysis I never realized the degree to which ‘competitive cities' were embedded in all thinking about cities. I totally underestimated the size of challenge that this represented. Whose got the biggest skyscraper has a very tangible answer – we can go and visit it, see a photograph of it, and read about its impressive physical dimensions. It is all about place, an attribute thereof. But relations between world cities are far less tangible; they can be seen as aerials for electronic communication on top of all large buildings but most have little or no impact on the observer on the ground, whether academic or blogger. This is all about myriad flows – ideas, information, knowledge, plans, instructions, advice, etc. from businesses in the buildings, all travelling via satellites in Castells' (1996) global space of flows. Gradually I came to appreciate that the key to pursuing a flows perspective on cities was to begin with process rather than place.
The way to remove a presumption is to change the question. The point about process is that many can occur simultaneously in the same place. Thus the idea that a city is a cacophony of processes appears unexceptional but the key point is that it allows full reign to relational thinking. For instance, focusing on Frankfurt as a place we might ask whether it is truly a ‘global city' with a debate rehearsing the pros and cons of this idea. But if we treat Frankfurt as a complex amalgam of many processes we bring forth a very different debate. To be sure the process of global city formation occurs in Frankfurt, especially in its financial district, but only alongside many other important processes that are continually reproducing the city. The question is now about how global city processes relate to others, how important are they? I used Frankfurt as my example because its ‘global city status' is seen as problematic, there is a debate about it, which I see as pretty pointless. Obviously there are global city processes operating there but they are less important than in London. But this archetypal ‘global city' is equally complex with global city formation just one of many processes, albeit a relatively more important one. The message is simple: think flows as process before place as outcome. This is how I have treated my alpha-beta-gamma discomfort.
Let us look at the city process within globalization in more detail. The vibrancy of world cities derives from two distinct but related urban processes. Each generates an externality, a benefit beyond the market, a positive bonus for locating economic activity in the city. First there are dense patterns of intra-city relations that create agglomeration effects and cluster advantages. Second there are strong flows of inter-city relations that create network effects and connectivity advantages. All successful cities combine these agglomeration externalities and network externalities to maximise the bonus of an urban location. There is a massive and sophisticated literature on agglomeration effects (e.g. Fujita and Thisse 2002), but network externality effects have been much less studied. I have concentrated on the latter as part of the Globalization and World Cities (GaWC) research programme where the study of network externalities has been pioneered.
Introducing city networks directly addresses the question of city competition (Taylor 2012). Networks can only be reproduced through mutuality, that is to say they operate through cooperation rather than competition, win-win situations rather than zero sum games (Thompson 2003). In city networks, cities need each other and all contribute to the wellbeing of the network (Taylor 2004). One way of thinking about networks in relation to hierarchies is that the former encompasses all links across all cities whereas hierarchies focus on just specific ‘vertical' links between cities at different levels: this argument has been theorised as a new ‘central flow theory' that complements traditional central place theory (Taylor et al 2010). Therefore, I accept that cities compete with one another; my concern is for the presumption that competitive behaviour is deemed to exhaust the possibilities of inter-city relations. The key point is that inter-city relations are more complex than this; in the more sophisticated studies of competitive cities, it is accepted that the competition exists alongside cooperative relations (Begg 1999; Sassen 1999). But this does not mean that they are of equal importance. The competitive process as hierarchy is so much simpler than the cooperative process as network. For instance, with a roster of 15 cities arranged as a hierarchy (1, 2, 4, 8) links are restricted to just 14 possible inter-city relations; a network of 15 cities encompasses a complete 105 possible links. Thus in this case, a hierarchical presumption would eliminate 91 links from consideration, that is to say, 87% of inter-city relations are ruled out of play before analysis begins. Of course, with larger numbers of cities, the percentage of inter-city relations that are missed rises appreciably through thinking hierarchically.
My conclusion is that the city hierarchy presumption (central place theory) should be replaced by a city network presumption (central flow theory). This means we treat city network processes as generic (it is how cities work), leaving city hierarchical processes as contingent. Examples of such contingency are cyclical downturns that enhance competition because of scarcity, and city gateway battles where there is only space for one city interfacing a region/country with the rest of a city network (Taylor 2012). Thus in today's cities in globalization there is a world city network with hierarchical tendencies, the latter enhanced by the current economic downturn and more generally by political boundaries. But this tendency cannot grow to dominate, that would destroy the mutuality of the network and thus its reproduction.
World City Network Analyses
I draw on Castells' (1996) concept of ‘network society' to theorise central flow theory and operationalise it as world city network analysis. Castells' (1996) network society is constituted as three spheres of flows: infrastructural (the logistics of communication and transport), social organizational (the work enabled by the network infrastructures), and transnational elite behavioural (billionaire leisure and wealth management). Network analyses have been conducted at each of these levels (Derudder et al 2012); I focus on the second level in which cities are the critical places. I present this through three arguments: first I rehearse specification of network process, second I indicate and present basic results from such analysis, and third I consider the implications of this approach.
Specification: The Interlocking Network Model
World city network formation is modelled as an inter-locking network. Networks usually consist of two layers, the net level and the node level, but in an interlocking network there is a third ‘subnodal' level. For cities operating as key components of the global economy, it is advanced producer service firms operating at this third level that are the network makers; they create the world city network through their everyday practices linking offices across the world. The three layers of the world city network are therefore the net level of the global economy, the node level of cities, and the additional sub-nodal level of service firms. And it is the latter that define the critical level: this is where the agents of network formation are found. Thus for studying the world city network it is service firms that are investigated in order to understand the city network as the outcome. In other words, it is through studying the location strategies of firms that it is possible to measure and analyse the world city network.
Why focus on these service firms? In the 1970s two separate industries, computers and communications, merged their technologies to enable work to be coordinated worldwide based upon simultaneous connections. Early on Sassen (1994) spotted two contrasting economic geography effects: first, a dispersal of production to cheaper labour locales, and second, a contrary trend towards concentration of management and business service industries. The latter were required to organize the new worldwide production and were concentrated in cities. As Sassen (1991) originally argued, it is concentration of management alongside financial, professional and creative services that characterises contemporary ‘global cities'. Of course, service firms have always clustered in cities to provide such services to their clients but under conditions of contemporary globalization those specialised services became worldwide with fundamental implications for work practices. Firms need a multiple office policy across many cities to provide a seamless service and protect global brand integrity by keeping all work in-house.
This is how it came to be that from the 1980s onwards there have been hundreds of large service firms with trans-national office networks, many of them global in scope. Each firm had its own location strategy – which cities to have offices in, what size and functions those offices will be, and how the offices will be organised. It is the work done in these offices that ‘interlock' various cities in projects that require multiple office inputs. Thus the inter-city relations in these servicing practices are numerous electronic communications – information, instruction, advice, planning, interpretation, strategy, knowledge, etc., some tele-conferencing as required, and probably travel for face-to-face meetings at a minimum for the beginning and end of a given project. These are the working flows that combined across numerous financial, professional and creative projects in multiple firms to constitute the world city network (Taylor 2001, 2004).
So we have to study service firms to describe and analyse the world city network but, unfortunately, there is no feasible way that data could be collected from firms on these working flows. As well as the obvious confidentiality issues with competing private firms, there is also a feasibility issue: the degree of research collaboration that would be needed from a large number of firms makes such a data collection exercise beyond reasonable social science research logistics. However, this is not a particularly rare situation in measurement practices: where direct measures cannot be obtained, there is the fall back position of carrying out indirect measurement. This requires access to more easily available data plus credible assumptions about how the firms operate.
As mentioned previously, service firms offer a seamless service across their office networks. This means that the geographical distribution of their offices, and their scope and range, are important selling points in attracting new clients. Hence such information is commonly available on service firms' web sites. This has been the main source of data for measuring the world city network: for each firm, offices are assessed individually by asking what is the importance of this office in this city within the firm's overall office network? Answers to this question are termed the service value of a city to a firm. These values are coded and become the quantitative input into the study: the coding ranges from 0 (a firm having no office in a city) to 5 (a city housing the headquarters of a firm); standard or typical offices of a firm score 2, minor and major offices 1 and 3, respectively, leaving 4 for scoring cities housing exceptionally important offices such as regional headquarters. The credible assumption that is made is that the more important an office the more working flows it will generate. Therefore two important offices will generate a much higher level of flow between their respective cities than two minor offices between their respective cities. These data and this assumption are combined to generate estimates of inter-office working flow levels between cities for each firm; they are not actual working flows, but potential working flows, indirect measures derived from the data and the model assumptions. Aggregating all potential working flows for all firms located in a city generates estimates of its working flow relations with other cities; when this is done for all cities it constitutes the world city network.
Network connectivity is the main measure of importance of a city in this model (Taylor 2001). It is computed from the products of service values for the city with each other city for all firms. Thus assuming m advanced producer service firms and n cities we can define a service value for firm j in city i as vij. The basic relational unit of measurement is given by
rab,j = vaj . vbj (1)
which defines the relation between cities a and b in terms of firm j. This is an elemental city-dyad interlock for one firm. The aggregate city-dyad interlock is then given by
rab = Σrab,j (2)
For each city there are n-1 such aggregate city-dyad interlocks and the network connectivity for a city is given by their sum:
Ca = Σrai (in which a ≠ i) (3)
where Ca is the network connectivity of city a. This relates city a to all other cities within the network through its firms and measures the degree of integration of the city into the world city network.
The data collection to operationalize this model and provide the results reported below was carried out in 2010. Office networks were coded for 175 advanced producer firms chosen as leading firms in their respective sectors: the top 25 in accountancy, advertising, law and management consultancy, and the top 75 firms for financial services (banking, insurance and diversified finance). Their offices were scrutinized across 526 cities worldwide. The end result is a 526 cities x 175 firms matrix with each cell indicating the importance of a specific city in the office network of a specific firm, 92,050 service values in all.
From this large amount of customized data we compute network connectivities of cities as defined in equation (3) to show the degree of a city's integration into the world city network. The values computed from equation (3) are relatively large and therefore to make them easier to interpret we present them as percentages of the highest scoring city.
Analysis: Chinese Cities in the World City Network
For the analysis I have chosen to focus on Chinese cities because they illustrate well the city network presumption operating in a network with hierarchical tendencies. As argued earlier, competition is enhanced through national boundaries; where there are many states there will be numerous ‘gateway cities' to link the national economic jurisdictions to the world economy. This is the case for Europe. In contrast China has no such need for so many gateways. This contrast is shown in Table 1, in which are listed the leading 20 cities in China and Europe in terms of their global network connectivities. The top three in each case are relatively similar but thereon there is a marked contrast: the remaining 17 European cities are all much more connected than China's fifth ranked city Guangzhou. The latter, with Shenzhen, represent moderate integration into the world city network; the remaining 14 Chinese cities are only weakly connected. Clearly business service network flows are much more concentrated in China than in Europe reflecting the latter's multiple states – 13 of Europe's top 20 are capital cities operating as gateways to their own ‘national economies'. Thus when comparing China and Europe in the world city network it is not that inland China is unconnected beyond the six leading eastern cities, but their levels are far less than equivalent cities in Europe. This also reflects China's much more recent growth as part of the world economy which means that many of its dynamic cities have yet to develop a strong global business servicing presence.
Table 1 Top 20 Chinese and European cities in the world city network
*Global network connectivity given as % of highest (i.e. London)
However it is unlikely that China will ever produce a national sub-network as multifarious as Europe's. This is because of China's sovereign unity, but the country's large size has an additional effect. Beyond Europe, economic globalization processes seemed to have favoured one city over its historic rival in several countries that traditionally had two leading cities – Sydney rising over Melbourne, Toronto over Montreal, Mumbai over Delhi, and Sao Paulo over Rio de Janeiro are all classic cases. But this does not seem to be happening in China; with three leading world cities, this country is apparently developing without such restriction. The case of Hong Kong is the archetypal example of how, without a mutuality presumption, commentators get cities spectacularly wrong. In 1997 Hong Kong reverted to Chinese sovereignty. The change of political authority was widely expected to result in the demise of Hong Kong as a major world city. In Pacific Asia generally, Singapore was identified as the main beneficiary, and within China, it was assumed that Shanghai would take over Hong Kong's role of linking the country to the rest of the world city network. Of course all this was surmised on the basis of inter-city relations being a zero-sum game: Hong Kong loses; other cities gain. But this is not how the world economy and the world city network work: as noted earlier the basic commercial process, as described by Jacobs (1992), operates through win-win market scenarios. Such a scenario has certainly unfolded for Hong Kong: since 1997 Shanghai, Singapore and Hong Kong have all massively prospered. The latter city, instead of declining, is consolidating its position as number three in the world city network, above Tokyo and Paris, and is even catching up London and New York (Derudder at al 2010). It appears that Hong Kong has a deeply embedded niche that has grown to complement work done in Shanghai and Singapore; any competition between these three cities is simply not that important.
I will focus on Beijing and Shanghai to illustrate the mutuality alternative to competition. These are of general interest because they have been the top two cities growing their connectivities within the world city network since 2000 (Derudder et al 2010). Here I will show that their respective increases in connectivity has not brought forth a competitive scenario, rather this appears to be another win-win process. Their growing complementarity is shown through two pieces of analysis that disaggregate their global network connectivities previously discussed in Table 1. The first disaggregation is by service sector and answers the question which of Beijing and Shanghai is the more connected in each of five sectors. The second disaggregation is by other cities to depict each city hinterland and answers the question which cities are more linked to Beijing and which to Shanghai. Both analyses set the two Chinese cities within the global network of world cites as global business centres.
The data upon which the connectivities are computed consist of five service sectors: financial services, legal services, advertising, accountancy, and management accountancy. Thus the overall connectivity can be broken down into these five component parts. When this is done we find that there are interesting differences between Beijing and Shanghai: in no cases are both cities ranked in the top ten, rather when one city has a top ten position the other is always outside the top ten. This means that the two cities are integrated into the world city network through different patterns of sector workflows based upon their different divisions of labour. The implication is that the two cities are more complementary than competitive in their global servicing relations.
Tables 2 and 3 show the evidence for this conclusion. The former table shows the two sectors where Shanghai is in the top ten: financial services and advertising. These are the modern core commercial services. Shanghai's highest ranking (7th ) is as a global financial centre (Table 2) even though Beijing is the headquarters of the leading Chinese banks. In transnational financial terms, Shanghai is the place to be in China. Similarly, in advertising, Shanghai as China's leading commercial city is chosen by transnational advertising firms for their new offices resulting in 8th position (Table 2). In this case Beijing drops to a lowly 18th – it seems that in China today, government is not as big a market for advertising as it has been in western countries.
Table 2 Connectivities where Shanghai is stronger than Beijing
FS is financial services
Beijing is in the top ten for the other three services (Table 3). Its highest ranking is 8th for accountancy (Table 3) for which Shanghai drops to 14th. For management services and law Beijing maintains a 10 th ranking (Table 3) but for these services Shanghai fares very differently: its lowest ranking is for the former (a lowly 23 rd ) but it is nearly level with Beijing for law (in 11th position). It would seem that at this time, firms offering accountancy and management consultancy services are attracted by government (and the market it provides) to make Beijing more of a global priority than Shanghai. Law is more interesting because both Beijing and Shanghai are similarly attractive to these firms. As can be seen in other cities in Table 2, law firms service both financial centres and political centres; thus from Europe both Frankfurt and Brussels feature in Table 2. But overall, the key point is that firms in different sectors tend to use Beijing and Shanghai in different ways thus providing them with complementary roles in the world city network.
Table 3 Connectivities where Beijing is stronger than Shanghai
AC is accountancy
Complementarities between Beijing and Shanghai are also reflected in their respective hinterworlds. Hinterworlds are the non-local (i.e. beyond a city's hinterland) links cities have with other cities. Using data on the office networks of 175 leading business service firms, estimated working flows between pairs of cities are computed as the city-dyad interlock defined in equation 2. In Table 4 the inter-city links between Beijing and Shanghai are compared by subtracting Shanghai dyad interlocks from Beijing dyad interlocks. Where the result is positive Beijing has the stronger connections, where it is negative Shanghai has the stronger connections. The political and commercial roles of the two cities are reflected in this listing. Beijing is strongly connected to New York (United Nations HQ) and major neighbouring Pacific Rim cities (plus neighbouring Eurasian Moscow). Taipei is, of course, a special case with its commercial links to Shanghai but without political links to Beijing. The contrast between Beijing's links to Italy's capital city Rome compared to Shanghai's links to Italy's ‘economic capital' Milan is particularly noteworthy. In fact, Rome (also with UN functions) is the only western European city where Shanghai does not have the stronger links. Other Asian and American cities that are important commercial centres such as Chicago and Mumbai, have closer connections to Shanghai. Thus do the hinterworlds of Beijing and Shanghai complement each other in the world city network.
Table 4 Comparing Beijing and Shanghai hinterworlds
Link difference is Beijing's link to a city minus Shanghai's link to that city
The policy implications of such analyses are difficult to implement. City government has jurisdiction over its administrative area only; it is states that have ultimate authority for external relations. There are ‘town twinning' and ‘sister city' arrangements in place but these focus on transferring policy practice (‘best-practice') with much less focus on commercial inter-city relations. The relations analysed above have not been developed through city governments but by business service firms interlocking cities to satisfy commercial needs. Some cities have been developing ‘foreign policy' by locating ‘development offices' in other cities but this is a minor process compared to world city network formation. What this analysis implies is that it is a city's Chambers of Commerce that should be forming alliances with other cities' Chambers of Commerce to explore and grow complementarities. This is to call for city-dyad economic governance practice in which city governments themselves would facilitate and monitor through democratic process.
Concluding Discussion: Special City-triad Connections
This paper has progressed as far as identifying the importance of city-dyads when thinking through world city network analyses. Of course, networks are much more than collections of nodal-dyads, just focusing on the latter itself presumes missing most network processes. This is where the analysis can become excessively complex. I will offer just one example of city-triads that suggest an important structure that appears to have evolved without any formal government policy prescription.
The development of the world city network seems to have involved creation of two city-triads each related to key stages of network formation and reproduction with respect to leading states. The initial globalizing impetus for network formation seems to have created a New York-London-Washington city triad with each city having a distinctive role. New York is the leading financial centre and Washington is the leading political centre (including finance governance of IMF and World Bank) both located within the leading state. London has the role of being the global platform outside the state's jurisdiction. The latter allows for certain processes that cannot occur in the leading state, for instance the creation of a ‘Euro-dollar' market in London as an early step towards globalization. The latest impetus for network development seems to have created a very similar city-triad featuring Shanghai, Beijing, and Hong Kong. Shanghai is the fastest growing finance centre and Beijing the fastest growing political centre (including finance governance of state banks) both within the fastest growing country. Hong Kong has the role of being the fastest growing global platform outside the state's jurisdiction. Again, the latter allows for certain processes that cannot occur within the state, which is the reason for China's ‘one country, two systems' policy and why Hong Kong continued to prosper after 1997. I think the parallel between the two city-triad processes is remarkable suggesting a necessary underlying structure that could only be recognised using relational thinking about win-win situations.
Beaverstock, J V, Smith, R G and Taylor P J (1999) ‘A roster of world cities' Cities 16, 445-58
Begg I (1999) ‘Cities and competitiveness', Urban Studies 36, 795-809
Castells, M (1996) The Rise of Network Society. Oxford: Blackwell
Christaller, W (1966) Central Places in Southern Germany. Englewood Cliffs, NJ: Prentice Hall
Derudder, B, Taylor, P J, Ni, P, de Vos, A, Hoyler, M, Hanssens, H, Bassens, D, Huang, J, Witlox, F, Shen, W and Yang, X (2010) ‘Pathways of change: shifting connectivities in the world city network, 2000-08' Urban Studies 47, 1861-77
Derudder, B, Hoyler, M, Taylor, P J and Witlox F (eds) (2012) International Handbook of Globalization and World Cities. Cheltenham, UK: Edward Elgar
Friedmann, J (1986) ‘The world city hypothesis', Development and Change 17, 69-83
Fujita, M and Thisse, J-F (2002) Economics of Agglomeration. Cambridge: Cambridge University Press
Jacobs, J (1992) Systems of Survival. New York: Vintage
Sassen, S (1991) The Global City. Princeton, NJ: Princeton University Press
Sassen S (1994) Cities in a World Economy. Thousand Oaks, CA: Pine Forge Press
Sassen S (1999b) ‘Global financial centers', Foreign Affairs 78 (1), 75-87
Taylor, P J (2001) ‘Specification of the world city network', Geographical Analysis 33, 181-94
Taylor, P J (2004) World City Network: A Global Urban Analysis. London: Routledge
Taylor, P J (2009) ‘Urban economists in thrall to Christaller: a misguided search for city hierarchies in external urban relations' Environment and Planning A 41, 2550-55
Taylor, P J (2012) ‘On city cooperation and city competition' in B Derudder, M Hoyler P J Taylor and F Witlox (eds) International Handbook of Globalization and World Cities. Cheltenham, UK: Edward Elgar, pp. 56-63
Taylor, P J, Hoyler, M and Verbruggen, R (2010b) ‘External Urban Relational Process: Introducing Central Flow Theory to Complement Central Place Theory', Urban Studies 47, 2803-18
Thompson, G F (2003) Between Hierarchies and Markets: the Logic and Limits of Network Forms of Organization. Oxford, UK: Oxford University Press