GaWC Research Bulletin 373

GaWC logo
  Gateways into GaWC


Mega-Cities in Theoretical Perspective

P.J. Taylor *


The concept of mega-city is a very simple one. Developed by UN institutions to describe ever-larger urban agglomerations, they are defined as cities with populations above a given high threshold. The latter has increased as city sizes have grown and currently the threshold is 10 million. Today this defines 26 cities as ‘mega’ and these are listed in Table 1. To put this into historical perspective, in 1900 there were no cities that would qualify as mega. In fact there were only 16 cities with populations over one million (Chandler 1987, 492); today there are 477 ‘millionaire cities’ (, accessed 21-02-11)!

Clearly the twentieth century has been a period of massive urbanization. But it is not just the sizes of cities that have changed; the geography of largest cities has fundamentally altered. Of the 16 millionaire cities in 1900, half were from the UK and USA (four each), with three more from continental Western Europe, two from Russia, and one each from China, India and Japan. In Table 1 only six cities are from the old economic core-region (two each from USA, Western Europe, and Japan) of the world-economy: half of today’s mega-cities are in the developing countries of Asia (13), with the remainder from Latin America (4), Africa (2), and Russia. Of course this geographical realignment has been long recognised and has been associated with a break in the determinants of large urban population growth (Castells 1978). The nineteenth century had also been identified as a period of massive urbanization (Weber 1899) seen as a consequence of large-scale industrialization. But large city growth in the twentieth century displayed a disjuncture between demographic growth and economic development: most of the cities listed in Table 1 were not the important hubs of the world-economy through the twentieth century. This ‘causal break’ is usually dated to around the middle of the twentieth century and is associated with increasing rates of urbanization leading to the situation in the early twenty first century where urban dwellers constitute a majority of humanity. 

The twenty first century will be an urban century: sometime in the middle of the century over three-quarters of the world’s population are expected to be living in cities. At the very least this means that the world academy should take theories about cities very seriously if we are to understand our new, historically unprecedented, urban world. But the uniqueness of our times should not preclude intelligent inquiry of how we got ourselves into this position. In this short essay I introduce one way of satisfying these urgent needs by drawing on the oeuvres of Jane Jacobs and Immanuel Wallerstein; this essay is part of an on-going project that attempts to reconcile their divergent and often opposing premises.  The argument is developed in five sections. First I begin with migration as the motor of urbanization; it is the changing nature of migrations that have produced our increasing urban modern world. Second, I present Jacob’s ideas on cities as process to provide a means of understanding city growth. Third, I present Wallerstein’s ideas as to why massive and increasing urban growth has characterised our modern world. Fourth, these two theorised processes are brought together to offer a perspective on mega-cities and twenty first century urbanization. It is the latter that is the focus of the final section identifying current processes and considering their future implications.

Migration, cities and the modern world-system

Historically, cities have been great demographic sinks: they lessened life expectancies. Thus cities only grew through migration. In other words rural-urban migration has been, literally, the lifeline of cities. And before the modern era there have been very large cities: three in particular can lay claim to be the first ‘millionaire city’: Imperial Rome, Caliphate Baghdad, and Ch’ing Peking. And these examples clearly indicate the forces behind growth of the largest cities. These great cities are reflections of the political power wielded by large world-empires; the centralization of power is accompanied by the concentration of wealth requiring in-migration to service both state function needs and satisfy market consumption wants. In these societies there were non-political cities of reasonable size with mainly economic functions – largely coastal and river ports – but they never challenged imperial capitals for sheer size.

However, this dominance of political cities was challenged with the transition to the modern world-system in Europe in the long sixteenth century (c. 1450-1650). Although Europe did not have a dominant world-empire before 1450, and its urban trajectory was led by the commercial cities of northern Italy, nevertheless Europe’s largest city at the peak of the ‘commercial revolution’ in 1300 was not Venice, with an estimated 110,000, but a political city, Paris, capital of the largest kingdom, which was more than twice this size with some 228,000 residents (Chandler 1987, 17). However in the transition there is the beginning of a new pattern with economic cities within the hegemonic state dominating urban growth (Taylor et al 2010). In this case, cities in Holland, led by Amsterdam, which was not then a capital city, show very fast growth rates relative to the rest of Europe (Israel). This is the first modern example of economic forces rivalling political forces as a maker of cities. This is exemplified by the British industrial revolution where Birmingham, Glasgow, Liverpool, and Manchester are the fastest growing cities of the eighteenth century (Taylor et al 2010). By 1900 seven of the 16 ‘millionaire cities’ previously referred do not have state capital functions: Birmingham, Boston, Chicago, Glasgow, Manchester, New York, and Philadelphia. All the latter cities grew by attracting migrants for the economic opportunities that were perceived as being available. The remaining millionaire cities at this time are all capital cities but ones that were themselves rapidly industrializing such as Berlin, London, Paris and Vienna. This is the first modern effect on urbanization that is unprecedented in history: the creation of great industrial cities as described by Weber (1899).

The second modern effect on urbanization unprecedented in history is the rise of mega-cities in the twentieth century. Whereas the rural-urban migration for both political and economic cities was largely based on the pull of the cities, urban opportunities seemingly outweighing the attractions of staying rural, with mega-cities the situation is much more complex. Certainly push factors relating to reorganizations of rural worlds seem to be just as important as city pull factors. This is especially the case where the mega-city in poorer countries is unable to provide anywhere near the formal jobs to match in-migration. The result has been the production of what Davis (2006) calls ‘mega-slums’ including many ‘millionaire slums’: first, second and third generation urban slum dwellers are becoming a major global demographic (Brugmann 2009). I argue that you need a combination of Jacobs’ city generic process and Wallerstein’s systemic modern process to understand what has been going on and to consider implications for the twenty first century.

Jacobs’ city generic process

Here I summarize the way in which Jacobs (1969, 1984) conceptualizes cities, personally embellished to highlight networks.

  1. The first key point is that cities are seen a process, as multi-scalar, multi-networked bundles of mechanisms.

  2. She focuses on economic mechanisms by which she means work.

  3. In this way she distinguishes between two types of city growth: expanding existing work (‘old work”), for instance increasing the size of a factory in a city, and generating ‘new work’ by introducing production novel to a city.

  4. The latter is the key to economic expansion because it results in enhancing the complexity of the city’s division of labour. This is city cluster formation and, for the city as a whole, produces agglomeration advantages (access to transient, tacit and dynamic knowledge).

  5. New work can come in various ways with two being particularly important: (a) spin-offs from existing old work to create new products; and (b) introduction of new work by copying work in other cities.

  6. Slow, gradual economic growth occurs when exports are increased through getting more out of old work.

  7. Rapid, ‘explosive’ economic growth happens when imports from other cities are replaced and shifted through new work. This is city network formation.

  8. The latter is what creates great cities; it is a generic process that has seen city networks lead economic expansion throughout history.

  9. But it is not all positive: economic growth requires material, labour and capital inputs that are taken from non-city places.

  10. She describes ‘five great forces’ that are continually remoulding the economic landscape for city needs: (a) market demand for raw materials, energy and food creates primary supply regions; (b) transplanting old work creates secondary and tertiary supply regions; (c) demand for labour is a pull factor in rural-urban migration; (d) new uses of rural land with less labour input produces ‘clearances’; a push factor in generating more labour in the city; (e) surplus city capital requires projects for investment (e.g. unnecessary dams) and capital circulation.

  11. In the local city-region these forces operate together and complement one another to extend the scale of the city process of economic expansion. This is city-region formation, which enhances advantages of agglomeration.

  12. However beyond the local, these forces tend to operate singly and destructively to create what Jacobs calls ‘economic grotesques’. In contrast to complex city economies these are simple economic arrangements of dependency and exploitation.

I call these processes ‘city-ness’ (Taylor 2006, Taylor et al 2010b); the mechanisms are graphically displayed in Figure 1. The overall process is generic: it works in ancient Mesopotamia, in ancient Zimbabwe, in Middle American Maya cities, in Ch’ing China, in British industrial cities, and in today’s world cities. But as such, it does not explain why urbanization built upon economic growth expanded in the modern era in ways never before experienced. For this we need to understand another macro-process that defines the specificities of modernity.

Figure 1: Generic Processes of City-ness

Figure 1

Wallerstein’s systemic modern process

This is my summary of how Wallerstein (1979, 2004) treats modernity, personally embellished to highlight cities.

  1. The basic organizations of social life are historical systems.

  2. They are historical because they have a beginning (transition from what was before), a period of ‘normal’ reproduction, and an end (transition to what follows); they are systemic because they have a distinctive ‘normal’ social mix of processes.

  3. There are two types of large (beyond local) systems: world-empires dominated by political elites and ‘world-economies’ where there is a balanced relationship between economic and political power. These are world-systems, not in the global worldwide sense, but as indicating distinctive social ‘worlds’.

  4. There have been many world-empires, such as Rome, Inca and the succession of Chinese world-empires, but only one successful (i.e. reproducing) world-economy: the modern world-system.

  5. As indicated earlier, the transition to this modern world was the ‘long sixteenth century’; it has subsequently been reproducing ‘normally’ through hegemonic cycles (Dutch mercantile modernity, British industrial modernity and US consumer modernity) for about four centuries and is now in its demise phase.

  6. The modern world-system’s normal mix of social processes pivots around ceaseless capital accumulation, an economic process that rewards highest profit taking: it is a capitalist world-economy.

  7. This capitalist world-economy process has created distinctive social times and social spaces.

  8. The social times of the modern world-system are multiple cycles of varying time span; the largest being Kondratieff (about 50 years), hegemonic (100-150 years), and the total systemic cycle (about half a millennium).

  9. The social spaces of the modern world-system are a core-periphery structure with core processes featuring relatively high tech, high value labour production and periphery processes featuring the obverse.

  10. In addition there is an inter-state system through which political power is challenged and whose integrity (multiple sovereignties) has been maintained by the hegemonic states, which has ensured no reversal to a world-empire (single sovereignty).

  11. It is in this modern world that cities were freed from world-empire restrictions to forge ahead and create an ever-expanding core zone replenished by an equally expanding periphery zone: hegemonic cycles were central to this process with their clusters of rapidly expanding cities (Holland’s cities in the late sixteenth and seventeenth century, northern British cities in the late eighteenth and nineteenth centuries, and American ‘manufacturing belt’ cities in the late nineteenth and twentieth centuries).

  12. But expansion cannot last forever; we are using up the Earth, which is why the twenty first century marks the demise of the system: a necessary transition to something new.

This is the specificity of the modern world-system as capitalist world-economy. This is the framework in which I suggest we can understand contemporary mega-cities as specific examples of city-ness in the demise phase of the modern era.

Mega-cities in a model of intersecting processes

Mega-city formation in the twentieth and current centuries is the result of an intersection between the generic city-ness process and the specific modern systemic process. I will show this process relationship at two levels, general and particular.

Generally, the coming of the modern world-system has accentuated city-ness. Political restrictions, inherent to world-empire structures, have been less obstructive and destructive as an emphasis on territory and place has been challenged by the necessity of flows for a vibrant economic process. This is shown in Figure 2 where all the dynamic mechanism arrows have been made more prominent. The culmination of this today is positively shown by the explicit generation of increasing ‘world-class’ agglomeration gains and city clusters (Fujita and Thisse 2002; Sassen 2001), by creation of an intensive world city network (Taylor 2004), and by development of global city-regions (Scott 2001; Hall and Pain 2006). But on the negative side, the arrows of exploitation have been equally enhanced as first the ‘new international division of labour’ followed by neo-liberal globalization has intensified both demands on the Earth (hence entering the demise) and transnational opportunities for super-capital accumulation through the five great forces. It is here that I locate the rise of mega-cities in poorer countries.

Particularly in Figure 2, mega-cities result from a combination of enhanced demand from primary supply regions with push factors favouring clearances of unwanted rural labour. For all of its history the modern world-system has acted as a great ‘social engine’ undermining peasant economies as cities continually remake rural areas for their ever-changing needs (Vanhaute 2008). It is this ‘de-peasantization’ that has come to ahead in the last half-century or more to create mega-cities. However the cities whose demands are primarily responsible for the final ‘global’ remaking of the rural are from the core-zone of the capitalist world-economy, whilst most of this remaking is happening in the non-core zones of the world-economy. But because of the mosaic of boundaries that is the inter-state system, most migration remains within states, with relatively little ‘international migration’. The end-result is the creation of mega-cities in poor countries where local city-dynamism cannot keep up with unprecedented levels of rural-urban migration. The key outcomes are mega-city slums as shown in Figure 2.

Figure 2: Processes of City-ness in the Modern World-System

Figure 1

This combining of Jacobs’ and Wallerstein’s ideas provides a theoretical perspective, a reasoned framework, for understanding mega-cities in contemporary globalization. But the reality will always be much more complex than such modelling allows for. In Figure 2 I have allocated mega-city slums to the category ‘excluded places’. This locates them outside the normal flows of the capitalist world-economy. This means that the creativity inherent in large city agglomerations is not able to generate a Jacobsean growth process, but rather has to be used for an economics of survival within an informal economy. I have argued that more creativity is required to survive in Kinshasa or Karachi than is needed to generate banking profits in London of New York (Taylor 2010). But this is not the only way to interpret the economics of mega-slums. There is growing evidence that new waves of economic growth are being generated in these seemingly inhospitable places: Neuwirth (2006) shows the new urban creativity in a much more positive economic light; Brugmann (2009) sees the making of new urban revolutions. From the perspective of this essay we can argue that this as a contest between Jacobs’ generic urban process and Wallerstein’s structural peripheral process. If the evidence continues to grow that Jacobsean processes are happening in these least likely places in the capitalist world-economy, this would signal a critical erosion of the modern world-system’s ‘normal’ reproduction and therefore provide clear support for Wallerstein’s assertion that we are in the demise phase of the system.

Implications for the twenty first century

The ultimate purpose of developing this new theoretical perspective is to provide possible intellectual tools for coping with the challenges looming ahead of us in the twenty first century. Obviously we must go beyond labelling the century as ‘demise phase’, our geohistory has to be pressed into understanding key processes in the present and speculate on key processes of the near-future. For the former I will briefly discuss the contemporary rise of China and its cities; for the latter I will consider city/state relations with respect to confronting climatic change.

Both Beijing and Shanghai feature as mega-cities in Table 1. More generally, in the last three decades, China has experienced the greatest rural-urban migration ever recorded resulting in massive urbanization: there are currently more than 50 ‘millionaire cities’ in China. But China does not feature in Davis’ (2006) list of leading mega-slums. Relative to mega-cities in other poorer countries, China seems to have coped well: in this case economic development has accompanied the demographic revolution. How has this been achieved? It would seem that the Chinese communist government has managed to open up their economy to world-economy processes while at the same time keeping control of their own domestic economic process. Their remarkable accomplishment has been to generate a supply of jobs that keeps up with the in-migration to cities. Such a huge task requires a global reach. They have kept up through what I have termed a ‘labour foreign policy’: taking jobs away from other poor countries across the world (Taylor 2011). The result has been the creation of more ‘new work’ than seemed possible in an export-led growth policy that is currently being transformed into an import replacement process: from a Jacobsean position, China’s ‘take off has only just begun! But this has almost certainly only been possible because of the size of the Chinese state and it is highly unlikely that it can be repeated (by, say, India). I interpret it as the last great imperialism of the modern era. From Wallerstein’s position, this unexpected and unusual ‘peripheral’ economic success is further evidence of the breakdown of normal reproduction of the system, which is therefore in demise.

Table 1: Mega-cities (as of 01-01-11)



Population (millions)


Mexico City
New York
Sao Paulo
Los Angeles
Buenos Aires
Rio de Janeiro


Source: The Principal Agglomerations of the World (, accessed 21-02-11)

The great travail of the twenty first century looks likely to be climatic change. The evidence suggests a series of scenarios all of which indicate the need for a social revolution: our consumer modernity will not survive, that is what demise means. But how do we get to a sustainable future? What is the path to a viable new world? My first response is that I would not start from here. This unfeasible pessimism is not just a reaction to ever more consumption fuelled by Chinese production but relates to a structural argument. The politics of the modern world-system operates through an inter-state system, a form of global governance that appears to be especially designed to make worldwide policy agreement impossible. Quite simply, the politics is inward looking towards the domestic needs of each state so that the free-rider potential scuppering any ‘global policy’ is immense. Hence, we experience on-going failure in policymaking to combat climatic change. Since this is a structural problem inherent to our world we should not expect political solutions to combat climatic change until it is too late. Hence we urgently need a Plan B. Inter-state negotiations will continue and may ameliorate the worse change, but serious environmental changes there will be. In Plan B the impossible conservative goal of sustainability is replaced by a more radical goal of resilience. We need to bounce forward to a new world. And this will be accomplished not through states but through cities. Historically cities are much more resilient that states; the latter come and go while cities remain and adapt: most cities are much older than the states they are currently part of. A resilient future may be built upon the creativity in cities, especially through harnessing the import replacement process to generate a more localised global world.


Brugmann, J (2009) Welcome to the Urban Revolution: How Cities are Changing the World. New York: Bloomsbury

Castells, M (1978) The Urban Question; a Marxist Approach. London: Edward Arnold

Chandler, T (1987) Four Thousand Years of Urban Growth: an Historical Census. New York: Edwin Mellen

Davis, M (2006) Planet of Slums. London: Verso

Fujita, M and Thisse, J-F (2002) Economics of Agglomeration: Cities, Industrial Location, and Regional Growth. Cambridge: Cambridge University Press

Hall, P and Pain, K (eds) (2006) The Polycentric Metropolis. London: Earthscan

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

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

Neuwirth, R (2006) Shadow Cities: a Billion Squaters, a New Urban World. London: Routledge

Sassen, S (2001) The Global City: New York, London, Tokyo. Princeton, NJ: Princeton University Press

Scott, A (ed) (2001) Global City-Regions. Oxford: Oxford University Press

Taylor, P J (2004) World City Network: a Global Urban Analysis. London: Routledge

Taylor, P J (2006) ‘Cities within spaces of flows: theses for a materialist understanding of the external relations of cities’ in P J Taylor, B. Derudder, P Saey and F Witlox (eds) Cities in Globalization. London: Routledge

Taylor, P J (2011) ‘Thesis on labour imperialism: how communist China used capitalist globalization to create the last great modern imperialism’ Political Geography 30,

Taylor, P J, Firth, A, Hoyler, M and Smith, D (2010a) ‘Explosive city growth in the modern world-system: an initial inventory derived from urban demographic changes’ Urban Geography 31, 865-84

Taylor, P J, Hoyler, M, and Verbruggen, R (2010b) ‘External relation process: introducing central flow theory to complement central place theory’ Urban Studies 31, 2803-18

Vanhaute, E (2008) ‘The end of the peasantries? Rethinking the role peasantries in a world-historical view’ Review (Fernand Braudel Center) 31, 39-60

Wallerstein, I (1979) The Capitalist World-Economy. Cambridge: Cambridge University Press

Wallerstein, I (204) World-Systems Analysis: an Introduction. Durham, NC: Duke University Press

Weber, A F (1899) The Growth of Cities in the Nineteenth Century: A Study in Statistics Ithaca, NY: Cornell University Press [1963 reprint]


* Peter J. Taylor, School of the Built and Natural Environment, Northumbria University, UK; email:

Edited and posted on the web on 28th February 2011