GISdevelopment.net --> Proceedings --> GISDECO
The use of GIS in informal settlement upgrading: Its role and impact on the community and on local government
AProfessor John Abbott
Department of Civil Engineering,
University of Cape Town,
Private Bag, 7701, Rondebosch,
South Africa.
Tel. ++27 21 650 2605
Email: abbott@ebe.uct.ac.za.
Abstract
The Urban GIS Group within the Department of Civil Engineering at the University of Cape Town has been coordinating a pilot informal settlement upgrading in Cape Town since 1998. The project objective has been the evolution of a model-based approach to informal settlement upgrading that is both structured and replicable. It was felt that the only way this could be achieved was through the use of a spatial data management system operated through a GIS system. The spatial database has been used for all facets of data collection and data process and forms the basis for all decision making. Thus it covers all physical data pertaining to the site, cadastral and shack data, demographic and socioeconomic data (with an in-depth review of every household) economic opportunities and physical planning and design data. The result is a comprehensive, integrated, settlement upgrading methodology that could not have been developed without the use of GIS and spatially referenced physical capable of being rapidly updated.
Introduction
In 1998, the Urban GIS Group within the Department of Civil Engineering at the University of Cape Town initiated a pilot informal settlement upgrading in two settlements in Cape Town with a combined population of 10,00 people. Working in partnership with the communities involved, and with the local authority, the objective was to explore the feasibility of creating a method-based approach to informal settlement upgrading that could be used as a replicable model for South Africa. This method is underpinned by a spatial data management system operated through a GIS interface and it is primarily this facet of the approach that will be described in this paper.
A key objective of this upgrading methodology is to empower the community, both through the provision of detailed information on the community and then by the use of that information to support their negotiations with the local authority. The ability to access all the relevant data on their community and settlement did strengthen the community, and gave them a power base that facilitated a partnership with local government. At the same time though, the extent to which the community could take control of the data management process (as opposed to taking ownership of the data itself) was limited, due to lack of knowledge and resources to take over the technology. This is an issue that is the subject of ongoing work.
Paralleling this community process is a second thrust focussed on (primarily local) government in the Western Cape. When this project began there was an hostile attitude to informal settlement from all spheres of government. It has been argued elsewhere (Abbott, 2000) that this is due, at least partially, to a fear of informality and the implications of informality on city government. Building a structured methodology, and particularly one based on geo-spatial information, has gone a long way towards overcoming this fear. As a result the majority view in both local and provincial government is now supportive of informal settlement upgrading. Furthermore, the use of geospatial information to describe the upgrading process has facilitated discussions on levels of service, social spatial relationships and tenure options. This has led to a move away from roads, towards a pedestrian-based movement network. It has provided flexibility in terms of the type of tenure that is most appropriate.
1 Background
1.1 Urban management in developing countries
In developing countries, cities are experiencing a real demographic explosion. Today they shelter more than 40% of the population of these countries and are subjected to the doubling of their size every twenty-five years on average, and every fifteen years in East and West Africa (UNCHS, 2001).
The decentralization policies, which are being generalized, are giving the main urban management authorities to the municipalities. However, with very limited means and unqualified staff, they are unable to handle all the tasks allotted to them. In such a context, problems arise: informal settlements, construction of the unoccupied land (agricultural zones, land reserves, land unfitted for settlements), poorly maintained and insufficient infrastructures and equipments, lack of waste and wastewater evacuation and processing, natural resources degradation, etc.
Numerous experiences have attempted to provide solutions to these problems. But following the analysis of the UNECA 1 (Fourie & Nino-Fluck, 1999), the success rate is low, either because the tools are not adapted to the field reality, or because the capable staff is missing for their implementation.
1.2 The city of Thiès and the project IMAP
Tab 1: Development indicators for the city of Thies
Thiès, the second largest agglomeration in Senegal in terms of its population, is a classical example of middle sized developing city, as presented in table 1. Being managed with an annual budget of two million euros (a hundred to a thousand times less than a European city of the same size), the town has at its disposal a staff limited to four executives: an administrator, an engineer and two technical managers. They are in charge of several crucial problems:
Tab 2: Partnership and funding
At the least it is an essential component of development. This paper though would go much further, and argue that it is central to development, to the extent that informal settlement upgrading cannot be mainstreamed without it.
For the reality is that informal settlements are extremely complex environments. To date the vast majority of upgrading initiatives have been sector-based. But these take place without any real understanding of their impact on other sectors, or even their long-term impact. And there is evidence that the impact can be marginal or even negative (Verma, 2000; Amis, 2001). The philosophy underpinning this methodology is that only through the use of an effective spatial data management system can different interventions can be related, their impact evaluated and, ultimately, their input integrated. Such a system, though, will only carry legitimacy and credibility if the process is under the overall control of the community. And that is the key issue of relevance to this conference.
The use of generic data
This methodology differs from others that use GIS, such as Belo Horizonte for example, in that it is driven by the need to manage spatial data effectively, rather than manage a physical plan. This makes the relationship between the community and the spatial dataset a key element of the upgrading process. In constructing the spatial data management system around this relationship, the first point to recognise is that there are two, distinct, core datasets. The first of these relates to the dwelling, and essentially represents the physical and spatial data. The second relates to the people living in the dwelling, and provides the social and economic data. Interestingly, the cadastre, or land boundary system, is not considered to be an independent dataset, but rather to be simply one element of the physical and spatial data.
Within these two data sets, there are different layers of data, some of which are primary and some secondary, with the distinction being that, while primary data is obtained from a clearly defined source, secondary data derives from a manipulation of different primary datasets. Not all datasets need to be available to all parties. For example, communities may conduct specific surveys to meet their own needs, and these would not have to be made available necessarily to other parties. Thus it is useful to draw a distinction between primary data that is available to all parties and primary data that is specific to one particular party’s needs. The former will be termed here as generic data. It is this generic data that provides the foundation for all analysis and development in informal settlement upgrading.
As mentioned previously, the basic structuring elements for the spatial data management system are the dwelling and the people living in the dwelling, with the specific identifier for the latter group being the head of household. In database terms, these two will be given a unique identifier, and all other data will be linked to one or the other. These two database-structuring elements will then be linked, where necessary, through the dwelling identifier. Within this contextual framework, the primary data can then be grouped or structured in a number of different ways. The system described here is based upon a detailed ‘data flow diagram’ which is shown in figure 1 below (Abbott, 2001).
Figure 1: The data flow diagram for informal settlement planning (source: Abbott, 2001)
In this system, the primary generic data is differentiated according to its nature, with three major data types being defined as base data, physical-spatial data and socio-economic data respectively. Of these, the first two will be linked to the dwelling and the last one to the head of household. The first data type (the base data) is constructed around raster images (most often in the form of aerial photographs) of varying resolution, which provide a visual backdrop. These raster images provide the basis for the shack vectors map of the site, to which shack numbers can be added to provide dwelling identification.
The second data type (the physical-spatial data) relates the site itself. Thus it comprises all data relating to the site on which the dwellings are situated and cadastral data, as well as the spatial planning framework elements that impinge on that site. The term ‘physical’ is used here to describe those attributes of the site that can be seen or, if below the ground, given geo-spatial definition. Thus the term covers the topography and the natural features above the ground, as well as any additional features such as water standpipes, latrine structure or telecommunication poles. Below ground it will show the engineering services and, through the use of geotechnical surveys, the nature of the underlying ground formation and the water table. The cadastral data covers all cadastral boundaries as well as servitudes. The term ‘spatial’ relates to the non-physical spatial definition of the site. This could include zoning boundaries or spatial structuring elements already proposed for the site. For New Rest some of these are shown in figures 2 below.
Figure 2: An example of physical and cadastral elements above ground (source: Abbott and Douglas, 2001:38)
The third data type relates to the people living in the settlement. Here the generic data comprises essentially demographic data. Now there is a great deal of additional data that can be gathered about communities. In New Rest a major household survey, covering every household, was funded by an overseas donor and carried out as a community-based activity. This provided a large amount of additional social and economic data. However, this data is not essential to the development of the generic integrated database and is the property of the community, although the output from any analysis may be shared with other parties. Hence this data is termed non-generic primary data. A typical demographic data output would be the number of people per dwelling.
Having identified all the primary data types, the next issue becomes the structuring of an appropriate database. As described earlier, the base dataset is straightforward, comprising a shack vector polygon and dwelling reference number. Similarly the physical and spatial data is relatively straightforward, as it can be grouped easily according to its source. However, the data related to the community is particularly complex, and needs to be structured with care. The complicating issue here is the multiple use of this data. The development of an appropriate database was an iterative process (Abbott and Douglas, 2001, Abbott, 2001), and improvements are still being made.
The current database structure comprises a main database that contains all the data relating to the head of household and spouse. This is then linked to separate databases covering dependents and tenants. Within the main database additional fields are provided for head of household gender and marital status. This then provides the flexibility to deal with gender analysis and single parent families, both of which of crucial to the long-term success of the upgrading project.
Supporting the local authority – shared decision-making,
Once the generic data is available then the upgrading project can move forward on a number of fronts. The ‘modus operandi’ of the upgrading project is one of partnership between the local authority and the community. As those working for NGOs will appreciate, the way in which the two parties interpret the term partnership can be quite different. Local government in South Africa is caught in a web of contradictory legislative. It is not a decision-making body in its own right (on a par with national and provincial government). Yet it is recognised by the constitution as one of the three spheres of government.
In practice there are two major constraints placed on local government. The first derives from a lack of clarity to the level of discretion, authority and initiative that municipalities should enjoy. (DBSA, 2000:36). The second is the balance between what the South African White Paper on Local Government calls developmental local government and cooperative government (Republic of South Africa, 1998:19). The problem here, when it comes to partnership, is that national government defines the term developmental, and the term, as it is currently used in South Africa, is synonymous with (physical) delivery. Furthermore, local government in South Africa has a legal responsibility to ensure the operation of infrastructure services. This makes it extremely difficult to have social sustainability and economic growth as the main drivers of development.
A methodology has to be able to deal with this reality if it is to achieve success. And the way in which this is done is through a clearly defined differentiation of responsibilities, using the participation framework linked to scale. To some extent this means drawing a distinction between the physical sustainability elements of the project and the social sustainability/economic livelihoods element. It is recognised that such a distinction is inherently false, particularly given that the stated primary goal of the project is social and economic development. However, there are a number of reasons why it was felt necessary to draw this distinction.
The solution has two parts to it. Firstly the upgrading methodology is linked to an understanding of scale, as indicated earlier in the paper. At the macro scale, which is where the physical and spatial issues tend to dominate, there are four key areas that have to be addressed if informal settlements are to be set on a road towards long-term sustainability. These are physical risk, physical/spatial integration with/into the surrounding areas, movement and access and environmental health (of which the first three will be discussed in greater detail). In all of these the local authority plays a major role, and the nature of the decision-making process is one of consensus between the parties. The second part of the solution is linked to how the local authority plays its role. Specifically a mechanism had to be found that enabled the local authority to separate its developmental role from its legislative function. This was achieved through the formation of a steering committee (representing the community, its advisors and different departments within the City) to discuss all facets of the development. The steering committee agreed the development process in the different sectors, and then passed this back internally for ratification and amendment to the legislative framework (where required). In each of the three areas of activity covered her spatial data management and GIS played a major role in facilitating progress.
Physical risk
Physical risk (as it relates to a specific site) derives from three sources. The first of these is the ground, the second water and the third fire. The potential impact, hence the specific risk, will be location specific. However, it is possible to define the scope of these risks, as indicated below.
Figure 3: the results of the Geotechnical survey for New Rest, superimposed on the shack vectors (source: Abbott and Douglas, 2001:40)
Physical/spatial integration
A major characteristic of all informal settlements is the (physical and social) discontinuity that exists between the settlement and the surrounding areas. This isolation of the informal settlement is often social, as well as physical and spatial. And it is often exaggerated by upgrading interventions, where the focus is turned inwards, towards what needs to be done internally. From the beginning, the project targeted key people in the surrounding areas (local civic movement, development forums, business people) (Abbott, 2000). And it was clear that the concerns were not only about competing claims for space. They also centred around the fear of informality mentioned previously. The GIS system was used to provide spatial linkages to the surrounding areas and to integrate the settlement into the formal city. In this way, the whole process of formalisation, when communicated to those outside (supported by graphical output), played a major role in creating a framework for integration.
Movement and access
South Africa has adopted an extremely conservative approach to access in its housing policy, if only by default, in that it has not challenged the dominant planning paradigm of one house – one plot and road access to every house. Yet this paradigm, particularly the access component, is simply not feasible in informal settlements where dwelling densities can be up to 200 dwellings per hectare. This approach provides formal roads (i.e. vehicular access) along the major routes already defined organically within the settlement, and introduces a network of footpaths elsewhere. The result is shown in figure 4 below. The result of this shift in thinking has been a major saving that allows services to be provided for a lower cost per dwelling (taking into account the additional cost of servicing marginal land) than is currently the case for low-income ‘greenfield’ sites. In addition it allows for densities of between three and six time higher than current low-income greenfields sites projects in Cape Town.
Figure 4: the movement network in New Rest (source: Abbott and Douglas, 2001:83)
Supporting civil society – community-based decision-making
The active and ongoing involvement of civil society is crucial to the long-term sustainability of an upgrading programme. The perennial issue, though, is what constitutes this type of involvement. This debate has been explored elsewhere (Abbott, 2002a; b) and will not be repeated here. The main point is that the current project seeks to build a partnership between the community and the local authority, within which there are significant areas of community management and control.
Within this context, there are three broad areas of involvement. The first is in the partnership process described in the previous section. The community, though its committee structures, is a full partner in this component of the decision-making process. However, it would be naïve to think that local authorities (both politicians and officials) accept this situation easily, in spite of the rhetoric. Hence it is crucial to have strong NGO support in this area. Community Organisations need full access to all information and the technical support needed to understand the implications of different decisions. GIS plays a valuable positive role in this respect. There is no doubt that the visual medium aids the process of assimilating and processing data (and information) enormously. The result in this project was that it enabled the community to support the University team (who in this case played the role of an NGO) even where issues were not fully understood and were taken on trust. This was aided by the openness that the visual display of situations and options brings to the process.
The second area of involvement is that of community management (relating primarily to physical assets). In fact both of these first two areas have a strong focus on physical space and assets. This second area is not covered here. Community management (of infrastructure in particular) is a large and growing area of research. Whilst extremely important, it is not the driving force in an upgrading project. In New Rest, therefore, it is seen as an area to be explored in greater detail once more fundamental decisions have been taken.
The third area of involvement is the one that will be discussed here. That is the social sustainability and economic empowerment of the community, wherein the local area physical development comprises a supporting element. The experience in Cape Town is that the community has to take responsibility for its own social and economic development. It can (and should) work with the local authority but, as mentioned earlier, the local authority is so delivery driven that it struggles to understand that delivery of physical assets alone is not sustainable unless it is underpinned by sustainable livelihoods. Hence this facet of development is managed by the Community and the University, with the support of Cordaid, a Netherlands-based NGO donor. In looking at the relationship between social and economic development and spatial data management, the paper will touch on three areas where the latter plays an important role. These areas are sustainable livelihoods, social-spatial relationships and social exclusion/ marginality.
Sustainable livelihoods
An economic opportunity analysis of New Rest classified the area as a low-income ‘dormitory’ settlement. Although a relatively high percentage of the population is in formal employment, this is primarily unskilled or semi-skilled and the mean income is low (R750/$65 per month). Given the income and expenditure patterns, and the limited opportunities for additional employment within New Rest, ‘supply-side’ interventions (e.g. training) are unlikely to help in the sort to medium term (Baumann, 2002:11). Instead the thrust of the sustainable livelihoods strategy is on redirecting expenditure inwards and leveraging existing skills and assets.
The survey that generated the above information also showed where individual businesses were being operated and what those businesses were. From this information three activity areas were identified: sewing, food and beverage. The last two current represent over 80% of all businesses. Whilst it is recognised that these businesses currently add little value to the area, providing mainly convenience, they nonetheless provide opportunities for a redirection of income flows inwards. There is also the potential of increased tourist spending, New Rest already being on the tourist map. This raises the possibility, using the GIS, of identifying and creating an entertainment block within New Rest. In terms of sewing, New Rest has a particular expertise in the production of traditional Xhosa dress. The option here is to create a cooperative/joint business among the women involved, where they would either share work or share elements of the production process, drawing on experiences gained in India. In all of these areas, geo-spatial referencing is providing a valuable tool in aiding these decisions.
Social spatial relationships
A major innovation in the Cape Town upgrading methodology is the way in which it deals with the issue of space, and recognises the importance of social-spatial relationships. Community decision-making processes revolve around the participatory planning approach. However, if it is to be effective, then this has to be initiated at the appropriate scale. It also has to recognise that spatial change has a social impact that needs to be identified. This means taking decisions down to the lowest level of collective spatial unit. This unit is going to vary in size. Initially, four areas that were chosen as pilot areas for the methodology. The objective is for people in each area to make decisions about the future spatial structure of their area, the housing system to be used, access and servicing. Initially there were just two larger blocks, with areas 1, 2 and 3 forming a single block and area 4 forming the second block. However, while area 4 was found to be suitable as a spatial entity for decision-making, combined block 1-3 was totally inappropriate.
Once the survey began it was found that areas 1, 2 and 3 represented three completely different situations and needed to be treated separately. Area 1 was occupied by members of the same extended family (i.e. brothers, sisters, cousins, aunts, uncles). Doors faced inwards and there was strong social interaction between the units. Area 2 on the other hand was occupied by individuals from diverse backgrounds (nationalities, origin). Doors faced outwards; there was no interaction between residents, and each house was considered an individual entity. Area 3 was different again. Because it had more space, the committee identified it as an area where relocated families could be settled. The very existence of this space though provided an opportunity for its retention. This would however require more of a cluster housing approach to be adopted. As a result of these differences, the approach to participatory planning was quite different for each of the three areas. Figure 5 shows how the individual houses were linked to the database.
Figure 5: A raster image of a shack, plus the shack details, linked to a specific shack polygon as a non-spatial attribute (source: Abbott and Douglas, 2001:42)
Individual exclusion/marginality
Because the New Rest upgrading is situated within the context of the national housing policy, this means that housing and services are ‘being imposed’ upon the community. The use of this term may sound strange, particularly given the high demand for housing in South Africa. The reality though is that not everyone necessarily wants to own a house, or even if they do, it may not necessarily be there. In understanding this dynamic, a distinction needs to be drawn between preference and affordability. Those who prefer (from choice) not to own a house will be identified in the participatory planning process.
Of greater concern are those who cannot afford a house and who will be marginalized by the upgrading process. This is a critical issue that is being totally overlooked by the current housing policy. In New Rest the socio-economic survey is used to identify and locate all families that have particular difficulties, for whatever reason. This will then allow them to be linked into the City’s and the Province’s social welfare system. At the same time, it allows the participatory planning facilitator to ensure that they are included in the planning process.
Conclusions
The ability to represent informal settlements spatially, through the medium of a GIS, is being seen increasingly as an important, if not essential , component of the upgrading process. In the methodology described here it goes much further than this. The GIS itself provides the underpinning technology for informal settlement upgrading, while geospatial information management provides the framework for the upgrading methodology. Seen in this light, geospatial information management has the potential to completely transform the way in which informal settlements are developed. However, geospatial information management will only achieve this goal if it is accompanied by a shift from the current development paradigm.
Informal settlement upgrading is, first and foremost, about the improvement in the conditions of an identifiable community leading, in the long-term, to the development of a sustainable settlement. Spatial data management, operated in the context of a people-centred approach, is the key to achieving this. The New Rest upgrading project has attempted to provide a framework for this. It has attempted to define, and then bring together, all the different elements of the upgrading process, and then sought to relate these to each other within the framework of an integrated spatial data management system. This is an ongoing process, and dependent for its long-term success on the community being able to play a full and meaningful role. This paper has described one way in which such a system might be constructed. Whilst work is still ongoing the results to date have been very encouraging, and indicate the potential of the approach to create a replicable methodology that could be adapted across a diversity of settlement types.
References
The use of GIS in informal settlement upgrading: Its role and impact on the community and on local government
AProfessor John Abbott
Department of Civil Engineering,
University of Cape Town,
Private Bag, 7701, Rondebosch,
South Africa.
Tel. ++27 21 650 2605
Email: abbott@ebe.uct.ac.za.
Abstract
The Urban GIS Group within the Department of Civil Engineering at the University of Cape Town has been coordinating a pilot informal settlement upgrading in Cape Town since 1998. The project objective has been the evolution of a model-based approach to informal settlement upgrading that is both structured and replicable. It was felt that the only way this could be achieved was through the use of a spatial data management system operated through a GIS system. The spatial database has been used for all facets of data collection and data process and forms the basis for all decision making. Thus it covers all physical data pertaining to the site, cadastral and shack data, demographic and socioeconomic data (with an in-depth review of every household) economic opportunities and physical planning and design data. The result is a comprehensive, integrated, settlement upgrading methodology that could not have been developed without the use of GIS and spatially referenced physical capable of being rapidly updated.
Introduction
In 1998, the Urban GIS Group within the Department of Civil Engineering at the University of Cape Town initiated a pilot informal settlement upgrading in two settlements in Cape Town with a combined population of 10,00 people. Working in partnership with the communities involved, and with the local authority, the objective was to explore the feasibility of creating a method-based approach to informal settlement upgrading that could be used as a replicable model for South Africa. This method is underpinned by a spatial data management system operated through a GIS interface and it is primarily this facet of the approach that will be described in this paper.
A key objective of this upgrading methodology is to empower the community, both through the provision of detailed information on the community and then by the use of that information to support their negotiations with the local authority. The ability to access all the relevant data on their community and settlement did strengthen the community, and gave them a power base that facilitated a partnership with local government. At the same time though, the extent to which the community could take control of the data management process (as opposed to taking ownership of the data itself) was limited, due to lack of knowledge and resources to take over the technology. This is an issue that is the subject of ongoing work.
Paralleling this community process is a second thrust focussed on (primarily local) government in the Western Cape. When this project began there was an hostile attitude to informal settlement from all spheres of government. It has been argued elsewhere (Abbott, 2000) that this is due, at least partially, to a fear of informality and the implications of informality on city government. Building a structured methodology, and particularly one based on geo-spatial information, has gone a long way towards overcoming this fear. As a result the majority view in both local and provincial government is now supportive of informal settlement upgrading. Furthermore, the use of geospatial information to describe the upgrading process has facilitated discussions on levels of service, social spatial relationships and tenure options. This has led to a move away from roads, towards a pedestrian-based movement network. It has provided flexibility in terms of the type of tenure that is most appropriate.
1 Background
1.1 Urban management in developing countries
In developing countries, cities are experiencing a real demographic explosion. Today they shelter more than 40% of the population of these countries and are subjected to the doubling of their size every twenty-five years on average, and every fifteen years in East and West Africa (UNCHS, 2001).
The decentralization policies, which are being generalized, are giving the main urban management authorities to the municipalities. However, with very limited means and unqualified staff, they are unable to handle all the tasks allotted to them. In such a context, problems arise: informal settlements, construction of the unoccupied land (agricultural zones, land reserves, land unfitted for settlements), poorly maintained and insufficient infrastructures and equipments, lack of waste and wastewater evacuation and processing, natural resources degradation, etc.
Numerous experiences have attempted to provide solutions to these problems. But following the analysis of the UNECA 1 (Fourie & Nino-Fluck, 1999), the success rate is low, either because the tools are not adapted to the field reality, or because the capable staff is missing for their implementation.
1.2 The city of Thiès and the project IMAP
Tab 1: Development indicators for the city of Thies
Thiès, the second largest agglomeration in Senegal in terms of its population, is a classical example of middle sized developing city, as presented in table 1. Being managed with an annual budget of two million euros (a hundred to a thousand times less than a European city of the same size), the town has at its disposal a staff limited to four executives: an administrator, an engineer and two technical managers. They are in charge of several crucial problems:
- Firstly, a constant demand for new settlements, which is inducing a multiplication of the informal settlements, a strong land speculation, and that is degenerating after some years into a conflict with the surrounding rural villages;
- A second conflict pits the city against the neighboring rural communities, due to the problem of city expansion that lacks consultation and transparency;
- Concerning the environment, the nonexistence of any dumps and the accumulation of wastes on the vacant land requires an urgent action;
- Finally, several town districts have become unhealthy because of the lack of wastewater evacuation systems and of the frequent flooding due to the deforestation of the surrounding hills.
Tab 2: Partnership and funding
At the least it is an essential component of development. This paper though would go much further, and argue that it is central to development, to the extent that informal settlement upgrading cannot be mainstreamed without it.
For the reality is that informal settlements are extremely complex environments. To date the vast majority of upgrading initiatives have been sector-based. But these take place without any real understanding of their impact on other sectors, or even their long-term impact. And there is evidence that the impact can be marginal or even negative (Verma, 2000; Amis, 2001). The philosophy underpinning this methodology is that only through the use of an effective spatial data management system can different interventions can be related, their impact evaluated and, ultimately, their input integrated. Such a system, though, will only carry legitimacy and credibility if the process is under the overall control of the community. And that is the key issue of relevance to this conference.
The use of generic data
This methodology differs from others that use GIS, such as Belo Horizonte for example, in that it is driven by the need to manage spatial data effectively, rather than manage a physical plan. This makes the relationship between the community and the spatial dataset a key element of the upgrading process. In constructing the spatial data management system around this relationship, the first point to recognise is that there are two, distinct, core datasets. The first of these relates to the dwelling, and essentially represents the physical and spatial data. The second relates to the people living in the dwelling, and provides the social and economic data. Interestingly, the cadastre, or land boundary system, is not considered to be an independent dataset, but rather to be simply one element of the physical and spatial data.
Within these two data sets, there are different layers of data, some of which are primary and some secondary, with the distinction being that, while primary data is obtained from a clearly defined source, secondary data derives from a manipulation of different primary datasets. Not all datasets need to be available to all parties. For example, communities may conduct specific surveys to meet their own needs, and these would not have to be made available necessarily to other parties. Thus it is useful to draw a distinction between primary data that is available to all parties and primary data that is specific to one particular party’s needs. The former will be termed here as generic data. It is this generic data that provides the foundation for all analysis and development in informal settlement upgrading.
As mentioned previously, the basic structuring elements for the spatial data management system are the dwelling and the people living in the dwelling, with the specific identifier for the latter group being the head of household. In database terms, these two will be given a unique identifier, and all other data will be linked to one or the other. These two database-structuring elements will then be linked, where necessary, through the dwelling identifier. Within this contextual framework, the primary data can then be grouped or structured in a number of different ways. The system described here is based upon a detailed ‘data flow diagram’ which is shown in figure 1 below (Abbott, 2001).
Figure 1: The data flow diagram for informal settlement planning (source: Abbott, 2001)
In this system, the primary generic data is differentiated according to its nature, with three major data types being defined as base data, physical-spatial data and socio-economic data respectively. Of these, the first two will be linked to the dwelling and the last one to the head of household. The first data type (the base data) is constructed around raster images (most often in the form of aerial photographs) of varying resolution, which provide a visual backdrop. These raster images provide the basis for the shack vectors map of the site, to which shack numbers can be added to provide dwelling identification.
The second data type (the physical-spatial data) relates the site itself. Thus it comprises all data relating to the site on which the dwellings are situated and cadastral data, as well as the spatial planning framework elements that impinge on that site. The term ‘physical’ is used here to describe those attributes of the site that can be seen or, if below the ground, given geo-spatial definition. Thus the term covers the topography and the natural features above the ground, as well as any additional features such as water standpipes, latrine structure or telecommunication poles. Below ground it will show the engineering services and, through the use of geotechnical surveys, the nature of the underlying ground formation and the water table. The cadastral data covers all cadastral boundaries as well as servitudes. The term ‘spatial’ relates to the non-physical spatial definition of the site. This could include zoning boundaries or spatial structuring elements already proposed for the site. For New Rest some of these are shown in figures 2 below.
Figure 2: An example of physical and cadastral elements above ground (source: Abbott and Douglas, 2001:38)
The third data type relates to the people living in the settlement. Here the generic data comprises essentially demographic data. Now there is a great deal of additional data that can be gathered about communities. In New Rest a major household survey, covering every household, was funded by an overseas donor and carried out as a community-based activity. This provided a large amount of additional social and economic data. However, this data is not essential to the development of the generic integrated database and is the property of the community, although the output from any analysis may be shared with other parties. Hence this data is termed non-generic primary data. A typical demographic data output would be the number of people per dwelling.
Having identified all the primary data types, the next issue becomes the structuring of an appropriate database. As described earlier, the base dataset is straightforward, comprising a shack vector polygon and dwelling reference number. Similarly the physical and spatial data is relatively straightforward, as it can be grouped easily according to its source. However, the data related to the community is particularly complex, and needs to be structured with care. The complicating issue here is the multiple use of this data. The development of an appropriate database was an iterative process (Abbott and Douglas, 2001, Abbott, 2001), and improvements are still being made.
The current database structure comprises a main database that contains all the data relating to the head of household and spouse. This is then linked to separate databases covering dependents and tenants. Within the main database additional fields are provided for head of household gender and marital status. This then provides the flexibility to deal with gender analysis and single parent families, both of which of crucial to the long-term success of the upgrading project.
Supporting the local authority – shared decision-making,
Once the generic data is available then the upgrading project can move forward on a number of fronts. The ‘modus operandi’ of the upgrading project is one of partnership between the local authority and the community. As those working for NGOs will appreciate, the way in which the two parties interpret the term partnership can be quite different. Local government in South Africa is caught in a web of contradictory legislative. It is not a decision-making body in its own right (on a par with national and provincial government). Yet it is recognised by the constitution as one of the three spheres of government.
In practice there are two major constraints placed on local government. The first derives from a lack of clarity to the level of discretion, authority and initiative that municipalities should enjoy. (DBSA, 2000:36). The second is the balance between what the South African White Paper on Local Government calls developmental local government and cooperative government (Republic of South Africa, 1998:19). The problem here, when it comes to partnership, is that national government defines the term developmental, and the term, as it is currently used in South Africa, is synonymous with (physical) delivery. Furthermore, local government in South Africa has a legal responsibility to ensure the operation of infrastructure services. This makes it extremely difficult to have social sustainability and economic growth as the main drivers of development.
A methodology has to be able to deal with this reality if it is to achieve success. And the way in which this is done is through a clearly defined differentiation of responsibilities, using the participation framework linked to scale. To some extent this means drawing a distinction between the physical sustainability elements of the project and the social sustainability/economic livelihoods element. It is recognised that such a distinction is inherently false, particularly given that the stated primary goal of the project is social and economic development. However, there are a number of reasons why it was felt necessary to draw this distinction.
- It has proven almost impossible for Council officials (and to a lesser extent politicians) to grasp the centrality of social and economic issues to long-term development, for the reasons described earlier.
- There are physical and spatial elements to the upgrading, which do need to be addressed. At the same time, by driving the agenda in the way that will be described, this limits the physical/spatial interventions, thereby preventing their having a potential negative impact.
- The approach is compatible with optimised community decision-making. It can be argued, in fact, that it enhances such decision-making, though the optimal utilisation of scale and its linkage with decision-making.
The solution has two parts to it. Firstly the upgrading methodology is linked to an understanding of scale, as indicated earlier in the paper. At the macro scale, which is where the physical and spatial issues tend to dominate, there are four key areas that have to be addressed if informal settlements are to be set on a road towards long-term sustainability. These are physical risk, physical/spatial integration with/into the surrounding areas, movement and access and environmental health (of which the first three will be discussed in greater detail). In all of these the local authority plays a major role, and the nature of the decision-making process is one of consensus between the parties. The second part of the solution is linked to how the local authority plays its role. Specifically a mechanism had to be found that enabled the local authority to separate its developmental role from its legislative function. This was achieved through the formation of a steering committee (representing the community, its advisors and different departments within the City) to discuss all facets of the development. The steering committee agreed the development process in the different sectors, and then passed this back internally for ratification and amendment to the legislative framework (where required). In each of the three areas of activity covered her spatial data management and GIS played a major role in facilitating progress.
Physical risk
Physical risk (as it relates to a specific site) derives from three sources. The first of these is the ground, the second water and the third fire. The potential impact, hence the specific risk, will be location specific. However, it is possible to define the scope of these risks, as indicated below.
- Risks deriving from ground conditions. There are macro scale risks, such as earthquakes, but the majority of ground related risks are more localised. These include the dangers associated with living on steep slopes, unstable ground (e.g. sinkholes) and marginal land (i.e. land that is poorly suited to human habitation, such as contaminated land or low-bearing capacity soils, whether natural or infill material).
- Risks deriving from water. The risks here can arise from flooding (whether from river or sea), or low-lying land subject to a fluctuating water table.
- Risks deriving from fire. The major risk here is associated with the use of poor quality, fire-prone building materials. Experience in Cape Town, where this is a major risk, indicates that the most effective way of controlling fires is to prevent them starting. This can be addressed through education programs without necessarily requiring people to be moved.
Figure 3: the results of the Geotechnical survey for New Rest, superimposed on the shack vectors (source: Abbott and Douglas, 2001:40)
Physical/spatial integration
A major characteristic of all informal settlements is the (physical and social) discontinuity that exists between the settlement and the surrounding areas. This isolation of the informal settlement is often social, as well as physical and spatial. And it is often exaggerated by upgrading interventions, where the focus is turned inwards, towards what needs to be done internally. From the beginning, the project targeted key people in the surrounding areas (local civic movement, development forums, business people) (Abbott, 2000). And it was clear that the concerns were not only about competing claims for space. They also centred around the fear of informality mentioned previously. The GIS system was used to provide spatial linkages to the surrounding areas and to integrate the settlement into the formal city. In this way, the whole process of formalisation, when communicated to those outside (supported by graphical output), played a major role in creating a framework for integration.
Movement and access
South Africa has adopted an extremely conservative approach to access in its housing policy, if only by default, in that it has not challenged the dominant planning paradigm of one house – one plot and road access to every house. Yet this paradigm, particularly the access component, is simply not feasible in informal settlements where dwelling densities can be up to 200 dwellings per hectare. This approach provides formal roads (i.e. vehicular access) along the major routes already defined organically within the settlement, and introduces a network of footpaths elsewhere. The result is shown in figure 4 below. The result of this shift in thinking has been a major saving that allows services to be provided for a lower cost per dwelling (taking into account the additional cost of servicing marginal land) than is currently the case for low-income ‘greenfield’ sites. In addition it allows for densities of between three and six time higher than current low-income greenfields sites projects in Cape Town.
Figure 4: the movement network in New Rest (source: Abbott and Douglas, 2001:83)
Supporting civil society – community-based decision-making
The active and ongoing involvement of civil society is crucial to the long-term sustainability of an upgrading programme. The perennial issue, though, is what constitutes this type of involvement. This debate has been explored elsewhere (Abbott, 2002a; b) and will not be repeated here. The main point is that the current project seeks to build a partnership between the community and the local authority, within which there are significant areas of community management and control.
Within this context, there are three broad areas of involvement. The first is in the partnership process described in the previous section. The community, though its committee structures, is a full partner in this component of the decision-making process. However, it would be naïve to think that local authorities (both politicians and officials) accept this situation easily, in spite of the rhetoric. Hence it is crucial to have strong NGO support in this area. Community Organisations need full access to all information and the technical support needed to understand the implications of different decisions. GIS plays a valuable positive role in this respect. There is no doubt that the visual medium aids the process of assimilating and processing data (and information) enormously. The result in this project was that it enabled the community to support the University team (who in this case played the role of an NGO) even where issues were not fully understood and were taken on trust. This was aided by the openness that the visual display of situations and options brings to the process.
The second area of involvement is that of community management (relating primarily to physical assets). In fact both of these first two areas have a strong focus on physical space and assets. This second area is not covered here. Community management (of infrastructure in particular) is a large and growing area of research. Whilst extremely important, it is not the driving force in an upgrading project. In New Rest, therefore, it is seen as an area to be explored in greater detail once more fundamental decisions have been taken.
The third area of involvement is the one that will be discussed here. That is the social sustainability and economic empowerment of the community, wherein the local area physical development comprises a supporting element. The experience in Cape Town is that the community has to take responsibility for its own social and economic development. It can (and should) work with the local authority but, as mentioned earlier, the local authority is so delivery driven that it struggles to understand that delivery of physical assets alone is not sustainable unless it is underpinned by sustainable livelihoods. Hence this facet of development is managed by the Community and the University, with the support of Cordaid, a Netherlands-based NGO donor. In looking at the relationship between social and economic development and spatial data management, the paper will touch on three areas where the latter plays an important role. These areas are sustainable livelihoods, social-spatial relationships and social exclusion/ marginality.
Sustainable livelihoods
An economic opportunity analysis of New Rest classified the area as a low-income ‘dormitory’ settlement. Although a relatively high percentage of the population is in formal employment, this is primarily unskilled or semi-skilled and the mean income is low (R750/$65 per month). Given the income and expenditure patterns, and the limited opportunities for additional employment within New Rest, ‘supply-side’ interventions (e.g. training) are unlikely to help in the sort to medium term (Baumann, 2002:11). Instead the thrust of the sustainable livelihoods strategy is on redirecting expenditure inwards and leveraging existing skills and assets.
The survey that generated the above information also showed where individual businesses were being operated and what those businesses were. From this information three activity areas were identified: sewing, food and beverage. The last two current represent over 80% of all businesses. Whilst it is recognised that these businesses currently add little value to the area, providing mainly convenience, they nonetheless provide opportunities for a redirection of income flows inwards. There is also the potential of increased tourist spending, New Rest already being on the tourist map. This raises the possibility, using the GIS, of identifying and creating an entertainment block within New Rest. In terms of sewing, New Rest has a particular expertise in the production of traditional Xhosa dress. The option here is to create a cooperative/joint business among the women involved, where they would either share work or share elements of the production process, drawing on experiences gained in India. In all of these areas, geo-spatial referencing is providing a valuable tool in aiding these decisions.
Social spatial relationships
A major innovation in the Cape Town upgrading methodology is the way in which it deals with the issue of space, and recognises the importance of social-spatial relationships. Community decision-making processes revolve around the participatory planning approach. However, if it is to be effective, then this has to be initiated at the appropriate scale. It also has to recognise that spatial change has a social impact that needs to be identified. This means taking decisions down to the lowest level of collective spatial unit. This unit is going to vary in size. Initially, four areas that were chosen as pilot areas for the methodology. The objective is for people in each area to make decisions about the future spatial structure of their area, the housing system to be used, access and servicing. Initially there were just two larger blocks, with areas 1, 2 and 3 forming a single block and area 4 forming the second block. However, while area 4 was found to be suitable as a spatial entity for decision-making, combined block 1-3 was totally inappropriate.
Once the survey began it was found that areas 1, 2 and 3 represented three completely different situations and needed to be treated separately. Area 1 was occupied by members of the same extended family (i.e. brothers, sisters, cousins, aunts, uncles). Doors faced inwards and there was strong social interaction between the units. Area 2 on the other hand was occupied by individuals from diverse backgrounds (nationalities, origin). Doors faced outwards; there was no interaction between residents, and each house was considered an individual entity. Area 3 was different again. Because it had more space, the committee identified it as an area where relocated families could be settled. The very existence of this space though provided an opportunity for its retention. This would however require more of a cluster housing approach to be adopted. As a result of these differences, the approach to participatory planning was quite different for each of the three areas. Figure 5 shows how the individual houses were linked to the database.
Figure 5: A raster image of a shack, plus the shack details, linked to a specific shack polygon as a non-spatial attribute (source: Abbott and Douglas, 2001:42)
Individual exclusion/marginality
Because the New Rest upgrading is situated within the context of the national housing policy, this means that housing and services are ‘being imposed’ upon the community. The use of this term may sound strange, particularly given the high demand for housing in South Africa. The reality though is that not everyone necessarily wants to own a house, or even if they do, it may not necessarily be there. In understanding this dynamic, a distinction needs to be drawn between preference and affordability. Those who prefer (from choice) not to own a house will be identified in the participatory planning process.
Of greater concern are those who cannot afford a house and who will be marginalized by the upgrading process. This is a critical issue that is being totally overlooked by the current housing policy. In New Rest the socio-economic survey is used to identify and locate all families that have particular difficulties, for whatever reason. This will then allow them to be linked into the City’s and the Province’s social welfare system. At the same time, it allows the participatory planning facilitator to ensure that they are included in the planning process.
Conclusions
The ability to represent informal settlements spatially, through the medium of a GIS, is being seen increasingly as an important, if not essential , component of the upgrading process. In the methodology described here it goes much further than this. The GIS itself provides the underpinning technology for informal settlement upgrading, while geospatial information management provides the framework for the upgrading methodology. Seen in this light, geospatial information management has the potential to completely transform the way in which informal settlements are developed. However, geospatial information management will only achieve this goal if it is accompanied by a shift from the current development paradigm.
Informal settlement upgrading is, first and foremost, about the improvement in the conditions of an identifiable community leading, in the long-term, to the development of a sustainable settlement. Spatial data management, operated in the context of a people-centred approach, is the key to achieving this. The New Rest upgrading project has attempted to provide a framework for this. It has attempted to define, and then bring together, all the different elements of the upgrading process, and then sought to relate these to each other within the framework of an integrated spatial data management system. This is an ongoing process, and dependent for its long-term success on the community being able to play a full and meaningful role. This paper has described one way in which such a system might be constructed. Whilst work is still ongoing the results to date have been very encouraging, and indicate the potential of the approach to create a replicable methodology that could be adapted across a diversity of settlement types.
References
- Abbott, J (2002a) “An Analysis Of Informal Settlement Upgrading And Critique Of Existing Methodological Approaches” Habitat International 26 (3) 303-315.
- Abbott, J (2002b) “The Development Of A Generic Integrated Approach To Informal Settlement Upgrading” Habitat International 26 (3) 317-333.
- Abbott, J (2001) “The use of Spatial Data to Support the Integration of Informal Settlements into the Formal City” International Journal of Applied Earth Observation and Geoinformation 3 (3) 267-277.
- Abbott, J (2000) “Upgrading Informal Settlements for integration into the formal city”. Conference on Formal and the Informal City - What Happens at the Interface? The Nordic Africa Institute, Uppsala, Centre for Development Research, Copenhagen and the Department of Human Settlements, Copenhagen in Copenhagen, Denmark, June 15 – 18, 2000
- Abbott, J. (1996) Sharing the City: Community Participation in Urban Management, London: Earthscan Publications Ltd. ISBN number 1 85383 323 1.
- Abbott, J. and Douglas, D. (2001) A Methodological Approach to the Upgrading, In-Situ, of Informal Settlements In South Africa, Pretoria: Water Research Commission. WRC Report No 786/2/01. ISBN No 1 86845 796 6. 89pp.
- Amis, P. (2001) Rethinking UK aid in urban India: reflections on an impact assessment study of slum improvement projects, Environment and Urbanisation, 13 (1), 101-114.
- Baumann (2002) Economic Opportunity Analysis of New Rest, Cape Town, unpublished report, 27 February 2002.
- DBSA 2000 Development Report: Building Developmental Local Government. Halfway House, RSA: Development Bank of Southern Africa.
- Republic of South Africa 1998 The White Paper on Local Government, Government Gazette notice 423 of 1998, vol. 393, no. 18739. Issued by the Ministry of Provincial Affairs and Constitutional Development, 13 March 1998.
- UNCHS (Habitat) (1996) Urban Poverty: a World Challenge - The Recife Declaration, March 1996, Habitat II, Recife International Meeting on Urban Poverty, Recife, Brazil, 17-21 March.
- Verma, G.D. (2000) Inodore Habitat Improvement Project: success or failure? Habitat International
24 (1) 91-117.
Acknowledgements
The development and evolution of the methodology for the New Rest upgrading project is funded by CORDAID of the Netherlands. This builds on a previous desk study funded by the Water Research Commission. I would like to acknowledge the contribution of the community, particularly Morrison Mvumvu, coordinator of the development trust and of my colleagues who have contributed over the years to this project.