DECISION MAKING FOR SUSTAINABILITY AND AFFORDABILITY

By Alistair McIntosh, McIntosh, Xaba and Associates Durban, South Africa
and Clarissa Fourie, School of Civil Engineering, Surveying and Construction, University of Natal (Durban), South Africa

Prepared for presentation at Strategies for a Sustainable Environment Conference, Johannesburg, South Africa, August, 2000.

Abstract
Sustainability and affordability issues cut across many sectors related to the built environment namely, services and infrastructure, land tenure, governance, funding, cost recovery, planning, housing, private and public sector and the informal sector, and justice.

Sustainability is critically linked to affordability, not only in relation to the capital costs of the programme, but also long term maintenance and cost recovery from users. Present approaches, linked to outdated notions about the capacity of the state or to short term goals of donors, tend to be focused on delivery rather than long term sustainability. Some ideas for a new approach will be suggested, which are likely to increase both the affordability and therefore the sustainability of initiatives.

Introduction
Both internationally and in South Africa there are problems with the sustainable management of the built environment. While the problem manifests itself in terms of slow land and service delivery, in fact the problem emanates from the wider governance approach which has been adopted by a country. That is, a major dilemma facing much of the developing world concerns the misalignment of law and administrative and management systems with the actual systems of delivering land, buildings and services. It is this misalignment which creates unsustainable land and service delivery.

Sustainability is critically linked to affordability, not only in relation to the capital costs of the programme, but also long term maintenance and cost recovery from users. Present approaches, linked to outdated notions about the capacity of the state tend to be focused on delivery rather than long term sustainability.

This paper will examine this dilemma cross-sectorially. That is, a range of sectors will be examined within this framework namely, services and infrastructure, land tenure, governance, funding, cost recovery, planning, housing, private and public sector and the informal sector, and justice. It will do this by first exploring some of the problems and solutions found internationally, and then focusing on how these issues manifest themselves in South Africa. Some suggestions will then be made as to the way forward in terms of institutional arrangements for affordable and sustainable land and service delivery in South Africa.

The international perspective: A statement of the problem
The United Nations has tried to document the size of the global problem in relation to informal land delivery, informal settlements, the currency of land records and the shortage of services. They estimate that "(t)he extent of informal... settlements varies from country to country, comprising 20-80% of urban growth and affecting 15-70% of the urban populations of developing countries" (UNCHS:1996:4). That is, 20-80% of land delivery in the cities of the developing world is informal. With respect to housing, the United Nations city agency, Habitat, calculates that about one third of the world's housing is informal (unauthorised and/or with illegal tenure). This increases in Sub Saharan Africa, where over 50% is informal, and in the Asia-Pacific where over 40% is informal (UNCHS:1997). These figures indicate that something is radically wrong with our systems. We are arguing that what is wrong is the misalignment of law and administrative and management systems with the non-formal self managed systems. This mismatch is not confined to land delivery, but extends right across the spectrum of sectors within both urban and rural contexts (including justice, water, housing and infrastructure provision).

With respect to services, in the developing world only 43% of urban households have a sewerage connection, 60% water and 73% electricity. In Sub Saharan Africa 13% have a sewerage connection, 37% water and 43% electricity (UNCHS:1997).

Although slow land and services delivery can be blamed to some extent for these figures, the problem can also be directly linked to the income profile of users and to the amount of funding available. From 1985 to 1995 for the whole of Africa there was a negative growth rate in GDP per capita. In 1986 GDP per capita was $714 and in 1995 it was $665, with most years being a negative growth rate (UNECA: 1997:13). Not all countries in the region are equally wealthy and this is a generalized picture. This picture also applies to other developing countries. About one third of the planet lives below the poverty datum line, with about 40% below this line in Sub Saharan Africa and Asia Pacific (UNCHS:1997). This means that in many developing countries a large proportion of the population cannot afford to pay for services.

With respect to the possibility of supplying services, again based on Habitat figures (1997), the average per capita funding available for the supply of infrastructure is about $22 in Sub Saharan Africa ($6 for roads) per year, with an average for all developing countries of $59 ($11 for roads). This is in comparison to $623 ($127 for roads) for industrialized countries, which make up about 14 percent of the population of the planet.

In regard to formal land delivery, there is no documentary evidence of title for up to 90 percent of the land parcels in developing countries, with an estimated less than 1 per cent of sub-Saharan Africa being covered by any kind of cadastral survey (UNCHS:1991:3, 1990:4). "A World Bank study on Africa.. (showed that) ..if no dispute occurs, the process of land registration takes an average of 15 to 18 months, and that normally, a period of two to seven years is not uncommon. This lengthy and costly procedure.. (means that).. tens of thousands of land titles.. (are usually).. pending." (UNCHS:1991:5).

This international picture serves to contextualise, and give perspective to, the South African situation. We will show that the kind of problems South Africa is experiencing conforms to the general developing world picture. Learning from international experience, we are arguing that we are not going to solve problems here by working more efficiently within a 'business as usual' framework because of the mismatch between formal and non formal systems. Rather we have to alter the way we are doing things to be able to develop an affordable, sustainable built environment.

That is, we need to come to terms with the reality, as shown by these figures, that large sectors of the population depend on non-formal systems of delivery which are essentially self-managed, and generally fall outside of the ambit of government, policy, planning and management systems. The extent of these systems, and their contribution to livelihoods and social order, is rarely acknowledged. What is also not acknowledged is that these systems provide access to resources and services for the poor; and usually cannot be replaced by government programmes and services. Generally government policies and programmes are not aligned with such systems. Government programmes tend to compete with these systems, rather than complement them. Also they do not adequately contribute to enhancing the effectiveness of these systems in promoting livelihoods and/or providing affordable services. Yet at the same time, formal service and delivery systems remain beyond the reach of the poor, and even when implemented are often not sustainable.

An international perspective: Best practices for affordability and sustainability
To ensure sustainability of the built environment a different approach to service design has to be adopted. A practical example is given here which demonstrates many of the component parts that make up a sustainable affordable approach; and points to the way forward to over-coming the misalignment of law and administrative and management systems.

Typically the planning of informal settlement regularisation has taken place on an ad hoc, settlement by settlement basis. The best approach however is a citywide approach and this has successfully been done with three Indian cities, including one with a population of over one million (Diacon:1997). Instead of focusing on individual settlements, or on the city limits, as the area for planning, the focus should be on the primary infrastructure networks, such as the water mains, road networks and/or sewerage system of the urban area.

Urban planning should create a holistic infrastructure supply, which facilitates individual household connections as and when they can afford it. This combines "..three forms of intervention: the production of primary infrastructure network, the incorporation of informal sub-dividers into the production process, and the progressive servicing of areas that are already occupied, with plenty of scope for community mobilisation and self-help.. The population contributes directly to the provision and management of infrastructure and services. Such approaches also favour a less centralised form of urban management, promoting community organisation at the settlement.. level" (Durand Lasserve:1998:241-2).

Using this approach, the design for the sewer system was made affordable by linking a range of innovative practices. Indore, a city of over a million people in India, had no underground sewers. Creating underground sewers for the whole city solved both the informal settlements' sewerage problem, as well as the rest of the city. As the city benefited as a whole it was possible to cross subsidise (Diacon:1997). One of the most expensive problems associated with underground sewers is the cleaning out of blockages. To do this special inspection chambers are usually built into the sewerage system, which pushes up the capital cost of the system. Also inspections have to be done by local authority officials, as does the removal of the blockage. This also increases maintenance costs and user charges.

In Indore, the system was designed so that each individual house was responsible for its own connection to the sewer network, as and when they could afford it. This connection included a gully trap at the front door of the house, where any blockages produced by the house would be caught. In this way the house residents were responsible for sorting out their own blockages. This meant that much fewer blockages entered the main underground sewerage system, making it much less costly to maintain and therefore affordable to the users. However, it would not be possible to introduce such a system successfully without extensive community capacity building around the maintenance of services (Diacon:1997).

This is just one example from Indore of innovative service design linked to community capacity. From this example it is possible to establish the link between on the one hand, community participation and on the other hand, planning, servicing and affordability. It is also possible to see how community participation and the building of capacity within the community are directly linked to the ongoing maintenance and sustainability of the services. The key to the success of the citywide regularization project was the slum dwellers themselves, who showed that they were willing to mobilize resources despite their poverty. They were partnered by government agencies, including local government, NGOs, including women's NGOs, and local professionals. This initiative was driven primarily through community control. Involvement of the community will:-

• Promote transparency and accountability in the planning and implementation of policies and programs (UNCHS:1996:6-8);
• Encourage affordable systems, self funding and cost recovery (Diacon:1997:63);
• Create a feeling of ownership by the community of policies and strategies (UNCHS:1996:6-8), which is important for long term maintenance in the area (Diacon:1997);

Planning decisions in which stakeholder participation, and especially community participation is critical, relates to the choice of services. Service affordability by the community and the ability of the local authority to cost recover should be key design characteristics for planners rather than national planning norms (Diacon:1997). The community may well prefer to have cement footpaths as an access to every house, and agree to prevent vehicles using those footpaths, than have to pay for the cost of upgrading the settlement to allow expensive vehicular access to every house.

Funding and cost recovery. Funding and cost recovery should become central design features when undertaking projects in the built environment. A number of lessons have been learned in relation to the funding of regularisation projects and the subsequent cost recovery:-

• Donor funding is common (Bestpractices:1999) in regularisation projects. Donors tend to fund the initial provision of the legal tenure and services, including the capital costs, and not the ongoing maintenance of the services and the land records. If the original design created unaffordable maintenance and user costs, sustainability can be compromised;

• Central government funding is common for ad hoc settlement by settlement regularisation (Azuela and Duhua:1998 -Mexico and El-Batran:1999a,b -Egypt). City wide coverage with full regularisation is rarely, if ever achieved, even in the largest regularisation programs in the world (Azuela and Duhua:1998 -Mexico). This is largely because they are so costly (Durand Lasserve:1998:240). Cost recovery from those being regularised is often not successful (Banerjee:1999a,b -India);

• Local authorities rarely have a large enough tax base to undertake the task on their own. Through partnerships, local authorities have been successful in undertaking extensive regularisation (Diacon:1997; Banerjee:1999a,b);

• Approaches whereby the community save towards the project before regularisation appears to work well. This is especially true where women are involved in the project. Communities have shown the capacity to mobilise money for regularisation quicker than the local authority (Diacon:1997);

• Projects where all households have to become involved and pay for service installation at the same time (Huchzermeyer:1999 -South Africa) are not as successful as projects, which allow households to choose when and if, they wish to acquire individual connections (Diacon:1997). That is, individual connection to services rather than mass servicing is more affordable for poorer households;

• Regularisation is often subsidised (Huchzermeyer:1999 -South Africa; El-Batran:1999a,b -Egypt), while the maintenance of the area is often on a cost recovery basis (El-Batran:1999a,b -Egypt). Best practices centre on cross subsidisation for the creation of infrastructure networks, but that individual household connection should be done with self-funding or community funding (Diacon:1997 and Bestpractices:1999 -Pakistan/KKB). In this way households take responsibility for the maintenance of the services and are more likely to pay user charges;

The costs created by the design of the services are critical to the affordability of the services, ongoing maintenance and sustainability. Community participation is crucial to the design and maintenance of affordable service. From another angle, designing affordable services, community/labour based inputs into the maintenance of the services and the cost of their maintenance is critical to whether a local authority is willing and can afford to get involved in the regularisation of informal settlements (Diacon:1997). This is especially true in relation to roads and underground sewers (de Castro:1999 -Brazil), which can be very costly to create and maintain.

Finally, the focus of funding informal settlement regularisation has generally been on the installation of services and the delivery of titles. For sustainability funding requirements need to be seen in terms of capital costs and maintenance costs, as well as affordability and recoverable user charges.

In conclusion, the lessons learnt from this infrastructure example can be applied across many other sectors. However, such approaches require new governance approaches, which are not premised on the notion that the state has the capacity to deliver land and services for all, sustainably and affordably. Rather, new governance approaches need to be developed which link with existing non-formal self-management approaches.

Planning
Planning systems are generally based on the assumption that there are instruments available, in terms of which development can be managed to promote economic growth, maximise the use of space, manage land use effectively and manage the use of environment and natural resources. These instruments range from prescriptive legislation which requires that development applications are assessed in terms of particular procedures and criteria, to development plans which guide public investment and provide criteria for the use and management of land at a local level.

Planning systems often fall short in assuming that such instruments can be applied within areas in which informal systems of land and service predominate. In South Africa planning is guided by the Development Facilitation Act that includes a number of principles according to which integrated development plans should be compiled, and in terms of which planning applications are meant to be judged. These include principles such as promoting compact cities, avoiding urban sprawl, allowing mixed-use etc.

The difficulty of applying prescriptive legal instruments within non-formal settlements is that intervention is only possible where formal rights are applied for, since in these cases the necessity for land use approval can be imposed. It is only in these instances that it is, or will be possible, to control land use. (Within many of the former homeland areas, for example, residential sites have not had to enter the formal PTO approvals system.) What this means is that to a large extent, informal land allocation and management systems cannot be controlled by formal planning systems and procedures in the same way that formal areas are controlled.

Where formal planning systems can influence the non-formal areas is through the location of public investment, whether infrastructure or public facilities. Since the state retains control over this process, spatial patterns of non-formal settlements can be influenced. However, this falls short of the very direct management controls anticipated by formal planning systems. These assume that the use and management of land can be directly controlled at a micro-level.

In short, the scope for state intervention lies in exercising indirect influence on settlement through strategic public investment, and where formal land management systems intersect with informal ones. Aside from these contexts, the capacity of government to play a role in informal systems of land administration will depend on the development of positive relationships between planning authorities and non-formal land management structures. Only in this way is it possible to negotiate land use and environmental management systems that are consistent with the formal planning systems.

However, this means making concessions to informal systems. There are two reasons for this. First, the state is unable to substitute for the social security and welfare functions of land accessed through informal systems. Secondly, the state does not have the capacity to replace informal systems of land allocation and management with formal planning procedures and mechanisms for land use control.

Governments have found it particularly difficult to reorient formal planning systems to provide sufficient incentives to informal systems to comply with planning and development goals. Reasons vary, from the fact that planning systems in many countries are really only geared to servicing the middle classes and middle class areas; to the fact that political and institutional compromises have to be made, to work with informal systems of land management. Political deals have to be struck with unelected (and often unsavory) local leaders. Typically fragmented line agency bureaucratic systems are also ill equipped to deal with the service and planning issues within informal settlements.

The issue of housing and finding an affordable place to live
Housing policies and programmes within many developing countries are not geared to the bottom end of the market, and generally do not complement or enhance predominantly informal systems of housing delivery (UNCHS:1996).

South Africa's housing policy has revolved around a R16000 grant on freehold land which is meant to cover both survey and infrastructure costs. Usually this grant will only cover the site and connections or part of a top structure. The pattern in applying the housing policy has been that state housing projects are not able to accommodate the very poor. Households who acquire houses at no cost to themselves have difficulty paying rates and service charges, and the very poor will usually find sanctuary in dense informal settlements on the outskirts of the city where cheap access to land and services can be secured. The consequence for a city like Durban, which has traditional authority land surrounding it, is the increasing dense informal settlements in the outlying areas. Formal controls and services cannot be provided in these areas without the support of the traditional authority or similar local land management structure.

In many of these tribal areas population pressure has resulted, because of the numbers of people seeking affordable access to opportunities provided by the city. Both this, and the pressures placed on the natural environment within some of these areas by the emergence of dense unplanned settlements, dictate:

• The necessity to provide affordable services to these areas, recognising that most conventional services (housing, reticulated water etc.) will remain outside of the reach of people residing in these areas. If one adopts a new approach to governance in terms of the affordability and sustainability paradigm, different choices will be made about which services should be prioritised, as well as the form that delivery should take. Potentially affordable services include community gardens, wood lots, boreholes and protected springs.

However even these types of services require much greater levels of participation and community management if they are to be provided affordably than conventional municipal services. Typically the mobilisation of communities in support of service delivery involves developing the support and co-operation of traditional authorities in these areas. Therefore a municipality wishing to provide affordable services must develop very close relationships with existing land management administrations and mechanisms to be able to mobilise community support for service delivery;

• Effective land use and environmental management within these areas. Negotiated arrangements with local land management/administration structures need to become a priority for the development of effective land use and environmental management systems, within these areas. These will have to be secured through co-operative agreements reached with traditional authorities and/or informal land management administrations.

The point that is being made here is that current government policies do not accommodate the bottom end of the market in terms of housing. At the same time, there are not appropriate policies and instruments which would be needed to manage those existing urban informal settlements, which cater for this sector of the population.

It should be emphasised however, that the issue of cheap and affordable housing is not solely an issue for the cities and small towns. A recent finding of a review of the Land Reform Programme in South Africa (which has had a primarily rural focus) also highlighted a tremendous demand for land for settlement for the very poor (Department of Land Affairs:1999). These are people who cannot afford formal housing opportunities and the related cost of services, and are unlikely to be in a position to take up commercial or semi-commercial agricultural ventures owing to shortages of labour, capital or skills. Since these groupings are not being catered for in terms of existing policies or programmes by any organ of government, the Department of Land Affairs was attempting to cater for these particular needs in its land reform projects. However, since the Department of Land Affairs has had a limited mandate to transfer land, it has not been able to orchestrate the development or management of such land with the consequence that many such settlements have not been sustainable.

A housing policy that recognises the contribution of non-formal housing delivery systems to affordable housing would have to deal with two major issues. Firstly, it would have to confront the tenure issue, as individual freehold title is the only option at present. Secondly, such a policy would have to find ways of accommodating the non-formal structures which effectively deliver much of the housing within the country at present.

Land tenure and land delivery systems
As a result of past apartheid policies many of South Africa's citizens have become used to living outside of the formal land registration system. This factor, together with the legacy of a skewed distribution of land ownership which limited Black South Africans (75 percent of the population), until 1991, to only 13 percent of the land, has direct implications for the country's land registration system and cadastral records. We are arguing that the currency of these records is poor in many developing areas of the country; and that this currency might not improve as much as is expected using current procedures. Such a lack of currency makes land management and land reform more difficult where it is needed the most.

Buckley, a recent past President of the statutory council of the South African survey industry, argued as early as 1995 that in many areas of South Africa the records do not mirror what is on the ground. For example, "(i)n Soweto ..(a)lthough freehold title has only recently been introduced, more than 15% of the recent transfers have not been registered. These figures are expected to escalate." (Buckley:1995:2). This statement accords with the findings of a number of other land surveyors around the country.

In many areas of South Africa there are multiple claims to the same parcel of land in terms of the restitution process. An example of multiple claims, including restitution claims, within the same area, albeit one of the more complex examples, is that of Cato Manor. This vacant area of land, in the heart of Durban, intended for the development of low income housing for 180,000 people, has a range of claims against it. Some of these being, the former landowners; the present landowners; the present squatters on the land; the squatters who lived there before they were removed in the late 1950s when the original landowners were expropriated for apartheid planning ideals; and the Cato Manor Development Association, which heads a President's project to develop the area for the poor (Makhathini:1994:61).

This example, though complex, is by no means unusual and demonstrates the enormous problems being faced in the country. Not only is land management and land delivery extremely difficult, but adjudication becomes a centre stage issue. South Africa at present has no legislation which adequately caters for this type of adjudication. The only adjudication legislation on the statute books was created during the apartheid era for the awarding of deeds of grant on state land (Statutes of the Republic of South Africa:1993).

Where titling and adjudication legislation and procedures are inadequate the weak, and especially women, are discriminated against. Lastarria-Cornhiel argues that "(d)uring the transition (be it through land reform, market forces, or a land titling project), men and particularly male heads of households acquire total, legal ownership of.. land. (The process can.. eliminate the few rights.. that women had to land.." (Lastarria-Cornhiel:1995:39).

In these circumstances if the registered male title holder dies then it is unlikely that the woman, who originally had some form of de facto land right, would take the trouble to transfer the land to her name. This aspect is important both for the customary tenure areas in South Africa, as well as for the burgeoning informal settlements which are in the process of being titled and serviced. A few of the major reasons why the cadastral (land surveying) records are losing currency in some of the developing areas are:

• The present registry is not accessible in terms of location;
• Transfers are too costly;
• Transfers do not take place because of conflicting claims to the land;
• The deeds registry system is not transparent to the poor and historically disadvantaged, who in some instances are still suspicious that the system is not looking after their interests;
• Conflict over women's de jure land rights despite their de facto rights;
• Weak local government structures and officials who have allowed development without land use control consent procedures or legal titling;
• The structural tension between group and individual rights to land -this occurs in customary tenure as well as formal townships;
• The slow pace of the legal redistribution of the land, because of multiple claims, the administrative and legal complexity of creating developable land and adjudication problems;

A way forward to solve these problems is by locating the land records at the local level for developing areas. The land records must also be held in such a way that they are readily available to a wide range of local officials and ordinary people to facilitate local land management (land administration). In this way local land management will be improved because the land information is readily available, but it is also improved because of the increased symmetry in this information. This increased symmetry comes about not only because the land information is available at the local level, but because it also sets up local linkages and builds the decision making capacity of local communities.

Water
Few people within rural areas or informal settlements can afford reticulated water systems. Most of the rural areas depend on protected springs and boreholes that depend to a large extent on voluntary committees to ensure their maintenance and continued operation. The capacity and authority of such committees invariably depends on firstly, the technical support which might be obtained from government (whether Water Affairs, a water authority or the municipality). However, secondly, their capacity and authority also depends on the mandate received from the land or community management structure which exists in these areas.

Failure to recognise this resulted in thousands of boreholes, windmills and springs being rendered non-functional during the apartheid period. Although the Department of Water Affairs has recognised the necessity to establish and train water committees to enable implementation of the its community water supply programmes, it has been slow to recognise existing self-management systems and structures (or the role of formal local government systems). It is possibly for this reason that their community water supply programmes have been much more expensive than anticipated, with the consequence that much larger amounts of money are required for facilitation and training than was originally anticipated (this argument needs to be verified).

The issue is not restricted to rural areas. Within urban informal settlements where stand-pipes have been provided, various systems to recover costs have been attempted, ranging from the 'kiosk system' in which individuals are mandated to sell water, to quarterly or monthly levies charged on each household within the settlement. All such systems rely on the cooperation and support of the land, or community management structure. In the case of the kiosks, they rely on local administration structures to allow individuals to trade in the sale of water, and in the case of monthly or quarterly collections, the water provider is totally dependent on the community management structure to create agreements, and develop systems for the collection of levies.

Failure to acknowledge and work with the land and community management structure, whether a warlord, and traditional authority or a civic generally means that costs cannot be recovered.

Energy and electrification programmes
It was anticipated that electrification programmes could be spread to outlying rural areas and to informal settlements through technological means. That is, the prepaid meter system could be used to build some of the infrastructure costs into the payment system for the use of electricity.

The major unanticipated shortcoming of the pre-paid metering system has been the limited amount of electricity used by each household, this is aside from problems of tampering in order to bypass payment. The tendency has been to use electricity only for lighting, whether in outlying rural areas, or in dense informal settlements. Cooking, heating and other needs have continued to rely on traditional sources of energy, especially wood.

The consequence of this is that, the electrification programmes in rural and informal settlement have had to be limited to areas with sufficient combinations of density and proximity to the electricity grid to make the programme economically viable. The important point however is that, even where electrification programmes can be successfully put in place, because of affordability issues, the availability of electricity does not obviate the need for other strategies which will enable sustainable use of traditional sources of energy.

The fact that other strategies (such as woodlots) typically require extensive self-management, and the approval of existing land management structures, highlights the point that technological innovations cannot, or can only partially, substitute for the necessity to engage with these bodies in service delivery.

Justice and dispute resolution
The extent to which traditional authority justice and dispute resolution systems are used in traditional areas varies between and within provinces in South Africa. Within provinces like KwaZulu Natal, Mpumulanga, the Northern and North Western Provinces they continue to be used extensively. Within the Eastern Cape these are still extensively used in parts of the former Transkei, but not in the former Ciskei or the Free State (former Qwa Qwa and ThabaNchu). Such support for traditional systems of dispute resolution and justice is in spite of negligible support for these functions by the various provincial governments or their homeland predecessors.

An important observation however, is that where traditional authorities have lost their power base to civics or warlords an adapted form of local justice continues to be used, often only with altered procedures and symbols. That is, the collapse of traditional justice systems does not imply that people will use the formal systems of justice. Rather, formal systems are seen as too cumbersome, expensive and beyond the reach of ordinary people.

Similar non-formal justice systems also emerge within urban informal settlements. In some cases they replicate the tribal court in terms of procedure and sentence, in others they take on the mantle of the 'civic' courts, which existed during the early period of South Africa's transition. The extent to which such informal systems of justice are growing in South Africa's urban areas, and the perceived ineffectualness of the formal criminal justice system, is evidenced in the phenomenal growth of vigilante movements such as Magopa in South Africa's northern provinces and Pagad in the Western Cape.

The shift in South Africa's criminal justice system towards 'community policing', with its emphasis on prevention and securing community support in the administration of justice, could facilitate a shift towards acknowledging and working with existing non-formal justice systems that operate within the law. This leap has not been made however, with the consequence that formal and informal justice systems compete with one another, to the detriment of crime prevention as a whole.

Institutional arrangements for sustainable service delivery
Three broad points may be made from the discussion thus far. Firstly, the form that non-formal systems of service take are substantially different from conventional municipal services, in that they require high levels of self-management and support and approval from informal land management structures. Second, engagement with such bodies is fundamentally political in nature. Local political authority of land management structures is founded on granting poor people access to land and other resources that the state cannot generally hope to provide. Securing support of these bodies is essentially contingent on securing a quid pro quo which will not undermine the existing power holders, or will provide them with an alternative (and perhaps more legitimate basis) for securing consent and support.

Two examples of the second point may give grist to the argument. The first is the observation that one of the first successful slum upgrading programmes in the Durban Metro was in an area controlled by one of the most notorious warlords on the outskirts of the city. His control over the settlement enabled many of the obstacles to regularisation in neighbouring communities, relating to community leadership, to be overcome, once the quid pro quo with the war lord had been achieved.

A second example concerns the reluctance of many traditional leaders in KwaZulu Natal to contemplate development initiatives in their areas, for fear of losing control over the process. This has been ascribed to the severe neglect of the institution of traditional leadership during the apartheid and post-apartheid periods, and the undermining of officials' functions outside of land allocation. Where the officials had been previously marginalised in developmental activities by government, para-statal and non-government organisations, they no longer felt able to allow such developments without losing ground to competing local interests. Widening their basis of legitimacy, through training and keeping them informed on development opportunities, would enable an improvement of their status as facilitators of development, rather than being reactionary conservatives.

The political nature of the necessary engagement between government and non-formal land management structures is one of the major stumbling blocks in realigning formal law and management systems with the predominant informal systems. At a national level in South Africa, this is firstly reflected in a continued lack of clarity about the role of local government and traditional leaders. Secondly, there is an ongoing uncertainty regarding a policy on land tenure within the former homeland areas. Thirdly, at a local government level, it is reflected in a reluctance to engage with traditional leaders, or other informal land management structures, at a political or even an administrative level. The implicit and as we have shown, fallacious assumption remains in place, that government can meet its constitutional and social obligations without recourse to traditional authorities or other (non formal) land management structures.

It is by no means being suggested that the choices that must be made in such negotiations are easy ones, particularly where there are competing contenders for government patronage. Where councillors, or officials, are not mandated to strike deals with such structures, (ideally within defined parameters set by such councils), it is unlikely that government will be able to contribute to the enhancement of informal systems of service delivery.

A third point to be made however, is that conventional municipal models comprising a range of functional departments generally providing conventional municipal services are typically poorly equipped to deal with the problems and service needs of such settlements. Conventional municipal models usually rely on paid professionals to provide services for which service charges are levied. In enhancing informal systems of service delivery, a priority is ensuring that liaison between communities and service-providers is not fragmented along sectoral lines. The consequence of such fragmentation is that the agreements reached around service delivery or infrastructure provision within one sector are contradicted by the activities of other departments that provide related services. Instead, mixed messages and conflict can be avoided where different service providers interface with communities through a common dedicated liaison service.

From another angle, the link between community liaison at a local level and political decision-making at the apex of the local government body becomes extremely critical, given the potentially political nature of this interaction. The purpose of local liaison will be subverted where service departments or liaison officers lack mandates from the apex to enter into agreements which will enable sustainable and affordable service delivery.

The need for co-ordinated liaison and committed political mandates are reflected in the Indian case study described earlier. This example highlights the importance of liaison and community participation at a local level in negotiating the choice of services, and the recovery of costs. It also highlights the need for extensive co-ordination across service branches, and political commitment at a city-wide level, in re-orienting what services and infrastructure are provided, and how this is done.

Conclusion
An overarching conclusion emerging from this paper is that a 'business as usual' approach to the built environment and its related services will not address the needs of the very poor within non-formal settlements, which are a significant proportion of any urban population. It is equally important to highlight the fact that the barriers to sustainable service delivery and development posed by affordability constraints, and the solutions that are required, are common across sectors. These relate both to the type of services that are provided, how such services are delivered, and the extent to which these support, rather than undermine, existing community-based service delivery and management systems. Sectorally-based literature and practice has generally not grasped this nettle, with the consequence that South Africa, together with much of the developing world has a long way to go in addressing the needs of the poor.

References
Azuela and Duhua, A. and E. Duhau 1998 Tenure Regularization, Private Property and Public Order in Mexico, in E.Fernandes and A.Varley (eds.), Illegal Cities Law and Urban Change in Developing Countries, Zed, London.
Banerjee, B. 1999a The role of researchers and practitioners in setting up a concerted urban management approach: lessons from South Asian municipalities, Paper presented at N-aerus International workshop on Concepts and Paradigms of Urban Management in the Context of Developing Countries, Venice, 11-12 March, http://obelix.polito.it/forum/n-aerus.
Banerjee, B.1999b Security of Tenure in Indian Cities: Overview of Policy and Practice, Paper presented at 'Tenure Security Policies in South African, Brazilian, Indian and Sub-Saharan African Cities: A Comparative Analysis', Centre of Applied Legal Studies, 27th-28th July, 1999, Johannesburg.
Buckley, R, 1995, Low level registries, Paper presented at the Commonwealth Association of Surveying and Land Economy (CASLE) and the International Federation of Surveyors (FIG) Conference on, Sustainable Development: Counting the cost and -maximising the value, Harare, Zimbabwe.
De Castro, S. 1999 Favela Bairro Programme: A brief institutional analysis of its land aspects, Paper presented at 'Tenure Security Policies in South African, Brazilian, Indian and Sub-Saharan African Cities: A Comparative Analysis', Centre of Applied Legal Studies, 27th-28th July, 1999, Johannesburg.
Department of Land Affairs, 1999, Review of South African Land Reform Programme, Pretoria.
Diacon, D. 1997 Slum Networking. An Innovative Approach to Urban Development, Building and Social Housing, Foundation, London.
Durand Lasserve, A. 1998 Trends and Issues, in E.Fernandes and A. Varley (eds.), Illegal Cities Law and Urban Change in Developing Countries, Zed, London.
El-Batran, M. 1999a Tenure Upgrading and Security of Tenure Policies in the Cairo Metropolitan Region, Paper presented at 'Tenure Security Policies in South African, Brazilian, Indian and Sub-Saharan African Cities: A Comparative Analysis' Centre of Applied Legal Studies, 27th-28th July, 1999, Johannesburg.
El-Batran, M. 1999b The Egyptian Experience of redefining property rights in the age of liberalization and privatization: a review, Paper presented at 'Law and Urban Space -Facing the paradox: Redefining Property Rights in the age of Liberalisation and Privatisation, International Research Group on Law and Urban Space (IRGLUS), and Centre for Applied Legal Studies (CALS), Johannesburg, South Africa, 29-30 July,1999.
Huchzermeyer, M. 1999 Tenure questions as understood in the debates in South African informal settlement and intervention literature of the 1990s -a critical review, Paper presented at 'Tenure Security Policies in South African, Brazilian, Indian and Sub-Saharan African Cities: A Comparative Analysis', Centre of Applied Legal Studies, 27th-28th July, 1999, Johannesburg.
Lastarria-Cornhiel, S, 1995 Impact of privatization on gender and property rights in Africa, Paper prepared for Gender-Prop, an international email conference on Gender and Property Rights, May-December 1995.
Makhathini, M, 1994, The case of Cato Manor, in Hindson, D. and J. McCarthy (eds.) Here to stay. Informal settlements in KwaZulu-Natal, Indicator, University of Natal.
South African Government, 1993, Distribution and transfer of certain state land act, No. 119 of 1993, Statutes of the Republic of South Africa -Land.
United Nations Centre for Human Settlement (Habitat) 1990 Guidelines for the Improvement of Land Registration and Land Information Systems in Developing Countries, Nairobi.
United Nations Centre for Human Settlement (Habitat) 1991 Report of the Workshop on Land Registration and Land Information Systems, Nairobi, Kenya (15-18 October 1990).
United Nations Centre for Human Settlement (Habitat) 1996 New Delhi, Habitat II Global Conference on Access to Land and Security of Tenure as a Condition for Sustainable Shelter and Urban Development, New Delhi, India (17-19 January, 1996).
United Nations Centre for Human Settlement (Habitat) 1997 Global Urban Observatory Monitoring Human Settlements with Urban Indicators (Draft) Guide.
United Nations Center for Human Settlement 1999 Amenagement et gestion fonciere urbaine et Gouvernance locale en Afrique francophone, draft report of the Meeting, Colloque des professionnels de Amenagement urbain en Afrique Francophone, Ouagadougou Avril 1999.
United Nations Economic Commission for Africa 1997 ECA Statistics Division, ECA/STAT/NAC.1/96/2, June 13, 1997. www.bestpractices.org 1999 United Nations Centre for Human Settlement,