This Research Bulletin has been translated into Italian and published as 'Milano nella rete delle città mondiali' in P. Perulli (ed) (2012) Nord. Una città-regione globale Bologna: il Mulino Studi e Ricerche, pp. 177-194.
Please refer to the published version when quoting the paper.
Introduction: cities, states and Jacobs
The Harvard economist Edward Glaeser (2011, 272) has recently described Jane Jacobs as a person who ‘bestrides the world of cities like a colossus’. In his endorsement of her ideas he is following in the footsteps of the economics Nobel Lauriat Paul Krugman (1995, 5) who offered an equally effusive assessment as ‘patron saint of the new growth theory’. Colossus or saint, Jacobs has had a long record of attracting interest in economics (Nowlan 1997) but I will argue that economist’s use of Jacobs is limited based upon a quite narrow appreciation of her very distinctive oeuvre. Specifically there is an emphasis on her The Economy of Cities (Jacobs 1969), at the expense of her later thinking in Cities and the Wealth of Nations (Jacobs 1984), and with little or no reference to her Nature of Economies (Jacobs 2000). In contrast, it was the middle volume of her trio of economics treatises that I used as the theoretical starting point for my World City Network (Taylor 2004). It is here that she develops more fully her ideas on the external relations of cities, a part of her thinking that seems not to have peculated into the consciousness of economists (Taylor 2009).
The relative neglect of Jacobs’ middle volume is also very important because it is here that she argues very strongly for studying ‘real’ economic systems – city-economies and city-regions – and not the ‘amorphous blurs known as national economies’ (Jacobs 1984, 44). She is very insistent on this:
Her grab bag description appears a ready-made fit for representing the Italian ‘national economy’. With the longevity of its city-economies and relative recency of its political construction, Italy is a classic example where cities should be to the fore in economic analysis. And this does not mean simply using the state administrative definitions of ‘cities’, as economic entities they have to be understood as city-regional economic systems. Statistical analyses of administrative units by economists are therefore problematic; there is an overriding need for economic-geographical definitions of the entities being analysed. In the context of this essay this means prioritizing Milan over Italy in how we understand the role of this city and its region in the world city network.
It is obvious why this position of Jacobs should be particularly attractive to me in my pursuit of understanding contemporary cities in globalization. But Jacobs’ ideas should most certainly not be read as a precursor of globalization theses that envisage the end of states. In a related study (Jacobs 1992), she is very insistent that the commercial work of cities requires some form of governance structures for successful reproduction. Thus although ‘national economies’ only exist in statistics, this does not mean that modern states are not important in the economic process. They may not define ‘economies’ but it is important to a city, which state it belongs to. For instance, the nineteenth century conversion of Venice from a city-state as the nodal apex of the Adriatic region to being on the periphery of first the Austrian state and then the Italian state seems to have sealed the city’s economic decline (Taylor 2007). In the modern world states define ‘spaces of economic regulation’ that are vital to city economies. From business-friendly domestic policies to trade-friendly foreign policy, cities can prosper or decline as the result of state actions. Jacobs’ (1984) concern for the ‘wealth of nations’ strongly features an indictment of how states’ hold back cities to their own economic detriment; Glaesner (2011, 249) similarly includes a plea for states to ‘give cities a level playing field’ in policymaking which he sees as singularly lacking in US policy since 1945.
What does ‘bringing states back in’ to analyses of cities in globalization mean in practise? For a start it does not herald a return to the national urban systems school of the 1960s and 70s that envisaged largely autonomous hierarchies of cities as the separated geographical structures of national economies. But we can study cities in globalization as groups being subject to particular national regulatory regimes. The USA may consist of multiple city-regions as also does China but those city-regions are all very distinctively American and Chinese respectively and this is vital for understanding them within globalization. This is for historical reasons – when and how a country’s cities got encompassed by globalization is critical to their current circumstances. Furthermore, national regulatory spaces define segmented market structures that will be recognised as such by the firms that are makers of city-economies. Since the middle of the last century popular consumption has been largely nationalized through the media of TV, which has been very much a national industry carefully regulated by states (Morley and Robins 1995). The result has been that city-economies of production are linked indelibly to national markets of consumption. This is clearly reflected in the advertising industry where ‘global campaigns’ have to be customised for culture and language to enable selling at the national scale (Faulconbridge et al 2010). For the purposes of this essay it means that we must not be tempted by extreme globalization rhetoric to think we can understand Milan as a city in globalization bereft of its Italian location.
To recognise national influences on cities and their economies while simultaneously eschewing description of national city hierarchies means ditching theory, specifically national urban systems thinking, and reverting to contingencies, the particular placement of cities in national spaces as the result of past political and military exigencies. This includes the simple matter of size of state, something that so exercised Jacobs (1984, 32) – she asked whether the USA, the Soviet Union and Canada could be ‘economic common denominators’ with Singapore, Ecuador and the Netherlands. Quite simply a city in small state must have an open economic policy to survive whereas cities in large states enclosing very many other city-regions are able to prosper under protectionist economic policies as happened in both USA and Germany in the second half of the nineteenth century. However Italy is a medium-sized country and the contingency I will explore in this essay is the situation of countries with rival leading cities. Like Australia (Sydney and Melbourne), Brazil (Sao Paulo and Rio de Janeiro), Canada (Toronto and Montreal), India (Mumbai and Delhi), Spain (Madrid and Barcelona), and to a lesser extent China (Shanghai and Beijing) and Russia (Moscow and St Petersburg), Italy has two leading cities, Milan and Rome. With the coming of globalization, there seems to be a degree of destabilization in the balancing of these national city-duos with the former of each pair listed above seemingly pulling away from its rival as a consequence of the new global economic opportunities. This is particularly noticeable in financial work, often indexed by the closure of rival city stock exchanges, but is more complicated where one of the two cities is the state capital as is the case with the Italian pair. However in the Italian case, like India, but not like Russia and Spain, it is the non-capital city that appears to have taken most advantage of globalization. In his Triumph of Cities, Glaeser (2011, 236-8) features Milan as a classic example of a contemporary city that is succeeding in reinventing itself - ‘roaring back in the postindustrial age’ is how he phrases it (p.236) – while contemporary Rome is conspicuous by its absence in the book. Thus in my exploring of Milan as a city in globalization within the Italian economic space, I will compare the city with Rome to glean a little about what has made Milan relatively so successful today.
Methodology: the interlocking network model
According to Jacobs (1969, 35):
I have used this generic statement to fashion a relational approach to understanding contemporary cities in globalization. In particular, I have melded Jacobs’ insight with Saskia Sassen’s (1991) seminal work The Global City where she finds the leading cities of the world to be marked by growth in provision advanced producer services: cities like New York, London and Tokyo are locales where both producers of, and markets for, these customized financial, professional and creative services are found. I have interpreted Sassen’s global cities as a case of Jacobs’ successful dynamic cities under conditions of contemporary globalization (Taylor 2004, 51-2). The result is a relational approach to understanding inter-city relations as world city network (Taylor 2004).
I specify the world city network as an interlocking network model (Taylor 2001). This is an approach that has three levels to its network: a net level of interacting cities in the world economy; a nodal level of cities through which the network is produced; and a sub-nodal level of advanced producer service firms who are the network-makers. The latter create the network through their routine work in servicing corporate capital using worldwide office networks. It is by measuring and analysing large numbers of such global office networks that the importance of cities can be assessed within the world city network. The basic assumption is that the larger and more important a city office, the more flows – information, instructions, plans, knowledge, video conferences, face-to-face meetings - to other cities are generated. Thus two cities with large offices will have proportionately more ‘potential work flows’ between them than two other cities both with relatively smaller offices. Operationalizing this model through data collection on firms’ office networks (Taylor et al 2002; Taylor 2004) produces a measure of network connectivity indicating the degree of a city’s integration into the world city network.
This methodology has history when it comes to Milan. Network connectivities were first measured for data collected in 2000 and Milan, perhaps surprisingly, was found to be ranked eighth in the world (Taylor et al 2002). According to Sara Gonzáles (2009, 35-6) this finding stimulated a major policy rethink by Milan elites through a new ‘focus and impetus to present a more outward-looking image of Milan as an internationally connected city’ (p. 38) to replace existing parochial thinking. This has subsequently been developed by research projects sponsored by Globus et Locus, a Milan-based thinktank, that have explored the external relations of contemporary Milan. At GaWC (the Globalization and World Cities Research Network – www.lboro.ac.uk/gawc), the original source of the 2000 world city network results, there is a programme of updating this research and in this paper I draw on the results for 2008 (Taylor et al 2011c), and specifically I build upon and extend interpretations of European and Italian case studies (Taylor et al 2011a; Catalano 2011).
The 2008 research project measured offices of 175 firms in 525 cities across the world. The service sectors covered were financial services (75 firms), accountancy (25), advertising (25), law (25), and management consultancy (25). Therefore the connectivity of Milan is derived from the importance of firms’ offices in the city in relation to other cities among the remaining 524 that also have offices of the Milan-located firms. In the event, Milan retains its eighth ranking (Taylor 2011, 24) although it is closely followed by the rapidly rising Shanghai and Beijing. In the next two sections I will interrogate this finding first through focusing on a sectoral disaggregation of the network connectivity, and second, through detailed analysis of Milan’s specific relations with other cities. As mentioned previously, both analyses will feature Milan in comparison with Rome.
Comparing the connectivity profiles of Milan and Rome
In 2008 Rome was far less integrated into the world city network than Milan, ranking thirtieth. However this still indicates that Rome is an important network node and confirms Italy as a country with two leading world cities (Italy’s third and fourth ranked cities are Bologna 162nd and Genoa 231st). In countries with dual world cities the interesting question is whether the services in the two cities mirror each other, implying competition, or differ sufficiently to suggest some element of complementarity. This can be ascertained through disaggregating the connectivity values into their constituent sectoral components.
The service disaggregations for Milan and Rome are shown in Table 1 alongside the disaggregation for combined New York and London (NYLON). The latter provides a sort of ‘global city standard’ service profile for comparison. But before comparing the cities something needs to be said about the different services. The first point to note is that for all three disaggregations, accountancy accounts for more of the connectivity than any other service. And this is despite there being only 25 accountancy firms in the data compared to three times as many financial service firms. Quite simply, this reflects the ubiquity of accountancy firms’ offices across the world – this is by far the most globalised of advanced producer services. In contrast law contributes the least to connectivities; this is because law firms tend to have quite selective city office networks, focusing on just the more important cities. Advertising is similar to accountancy with firms having offices in nearly all countries to tap into their markets but they are more selective than accountancy in focusing on cities housing national media. Management consultancy is the most recent of the services and still has a bias towards the service’s country of origin, USA. However, at the core of all this servicing lies finance, the key constituent of servicing global capital (hence 75 firms used). Financial firms tend to be larger than other service firms and, although retail banking is ubiquitous, our concentration on wholesale banking (i.e. producer services rather than consumer services) means that their worldwide office networks are relatively selective: they have 43% of firms in the data but do not approach anywhere near this level of contribution to the connectivity values in Table 1.
The empirical findings to be derived from Table 1 can be listed as follows:
Interpretation of these findings is relatively clear-cut: Milan and Rome are not mirror images in terms of their global servicing profiles. They appear to be complementary in their linking Italian economic space to the world city network with Milan centred on its financial leadership and Rome taking advantage of capital city functions.
Milan and Rome in leading world city-dyads
Most world city network analysis since 2000 has focussed on the network connectivities of individual cities. However recently we have begun to refocus our work on world-city dyads, specifically on measures of the degree of connectivity between pairs of cities (Taylor et al 2011b). These are, of course, the building blocks of overall network connectivity measures and it seems sensible that in relational analyses special consideration should be given to these dyads. We could go further and suggest that world city dyads are the prime units of world city network analyses, not the individual cities. Presently only NYLON exists as a recognisable city dyad with its own distinctive name, although there has been mention of PAR-LON for Paris-London (Halbert and Pain 2009). I do not think anybody has ever suggested MIRO for the Milan-Rome dyad but this may be the way of the future. In the meantime I will provide some initial results to show what world city-dyad analysis looks like.
From the 2008 data we can produce a 525 x 525 inter-city matrix showing the size of the connectivity between every pair of cities. This produces 137,550 separate city-dyads scores, most of which are quite small; low scores may be suspect because they are derived from an insufficient number of firms. Thus we concentrate on the 570 city-dyads with the largest scores, specifically over 450. A number this size suggests a pair of cities shares about 100 firms that will contribute to their dyad connectivity score. Of these 570 leading city-dyads, Milan and Rome feature in fifty of them. These are listed in Table 2 along with their world rankings.
The first point to note about this list is that Milan features far more often than Rome, in 38 dyads against 13 to be precise. And Milan’s dyads appear higher in the rankings. Both of these finding merely reflect the relative overall connectivity rankings of the two cities (i.e. Milan 8th, Rome 30th). However it does bring home the meaning of a higher overall ranking as more intensive integration into the world city network. But the relative rankings for each city are more interesting. The list in Table 2 begins with Milan’s top six dyad memberships and these directly follow the overall connectivity ranking from London down to Tokyo (Taylor 2011, 24). Rome follows this sequence for the top five (down to Singapore), but then we find Beijing listed as equal to Tokyo. In the general rankings Beijing is 10th, after Shanghai. For Milan its Shanghai dyad ranks above its Beijing dyad. Here we get a hint of the differences in how Milan and Rome differ in their inter-city relations: Milan links more with China’s financial centre, Rome with its capital city.
Before leaving Table 2, I should comment on the Milan-Rome dyad. At first sight it might seem odd that given the two cities are geographically not that far apart and in the same country, their dyad should be larger and therefore rank higher for each city. However, in terms of relative dyad ranking, the dyad shows higher than expected ranking for both cities: in overall connectivity Milan ranks eighth but it is Rome’s sixth ranked dyad; in overall connectivity Rome ranks 30th but it is Milan’s 28th ranked dyad. This is a further hint of the complementarity between the two cities rather than competition.
City-dyad analysis can be taken further using the concept of a city’s hinterworld – the geography of a city’s connections across the world (Taylor 2004). This is, in effect, a map of a city’s dyads. In analyses of the 2008 data we have selected particular hinterworld orientations to specific places to illustrate where a city is ‘over-linked’ and where it is ‘under-linked’. These measures are relative to all other cities and standardized to make comparison easy. Thus an orientation score of zero means that a city’s links to that place is exactly as expected for its overall level of connectivity. It follows that over-linkage and under-linkage show more and less than expected connectivity to a given place respectively. Some results from these analyses are shown in Table 3 for Milan and Rome. Six orientations feature in the analyses, occurring as three pairings. The first orientation is Europe so that the question being asked is what degree of concentration towards other European cities is to be found in Milan’s and Rome’s dyad values? The second concentrations of dyads are measured for relations with other Italian cities. Both of these measures indicate how ‘local’ (first world-region, then country) each city’s hinterworld is within the world city network. The other two pairs of places feature alternative world regions as leading globalization arenas – North America and Pacific Asia, each paired with the key cities of the region – NYLON and the Chinese city-triad of Beijing-Hong Kong-Shanghai. In this argument NYLON represents the existing ‘old’ globalization core, and the Chinese city-triad represents the potential ‘new’ globalization core.
Source: Taylor et al (2011a, 130-34); Catalano (2011, 242-43)
The findings from Table 3 can be listed as follows:
These findings are consistent in one key respect: Milan’s hinterworld is more global than Rome’s. It’s orientation is less European and Italian than Rome’s, and it is more orientated towards North America and Pacific Asia, and towards the rival globalization cores of NYLON and the Chinese triad than is Rome.
Conclusion and caveat
The conclusion to this paper is quite straightforward: Milan remains Italy’s prime world city in terms of advanced producer services. Let us rehearse exactly what this means. Our analyses have measured a particularly important economic process in contemporary globalization: world city network formation. This is interpreted as an ongoing example of Jacobs’ dynamic city growth model and therefore it tells us why Milan has been identified as a current urban success story. Note that we have measured a process not a city. Cities are complex bundles of myriad processes; we have measured just one process, albeit a very important one, but obviously Milan, and also Rome, are so much more than economic growth machines.
Of course, it is not at all surprising that Milan is currently more economically successful than Rome in the world city network. This city network pivots about financial services where Milan is particularly strong: contemporary financialization of the world economy gives Milan a major advantage. But there are many other city networks being formed within contemporary globalization, many where political processes are more salient (e.g. NGOs as city network-makers). Often within these networks Rome appears more important than Milan (Taylor 2005). However, today, it is economic networks to the fore reflecting the dominance of neo-liberal globalization over the last quarter of a century. Nevertheless, it is not inconceivable that as the twenty first century unfolds in an era of multiple crises (energy, climate, inequalities, identities, etc.), political processes might well return to outrank narrow economic priorities. A growing necessity for the some form of global governance could well turn the tables with Rome having new future advantages as a world city in comparison to Milan’s less useful economic prowess.
The data used in this paper was produced by a consortium of researchers at the Department of Geography, Loughborough University (UK), the Chinese Academy of Social Sciences (Beijing), and the Department of Geography, Ghent University (Belgium). The dyad analysis derives from a research project funded by ESRC (UK) at Loughborough University: ‘Benchmarking the world city network: city connectivities on the eve of the current financial crisis’ (Res-000-22-3575).
Catalano, G. (2011) “Italian cities’ in P. J. Taylor et al (eds.) Global Urban Analysis: a Survey of Cities in Globalization. London: Earthscan, pp. 241-44.
Faulconbridge, J., Beaverstock, J., Nativel, C. and Taylor, P. J. (2010) The Globalization of Advertizing: Agencies, Cities and Spaces of Creativity. London: Routledge.
Gonzáles, S. (2009) ‘(Dis)connecting Milan(ese): deterritorialised urbanism and disempowering politics in globalising cities’, Environment and Planning A 41, 31-47.
Glaeser, E. (2011) Triumph of the City. New York: Macmillan.
Halbert, L. and Pain, K. (2010) ‘PAR-LON – doing business in knowledge-based services in Paris and London: a tale of one city?’ GaWC Research Bulletin No. 307(A).
Jacobs, J. (1969) The Economy of Cities. Nrw York: Vintage.
Jacobs, J. (1984) Cities and the Wealth of Nations. New York: Vintage.
Jacobs, J. (1992) Systems of Survival. New York: Vintage.
Jacobs, J. (2000) The Nature of Economies. New York: Vintage.
Krugman, P. (1995) Development, Geography, and Economic Theory. Cambridge, MA: MIT Press.
Morley, D. and Robins, K. (1995) Spaces of Identity. London: Routledge.
Nowlan, D. (1997) ‘Jane Jacobs among the economists’ in M. Allen (ed.) Ideas that Matter: the Worlds of Jane Jacobs. Owen Sound, Ontario: Ginger Press, pp. 111-13.
Sassen, S. (1991) The Global City: New York, London, Tokyo. Princeton, NJ: Princeton University Press
Taylor, P. J. (2001) ‘Specification of the world city network’, Geographical Analysis 33, 181-94.
Taylor, P. J. World City Network: a Global Urban Analysis. London: Routledge.
Taylor, P. J. (2005) ‘Leading world cities: empirical evaluations of urban nodes in multiple networks’ Urban Studies 42, 1593-1608.
Taylor, P. J. (2007) ‘Problematizing city/state relations: towards a geohistorical understanding of contemporary globalization’, Transactions, Institute of British Geographers 32, 133-50.
Taylor, P. J. (2009) ‘Urban economics in thrall to Christaller: a misguided search for city hierarchies in external urban relations’, Environment and Planning A 41, 2550-55.
Taylor, P. J., Catalano, G., and Walker, D. R. F. (2002) ‘Measurement of the world city network’, Urban Studies 39, 2367-76.
Taylor, P. J. (2011) ‘Advanced producer services in the world economy’ in P. J. Taylor et al (eds.) Global Urban Analysis: a Survey of Cities in Globalization. London: Earthscan, pp. 22-39.
Taylor, P. J. Derudder, B., Hoyler, M., Pain, K., and Witlox, F. (2011a) ‘European cities in globalization’ in P. J. Taylor et al (eds.) Global Urban Analysis: a Survey of Cities in Globalization. London: Earthscan, pp. 114-36.
Taylor, P. J., Hoyler, M., Pain, K., and Vinciguerra, S. (2011b) “Extensive and intensive globalizations: explicating the low connectivity puzzle of US cities using city-dyad analysis’, GaWC Research Bulletin No. 369(A).
Taylor, P. J., Ni, P., Derudder, B., Hoyler, M., Huang, J., and Witlox, F. (eds.) (2011c) Global Urban Analysis: a Survey of Cities in Globalization. London: Earthscan
Note: This Research Bulletin has been translated into Italian and published as 'Milano nella rete delle città mondiali' in P. Perulli (ed) (2012) Nord. Una città-regione globale Bologna: il Mulino Studi e Ricerche, pp. 177-194.