GaWC Research Bulletin 457

(A)

Promiscuous Agglomerations: Towards Integrating Urban Agglomerations with Urban Networks

Peter J. Taylor^*

 

Introduction

Lindisfarne is a small island off the north east coast of England. Its remoteness attracted the early Celtic church, and a monastery was built there in the seventh century. It is famous for the production a century later of the beautiful Lindisfarne Gospels, which are usually attributed to a monk called Eadfrith. In creating this work Eadfrith is credited with inventing, among other things, the first lightbox: before painting each page he sketched his designs on transparent material stretched over a frame and shone a bright light from behind to guide his brushwork. However, ‘he worked alone so his inventions went no further at the time, they were as hidden as he was’ (Pye 2014, 54). It was to be about a millennium later before lightbox devices reappeared in Europe. The key point of this story is that Eadfrith worked in as ‘un-urban’ a social environment as can be imagined. Working on his own meant no urban agglomeration and doing it on a remote island meant no access to urban networks. Quite simply, eighth century Lindisfarne was no place for conversion of invention into innovation and its wider diffusion.

I begin with this example to emphasise the necessity of an enabling social context for economic development building upon new working practices. Economic development is a collective endeavour and the mutuality is provided, as Jacobs (1969) long ago attested, by cities. More recently city dwellers have been described as ‘smart people’ learning from other smart people (Glaeser 2012, 7) but this is a smartness feeding off a special social context; as an individual Eadfrith was evidently very clever but also so very rural. As suggested above, the specialness of an urban context is based upon two spatial characteristics of cities; the density of urban activities which is denoted as agglomeration effects, and the external linkages of urban activities which is denoted as network effects. In basic economic terms these can be understood as local and non-local externalities, two commercial advantages provided by cities beyond ordinary market transactions (Hicks 1969). The purpose of this chapter is to explore how agglomeration and network externalities translate into expressions of city power, the construction of economic spaces within and between cities to satisfy the economic demands of those cities.

The argument is presented as four parts. The first two parts provide rudimentary outlines of the two externalities starting with agglomeration. This topic has a relatively large economics literature tracing back to one of the modern founding fathers of the discipline, Alfred Marshall’s (1890) famous detection of an ‘industrial atmosphere’.  In contrast urban networks are the subject of a much more disparate, not to say disjointed, literature, mainly led by geography but without any strong theoretical pedigree. In both cases Saskia Sassen’s (1991)concept of global city is used to illustrate an important current instance of these city effects. In addition there is one scholar who works occurs strategically across both externalities; Jane Jacobs’ (1969, 1984, 2000) theory of economic development through cities is used tie agglomeration and network externalities together as a gross economic effect of awesome power. But what is this power? In the third section, this question is explored using John Allen’s (2010) ideas on indirect expressions of power through cities that contrasts with the more familiar direct expressions of power as typically executed by states. The final section mixes up the array of previous ideas through concrete examples of contemporary city power configurations. This is not a matter of ‘Mayors ruling the world’ but rather a more subtle insertion of city powers into corporate globalization generating variegated outcomes.

The basic tenet upon which this chapter is written is to promote understanding of the relations between urban agglomerations and urban networks. In many cases the firms producing agglomeration effects within a city are the same firms who are generating network effects with other cities. Agglomeration is always a dynamic process within a city but this dynamism also extends beyond the city: no urban agglomeration is an island. It is this promiscuous agglomeration thesis that demands studying urban networks alongside urban agglomerations.

Agglomerations and Clusters

In one of the standard texts on economic agglomeration Fujita and Thisse (2002, 1) begin by emphasizing that their subject is spatial concentrations of economic activities in ‘real world situations’ that occur at a variety of scales from the global (core-periphery) to the urban. It was at the latter scale that Marshall (1890) first identified ‘industrial districts’, specifically Sheffield’s steel industry and Manchester’s textile industry, where linkages between firms created unique conditions for a given industry’s growth and development. Latterly at the intra-urban scale agglomerations are commonly referred to as economic clusters so that agglomeration is treated as a clustering of industries within a city.

The standard definition of a cluster given by Porter (1998, 197-98) is as a ‘geographic concentrations of interconnected companies, specialised suppliers, firms in related industries, and associated institutions (for example universities, standards agencies and trade associations) in fields that compete but also cooperate’. Note that this includes more than just the firms so that clusters are constituted as complex amalgams of industrial organization. As such they offer firms a dual economic advantage: both large-scale effect of reducing costs and small-scale effect of increased flexibility. Not surprisingly therefore, firms within clusters grow faster than average.

Not all clusters are the same and numerous types of clusters can be identified depending upon which cluster processes are dominant – firm size, supplier/customers relations, internal/external orientation, and degree of cost reductions (Markusen 1996). But all are predicated on dense knowledge circulation both overt and tacit. At the heart of this is a strong demand and supply local nexus. On the demand side, knowledge of customer needs, often quite sophisticated, is accessible through building up relations of trust based upon past business and current reputation. On the supply side the key benefit is knowledge spillovers to create a fertile environment for innovation. Further there is access to specialised inputs, notably labour. The end result is that strong clusters are the place to be for any aspiring firm. And importantly, all this complexity is very difficult to replicate in a new location so that clusters are difficult to emulate and are resilient over medium-term time cycles.

But this clustering is not the only urban agglomeration effect. It is generally termed localization economies to distinguish it from the broader urbanization economies. The latter relates to the advantages of being within the encompassing city region. The larger the city economy, the goods and services for business will be more diverse and specialised (e.g. in law firms). These are commonly referred to as Jacobs’ externalities on account of her championing these city advantages in her classic work on the economy of cities (Jacobs 1969). Whereas other scholars have identified a downside to city growth, notably congestion and pollution, Jacobs memorably entitled her chapter on these problems “The Valuable Inefficiencies and Impracticalities of Cities” (pp. 85-121). They represent the complexity that is cities and their solutions are the stuff of city innovation (e.g. subway trains; skyscraper buildings). Thus rather than theoretically searching for the ‘optimal size’ of cities by making them economically ‘efficient’, Jacobs treats problems as ‘valuable’ and even argues that they are inherent to what it is to be a city. Inefficiencies have always stimulated key innovations and without them cities would perforce cease to develop, resulting in further urbanization economies being stymied.

The question naturally arises as to the efficacy of localization and urbanization effects. Glaeser et al (1992) posed this question as two oppositions. Are knowledge spillovers primarily within one industry or between many industries? And what is the impact of the market structure, competitive or oligarchic? They identified three positions from these queries: (i) the original Marshall theory of one industry and oligarchy; (ii) the Porter theory of one industry and competition; and (iii) the Jacobs theory of diversity of industries and competition. Glaeser et al.’s analysis of employment growth across 170 city regions from 1956 to 1987 firmly supported Jacobs’ position: the competition and diversity variables were strong, positive and significantly related to growth. Such empirical studies do not ‘disprove’ localization effects. They employ an extensive statistical analysis that picks out general trends in the data; this does not mean that localization effects cannot be found in specific case studies of particular industries.

Although not framed in Jacobs’ terms, Sassen’s (1991) detailed description of global city formation portrays a textbook example of a dynamic agglomeration process under conditions of contemporary globalization. With myriad large corporations requiring specialist aid in navigating their business at a transnational, often global, scale, existing advanced producer service firms diversified their work and offered new financial, professional and creative inputs thus enabling global economic expansions. Centred on financial support services but intermingled with a vast array of other key services (such as commercial law, management consultancy, worldwide advertising), massive economies of agglomeration were created resulting in improved flows of information, greater efficiency and higher liquidity. Sassen argued that this process was to be found in a few strategic cities – global cities (she studied New York, London and Tokyo) - where supply coexisted with demand (corporate headquarters) facilitating close supplier/customer relationships. Based upon trust built upon face-to-face interactions such relationships are themselves a major source of innovations. This example illustrates the porous boundary between localization and urbanization effects: financial sector clustering is critical as shown by Beaverstock et al (2003) for London while city-size also appears to be vital as shown by Frankfurt having financial clustering but lacking sufficient size (robust Jacobs externalities) to qualify as a ‘global city’ (Taylor et al. 2001).

The empirical evidence for Jacobs externalities supports her foregrounding of cities, but it is only a partial use of her theory of economic development. The crucial point of her theory is the linking of internal urban processes to inter-urban relations, which is essential for understanding cities in today’s globalization (Taylor and Derudder 2015).

Networks and Connectivities

The network externalities of cities are much less well understood than agglomeration externalities. One of the reasons for this is that the substantive elements of potential network effects – e.g. communication, transport, logistics, trade, migration  - are studied separately where cities are not always the main focus of attention. In addition, whereas the external relations of localization and cluster processes can be studied as specific industrial supply chains, for the urbanization effect the external dimension has been neglected.  For instance in their standard text, Fujita and Thisse (2002, 115-19) devote just four pages (less than 1% of the total) to ‘systems of cities’. They reprise Henderson’s (1987) work on trade as a product of economic policy agency (city government) rather than being part of, or deriving from, the process of agglomeration dynamism as argued by Jacobs (1969) (see also Abdel-Rahman and Anas 2004, 2319-22). As such, this trade approach does not address network externalities. More typically, matters stay largely at the single city level perhaps with brief references to central place theory, which deals only with city hinterlands in hierarchies. The latter theory has both economics (Losch 1940) and geography disciplinary origins (Christaller 1933) and was converted from a service sector model into a ‘national urban system’ format (Berry and Horton 1970) in the 1960s. This briefly aspired to national planning applications in the 1970s (Bourne 1975, Bourne and Simmons 1978). However this line of thinking also never addressed network externalities; the nature of its conceptualization precluded contribution to understanding city development: as Jacobs (1969, 33) pointed out, no city ever grew by just trading with its hinterland. Krugman (1987) dismissed it as ‘Germanic geometry’, providing only an interesting organization frame. But there remains a need for a model of inter-city relations that can begin to address the question of network externalities.

Not fully taken on board by economists (Nowlan 1997), Jacobs’ (1969) theory of economic development through cities is based on the concept of new work. She uses this to distinguish development from mere growth: adding more work to an existing production (old work) will make an economy grow larger (e.g. by doubling the size of a factory) but the city’s division of labour stays the same; in contrast adding new work will make the division of labour more complex, which for Jacobs indicates economic development. There are two ways in which new work enters a city economy. First, the straightforward way is by innovation, a new product or way of making that product. This occurs through localization and urbanization agglomeration processes operating in a particular city as described above. Second, and innumerably more common, there is import replacement: applying innovations emanating from other cities to produce locally what had previously been traded from outside. She argues that this process, diffusion of innovations between cities, is a potent economic force whose promiscuity within economic agglomerations leads to rapid growth spurts in cities. These latter have constituted the leading edge of economic development bringing together internal and external relations of cities. And under conditions of contemporary globalization it is this conjoining agglomeration and network effects, local and non-local externalities, which has become so conspicuously important for understanding the power of cities today.

Although global inter-city relations are explicit in Sassen’s (1991) argument, references to them are suggestive but not definitive. However, this is rectified in Castell’s (1996) conception of a network society based upon a global space of flows, which he sees as replacing industrial society based upon spaces of places. This new dynamic ‘social space’ operates at three levels in a kind of sandwich structure. The bottom layer consists of basic infrastructure support networks (communication, logistics, transport), and the top layer is constituted by the often opaque spaces of the global rich (tax havens, guarded play spaces, luxury city zones). In between there is the crucial work of everyday reproduction of globalization made up of myriad networks of the social practices. Castells argues that Sassen’s global cities are a crucial component of such formative practices, providing a new spatial organization for corporate globalization. But his conception envisages many more cities than Sassen’s very selective global city concept. It is his conception of inter-city relations that has come to be modelled as the world city network (Taylor 2001) and wherein a search for contemporary network externalities can begin.

In the recent past (i.e. in Castells’ industrial society) the business services that Sassen describes were typically associated with their city of origin – a Boston law firm, a New York advertising agency, a San Francisco bank, etc.. Providing financial, professional and creative services to their home city clients they combined specialist knowledge with local acumen. If their clients had international service needs these were accommodated through ad hoc means: banks would direct clients to their designated ‘correspondence banks’ with whom they had prior agreements; law firms had more informal arrangements, etc.. However when these clients began to globalise their businesses at new unprecedented levels (network society) many service firms followed their clients to pastures new. Finding themselves in other city markets they devised their own inter-city strategies explicitly represented in new worldwide networks of offices. This location strategy – keeping the servicing in house - had the additional advantage of brand protection compared to the previous inter-city practices. Thus was new servicing work diffused across cities around the world commonly housed in the new office tower blocks that became the agglomeration centres of global or, more generally, world cities.

The world city network is conceived as the amalgam of all the vast number of office networks of advanced producer service firms. The everyday work in a firm’s office that engages with other offices of the firm generates flows of information and knowledge (instruction, assessment, strategy, advice, coordination, specialist inputs (professional and local), analysis, planning, permissions, etc. etc.) between cities. This is modelled as an interlocking network of inter-city relations that has services firms as the agents (through ‘interlocking’ cities) of world city network formation. Once specified as a model and fed with data on firms’ worldwide office networks, this approach provides estimates of information and knowledge flows between cities to provide a powerful descriptive tool for understanding cities in globalization. The key measure is global network connectivity that shows the degree of integration of a city into the world city network. This is a gross measure of how much work in city agglomerations are linked together, city by city. But just like agglomeration effects, there is variety in network effects and these will be discussed in the fourth section below where externalities are linked to power.

However there is an important matter that should be addressed before this section concludes. It can be argued that because this network approach is based upon just one economic sector (advanced producer services) it can only lead to a partial understanding of inter-city relations. Furthermore although some service firms (notably banks) are themselves very large corporations, most business service providers are relatively small corporations in the global economy (e.g. are missing from Forbes lists of largest firms). This means that although their offices typically dominate the skylines of contemporary city development across the world, advanced producer services constitute only a relatively small portion of many cities’ economies. Nevertheless there is a crucial reason why focus on business services is a credible way of understanding contemporary city economics. These firms are not only providing global new work (service innovations); they indicate where economies are developing (other new work) so as to require these services. As such they can be considered as an ‘indicator sector’, signifying variable urban conditions across the global economy, analogous to ‘indicator species’ used to indicate vitality within eco-systems. Such indicator entities are not the biggest but they are special for their strategic position. This key notion implies the existence of city power.

Instrumental and Facilitative Powers

It is commonplace to view cities as organized into hierarchies based on their importance, usually indicated by their population size. This way of thinking promotes inter-city relations as essentially competitive: successful cities ‘climb’ the hierarchy at the expense of less successful cities. This depiction of cities engaged in a zero-sum game presumes a certain conception of power, domination: those at the top have more resources which can be deployed to compel those below to comply with their needs. However if cities are viewed as organized into networks a quite different outcome follows. This way of thinking promotes inter-city relations as essentially cooperative: successful cities need each other and feed off each other’s success. This is a positive sum game and such mutuality is at the heart of Jacobs’ theory of economic development. It does not follow that understanding city power is a simple choice between these two conceptions; in the messy real world cities are best viewed generally as ‘networks with hierarchical tendencies’. In this section the two forms of city power will be treated in turn but with more emphasis on the more complex cooperative dimension.

Power as domination is called instrumental power. There are three main operations of this power involving cities. Firstly, political power operates through capital cities to control territorial sovereignty, which includes economic jurisdiction over flows of factors of production. This is a special case of power deployed in a city but the source of that power is elsewhere, in the state. Although in journalistic language there are references to “Berlin doing this’ or ‘Washington doing that’, this is not city power and will not be considered further. However secondly, conceptualization of the spatiality of national economies as national urban systems with their national urban hierarchies introduces an economic instrumental power. This initiates inter-city competition where capital cities are often but not always at the top (cf. New York, or competitions for first place – Sydney/Melbourne, Toronto/Montreal). Here the power is city-based largely through corporate organization whereby headquarter decisions mainly in bigger cities determine economic activities mainly in smaller cities. Thirdly, this latter city power has been stretched to a global scale in early treatments of cities in globalization, initially by Hymer (1972) and most famously by John Friedmann (1986) in his world city hierarchy. The latter presented a competitive world of city powers defined by their geographical scopes topped by three dominant cities: New York, London and Tokyo (i.e. the same cities Sassen (1991) later researched as ‘global cities’). Cities in the hierarchy were viewed as ‘command and control centres’ on account of their housing the corporate headquarters whose global production strategies (searching out cheap labour) were creating the newly observed ‘new international division of labour’ (Frobel et al. 1980).

Power through networks is called facilitative power. Instead of elites holding power, in this form of power, elites produce power. Networked power is an effect of social interaction, a nexus of complex dynamic relations where power operates subtly but unmistakeably. A key power that cities possess is to be found in their capacity to hold networks together, they grow multifarious connections to become sort of ‘brokers’ for the world economy. Compelling is replaced by obligation, a need to work in a certain way because of social configurations that make it necessary. Thus Allen (2010) refers to cities as ‘obligatory passage points’ for the operation and development of significant work practices. For large American investment banks London is the place to be because of the knowledge nexus that that makes it an essential global economic platform for bridging US finance to the rest of the financial world. Similarly for many years firms wishing to enter the Chinese market were seemingly obliged to work via Hong Kong. It is through these operations that network power is shown to be more than a simple spatial of extension agglomeration. Urban agglomerations are promiscuous and integrate with urban networks as the source of city power that derives from position within the world city network. It is what Sassen (1994) was alluding to when she referred to her global cities as ‘strategic places’ and Castells (1996) extended as networked power.

Instrumental and facilitative forms of power have two contrasting theoretical expressions. Inter-city competition is manifest in the hierarchies of central place theory. But this is modelled as cities’ relations to their respective hinterlands. Although varying in scale up the hierarchy, this economic process remains a local one: it precludes consideration of non-local ‘horizontal’ relations between cities. But these are the very relations that underpin Jacobs’ theory of economic development. The mutuality embodied in the interlocking network model can be interpreted as ‘central flow theory’, a non-local spatial interaction theory to complement the local spatial interactions of central place theory (Taylor et al 2010). Thus do cities express both instrumental and facilitative powers simultaneously, deploying simple vertical dominations along side the continuous making of complex horizontal obligations. This is the dual power of cities in networks with hierarchical tendencies.

Cities in Corporate Globalization

How is the power of urban agglomerations and urban networks manifest in contemporary globalization? To answer this question it is necessary to decide how to portray today’s globalization. For several decades globalization has been linked to neo-liberalism on account of the neo-liberal policies of states (from Thatcherism, Reaganomics and IMF structural adjustments onwards) promoting free trade and international investment. Consolidated by the creation of the World Trade Organization (WTO) in 1994 (and confirmed by China’s membership in 2001), neo-liberalism is clearly the political and ideological instigator of contemporary globalization. But there is another way of considering globalization through focusing on the outcome in terms of power. Although set up by states, the direct beneficiary of the WTO are corporations: its workings can be interpreted as a sort of ‘Declaration of Corporate Rights’ (that is overriding state sovereignty) analogous to the 1948 ‘Declaration of Human Rights” (also overriding state sovereignty). In this argument neo-liberalism is the means to an end, which is international corporate power. Thus use of the label ‘corporate globalization’ within which cities are recognised as key players in the current world economy.

Today’s global space economy is primarily constituted as a bi-layered economic process. This is city-centred but in which states continue to have an important role. Thus the two layers are (i) a commercial space of flows largely articulated through cities as urban agglomerations and (ii) an international space of places that are state economic jurisdictions over bounded economic factors. The interrelations between these two global spaces is sorely under-researched; there are no models of their interaction due to simultaneous dearth of both joined up theory and relevant data. In such circumstances it is only possible to provide glimpses of the bi-layered process through concrete examples of its operations. Five examples of city configurations are discussed by way of illustration. The first highlights a common problem in current studies of cities in globalization: a case study that ignores process and thereby generates problematic empirical findings. The second counters this empiricist trend by focusing on results from world city network analysis that focuses on strategic places. The other three examples treat power expressed through specific configurations: triangular bridges in China and the north Atlantic; economic shadows throughout the world city network; and critical London/Frankfurt relations in Europe.

Financial Centres with and without Agglomerations

Billed as listing the ‘most powerful financial centres in the world’ (uk.businessinsider.com/most-powerful-financial-centres-gfci-index-for-2017-2017-9), the Global Financial Centres Index (GFCI) has been ranking financial centres since 2007. The methodology combines existing data on each centres ‘competitiveness’ with the responses of ‘international financial professionals’ to a questionnaire also firmly framed in the competitive paradigm. Data is organized as five ‘factors of competitiveness’ (Business Environment, Financial Development, Infrastructure, Human Capital and Reputation) wherein each covers four topics. Of the resulting 20 topics only four directly relate to urban agglomerations and urban networks: cluster depth and breathe under Financial Development, transport and ICT under Infrastructure, and level of innovation under Reputation. This is what Jacobs (2000) calls a ‘thing theory’ whereby process is described by the attributes of a city rather than their actual intra- and inter-relations. In this case Connectivity is actually mentioned but only in terms of how well known a centre is to those respondents in other centres; the information is simply used to scale centres into ‘global’, ‘transnational’ and ‘local’.

The latest headline result of this exercise (for 2017) shows the top 20 financial centres out of 85 studied (Table1). These rankings are based on competitiveness and should, therefore, represent instrumental power. The top five centres seem to fit expectations but beyond that there are definitely surprises. In Europe, for instance, after London only four centres appear: Zurich, Frankfurt, Luxembourg and Geneva – no Paris, Amsterdam, Milan or Madrid in the top 20. More globally, Montreal, Melbourne, Boston and Shenzhen are also top 20 surprises given the absence of Chicago, Washington and Seoul in the top 20. What is being picked up here is the difference between major cities with their multiple agglomerations and a narrow focus on the financial function. This is confirmed in the rankings below 20 where ten islands feature - in rank order: Bermuda, Cayman Islands, British Virgin Islands, Jersey, Guernsey, Isle of Man, Trinidad and Tobago, Gibraltar, the Bahamas, and Malta. These are clearly tax havens, small countries selling their political sovereignty to provide financial advantages to transnational clients. In this reading of the world of finance, Bermuda (ranked 29) and the Cayman Islands (31), for instance, are more ‘powerful’ than Amsterdam (33). Further, the British Virgin Islands out-ranks Vienna, Johannesburg, Milan, Brussels, Madrid, Sao Paulo, and Moscow among many other cities with important financial functions. Remote islands but not exactly Eadfrith’s Lindisfarne, Figure 1 shows a picture of the British Virgin Islands’ opaque economy. Each drawer probable represents a company that has connections to the rest of the world but, although spatially very, very, very close together, this is most certainly not a financial cluster. Connectivity without agglomeration, this is a simple state process (the obverse of urbanization in a autarchic state) far from the complexity of city process.

Table 1           Top 20 Global Financial Centres

* Global Financial Centre Index expressed as % of highest rank

Figure 1          British Virgin Islands’ non-agglomeration economy

Nevertheless several important points can be made. First, collecting data on individual items and then amalgamating them to create a measure of process is a recipe for confusing and compounding different processes. Second, the seemingly ubiquitous assumption about competition is seriously misleading. Tax havens, like cities, are not only competing, they work together, often carrying out complementary tasks in overall projects (Zucman 2015, 23-29). For instance, most ‘shell’ companies domiciled in the British Virgin Islands are created in Geneva (p. 28) because ‘nothing happens in the Virgin Islands’ (p. 73). Agglomeration is still important but happens elsewhere: “Most money managers still work in New York, Paris or London - close to their clientele – but the funds are subjected to the laws of the tax haven” (p. 27). Third, note that to unravel these processes requires a focus on flows between places not just places. This extends to a larger scale of presentation where the GFCI is analysed regionally. Thus are London and New York separated, the first discussed within Europe, the second within North America. This simplistic regional framing immediately forecloses on what is globally by far the most important inter-city link, the only one with a name: NYLON. Quite simply it is impossible to understand London in the world economy separate from New York and vice versa, but to understand requires introduction of facilitative powers.

Positionality in the World City Network: Promiscuous Agglomerations

As previously noted, global network connectivity is the basic measure of a city’s integration into the world city network. Table 2 shows the top 20 cities on this measure; it is in the same format as Table 1 to aid comparison. These cities can be interpreted as constituted by  the most important agglomerations that facilitate the operation of a world city network. The most obvious difference with the GFCI results is that although London and New York are the top ranks in both tables, with connectivity measure they are much more important relative to the rest: they have immense facilitative powers within and between them. In addition there is a different world geography that emerges in Table 2. Table 1 consists of cities from rich countries (especially English-speaking) plus China whereas leading cities in the ‘Global South’ feature in Table 2: Sao Paulo, Mumbai and Mexico City are important facilitators of the world city network. Furthermore, Zurich, Luxembourg and Geneva are conspicuous by their absence in the top connectivity results; Table 1 picks up a particular operational importance they have in wealth management but this does not translate into being so important as broader network facilitators when actual office location decisions (where financial work is done) are considered as shown in Table 2 (they actually rank 32nd, 66th and 88th respectively in GNC). This shows the advantage of modelling a process over compilation in simple ‘thing theory’ for understanding city powers.

Table 2      Global Network Connectivity

* Global network connectivity expressed as % of highest rank

World city network analysis incorporates much more than the basic measure of global network connectivity (Taylor and Derudder 2015). One type of analysis is particularly pertinent to finding facilitative powers. Global network connectivity indicates the quantity of work facilitating network production but this is not the same as strategic positioning in the world city network. The concept of strategic implies a special positionality in the way the network develops. This has been explored in Taylor et al (2014) in which a measure of strategicness - strategic network connectivity - is devised. Three steps are involved. (i) Principal components analysis of the whole data set (175 cities in 526 cities) shows two distinct location strategies: intensive globalization featuring dense connections between cities in rich countries plus China and extensive globalization which brings in key cities in the ‘Global South’. Both are anchored in the New York-London connection but otherwise are quite different. (ii) 25 firms are identified whose office networks intersect the two globalizations: these are designated strategic networks. (iii) 45 cities where these firms are particularly active are designated strategic cities because of this special positionality. Straddling the two globalizations, these cities’ agglomeration effects can be said to be globally promiscuous, their alternative mutualities making them key brokerage places.

Table 3 shows the top 20 cities for strategic network connectivity with their ranking compared to global network connectivities reported in Table 2. Four cities that drop out of the top 20 in this exercise are also listed to provide a full comparison.  Note the swapping of positions between London and New York – their critical inter-city relation will be addressed in the next section. Here focus is upon cities with large changes in rank. There are seven cities where ranking increases five places or more and five of them are US cities. This is the most important finding: as advanced producer firms have been globalizing their office networks the special strategic ones have kept a strong geographical anchor in the USA, still the powerhouse state of the world economy. Thus Chicago is now 3rd and Los Angeles 7th separated by San Francisco coming from outside the top 20 to 6th. Perhaps less surprising are Miami with its Latin American links at 16th and Washington DC with its political clout just entering the table. The other two cities in this rise category are one that is expected – Frankfurt as the EU brokerage city; and one quite unexpected – Bangkok used, possibly, as strategic city between East and South Asia. Of the large decliners, the four that actually leave the top 20 are very different cities but all are dominant in their own countries; evidently this is not projected strategically across the world. Dubai is a different case because it is often viewed as strategic given its geographical position between Europe and Asia. But Bassens (2013) refers to it as a gateway only to itself. It operates as a financial outlier of London through which investment is channelled for the city’s own rapid development thereby minimizing its importance as a global strategic site.

Table 3      Strategic Network Connectivity

Cities changing rank by 5 or more places are emboldened (strategic rises)
and italicized (strategic declines)

This analysis is certainly not the last word on Sassen’s global cities as new strategic places (see Goerzen et al 2013) but it does provide a unique worldwide assessment of strategicness in contemporary globalization. For the moment more can be learned from particular inter-city configurations.

Triangular Bridges of Shared Powers

China is the only country with three cities in the top ten for global network connectivity (Table 2). Lai (2012) has studied the mutualities between Beijing, Hong Kong and Shanghai: she sees ‘a dual headquarter strategy’ for Beijing-Shanghai relations and ‘parallel markets’ for Hong Kong-Shanghai relations’. Beijing-Hong Kong relations are not so explicitly characterised but clearly relate to Beijing’s political role in relation to Hong Kong’s outsider role.

There is a clear functional division of powers between the three cities that creates a win-win triangular configuration. As the capital city Beijing’s has a command and control focus dominating through state planning and policy-making plus housing headquarters of state corporations including banks. In contrast Shanghai is more innovation orientated: it ‘is tasked with testing new products, developing new markets and financial innovation’ (Lai 2012, 1283). Hong Kong, on the other hand, has developed as the crucial conduit ‘connecting global capital to China’ using its relative political autonomy as the country’s ‘offshore financial centre’ (p. 175). This triangular bridging arrangement appears to have evolved generically – there is no mention of it in any of China’s five-year plans. This is because this triangulation is largely a product of facilitative powers, an economic process by service firms operating through all three cities to harness a mix of agglomeration and network externalities. Thus have these shared city powers coordinated China’s integration into the world economy and thereby enabled the country’s rapid economic development.

There does seem to have been a precedent to this high-level, power-sharing triangulation: Washington as political capital, New York as financial innovation centre, and London as the financial offshore piece of the jigsaw. Differences between the China and US states means that Washington is not as powerful as Beijing but the city has become a major strategic node in the world city network (Table 3) not least because of the concentration of global legal services and more general lobbying activities in the US capital. Contemporary ‘financial NYLON’ derives from the scarcity of dollars in Europe after World War II resulting in the development of the euro-dollar market in London. Subsequent liberalising of London’s financial industry in the 1980s (“Big Bang”) attracted New York investment banks to create ‘a cooperative division of labor’ between the two cities (Sassen 1999, 81). This consists of London’s ‘denationalized platform for global operations’ and New York’s ‘brilliant financial engineering’ (p. 83-4). Smith (2012, 421) goes as far as saying they operate like a single city – ‘a transatlantic metropolis’. It can be added that New York has the deeper agglomeration effects whereas London contributes the necessary global network effects. Thus New York investment banks and other leading service firms are able to simultaneously harness both agglomeration and network externalities.

Territorial Power Shadows: States Influencing the World City Network

Some years ago Hill and Fujita (1995) pointed out that Osaka’s economic development was constrained by what they called its ‘Tokyo problem’. By this they meant the growing primacy of Japan’s capital city due to increasing concentration of corporate headquarters, information and financial services there. This was the global city formation that Sassen (1991) described as happening in the city. It represents the most explicit way in which states affect the world city network. States with their territorial jurisdictions over economic practices create a world mosaic of national markets in which firms operate. This directly influences firm’s global location strategies so that it is common for there to be just one major city, typically the capital city, that houses the offices of, say, transnational financial service firms. In addition, global law firms have to negotiate this mosaic through the multiple national legal codes, and advertising agencies have to be cognizant of language and cultural differences of their consumers across multiple countries. Both of these service firms are typically located in just one city per country.  Thus Osaka is not the only ‘second city’ to be affected by a corporate globalization that enhances city primacy within states: Mombasa has its Nairobi problem, Pusan has its Seoul problem, Cordoba has its Buenos Aires problem, Krakow has its Warsaw problem, Arhus has its Copenhagen problem and so on (Taylor and Derudder 2015, 76).

This influence of states on the world city network extends beyond second cities and is termed the territorial shadow effect. Even in the case of the USA, it seems that New York casts its economic shadow over the other major cities (Taylor and Lang 2005). One unexpected finding of the early world city network analyses was the relatively low ranking of US cities for global network connectivity. This was found to be due to the size and sophistication of the American business services sector that resulted from two location decision-making consequences. On the one hand, most foreign service firms decided they need a presence in the USA but did not attempt break into existing well-serviced markets. For this limited purpose they overwhelmingly concentrated in New York. On the other hand, most US service firms prospered in the largest service market in the world and were adverse to risk by moving to additional less known markets. But we now know from Table 3 that firms straddling the two service globalizations bucked these two tendencies to become ‘strategic firms’ generating important ‘strategic global cities’ in the USA beyond New York.

However the USA is a special case in globalization. In other countries in which there were traditionally several important cities, globalization has enhanced primacy without any strategic adjustments. Beyond the obvious cases of the success of London and Paris in the world city network, there is evidence of globalization tilting old city rivalries towards Sydney, Toronto, Milan, Sao Paulo, Mumbai, Johannesburg over Melbourne, Montreal, Rome, Rio de Janeiro, Delhi and Cape Town respectively, often indexed by the national centralization of stock exchanges in the former cities. Nevertheless there is nothing inevitable about this tendency; an interesting exception to this outcome tells us something further of state effects in changing circumstances. With the reunification of Germany in 1989 and the return of Berlin as federal capital, this city embarked on policies to make it central Europe’s ‘global city’ harking back to its historical pre-eminence in the region. But it was unable to fulfil this ‘global city vision’ not least because the city developments in former West Germany had already globalised without a primate centrepiece (Kratke 2001). There was, for example, no possibility of Berlin replacing the facilitative power of Frankfurt as Germany’s leading global financial centre. This example clearly shows that agglomeration and network effects are not simple political policy instruments but are complex economic processes made and maintained by firms, local and non-local.

London/Frankfurt and Europe in the Global Space-Economy

London’s relations with New York and Frankfurt’s relations with Berlin have been briefly treated but what of the relations between these two European financial centres? London-Frankfurt relations became a matter of general public interest in 2001 with the launch of the new euro currency, which the UK did not join. The financial press portrayed this as a competitive contest between the two cities as to which would now become Europe’s premier financial centre: the establishment of the European Central Bank in Frankfurt was taken as evidence for Frankfurt ‘catching up’ (Beaverstock et al 2001, 6-8). Simple investigation, however, quickly found that this was not what was happening; all transnational firms studied were found to have branches in both cities. It was in their interests for London and Frankfurt both to be successful, with different functions allocated to each city. Thus the inter-city relationship was one of cooperation; the journalistic debate had fastened on to UK-German political rivalry within the EU and incorrectly projected this on to an economic inter-city relation.

This case study is interesting for its use by John Allen (2010, 2901-04) to illustrate facilitative power relations between cities. He interprets the processes described as an example of a positive-sum city network based upon a ‘beneficial division of labour’. The basic mutuality can be stated as follows: London’s importance in global capital markets and associated work skills benefited Frankfurt through bridging to its servicing of German and European markets, while Frankfurt benefited London by providing access to new developing business opportunities in an expanding Europe. Allen argues that this thick and dense interpenetration of financial and professional elites is how relations between cities now operates. Again there is nothing inevitable about this; it has to be continuously worked on. This involved both brokering new relations (contacts, visits) plus maintaining the network by building on past brokering. All this work is enabled by electronic connectivity but it is constructed though co-presence, reducing risk over distance in mobilising and managing funds. This is an awesome power, which Allen illustrates specifically by the conversion of Sydney from a global fringe position as gateway city into the Australian economy to a global central position as key broker in Asian, North American and European city networks.

Returning to the London-Frankfurt relation, this is back in the news due to Brexit. Now it is the UK leaving the EU that is seen as a threat to London’s pre-eminence as Europe’s financial centre. Stories of firms leaving London plays into a political narrative not unlike the euro case study. Firms, or parts of firms, leaving presumably do not need, or no longer need, what London presently has to offer. Agglomerations are essentially dynamic, change is normal, but their innate power, including their promiscuity, makes them strongly resilient. For instance, after the 9/11 attacks on New York, affected firms (i.e. totally destroyed offices) were up and working again, using their links and contacts, within a day (Page 2005). If there were a concerted political effort to reduce London’s importance in a competitive contest through EU political opportunism, the winner would likely not be Frankfurt, nor Paris, nor Dublin, nor Amsterdam but Singapore! That is to say, a win-win situation would be converted into a lose-lose outcome.

Concluding Remarks

This chapter has attempted to bring together urban agglomerations and urban networks into a single argument about the power of cities. Largely focussing on their facilitative powers, the seminal work of Jane Jacobs (1969) on economic development through and between cities has been employed: local and non-local city externalities are interpreted as promiscuous agglomerations, clearly evident in corporate globalization.

The contemporary world economy, with its vast quantities of flows, has been likened by Nigel Thrift (1999, 272) to ‘a blizzard of transactions’ of ‘almost unimaginable complexity’. Blizzards are experienced as chaotic but are actually expressions of a wider organized complexity. Jacobs (1961) famously categorised cities as expressions of organized complexity and the introduction of the concept of promiscuous agglomerations in this chapter has tried to introduce some order of comprehension to this economic blizzard. Bypassing central place theory and linking central flow theory to economic understandings of agglomerations has enabled some descriptive insights but no theoretical extension is provided. Central flow theory is not the geometric stasis that is central place theory but it remains a theoretical framework only, just a useful tool to aid empirical investigations. It was noted in the introduction that economists’ use of Jacobs’ work has been limited to urban agglomeration (i.e. a ‘half-Jacobs’ as it were) and this has been righted here in a most preliminary way. There is immense work still to do.

REFERENCES

Abdel-Rahman, H. M. and Anas, A., (2004) Theories of systems of cities. In J. V. Henderson and J-F. Thisse (eds), Handbook of Regional and Urban Economics, (Ed. 1, Vol. 4). London: Elsevier.

Allen, J. (2010) Powerful city networks: more than connections, less than domination and control.
Urban Studies 47, 2895-2911.

Bassens, D. (2013) The city-upon-the-gulf: The relational growth and decline of ‘world city’ Dubai. In M. Acuto and W. Steele (eds) Global City Challenges. London: Palgrave.

Beaverstock, J. V., Hoyler, M., Pain, K. and Taylor, P. J. (2001) Comparing London and Frankfurt as World Cities: A Relational Study of Contemporary Urban Change. London: Anglo-German Foundation.

Berry, B. J. L. and Horton, F. E. (eds) (1970) Geographic Perspectives on Urban Systems. Englewood Cliffs, NJ: Prentice Hall.

Bourne, L. S. (1975) Urban Systems: Strategies for Regulation. Oxford: Clarendon Press.

Bourne, L. S. and Simmons, J. W. (eds) (1978) Systems of Cities. New York: Oxford University Press.

Castells, M. (1996) The Rise of Network Society. Oxford: Blackwell.

Christaller, W. (1933/66) Central Places of Southern Germany. Englewood Cliffs, NJ: Prentice Hall.

Friedmann, J. (1986) The world city hypothesis. Development and Change 17, 69-84.

Frobel, F., Heinrich, J. and Kreye O. (1980) The New International Division of Labour. Cambridge: Cambridge University Press.

Frujita, M. and Thisse, J-F. (2002) Economics of Agglomeration. Cambridge: Cambridge University Press.

Goerzen, A., Asmussen, C. G. and Nielson, B. B. (2013) Global cities and multinational enterprise location strategy. Journal of International Business Studies 44, 427-50.

Glaeser, E. L. (2012) The Triumph of the City. London: Macmillan.

Glaeser, E. L., Kallal, H. D., Scheinkman, J. A. and Shleifer, A. (2002) Growth in cities. Journal of Political Economy 100, 1126-52.

Henderson, J. V. (1987) Systems of cities and inter-city trade. In P. Hansen, M. Labbe, D. Peeters, J-F. Thisse and J. V. Henerson (eds) Systems of Cities and Facility Location. Chur: Harwood Academic Publishers.

Hicks, J. (1969) A Theory of Economic History. Oxford: Clarendon Press.

Hill, R. C. and Fujita, K. (2003) Osaka’s Tokyo problem. International Journal of Urban and Regional Research 19, 181-93.

Hymer, S. H. (1972) The multinational corporation and the law od uneven development. In Bhagwati, J. (ed) Economics and World Order. London: Collier-Macmillan.

Jacobs, J. (1961) The Death and Life of Great American Cities. New York: Random House

Jacobs, J. (1969) The Economy of Cities. New York: Vintage.

Jacobs, J. (1984) Cities and the Wealth of Nations. New York: Vintage.

Jacobs, J. (2000) The Nature of Economies. New York: Vintage.

Kratke, S. (2001) Berlin: Towards a global city? Urban Studies 38, 1777-99.

Krugman, P. (1987) Development, Geography, and Economic Theory. Cambridge, MA: MIT Press

Lai, K. (2012) Differentiated markets: Shanghai, Beijing and Hong Kong in China’s financial centre network. Urban Studies 49, 1275-96.

Losch, A. (1940/54) The Economics of Location. New Haven, CN: Yale University Press.

Markusen, A. (1996) Sticky places in slippery space: A typology of industrial districts. Economic Geography 72, 293-313.

Nowlan D. (1997) Jane Jacobs among the economists. In M. Allen (ed) Ideas that Matter: The Worlds of Jane Jacobs. Owen Sound, ON: The Ginger Press.

Page, M. (2005) The city’s end: Past and present narratives of New York’d destruction. In L. J. Vale and T. J. Campanella (eds) The Resilient City. New York: Oxford University Press.

Porter, M. E. On Competition. Brighton, MA: HBS Press.

Pye, M. (2014) The Edge of the World. London: Penguin.

Sassen, S. (1991) The Global City. Princeton, NJ: Princeton University Press.

Sassen, S. (1994) Cities in a World Economy. Thousand Oaks, CA: Pine Forge.

Sassen, S. (1999) Global financial centers. Foreign Affairs 78, 75-87.

Smith, R. G. (2012) NY-LON. In B. Derudder, M. Hoyler, P. J. Taylor and F. Witlox (eds) International Handbook of Globalization and World Cities. Cheltenham, UK: Edward Elgar.

Taylor, P. J. (2001) Specification of the world city network. Geographical Analysis 33, 181-94.

Taylor, P. J. and Derudder, B. (2015) World City Network. London: Routledge.

Taylor, P. J., Derudder, B., Faulconbridge, J., Hoyler, M. and Ni, P. (2014) Advanced producer services as strategic networks, global cities as strategic places. Economic Geography 90, 267-91.

Taylor, P. J., Hoyler, M. and Verbruggen, R. (2010) External relation process: Introducing central flow theory to complement central place theory. Urban Studies 47, 2803-18.

Taylor, P. J. and Lang, R. E. (2005) US cities in the world city network. Metropolitan Policy Program Survey Series. Washington, DC: Brookings Institution.

Thrift, N. (1999) Cities and economic change: Global governance? In J. Allen, D. Massey and M. Pryke (eds) Unsettling Cities. London: Routledge.

Yeandle, M. (2016) The Global Financial Centres Index 20. London: Z/Y Group Ltd.

Zucman, G. (2015) The Hidden Wealth of Nations. Chicago: University of Chicago Press.

NOTES

^* Peter J. Taylor, (email: crogfam@gmail.com)