In considerations of the apartheid system in South Africa, urban studies loom very large, particularly from the 1960s. The roots of spatial separation, the cultural discourse whereby legitimate urban citizens were white, the phases of legislation authorising the forced removal of people so that race groups could be separated and the impact in particular of the Group Areas Act of 1950, all figure prominently in the relevant literature (Bonner et al. 2001; Gervais-Lambony et al. 1999; Maylam 1995; Rogerson & McCarthy 1992; Smith 1992; Swilling et al. 1991 inter alia). We know as well that much of the root-and-branch resistance to the state was generated in the black urban townships, where demands for better schools, housing, transport, employment opportunities, etc. played a fundamental role in constructing mass resistance. This was a particularly widespread phenomenon of the 1980s, when civic organisations proliferated and the United Democratic Front (UDF) affiliated hundreds of branches, touching far larger numbers of participants than had ever before confronted the state at one time (Cobbett & Cohen 1988; Seekings 1991; Van Kessel 1995).

This paper tries to consider whether this resistance laid the foundation of a new and different social order in South African cities. Most of the critical scholarly literature has been quite sceptical, some of it scathingly sceptical, of any claims for more than superficial change (Bond 2000). And yet the state has certainly devoted large amounts of time and effort to attacking the problem of the cities. Here, we try to consider what has changed and what has not with a broad structural brushstroke. There is certainly scope for engaged activists to continue to take up the cause of the impoverished and the excluded, who still very much do exist. However, the terms for best understanding the South African system have changed in important ways; this is no longer an urban setting which can be evaluated in terms of whether it does or does not continue the apartheid framework.

In a recent article, I have made the case for acknowledging key elements of change while paying attention to their limitations (Freund 2006; but see also Pillay et al. 2006). There, I stressed three elements. The first was a dramatic deracialization and rationalisation of local government in the course of restructuring boundaries and authority structures. This has meant not only getting rid of racial demarcation and racially exclusive remits of services and representation; it has also ended the autonomy of small suburban communities on the urban periphery and, in certain cities, the intrusive role of black homeland governments that were intended to bring about racial autonomy and even political independence. Thus in Durban, one of the most complex of such municipal regions, more than 40 separate jurisdictions existed before 1994 which have since been melded into one, in the establishment of the Ethekweni Metro in 2001.

As an administrative exercise, this was a difficult and exhaustive process that went through a variety of stages. First came the furtive introduction of metropolitan structures in the late apartheid period, the shadowy Regional Services Councils, where the political and economic dead ends of apartheid policy forced change on the government. Then in the dangerous interim government period of 1990–1993, when existing authority was essentially a lame duck, informal coalitions involving business, the local state and insurgent black political forces came together to hold in existing structures while marking the foundations of a new order. This was followed by the creation in 1994–1996 of elected local governments which over-represented minorities based on the so-called sunset clause in the negotiated settlement, and where the big cities enjoyed two-tier government. However, the contradictions between old and new led to crisis, which was especially marked in the financial sphere. In the late 1990s, the African National Congress (ANC) moved towards the creation of so-called Unicities; here, we can talk of renewed, indeed unprecedented, centralization within local government intended to promote transformation and the equalisation of services. A remarkable feature of the new urban order was the rationalisation of local governments to a total of only 284 municipalities blanketing the entire nation. Most of these include both urban and rural components; even large cities such as Durban or Bloemfontein include substantial areas of low population density and predominantly agricultural land use (SACN 2004).

The efficacy of the new order compared to the pre-2001 transitional period is striking in the big cities, but its limitations need in turn to be underscored. While the rationalisation of local government units was intended to diffuse effective, democratic government into small towns with their extreme and visible racial hierarchies, and into the former homeland jurisdictions which had been administered with high levels of corruption, malfeasance and incompetence, in reality the huge disparities continue to exist spatially. The big metropolitan areas continue to have the advantage of the rates systems of local taxation providing substantial revenues and potentially some autonomy (Lootvoet & Freund 2005). They attract disproportionately well educated and trained people as administrators. While some small communities are effectively run, the majority are in serious difficulty.

In addition, the top-down, centralising strategy enshrined in ANC practice creates a real dearth of initiative from below. There is a striking gulf that yawns between popular perceptions of needs and the discourse of the state. Institutional attempts to fill this gulf have so far not been very successful (Parnell et al. 2002; Transformation 2008: 66–67) Councillors represent their parties and are first and foremost concerned to attract the attention of party power brokers so they can rise in their service; they function as little more than bureaucrats without genuine qualifications. This is certainly the case with the ANC, which is in an uncontested position in most municipalities. However, in some centres in the Western Cape, where the large white and coloured populations in good part support opposition parties (Jolobe 2007), and in KwaZulu-Natal, where the Inkatha Freedom Party of Mangosuthu Buthelezi remains important in many smaller municipalities, conditions are not very different. At best, it can be said that nationally driven service delivery has not really been blocked for partisan reasons; electoral competition has perhaps helped to push it along.

In the view of numerous anxious ANC observers, the problems originated with the dismissal of the UDF after 1994 and the effective dismantling of community organisations. Here lay the legitimate potential for democratic governance intimately linked to the struggle period. The civic organisations have continued to live on under the umbrella body the South African National Civic Organization, but they have been driven with corruption and with fundamental weaknesses vis-à-vis the central state which controls the purse (Heller 2003; Zuern 2006). The result is that local government has, despite attempts at creating democratic spaces, been heavily administrative and top-down, even though the government itself is dominated by those dependent on the votes of the poor and previously disenfranchised.

The second major shift, at least in theory, is that South African city administrations are now tasked with complex developmental chores including originating and developing their own integrated planning mechanisms. In practice, this has been very much less successful than the shaping of administrative unity. It is true that cities in South Africa have long been involved with economic growth initiatives, above all through the manipulation of land rentals and sales (Freund 2002). It is not only that this process is now difficult to implement due to lack of capacity in many smaller communities; it is also true that the relationship between the public and private sectors has changed. In the past, the local state was able to facilitate, and even in certain respects direct, private sector development. This capacity is today much less present. The idealisation of public-private partnerships depends on an abstract, idealised and unreal projection of the actual private sector, now so dominant, and its relation to the public.

Let us take the case of tourism. In some areas, this has been a significant factor for economic growth. The most dramatic example is Cape Town, which was estimated in 2007 to be growing at a rate of perhaps 7% p.a., in good part due to its attractiveness as a tourist destination. It is questionable, however, to what extent the metropolitan government has been responsible for this growth, including highly prized tourism from overseas, much of which owes itself to the physical endowments of the city, or has shown much capacity to regulate or control it in, for instance, ways that are needed to preserve the environment required for the tourism business itself. There are certainly other municipalities than Cape Town where tourism has real potential and tourist attractions can be fashioned out of a range of different ingredients. However, tourism is assumed to be a potential source of wealth in every municipality, beyond where the grain of common sense could take it. This is not Europe, where tourist revenues can feed off the leisure activities of local pensioners and subsidised school parties. It is also very much less clear what the economic advantages of tourism are beyond a limited trickle-down supply of service jobs for South Africa’s low-skill population with little or no access to capital. Calls for tourism to be the engine of development in every municipality are more pathetic than convincing.

A third challenge for the cities, with serious consequences, concerns the role of the municipality in what South Africans term ‘delivery’—the availability of housing, electrification, water reticulation and other features in poor neighbourhoods. Here, the municipalities are really only agents of state policies which are decided nationally. However, in the eyes of ordinary men and women, it is also here that the municipality is most visible and most acutely judged. There is no question that the delivery packages are and continue to be taken very seriously by the ANC, for obvious political reasons; it is their potential voters who are desperately concerned with these issues. It is also true that these systems and objects (as gifts from the state) have been far more widely diffused than before, according both to government sources and to surveys of opinion (Hemson & O’Donovan 2006).

However the delivery process, with its sense of free entitlement from the state rather than a deeper social transformation in the lives of the poor, has engendered a situation in which those who are falling behind in the queue are prepared to make trouble in order to advance their places and make sure they are not left out. For an insight into popular politics in South Africa, polls of trade union members who keep saying that they are totally committed to the ANC, which has basically good policies and the only leaders they are prepared to obey, are striking evidence (Buhlungu 2005). The ANC makes mistakes, according to this view, and is always in danger of moral degeneration, but the solution lies in reminding rulers of their obligations with strikes, public demonstrations and other forms of expression, not in order to unseat authority but to remind it of the vox populi. The possibility of destruction of property and even violence is not ruled out. This is quite different from the political rhythm aimed at unseating authority or finding an alternative to it through the ideology of ‘good governance’.

There are moreover deep antagonisms over the fees required for maintenance of water and electricity in particular, and over the poor quality of the public housing on offer, generally located at a considerable distance from the urban centre and consisting of very small detached units of limited utility for reconstructing stable family life. The cities have witnessed the generation of social movements which are sometimes willing to resort to desperate measures to reinstitute water and electricity services or to avoid eviction (Ballard et al. 2006; McDonald & Ruiters 2005; Morange 2004). This conundrum stems in good part from the prevalence of unemployment and the static or declining role of industrial employment as a means of stabilising and permitting the deep organisation of an urban working class. In fact, the social movements in areas inhabited by poor people are present precisely in the absence of disciplined organisation which could create systematic responses and engage in bargaining, including forms of bargaining that involve systematic class alliances (Barchiesi 2004).

Either representativity that is effective or development efforts based on local co-operation between different levels of government require some tolerance of shared power on the part of the centre. However, the ANC is not very comfortable with either option. A situation, say in Cape Town, whereby the opposition Democratic Alliance might potentially represent a dynamic bourgeoisie coming together with a reasonably well-represented coloured working class to create an autonomous growth strategy that would largely exclude the ANC and its African clientele, would be truly poisonous to the authorities in Pretoria. In Durban, it is possible to locate in the formation of Operation Jumpstart, an effective alliance of business leaders, city bureaucrats and ANC cadres stemming from the transitional period, the seeds of a distinctive growth strategy. However, efforts to build on and sustain such an alliance have been institutionally cut off since 2001 because they were too independent of national ANC control. This works directly contrary to contemporary urban literature on Western cities which assumes a decline in the efficacy of central state intervention and emerging autonomous ‘growth coalitions’.

At the same time, the possibilities of intervention through delivery from above are both limited and charged with increasing state dependence, if unaccompanied by improved chances for individuals to hold down stable jobs or sustain their own improved life chances. Without this, redistribution degenerates into distribution of a dole (De Oliveira 2006). The halting character of this kind of improvement will remain until trade unions or other forms of disciplined and stable popular organisations are themselves able to act from below and interact with the necessary interventions from the state at various levels.

Adaptability

If we turn away from policy issues to try and ‘read’ the South African city, it is of course wise first to recall that massive and perhaps traumatic change such as what happened in Russian cities after 1917 or in Algeria after 1962 is not very common. More typical is the situation where layered change affects cities, where structure and form from one era inevitably affect what follows and have an impact on later change. In this sense, we might more accurately look for adaptability and change by accretion rather than dramatic shift.

The South African transition of 1994 might be characterised as a wary compromise. The possibility for power to come into the hands of those elected by the majority was one aspect of this transition. However, a system of rules created through the Constitution and thereafter through the construction of policies limited the impact of change substantially. In effect, there was not merely a race compromise but a class compromise. The relevant literature often refers to a shift from the Reconstruction and Development Programme (RDP) phase of 1994–1996, when a programme calling for wide-ranging social change was official policy, to the Growth Employment and Redistribution (GEAR) policy phase thereafter in which tight monetary policies governed the economy and limited the potential for redistribution (Padayachee 1998). However, the least one could say was that the political compromise already presupposed those limits, whilst GEAR aimed for rapid economic growth that might make redistribution possible anyway. South African cities are inevitably a product of that 1994 bargain.

One theme that can be explored under the rubric of adaptability is the South African response to the greatly increased vulnerability of middle class neighbourhoods, as well as public transport and city centres, to crime (Palmary et al. 2003). The cities became much more open to their poorer inhabitants with the abolition of the pass laws in 1985 and changes in policing. One result, given gross economic inequality and the diminishing belief in, and fear of, authority on the part of the bold and the desperate is the wide diffusion of crime against property and violent crime against the person. These were already dire in the apartheid townships. High levels of crime genuinely rule out or greatly limit many potential aspects of social and economic development that assume a more open city. There is also no question that crime threatens some ways of life particularly: a classic case would be the presence of isolated family units, especially of elderly people, on small properties on the city’s edge. However, the detached suburban housing unit and the way of life that purportedly goes with the suburban ideal, much enshrined in South African middle class practice, is also seriously affected. The advertising-driven media conspire with state propaganda to suggest that limitless empowerment is possible for everyone. While improved economic conditions have reduced levels of some forms of crime, the combination of poverty and dreams of empowerment creates a deep and so far ineradicable social foundation for criminality.

In townships, let alone informal settlements, crime plays a huge role in limiting human movement and reduces the potential for business and social activity while private households are vulnerable. Qualitative and quantitative improvements in policing (the numbers of police fell considerably during the 1990s, although they have now recovered and will soon reach new numerical heights) can improve but not end this situation. There is certainly a strong tendency for policing to be intensified in certain areas considered particularly valuable because they attract massive numbers of shoppers or tourists, while other areas, for instance parks and places of natural beauty frequented by small numbers of local people, fall into neglect and become dangerous. The unevenness of policing is highlighted by the army of private policing personnel who have come to form a major new source of employment, to the benefit of those who can afford their services.

Without the means to assess whether the situation with regard to crime has changed in the townships, one can however suggest overall responses in suburbia. A first reaction has been one of exit. Estimates suggest that more than 10% of the white population, plus many coloured and Indian individuals have left the country since 1994, and usually insecurity is given as the most important causal factor that can be identified. Beyond that, two reactions can be delineated. One might imagine a decline in suburban detached family living, going together with the growing percentage of the population living outside nuclear family situations. However in the South African situation secure public space, while it exists at the level of the shopping mall and in some types of urban context, is difficult to assure. The balance in, to take an extreme example, the notorious inner city neighbourhood of Hillbrow in Johannesburg, once all-white and still densely peopled, tilts rather towards danger and public indifference amidst degradation of the built environment. As a result this option is at best very qualified.

The second and far more common South African reaction is the retention of the suburban ideal, now coupled with exclusion through walling and other devices such as security gates in streets, guardhouses, and the construction of gated suburban housing tracts as well as electronic alarm and automobile tracker systems (Bremner 2002; Czegledy 2003). In Johannesburg, to an important extent, capital has moved out of the very extensive traditional city centre and created a new and far more controlled access centre in once suburban Sandton.Footnote 1

The recent ‘xenophobic’ violence of May 2008 highlights the unprecedented arrival in South African cities of post-apartheid international migration, largely from the southern half of Africa. A former city official has estimated verbally that in Durban there is a population of more than 100,000 people of this description, far fewer than in the typical scare literature of the mass media but still perhaps 5% of the urban integument. The May 2008 events (less severe in Durban by far than in Johannesburg or Cape Town, where the numbers of foreigners are certainly considerably higher) highlighted the contradictions in state policy. While the state is eager to proclaim its intended benevolence towards Africans in general, there is a half-hearted policy of sometimes expelling but usually tolerating those who lack proper papers to work in South Africa, a situation that inevitably becomes a source of massive petty corruption. The foreigners find a hostile reception in settled townships in general, and come to live in the spaces of least resistance—new and uncontrolled squatter areas and the flatland fringes of the city centre. In the cities, they now have a strong presence in informal service activities, but shelter themselves with difficulty from the hostility of the bulk of the population for whom their competition as traders and jobseekers is unwelcome. Nonetheless, they are beginning to play characteristically important roles in the many informal niches of the urban economy.

Durban: A Tour of Post-apartheid Changes

To examine what has changed, let us use the example of the city of Durban, South Africa’s principal port and major coastal city alongside Cape Town (Freund & Padayachee 2002). Here, we can outline a number of significant changes.

Integration of the Golden Mile

Once defined as a space for white leisure, then divided—very unequally—between the races, since the late 1980s this stretch of sea frontage, decades from the harbour on the other side of the city centre, has opened to large sectors of the population and experiences dense policing. On a normal day, a pedestrian will quickly observe the beachfront used and enjoyed by members of all four of apartheid’s race groups as well as pockets of foreign tourists, in a setting which is probably the symbolic heart of Durban. At times, they are overwhelmed by the mass of black Durbanites and poor visitors from further afield, who outnumber everyone else at peak holiday season but not otherwise. At the north end of the beachfront is the Sun Coast casino shopping complex, a clear planning success story of the new South Africa. Here, a densely packed and racially very mixed crowd of middle class and lower middle class people can be found on any given evening.

The Central Business District (CBD)

While the affluent have withdrawn from its use, the CBD has retained important functions. The city government occupies much space here. Law firms and many firms linked to harbour activity also remain, as do a cluster of hotels on the seafront. The kinds of shops that cater to passersby have changed, but the streets remain crowded with throngs of very largely employed, if rarely white, people.

Flatland

Behind the CBD, just inland from the harbour, the little neighbourhood around Albert Park, the nearest residential area to the centre, has truly decayed and its partial squalor serves as a miniature reminder of Johannesburg’s Hillbrow. Further up, Berea Road marks the passage up to Tollgate atop Berea Ridge and the beginning of the highway inland to Pietermaritzburg and Johannesburg. Previously, a white group area with many inexpensive apartment blocks, this neighbourhood contains intermittent pockets of true decay juxtaposed with islands of stability serving those who want or need a functioning urban rental community, as well as apartheid era survivors trapped in what they would see as a woeful potential danger spot.Footnote 2 Prostitutes patrol some of the streets. Many black office workers who seek a different lifestyle from that of the townships have come to live here, and the streets are thronged with young black people in the early evening but a large number of Indians, coloureds and whites remain as well. It looks like a fairly accurate representation of the urban South African lower middle class today, an unlovely but vibrant section of town (Fig. 1).

Fig. 1
figure 1

Inner Durban. Source: map drawn by Frank Sokolic

Full size image

Gentrification

If you turn off Berea Road on the lower reaches of the Berea, you will notice big changes—in lower Umbilo, around the Botanical Gardens and on up into the beginnings of Morningside on Lower Florida Road. Here, you find what you might have hoped to see lower down in the CBD: a massive expansion of contemporary offices, often converted suburban homes, where lawyers, doctors, dentists, architects and other building services professionals, information technology firms, book shops, antique shops and suchlike have invaded and dramatically expanded the limited business premises of old white suburbia. Restaurants and outdoor cafes and indeed public life generally were not part of old Durban: whites preferred to spend their leisure time in private clubs and on sports fields.

Fig. 2
figure 2

Ethekwini Metro. Source: map drawn by Frank Sokolic

Full size image

Demography and integration

The larger expanse of the city has also experienced change. For one thing, there have been significant demographic changes which are now able to be reflected in physical movement (Hindson & O’Leary 2003; Vermeulin 2006).Footnote 3 Although there has been controversy about census figures, there can be little doubt that the white population of Durban has fallen considerably in the last 20 years. At the same time, the large Indian population has stagnated or increased only very slightly while the black population has grown substantially.Footnote 4 There has been, partly as a direct consequence of this, a considerable move of the Indian and black population in a kind of spillover effect into neighbourhoods formally closed to them in the past. Thus, the working class white suburbs that stretch south of the Berea ridge inland, along the old line of rail into the interior, have changed their racial composition very substantially. The densely built-up but affluent suburban mix around Musgrave Centre on the central Berea, once entirely white, has rapidly acquired a large Indian, particularly Muslim, population. Here are many new luxury apartment blocks and complexes.

Township changes

Southwest of the centre, the large Indian region of Chatsworth (Fig. 2), built up in the 1960s, started life as a bleak and largely undifferentiated residential township. Today, it has a complex economic life to it with a large amount of informal commercial and industrial activity, expanded shopping centre life and intensified differentiation between rich and poor neighbourhoods. The distinction between economic and residential activity, so rigidly maintained in apartheid planning, has broken down, reproducing the character of predominantly Indian settlements in the vicinity of the city in the pre-apartheid past. The poorer Indian areas have become the home of many African people, where intense conflicts over evictions and service delivery have taken place pitting both Indians and Africans against the local state (Desai 2002).

These kinds of changes are also notable in the established African townships. The local state has established a programme, inspired in theory by Local Economic Development (LED) planning, called INKS (Inanda-Ntuzuma-KwaMashu) to improve a large swathe of the northern part of metropolitan Durban. INKS has not made massive changes; it has largely failed to interest private capital, but the LED programme has brought about many integrated and improved services, extended urban agriculture and is helping to create small business centres which begin to take on a life of their own, notably through the presence of shops specialising in building equipment and household goods. At the same time, backyard shacks and extensions have multiplied in some African areas, highlighting growing internal class differentiation. Local government has also been responsible for filling in the large swathe of once empty land in mid-Durban called Cato Manor Farm under special conditions of land invasion crisis and conflict since 1994, together with important funding from the European Union and imaginative leadership (Freund 2007).

Flight

Much of the economic dynamism to be found in Durban, however, comes from the shift of the population and business activity further northwards along the coastline towards Umhlanga.Footnote 5 On the beach and above it, we can find large numbers of ostentatious new properties whose inhabitants are probably a representative mix of the richest 5% of the population. The movement to Umhlanga partly represents flight away from the presence of the poor, now so much more able to access central areas of the city, but no longer exclusively on the part of whites. Moreover it is flight to a new attraction similar to the so-called edge city phenomenon in America, powered by the highway, the car and the businessman’s search for greenfields development as the most lucrative way of developing real estate. Around the Gateway shopping complex, such developers are constructing numerous blocks of flats which will have all the essential safety features. Unlike the US ‘edge cities’, Umhlanga is part of the Ethekwini Metro and is controlled by the same fiscal regime as applies in poor neighbourhoods, but the logic of urban post-apartheid development is such that the local state tolerates and even in part facilitates the expansion of this apparent partial escape of the Durban bourgeoisie to appealing distant pastures.

This virtual tour tends towards the conclusion that there has in fact been a lot of visible change in the built environment of the core of a typical South African city, but it is change that reflects the changes in society as a whole: the gradual deracialization of the upper middle and wealthy strata, the considerable growth of a lower middle class no longer predominantly white, the deregulation of the apartheid city, and the growing presence of informal economic and social activity of every sort.

Limits to Change

On the other side of the coin are the limits to change. Here, I wish to focus on two related issues, first of all, the relationship between the public and the private. There is a lot of essentially mistaken normative writing today about the so-called win-win public-private partnership. This is supposed to be of the essence, for instance, in planning along the lines of LED, according to international best practice. At the heart of the relation between the public and private sectors is the reality that the ANC has essentially demanded from the private sector changes in the racial composition of management and in return it has abandoned most efforts—that were already in the process of being abandoned by its National Party predecessor before 1990—of disciplining or directing capital. In return, the private sector is at once anxious to take advantage of profitable opportunities in the South African city, reluctant about long-term investment in general, and recalcitrant about well-meaning schemes that are not on the face of it profitable. In South Africa, it is the banks and other investment companies which determine the direction of development rather than the state, and they do this in ways which reinforce the class patterns of the past. While investment rates have gently risen in recent years, South Africa fits very neatly with the South American countries with low national investment rate figures, as opposed to the ‘developmental states’ of Asia characterised by high investment rates. The essence of South African economic policy, at least during the presidency of Thabo Mbeki, was what is usually called Black Economic Empowerment. This marries the creation, through a variety of mechanisms, of a black business elite with more widespread affirmative action measures bringing black workers into state employment and black families into the middle class, while providing a range of dole-like payments and basic amenities to the mass of the poor. It implies that a new black bourgeoisie must form itself through integrating with existing institutions and networks and through nesting within, without destroying, the white-run business world.

There is substantial evidence to show that the state seeks to improve the lot of the poor (Hemson & O’Donovan 2006). It has constructed a great deal of housing in new neighbourhoods with new community facilities, it has extended the provision of electricity and water substantially, and it has created or improved amenities in established black townships. However, it has, as elsewhere, experienced difficulties in getting the poor to pay for the maintenance of new facilities and infrastructure and it has overwhelmingly preferred to construct new neighbourhoods far from the centre and far from the desirable suburbs where most wealth lies, often on state-owned land and on the outer edges of existing black townships (Todes 2003).

This is critical for an understanding of the contemporary form of urban development. Change in the built environment will focus on high-profit construction for the well resourced; it will rarely mix the affluent with the impoverished. Public housing as a means of social promotion for poor people is also limited by the extent to which the poor remain un- or underemployed: it is going to be constructed at minimum cost, on land available cheaply on the urban periphery, and it is not going to interest the business sector as a significant source of investment.

Nothing is more critical to the form of South African cities than the dominance of the motor car, which has become ever more pronounced since World War II. By the 1970s, whites were ceasing to make any use of public transport, a pattern more generalised in the middle class as a whole today (Durban Metro 1990). The poor depend on private buses and taxis, typically owned by small-scale black entrepreneurs. It is noticeable that public transport, which could be the means through which urban form alters, has hardly been of interest to the ANC. If one returns to the text of the RDP of 1994, there is a clear if very abstract call for this to be reversed, but so far public transport policy has largely been confined to blueprints (Behrens & Wilkinson 2003).

There is now a great deal at stake in retaining the dominance of the car. South African transport at this stage is intimately and totally tied in with the motor vehicle. On the one hand, motor vehicles and motor vehicles components (through agreements with foreign, largely German-owned companies) have become the country’s one really significant industrial export, while remaining an immense, even more costly, import at the same time. On the other, a very large part of the small business sector in any South African city is involved with the maintenance, repair and adaptation of motor vehicles. Everywhere, new developments are characterised by sprawl that is made possible by an excellent highway system which allows large distances to be traversed very quickly, through very generous use of fossil fuels (Mabin 2005; Turok 2001). Furthermore, since the end of the 1970s, black-owned taxis have come to ferry the majority of the population around the cities, and they are a powerful, lobby for the status quo, difficult to control. Public works are now steadily improving the quality of the national road grid after a hiatus during the early years of ANC rule.

Conclusion

I would therefore posit that the balance of economic forces in South Africa determines that urban investment and development will:

  1. 1.

    favour the interests of the private sector, which seeks insulation for preferred activities and preferred actors through forms of exclusion that are understandable, given the scale of poverty and the extent of crime;

  2. 2.

    support the kinds of investments that are familiar and characteristically profitable, such as high value housing and consumption projects, shopping malls linked to road transport and greenfields development; and that consequently

  3. 3.

    social development aimed at upgrading the lives of the poor will depend heavily on the initiatives of the state, which in turn feels obliged to sustain private sector initiatives whenever possible.

Another important area, however, is that of urbanisation itself and here the continued dominance of business interests is not to blame. The reality is that the ANC state is not much more interested than its predecessor, despite the abrogation of racially exclusive legislation and the absence of massive removal schemes, in the systematic urbanisation of the South African population in what used to be called an orderly way. Surely, by far, the most important aspect of land reform policy ought to be the systematic making available of land for an incoming population through expansion on the edges of the city as it grows, but there is nothing of the kind. The biggest example of a South African land invasion so far has characteristically been not on a profitable farm, but on state land at the city’s edge, in Bredell on the north-east outskirts of Johannesburg, and the perpetrators were generally already urban shack or backyard dwellers (James 2007). Apart from housing, there is equally little thinking about health, education and other relevant social needs for the incoming and impoverished. The heart of this should surely be the availability of land, and a modicum of security that can be a platform for some kind of sustainable, stable way of life. The reality is that there are few if any plans for systematic urban expansion recognising this need. Indeed, it remains so difficult, expensive and probably dangerous for poor people to leave land to which they do have access for the city that the rapid urban growth one would have expected to follow the removal of the pass laws and other related restrictions has actually been restrained, and mostly confined to a few centres where wealth is obviously being rapidly created (SACN 2004).

In conclusion, this paper suggests that post-apartheid South African cities may not live up to dreams of a rainbow nation, nor are they marked by reduced levels of economic and social differentiation. However, a process of deracialization has begun and they do demonstrate a considerable dynamism. Here, investment follows lines that are very well established: dominance of the motor vehicle, concentration on housing and commercial activities that serve the affluent, low-density, highway-dependent development—aspects of the past, but not directly connected to apartheid. There is adaptability—following the imperative of physically isolating the affluent through a range of security considerations from the forces that are no longer pent-up by the systems of segregation/apartheid as in the past. At the same time, informal sector activities take advantage of the declining policing capacity of the state in this regard. Counter-hegemonic intervention depends heavily on the state, which takes important and interesting but limited and somewhat contradictory initiatives. The state is very far from being the disabled or irrelevant dinosaur pictured in much ideologically driven contemporary literature on Africa.Footnote 6 This reflects the nature of the bargain struck in the pre-1994 transition period between those likely to take over the state, as well as the enduring staying power of the status quo in ways other than those related to race.