GaWC Research Bulletin 52

GaWC logo
  Gateways into GaWC

This Research Bulletin has been published in A Rogers and H A Viles (eds) (2002) The Student's Companion to Geography 2nd ed. Oxford: Blackwell, pp. 113-7.

Please refer to the published version when quoting the paper.


New Political Geographies 'Twixt Places and Flows

P.J. Taylor

The USA is a nuisance. I mean this in a loosely legal sense. In English law behaviour constitutes a nuisance when it causes serious discomfort or inconvenience to others: a factory emitting noxious fumes over a neighbourhood is a dangerous nuisance. Extrapolating to the global scale, the USA is by far the world's greatest polluter who, by refusing to ratify the Kyoto Agreement on restricting greenhouse gas emissions, condemns the rest of the world to suffer the effect of its excesses. The USA is able to conduct itself in this obnoxious way because of its political power. As 'lone superpower' it looks after its own supposed interests first, and only then considers the consequencies for others.

Such selfish behaviour is normal in the traditional conduct of international relations. In the competitive world of rivalry between states, each has had to look to its own ends and use any means available to satisfy them. Hence the twentieth century was a century of many wars both great and small. On several occasions during the Cold War this nearly led to nuclear catastrophe, not just for the USA and USSR, but for the whole of humanity. That dangerous world is now past. A divided political world has been succeeded by an integrated economic world: the Cold War has been replaced by globalization. In the new circumstances, the behaviour of states is much more circumscribed by the world economy than heretofore. Hence, rather than using the term lone superpower, a throwback to Cold War terminology, the USA's nuisance behaviour might be better described as the actions of the last state with a large degree of national sovereignty (Martin and Schumann 1997, p. 216) in a new globally connected world. This is a different sort of world, one where we need to think about politics and space in new exciting ways.


The traditional study of political geography has focussed on the spatial structure of states and their aggressive competitive for territory. This was to view the world as a mosaic of political spaces, sovereign state territories, which the Cold War simplified into a three-part mosaic of 'first', 'second' and 'third' worlds. Such 'mosaic thinking' is no longer appropriate for understanding political geography under conditions of contemporary globalization (Taylor 2000, 2001).

A new spatial framework is required. This has been provided by Castells' (1996) in his depiction of a new global 'network society'. His analysis depends upon a distinction between 'spaces of places' and 'spaces of flows'. The mosaic of territorial states is an example of the former, the global transactions of financial markets are an example of the latter. Geographically, globalization is a result of a change in the balance of importance between these two different types of space. Combining the technologies of communications and computers has facilitated massive growth in connections and linkages across the world and in the process has often undermined the territorial integrity of states. A good example is the Internet where governments are finding it very difficult to censor material within their own territories. The rise of network society is built upon such spaces of flows.

It is important not to think that spaces of flows are replacing spaces of places. Both are, and always have been, necessary for the each other's existence. The key word in the above discussion is, therefore, balance, reflecting the changing saliences of the two types of space. Thus network society does not mean 'the end of the state', states remain important institutions embedded in our modern world. Globalization is forcing states to restructure their activities as they are seriously challenged in some key functions (e.g. regulating financial markets). The development and deepening of regional institutions such as the North American Free Trade Association and the European Union are the clearest examples of spatial restructuring, providing 'havens' for states in the threatening turmoil of the world economy. In this way states remain important but they are no longer omnipotent in international relations. World cities have emerged as important rivals in many ways. As nodes in a world city network, the major cities of the world have found a new lease of life in network society. Little more than two decades ago great cities such as New York were seen as massive political and economic problems, losing population and jobs and going bankrupt. Today New York is an archetypal global city (Sassen 2001), one of the most important nodes in the world city network. This remarkable turnaround in the fortunes of large cities is probably the key geographical effect of contemporary globalization.

New political geographies for the twenty-first century, therefore, will be about how power in the world-economy is organised between spaces of places (notably states) and places of flows (notable cities). I will illustrate this by taking a fresh look at three topics that have been central to traditional political geography - boundaries, capital cities and federalism - to show how they fare in a political geography beyond simple mosaics.


Sovereign boundaries define territorial states. They are where flows - people, commodities, money - into and out of the state are traditionally controlled. But with globalization, boundaries are being seen less as spatial barriers and more as places of contact. Thus communities that were once resigned to being at the edge of the all-important state are now developing the idea of being the meeting place of nations.

Such opportunities are particularly evident in the European Union (Murphy 1995) and one particularly interesting example is along the north-west Mediterranean shore that straddles France and Spain (Morata 1997). Here there are two rival organisations promoting cross-border co-operation. The first is a regional alliance between Catalonia and its two French neighbours, Languedoc-Roussillion and Midi-Perenes. This 'West Mediterranean Euro-region' defines a cross-border territory. The second is the C-6 network, an alliance of six cities, four in Spain (Barcelona, Palma de Mallorca, Valencia and Zaragoza) and two in France (Montpellier and Toulouse). This defines a cross-border network. Both arrangements are promoting cross-border economic development through political co-operation but they are based upon very different forms of spatial organisation. The Euro-region is a traditional 'space of places' strategy operating through a defined area, the C-6 network is an innovative 'space of flows' strategy operating through local nodes in the European city network. This is probably the clearest example of these alternative organisations of space operating in the same region. And yet, the two organisations compete rather than co-operate, an expression of the places-flows tensions we can expect in the political geography of the twenty-first century.


The importance of capital cities grew with the increasing power of the nation-state. By the end of the nineteenth century the great cities of Europe were the 'imperial capitals' of the great powers of Europe: London, Berlin, Paris and Vienna. These cities totally dominated their national/imperial territories: they were firmly planted at the top of national city hierarchies with economic growth and cultural flowering complementing their political centrality. This all-encompassing model of the capital city contrasts with the political specialisation model in which the capital city has few functions beyond its capital role. Such capital cities are chosen or developed specifically to promote the territorial integration of the state. Cities selected to replace old capitals are usually more central (e.g. to Moscow from St Peterburg, to Ankara from Istanbul, to New Delhi from Calcutta); new capital city developments are usually located as spatial compromises (e.g Washington, DC between North and South, Canberra between Sydney and Melbourne, Ottawa between Ontario and Quebec, Brazilia between Amazon region and the rest of the country). How do these different capital city strategies fare under new conditions of globalization?

Specialising in national politics is not a good springboard to take advantage of the new opportunities of globalisation. Locked into a territorial politics in a space of places, most specialist capitals have not fared well. Ankara is no match for Istanbul in the global stakes and the same can be said for Brazilia (with respect to Sao Paulo), Canberra (with respect to Sydney), Ottawa (with respect to Toronto), New Delhi (with respect to Mumbai). In contrast the all-encompassing capital cities have been able to convert to world city status relatively easily. London, Tokyo and Paris are prime examples. Where this has not happened, specific reasons relate to outcomes of twentieth century wars: Vienna as a victim of the defeat of Austria-Hungary in World War I and Berlin as a victim of the defeat of Germany in World War II.

However there are two specialist capital cities that have been able to make the transition to world city status. Both are unusual examples. First, Washington, DC has gradually broadened its functions and climbed the US urban hierarchy (Abbott 1996). Being the capital city of a twentieth century superpower with by far the largest national economy in the world created a situation for the growth of a capital city economy for firms needing to mix domestic and international contacts (e.g the lobbying activities of global law firms). Second, Brussels, the unofficial capital of the European Union as the home of its most important institutions, was chosen on compromise grounds (being in a small and unimportant country, Belgium). However, as the decision making centre of the largest economic-market area in the world, Brussels has attracted many corporations who wish to operate their European activities from near the political centre of the continent (Elmhorn 2001). Although Washington does not rival New York, Chicago or Los Angeles, and Brussels does not rival London, Paris and Milan, both cities have developed world city status as distinctive nodes in the world city network where economic and political flows meet in an interaction of the national and the global.


Federalism has been termed the most 'geographical' form of state because sovereignty is split between the central state and its federal units. This allows for different territorial interests to be represented countering separatist tendencies within the state. This spatial strategy failed in the aftermath of the Cold War with the demise of federal USSR and the disintegration of Yugoslavia. However these represent very special conditions - the collapse of communism - that put unsustainable pressure on the existing state structures. In the rest of the world federalism has been successful in maintaining the spatial integrity of most large states across all continents including the USA, India, Brazil, Germany, Canada, Nigeria and Australia. Under conditions of contemporary globalization, however, there is an unintended consequence of this spatial flexibility in terms of the configuration of the world city network.

In a federal system there is decentralization of power to lower units whereas in a centralised state the power is concentrated. These two alternative spatial arrangements will create different forms of national urban hierarchies: with the latter one city will be pre-eminent whereas decentralisation creates a situation in which many important cities can flourish. This difference might well be accentuated with globalization since the advantages of national power concentration in one city should translate into a special advantage in global competition. For instance, in highly centralised Japan, it has been argued that the main constraint on Osaka in its aspiration for world city status is its 'Tokyo problem' (Hill and Fujita 1995). In Table 1 the ratio of the connectivity in the world city network (as measured in Taylor et al. (2001) using global-scale business services) between the first and second ranked cities is shown for a selection of federal and centralised states. The contrast between the two types of state is stark: in federal states the first city is about one and a half more connected than its closest rival, in centralised states this increases to ratios of over 3. It seems that not only does Osaka have its 'Tokyo problem', but Manchester has its 'London problem' and other second cities in centralised states are similarly afflicted. The obverse is that federal states are more conducive to more widespread world city formation. The resulting 'unbalanced growth' is leading to demands for decentralisation in some states such as in the UK with its recent devolution and plans for regional assemblies and in Indonesia with the provincial resistance to rule from Jakarta.


These arguments have introduced new ways of thinking in political geography whereby traditional spaces of places are intertwined with new spaces of flows. Under conditions of contemporary globalization, simple competition between territorial states has been combined with the more complex mutual relations between networked cities. This is a crucial agenda for political geographers in the twenty first century but there are others of importance. I have concentrated on just a few selected themes - boundaries, capital cities, federalism - in the context of globalization and there are other exciting research agendas notably in quantitative electoral geography, critical geopolitics and post-colonial political geographies. These indicate the range of topics covered in contemporary political geography. There are two contrasting overviews of the sub-discipline that readers will find instructive: Agnew (1997) has brought together readings to illustrate alternative approaches to political geography, and Taylor and Flint (2000) integrate topics using a single framework to create a world-systems political geography.


Abbott, C. 1996: The internationalization of Washington, D.C.. Urban Affairs Review, 31 (5), 571-94.

Agnew, J. 1997: Political Geography: A Reader. London: Arnold

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

Elmhorn, C. 2001: Brussels: a Reflexive World City. Stockholm: Almqvist and Wiksell International

Hill, R. C. and Fujita, K. 1995: Osaka's Tokyo problem. International Journal of Urban and Regional Research, 19 (2), 181-94

Martin, H-P. and Schumann, H. 1997: The Global Trap. London: Zed

Morata, F. 1997: The Euro-region and the C-6 network: the new politics of sub-national co-operation in the West-Mediterranean area. In M. Keating and J. Loughlin (eds.) The Political Economy of Regionalism, London: Frank Cass, 292-305

Murphy, A. 1994: Emerging regional linkages within the European Community: challenging the dominance of the state'. Tijdschrift voor Economische en Sociale Geografie, 94, ***-***

Sassen, S. 2001: The Global City. Princeton, NJ: Princeton University Press (second edition)

Taylor, P. J. 2000: World cities and territorial states under conditions of contemporary globalization. Political Geography, 19, 5-32

Taylor, P. J. 2001: World cities and territorial states under conditions of contemporary globalization: looking forward, looking ahead. GeoJournal, **, ***-***

Taylor, P. J., Catalano, G., and Walker D. R. F. 2001: Measurement of the world-city network. GaWC Research Bulletin No. 43

Taylor, P. J. and Flint, C. 2000: Political Geography: World-Economy, Nation-State, Locality. London: Prentice Hall (fourth edition)

Table 1: First and second city global connectivity ratios by type of state
Centralised states Federated states
Top two national pairs Ratio Top two national pairs Ratio












New York/Chicago












Edited and posted on the web on 20th August 2001

Note: This Research Bulletin has been published in A Rogers and H A Viles (eds) (2002) The Student's Companion to Geography 2nd ed. Oxford: Blackwell, pp. 113-7