POTTER, ROBERT; LLOYD-EVANS, SALLY

The city in the developing world. Chapter 6: The structure and morphology of cities in developing areas: can we generalise? Longman. 1998.

As noted in the previous chapter, many cities in the developing world have undergone major structural transformations in the last 50 years, not least due to the rapid process decolonisation which has taken place in Asia, Africa and the Caribbean. As more people are being drawn closer together through new transport and communication links, and changes in commerce and the media, the notion of global cities has attracted much geographical attention (Hamnett, 1995; Knox, 1996), as stressed in Chapter 3. Global or world cities, as they are also known (Hall, 1966), are those deemed to be places of intensive interaction which undertake major economic, social and political roles in the accumulation of capital. Global city status is ascribed to those cities which are centres of power and where 'a disproportionate part of the world's most important business is conducted' (Hall, 1966: 7).

As noted in Chapter 3, not all global cities are located in the developed world. Although the term is more likely to include cities in North America and Europe, the prominence of Asian cities and the rising importance of some Latin American centres has added a new dimension to theorising about Third World cities. As cities are dynamic entities, undergoing constant transformation according to prevailing macro-economic and social conditions, it is not surprising that the question of how globalisation affects cities in the developing world has received increasing attention in the last decade (Simon, 1993).

In recent years, the issue of globalisation has attracted much interest in urban studies, through an attempt to analyse the relationship between the structural development of global cities and macro-economic policies (Hamnett, 1995). This new agenda has seen a shift away from theorising which focuses solely on the internal processes affecting city structure, to an approach which aims to enhance our understanding of the ways in which global economic and social processes, discussed in previous chapters, influence city structure and the built environment, and the resulting effect on the social and political organisation of the city.

Much of this recent work, however, has been undertaken in the context of the developed world, and has placed emphasis on western inner city decline, decentralisation and suburbanisation (Sassen, 1991). In global cities such as London, Tokyo and New York, the reconceptualisation has focused on the globalisation of finance and property markets, which has resulted in new waves of urban investment in services and the gentrification of derelict inner city areas. Far less work has attempted to investigate the effects of global economic change on developing world cities, which are also affected by changing patterns of capital accumulation, investment and consumption. In this context, the question of how globalisation is affecting the structure of developing world cities remains more elusive.

Developing world cities are indeed changing, but it is the nature of this change which is often misunderstood. On the one hand, there are common perspectives which argue that as a result of globalisation, particularly in relation to global time—space compression, developing world cities are becoming more like their western counterparts, particularly in respect of their consumption patterns, popular culture and the built environment. Evidence to support this argument is apparent in the spread of western consumer goods through 'coca-colonisation', western films, music and multinational enterprises throughout Third World cities across the globe (King, 1991; Sklair, 1991). Such visible outcomes of the globalisation of culture have led many protagonists to argue that western cultures are slowly dominating urban lifestyles, and leading to homogenisation across the developing world. At another level, however, it is apparent that western products and popular culture are not readily accessible to all sectors of the population, an argument reviewed in Chapter 3. The key point here is that developing world cities are experiencing a different type of development as a result of their greater internal inequalities. Indeed, the level of social inequality and rapid growth of developing world cities are two factors which distinguish them from advanced capitalist cities.

There are increasing similarities in the central business districts (CBDs) and residential suburbs of many developing world cities due to modern high rise office blocks, impressive multinational headquarters, and the prominence of western retailers (see Plates 6.1 and 6.2). However, the power of western advertising and multinational dominance, which is so readily accepted as a homogenising force, can be questioned by a closer examination of the ways in which globalisation affects the internal structure and social organisation of the city. For example, whilst trends in some developed world cities have been towards decentralisation, suburbanisation and associated inner city decline, there is little evidence to suggest that this is happening on any major scale in the developing world. Indeed, all too frequently, densities are increasing throughout the city. As the present chapter will outline, city centres in many developing regions are still vibrant in relation to their socio-cultural and economic roles (Ward, 1993).

This chapter, therefore, has two main aims. Firstly, it attempts to re-address the relatively neglected theme of the role of developing world cities in the global economy through an examination of the processes which affect city development. It aims to address the issue as to whether developing world cities are becoming more like western cities due to macro-economic forces. For example, how have new transport systems, modern technologies and global capital accumulation affected cities in the developing world? Secondly, through a detailed study of recent work on urban morphology, the chapter investigates whether we can generalise about the structure of Third World cities. Through a critical examination of a number of models which have presented generalisations about African, Asian and Latin American cities, we explore whether a basic model of Third World city structure can be identified.

The chapter begins with a brief account of the social, economic and political factors which are affecting urban morphology and social organisation. Here, attention is paid to understanding the ways in which global processes have influenced city structure. The discussion then deals with the historical dimensions of Third World city structure, through an analysis of pre-industrial and colonial city forms. A contemporary analysis provides examination of a number of regional city studies and examples, before attempting to draw conclusions. Following on from Chapter 5, the underlying argument focuses on the interactive relationship between urban structure and social process, which has manifested itself in a variety of different urban environments through time.

Cities in the developing world: convergence or divergence?

-Factors affecting city structure in the developing world

There are two main factors which serve to distinguish developing world cities from advanced capitalist cities. The first is rapid recent growth as a result of rural-urban migration and natural population increase. The second factor is a history of marked social inequality and poverty, where extremes of wealth and poverty are often found in close juxtaposition. Centuries of inequality, often as a result of colonialism, have led to the development of distinct traditional and modern urban modes of production in relation to industry, employment, transport and housing (see Chapters 7 and 8). As earlier chapters have elucidated, developing world cities are increasingly diverse. They have their own unique histories of development which have been influenced by different stages of development, from colonial rule through to their current position in the post-colonial global economy. Thus, we should expect urban morphology to be diverse.

As this chapter will demonstrate, colonialism has played a major role in shaping the internal organisation of cities across the developing world. Socially segregated cities were developed by various European powers from the 1500s onwards, as trading or administrative centres, depending on their size and strategic location. In the post-colonial era, however, cities have been affected by their economic role in respect of their national, regional and global positions as industrial, financial or political centres. Of course, we must not underestimate the importance of macro-economic processes, such as the New International Division of Labour and structural adjustment, in shaping the built environment and the social stratification of cities.

Recently, cities in the Newly Industrialising Countries (NICs) of Asia and Latin America have become focal points for Foreign Direct Investment (FDI), manufacturing branch plants and the headquarters of multinational enterprises, as covered in Chapter 3. Growing consumer markets are demanding modern transport and better urban services. City structure also differs according to a country's ability to undertake the structural transformation, in terms of employment, housing and infrastructure, needed to accommodate increased population. The 1980s debt crisis halted many new plans for urban redevelopment along with the expansion of infrastructure.

An often under-explored facet of urbanisation has been the introduction of modern transport systems which fundamentally alter the location of industry and residential areas, and disrupt traditional social structure (Chokor, 1989). Most cities have a heterogeneous range of formal and informal transport systems which serve different sectors of the population (Hilling, 1996; Simon, 1996). As Simon (1996: 93) argues, transport systems 'need to be understood in terms of the history, culture and functions of each city'. For example, the development of international airports, motorways and mass rapid transit systems (MRTSs) has dramatically altered the social and economic structure of some large cities in recent decades.

In considering the influence of globalisation on urban transport systems, Simon (1996) argues that despite the obvious diversity of cities, the transport systems and management policies being adopted in large cities are converging, as they all generally experienced large-scale motorway and rail developments in recent decades. Increasing car ownership has led to widespread environmental pollution and traffic congestion in Asian and Latin American 'mega cities'. Internationally designed, and in many cases highly advanced, metro systems have become a symbol of modernisation in the NICs of Asia and Latin America.

In Singapore, Hong Kong, Sáo Paulo, Mexico City, Caracas and Seoul, to name a few, MRTSs have become a key feature of city expansion. Today, many Southeast Asian cities such Kuala Lumpur and Jakarta are investing huge sums of money in upgrading their current rail systems in preparation for the development of Light Rail Transits (LRTs) which are designed to connect the. CBDs with the suburbs. Planning an affordable, sustainable and comprehensive transport system which is accessible to all sectors of the population represents a major challenge to city planners and governments.

-Globalisation and developing world cities: convergence or divergence?

A useful initial theoretical framework to employ in exploring the changing role of cities in the developing world is that provided by Armstrong and McGee's (1985) influential work which identifies cities as 'theatres of accumulation' and 'centres of diffusion'. As reviewed in Chapter 3 at the urban systems scale, Armstrong and McGee argued that cities play a key role in accumulating capital, international investment, elite populations, modern employment and services. Indeed, the preference of Third World governments and large firms for promoting industrial development in, or close by, large cities has resulted m an important relationship between industrial development and urban structure. Whilst cities provide a plentiful supply of skilled labour and other means of production such as infrastructure, they also house the seats of economic and political power. Cities, however, also play another important role as 'centres of diffusion' with regard to culture, urbanism, modernity, western tastes and consumerism, which promotes further capital accumulation in the city.

This trend is also part of a pattern of 'convergence' in consumption brought about by elite groups, partly related to the introduction of rapid transit systems, the growth of suburban developments and western lifestyles, whereby developing world cities are becoming more 'westernised' (Harvey, 1989; Potter, 1990, 1993b, 1997). Yet, in respect of many other factors such as demography, social structure, housing quality and productive activities, cities in the developing world are becoming increasingly divergent. This argument provides an intra-urban extension to that presented at the urban systems level in Chapter 3.

Thus, an important characteristic of urban structure, both within and between countries in the developing world, is its diversity. On the one hand, urbanisation has resulted in the development of 'mega cities' which have undergone dramatic changes in recent decades, but on the other, smaller cities have also undergone transformations, albeit of a less dramatic nature (Hardoy and Satterthwaite, 1986; Lowder, 1991; Grant, 1995; Gilbert, 1996). It is important to note that a very significant proportion of the rural population in the developing world has migrated to relatively small cities of under 2 million, rather than to the huge '10 million' cities which have captured the attention of the media. Indeed, there are striking contrasts in the scale and extent of city development between continents. As a result, a number of authors have questioned whether a single model of city structure can be widely applied to cities in different parts of the world.

The growth and structure of developing world cities

-Historical development and the pre-industrial city

It has frequently been argued that 'in origin Third World cities and towns -both pre-colonial and colonial- are quintessentially pre-industrial' (Dickenson et al., 1996: 219; Potter, 1985). The concept of the pre-industrial city as a distinct urban form was explored by Sjoberg (1960) through a considered analysis of a number of common characteristics shared by developing world cities before industrialisation. In his model, shown in Figure 6.1, Sjoberg identified three main aspects of land use which distinguished the pre-industrial city from industrial cities in developed nations. Firstly, the urban centre tends to be pre-eminent over the periphery, particularly in the spatial distribution of different social classes, as the elites reside in, or near to, the central core, which housed important religious monuments, administrative centres and symbols of prestige. Here, a noticeable characteristic of the pre-industrial city, as distinguished from North American and European cities, is the absence of a Central Business Distric (CBD). Instead, social patterns in the pre-industrial city are explained by cultural values which identify residence in the historic core as high status. A declining social gradient of residence from the core relegated the more disadvantaged groups to peripheral settlements, a pattern which is maintained in many developing world cities today.

Secondly, these broad social areas were further differentiated into distinct quarters or precincts according to occupational status by guild, or to family and ethnic ties. Sjoberg's third characteristic relates to the lack of functional differentiation in land use patterns in the city, where plots were used for multiple functions such as housing and work. Underlying this pattern of mixed land use was the lack of modern transport which predicated not only the use of residences for work, but also explained the elites' desire to live near to the heart of city life. Other common attributes of the pre-industrial city included the prominence of a walled structure and narrow streets which preserved a distinct social pattern, characteristics which are still visible in many cities built before industrialisation and motorised transport. Kano and Ibadan provide good examples of cities which still display many pre-industrial characteristics, but which are increasingly being brought under pressure from modern development (Plate 6.3, overleaf). In both examples, major roads now pierce the traditional city core in order to create modern access for peripheral communities.

In an informed discussion on Third World city structure, Chokor (1989) also maintains that developing world cities are inherently structured around their pre-industrial foundations, with traditional residential communities surrounding a city core, although these have expanded under colonial and post-industrial development. As a result of modern transport and industrial development, most cities now have congested, decaying cores, but they nevertheless maintain a distinct cultural and social structure which is linked to the built environment. As Chokor (1989: 318) argues, 'undoubted unity, social harmony and bonds which exist between physical form and people is thus a key characteristic of traditional city settings'. Examples of this traditional integrative social design can be seen in most developing world cities, whether in the building of walled homes separated by small pathways to give privacy in traditional Islamic and Indian cities such as Hyderabad, the communal courtyards of western Nigeria or the importance of the 'yard' in the Caribbean. As previously discussed, agents of change in the form of western-style planning, particularly in the building of roads and motorways, threaten traditional community structures and ways of life which are essential to the economic and cultural survival of large sectors of the city.

In assessing the overall applicability of the pre-industrial city model to present-day Third World cities, we can see that elements of a pre-industrial land-use structure still exist. It is important to note, however, that Sjoberg's work was based on relatively very small cities of less than 100,000 residents. In this context, Dwyer (1975) has argued that the pre-industrial social pattern is still observable in many small developing world cities, particularly those in Asia and Africa. But other authors have noted that the pre-industrial concept is fast becoming less typical in the rapidly growing 'mega cities' of Latin America.

In many developing world cities, the desire to reside near the city centre is relatively strong, although suburbanisation has increased in recent decades. Amato (1970, 1971), in a series of studies based on Bogotá, Quito, Lima and Santiago, observed that the elite groups have deserted the original city core and moved to one or more sectors of the outer urban fringe. As this chapter will discuss further, trends of suburbanisation by the elites have also been identified in Caracas by Morris (1978) and in Port of Spain by Conway (1981), Potter (1993b) and Potter and O'Flaherty (1995). This process is highly reminiscent of changes which have taken place in western cities. Of course, the fact that developing world cities havc now been subject to industrialisation, even if the process has not been uniform, adds a further dimension to Sjoberg's model. In summary, therefore, it is apparent that Third World cities do share some of the common attributes expressed in the pre-industrial model of urban city structure. There is, however, a need for more precise explanations of structure which take account of regional diversity and historical development, a task which has been undertaken by other authors (Brunn and Williams, 1983).

-Perspectives on the colonial city

Griffin and Ford (1980) have argued that a contrast exists between thinking about the internal structure of developing world cities and generalised descriptive models of the pre-industrial city and models of modern western cities. It is often presumed that pre-industrial cities will eventually become like western cities, but this line of thought ignores both the distinctive nature of the processes affecting Third World cities, and the diversity of cultural representation in the built environment. One important factor relates to the historical role undertaken by cities in different parts of the globe. In particular, a major criticism of Sjoberg's pre-industrial model is that it fails to differentiate between the diverse colonial experiences of developing nations which comprised distinct social structures (Simon, 1989a).

Horvath (1969, 1972) classified the colonial city as a distinct urban form resulting from the domination of an indigenous civilisation by colonial settlers. European-style churches, palaces and administrative buildings fashioned on the prevailing architectural style of the time are visible in most cities. In particular, it is the unique cultural contact between two civilisations which King (1976) identifies as pivotal. According to Horvath (1969: 76), 'the colonial city is the political, military, economic, religious, social and intellectual entrepôt between the colonisers and the colonised'.

Colonial cities were often developed for their commercial functions through the mixing of European urban forms with indigenous tradition, religious and cultural practices. Some colonial cities have been identified as dual cities following the development of a new European city next to the old indigenous city, as in New and Old Delhi. The processes through which colonial cities developed have been referred to as dependent urbanisation, as cities were reliant on industrialisation located in the metropolitan country (Castells, 1977). The structure of the colonial city is also rather place specific, because it was determined not only by the urban ideology of the colonisers but also by the role of individual cities as agents of imperialism. Some cities were purely administrative centres, whilst others were transport nodes or active in the creation of social surplus product. The structure of the colonial city also differed according to the culture of the colonisers. In the African context, Simon (1989a) states that Portuguese colonial cities such as Luanda and Maputo were difrerent from the British-controlled cities of Lusaka and Harare in respect of their architecture and social organisation.

In respect of internal structure, the colonial settlers rctaincd their position through the development of culturally and ethnically distinct quarters. As a result, colonial cities often maintained interracial social distance which was exemplified in extreme variations in housing quality, the provision of social amenities and access to employment. Elite housing areas for the settlers were often spacious, paved and well serviced, in comparison to the high-density settlements of indigenous populations. Juxtaposed modes of production and employment, consisting of the traditional and modern, remain visible in post-colonial cities today. In fact, it has been suggested by Horvath (1969) and McGee (1971) that cities in Africa and Asia are more likely to have remained traditional than cities in Latin America. As a result of this diversity, a number of authors have attempted to develop regionally specific models of internal city structure.

-The African city

Sub-Saharan Africa is still the least urbanised world region, although rates of urbanisation have increased markedly in the last 40 years. On average, African cities are smaller than those of Latin America and Asia, often consisting of between one and two million residents. It has been argued by a number of authors that there is a lack of conceptual studies which attempt to explain the structure and form of cities in an African context (Simon, 1989a). Much theorising presumes that the African city conforms closely to the pre-industrial model, an assumption which underplays the post-colonial development of African cities. Although McGee (1971) argued that smaller African cities are likely to provide the best examples of colonial city structure, it is important to understand the distinct development of African cities under the various versions of European settlement.

An attempt to develop a distinct model of the African city was undertaken by the United Nations in 1973. As highlighted in Figure 6.2, the model was based on the existence of an indigenous core and the organisation of different ethnic groups through density gradients in a pattern which ascribed low-density land use to the administrative and housing requirements of the colonial elites, and high density to indigenous populations. In Southern and East African cities, segregation was maintained through strict legislation, whereas divisions were less formal in West African cities, due to the smaller numbers of settlers involved.

The most extreme form of social segregation and economic division was found in South Africa's apartheid cities, where the concept of 'separate development' led to the forced removal of the black population from the inner city to townships located on the periphery. Elsewhere in the city, access to urban amenities and employment was segregated according to race and wealth. In the post-apartheid period, attempts to restructure the city and redistribute wealth pose a major challenge for the millennium (Smith, 1992, 1995).

Criticisms of the United Nations' model centre around its failure to recognise the post-colonial transformations which have affected African cities. Post-colonial African cities can be characterised by the mixing of various modes of economic production and housing, such as traditional and modern, informal and formal (see Chapters 7 and 8), which have resulted in diverse structures (O'Connor, 1983). In this context, however, Simon (1989a) questions whether it is still useful to talk about the 'colonial' or 'post-colonial' city. Instead, he suggests that we might turn our attention to examining the influence of macro-economic forces, such as structural adjustment and neo-liberal trade arrangements, on the future development of African cities.

-The South Asian and Southeast Asian City

A great diversity of traditional and modern city forms can be recognised in the great cities of Asia. In particular, there appears to be a distinct difference in the fundamental organisation and structure of the South Asian city compared to cities in Southeast Asia. In South Asia cities such as Delhi, Calcutta, Bombay and Madras were traditional colonial cities built for administrative purposes, or as ports. Although there is no comprehensive model of the South Asian city, it is argued that they are hybrids of a specific colonial model (Brunn and Williams, 1983; Lowder, 1986), as shown in Figure 6.3 (overleaf). Often built as a port location, South Asian cities frequently contained a walled fort which was surrounded by open space for security. They developed western-styled CBDs, around which the European residences were located. Whilst Europeans lived in well-serviced spacious neighbourhoods, the indigenous population lived in ill-planned streets and high-density housing away from the European population. Although there were some intermediate locations of mixed race, it has been argued that this colonial structure still underlies many modern cities in India.

In colonial times, urban areas in Southeast Asia were essentially developed as trading cities with central cores consisting of local bazaars and high-density commercial uses. Although there were marked contrasts between 'western' and 'non-western' sectors of the city, it is argued that Southeast Asian cities have a more mixed and less uniform structure in terms of land use. In an early model of the Southeast Asian city, McGee (1967) paid greater attention to identifying the various ethnic groups that had been involved in restructuring the colonial city (see Figure 6.4). Although there is a western-styled CBD, distinct commercial segments have been developed according to whether the entrepreneurs are Chinese, Indian or European. Although there are zones of squatter settlements along the city's periphery, new industrial zones and agriculture are also located in this semi-rural periphery known as the 'desakota'. McGee and Greenberg (1992) have examined the way in which these extended metropolitan regions are absorbing village settlements and blurring rural-urban contrasts.

Also of great importance is the mixed zoning of residential areas between modern and traditional types of housing and social structure. For example, in Jakarta, spontaneous as well as traditional villages known as 'kampungs' occur throughout the city (Krausse, 1978). Kampungs represent a key component of Indonesian development, and similar traditional structures can be seen in other regions. Jakarta's urban ecology broadly conforms to this model, with mixed inner-city land use and a distinct 'Chinatown' in the west near the port. As in other developing world cities in Africa and Latin America, the elites are starting to suburbanise. This process has been made easier by the development of mass rapid transit systems in many cities.

In recent times, given the economic prosperity of the region, Southeast Asian cities have been the principal centres of growth and planning for social and economic development (Dwyer, 1972, 1995). Singapore, for example, is now classified as an important global city. Southeast Asian cities have been seen as catalysts for change, and they have benefited from urban-centred development policies. Despite rapid technological change, however, the majority of the population still live and work in the non-western areas. This is illustrated by the existence of large areas of indigenous buildings throughout cities, many of which still lack modern infrastructure. Whether the 'economic miracle' will bring about major structural change for the majority of the population remains to be seen. In essence, therefore, the modern Southeast Asian city is still based on a dichotomous form of development, but one which is less characterised by distinct sectional zoning than is the case in the Latin American NICs.

-The Latin American city

As noted in Chapter 1, Latin America has undergone rapid urbanisation and industrialisation in the post-war period (Gilbert, 1990). In 1930 most of the Latin American population lived in rural areas, but by 1980 over half the population was urban. By the millennium, around one-third of the Latin American population will live in cities of over one million in size, a factor which clearly highlights the rapid nature of rural-urban migration in recent decades. Owing to this unprecedented growth and the effort to achieve modern industrial development in most countries, questions over whether Latin American cities are following the same path as western cities have frequently been raised.

Primacy, the prominence of mega cities and rapid growth from rural-urban migration are all characteristic of Latin American cities. Although there are elements of their development which compare with western cities, global economic processes and national decision making have influenced their internal city structure in a distinctive way. Ward (1993) argues that Latin American cities are inherently different from western cities for a number of reasons. Firstly, at the macro level, Latin American cities have a different demographic structure as a result of post-war rural-urban migration. Secondly, the relationship between macro-economic processes and urban structure has resulted in a different set of outcomes. In particular, Latin American cities have not experienced the same cycle of investment/disinvestment that most western cities have. Through the adoption of industrialisation via import substitution in the 1940s and 1950s, to export-oriented production in more recent decades, the higher level of industrialisation in Latin American cities is also a distinct feature. Furthermore, the rapid social polarisation which has been identified as a major effect of globalisation in western cities has always been inherent in the social geography of developing world cities.

As noted earlier, Latin American cities may be regarded as both 'theatres of accumulation and centres of diffusion' (Armstrong and McGee, 1985) which are currently displaying convergence in respect of the wealth and the consumption patterns of elites, and divergence in terms of social inequality and production possibilities. There has been little evidence of gentrification or large-scale urban redevelopment at the micro scale. Although there have been many plans to redevelop Latin American inner cities, processes of recentralisation were largely halted by economic austerity in the 1980s which curtailed public spending and private investment. It is uncertain whether these plans will resurface if Latin American nations prosper in the new millennium.

Latin American cities are also seen to possess vestigial characteristics of colonial pre-industrial cities. In large conurbations, such as Bogota or Sao Paulo, this traditional structure has been superimposed by modern industrialisation, modern transport, the influx of millions of new dwellers and the land-use demands of the elites and emerging middle classes. In an attempt to combine elements of the pre-industrial model with modern developments, Griffin and Ford (1980) devised a model of Latin American city structure in which 'traditional elements of Latin American culture have been merged with the modernising processes altering them'. The original model, shown in Figure 6.5, was based on the Mexican border city of Tijuana, and demonstrates several principal features.

According to Griffin and Ford, the dominant characteristic of the Latin American city is a prominent commercial spine/sector which extends out from the CBD, and which houses the city's most important economic, social and cultural amenities, and a substantial proportion of high-income and wellserviced residences (Plate 6.4, overleaf). Residential areas and facilities such as theatres, hotels, restaurants, prestigious offices, private hospitals, museums and leisure facilities are located on, or near to, a 'tree-lined boulevard', which Griffin and Ford argue is present in nearly all Latin American cities. As shown in Figure 6.5, the elite residential sector is a wedge, in the Hoytian sense and combines western-style amenities with a Latin American desire for centrality. Increasingly, the elites have begun to decentralise outwards to western-style suburbs (Plate 6.5, overleaf).

Away from the spine, there is a series of concentric zones with socio-economic characteristics that are almost opposite to those of western cities. Here, socio-economic levels and housing quality decrease with distance from the city centre. Three distinctive zones can be identified: a zone of maturity, a zone of in-situ accretion, and a zone of peripheral squatter settlements. The zone of maturity refers to rcsidcntial areas which were areas of elite residence, perhaps in colonial times, which have filtered down to middle-income groups when elites moved elsewhere, or 'zones of gradual improvement' where self-built housing has gradually been improved and consolidated over time. These residential areas are also likely to be better serviced than other city districts, with regards to sanitation, transport, electricity and amenities such as street lighting. In many Latin American cities, old colonial residences around the zocalo (colonial plaza) have suffered from a lack of investment and upkeep as the relatively wealthy have moved out along the spine, and such dwellings have become rented tenements (Gilbert and Ward, 1982; Gilbert, 1990).

Unlike the relative stability of the zone of maturity, the zone of in-situ accretion is in a constant state of change as residents move in and out according to lifecycle and status. This zone is not uniform, and it comprises a great variety of housing types of a modest quality, often next to commerce and informal economic activity. Some districts are moving towards maturity, whilst others are the recipients of new developments, often self-built or government-maintained. Provision of services such as water and electricity can vary from street to street. Although Griffin and Ford argued that this zone would gradually improve over time, it is unlikely that dramatic improvements have been made in these areas since the 1980s due to. economic austerity.

On the periphery of the city are zones of squatter settlements which are characteristic of most developing world cities. As will be discussed in the case of Mexico City, some squatter settlements are so large that they could constitute cities in their own right. In addition to these main zones, Latin American cities also contain sectors of disamenity. These are areas of the city which have not been consolidated over time, and are unlikely to be upgraded or improved. Instead, they remain areas of slums and rented tenements such as the 'favelas' of Rio de Janeiro, where new arrivals to the city are often housed. Major industrial and environmentally polluting activities are also likely to be found here.

With the exception of the commercial spine, Griffin and Ford argue that these concentric zones are essentially pre-industrial, suggesting that if the elites were subtracted, Latin American cities would indeed be pre-industrial. Criticisms of this model have centred on the fact that in many large Latin American cities, elites have begun to suburbanise at a rapid rate due to increasing pollution, crime levels, high-density land use and traffic congestion (Ward, 1990; Gilbert and Varley, 1991). This outward movement of the elites, however, is likely to be restricted by the level of transportation and motorway access to the CBD. Unlike western cities, however, selected suburbanisation has not led to the demise of the inner city which, although suffering from decay in some cases, still retains its overall vibrancy and importance. Ward (1993) has observed that the growth of the urban informal sector is partly responsible for keeping Latin American city centres alive.

In La Paz, Bolivia, a city which is dramatically affected by altitude and terrain, the elites have gradually moved westwards to enjoy a better climate and lower altitude. New shopping malls, property developments and leisure facilities have accompanied this move, although the city centre is still the commercial and cultural heart. In the last decade, a new low-income district of La Paz has sprung up to the east, thus dividing the city into two distinct sections. 'El Alto' is a rapidly expanding low-income settlement which has been built by low-income workers on the eastern plateau around the International Airport. As the name suggests, the settlement is built at an altitude of 4000 metres, giving rise to extreme weather conditions and altitude sickness, in contrast to the western elite suburbs which lie at 2300 metres. Workers from El Alto commute into downtown La Paz, using a variety of formal and informal modes of transport. As a result of this expansion. La Paz is essentially composed of two cities which are markedly different in terms of physical factors such as altitude and climate, and socio-economic factors such as wealth and residential quality. This example further serves to highlight how different cities react to socio-economic change.

The following section critically examines the example of Mexico City, not only in respect of Griffin and Ford's work but also in relation to the effects of recent macro-economic conditions (see Ward, 1990). This examination of Mexico City is based on observations and fieldwork in the 1980s, and also draws on recent research by other authors (Eckstein, 1988; Ward, 1990).

-Latin American city structure: the case of Mexico City

Mexico City, perhaps Latin America's premier example of a primary city, is the largest city in the world, with a current population of around 19 million which is forecast to grow to around 30 million in 2010. Mexico City is also the economic and political heart of the country, a factor which is partly attributed to its centralised political system. It is also the world's largest labour market, where around 40 per cent of the population work in the city's informal sector. In addition, 43 per cent of the nation's capital is invested here, and in 1984 it had one-third of the nation's manufacturing and commercial jobs (Dickenson et al., 1996).

In the 1980s, Mexico was adversely affected by rapid rural-urban migration, structural adjustment and declining oil revenues (Eckstein, 1988). As a result, industrial development, employment and urban service provision have all been restricted, resulting in increasing poverty for a large proportion of the population. Despite some optimism concerning increasing economic prosperity in the 1990s, modern Mexico City serves as a good example of skewed j urban development, displaying the poverty of the masses and the extreme wealth of the few.

Mexico City occupies the same site as the prominent Aztec city of Tenochitlan, which was built in the early 1300s on the lake of Texaco at an altitude of 2238 metres above sea level. Under Aztec rule, its advanced structure comprised a series of compact rectangular islands and pyramids, which were separated by navigable canals. When the Spanish arrived in 1519, they found a city of over 100,000 inhabitants, with a complex urban social structure. The great city of Tenochitlan was conquered and destroyed by the Spanish invasion in 1521. A new colonial capital was built on the ruins of the Aztec city, and was renamed Mexico. The colonial city was based on a rectangular grid with a focus on the central Zocalo which contained a large plaza, the national palace and cathedrals, all of which remained prominent after Mexico's independence in 1821.

Throughout history, the Zocalo has remained the historical and administrative centre of the city (see Figure 6.6). In the 1900s, with a population of 345,000, the city's growth largely accrued around the Zocalo, with elite residences being built westwards near Chapultepec Park, and to the north of the city (Ward, 1990). In the 1920s, developments began to extend southwards, and in the 1930s, when the population reached one million, many of the elites built residences in Lomas de Chapultepec, along the tree-lined boulevard of the Avenida de la Reforma.

Since then, and as a result of rapid population growth, the city has constantly expanded its borders to incorporate many old neighbouring municipalities. After the Second World War, the city expanded in all directions, but particularly to the east, incorporating the dry bed of Lake Texaco. This unstable eastern district, which was mainly occupied by lower-income settlements, was the main area of devastation in the 1985 earthquake.

In the 1980s, the city covered over 1000 square km, housing 20 per cent of the national population in 0.05 per cent of the country's territory. The CBD has expanded westwards for over a mile along the Paseo de la Reforma which is known locally as the 'Zona Rosa' (pink zone). This linear CBD houses the city's main commercial offices and skyscrapers, hotels, theatres, exclusive retail outlets and cosmopolitan restaurants. It serves as a vibrant and dynamic centre for the wealthy.

In a pattern which resembles other Latin American cities, the historical Zocalo has gradually filtered down the socio-economic scale, although it still retains an important religious and administrative function. Today, many colonial buildings have been subdivided for rented tenements (vecinidades) and low-income requirements. It is, however, still a culturally an economically important centre for lower-income communities, with its informal markets, fiestas and religious services. In attempting to draw upon Griffin and Ford's ideas on Latin American city structure, therefore, Mexico City's spatial development broadly conforms to some generalisations, but as noted by other authors (Ward, 1990, 1993), there are other processes affecting the city which are not considered in their analysis.

As stated above, Mexico City does have a commercial spine which extends from the CBD westwards along the tree-lined boulevard of Paseo de la Reforma where many of the city's elite residences of Lomas de Chapultepec and western-styled amenities are located. In the 1970s, this sector was extended farther westwards to incorporate the pleasant, if rather hilly elite suburban developments of Bosques de las Lomas complete with American-style shopping malls, golf and health clubs, and private security patrols. Areas such as Polanco, which consists of houses and apartments close to the CBD, appear to be filtering down slightly to upper-middle class residents. Other elite residential areas include San Angel and Jardines de Pedregal, which offer old colonial-style recidences in the peripheral southwest of the city. The congestion, pollution and crime of central Mexico City are leading many elites to move to greener peripheries. Western expansion, however, is limited by the mountain range, although there has been some movement outside the city to Puebla, Cuernavaca and Querétaro. In 1981, the government invested in a plan to electrify the rail link between Querétaro and Mexico City to enable commuters to move out of the city, but this was halted by the ensuing debt crisis. The extent to which elite groups will continue to suburbanise remains to be seen. One major obstacle is the lack of services, transport and economic activity to meet their demands outside the city.

With some exceptions, the quality of Mexico City's housing stock is extremely poor outside the western-southern section and presents a complex social structure. In 1980, 30 per cent of homes lacked an internal water supply and 40 per cent were without sewerage facilities (Ward, 1986). As Figure 6.6 (page 133) highlights, there are zones of maturity around the central city. Established areas of squatter settlements known as colonias populares (people's neighbourhoods) constructed in the 1950s have now become part of the formal city structure, and are reasonably well serviced. Elsewhere, new colonias populares are growing on the city's periphery, without access to urban services and employment. Generally, lower-income neighbourhoods, colonias proletarias, are situated to the north and east of the city where the airport and major industry are located.

Areas of slums and shack developments, known as ciudades perdidas (lost cities), also constitute part of these developments. The growing middle class is located mainly around the southwestern wedge, such as Coyocan, along the Insurgentes Sur motorway where status increases with distance from the city. In the 1970s, satellite developments were built to the west and north of the city for middle and upper-middle class residents in Satelite, Tlalpan and Lindavista which were linked by El Periferico to the CBD.

Satelite was the site for Mexico's first large shopping mall development and leisure complex, but since the 1980s the town has been encroached upon by colonias populares and is now part of the city. Many areas of the city that were once peripheral and rural in character have since been absorbed by the city's expansion. Some colonias populares on the city's periphery are the size of many developing world cities. For example, Naucalpan in the northwest has grown from 30,000 in 1950 to 408,000 in 1970 and to 1 million in the 1980s. Since the mid-1980s, more rural areas to the north and east of the city have been consolidated by low-income dwellers at an alarming rate.

From this example, it can be seen that although Latin American elites are showing some signs of convergence in their residential and consumption patterns, Latin American cities are evolving in a culturally specific form. Furthermore, the scale or inequality and the sheer extent of urban deprivation in places like Mexico City have not been seen elsewhere. Since the 1980s, increasing environmental degradation, poverty and congestion in Mexico City have pointed to decentralisation as a planning policy, but whether this can really be achieved remains debatable.

The structure of developing world cities: final comments

This chapter has focused on the changing internal structure and social organisation of cities in the developing world. The key theme has been the recognition that developing world cities, despite some evidence of westernisation, have reacted in diverse ways to the forces of globalisation and macro-economic change. Whilst the applicability of various models of urban morphology to cities in Africa, Asia and Latin America is highly debatable, these models do highlight a number of common similarities which serve to enhance our understanding of traditional city structure.

Since the 1960s, rapid rural-urban migration has changed the character of cities in the developing world. Although it can be argued that some cities maintain various characteristics of their pre-industrial form, it is this recent growth which has had the most profound effect on city structure. In the postmodern era, many cities are experiencing some convergence in terms of architectural style, modern CBDs, leisure facilities, transport developments and popular culture, but this change is expressed in many different ways according to distinct social and cultural environments. In terms of their social structure, most cities in the developing world are still socially segregated in terms of access to wealth, urban services, employment and adequate shelter. Thus, developing world cities are not following exactly the same path as western cities, If nothing else, the existence of marked social inequalities, which are visible in the massive squatter settlements surrounding such cities, make them distinct urban forms in their own right.