Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Informal settlement upgrading in Durban, South Africa : building institutional capacity to sustain the… Van Horen, Basil 1996

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1996-148475.pdf [ 31.63MB ]
JSON: 831-1.0058290.json
JSON-LD: 831-1.0058290-ld.json
RDF/XML (Pretty): 831-1.0058290-rdf.xml
RDF/JSON: 831-1.0058290-rdf.json
Turtle: 831-1.0058290-turtle.txt
N-Triples: 831-1.0058290-rdf-ntriples.txt
Original Record: 831-1.0058290-source.json
Full Text

Full Text

INFORMAL SETTLEMENT UPGRADING IN DURBAN, SOUTH AFRICA: BUILDING INSTITUTIONAL CAPACITY TO SUSTAIN THE IMPROVEMENT PROCESS  by BASIL VAN HOREN  B.Com (Hons)., University of Natal, 1985 M.T.R.P., University of Natal, 1988  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR IN PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (School of Community and Regional Planning)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA August 1996 © Basil van Horen, 1996  In  presenting this  degree at the  thesis  in  University of  partial  fulfilment  of  the  requirements  British Columbia, I agree that the  for  an advanced  Library shall make it  freely available for reference and study. I further agree that permission for extensive copying  of  department  this thesis for or  by  his  or  scholarly purposes may be granted her  representatives.  It  is  by the  understood  that  head of copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  The University of British Columbia Vancouver, Canada  DE-6 (2/88)  11 ABSTRACT This research examines the planning of the Besters Camp informal settlement in-situ upgrading project in Durban, South Africa.  International experience indicates that in-situ upgrading has  generally been successful in delivering services to the urban poor. It also confirms, however, that benefits of upgrading are mostly confined to project localities, and that the institutional capacity necessary for ongoing improvement in the lives of settlement residents is seldom created. Drawing from experience as a planner in some of South Africa's first generation of settlement upgrading projects, this research focuses on the question of how informal settlement upgrading can be planned in order to contribute to the continuity of the settlement improvement process. A n important implication of the 'informalization' process in many Third World cities is that parts of cities in which the de facto rules of the game are dominant are growing considerably more rapidly than the 'formal city', which is governed according to the de jure instruments of control, such as statutory law, planning regulation, and legal administrative authority. In this context, the most common approach to planning the upgrading of shack settlements is to attempt to legalize, or 'formalize the informal'. Experience in Besters Camp upgrade elucidates ways of moving beyond a narrow focus on legalization, and onto more flexible regularization strategies. These strategies attempt to bridge the gap between de jure and de facto systems by integrating elements of both into the planning process, and thereby contributing to an amended legal regulatory framework appropriate to planning in informal settlements. Regularization involves putting in place the institutions - that is, the norms and structures - that are economically and politically viable, and which have the potential to carry the consolidation process forward into the longer term. Regularization takes as its starting point the delivery of services, which provides an activity into which capacity building initiatives can be integrated. Importantly, in order to sustain the impacts of upgrading projects, it is necessary to connect local settlement-level institutions and metropolitan-level institutions in such a way as to provide a relationship of complementary autonomy at both levels.  Ill  TABLE O F CONTENTS  ABSTRACT TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES ABBREVIATIONS NOTE ON ETHNIC/RACIAL TERMINOLOGY AND CURRENCY ACKNOWLEDGEMENTS  .  C H A P T E R 1: I N T R O D U C T I O N : T H E P L A N N I N G O F I N F O R M A L  ii iii viii ' ix x xi xii  SETTLEMENT  IN-SITL U P G R A D I N G  1.1 PROBLEM STATEMENT AND RESEARCH QUESTION 4 1.1.1 The changing international context: major transformations 4 1.1.2 The planning environment: informal settlements 6 1.1.3 The activity: informal settlement in-situ upgrading 10 1.1.4 The problem: localized, short-term project impacts 12 1.1.5 Research question 14 1.2 CONCEPTUAL FRAMEWORK 15 1.2.1 The nature of informal settlements .16 1.2.2 'Informatization' and shades of informality ..17 1.2.3 Informal settlement policy formulation 19 1.2.4 Regularization and bridging the gap between de jure and de facto 21 1.2.5 Institutional capacity building and the continuity of the upgrading process 22 1.3 RESEARCH METHODOLOGY ....24 1.3.1 Case study as the strategy of inquiry 24 1.3.2 Data collection 30 1.4 OVERVIEW OF THE CHAPTERS 32 CHAPTER  2:  INFORMAL  METROPOLITAN  SETTLEMENTS  WITHIN THEIR NATIONAL A N D  CONTEXT  2.1 THE SOUTH AFRICAN NATIONAL CONTEXT FOR INFORMAL SETTLEMENT SETTLEMENT GROWTH 35 2.1.1 Pre 1990: early urban settlement ..36 2.1.2 Informal settlement growth until 1948 38 2.1.3 1948-1967: mass public housing and destruction of informal settlements 40 2.1.4 1967-1980: influx control and the re-emergence of informal settlement 42 2.1.5 1980-1990: rapid growth of informal settlement and political mobilization 45  iv 2.1.6 Political and urban transition in the 1990s .48 2.2 INFORMAL SETTLEMENT GROWTH IN DURBAN 56 2.2.1 Pre-1960: early settlement and bulldozing 56 2.2.2 Informal settlement growth between 1960 and 1985 61 2.2.3 1985 - 1990: transitionfromsurrogate state control to adversarial control 67 2.2.4 The metropolitan context for development in the 1990s 71 2.3 INFORMAL SETTLEMENT INTEGRATION AND MARGINALIZATION 75 CHAPTER 3 T H E L O C A L SETTLEMENT CONTEXT FOR UPGRADING  3.1 A 'VIEW FROM THE INSIDE': THE INTERNAL FABRIC OF INFORMAL SETTLEMENTS 77 3.1.1 'Community', and locality 78 3.1.2 'Slums of hope' 81 3.1.3 'Slums of despair' 83 3.2 BESTERS CAMP AS A MICROCOSM OF NATIONAL POLITICALECONOMIC CHANGE 86 3.2.1 Pre-1960s: Early settlement 87 3.2.2 1960-1974: Controlled settlement and the emergence of shackfarming 89 3.2.3 1975-1985: Heyday of the shackfarmers 92 3.2.4 1985-1989: Land invasions and the dominance of Inkatha warlords 96 3.2.5 Violent transfer of control and new power struggles ....104 3.2.6 Emergent participative democracy or social disorganization? 107 3.2.7 Organizational and leadership forms 110 3.2.8 External influences and interventions 113 3.3 THE DE FACTO AND DEJURE 'RULES OF THE GAME' 117 CHAPTER 4: UPGRADING  T H E PLANNING OF INFORMAL  SETTLEMENT IN-SITU  4.1 PARADIGM SHIFTS IN POLICY AND PRACTICE 122 4.1.1 Settlement bulldozing and public housing construction 122 4.1.2 The Turnerist concept of'self-help' 124 4.1.3 Sites-and-services, and in-situ upgrading 126 4.1.4 Legalization and formalization 128 4.2 SOME UNDERPINNINGS OF THE DOMINANT PLANNING ORIENTATION IN INFORMAL SETTLEMENTS .130 4.2.1 An ideological and political dimension: the informal 'other' 130 4.2.2 Rational Comprehensive planning theory and Master Planning 132 4.2.3 Links between knowledge and action 134 4.3 SO WHY NOT FORMALIZE THE INFORMAL? 137 4.3.1 Formal and informal planning processes, and different rules of the game 137  4.3.2 Reinforced marginalization of informal settlements 4.3.3 Breakdown of the social fabric 4.3.4 Missed opportunities: building on the positives 4.3.5 The challenge: a paradox 4.4 REGULARIZATION AND CONTINUITY OF THE UPGRADING PROCESS 4.4.1 Sustainability: an indulgence? 4.4.2 Integration or disintegration of the de facto informal? 4.4.3 Balance between economic, social and physical development 4.4.4 Impact on policy: linking the local and city levels 4.4.5 Economic replicability and feasibility 4.4.6 Political feasibility  140 141 142 143 144 144 146 147 ...148 150 151  CHAPTER 5: INITIATION AND CONCEPTUALIZATION OF T H E BESTERS CAMP UPGRADE  5.1 BESTERS CAMP SHACK SETTLEMENT IN 1989 5.1.1 Physical and social infrastructure 5.1.2 Demographic characteristics 5.1.3 Economic profile 5.1.4 Movement patterns 5.1.5 Legal and jurisdictional complexity 5.1.6 Political antagonisms 5.2 PROJECT INITIATION 5.2.1 The policy formulation process 5.2.2 Project identification 5.2.3 Principles underpinning the initial relationship between the UFISD and Besters residents 5.2.4 Incorporation of Nhlungwane and parochial autonomy 5.2.5 Ezimangweni and limited effective organization 5.2.6 Separation of de jure authoritiesfromproject decision-making 5.2.7 Relationship with funding agencies: Durban City Council and the IDT 5.3 THE UPGRADE AS CONCEPTUALIZED AND PLANNED ..: 5.3.1 Project components: basic health and safety, and consolidation 5.3.2 Planning principles 5.3.3 Planning procedures  155 155 157 159 161 162 163 166 166 168 169 174 176 177 178 181 181 183 187  CHAPTER 6: L O C A L SETTLEMENT L E V E L IMPACTS  6.1 SERVICE DELIVERY 6.1.1 Physical infrastructure provision: securing the conditions for basic health and safety 6.1.2 Private consolidation: tenure delivery  194 194 201  VI  6.1.3 Private consolidation: shelter upgrading 203 6.1.4 Public consolidation 205 6.1.5 Economic development and labour based construction 206 6.1.6 Cost recovery 209 6.1.7 Maintenance and ongoing management of services 213 6.2 PARTICIPATION AND PROJECT DECISION-MAKING INSTITUTIONS 215 6.2.1 Development Committees and local level control 216 6.2.2 Introduction of the 'Community Development Trust' concept 219 6.2.3 The establishment of the Inanda Community Development Trust (ICDT) 221 6.2.4 Ongoing decision-making: the ICDT as Consolidation Vehicle 223 6.2.5 Dependence relationships and capacity building benefits to Civics and NGOs 226 6.2.6 The role of women 231 6.2.7 The transition from local level control to political legitimacy 233 CHAPTER 7: THE UPGRADING PROJECT AND ITS EXTERNAL URBAN CONTEXT  7.1 'RULES OF THE GAME' IMPOSED BY DEVELOPMENT FINANCE 7.1.1 Principles underpinning the developmentfinanceregime 7.1.2 Negotiating the non-negotiable 7.1.3 The relationship of social to physical development 7.1.4 Incorporating institutional development without compromising efficiency 7.1.5 Reorganization of project finance 7.1.6 The centrality of development finance 7.2 REPLICABILITY AND ECONOMIC SUSTAINABILITY AT A CITY-LEVEL 7.2.1 Policy and organizational impacts 7.2.2 Local government and complementary capacity building 7.2.3 Costs of replicating the upgrading process 7.2.4 Funding a replicable upgrading process in the DMA 7.2.5 Constraints to economic sustainability 7.2.6 Upgrading as contributing to integration or reinforced segregation? 7.2.7 Concluding comments  236 23 6 239 240 ....242 244 247 249 249 252 254 257 263 266 268  CHAPTER 8: PLANNING TOWARD CONTINUITY OF THE UPGRADING PROCESS  8.1 CONTRIBUTIONS TO THE CONTINUITY OF THE UPGRADING PROCESS .....270 8.1.1 Basic service delivery first 270 8.1.2 The planning process 271 8.1.3 Project decision-making ........273 8.1.4 Consolidation vehicle and institutional relationships 274 8.1.5 Impact on policy 275 8.1.6 Practice-based policy 275 8.1.7 Building a skills base that can contribute to ongoing upgrading 276  vu  8.1.8 Integration of social and institutional development programmes 277 8.1.9 Land management and tenure delivery .278 8.1.10 Economically replicable and sustainable 279 8.2 REGULARIZATION AND MOVING BEYOND FORMALIZING THE INFORMAL....279 WORKS CITED  283  APPENDIX  312  viii  LIST OF TABLES Table 3.1: Table 3.2: Table 5.1: Table 5.2: Table 6.1: Table 6.2: Table 7.1: Table 7.2:  Number of shacks in Besters, Nhlungwane, and Ezimangweni 1960 - 1990 99 Overlaps and differences in the de jure and de facto 'rules of the game' 119 Jurisdictional makeup of Besters Camp settlement 162 Resident-defined areas in relation to legal jurisdictions 185 Comparison of project cost breakdown 197 Infrastructure maintenance in Besters Camp 215 Basic service level provided in Besters Camp ....255 Total capital investment required for a 10-year Durban Metropolitan Area informal settlement upgrading programme 256 Table Al: Typology of shelter in the Durban Metropolitan Area 313 Table A2: Informants by organizational affiliation, specialization and rank 314 Table A3: Services delivered in Besters Camp, 1990-1995 315 Table A4: Ongoing running costs for Besters Camp, 1995 316 Table A5: Durban's Metro Housing Development Account 1991/1-1995/6 317 Table A6: Number of deaths in political violence in South Africa, 1987-1994 318 Table A7: Capital cost breakdown 321 Table A8: Breakdown of investment components and cash flow for 10 year upgrading programme 322  ix LIST OF FIGURES Figure 1.1: Figure 1.2: Figure 1.3: Figure 1.4: Figure 2.1: Figure 2.2: Figure 2.3: Figure 2.4: Figure 2.5: Figure 3.1:  Locality: Durban, KwaZulu/Natal, South Africa 3 Informal settlements and the 'apartheid city' 8 Informal settlements and the de facto 'rules of the game' 19 Regularization and the building of institutional capacity 23 The 'apartheid city' in transition 50 Major actors in the urban development arena in South Africa, 1990 52 Chesterville and Cato Manor before removals in 1958 62 Chesterville and Cato Manor after removals 63 Informal settlements in the Durban Functional Region 71 Locality of Besters Camp within Inanda and northern Durban Functional Region 86 Figure 3.2: Mahatma Ghandi's Pheonix, Dube's and Shembe settlements in 1960 90 Figure 3.3: Dalmeny, Nhlungwane and Ezimangweni in 1977 93 Figure 3.4: Ekuthuleni, Nhlungwane and Ezimangweni in 1986 97 Figure 3.5: Ekuthuleni portion of Besters Camp in 1985 100 Figure 3.6: Ekuthuleni portion of Besters Camp in 1989 after land invasion 101 Figure 3.7: Besters, Nhlungwane, and Ezimangweni in 1990 109 Figure 5.1: Besters Camp and adjacent townships and settlements in 1990 156 Figure 5.2: Jurisdictional boundaries in relation to Besters Camp in 1990 164 Figure 5.3: Political and organizational relationships in Besters, 1990 173 Figure 5.4: Conceptual description of planning process 188 Figure 6.1: Area 2 of Besters after service installation in 1995 .....199 Figure 6.2: Project decision-making structures, 1993 222 Figure 6.3: Relationship of ICDT to other interest groups in ongoing consolidation, 1996 224 Figure 6.4: Relationship between interest groups in typical upgrading project 230 Figure 6.5: NGOs and CBOs as project decision-making intermediaries .....231 Figure 7.1: Annual capital expenditure and net cash flow of a 10-year metropolitan-wide upgrading programme 260 Figure 7.2: Annual and cumulative financing requirement of a 10-year metropolitan-wide upgrading programme 261 Figure 8.1: Regularization and the building of institutional capacity 270 Figure Al: Steps in planning procedure ..319 Figure A2: DFR, DMA and Durban municipal boundaries 320 Figure A3: Phasing of major investment components for 10 year upgrading programme ....322  ABBREVIATIONS ANC AZAPO BLA CBO CCDF CCLS CHESS COSATU CSS DBS A DCC DDA DFR DMA DMHDA GNU ICDT IDT KIP KZ KZN LFTE NCFFW NGO NPA PAC PRISM READ RDP RKDP RSA SACP SAIRR SANCO TREE IDT UDF UFISD UF UNCHS UND UWUSA  African National Congress Azanian People's Organization Black Local Authority Community Based Organization Consolidated Capital Development Fund Centre for Community and Labour Studies Centre for Health and Social Studies Congress of South African Trade Unions Capital Subsidy Scheme Development Bank of Southern Africa Durban City Council Department of Development Aid Durban Functional Region Durban Metropolitan Area Durban Metropolitan Housing Development Account Government of National Unity Inanda Community Development Trust Independent Development Trust Kampung Improvement Programme KwaZulu KwaZulu-Natal Less Formal Township Establishment Act National Council for Child and Family Welfare Non governmental organization Natal Provincial Administration Pan Africanist Congress Planned Route into Science and Mathematics Read Educational Trust Reconstruction and Development Programme RSA/KwaZulu Development Project Republic of South Africa South African Communist Party South African Institute of Race Relations South African National Civic Congress Training and Resources for Early Education Independent Development Trust United Democratic Front Urban Foundation Informal Settlements Division Urban Foundation United Nations Centre for Human Settlements University of Natal, Durban United Workers Union of South Africa  xi ETHNIC/RACIAL TERMINOLOGY The nature of South Africa's political history makes reference to racial terminology unavoidable. Frequent reference to the ethnic and racial groups as classified under the now repealed Population Registration Act of 1950 is made throughout the text, but this does not imply acceptance of such classification. Using currently accepted nomenclature, the term 'black' is used to include all those previously classified as 'coloureds', 'Asians' as well as 'blacks'. 'Indian' is preferred to the old official term 'Asian', and 'African' is used rather than the old official term 'black'.  CURRENCY The value of the Rand has dropped considerably since its creation at US $2 = Rl in 1961, to $1.22 = Rl in 1980, falling to $0.39 = Rl in 1990.  xii ACKNOWLEDGMENT  This study owes much to my work colleagues and the residents of the settlements in which I worked who provided me with an education in settlement life, politics, and development interventions. These lessons were complemented and refined by my supervisory Committee at the University of British Columbia - my supervisor Prod Laquian, Peter Boothroyd, Michael Leaf and Terry McGee - to whom considerable thanks are due for invaluable advice and guidance.  Funding for the research was provided by a grantfromthe Canadian International Development Agency (CIDA) and the Centre for Human Settlements at the University of British Columbia, and the Social Sciences Research Council of Canada, whose assistance is gratefully acknowledged.  Many thanks also to all the friends and family whose support greatly contributed to the completion of this endeavour.  1 CHAPTER ONE INTRODUCTION: THE PLANNING OF INFORMAL SETTLEMENT IN-SITU UPGRADING  This research concerns the planning of upgrading interventions in informal settlements in the cities of developing countries. Case study material is provided by an in-situ upgrading project in Durban, South Africa.  As in many other developing countries, South African cities are undergoing a process of 'informalization'. An important implication of this process is that parts of cities in which de facto rules of the game are dominant, are growing considerably more rapidly than the 'formal city' that is governed according to the de jure instruments of control such as statutory law, planning regulation, and legal administrative authority. Given the mostly 'illegal' status of so-called informally housed people in most developing countries, the role of governments has historically ranged, at worst, from the destruction of informal housing without providing any alternative in urban areas, through to attempts at upgrading, at best. The 1990s has however marked an increasingly widespread acceptance of informal settlement in-situ upgrading both as a major component of housing policy in most developing countries, and as a strategy to engage the process of informalization.  International experience has however indicated that while informal settlement upgrading has generally been successful in delivering physical services and infrastructure to the urban poor within project localities, it has been less successful in building the institutional capacity necessary to facilitate the continuity of the improvement process into the longer-term. Central to current  2  approaches to upgrading is a conception of legality and illegality which has given rise to a vogue of legalization programmes. In these terms, planners attempt to draw new boundaries around and through informal settlements in an attempt to include them in the statutorily regulated city with which they are most familiar and comfortable. Put differently, the dominant approach is to attempt to formalize the informal, or endow settlements with a status that implies legitimacy in terms of statutory law. Whilst this could take the form of non-interference by the authorities, the  delivery of physical services, or explicit legitimization via the delivery of legal tenure, the focus remains narrowly on legal status.  In order to move beyond simply attempting to formalize the informal, the question then arises as to how most effectively to engage environments that not only exist in defiance of statutory planning norms and regulations, but indeed are environments that were often established precisely so as not conform to those planning regulations in thefirstinstance This is a question that will be explored in this research via an examination of recent upgrading experience in Durban, South Africa. The case study upgrade was a precursor to South Africa's first generation of upgrades. The project traversed the 1989-1996, which was a period of rapid and dramatic political transformation in South Africa.  Although this dissertation is focused on the contemporaty policy and practice of planning in informal settlements in South Africa, it is important to be aware of the dramatic and turbulent context within which the case study is located. The suppression of challenges to the status quo throughout Nationalist Party government rule, the violent upheaval and popular protest of the 1980s, the internationally unprecedented handing over of power to the country'sfirstpopularly elected government, and the difficult shiftfromthe politics of protest to the politics of  3  reconstruction o f the 1990s provide a dramatic and rapidly changing backdrop to the subject o f this study - improving the lives of people living in particularly marginalized and deprived circumstances.  The technical and political intricacies o f upgrading are extremely complex in the South African context, where the current phase of rapid political transition presents many challenges in addressing the country's racially distorted socio-economic fabric - distortions which find particularly acute expression in the rapidly expanding shantytowns and informal settlements o f the major urban areas. Notably, over half of South Africa's African metropolitan population lives in 'informal' housing.  Figure 1.1 Locality: Durban, KwaZulu-Natal, South Africa  A l o n g with the assumption o f political power in M a y 1994, the new Government o f National Unity ( G N U ) inherited a housing sector oriented primarily toward private sector delivery of middle and upper income housing. Also inherited was the legacy o f South Africa never having  4  had a housing policy that explicitly recognized informal settlements as permanent parts of the urban fabric, and thus being entitled to resource inputs adequate to address the poor material and social conditions within the settlements. The end of apartheid rule, and with it the end of the politically sanctioned social marginalization of the majority of the population, has been a necessary but not a sufficient condition in improving the quality of life of people living in informal settlements. Central to addressing the social and economic relationship between informal settlements and the 'formal city' - and thus of specific interest in this research - is the manner in which the conceptual and operational boundaries between them are either reinforced, or amended.  1.1 PROBLEM STATEMENT AND RESEARCH QUESTION 1.1.1 The changing international context: major transformations  In understanding the changing international context, there are at leastfivemajor transformations that require consideration. With regard to thefirstfour, Perlman (1993, 20-21) provides a useful description of the process of'global urbanization' as it has manifested itself in the following transformations with respect to human settlements. • First, the trend from rural to urban is indicative of a world that is becoming predominantly urban. During the forty year period between 1950 and 1990, the world's urban population more than trebled, from 730 million to 2.3 billion. Between 1990 and 2025, urban population is projected to more than double again, to over 5.1 billion. Approximately 90 per cent of this  increase is anticipated to occur in cities of the Third World - an increase of 2.6 billion people • *• (United Nations 1993). By the year 2000, more than 50 per cent of the world's population will be living in cities, and most future population growth is projected to occur in urban areas, not rural. This trend is most pronounced in cities of the Third World, which will account for over 90 per cent of the increase in world urban population between 1995 and 2025 (ibid).  5  • Second, a shift in population growth ratesfromnorth to south is suggested by cities in industrialized countries facing stabilizing or even declining populations, while urban population growth in developing countries is dramatic. By the year 2000, the urban population of developing countries will be almost twice that of First World countries, and will be almost four times as large by the year 2025 (ibid). • Third, there is a trendfromformal to informal, which is not distributed equally throughout the urban fabric. Currently, between one-third and two-thirds of the population of Third World cities live in various conditions of 'illegal' or irregular settlements. Significantly, the trend is that the numbers of people living in squatter settlements, shantytowns, illegal subdivisions, or tenements in declining neighbourhoods, is growing at twice the rate of growth of the 'formal city' existing within formally planned, legalframeworks(Perlman 1993, 21; Hardoy and Satterthwaite 1989, 12). • Fourth, there is an increasing phenomenon of the transformation of cities to Mega Cities. Cities are reaching sizes that are unprecedented in history. By the year 2010, there will be 26 cities with populations of 10 million or more, as compared with only one city of that size fifty years ago. Twenty-one of these mega-cities will be in developing countries (UN 1993). • The context for these changes is provided by afifthmajor transformation that is currently underway. The process of globalization', which refers to the increased integration of various parts of the world to a globalfinancial,political, and cultural system, has been argued to be one of the major structural trends of the current period (Dielemann and Hamnett 1994). The increased sophistication and accessibility of transportation systems and technology, particularly information and telecommunication systems such as telephone, fax, and computer networks using satellites andfibreoptics, have facilitated the rapid integration of various parts of the world economy. The term 'globalization' is however somewhat deceiving since it is not the global  6  economy as a whole that is becoming increasingly integrated; rather, it is the 'formal city' which is effectively linked into international economic andfinancialsystems, while the 'informal city' is not.  Since informal settlements simply do not have access to even the most basic services, let alone any of the sophisticated information technology, they are at risk of becoming increasingly marginalizedfromthe urban systems of which they are part. The magnitude of the phenomenon becomes particularly significant in the context of the process of informalization in Third World urban centres noted earlier in terms of which the informal settlements, shantytowns, and illegal subdivisions are growing twice as rapidly as the 'formal city'. The link between these problems and the process of globalization is however highly contingent upon the specifics of the urban, social, and political context of the cities under examination, and the attendant form and degree of state intervention designed to address those problems.  1.1.2 The planning environment: informal settlements  Housing that exists outside of governmental regulatoryframeworkshave been described in housing literature in a variety of ways: variously being referred to as shanties, slums, shack, irregular, squatter, informal, self-help, and spontaneous settlements. Without entering into the long-standing debate on labels, and while recognizing that each description is open to criticism (see, for example, Marris 1981; Gilbert and Gugler 1992, 123), it is necessary to clarify the terms that are to be used in this research. To this end, this section begins with a clarification of the nature of settlements that are the focus of this research, followed by a discussion of definitions of upgrading.  7  In Durban, South Africa, the term imijondolo refers to a 'shack', typically constructed from wattle and daub, or alternativelyfrommaterials such as plastic, wood, corrugated iron or any combination of these, depending on the availability of materials. Imijondolo however refers to the singular, there being no indigenous term describing aggregations of shacks - as in the case offavela or barrio in other developing countries. The residents of such settlements refer to their settlements by locality name - for example, the residents' name for one of the settlements in this case study is Ezimangweni, which means 'the place of the mangos'. At a more generalized level, South African policy and research literature most commonly uses the generic term, 'informal settlements' to describe suchfree-standingclusters of shacks. Whilst not intended to imply a dualistic conception of informality, the term 'informal settlements', to be used interchangeably with'shack settlements', are therefore the terms used in this research.  Informal settlements, which are spatially discrete or separatefromformally planned housing developments, may be located on vacant tracts of land within the boundaries of formal townships and developments, or on vacant land on the periphery of urban areas. Informal settlements may exist inside or outside of local government jurisdictions, and may be located on tracts of privately or publicly owned land. However, one of the primary defining characteristics of informal settlements is that the land occupation process, the shelter construction process, and continued existence of such settlements takes place outside of the boundaries of any statutory planning or administrative frameworks. It is this category of shelter - informal settlements - that are the primary focus of this research. Informal settlement' - is preferred to the term 'informal housing' because it is a more inclusive term, incorporating the physical structure of housing in addition to other dimensions of the physical, political, social and economic environment. Whilst interlinked in terms of the broader housing problem, informal settlements are distinguished from  8  other types of informal housing - such as backyard shacks, illegal subdivisions on private estates, and housing constructed in sites-and-services schemes (see Table Al in Appendix) - not only in terms of the physical and social makeup, but as is argued here, also in terms of the planning process most appropriate in addressing the respective environments.  Providing a framework within which to examine informal settlements in the South African context is the so-called 'apartheid city' (see Figure 1.2 below). First theorized by Davies (1981, 1991), the 'apartheid city' has subsequently become widespread colloquial (for example, Lemon 1991; Smith etal. 1992; Swilling etal. 1992), referring to the following typical features of South African cities: • Residential racial segregation. • Economic segregation. This was underpinned by the premise of the Nationalist Party government that Africans were in the cities to serve as a labour pool for mining, industry, located in white designated parts of towns. This, combined with the numerous prohibitions on African businesses in the locations or townships, meant that there was very limited development of African owned business in townships and informal settlements, and these areas thus lacked a viable economic andfiscalbase. • Fragmented local government. Local government was structured according to the notion that urban areas could be compartmentalized into separate racial units presided over by racially separate local governments, with their own fiscal, legal, and a<jniirastrative systems. This gave rise to the Black Local Authority (BLA) system, which was to become a focal point of urban protest, and rents and service payment boycotts in the 1980s. In the case of Durban, this also meant the larger low-income African townships and most informal settlements falling under separate homeland jurisdiction.  9  t  \ HIGH IlNCOME:;  HOMELAND"  boundary  HOMELAND"  boundary  INFORMAL SETTLEMENTS  WHITE C.B.D. A N D EDGE  AFRICAN MUNICIPAL TOWNSHIPS  INDUSTRIAL  INDIAN/COLOURED TOWNSHIPS  SINGLE SEX MIGRANT HOSTELS  WHITE G R O U P AREAS (UPPER/MIDDLE/LOW I N C O M E )  MAJOR ROADS  Figure 1.2 Informal settlements and the 'apartheid city' (source: adaptedfrom Davies 1981)  • Separate housing delivery systems. Africans' options were either rental or leasehold status in overcrowded formal townships, or informal housing, as compared to the heavily subsidized systems catering for other race groups, particularly whites.  10  • Municipal infrastructure and services. Either limited, or no, provision was made in African designated areas. • Spatial sprawl. Service provision reinforced urban sprawl and dispersed settlement patterns, with the implication for Africans of increased transport costs, reduced access to formal sector jobs, and diminished small business opportunities due to scale and threshold limitations.  While the local government, housing delivery system, and municipal infrastructure and services provision elements have been a particular focus of political attention in current attempts to address historical inequities via post-apartheid urban reconstruction efforts (RSA 1994; RSA 1996), the fundamental spatial and economic organization of urban areas have experienced less dramatic change. By tracing the growth of informal settlements in relation to the evolution of the 'apartheid city' - itself an aberration of the African colonial city structure described by Mabogunje (1990) - it is seen that through a number of political dispensations until the present, informal settlements have increasingly grown to exist outside of the direct control of the state - thus posing a further challenge to government initiated urban reconstruction efforts.  1.1.3 The activity: informal settlement in-situ upgrading Grounded in changes in housing policy orientation - reviewed, for example, in Perlman (1987), Palmer and Patton (1988), and Hardoy and Satterthwaite (1993) - there have been several major international trends in response to the growth of informal housing. These have moved through initial attempts to restrict the influx of people into urban areas and to prevent the growth of informal settlements, to a focus on the provision of public housing, through to the site and service and in-situ upgrading projects in the 1970s and 1980s. While there is currently  11  widespread acceptance of informal settlement in-situ upgrading as a component of housing policy in Third World cities, upgrading programmes are increasingly reflecting attempts to address capacity building imperatives via attention to improving access to credit, improving urban management systems, strengthening local government and on city-wide policy change.  Informal settlement upgrading can be defined in a number of ways, many of which build upon the contributions of John Turner (1965, 1968, 1976, and Turner and Fichter 1972). In its narrowest sense, 'upgrading' of informal settlements refers to the securing of basic health and safety of the poorest urban dwellers (primarily those below the 20th income percentile) through the provision of basic infrastructure and which may include potable water, sanitation, refuse removal, roads, stormwater drainage, and sometimes electricity. In-situ upgrading, which has an emphasis on varying degrees of settlement residents' involvement in the decision-making process, also implies minimum displacement of households in the development process. The settlement improvement in Ashok Nagar, Madras, provides an example of such an approach to upgrading (Robben 1987).  A broader definition of upgrading would, in addition, include activities oriented toward the longer-term facilitation of 'consolidation' processes which create the momentum for sustainable improvement in both physical and social development terms. Such activities would include tenure delivery, economic enterprise development, health and education programmes, and housing upgrading. For example, the massive Lusaka upgrade involved infrastructure upgrading, tenure delivery, and housing unit upgrading support (Pasteur 1979; Rakodi 1986), as did the Old Naledi upgrade in Botswana (von Nostrand 1982).  12  While all these activities most often take place within upgrading 'project' parameters, a broader definition of upgrading would, in addition, be inclusive of the longer-term processes in terms of which the politically and economically marginalized status of informal settlements is reversed, or 'upgraded'. Although oriented toward the longer-term, many of these 'programmatic' mechanisms and policies need to be put in place as part of the initial project endeavour. Notably though, very few upgrading initiatives have successfully moved beyond a project-level focus, toward the achievement of city-wide impacts. A notable exception is the Kampung Improvement Program in Indonesia which Taylor (1987) and Silas (1984, 1982) describe as not only having achieved city-wide impacts in the mega-urban region of Jakarta, but also as having become a nationwide approach covering over 400 urban centres in which basic services have been delivered to over 3 million low-income kampung residents.  1.1.4 The problem: localized, short-term project impacts  My interest in planning informal settlement upgrading stemsfromdirect involvement as a planner in Durban, South Africa. Supplementing this personal experience, three broad subcategories of literature were reviewed in the formulation of the problem statement. These included: (1) literature that attempts to improve the understanding of Third World development issues via the formulation of theoretical constructs; (2) empirical analyses of aspects of Third World urban development issues, that have a policy focus; and (3) literature that has as its focus the formulation of the most effective way of implementing policies designed to address problems associated with housing and urban services in developing countries. Relevant aspects of these three categories are examined in the next two chapters (Two and Three), which outline the growth of informal settlements within a South African 'apartheid city' context.  13  International experience has indicated that informal settlement upgrading has generally been successful in delivering physical services and infrastructure to the urban poor within project localities, but not in creating the institutional capacity necessary for sustaining the momentum for improvement in the quality of peoples lives into the long-term. A common theme in evaluations of upgrading initiatives - such as contributions in Laquian (1983 a), Payne (1984), Skinner et al. (1987), Durrand-Lasserve and Pajoni (1993) - are that upgrading practice has had limited impacts on city-wide housing policy, or on institutional changes at a metropolitan level, which are essential changes if the 'upgrading' process is to be sustained into the longer-term. Whilst having identified the failure to create institutional capacity as a problem, a gap exists in the literature insofar as there is no coherent framework designed to build that capacity both at local settlement and at city-wide levels.  In an attempt to address the failure of current upgrading approaches to create the capacity necessary to sustain the upgrading process, and drawing from many of the most recent international policy and research contributions in the arena of informal settlement upgrading, some advances have recently been made in the implementation of a first generation of informal settlement in-situ upgrading projects in South Africa. These interventions nonetheless still reflect deficiencies in certain respects in terms of achieving a sustainable impact. Based on the literature and my experience in the upgrading of informal settlements, it is argued that the inadequacy of current approaches to the implementation of informal settlement upgrading are reflective of a disjunction between planning theory, policy and practice, which is rooted in problems in respect (1) of the conception of informal settlements both as entities within themselves, and in relation to the broader urban context within which they exist, (2) of the policy formulation process; and (3) of the translation of policy into practice, or implementation.  14  At the level of planning theory that informs practice, it is argued that mainstream plarining theory is characterized by a problematic conceptualization of informal settlements, with consequent problems for planning practice in which the dominant orientation is to attempt to (oxymoron-ishly) formalize the informal. It has been argued elsewhere that one of the central problems with current approaches to upgrading is that the preoccupation with, and narrow conception of, legality and illegality has given rise to a vogue of legalization programmes that fail to address what is after all the issue of primary importance:findingways of improving the quality of life of people living in informal settlements (Varley 1989; Farvacque and McAuslan 1992). The focus of mainstream planning theory which is on the statutorily planned and regulated city provides few clues as to how to engage environments that not only exist in defiance of (statutory) planning norms and regulations, but indeed which are environments that were established precisely so as not conform to those planning regulations in thefirstinstance.  1.1.5 Research question This research addresses one primary question. In addressing the question of how informal settlement upgrading can be planned so as to build the institutional capacity that is necessary to ensure the continuity of the settlement improvement process, attention will focus on  innovations as well as on problems or shortcomings in current policy and practice.  The distinction between three major types of planning theory - theory in planning, theory of planning and theory for planning (McConnell 1981) - provides a useful starting point in addressing this question. Firstly, this research involves a consideration of theory in planning, or the substantive aspects of the urban environment which planning is to affect: i.e. the internal fabric of informal settlements, as well as their relationship to the broader urban systems of which they  15  are part. Secondly, and having constructed a picture of the nature of informal settlements, attention is focused on theory for planning - which addresses the context within which planning takes place, and attempts to explain why, in that context, planning is as it is. Finally, this then provides the basis upon which to make prescriptive or normative contributions - both of a substantive and procedural (theory of) nature - in respect of planning strategies designed to build institutional capacity and thereby to sustain the upgrading process.  1.2 CONCEPTUAL FRAMEWORK A central issue considered in this research concerns the role of planning in the context of environments that exist outside of the boundaries of legal, formally planned frameworks created and regulated by governments in developing countries. As noted in the problem statement above, the theory-policy-practice disjunction finds clear expression in the tension between theoretical and policy tools designed for formally regulated contexts, and their implementation in informal environments in which a different and sometimes incompatible set of regulatory systems operate. A central difficulty for planners working in informal settlements is that in urban systems where the formal and informal are highly integrated, planners effectively straddle the two 'worlds' but have generally only been educated and equipped with tools to function at the end of the continuum where statutory regulation applies.  In order to effectively address this theory-policy-practice disjunction, planners need to understand how informal settlements function, both internally and in relation to the urban context within which they exist. This requires an understanding of the rules of the game applicable in both the formal and informal worlds. At the most fundamental level, the rules of the game in the  16  informal are defined by de facto social relations. By contrast, in the formal city, the rules of the game are defined and function according to frameworks created and regulated the state, and which are accorded 'formal' status in statutory law. In engaging the tensions between these different 'rules of the game', planners need to focus both on the potential strengths and difficulties characterizing the internal social fabric of informal settlements themselves, as well as on the structural context within which informal settlements exist. The following levels are thus examined:  1.2.1 The nature of informal settlements An understanding of the nature of informal settlements requires an examination of the internal make-up, as well as of the broader urban context within which informal settlements exist. In constructing a picture of the nature of informal settlements to be examined in this research, two complementary perspectives are considered in Chapters 2 and 3.  First, much of the literature with a focus on informal settlements in the Third World was written by authors located some distance away - both in body and life experience -fromthe day to day realities of these settlements. The authors of what can be described as 'A View from the Outside' are typically researchers based in universities, and policy makers based in international  development agencies and governments. Contributionsfromthis mainstream body of literature dominate informal settlement policy debates and policy formulation.  There is another substantial body of literature that attempts to describe and explain the workings of these same informal settlements, but which proceedsfroma starting point which focuses on the internal social structure of these settlements. It was this group, comprising people  17  who have lived and worked extensively in informal settlements and who display a sensitivity to sociological and anthropological factors, thatfirststudied and reported on the internal complexities of settlements in an attempt to construct what is described here as 'A Viewfromthe Inside'. An examination of the social history of informal settlements and their attendant forms  of social (dis)organization as well as an analysis revolving around issues such as parochialism, patronage and clientelism, perceptions, and culture, are useful in informing a more adequate and complete understanding of the very complex make-up of informal settlements.  Notwithstanding extensive critiques of dualistic conceptualizations of the 'formal:informal  1  relationship, informal settlements are still often treated - both in planning policy and practice - as somewhat 'separate' and distinct from the formal urban areas of which they are part. A useful starting point in locating an understanding of the nature informal settlements is thus via a clarification of the concept of informality as it is to be here.  1.2.2 'Informalization' and shades of informality  Consideration of urban dualism in Third World urban development literature has been voluminous. The unevenness of development in the respective worlds of the 'have's and havenot's' has been a central theme in examinations of the various dimensions of urban economic duality (reviewed, for example in Bromley 1979; Peattie 1987; Rakowski 1994). The terms of the debate have shifted from the 1970s focus on the problems of definition, terminology and whether the informal was half of a dichotomy or an extreme on a continuum, to a focus in the 1980s on quantification of informal sector activities, and more recently, on attempting to understand the nature of the phenomenon rather than the label used.  18  A point of departure for this research follows Peattie's argument that rather than placing emphasis on attempting to define the term 'informal', effort should rather be expended on understanding institutions and the way these institutions interact as a way of informing the policy issues or problems to be solved (Peattie 1987, 858). It is thus the nature of, and the relationships between, the institutions - meaning norms and structures - in such low income settlements and their urban context, that is of primary concern here.  A conceptualization that moves beyond a rigid formal:informal dualism, is the concept of 'informalization' which explains the informal as a fluid and ever-changing process, rather than as an object or a discrete sector. This drawsfromCastells and Portes' (1989) 'informalization' concept which, although focusing specifically on the informal economy, can be extended to be inclusive of the other dimensions of the urban environment - as applied in the South African context, for example, by Rogerson and Preston-Whyte (1991) and Swilling (1994).  At a superficial level, the existence of informal settlements is defined by their relationship to the formally planned city, which in turn is characterized by statutory regulation. While informal settlements are unregulated by the statutory institutions of society, they are nonetheless highly integrated into the broader urban system of which they are part. Importantly, informal settlements are also regulated, albeit by a set of norms and structures that are, in many respects, quite different to those regulating the formal city. The key issue here is the nature of regulation, and the power relationships underpinning the regulatory systems operating in the respective environments.  There are many shades of informality which cover a wide spectrum of cases. At the one  19  extreme, in a situation with no legal regulation at all (such as in squatter settlements which ignore all legal regulations), all activities would be performed in a manner referred to as 'iriformaF. Along the continuum though, increasing degrees of statutory regulation are applicable, until the other extreme is reached where, in the case of formally planned human settlement for example, the planning, servicing, building of shelter, and occupation all takes place according to statutory regulation. Even in the formulation, promulgation, and implementation of such statutory regulation however, the letter of the law is seldom strictly adhered to, and there is invariably a degree of'off the record' or 'behind the scenes' manoeuvring. Formality in an absolute sense therefore seldom exists. Rather, there is a constantly changing interplay of the shades of both formality and informality, both competing for control of social relations at all levels of the continuum.  t  Informal settlements Figure 1.3. Informal settlements and the de facto 'rules of the game'  Given that shack settlements are highly integrated within the different economic, political, social/cultural, environmental, and spatial layers of the urban system, the respective sets of de jure and de facto rules of the game also overlap in many areas, whilst in other areas they contradict  20  each other or are mutually exclusive and separate. As represented by the diagonal line in Figure 1.3 above, the interface between the spheres in which the respective rules of the game apply shifts continually, reflecting changes in the relationships and power balance between the state and civil society at a broader level, and between the 'formal city' and the 'informal city' at an urban level. It is at this interface that planners working in shack settlements are located.  1.2.3 Informal settlement policy formulation  Ultimately, policy formulation centres on competition and conflict over access to, and control over, resources. This is both a technical issue, and also an economic and political question at the root of which lie the complex sets of power relations characterizing the society in question. There are a number of issues addressed here, which include the implications of the conceptual separation between formal and informal, and of policy being formulated at a level that is detachedfromthe on-the-ground realities of informal settlements. In addition, the question of whose interests are ultimately served by policy, the sets of competing policy propositions generated by various interest groups, and the institutional relationships between those interest groups in the policy formulation process are considered.  Although development theory and empirical research has long asserted the highly integrated nature of formal and informal (Lomnitz 1978; Moser 1984), this conceptual integration appears not to havefilteredthrough to the level of housing policy and practice. In many Third World cities, laws and policies originally designed specifically for higher-income formally planned areas continue to be uncritically applied to informal settlements without taking into account their specific characteristics (Hardoy and Satterthwaite 1989). When such policy is translated into practice, it serves to further reinforce the separation and marginalization of informal settlements  21  in terms of access to urban resources and power (Gilbert 1990).  The conceptual separation between formal and informal, when translated to the level of practical intervention in informal areas - whether in terms of development intervention, or authority regulation and controls - manifests itself in a pattern where one set of practices and 'rules of the game' are applied in informal settlements, while another is applied in the formal. Alternatively, and where no policy exists which explicitly and comprehensively addresses the development of informal settlements as has been the case in South Africa, then regulations, standards, and procedures that were designed for higher income formal areas are often applied uncritically in informal settlements.  1.2.4 Regularization and bridging the gap between de jure and de facto  Planning processes that have traditionally exhibited a narrow focus on legal status, need therefore to move beyond a cursory attention to the role of de facto institutions in shack settlements, and along the lines argued by Mabogunje (1991,1992), to focus on ways of integrating those norms and structures into the planning process as a means of building institutional capacity. Based on a review of upgrading experience in Asia, Latin America, and Africa, van der Linden (1994) points out that describing an approach that imitates or integrates the characteristics of informal settlements into the planning process as a new 'model' is perhaps over-stretching the argument, and can optimistically at best, be described as an emergent direction in which low income housing solutions may be evolving.  This research explores such ways of moving beyond a narrow concern with legalization. In this regard, assistance in this regard is provided by the broader concept of informal settlement  22  'regularization'. Importantly, regularization may include a gradation of upgrading strategies that legitimize and integrate aspects of settlements' de facto institutions into the planning process, thereby also contributing to a legal regulatory framework that is more appropriate to informal settlements. A regularization process that builds on the positives of both de facto and de jure sets of rules, will mean that some of the de facto norms and structures may be eliminated completely, some may be amended, and those which contribute to a more effectively functioning urban system will be retained and integrated into the overall metropolitan governance system.  1.2.5 Institutional capacity building and the continuity of the upgrading process  In assessing the extent to which the project built the capacity of institutions necessary to sustain the improvement process, three sets of criteria are evaluated. First, regarding criteria most commonly used in evaluating upgrading projects, the upgrade is analysed with reference to physical infrastructure upgrading (especially water, roads, sanitation), employment impacts, housing upgrading, public participation, land tenure delivery, and institutional change (Laquian 1983a; Payne 1984; Skinner 1987; Durand-Lasserve and Pajoni 1993). Second, the upgrade is analysed in terms of project objectives. The implementation agents' own criteria are of significance given that they articulate the outcome of a lengthy and intensive public participation process. The upgrade, then, is also analysed in terms of the extent to which it achieved its own objectives of "permanence, replicability, upgradeability, affordability, 'community' empowerment, and sustainability" (UFISD 1990a). Third, the upgrade is analysed in terms of whether it built the institutions necessary to sustain the upgrading process. This involves an examination of the factors that inhibited or contributed to the building of institutional capacity.  23 local settlement, and relationship to urban system  de jure & de facto regulatory frameworks  nature of informal settlements  planning of settl = ment upgrading  REGULA RIZATION  norms  « • Institu onal capacity building  structures, organizations  continuity of improvement process Figure J. 4  Regularization and the building of institutional capacity  These evaluative criteria will be organized with reference to the extent to which the upgrading process resulted in improvements in the following areas: • First, immediate improvements in the material and social conditions of life for ordinary informal settlement residents are assessed. Addressing basic needs is a fundamental prerequisite to the viability of any longer term programmes.  24  • Second, whether the project built the institutions - meaning norms and structures - to sustain the improvement process is assessed (see Figure 1.4 above). In this regard, the capacity of settlement residents to engage their built and social environment via taking control of decisionmaking control is examined. In addition, the capacity of external agents such as local government and NGOs to engage informal settlements with an understanding and sensitivity to the de facto norms and 'rules of the game' in terms of which the informal functions, rather than exclusively in terms of the de jure rules applicable in the formal city as has traditionally been the case, are assessed. Finally, the relationship between the local settlement and city levels are analysed with reference to the political legitimacy and the economic rephcability and sustainability of the upgrading process. The building of institutional capacity is thus analysed not as an end in itself, but rather in terms of it being a means to the end of sustaining the continuity of the upgrading process, and thereby, of the quality of settlement residents lives.  1.3 RESEARCH METHODOLOGY 1.3.1 Case study as the strategy of inquiry  Having described the research design, the subject of study, as well as the elements of the conceptual framework, it now becomes necessary to outline the strategy for the collection of empirical materials. At the bottom line, the strategy selected is determined by the skills, practices and techniques available when movingfromthe paradigm and research design, to the collection of empirical materials necessary to answer the research question (Denzin and Lincoln 1994, 202).  The strategy most effective in obtaining the information necessary to answer the research question presented above, is that of case study. Given that the case is something that operates or functions, and the study is the observation of those operations, the case study in this research  25  thus refers to the observation and analysis of the in-situ upgrading operation or process.  In this research, one detailed case study is more useful in helping formulate an understanding of the social reality of upgrading than rapid comparisons with multiple cases. The type of case study employed here is an instrumental case study, which refers to the examination of a particular case in order to provide insight into an issue or a refinement of theory. While the case is examined in detail, its context scrutinised, and its ordinary activities detailed, this is in order to inform the process of answering the research question (Stake 1994, 237). The case itself is of secondary interest: it plays a supportive role, facilitating an understanding of the subject of inquiry.  Drawing generalizable conclusionsfroma single case is, however, legitimate provided the impact of contextually specific variables on the in-situ upgrading process are recognized. Such general conclusions are further strengthened by conducting both micro and macro level analysis via an examination of the institutional mechanisms and relationships that connect and shape what happens at both levels. This research thus traverses levelsfromthe national political-economy, to the provincial, city, and local informal settlement levels.  Comparison further strengthens general conclusions that may be drawnfroma single case. However, comparison is, in itself, less important than the insights derivingfromthe case itself. Although the case is selected so as to represent the phenomenon more generally, the selection of the case is primarily informed by the opportunity to learn. It is optimal therefore to select a casefromwhich we can derive the most useful insights in refining or shaping theory and policy in respect of upgrading: "potential for learning is a different and sometimes superior criterion to  26  representativeness. Often it is better to learn a lot from an atypical case than a littlefroma magnificently typical case" (Stake 1994, 243).  Detailed description, or what Clifford Geertz (1973) calls "thick description" of a single case, provides the opportunity to dissect the upgrading process and examine selected aspects such as service delivery, public or private consolidation. This layering of the description and analysis within a single case thereby provides a basis upon which to make comparisons with selected aspects both of upgrading projects in other contexts, or for future studies of the same case. Such comparisons are made at the quantitative level, and with reference to the more complex social underpinnings of the upgrading operation. To the extent that comparison is employed as a component of this research, this is done in order to compare aspects of the overall approach employed in the case under examination to other commonly used approaches (specifically, regularization versus legalization). Ultimately, the purpose of such comparisons is to support or refine some of the insights derivedfromanalysis of the primary case. Case study selection  Research for this study was carried out in Besters Camp, an informal settlement in Durban, South Africa. The Besters Camp case study examines thefirstsuccessful attempt at implementing a large-scale informal settlement in-situ upgrading project in South Africa. Why this project as case study? First, the case study is a current (post-1990) attempt at implementing a comprehensive approach to upgrading, and therefore provides up to date material on policy and practice in this arena. Second, the approach adopted in this upgrade was informed by recent international experience and literature in the informal settlement upgrading literature, and was conceptualized with a view to constructively engaging many of the problems cited in pre-1990  27  evaluations of other upgrading projects. Third, the upgrade was a demonstration project and was designed with the explicit intention of providing generalizable lessons both in the South African context, and internationally. Fourth, the period during which the upgrade has taken place is of unique interest both in terms of South African and international history, spanning its conception during a State of Emergency in the 1980s, and its implementation and early functioning of South Africa'sfirstpopularly elected government at national, provincial and municipal levels.  Tracing the social history of the settlement, it is seen that the growth of Besters Camp was in many respects a microcosm reflective of changes in the broader South African political economy. The case study then, provides an opportunity to examine the extent to which many of the most recent policy contributions in the literature were effectively integrated in the project conceptualization, and the extent to which they were followed through in implementation and longer-term institutional capacity building programmes. Case study context and description: South African informal settlement growth  Between 1950 and 1990, natural population growth has accounted for South Africa's total population almost trebling,from13.8 million to 37.8 million. The total population is anticipated to reach 73.2 million by the year 2025, reflecting a small decline in annual rates of total population growthfromaveraging between 2 and 3 per cent between 1950 and the year 2000, down to between one and two per cent between 2000 and 2025 (XJN 1993). Between 1950 and 1990, South Africa's total urban population grew more than threefold,from5.9 million to 18.7 million. Between 1990 and 2025, it is projected to more than double again, to 50.2 million reflecting average annual urban population growth rates of 3.1 per cent until 2010, and then declining to 2.13 per cent by the year 2025. This is indicative of a changing population  28  distribution such that 43 per cent of the South African population was urbanized in 1950, increasing to 50 per cent in 1990, and to 68 per cent by the year 2025 (ibid). As is consistent with trends in many other developing countries (Preston 1988), these rates of increase are primarily a result of natural increases of the urban population, and to a lesser extent,fromchanges in the rates of rural-urban migration.  While South Africa's apartheid urban policies were relatively ineffective in halting the process of urbanization, they had a significant impact on urban settlement patterns. One of the consequences of South Africa's colonial and apartheid history is that over 8 million African people currently live in informal housing. This is spatially concentrated such that over 70 per cent of the country's informally housed people live in South Africa'sfivelargest metropolitan areas Johannesburg, Durban, Cape Town, Port Elizabeth and East London. One infiveSouth Africans live in informal housing. This includes over 51 per cent of the African metropolitan population of the country (UF 1991, 7; RSA 1996).  A central component of the post-apartheid government's Reconstruction and Development Plan (RDP) is the provision of formal housing for all, involving a stated commitment to the delivery of one million formal houses by the year 1999. However, given the extent of the backlog, the state of the economy, and the need to put the institutionalframeworkfor such a vast delivery framework in place, the ability to deliver on this promise may be questioned. More fundamentally, not only does the future of existing informal settlements need to be explicitly addressed, but so too does the relationship between formal and informal housing delivery processes need to be clarified.  29 Besters Camp informal settlement in the Durban Metropolitan Area  The upgrading project that is the focus of this research is located in the Besters Camp informal settlement in Durban, which is South Africa's second largest metropolitan area. After the extended Johannesburg-Pretoria metropolitan complex within Gauteng which has a population of over 8 million people, the Durban Functional Region is the second largest urban area in South Africa with a population of 4 million. Of the Durban Functional Region population of four million people, approximately 1.8 million live in informal settlements, mostly on the periphery of the metropolitan area.  The Besters Camp informal settlement is located to the north of the Durban Central Business District, within the sprawling shack area of Inanda. Inanda is the largest informal settlement in the Durban metropolitan area, accommodating over three-quarters of a million people. The area referred to as Besters Camp refers to a 160 hectare settlement within Inanda, comprising four smaller localities as defined by settlement residents themselves - Besters, Mzomusha, Nhlungwane and Ezimangweni - each with its own unique history, social organization, and physical characteristics which are widely divergentfromlegally defined authority structures and cadastral boundaries.  Prior to the upgrading intervention, Besters Camp settlement was characterized by extremely low levels of services and infrastructure - the four water taps in Nhlungwane and Ezimangweni were the only water points serving the 50,000 people Uving in the area who otherwise purchased waterfromresidents of adjacent formal townships. There were no formally provided internal roads, nor stormwater control, sullage control, or electricity. Regarding social facilities, there were no formally provided facilities such as schools, creches, clinics, meeting halls  30  or churches. Furthermore, none of the residents were in possession of legal tenure to the land they occupied. The housing stock was informally constructed, at densities of up to 90 dwellings per hectare in some areas, with construction on very steep gradients.  At the time of project initiation in 1989, there had never been a large-scale in-situ upgrade successfully implemented in South Africa, and it was in this context that an in-situ upgrading project was planned by The Urban Foundation, a South African Non Government Organization, as a pilot 'demonstration' project. The project represented an attempt firstly to arrest further downgrading and to secure basic health and safety via the installation of basic infrastructure including water, sanitation, refuse removal, roads, stormwater removal and electricity. Second, planning interventions were concerned with how to facilitate a process of'consolidation' where the general living environment was to be improved. This involved the provision of freehold tenure, mobilizing other support actors (both state and NGOs), providing access to building materials and technical advice, stimulating local informal enterprises, facilitating investment in public (particularly health and education) facilities, and formulating 'appropriate' local regulations. Attention was also to be given to the question of informal settlement governance and the institutional framework within which this should take place, at a broader metropolitan level. The planning process was initiated by the Urban Foundation Informal Settlements Division (UFISD) in 1989, and initial funding was secured and implementation commenced in 1990. Major infrastructure installation and tenure delivery was completed in 1995.  1.3.2 Data collection Written materials •  Project management documentation. This included all the development agent's original  31  policy, technical, social/political, financial, and legal project management documents. This original project documentation provides a detailed 'inside view' of various dimensions of the upgrade as it progressed both prior to, and throughout, the implementation of the project, from the perspective of the interest groups directly involved, including consultants (engineering, planning, land survey, legal, financial), authorities (municipal, provincial government), and UFISD project staff. •  Policy documentsfromthe Urban Foundation, as well as policy documentsfromofficial  bodies (local, provincial, homeland and central government), and funding agencies. • Publications, conference papers and newspaper articles which focus on the settlement and the  upgrading project. • Surveys. A number of socio-economic and health surveys were conducted before and during  the upgrade. Participant observation  Participant observation - which refers to observation conducted when the researcher is playing an established role in the instance being studied - has been widely used in the context of Third World development initiatives, and in upgrading projects (Salmen 1987; Atkinson and Ffammersley 1994).  Participant observation provides the primary source of information for this research. Participation took the form of a direct involvement as a planner in the informal settlements of Inanda between 1988 and 1993. The most intensive period of involvement in the area was between 1989 and 1992, when I was part of the project team responsible for coordinating the planning, initiation, and implementation of the Besters Camp upgrading project. My full-time  32  office location was in the Besters Camp informal settlement. Unless otherwise indicated, Besters Camp social history and upgrading project data was sourced from participant observation during this period. Interviews Given the fluid and complex social fabric of the Besters Camp informal settlement, the interview style considered most appropriate due to its high degree of flexibility and latitude in the interaction between researcher and the respondent, was the semi-structured interview. In order to develop a more informed understanding of the issues as perceived by the major actors both directly and indirectly involved in the upgrade, 25 semi-structured interviews of residents of the Besters Camp settlement, members of Community-Based and political organizations, government officials and politicians, funding agencies, Non Governmental Organizations, and development agents were conducted between April and June 1995 (see Table A2 in Appendix). The interviews were open ended and semi-structured. Having started with a list of general questions, I narrowed down the questions as thefieldworkprogressed. Maintaining a consistent set of questions however enabled cross-checking and verificationfrominterview to interview. Interviews were conducted as informal dialogues, rather than as formal question-answer sessions.  1.4 OVERVIEW OF THE CHAPTERS Chapters Two and Three deal with theory in planning, or the substantive aspects of the urban environment which planning affected. Given that many development interventions in informal settlements are grounded in a problematic understanding of these environments, the purpose of Chapters Two and Three is to construct such a picture of the complex internal fabric of informal settlements, and their relationship to their urban context, so as to provide a starting  33  point for the ensuing analysis. Drawingfromboth international and South African literature in which broader structural political, economic, and legal dimensions of settlements are discussed, Chapter Two provides a description of the historical context for the case study, at both national and metropolitan levels. Following a discussion of urbanization and informal settlement growth in South Africa, attention is focused on housing policy with particular reference to informal settlements. The growth of informal settlements in Durban is described.  Chapter Three is focused on the internal fabric of settlements. This includes a consideration of the meaning and usefulness of the 'community' concept as a preface to a more detailed examination of what is described here as 'a viewfromthe inside' - perspectives focusing on sociological and anthropological aspects of informal settlements. This involves a brief history of the growth of the Besters Camp case study settlement, up to initiation of upgrading in 1990. This history of Besters Camp indicates how changes at the settlement level were a microcosm of national level political changes.  Having constructed a picture of the nature of informal settlements generally, and of the case study settlement specifically, Chapter Four is focused on theory for planning - which addresses the context within which planning takes place, and attempts to explain why, in that context, planning is as it is. This involves a critical examination of existing approaches to upgrading intervention in informal settlements. In order to provide a context for this discussion, the chapter begins with an outline of the paradigm shifts in the housing literature. The currently dominant approach to planning in informal settlements - described here as 'formalizing the informal' - is critiqued. This is followed by an outline of key evaluative criteria that need to be considered in the construction of aframeworkfor planning in informal settlements.  34  This provides the context for the following chapters which focus on what the upgrade delivered, how the delivery took place, and why delivery took place the way that it did - all with reference to the theme of whether the delivery process either inhibited or contributed to the continuity of the improvement process. Chapter 5 begins with a brief description of the profile of Besters Camp immediately prior to upgrading in 1990, following which the relationship between policy formulation and implementation that was applied on the upgrade, is elucidated. The shaping of the participation strategy and how this informed the initial conceptualization of the project planning process is then considered. This provides the basis for the examination in Chapter 6 of the local settlement level impacts of project implementation.  Chapter 6 considers local level impacts of what the project delivered, and via an examination of the primary institutions involved in the upgrade and their relationships with each other, it examines how project decision-making evolved throughout the 1990 - 1995 period. This provides a basis upon which to examine the factors informing the upgrading process, in which particular attention is paid to the significant role played by developmentfinance.Complementing the analysis of local settlement-level impacts, the following chapter - Chapter 7 - examines the major external impacts on the project, as well as the impacts of the project on the urban level, in which economic replicability, as well as political, policy and organizational impacts bear consideration.  Thefinalchapter, in which contributions are primarily normative and prescriptive, presents a synthesis of the lessons learned from an examination of the project. Areas in which upgrading projects can make a contribution to the continuity of the upgrading process are presented.  35  CHAPTER TWO INFORMAL SETTLEMENTS WITHIN THEIR NATIONAL AND METROPOLITAN CONTEXT  As suggested in the conceptual framework in Chapter 1, a starting point in developing an understanding of the nature of informal settlements is via an examination of the national and urban context within which settlements exist. A central theme of this Chapter concerns the 'asymmetrical integration' of informal settlements within their urban context - an asymmetry that has been particularly acute in the South African context. The growth of informal settlements, with particular reference to the tensions and evolving relationships between the de facto and de jure, are described first at a national level in South Africa, and then at the level of the Durban  metropolitan area. This Chapter thus provides the backdrop for the discussion in Chapter Three which moves down to the local context for upgrading, and considers the changing nature of the internal fabric of settlements.  2.1 THE SOUTH AFRICAN NATIONAL CONTEXT FOR INFORMAL SETTLEMENT GROWTH  While informal settlements are often treated as being separatefromthe formal city, it is important to recognize that "...settlements cannot empirically or theoretically be understood unless they are looked at as parts of a larger system and as products of the operation of the system" (Leeds A and Leeds E 1970, 230). Or more simply put, "...the two worlds are integrated within one system" (McGee 1979, 2). Informal settlements in South Africa are best understood via an overview of their growth within the context of the evolution of South Africa's particular form of urban residential segregation and control according to racial criteria. This not only  36  predates the post-world war II 'apartheid' era, but also predates the discovery of gold and diamonds in the 1880s. A centrally important aspect of the growth of informal settlements described in the following chapters is related to the process of informalization whereby informal settlements - whilst simultaneously spatially segregated and yet economically highly integrated within urban systems - have developed a growth momentum and sets of governing institutions that are quite distinct from those in the remainder of the urban areas of which they are part.  2.1.1 Pre-1900: early urban settlement  "...the government is like a man who has a cornfield which is invaded by birds. He chases the birdsfromone part of thefieldand they alight in another part of the field... we squatters are the birds. The government sends its policemen to chase us away and we move off and occupy another spot. We shall see whether it is the farmer or the birds who get tiredfirst..."(Oriel Monogoaba, squatter leader, quoted in Stadler, 1979, 19). At the time of the legal abolition of slavery by the British in 1834, there were few towns in Southern Africa (Mabin 1992). The largest non-rural settlements were the massive residential villages of indigenous chiefdoms, but these were devoid of thefinancialand commercial institutions which grew rapidly in the colonial ports such as Cape Town and Port Elizabeth from the 1830s onward, and which were replicated in Boer centres such as Potchefstroom after the 1840s. Land dispossession, military exigency and the increasing commercialization of exchange relationships had the implication that many African settlements indigenous to areas experiencing colonial and Boer expansion were rendered increasingly susceptible to involvement in the growing wage labour economy by the 1850s (ibid:\3-\4). Many of the early shack settlements located on the peripheries of urban centres such as Cape Town were set up by these in-migrants, as well as ex-slaves, by then employed in the wage economy (Western 1981). Notably though, even the major 'urban centres' such as Kimberley and Johannesburg were themselves initially shantytowns,  37  and were only later formalized (Smit 1990).  While substantial rural to urban migration took place throughout the 1800s in order for migrants to take up employment in the growing towns, it was the deliberate policies of the authorities to extract labour for white-owned farms and mines late in the nineteenth century that provided the major impetus to the decline of African agriculture in thefirstinstance, and to the growth of rural-to-urban migration in the second.  Throughout the nineteenth and twentieth centuries, migration patterns have remained essentially circulatory in nature. While state policy contributed to, and reinforced, circulatory migration, Mabin (1990) points out that these movement patterns cannot be understood solely as the outcome of state policy. They were, in addition, influenced by factors such as the relatively superior quality of life possible in some rural areas for poorer people, the importance of rural bases to historical, cultural and spiritual origins, the social significance of a 'home of one's own', and the highly segmented urban and rural labour markets. In this context, informal settlements have continued to serve as transit points for rural in-migrants, as permanent homes for both rural in-migrants as well as for second and third generation urban dwellers who have moved out of overcrowded formal townships, and as second (temporary or permanent) homes for seasonal migrants who maintain a rural base (ibid, 318).  The post-1880 period was significant largely due to the magnitude of economic expansion, foreign investment, and the increased centralization of control of the newly established diamond and gold mining industries. The impact in terms of urbanization revolved around the incorporation of all parts of rural South Africa into the rapidly growing wage economy. African  38 people were subjected to increasing pressures designed to dispossess them of their land: hut taxes, poll taxes, pass and vagrancy laws, and the demarcation of pockets of (usually marginal) land as 'reserves' within which only African people could live (Blundy 1992). Although most rural households found it difficult to avoid participation in the urban based economy through selling their labour, most remained linked to the rural areas, in many instances via the occupation of their land by part of the family remaining in the rural area (Mabin 1992, 15).  The state assumed a more coherent and powerful form with the establishment of the Union of South Africa following the defeat of the Boer republics in the War of 1899-1902 and their annexation to Britain. The foundation of 'native policy was to be a land policy which 1  focused on the racial separation of residential areas, and legislation was enacted to enforce this segregation (Davenport 1991, 1). Within towns, it was the introduction of the compound system in Kimberley's diamond mines that represented one of the primary elements shaping urban 'native policy and of modern strategies of control. This was the first rigid form of residential segregation 1  applied in South African cities (Rex 1974; Crush and James 1991). By the turn of the century, Smit (1990) notes that aside from those living in overcrowded mining compounds, the majority of Johannesburg's black population lived in backyard shacks and pockets of informal settlement.  2.1.2 Informal settlement growth until 1948 The 1913 Land Act continued to add land to the 'reserves', outside of which Africans were prohibitedfromowning land, while in the urban areas, the African population continued to be accommodated either in compounds, backyard shacks or informal settlements - most of which were considered health hazards by the authorities. By the 1920s, the numbers of Africans living outside of the compounds had increased such that, for example, in Johannesburg, approximately  39  5,500 were accommodated in formal municipal locations, 6,000 in freehold or leasehold townships on privately owned land, and approximately 50,000 lived as tenants on white-owned properties, occupying shanties, out rooms, or disused buildings let by slum landlords (Maylam 1990, 6).  Outbreaks of bubonic plague, influenza epidemics and the plague in some of these shanty and slum areas not only brought into focus the poor living conditions of urban Africans, but also compounded the authorities' antipathy to informal housing. One consequence of this hostility to shack housing was the destruction of many of the 'slums', and their replacement with formal townships which were usually located on the periphery of the urban centres (Morris 1981). During this period, one of the primary rationales used to justify this policy was a 'public health', or 'sanitation syndrome' argument - the alleged risk of disease emanating from slums and shantytowns (Swanson 1977; Torr 1987).  The principles underpinning the planned separation of racial groups were formalized following the Stallard Commission of 1918 (the Transvaal Local Government Commission), which articulated quite clearly that blacks should only be allowed in urban areas on a temporary basis and only for as long as their labour was required by white employers. This was summed up in the oft-quoted statement of the 'Stallardist doctine' that "the native should only be allowed to enter the urban areas, which are essentially the white mans's creation, when he is willing to enter and minister to the needs of the white man, and should depart therefrom when he ceases so to minister" (Transvaal, 1922, para.42).  Legislation embodying Stallardist principles and designed to deal with the changing urban  40  environment was enacted in 1923 in the form of the Natives (Urban Areas) Act. This legislation was significant in its attempt to restrict the influx of African people into urban areas, and the empowering of local authorities to allocate land and provide formal public housing for Africans within segregated residential areas, referred to as locations. Although some local authorities proceeded with the construction of locations, no serious attempt was made to control the influx of Africans into most towns during the 1930s (Davenport 1991, 8). As a consequence, and notwithstanding the harsh measures contained in the influx control legislation, Mabin (1992) points out that informal settlements grew rapidly on the peripheries of many cities and towns, often immediately outside local authority boundaries.  During the Second World War, urbanization continued to increase rapidly, largely as a consequence of rapid industrialization and declining rural economies. While the number of urban Africans almost quadrupled from 587,000 in 1921, to 2,329,000 in 1951 ~ representing an increase in the percentage of urbanized Africans to 28 per cent ~ this influx was matched with little formal housing construction by the state. During this period, only 99,250 houses were built, and only half of them for Africans (Maylam 1990; Lemon 1991, 4). Notwithstanding the consequent increased confrontation over the implementation of municipally controlled pass laws (Davenport 1991; McCarthy and Smit 1989), there was rapid growth of informal settlement in most cities and towns such that by 1948, approximately 58 per cent of South Africa's urban African population were 'squatters' on unserviced land (Posel 1991a, 20).  2.1.3 1948-1967: mass public housing and destruction of informal settlements  One of the Nationalist Party government priorities upon coming to power in 1948 was to formulate a response to what it perceived to be 'the urban native' problem. Reflecting many of  41  the notions underpinning the Stallard doctrine, legislation designed to create the framework for the 'racial and ethnic spatial engineering' that was to characterize the next four decades, was enacted. The cornerstones of this legislation were the Group Areas Act of 1950, the Prevention of Illegal Squatting Act of 1951, and legislation designed to enforce influx control and the urban labour preference policy.  A primary component of the solution, according to the Nationalist government, was to restrict the growth of the urban African-population, and ekminate urban unemployment by means of an influx control policy which would restructure the urban labour market in accordance with an 'urban labour preference principle' and the 'differentiated labour power' requirements of employers (Hindson 1987, Posel 1991a). In this context, influx control policy stipulated strict conditions under which Africans could live and work in urban areas, and it wasfromthis limited pool of labour that employers were required to draw their workers. Furthermore, afreezeon the permanent influx of rural in-migrants was enforced whereby their right to be in an urban area was terminated once their employment contract had lapsed. An extensive (although inefficient) administrative machinery was set up in order to implement the urban labour preference policy and to restructure prevailing employment patterns by legal and administrative means - even though the logic and momentum of this administrative system was often out of synchronization with the changing labour requirements of employers (Posel 1991a, 23-27; Posel 1991b).  Part of the imperative of control of the urban African population was clearly reflected in the bulldozing of informal housing, and the implementation of mass public housing schemes. During the 1950s the government began the systematic destruction of shanty-towns and inner city 'slums', among the largest of which included Cato Manor in Durban, District Six in Cape Town,  42  and Sophiatown in Johannesburg. Platzky and Walker (1985) point out that 730,000 urban Africans were resettled between 1960 and 1983. These removals were complemented by mass public housing construction programmes designed to accommodate people who were relocated from inner city areas, in-migrants, as well as the rapidly growing urban population. The massive townships of Soweto (Johannesburg), KwaMashu and Umlazi (Durban), and Nyanga and Gugulethu (Cape Town) were constructed during the 1950s and 1960s.  2.1.4 1967-1980: influx control and the re-emergence of informal settlements  By the mid-1960s, informal settlement within and around most South Africa's towns and cities had largely been eliminated, and in their place were the sprawling formally planned townships (Harrison 1992, 16). Linked to the urban component of government policy was an ideological underpinning of the segregationist 'homeland' policy which asserted that all Africans were basically 'tribalized' at heart, and therefore spiritually and culturally rooted in the ethnically defined 'homelands' in rural areas (Posel 1991a). Not coincidentally, these homelands also comprised the most economically marginal land.  In 1967, the government made the decision to stop all public housing construction outside the 'homeland' boundaries - afreezewhich remained in place until the 1990s. This policy was linked to the state's intention that all further public housing construction was to take place within homeland boundaries - where 66 new 'towns' were established between 1960 and 1970 (Harrison 1992, 16) - thereby also displacing the responsibility for the ongoing administration and maintenance of that public housing stock onto homeland authorities. Notwithstanding attempts at stringent implementation of influx control legislation, the extent of urbanization continued to accelerate throughout the 1960s (Simkins 1983; Hindson 1987), with the consequence that  43  informal settlements began re-emerging on the peripheries of urban areas, particularly those in close proximity to homeland boundaries.  Those who were excludedfromlegal' status in the formal townships proceeded to build shelter for themselves in increasing numbers, and so whereas in 1960 informal settlements on the periphery of urban areas had been virtually eliminated, by 1980 they accommodated approximately 3.7 million people - all of whom were Africans (Simkins 1983). The significance of the growth of settlements immediately outside of municipal boundaries is that by 'escaping' from state regulation of the economy (in terms of taxes and legislation, as well as control over the production and exchange of licit and illicit goods), the politically and economically disenfranchised are in some instances able to create a space for themselves in which to operate with greater benefit to themselves. In other instances, however, they do so because they have no alternative (Peil 1976; Lustig 1990).  Given the inability of the state to effectively control the increased growth of settlements, it is important to understand some of the factors underpinning not only the genesis, but also persistence of the informal settlement process. International experience indicates a variety of forms via which initial settlement may take place including purchases of landfromdevelopers, illegal subdivisions, and land allocations by politicians in terms of client-patronage relationships (Gilbert 1990). In addition, South African settlement growth during this period was typically via the expansion and densification of tenant occupation of land via 'shackfarming' (Hughes 1987), and increasingly during the 1980s, via land invasions (Mabin 1988; Rogerson 1989).  Governments reactions to such settlement have been varied. Under civil or military  44  dictatorships these pressures are generally ignored or result in evictions - as was the case in South Africa during this period. Stren (1992) points out that by contrast, democratically elected governments have, with limited success, attempted different approaches rangingfromattempts to legalize settlements, releasing public lands through land banks, and exchanges of public lands for illegally occupied land.  The survival and persistence of informal settlements is to a large extent a function of how they are located in relation to the political and economic structures of society. Informal settlements have not only been tolerated and permitted to exist, but in various ways, have been argued to serve the interests of the state and capital (Ward 1982). The private sector is argued to benefit in that the accommodation of workers in informal housing represents a subsidy to employers who pay workers lower subsistence wages than would be the case if workers were accommodated in more expensive, formal housing (Burgess 1982). Also, the large 'reserve army' of labour accommodated in informal settlements adds to downward pressure on wages, and provides an adequate labour supply for employers when needed. In addition, the relationship betweenfirmsand the unemployed to organized labour 'aristocracies' is often such that the informal sector provides both cheap labour and inputs into the production process, with the implication of undermining the power of organized labour. Finally, the existence of such large informal income earning opportunities allows governments to cut down on social welfare investment directed at the urban poor. In response to pressure to increase expenditure on collective consumption items such as housing, informal settlements represent an opportunity for the state to displace responsibility for the provision of these items onto the poor themselves (McGee 1979; Burgess 1982). Whilst the economic logic of these arguments is valid in the South African instance, such a functionalist logic is however seen to have become less applicable  45  in the political realm toward the end of the 1980s as settlements became increasingly volatile and challenging to the political status quo.  2.1.5 1980 -1990: rapid growth of informal settlement and political mobilization  In 1982, the Nationalist Party government announced a new strategy for low-income African housing, whereby it was to restrict its role to providing infrastructure and services, while the responsibility for the provision of housing top-structures was to lie with the private sector and individuals. Given that the private sector had historically concentrated its house building efforts on the emergent black middle class (Hendler 1987), the bottom end of the income spectrum continued to be uncatered for other than by people building their own shelter outside of existing legal frameworks - either backyard shacks in the formal townships, or in free-standing informal settlements. By 1985, while there was an estimated surplus of 37,000 houses in areas designated for whites (Lemon 1991, 15), there were an estimated 5 million urban Africans living in shacks throughout the country (UF 1991a).  Throughout the 1980s, the increased acceptance of the 'inevitability' of urbanization as a permanent part of the urban fabric was a common element of South African government rhetoric. Notably, the urbanization strategy articulated in the 1980s - and specifically the Government's 1986 White Paper on Urbanization - represented a shift awayfromrigid attempts to enforce influx control policies. However, the removal of influx control did not result in a massive increased influx of Africans to the urban areas. As noted above, that influx had been taking place notwithstanding influx control measures anyway, such that by 1980, approximately 42 per cent of Africans were urbanized (Simkins 1985).  46  In the context of continually deteriorating living conditions, and increased cost for urban services, mass urban-based protest increased in the 1980s. The immediate trigger to the escalation of violence was the implementation of the tricameral parliamentary system which sought to retain a racially divided governmental system with whites retaining ultimate control. During this period, a central role in the mobilization against apartheid was played by civic organizations - which are a local CBO variant of the Castellsian formulation of urban social movements (Castells 1977, 1983), and popularized in the South African context by contributions such as McCarthy and Smit (1984, 168-179). Most township civics were affiliated to the mass based anti-apartheid social movement known as the United Democratic Front, or UDF - which was effectively the internal arm of the banned African National Congress (ANC).  The state's response to increased levels of social upheaval and protest, centring primarily within formal urban townships in the major urban centres in the 1980s, took the form of increased repression and the declaration of successive States of Emergency. In contrast to the high levels of social mobilization and violence in the formal townships (Cock 1990), levels of political organization and activism were lower in most informal settlements. Few informal settlements were therefore targeted by the state's National Security Management System as part of the "Winning Hearts and Minds" (WHAM) strategy - a system not dissimilar to Latin American military governments' strategies for managing urban protest in the 1970s and 1980s. In South Africa, the strategy had the dual objectives of being designed to eliminate anti-apartheid political organization and leadership, whilst simultaneously addressing selected socio-economic roots of discontent by providing or upgrading basic services and infrastructure (Boraine 1989). Those settlements that were the focus of resistance to the state thus became the subject of governmentinitiated attempts to eliminate that resistance either by destroying the settlements themselves, or  47  by replacing anti-apartheid leadership with conservative and compliant leadership.  Experience in other countries has shown that no matter how well organized settlement residents are, unless that organization is linked to the de jure institutions of power — such as local government — then it becomes a "...ritual, powerless to effect action" (Peattie 1968, 67). The prospects for survival lie thus in alliances with influential politicians and patronage relationships with dominant groups, although these often lead to the co-option of leadership (Ray 1969). Ultimately though, while illegal settlers may enter into negotiations with government in an attempt to secure their status (Peattie 1990a), as soon as a settlement or organization represents a serious threat to the status quo, it is likely to come under threat in the form of elimination of leadership, or of eviction or removal (Ward and Chant 1987, 95). In South Africa, in addition to the elimination of activists and civic leadership by right-wing vigilantes (Haysom 1986), among the more notable examples of settlement elimination was the destruction of Cape Town's massive Crossroads informal settlement by government-aligned vigilantes, and the consequent displacement of 70,000 people in less than four weeks between May and June 1986 (Cole 1987).  Although generally having remained outside of the locus of political mobilization and upheaval, South African informal settlements increasingly entered the arena of political contestation and conflict toward the end of the 1980s. By the end of the 1980s, de facto political control in many settlements had shifted awayfromgovernment aligned leadership or warlords attempting to preserve the status quo, to a more militant leadership which was acting under the political banner of the United Democratic Front (UDF). Even so, organization within settlements was generally weak and parochial.  48  This is consistent with experience in other Third World countries which suggests that notions of well organized informal settlements are somewhat idealistic insofar as they are often underpinned by the "...romantic notion that the poor automatically organize themselves into communities, which often assumes a homogeneity of interests between different housing users which, more often than not, is short-lived or non-existent" (Moser 1989, 127). Also, given the limited signs of radical ideology, or propensity for revolutionary action in many shantytowns (Cornelius 1975), it has been argued that informal settlement residents' support for mainstream political parties or movements is conditioned more by the prospect of the receipt of material benefits via patronage arrangements, than due to being wedded to any ideological or political position (Perlman 1976, 1987) - thus the masses 'go with theflow'as political control of settlements changes. This is seen to be typical of the transition of many of Durban's settlements through various political regimes - about which more later.  In 1988, the state attempted to simultaneously tighten up control on squatting (particularly given the increased incidence of land invasions), and for thefirsttime to provide a legal mechanism with which to legalize some existing settlements and to establish sites-and-service schemes, via amendments to the Prevention of Illegal Squatting Act. Nonetheless, government policy favoured the implementation of site-and-service schemes, which continued to be located on the periphery of urban centres. This served to reinforce the inverted structure of the 'apartheid city' with the poor living furthestfromeconomic opportunities, and middle and upper income people (mainly whites) living closest to the city core.  2.1.6  Political and urban transition in the 1990s  The unbanning of anti-apartheid organizations and activists, the release of political  49  prisoners, and freeing up of political activity after 2 February 1990 marked a turning point in South Africa's political and economic landscape, following which a number of changes have taken place which impact directly on urbanization and informal settlements. Among the more significant of these, as Harrison (1992, 19) notes, was the 1991 White Paper on Land Reform which stressed that land rights of established informal settlements would be protected, whilst at the same time (and for thefirsttime) recognizing that the government had a responsibility toward people who were homeless as a "cumulative result of discriminatory measures and an inadequate housing policy". Furthermore, the 1991 Abolition of Racially Based Land Measures Act and the Land Acts implied that the process of residential desegregation became legal. Also, the scrapping of the Group Areas Act, the Land Act and other racially restrictive legislation, was complemented by legislation in the form of the Less Formal Township Establishment Act of 1991 which - also for thefirsttime - made legal provision for informal settlement upgrading.  Within the context of these changes, a number of impacts on the so-called 'apartheid city' can be identified. Importantly, these changes marked the entry of the ANC and other previously non-establishment groups into the policy debate which, outside of government circles, had been dominated by the private-sector funded Urban Foundation since the mid-1970s. Negotiations were set in motion directed toward the establishment of non-racial local authorities with a single tax base within which informal settlements would be represented, as well as the creation of funding mechanisms to provide for serviced-site subsidies targeted at lowest-income groups. This implied moves toward the replacement of the complex and often contradictory jurisdictional arrangements, with uniform systems of government within metropolitan areas.  Furthermore, and as indicated in Figure 2.1 below, changes to the racial structure of cities  III!  INFORMAL SETTLEMENTS  'GREYING' C B . D . A N D EDGE  HIGH DENSITY L O W - I N C O M E TOWNSHIPS  INDUSTRIAL  EX-INDIAN/COLOURED G R O U P AREAS (UPPER/MIDDLE/LOW I N C O M E ) EX-WHITE G R O U P AREAS (UPPER/MIDDLE/LOW I N C O M E ) SMALL POCKETS OF INFORMAL SETTLEMENT  Figure 2.1  The 'apartheid city' in transition  • "J  SINGLE SEX/ FAMILY MIGRANT HOSTELS MAJOR ROADS  51  have manifested themselves in a number of ways. Most visible has been the desegregation of inner-city and many previously 'white' suburbs, a process that had begun toward the end of the 1980s but which accelerated after the repeal of the Group Areas Act. This desegregation is essentially class based with access determined by wealth. In addition, there has been a growth of pockets of informal settlements within the boundaries of more affluent (often already desegregating) suburbs. Finally, informal settlements on urban peripheries have continued to grow, although more recently also on land previously designated for occupation by other race groups. According to the UF (1991c, 6), by 1990 there were over 7 million people living in urban informal housing, which was slightly over half of South Africa's African metropolitan population.  Regarding the major interest groups in the urban development arena (see Figure 2.2 below), the exclusive control exerted by the state until the 1970s came under increasing challenge from various directions following the establishment of the Urban Foundation (UF) in the mid1970s, and the growth of anti-establishment NGO service organizations and the anti-apartheid civic and political organizations mobilizing around urban development issues in the 1980s . The UF was established in the aftermath of the Soweto riots in 1976 as a non-profit organization, the mission of which was to act as change agent and development agent committed to an antiapartheid deracialization programme focused particularly on addressing the housing needs of the African urban poor. By 1990, the UF had grown from its origins as a small policy think-tank in Johannesburg, to an extensive national organization. The UF was divided between a policyoriented Johannesburg Head Office which remained focused on challenging and influencing government policy, andfiveprovincial regional branches which, by virtue of their implementation focus, had closer links to civic and political organizations in townships and informal settlements. These relationships however variedfromregion to region.  52  Particularly during the period of extreme political polarization of the 1980s, the articulation by the UF of its anti-apartheid deracialization stance embedded within a market based ideology, had the implication of radical critics considering it an 'establishment' actor trying to effect change from within 'the system'. Along these lines, Swilling (1990) for example, argued that the UF's approach failed to address the key economic roots of inequality such as the role of the market, private ownership of property, and access to and control over development capital. SYSTEM TRANSFORMATION B.C./Afrlcanlst Non-racial  Trade Unions  NACTU  SYSTEM REFORM liberal / pluralist  STATUS QUO PRESERVATION segregationist  COSATU  UWUSA  ANC  INKATHA  SACP Liberation movements  PAC/AZAPO  Liberal big business  'estobUshmenf  Community Based Organization^ (CBOs)  CIVICS  NonGovernmental Organizations (NGOs)  Service Organizations  Funding agencies (foreign)  The Urban Foundation Independent Development Trust (IDT)  Funding agencies (domestic) Western govts.  Kaglso  'White' ultra-right wing  Afrikaner big business  STATE AGENCIES  Hevos  Figure 2.2 Major actors in the urban development arena in South Africa, 1990  The role of the UF was thus in contrast to the anti-establishment service organization NGOs, whose focus was to develop closer advocacy links to civic and political organizations in low-income townships throughout the country. Given their strong links to civic organizations  53  with a history of active opposition to the South African government, these NGOs were reliant upon funding from foreign governments and agencies.  While interest groups on both left and right extremes of the political spectrum were not development actors per se, they played an important political role in townships and informal settlements. Of particular note were Inkatha, given its political significance in many townships, as well as trade unions such as affiliates to the Congress of South African Trade Unions (COSATU), many of which asserted the link between spheres of production and consumption.  With the unbannings and freeing up on political activity in February 1990, the relationships described in Figure 2.2 underwent significant shifts, with both the state and the ANC moving closer to the centre of the political spectrum, and Inkatha retreating to, and consolidating, its regional base in KwaZulu/Natal. Foreign funding directed at many NGO Service Organizations was reduced and terminated. Also, the orientation of anti-apartheid social movements and advocacy groups began a slow and uneasy shift awayfromprotest, toward reconstruction. Furthermore, the lifting of the State of Emergency marked the opening of space within which civics could consolidate their organizational structures, and for thefirsttime, coordinate regional and national level leadership within the civic movement, culminating in the formation of the South African National Civic Organization (SANCO) in March 1992. The UF began contemplating its own dismantling, with its policy units requiring reorientation, and with plans for implementation units to be set up as autonomous andfinanciallyself-reliant operating entities.  An important change in the development funding arena was the establishment of the Independent Development Trust (IDT) in 1990, after a one-off R2 billion allocation was made  54  by the governmentfroma Strategic Oil Reserve Fund. The Trust was to be set up under the direction of a previous Judge and chair of the Urban Foundation, Jan Steyn. In order to ensure political legitimacy, the Trustees who had autonomous decision-making power over allocation of funds, included no government representatives, nor politicians, but rather coopted non partisan individuals many of whom had some expertise in urban or rural development, and who were only appointed once confirmed as being acceptable to all major extra-parliamentary political interest groups.  In March 1991, the IDT announced the allocation of R750 million to the Capital Subsidy Scheme, in terms of which it would provide approximately 100,000 subsidies of R7,500 per serviced site for ownership by breadwinners earning less than R1,000 per month across the country. The Capital Subsidy Scheme, which was focused on the delivery of serviced sites to lowest income percentiles, was seen as afirststep in a process that would ultimately address the provision of shelter itself. The Capital Subsidy Scheme was to be implemented quickly, with project applications, selection, negotiations, and implementation to be completed within a five year period, thereby providing access to serviced, tenured sites to approximately 700,000 people. Furthermore, being administered as a single tranche meant that an ongoingfiscaldrain on government would be avoided.  The changing form and role of the South African state following the election of the new Government of National Unity (GNU) in 1994 signalled an important shift in its role in respect of the urban poor, and is one that runs contrary to international patterns. The role of the state in the Third World housing debate has variously been characterized as being located at points along a continuum between 'minimalist versus interventionist' (Mabogunje 1990), or 'support  55  versus provider' (Hambdi and Goethert 1989). In this context the widespread 'rolling back of the state' assumes considerable significance, given the increasing evidence that governments in most Third World countries have, in fact, for some time now been withdrawingfromformal housing provision anyway (Pugh 1989; Ramirez 1990; Potter and Salau 1990). Given the limited success of international self-help projects in reaching the poorest of the urban poor, various authors (Linn 1983; Dunkerley 1983; Mayo et al. 1986) note the trend in the 1980s toward reduced subsidies, minimizing direct production of housing, and focusing on supply-side solutions, increasing private sector supply of land and low-income housing, and increasing cost recovery.  The current trend toward 'the rolling back of the state' has been given added momentum by extensive evidence of problems associated with over-bureaucratization (Cohen 1990), corruption (Peil 1991), and breakdowns in administration of basic services and infrastructure (Stren and White 1989). In this context, Stren points out that the role of the state has shifted awayfromthe initial emphasis on housing 'provision', through an assisted self-help phase, to the present phase where the emphasis is on the 'proper management' of services and infrastructure (1990, 49). Even though pressure on governments to provide adequate shelter and services to the urban poor is increasing, many governments simply do not have the resources to adequately intervene even if they wanted to (Stren 1990; Cohen 1990).  The emphasis on 'proper management' has been given further impetus by the World Bank's current focus on the raising of urban productivity in which infrastructure deficiencies, amendment of regulatoryframeworks,transferring or decentralizing central government functions to local government, and developing thefinancialsector in Third World countries are the central  56  elements (World Bank 1991). However, this reform is being encouraged in the context of economic decline, administrative weakness and resource limitations in most developing countries. An important implication of these international economic trends is that housing is treated by governments in most developing countries as a market commodity primarily to be delivered by the private sector and to be fully paid for by users. In South Africa however, while there is tacit acceptance of a mixed low-income housing delivery system comprising informal settlement upgrading, sites and services, and public housing provision, there remains a strong political commitment to a provider role for the government via the "million (formal) houses in five years" promised by the Reconstruction and Development Programme (RDP) of the current government.  These changes at a national level provide the context for consideration of informal settlement growth at the metropolitan level. The next section thus moves down a level and traces the growth of informal settlements in the Durban metropolitan area.  2.2 INFORMAL SETTLEMENT GROWTH IN DURBAN 2.2.1 Pre-1960s: early settlement and bulldozing "the most serious threat to Durban's health and racial harmony lies in her slums and vast shack settlement, the breeding grounds of disease, crime and despair made even more dangerous by neglect" (UNP 1952, 341). Durban has the dubious distinction of being thefirsturban centre in South Africa in which patterns of settlement were organized on a racially segregated basis, and which established mechanisms which were subsequently copied and applied in other urban areas. Among the first of these mechanisms was the so-called Shepstonian system - which refers to the demarcation and entrenchment of communal and tribal land tenure rights for African people within the reserve areas designated for their exclusive occupation. Not only was the land allocated as African  57  reserves the least fertile land in rural areas, but the limited land allocations contributed further to what was already an acute land shortage for African people (Brookes and Webb 1979).  In Durban, thefirst'native' locations of Umlazi, Umvoti and Inanda were demarcated in 1847 in accordance with the locations policy of Shepstone. Given that the boundaries of these locations ranged, at their closest points, between 10 and 20 kilometresfromthe Durban city centre, they were later to be significant in influencing the pattern of informal settlement on the periphery of the expanding twentieth century Durban (ibid). Minnaar (1992, 5) notes that a further influence was the Durban Town Council's practice in the 1860s and 1870s of restricting land sales in favour of leaseholds which had the impact of discouraging the building of permanent, structures on leased land within the municipal boundaries. A consequence of this practice was that unregulated settlement occurred immediately outside municipal boundaries.  In the 1890s and early 1900s, the Durban Town Council attempted to address White ratepayers' objections to the increased influx and uncontrolled settlement of Africans via the institution of the so-called 'Durban System' which, in many respects, was the forerunner of segregated urban structuring in other major urban centres in South Africa. The 'Durban System' comprised a systematic set of policies which involved separation and registration - separation via the accommodation of African workers in hostels and barracks constructed specifically for them, and registration via the requirement that all Africans had to carry a pass reflecting employment status as per the 'togt' or daily paid labour system (Swanson 1976). These mechanisms were later incorporated in national legislation in the form of the Natives (Urban Areas) Act No.21 of 1923 and copied throughout the country.  58  The 'Durban System' was formalized by the enactment of a number of legislative measures which were to provide the basis for controlling the residence of Africans in urban areas. These measures included the Togt Labour Amendment Act of 1902 which made provision for the licensing of private compounds or barracks, The Native Locations Act no.2 of 1904 which enabled municipalities to establish locations within which Africans could be compelled to live, and The Native Beer Act of 1908 allowed the Town Council to operate a beer monopoly, the profits from which were supposed to be paid into a Native Administration Fund which was to have funded the establishment of locations (Swanson 1964). Durban did not set up locations as quickly as many of the other urban centres, preferring to rely on its 'togt' system and the licensed compounds to address the shortage of accommodation for African people within the city limits, which by 1904 numbered approximately 19,000, or about a quarter of the town's population (Swanson 1976, 161).  However, the continued influx of Africans to Durban and the inability of municipally provided compounds and barracks to accommodate the urban African population - only 30 per cent were housed in municipal quarters by 1920 - resulted in a shift in Durban Town Council policy toward acceptance of the need to construct additional housing for what was becoming a permanently urbanized African population. Due to the poor conditions within the hostels, most workers continued to live in accommodation informally provided by their employers. In an attempt to restrict the increasing numbers of Africans living in shacks, a series of bye-laws which formed the basis of Durban's regulations for the management of locations in future years were passed by the Town Council in 1916 (Minnaar 1992, 7).  These strict controls were however only applicable within the municipal boundaries, and  59  as a consequence, increasing numbers of African and some Indians settled immediately outside of the Durban borough boundaries, thereby escaping municipal control. In an attempt to extend control over these informally settled areas, the municipal boundaries were extended in 1932 to include these areas as part of the Borough, and in so doing added 92,000 to the city's population (Maylam 1983, 414)  In the 1930s and 1940s, thefirstformal townships of Lamontville and Chesterville, in addition tofivesingle sex migrant worker hostels, were constructed by the Durban municipality (TJNP 1952). These additions to the formal housing stock were however still inadequate for the expanding urban African population - an influx that was increasing due to the industrialization associated with World War II production, and the declining subsistence base of homeland areas (Maylam 1983, 416). Whereas informal housing within the Borough boundaries had, until the 1930s, taken the form of individual dwellings in the less visible backyards of private premises, a number offreestanding informal settlements emergedfrom1930 onward. The largest of these was the settlement of Cato Manor, located only 5 kilometresfromthe Durban's city centre.  According to Maasdorp and Humphries (1975, 14), the population of Cato Manor informal settlement increasedfromaround 400 in 1932, to approximately 17,000 in 1942. More significant in terms of future informal settlement growth however, were the informal settlements on the periphery of the city, which grew most rapidly in the 1940s, and became officially referred to as Durban's "black belt' (Maylam 1983,413). According to the 1951 government Census, over 45 per cent of the Durban population were classified as squatters or shack dwellers (Minnaar 1992,13). The response of Durban's municipal authorities was to label the shack settlements as the centre of illegal activities, crime, and disease.  60  The practices of some landlords contributed to the rapid growth of irregular settlement on the city periphery. The agricultural land on Durban's periphery had been acquired over many years by small-scale market gardeners and sugar-cane growers, but it became considerably more lucrative for these landowners to engage in what became known as 'shack-farming' where they rented out small parcels of land. The high rents charged often resulted in the original tenants subletting in order to be able to make rental payments, thereby intensifying the densification process (Maylam 1983,417). The landowners were able to avoid adherence to most health and municipal bye-laws given that their land was either located outside the municipal boundaries, or it was rated as 'agricultural'. Landowners engaging in shackfarming were also under no legal obligation to provide any infrastructural services such as water, roads, electricity or sewerage. Consequently, conditions in these settlements rapidly became increasingly overcrowded and unhealthy.  Tensions grew within settlements closer to the city centre not only as a result of the deteriorating physical and social conditions under which people were living, but also because of interventions such as police raids to destroy stocks of illegally brewed liquor and to check passes. In January 1949, so-called 'race riots' broke out in the Cato Manor settlement when African shack dwellers attacked the police and municipal officials, as well as municipal beer halls and Indian traders and landlords. In this violence, Webster (1977) notes that 138 people were killed, over a thousand injured, many buildings destroyed or damaged, and Cato Manor's surviving Indian residents were forced to evacuate the settlement.  The respective responses of the African population in thefirstinstance, and the municipal authorities in the second, were quite different. Many people who had lived in backyard shacks  61  or in formal accommodation in peripheral locations, moved into Cato Manor such that by the mid-1950s approximately 75,000 people were living in the area in the absence of any effective control by the authorities. The municipal authorities, who had lost effective control over the settlement, on the other had, began to make preparations for the demolition of the entire Cato Manor settlement, although at this time the legal mechanisms to expedite the relocation of the residents of Cato Manor did not yet exist. However, the election victory of the Nationalist Party in 1948, and its promulgation of the Group Areas Act of 1950, the Prevention of Illegal Squatting Act of 1951 and the Bantu Resettlement Act of 1954, provided precisely the mechanisms with which the Durban City Council was to proceed with in its major restructuring of the city.  2.2.2 Informal settlement growth between 1960 and 1985 The two primary components of the Durban City Council strategy in the early 1960s, were the systematic destruction of informal settlements within municipal boundaries in the first instance, and the construction of mass public housing on the city periphery in the second. Informal settlement destruction began with the removal of Cato Manor's population in 1958 (see Figures 2.3 and 2.4 below). Opposition to the removals flared up into violence during a police liquor raid in 1959 when nine policemen were killed during conflict with illicit liquor brewers and dealers whose operations thrived in the shack settlement (Ladlau 1975). Following this violence, the rate of removals escalated, and by 1966 virtually every shack had been cleared - a total of 82,000 peoplefromCato Manor, and another 13,000fromother informal settlements having been removed (Maasdorp and Humphries 1975). Whilst the majority of people removedfromCato Manor were relocated to the newly constructed formal township of KwaMashu twenty kilometres further awayfromthe city centre, many people chose to move to areas such as Inanda and live in shacks, thereby being able to access the benefits of nearby urban amenities and services, whilst  Figure 2.3 Chesterville township and Cato Manor before removals in 1958  Figure 2.4 Chesterville township and Cato Manor after removals  64  remaining outside of any legal system of control and regulation (Hughes 1987; Maylam 1983, 414).  In 1972, the KwaZulu homeland was formally established and took control of all African reserves in the province of Natal. In Durban, the townships of KwaMashu and Umlazi were also included within KwaZulu. In addition, a number of so-called 'Released Areas' were designated, with the intention that once non-African landowners had been relinquished of their ownership either by negotiation or expropriation, those areas would also be incorporated into KwaZulu. With national government having withdrawnfromformal housing provision, and with KwaZulu being financially unable to fund any large-scale housing programme, there was a proliferation of uncontrolled irregular settlement concentrated on land falling within the boundaries of KwaZulu.  A variant of Varley's (1989) examination of the treatment of ejidos in Mexico City as an example of governments manipulating the concept of'illegality' to suit their own ideological and political ends, is provided by the KwaZulu government's approach to informal settlements. At one level, informal settlement was not actively discouraged by the KwaZulu homeland government's refusal to implement RSA government's influx control measures within the homeland given KwaZulu's rejection of the RSA government definition of 'squatting'. In KwaZulu an African could not be squatting on someone else's land: legally, all a person needed was the permission of the local 'induna' or tribal chief (a KwaZulu appointed leader) who would allocate a piece of land and grant permission on which to build a shelter. Indunas or chiefs exerted extremely strict control over influx or occupation of tribal land under their control (Minnaar 1992, 25). Tolerance of informal settlement was to be one of the most significant factors shaping the spatial structure the Durban metropolitan area, given that most early informal  65 settlement occurred on land falling within the KwaZulu homeland, which bordered on Durban municipal boundaries.  At a second level, however, while verbally opposed to forced removals and resettlements in an attempt to protect their own credibility, homeland leaders such as KwaZulu's Chief Minister Gatsha Buthulezi remained implicated not only by virtue of their status as creations of the RSA government on whom they depended for most of theirfinancialand technical support, but also because their existence provided the government with a place to which people could be 'dumped' - all in the name of ethnically based 'nation-building' and the maintenance of supposedly 'traditional' structures (Platzky and Walker 1985, 183; Mare and Hamilton 1987). Furthermore, the proliferation of informal settlement under control of Inkatha (an ethnic populist movement revived by Buthulezi in 1975, and which drew primarily from a traditionalist Zulu support base) served KwaZulu's political interests in its power struggle with the UDF and ANC.  These contradictory approaches to the concept of'illegality' of informal settlement were particularly significant in the case of Durban given that it was the only large urban area in South Africa which had a homeland contiguous to its municipal boundaries. This further complicated the already complex mosaic of uncoordinated local authorities, chieftaincies, Released Areas, and homeland areas overlaid on authority structures established to deal with the four statutorily defined racial groups in Natal. Asidefromthe city government structures for the four racial groups within the Durban municipal area, the Durban Functional Region was characterised by four urban local authorities, andfifteenchieftaincies within the adjoining parts of KwaZulu, as well as one township committee and one advisory board for land designated for incorporation into KwaZulu (Bekker 1992, 113-114). Notably, none of these authority structures had a clearly  66 formulated development policy in respect of informal settlements within their jurisdiction.  Urbanization continued to increase throughout the 1960s and 1970s. Pressure on rural Africans to move toward the urban areas in the hope of securing employment and a better quality of life increased during this period due to a number of factors. These included increased incidence of forced removals of Africansfromso-called "black spots' in rural areas, the decreasing labour demand for cheap labour on White-owned farms due to the increased mechanization of agriculture, and the increased overcrowding and depletion of the natural resource base of the reserves. The combination of this influx with natural population growth resulted in Durban being one of the faster growing cities in the world between 1970 and 1980 when its population more than doubled,fromone to over two million (UF 1991b, 13). Most of this population growth manifested itself in the form of the construction of backyard shacks in formal townships, and the growth of sprawlingfree-standinginformal settlements on the urban periphery. Minnaar (1992, 21) notes that throughout the 1980s, the inflow of mostly rural migrants into Durban numbered about 100,000 per year, and combined with natural population growth and formal township overflow, the total number of people living in informal settlements had reached one million by 1985.  In the 1980s, the first strategically selective additions to the formal housing stock took place in the form of the site-and-services schemes at Inanda Newtown and Folweni. These initiatives are notablefirstlygiven their large scale and secondly insofar as they were the first siteand-service initiatives in Durban. The Inanda Newtown project was prompted by serious outbreaks of typhoid in the greater Inanda region, which followed a severe drought and water shortages. The water and health crises were however also used by the state as a pretext for a  67  series of forced removalsfromother settlements to the Newtown site-and-service scheme (Hughes 1987, 344; Tait 1994).  Until the mid-1980s, Durban's informal settlements were generally very socially fractured, in which welfare and political organizations found it difficult to mobilize support (Hughes 1987, 343). Since their formation in 1983, the United Democratic Front (UDF) - a coalition of antiapartheid organizations - had a negligible presence in informal settlements in Durban, focusing their mobilization efforts on townships. Even Inkatha, with its appeal to a more conservative traditionalist Zulu support base, had only secured marginally more support in the same urban informal settlements. Rather, social organization and settlement patterns were primarily defined by landlord:tenant relationships in which the practice of'shackfarming' had expanded. Landlords were able to exert strict control over new settlement, and via patronage arrangements with strongmen within settlements who were often also appointed as rent collectors, were relatively easily able to assert their authority over tenants.  2.2.3 1985 - 1990: transition from surrogate state control to adversarial control  The well documented (Meer 1985; Sutcliffe andWellings 1985; Sitas 1986; Hughes 1987) extensive violence and social upheavals of August 1985 represented a landmark in Durban's informal settlements, given the accompanying shift in social and political organization of the settlements. Until August 1985, Natal and Durban's townships and informal settlements had been relatively 'quiescent', and avoided the popular unrest and conflict that had been occurring elsewhere in South Africa. However, the assassination of a prominent humanrightsattorney sparked widespread organized protests, boycotts and marches, as well as looting and arson which was initially targeted at what appeared to be obvious symbols of'the system' such as policemen,  68  their houses, government buildings. The targets however soon changed to attacks on businesses and trading stores.  Sitas (1986, 109) notes evidence of a racial 'psychosis', aggravated by the mainstream media's coverage of events - which made much of the hostility of Africans toward Indians - and which was reminiscent of the 1949 riots. In areas such as Inanda (most of which was located within a 'Released Area designated for incorporation into KwaZulu), much of the violence was 1  targeted at Indian landowners. All of these landowners - who had been refusing to relinquish their land to the state - abandoned their properties and fled the violence. Inkatha vigilantes initially mobilized purportedly with a view to protecting businesses and property, but also used the opportunity to flush out UDF activists, and in the process, to assert effective control over many townships. Many of the vigilantes were transported infromrural areas and, following wellorganized land invasions, settled in informal settlements adjacent to the formal townships, which then also fell under the control of Inkatha warlords.  Warlordism in the region was such that informal settlement warlords in Durban functioned in terms of mutually reinforcing relationships with other warlord-types in the province: the rural/semi-rural traditional 'induna', KwaZulu or Black Local Authority urban town councillors, and 'hired-guns' (Minnaar 1992, 38-40). These warlords organized their own vigilantes to exercise control over settlements under their de facto jurisdiction, to enforce the payment of taxes or levies which were essentially rental, and which included a 'protection fee' as well as a membership fee to a political organization - in this case, Inkatha. In addition, warlords from different settlements intervened in support of each other via mobilization of vigilantes who ruthlessly eliminated challenges to Inkatha control of each others' townships and settlements  69  (Haysom 1986).  Among the reasons explaining the ease with which Inkatha was able to take control of most of Durban's settlements during the post 1985 period, was that the township and informal settlement leadership of the UDF was preoccupied with national and metropolitan level politics, rather than day-to-day local settlement level concerns, and as a consequence had not built effective grassroots organizational structures.  Inkatha warlord control over informal settlements was however short-lived. By 1989, and with the gathering momentum of anti-apartheid protest throughout the country, many Inkatha warlords in Durban were facing challenges to their control. Underpinning the fragility of warlords' control over many of Durban's informal settlements was the contradictory process of them continuing to allocate sites and thus further densifying settlements, but simultaneously being unable to facilitate the delivery of even the most rudimentary services so as to address the rapidly deteriorating living conditions in settlements under their control. In the absence of organized supportfromKwaZulu, many warlords fled their settlements or were killed. By 1990, effective control shifted to the UDF.  This shift in control took place during a National State of Emergency during which the attention of many UDF affiliates was focused not only on political representation, but on service delivery and maintenance systems, thereby placing attention on an area in which Inkatha warlords could not deliver. Also important was the fact that a high proportion of informal settlement residents were themselves second or third-generation urban dwellers, with consequent strong links to social networks in the more politicized formal townships.  70  With the demise of most Inkatha warlords which had acted as an effective political control mechanism for the state in informal settlements, and the replacement with the hostile UDF leadership, the state not only lost an alliance with the politically compliant leadership within settlements, but also lost their indirect control over settlement patterns, which until then had conformed largely to cadastral boundaries. By 1989, the gap between de jure and de facto boundaries not only widened, but increasingly cut across and contradicted each other. Irregular settlement had spread such that settlement boundaries (as defined by residents themselves) neither bore any resemblance to individual subdivisions, nor to the jurisdictional boundaries of local authorities. This disjunction applied also to the gap between the political structures within settlements and those of the rest of the city. Byerley and Mcintosh (1994) point out that informal settlements in rural areas remained subject to tribal authority, both in de facto and de jure terms. By contrast, urban informal settlements, and indeed many formal townships, were under the de facto control of civic structures reflecting varying degrees of organization and sophistication, but  which existed outside, and often in open conflict with legal or statutory structures.  Acknowledging the rapid growth of informal settlement on the urban periphery, the Durban City Council approved a resolution on 2 June 1987 not only indicating acceptance of the presence and permanence of informal settlement on the urban periphery, but also recognizing the need for their upgrading (DCC 1991,1). However, the resolution also stated that shacks and informal settlement within the municipal boundary were undesirable, and that all shacks should be identified and any new shacks destroyed and their occupants removed to alternative (as yet unidentified) land.  71  2.2.4 The metropolitan context for development in the 1990s B y 1990, the vast majority of Durban's informal settlements were clustered on the periphery of the city, extending in a broad semi-circular belt approximately 25 kilometres from  H M  L O W - I N C O M E HIGH-DENSITY AFRICAN TOWNSHIPS  O L D KWAZULU BOUNDARY  Figure 2.5  0  5  1  1  Informal settlements in the Durban Functional Region  10 J kilometres  72  the Durban CBD at their nearest point. As indicated in Figure 2.5 above, the major informal settlement concentration was the Greater Inanda area to the north of Durban's CBD which accommodated approximately 750,000 people, making it one of the largestfreestanding informal settlements on the sub-continent.  To the south, shack settlements such as Malukazi, Mgaga, and Ezimbokodweni accounted for a total of approximately 320,000 people, while to the west, there were large settlements in the Greater Marianhill and Mpumulanga areas (Tongaat-Hulett 1989; UF 1991, 14). However, the beginnings of a process of blurring the clear spatial separation between informal settlements and formally planned areas that had been strictly enforced by the authorities was evidenced by the increasing emergence of pockets of smaller informal settlement closer to the Durban city centre, and immediately adjacent to the well-located upper income - previously White - residential areas (Cross et al. 1992b).  By 1990, the majority of settlements in Durban were under the de facto control of the ANC, or groups broadly sympathetic to it. This control was howeverfragile,and asserted in the context of intense ongoing violent conflict (Byerley 1989; Minnaar 1992). Out of a total of 9,089 deaths in political violence in South Africa between 1985 and 1990, more than half of those deaths (4,671) occurred in conflict in Natal (SAIRR 1992, 486). The racial geography of Durban meant that the spatial focus for much of the violence was centred both within, and between, shack settlements on the metropolitan periphery in the first instance, the spatially and economically segregated established working class in the old townships in the second, and the emergent middle income group with the means to move into private housing estates in the third. Much of the violence was centred around conflict over scarce physical resources,  73  economic opportunities, and access to land. Morris and Hindson (1994) note that this competition for scarce resources in the context of the collapse or non-existent local government in peripheral informal settlements provided ideal conditions for mobilization to control land and other resources, thereby feeding into political mobilization and conflict which was organized around the localized power centres of Inkatha and the ANC - further fuelled by clandestine support for Inkatha by the police and military.  The political orientation of settlement leadership contradicted the political orientation of the de jure authorities in which settlements were located: the legal authority either being Durban's all-White (liberal) City Council, the Inkatha-oriented KwaZulu local authorities, or the provincial government which was an appendage of the Nationalist Party central government. Until the 1990s, settlements had seldom been the recipients of more than very limited patronage disbursements of development infrastructure orfinance.Many settlements, particularly smaller ones, and those that were poorly organized and unable to secure the support of advocacy or NGO  groups, remained subject to threat of bulldozing and removal.  The majority however, and particularly the more extensive and better organized who had been able to resist removal, 'enjoyed' a tenuous permanence.  Notwithstanding the fact that many settlements had established a degree of political autonomy as reflected in their effective control variously by comrades, 'comtsotsis'(gangsters operating under the pretence of being comrades), and internal civic structures, they remained highly integrated in the economy of the metropolitan area. Wilkins and Hofmeyer (1994, 114) indicate that in 1991, although approximately 57 per cent of adult settlement residents were formally unemployed, 33 per cent had formal sector employment, while the remainder  74  reported themselves as either engaged in part-time or informal sector employment. Nonetheless, it is likely that a substantial proportion of the 57 per cent of adults who were neither 'in the labour force, nor unemployed but seeking work', were economically active. For example, reflecting clear parallels to the case of drug operators in La Paz, Bolivia (Jiminez 1989), a settlement such as Bhambayi in Inanda is a case in point of a settlement in which the primary economic base revolves around economic activities located outside of the bounds of the law and therefore unlikely to be reported in socio-economic surveys.  In the absence of any clearly articulated informal housing policy - either at a national or metropolitan level - physical conditions within the increasingly overcrowded settlements continued to deteriorate. By 1990, there had not been any successful upgrading interventions completed in any of Durban's settlements. Consequently, at the turn of the decade, settlements typically had a very limited potable water supply system, no formally provided sanitation system, no household refuse removal, electricity, no community facilities such as schools, health facilities, and people had no legal title to the land they occupied. Hardened roads that existed were mainly bus routes for transporting workersfromperipheral settlements to places of employment. Internal road networks were non-existent.  Settlements lacked access to even the most basic communication technology such as telephones (DDA 1991). The growth and densification of informal settlements on the periphery of the city reinforced Durban's 'apartheid city' structure, with the majority of the urban poor still living at very high densities furthest awayfromplaces of employment (UND 1990; Tongaat-Hulett 1989, 41-43). Notwithstanding much policy debate, and other than sporadic allocations by government, there were no capital development or end-user finance  75  mechanisms in place to finance the upgrading of informal settlements by 1990.  2.3 INFORMAL SETTLEMENT INTEGRATION AND MARGINALIZATION  "I classify [the city] this way: the Governor's Palace is the living room. The mayor's office is the dining room and the city is the garden. And the favela is the backyard where they throw the garbage" (De Jesus 1962, 40). These comments in the diary of a faveladofromSao Paulo - The Diary of Carolina Maria de Jesus - provides an accurate insight into the relationship between informal  settlements and the urban systems of which they are part. In the South African context, at one end of the continuum there has historically been a very low level of integration, or distinct separation and difference between the statutorily regulated city and informal settlements, most particularly in terms of spatial segregation, visual difference, and political organization. At the other end of the continuum however, informal settlements have grown to be highly integrated within urban systems - most clearly in environmental and economic terms, as well as in terms of associated connecting infrastructure such transportation routes. Ultimately though, the nature of that integration is such that "...urban squatters are not marginal but integrated into the society, albeit in a manner detrimental to their own interests. They are not separatefrom,or on the margin of the system, but are tightly bound into it in a severely radically asymmetrical form" (Perlman 1987, 187).  The high degree of economic and environmental integration between settlements and their urban context is reflected in externalities that contribute to increasingly dysfunctional urban systems characterized by inefficient urban spatial structures, transportation systems, inappropriate land use arrangements, and the depletion of the natural resource base. Whilst official responses to these problems have been seen to focus on narrow 'health syndrome'  76  arguments, problems within settlements are further exacerbated because informal settlement residents often have no choice but to opt for survival strategies or immediate economic benefits at the expense of the long-term sustainability of their environment and hence of their livelihoods.  The high levels of environmental and economic integration into the urban economy, as well as kinship links into the broader urban fabric, can be contrasted with the lack of political integration. In representative 'democracies' where the strength of settlement populations acting as a group may hold the balance of power at election time, official mainstream politicians are more likely to respond, on a sporadic basis, to settlement residents' demands in exchange for votes (Ray 1969, 161-179). The disenfranchisement of Black people in South Africa prior to the 1990s however meant that informal settlement residents were even more isolatedfromthe mainstream of political activity than in other countries, thereby further diminishing their ability to secure government response to material or political demands.  Within Durban, the differences between informal settlements and the 'formal city' were therefore further exaggerated by the segregationist environment in which the city has grown. At the turn of the last decade, a vacuum existed in respect of housing policy and local government structures equipped to engage the complexities of informal settlements. Importantly, informal settlements simultaneously developed an internal dynamic of their own, as manifested in a set of regulatory institutions quite different to those in the formal city and which, over time, implied an increasing degree of autonomyfromgovernment control. It is to an examination of the growth of these institutions at the local settlement level, that the discussion in the next Chapter turns.  77  CHAPTER THREE T H E L O C A L SETTLEMENT L E V E L CONTEXT FOR UPGRADING  This Chapter takes as its starting point what is, after all, the core of the subject at hand ~ the settlements themselves and the people living in them. Following the introduction of some basic concepts, this chapter attempts to develop an understanding of the nature of the informal settlements via an examination of the history and growth of the Besters Camp settlement in Durban, South Africa.  3.1  A VIEW FROM THE INSIDE': THE INTERNAL FABRIC OF INFORMAL  SETTLEMENTS  Regarding the local-level nature of many case studies of informal settlements, an early anthropological study in Venezuela, The View from the Barrio, is prefaced with an acknowledgment that while there is a tendency on the part of anthropologists to think small, "...there is an element of the ridiculous in this procedure, like a child trying to understand a dinner party by sitting under the table. But there is something to be said for it as well. The conceptualizers and planners have to operate on a large scale...but in the end, all these generalizations and plans are summaries of what individual human beings are doing with themselves or will be doing in the future...this is an attempt to take a look at some large-scale social processes from the bottom, working out from a single case. It might be called an ethnography of urbanization and economic development" (Peattie 1968, 1-2). The settlement characteristics described in this chapter reflect some of the more common themes and characteristics emergingfroma large number of such anthropological and sociological studies of informal settlements. It is recognized however, that given their contingency on a variety of context-specific political, economic, and social factors, it is impossible to generalize the incidence of these characteristics across all informal settlements.  78  The study of social phenomena in their historical context needs to involve reference to historical documents, written records of the past, popular literature and cultural documents, but also needs to recognize that stories that tell history are always biased as they represent reality from the perspective of those articulating that history. It is recognized that representation is not simply referential, but always speaks for 'us' or for 'them' in the context of differential access to power. "Written history both reflects and creates relations of power. Its standards of inclusion and exclusion, measures of importance, and rules of evaluation are not objective criteria, but politically produced conventions"; there is no single standard by which we can identify 'true' historical knowledge. Rather, there are contests, more and less conflictual, more or less explicit, about the substance, uses, and meanings of the knowledge that we call history. This process is about the establishment and protection of hegemonic definitions of history (Scott 1989, 681).  Speaking for 'us' or 'them' (however those categories are defined) applies not only to those articulating their history, but also to the 'montage' that is constructed in any research exercise and which is reflective of the external researcher's interpreted representation and construction of that history (Tuchmann 1994). Debates about whether the perceptions as articulated, for example, by informal settlement residents are an accurate reflection of 'truth' are of less consequence than the fact that the stories reflect settlement residents' perceptions of reality, and therefore inform their norms, social institutions, and actions within that perceived reality - thus the importance, as a starting point, of examining views 'from the inside'.  3.1.1 'Community' and locality Before considering the internal fabric of informal settlements, some clarification of  79  terminology to be used here is in order. Some of the earlier studies of informal settlements were written in the 1960s during a period of intense hostility on the part of many governments toward informal settlements, and to an extent represented an attempt to counter the prevailing myths and negativity in respect of people living outside of the state's regulated frameworks.  To some, this literature has led to an overly idealized and somewhat romantic notion of the unity and cohesiveness of informal settlement 'communities', although this is a comment more in respect of citations of these earlier works (see for example Anthony 1971, 55), rather than of their specific content. The value-ladenness of the popular usage of the term 'community' is evidenced by it often being understood to imply high levels of organization, unity of purpose, coherence, and an adherence to a common sets of norms and beliefs by members of the 'community. However, given that 'community' as described in these terms is often non-existent 1  in informal settlements - or in most human settlements for that matter - some clarification of the meaning that is to be attached to the term here, is in order.  At the simplest level, community can be defined as a collection of people in a geographic area having a particular identifiable social structure, exhibiting some sense of common belonging, spirit, or consciousness. However, given the complexity of the concept, there is little consensus among sociologists of a common definition of the term. Frequently cited in this regard is Hillery's (1955) review of ninety-five definitions of the term, which concluded that beyond the recognition that people are involved in community, there is little consensus on the use of the term.  The tradition of constructing typologies can be traced back to writers such as Confucius, Plato and Aristotle. The typological tradition in more contemporary sociology is often traced back to  80  Ferdinand Tonnies, whose Gemeinschaft und Gesellschaft is often referred to as a predecessor to the works, inter alia, of Weber, Durldieirn, Becker as well as Robert Redfield's anthropological constructed type of community life in rural Mexico (1941) - all of whom developed type construction as a tool for the analysis of social systems. If, however, analysis moves beyond idealized conceptions of elements of society or 'community', and focuses simultaneously on the different elements of society, and on the nature of their integration and articulation with each other, it then becomes unnecessary to rely on "...postulating a minimal or maximal unit of organization like 'the community', nor arguing about its ontological status. We need only develop adequate and relevant tools to deal with its empirical description" (Leeds 1994, 215). To this end, the concept of locality provides some assistance.  Localities, in the context of human geographic distributions, refer to "...sensorily distinct loci  of settlement characterized by such things as more or less stable aggregates of people or inventories of houses, generally surrounded by and including relatively empty, though not unused, spaces" (Leeds 1994, 214). Thus, a city, a town, a mega-urban region, a village, as well as visually distinct sub-areas of urban areas such as informal settlements, fall under this definition. Localities comprise nodes which are points of the greatest density and widest variety of human behaviours in relation to the space in between them and the next locality.  However, while informal settlements may exhibit characteristics that make them 'sensorily distinct' places, it is important to stress the linkages between localities with their broader context. As Massey points out, links with other places are really relations of interdependence and, moreover, of an interdependence which is rarely equal. "What need to be analyzed are the aspects of domination, subordination, influence and power which these links embody. Through  81  an analysis of these one can understand the reality of a locality's 'place in the world' and a good deal of what makes it what it is, and gives it its identity" (Massey 1993, 145). Replacing the nebulous concept 'community' in this research then, is the concept locality - which retains the people, or human, dimension as central to analysis in terms of the interdependent and yet unique "nets of social relations" {ibid, 148) within localities - but situates this within an understanding of the political, economic, cultural and spatial relationships within informal settlements, and between settlements and their broader context.  3.1.2'Slums of hope' The classification of 'slums of hope' and 'slums of despair' wasfirstused in reference to 'slums' in First World countries. It was argued that 'slums of hope' had the capacity for social improvement, and could integrate themselves into the broader society of which they were part, whereas the 'slums of despair' were inhabited by socially negative and aggressive people who were a threat to each other and to the broader society (Stokes 1962). These themes repeat themselves throughout the literature on informal settlements and will be seen to have very different implications for the proposed role of governments in respect of development interventions.  The positive aspects of informal settlements were focused upon by John Turner (1965; 1968; 1976; Turner and Fichter 1972) who, together with anthropologist William Mangin (1967a; 1967b; Mangin and Turner 1969), argued in support of facilitating the development of the creative potential of the shack dwellers. This position was rooted in a conception of informal settlement residents as creative, innovative, and effective producers and controllers of large components of the housing process. It was argued that this dynamism is inhibited and frustrated  82  by governments. The positive dimension in respect of traditional neighbourhood organizations, provide confirmation of the positive angle which reinforces the notion that marginalized settlements are able to mobilize their own human and material resources with a view to physical and social improvement.  According to this perspective, the poor know their own needs far better than government officials, and given adequate support, will build housing that caters to peoples' needs far more adequately than government housing projects could ever do. The poor are, furthermore, seen to be highly resourceful, and make more efficient and effective use of the very scarce resources to which they have access than would be possible if they pursued conventional highly (and inappropriately) regulated approaches. Given the relationship between home living and work, which are conducted in the same building or at least in close proximity to each other, the builders of'popular housing' are in a more favourable position to improve their economic situation via, for example, opportunities for training and income-earning activities, than they would be in the context of conventional state-initiated mass housing estates.  These arguments are supported in a systematic engagement of conventional negative 'wisdoms' about informal settlements, in which a positive picture of informal settlements has been painted where it is argued that at the social, cultural, economic, and political levels, the residents of squatter settlements "...have the aspirations of the bourgeoisie, the perseverance of pioneers, and the values of patriots. What they do not have is the opportunity to fulfil their aspirations" (Perlman 1976,243). More specifically, socially, the squatters are generally argued to be quite well organized and cohesive. Culturally, residents of informal settlements are highly optimistic and aspire to better their lot in life, by working toward providing a better education for their  V  83  children, and improving the condition of their houses via spontaneous upgrading. Economically, they are hard working and productive; they not only build their own houses but also the overall informal physical and social infrastructure. Politically, they are neither apathetic nor radical, although keenly involved in the aspects of politics that directly affect their daily lives (Perlman 1987).  3.1.3  'Slums of despair'  "... that night I was in the favela breathing the smell of excrement mixed with the rotten earth. When I am in the city I have the impression that I am in a living room with crystal chandeliers, rugs of velvet, and satin cushions. And when I'm in the favela I have the impression that I'm a useless object, destined to be forever in a garbage dump" (De Jesus 1962, 44) In attempting to explain the perpetuation of poverty, the link between urbanization and social disorganization and change has been the focus of much attention in the literature. In terms of neoclassical economic and modernization theory, for example, the transitionfroma 'traditional' to a mature industrial society, involves a necessary accompanying process of 'cultural modernization' which implies the breaking down and replacing traditional values and social relationships with Western value patterns (Hozelitz 1960; Kuznets 1973). In contrast to the economic determinism of these arguments, alternative perspectives argue that rural migrants' movement to cities implies a necessary process of breakdown of cultural patterns and social organization (Wirth 1938; Redfield 1941). These arguments have, in turn, also been extensively challenged on grounds of inadequate consideration of the ways in which poor urban residents adapt to their changing urban circumstances without necessarily implying a process of social breakdown (Lewis 1952; Gans 1962; Mangin 1970).  A further explanation for the perpetuation of poverty in informal localities is provided by  84 Oscar Lewis's controversial 'Culture of poverty' thesis which argues that the persistence of poverty is at least partially attributable to the culture of the poor. This theory does not explain why poverty exists in thefirstinstance, but rather attempts to explain its persistence. The argument is that their culture enables the poor to survive under great duress, but at the same time perpetuates that poverty. The poor seldom maintain their jobs for very long, and the little they earn is spent on necessities, while anything left over is spent on luxuries shared with others. Given that most people become fatalistic about their marginal position in society, they seldom invest in ways that would raise future incomes. Human relationships are fragile, as are traditional units such as the family, and education opportunities are minimal. This way of life is passed on from generation to generation and becomes a self-perpetuating phenomenon, a vicious circle from which it is argued to be difficult to escape (Lewis 1970). This thesis has attracted considerable controversy, having been widely criticized, particularly by those (such as Leeds 1971; Griffin 1978) who interpreted the thesis as implying that the poor are to blame for their own poverty.  However, in an study of a poor urban locality in Jakarta, Indonesia, conducted over a fifteen year period (Jellinek 1991), considerable evidence was found to support both the elements of the 'culture of poverty' thesis as described above, and the structuralist perspective as articulated, for example, by Gilbert and Gugler (1992) and Peattie (1968,136-141), which argues that the political and economic structures in society constrain the ability of the poor to escapefromthe poverty cycle.  This latter structuralist perspective argues that the poor have the initiative, ability, and desire to improve their position, but are constrained by the economic and political structures of society. The insecurity associated with their jobs and incomes makes all other aspects of their lives  85 vulnerable. Even those who have regular jobs receive so little income that they are barely able to pay for their basic daily needs, let alone save or invest for the future. They are consequently priced out of education, homes, and amenities. Their informal enterprises are destroyed by larger firms, and their low-cost self-built shelters are destroyed in order to give way to large scale formal developments which accommodate higher income groups. The few who are able to adequately eduate their childrenfindthat they lack the necessary capital and contacts to secure well-paid jobs. The lack of an adequate and secure income reduces their mobility and reinforces their restricted view of the world, compounding their lack of awareness of alternative choices and opportunities which, even if they were aware of them, would be beyond their reach.  The poor are thus argued to become trapped by the structure of society,fromwhich few are able to escape. It is in this context that a revolving 'wheel of fortune' provides a useful metaphor in which the urban poor see their lives as sometimes being on top of, and sometimes underneath, the wheel, but ultimately: "...we come into the world with nothing and we leave with nothing" (an Indonesian kampung dweller quoted in Jellinek 1991, 181). The extent to which informal settlement residents are able to assert themselves upon the urban system of which they are part versus being trapped by political and economic structural relationships, is thus a function of the nature of both the internal fabric of those settlements as well as their relationship to their urban context. It is to the historical evolution of these relationships in the case of Besters Camp that attention is now directed.  86 3.2  B E S T E R S C A M P AS A M I C R O C O S M O F N A T I O N A L P O L I T I C A L - E C O N O M I C  CHANGE  The informal settlement now known as Besters Camp is located at the edge of the sprawling informal shack settlement of Inanda on the northern periphery of the Durban Metropolitan Area.  \  Redcllffe /  verulai  Umdlol  KWAZULU •BuffelsdraaJ-  ***** Released I Area 91  Mission Reserve 4579  Inanda Dan  Phoenix  BESTERS QAMP ?ROJECl?LOCALITY  \  KwaMasrau  Ntuzuma  ,, Waterfall  •'•..Umhlar/ga J?pcks  ,.-Released) Area / ,90  >GlenAnil':  •/  Durban/ North  INDIAN OCEAN  Newlands KwaDabeka  Everton /  Kloof  x  Clermont : : New i Ge^any  Reservoir Hills  'wesrvllle --. 1  KWAZULU BOUNDARY  N  GREATER INANDA BOUNDARY L O C A L AUTHORITY BOUNDARY  0 1  1 2 i  3 i  4  i =J  km  INANDA - RELEASED AREA 33  Figure 3.1 Locality of Besters Camp within Inanda and northern Durban Functional Region. (Source: adaptedfrom Cross et al. 1992a).  87  By 1990, Inanda accommodated approximately three-quarters of a million people, of which the Besters Camp settlement accounted for 50,000 people. As defined by settlement residents, the overall Besters Camp case study area currently comprises four sub-areas: Besters (itself comprising areas known as Ekuthuleni and Dalmeny during the pre-1990 period), Mzomusha, Nhlungwane, and Ezimangweni - each with its own unique history, social organization, and physical characteristics. While reflecting many parallels to the social and political fabric of settlements such as that described in West Kingston, Jamaica (Gayle 1983), a centrally important point in respect of Besters Camp is seen to be the extent to which, over time, it increasingly grew outside the direct control of the state and its legal regulatory frameworks.  3.2.1 Pre-1960s: Early settlement  Inanda has strong political, cultural and religious roots dating back to the late 19th century. The boundaries of the Inanda Mission Reserve were first demarcated at the time of the proclamation of magisterial districts in 1846, shortly before the first proclamation of the municipality of Durban in 1854 (UF 1982, 10). Although early settlement in the Inanda area dates back to the previous century, such early settlement was very limited in extent until the 1960s and consisted primarily of homesteads linked to the agricultural economic base of the area. Settlement patterns were rural in nature with dispersed homesteads organized in various capacities as managers or workers on the sugar cane and banana plantations, as well as in activities oriented around maize and vegetable growing.  By the 1870s all farms in Inanda were owned by individual or company speculators rather than producers. African people on this land thus experienced a rapid transformation in their status from being independent producers to becoming tenants on privately owned land. From the mid-  88 1880s onwards, African tenants began to be displaced by Indian workers who had completed their periods of indenture on sugar cane farms and had elected to stay in the area. ThesefirstIndian producers were also tenants, leasing small plots from absentee landlords, and complemented agricultural production with hawking, and keeping small stores which catered particularly for trade with Africans. In a relatively short time, and supported by favourable access to credit and land rentals and taxes, Indian producers were achieving great success and acquired increasing land holdings from speculators (Hughes 1987, 336-337).  In the early 1900s, before the 1913 Land Act put a stop to African land purchases in 'white' South Africa, a few large land holdings in Inanda became concentrated in the hands of some prominent members of Durban's Black economic and political elite. For example, Dr. J. L. Dube, (who later became thefirstnational president of the African National Congress), acquired 200 acres in 1901 for his self-help Christian Industrial School, in the area currently known as 'Dube's Farm'. Nearby, Isaiah Shembe acquired a portion of land for the headquarters of the national Zionist church, known as Ekuphakameni ('the elevated place'). In addition, close to both Dube's and Shembe's land, Mahatma Ghandi purchased land on which he established his Phoenix Settlement. Ghandi's settlement was an experiment in co-operative self-sufficient living based on his philosophy of satyagraha, meaning 'truth force', or 'call to action without violence', which was later to form a central component of the passive resistance campaigns against the South African government as well as the British colonial government in India (Phoenix Settlement Trust 1990, Swan 1985). Other important landowners included A.W.G. Champion (Natal chairperson of the ANC), and R. Ngcobo, whose family was also to become involved in senior positions in national and local politics (UF 1982, 10).  89  A legaltamingpoint influencing future land ownership patterns was the 1936 Land Act which designated Inanda a 'Released Area' intended for incorporation into the neighbouring African reserve once the state had acquired all the necessary land from non-African landowners. Hughes (1987) points out that the Act not only made future purchases of land by Indian people extremely difficult, butfromthat date onward, it was African landowners and those attached to power structures in the reserve who would purchase landfromIndian sellers struggling to maintain footholds on land in the area.  By 1960, the areas of Shembe's, and to a lesser extent, Phoenix, were characterized by limited levels of settlement, with Ekuthuleni, Nhlungwane and Ezimangweni still being unsettled. Settlement in Inanda during the pre-1960 period comprised a total of less than one thousand shacks concentrated within and around four major foci: the Inanda Seminary with its religious focus and, as indicated in Figure 3.2 below, the school at Dube's Farm, Shembe's church, and Ghandi's Phoenix settlement. Between these population concentrations, the unsettled land remained under agricultural cultivation.  3.2.2 1960-1974: Controlled settlement and the emergence of shackfarming  Settlement in Inanda generally, and in Nhlungwane and Ezimangweni specifically, increased during the post-1960 period due to a number of factors. Commencing in 1957, the formal township of KwaMashu was constructed by the Durban City Council primarily in order to accommodate the majority of the approximately 100,000 people who had been relocated from Cato Manor and other settlements between 1959 and 1966. In addition, the construction of Ntuzuma township by the Durban City Council began in 1973. Ntuzuma was located contiguous  90  Dube's  Mahatma  Ghandi's  Phoenix settlement Shembe settternent. Ezimangweni  Jeeha's  Ramdarie's  Figure 3.2  Mahatma Ghandi's Phoenix, Dube's, and Shembe settlements in I960  91  to KwaMashu and Inanda Released Area 33. This stimulated further informal settlement in the vicinity, firstly by people who did not wish to live in the formal townships but who wished to have access to amenities in those townships, and secondly by people who settled in the hope of being allocated a site in one of the townships.  In the absence of adequate administrative and rent collection mechanisms to deal with the increasing numbers of people settling adjacent to KwaMashu and Ntuzuma, people suspended rental payments to the Verulam Magistrates Court. As it became more widely known that people were living in the area without paying rent, further influxes of people into the area occurred. In addition to the large proportion of ex-Cato Manor residents, many people living in the area were rural in-migrants, but with extended family links with formal township residents.  New informal settlement during this period took place within legal jurisdictional and subdivisional boundaries, and occurred within Released Area 33. Shack dwellers in Inanda enjoyed immunity to the government's usually harsh forced removals accorded settlements in most other parts of South Africa. However, no local authority accepted responsibility for their social or physical well-being. During the pre-1975 period, many residents perceived themselves to be subject to the political authority of tribal chief Mzonjani Ngcobo of the Qadi District of KwaZulu. This chief, although having no legal jurisdiction over Released Area 33 given that it was outside of KwaZulu, appointed his 'indunas' (headmen) in Nhlungwane and Ezimangweni who sporadically collected nominalfinancial'contributions'fromshack dwellers. However, without legal authority, or any administrative orfinancialresources, no services were forthcoming from that perceived 'authority'.  92  The primary social control mechanism during this period was exercised by private landowners with freehold title, many of whom practiced shackfarming as an income-generating pursuit in portions of Nhlungwane and Ezimangweni. This did not apply in Ekuthuleni and Dalmeny where, other than homesteads accommodating farm managers or labourers working on agricultural estates, there was no new settlement.  The legal landowners in Nhlungwane and Ezimangweni who were also prominent businesspeople in the respective areas, joined the Inanda Residents' Association which had been formed in 1973. This organization was later renamed the Inanda Property Owners' Association, thereby more accurately reflecting its membership and agenda. Its stated objective was to lobby for a legally constituted local authority for Inanda which would be responsible for the planning and servicing of the area. At the same time, with the influx of people into Inanda, the prospect of selling or developing their major land holdings became an increasingly financially attractive option for landowners, and thus became the primary focus of their lobbying efforts with government.  3.2.3 1975 - 1985: Heyday of the shackfarmers By the mid-1970s, only a small number of landowners were pursuing the financially lucrative practice of shackfarming. Settlement patterns were consequently very uneven, strictly adhering to cadastral subdivisional boundaries, with densities rangingfromapproximately 40 dwellings per hectare in some subdivisions owned by landowners engaged in shackfarming, to zero dwellings in contiguous subdivisions owned by landowners who kept their land clear of settlement (see Figure 3.3 below). On unsettled land, agricultural land uses remained prevalent.  93  Phoenix settlement  Ezimangweni Phoenix township unOer construction  Figure 3.3  Dalmeny, Nhlungwane, and Ezimangweni in 1977  94  Following the formal reconstitution of Inkatha in 1975, successive Inkatha Committees were established in the various wards in the areas then referred to as Dalmeny, Nhlungwane and Ezimangweni. Although these Committees operated under the Inkatha banner, levels of mobilization and organization were negligible. During this period there was no competing organization such as civics or the ANC.  Landowners, and the associated expanding small trader  class, remained the most significant influences on settlement patterns and social organization.  Following a serious drought in Natalfrom1978 onwards, and a number of deaths due to typhoid in 1980, the period between 1980 and 1985 saw the initiation of the Inanda Newtown developments. These involved an intervention in which the Urban Foundation laid out 4019 sites with rudimentary services, and core houses ranging infloorspacefrom18 square metres to 87 square metres, but designed to enable incremental upgrading (Tait 1984). The Inanda Newtown development was largely a crisis management response to the deteriorating conditions in Inanda generally, and fears on the part of the state of a typhoid epidemic specifically. In addition, Seneque (1981) notes that it served as a pretext for a series of removals of some African tenants from Indian owned land.  As part of the national launch of the United Democratic Front (UDF), thefirstmeeting of the executive of the organization took place in the adjacent Phoenix settlement in September 1983. The emergence of UDF  affiliated civics particularly in the adjacent formal townships of  KwaMashu and Ntuzuma were significant in that they represented a challenge to Inkatha's influence in the settlements. Although the UDF affiliated civics initially targeted their political efforts on formal townships with a history of politicization (often related to resistance to forced removal) rather than the newer, less politicized, informal settlements, this emergence of a  95  competing political force to Inkatha resulted in Inkatha beginning to assert itself more aggressively in Natal generally, and in this area specifically, during the post-1985 period.  There was no further settlement in Ekuthuleni during this period. In Nhlungwane, violent conflict between Zulus and Pondos erupted in 1979 and 1980, notwithstanding their peaceful coexistence in the area for a number of years. Whilst some of the conflict was centred on access to scarce resources such as water, Haysom (1986) points out that inflammatory anti-Pondo statements by KwaZulu government leadership served to fuel this ethnic conflict. Many Pondos fled Nhlungwane and settled in the neighbouring area of Ezimangweni.  Notwithstanding the fact that there were numerous private landowners in Nhlungwane and Ezimangweni, families such as Ramdarie and Jeena were widely perceived by the resident population to be the major, if not the only, landowners in the respective areas. This perception may be partly explained by the prominence of the respective families: Ramdarie, for example, occupied a large colonial style mansion at the highest point and visible to the entire settlement, both werefinanciallysuccessful businesspeople in the settlement, and probably most importantly, they were most active in shackfarming. With legal landowners dominance as yet unchallenged, neither area had any form of civic organization, other than Inkatha committees which were based on tenuous notions of traditional leadership, and which occasionally acted as local level functionaries for landlords.  Between 1977 and 1985, the population of Inanda increased from 68,000 to 250,000 (Hughes 1987,342). During this period, aerial photographs indicate that very little settlement took place in Ekuthuleni or Dalmeny, while the number of shacks in Nhlungwane more than tripled, and  96  those in Ezimangweni increased more thanfivefold.Notwithstanding the influx, by 1985 settlement patterns remained within cadastral boundaries, with landowners maintaining strict control over site allocation. Although sub-letting could have performed a role of capital accumulation for existing shack dwellers, the availability of vacant space for new settlers meant that subletting had not yet emerged as a phenomenon in Nhlungwane or Ezimangweni during this period.  3.2.4 1985-1989: Land invasions and the dominance of Inkatha warlords  The upheavals of August 1985 noted in Chapter 2 involved at least three significant processes that were to influence the changing fabric of the settlement in the area. First, much of the violence in August 1985 was targeted at landowners. All Indian landowners consequently fled Inanda during the violence, vacating their shops, homes and properties which were plundered and colonized by former tenants. This included Ghandi's home, clinic and printing press in the Phoenix Settlement. The area was invaded and the settlement became known as Bhambayi (a variant of Bombay, India). Second, Inkatha extended its sphere of influence and was able to assert its version of 'order' in Inanda generally, and Ekuthuleni, Dalmeny, Nhlungwane and Ezimangweni specifically. Notwithstanding increasing challenges to the political status quo at a national level, by 1986 a more politically aggressive Inkatha leadership had assumed control of Ekuthuleni, Dalmeny, Nhlungwane and Ezimangweni, after deposing more passive leadership individuals and committees notwithstanding that they had also operated under the Inkatha banner. Third, a precedent was set regarding non-intervention by the state following invasions of both private and state owned land, thereby implying acceptance of de facto property rights by virtue of occupation.  97  Figure 3.4 Ekuthuleni, Nhlungwane, and Ezimangweni in 1986  98  In Nhlungwane, the Ramdarie mansion and all other landlords' buildings and homes were either destroyed or colonized. The small shopping node was plundered and looted; the only supermarket in the settlement was plundered, and then colonized by a spaza (informal) shop owner from the area. Similarly, in Ezimangweni, Jeena's shopping area was plundered, and landlordsfledthe area. Subsequent attempts were made by Indians to regain possession of their abandoned houses and land, initially via appealing to the police and army who were present but did not intervene when the attacks took place, and then later via the courts and via the state department responsible for the area. However, none of these attempts met with any success.  Analyses attempting to explain the causes of the upheavals of August 1985 have variously focused on a combination of socio-economic conditions, Inkatha:UDF conflict, landlord :tenant relationships, racial relationships, and the role of the state as agent provocateur (Meer 1985; Sutcliffe and Wellings 1985; Sitas 1986; Hughes 1987). Whether or not the state had been instrumental in initiating the violence remains a moot point, but what was clear was that the upheavals served as an effective means of dispossessing Indian landowners - thereby fulfilling the government's Group Areas objectives, and via Inkatha, of gaining ensuring of a substantial proportion of Durban's population.  One of the important social outcomes of the upheavals in Inanda was the rise to dominance of Inkatha warlords' - referred to by settlement residents as 'councillors'. Dauskardt (1991) and Minnaar (1992, 38) caricature the nature of the warlord-defined social relations as being characterized by a territorial basefromwhich they were able to tax and recruit; arbitrary ruthlessness against which there was no recourse; the blackmailing of businesses and the running of protection rackets; the extensive use of assassins to enforce their will against economic and  99  political competitors; and the feudal-style loyalty reminiscent of medieval times. Also, the medieval notion that mercenary soldiers or military retainers be paid and provided through tribute extracted from peasants either in money or kind, was similar to the system employed by some warlords. In Ekuthuleni, Nhlungwane and Ezimangweni, site sizes allocated to warlords' personal bodyguards and 'hit-men' which were up tofivetimes the extent those allocated to other settlers, still remain as a superficial testimony to such a system.  1960  1977  1985  1990  BESTERS  20  79  79  3465  NHLUNGWANE  15  220  765  1021  EZIMANGWENI  37  270  1500  2063  TOTAL SHACKS  72  569  2344  6549  Table 3.1 Number of shacks in Besters, Nhlungwane and Ezimangweni: 1960-1990  Settlement of Ekuthuleni took the form of an invasion of vacant public and privately owned land (Figures 3.5 and 3.6 below). New settlers in Ekuthuleni were required to pay Inkatha membership fees of R5, plus between R20 and R30 per site depending on its size, gradient and location, a monthly amount of R2 as a 'protection fee' (ostensibly as security against external threat), and between R5 and R10 for the 'services' of the warlord's functionaries who were sent out to demarcate new sites which paid no attention to legal subdivisional boundaries. Ekuthuleni was divided up into four wards, each controlled by a ward chairman appointed by the warlord.  By contrast, in areas such as Nhlungwane and Ezimangweni the settlement that had been  Figure 3.5  Ekuthuleni portion of Besters Camp in 1985  101  Phoenix  Newtown  Nhlungwane Ekuthuleni  KwaMashu  Figure 3.6  Ekuthuleni portion of Besters Camp in 1989 after land invasion  102  initiated under the shackfarming regime by legal landowners was inherited by warlords, whose 'contribution' to densification was essentially an infill exercise. Further complicating settlement patterns was the movement between areas such as Nhlungwane and Ezimangweni. For example, in December 1985, and coinciding with Inkatha's assertion of its increasingly bellicose Zulu nationalism, further violent conflict erupted between Zulus and Pondos, following which many of Nhlungwane's remaining Pondos fled to neighbouring Ezimangweni.  The rapidity of settlement in Ekuthuleni, as well as the densification of Nhlungwane and Ezimangweni can be explained by a number of factors. At a general level, the imperative to increase settlement levels was associated with the obviousfinancialbenefit deriving from payments being extractedfromsettlers, and which increased in proportion to the number of people settled in each ward. This was further reinforced by warlords 'power' on the ground being given quantitative expression by the number of people in wards under their control, and therefore able to be mobilized in the event of conflict with Inkatha's political opponents such as the UDF. Furthermore, during this period the primary external support base for these warlords was the KwaZulu government. The extent of KwaZulu support for each warlord was a function of the power he wielded, and thus his obvious value not only in the context of a political conflict with the UDF, but also in the broader party political 'numbers game'.  The symbiotic relationship between the Inkatha warlords and the state, in both its RS A and KwaZulu forms, were such that the warlords represented a control mechanism during a period of mass challenge to the status quo throughout the country. Within KwaZulu, the relationship between the warlords and the KwaZulu government was particularly strong in a mutually supportive sense - warlords who flew the Inkatha banner in settlements under their control, were  103  accorded leadership status by the KwaZulu government. For the more powerful such as Thomas Shabalala (the warlord controlling Lindelani settlement in Inanda) and Rogers Ngcobo (both legal landowner and shackfarmer in Congo settlement), this meant being accorded senior positions in Inkatha. For the emergent and as yet lower profile leaders such as those in Ekuthuleni, Dalmeny, Nhlungwane and Ezimangweni, support meant the clandestine provision of weapons, military training of members in camps in Northern Natal, as well as limited provision of temporary services such as water tankers being brought into the settlements, and mobile health services. The warlords thus represented a relatively cheap political control mechanism over informal settlements, involving limited resource commitment by the state.  Adjacent formal townships such as KwaMashu, Ntuzuma and Inanda Newtown experienced high levels of political activism by civic organizations affiliated to the UDF during the 1985 to 1990  period. Importantly though, informal settlements such as Ekuthuleni, Dalmeny,  Nhlungwane, and Ezimangweni remained 'outsiders' in this process of political development and maturation. Relationships between township and informal settlements were primarily associated with day-to-day issues, the most important of which was the shortage of clean drinking water in the informal settlements. Tension and conflict were common in interactions over access to these resources. For example, where water was purchased from individual households in KwaMashu, rates charged variedfroma flat rate of R2 per month to exorbitantly high prices - up to ten times the tariff rate, notwithstanding KwaMashu residents boycotting service charges and not paying municipal authorities for water themselves. Alternatively, water was taken without payment from communal standpipes in Inanda Newtown where residents were generally also not paying for water.  104  The extremely secular and parochial nature of social organization under the warlords was such that horizontal links between areas, and vertical links to organizational structures within civil society were virtually non-existent. Furthermore, warlords' absolute de facto control over land, as well as over the people occupying that land, had sidelined the influence of legal landowners in Inanda.  3.2.5 Violent transfer of control and new power struggles  Whilst the Ezimangweni, Nhlungwane, and Ekuthuleni areas were under Inkatha control - as was the whole of Inanda - the majority of the inhabitants of these areas were members of families living in formal townships in which the UDF had a significant presence. Whilst open expression of political sentiments that contradicted those of the warlords was not tolerated, the large number of informal settlement residents whose political experience included exposure to politics that was in opposition to Inkatha, became important toward the end of the 1980s when the balance of power changed significantly. In addition, as densities within informal settlements increased, implying increased pressure on the limited resource base, and as warlords' promises of resource delivery remained unfulfilled, so the degree of stability became increasingly fragile.  By 1989, there were approximately 50,000 people living in Ekuthuleni, Dalmeny, Nhlungwane and Ezimangweni. This quantitative manifestation of Inkatha control and power in the area however belied the fragility of that control. The fragility of warlords' control derivedfroma number of factors including the particular form of social organization that they had imposed, in addition to their inability to deliver resources to increasing numbers of people as promised at the time of allocating sites.  105  Furthermore, an underlying resentment was generated not only from their corrupt and often violent mode of control, but alsofromtheir visible accumulation of wealth in the context of extreme poverty. The social and political transition in the area in 1989 and 1990 was a function of these factors as well as of broader dynamics within the context of social upheavals and change nationally experienced toward the end of the 1980s. Why exactly the state abandoned the warlords when they were under threat, and thereby relinquished effective control over these vast settlements at this time, is unclear. What is clear however, is that the historic neglect by the state of informal settlements, both in terms of political intervention and service provision, opened up the space for the UDF, and subsequently the ANC, to mobilize.  In December 1989 and January 1990, the political balance of power in the Ekuthuleni, Dalmeny, Nhlungwane, and Ezimangweni settlements changed insofar as the last of the Inkatha warlords in the area were killed, had fledfromthe area, or had 'turned' political allegiance from previously dominant Inkatha, to the newly dominant 'comrades' operating under the UDF banner. The limited resources that had been introduced into the area, the most noteworthy of which were the water tankers and the mobile health services, were withdrawn after the demise of the Inkatha warlords, ostensibly on grounds of'safety' considerations.  In Ekuthuleni, the warlord was stabbed to death in December 1989 by a group of his 'lieutenants' who had 'turned' and became UDF supporters. One of the other ward chairmen was killed, and the others fledfromthe settlement. In Nhlungwane, the warlord fled the area, and moved to an inner city location in Durban. One of his ward chairmen had been shot dead earlier in 1988, while three other ward chairmen 'turned', or changed sides, and became supporters of the UDF, assuming leadership positions in the Civic organization that was set up and took control  106  of Nhlungwane. Ezimangweni's transition from an Inkatha control by four Committee members was marked by a mass public meeting at which political control shifted to the UDF following the warlord and his closest supporters having fledfromthe settlement. At that meeting a committee of twenty-four people was 'elected' with the responsibility of running the civic affairs of Ezimangweni. In all areas, remaining active Inkatha supporters either became the target of attacks or fled the settlements.  The shiftfromInkatha warlordism to control by the more politically radical 'comrades' is significant given the shift that this implied in the relationship between settlement leadership and the state. While Inkatha warlords had effectively displaced legal landownersfromde facto control of their land and settlement upon it, their relationship with the state was akin to Ward's (1989, 147-150) typology of a combination of patron-clientism, and limited cooption/ incorporation. By contrast, the 'comrades' broke even those tenuous links and asserted an autonomous and confrontational stance toward legal authority.  Particularly importantfromthe perspective of future upgrading initiatives though, was the organizational turmoil that existed in Besters Camp immediately subsequent to the demise of the warlords. The settlements were controlled by the young 'comrades' and the emergent but weak civic structures attempting to organize and assert themselves. This attempt at organization took place within the broader context of intense political competitionfromthe more experienced civic and political structuresfromthe adjacent formal townships of KwaMashu and Inanda Newtown vying for control of the informal settlements. Most notably, new sets of power struggles within informal settlements found expression at the local level between comtsotsis, women's groups, the youth, unemployed people, and the so-called 'Disciplinary Committees' which were local variant  107  of peoples courts'.  The internal self-identity of sub-areas remained extremely parochial, a consequence of which was that organized horizontal links between areas were very weak, and in many instances nonexistent. Links that did exist were kinship and familial ties to formal township residents that had been in place prior to the change in political regime anyway. Furthermore, in the context of a fractured and embryonic regional civic movement in the area, vertical links between local organizational structures in Besters Camp and broader civic and political structures remained tentative.  An important change in the functioning of the informal land market which had its roots in the allocation practices of the warlords, was that under the new regime payment was no longer made for initial site allocation or purchase, and neither were rentals paid for land occupancy. Subletting of rooms was negligible. Also, ownership transfers of shacks continued to be accompanied by monetary exchange.  3.2.6 Emergent participative democracy or social disorganization?  The 2nd of February 1990 marked the unbanning of the ANC and the lifting of the State of Emergency which implied, inter alia, that the banner flown by the power bloc dominant on the ground changedfromthe UDF to the ANC. Although most of the warlords were no longer in control of settlements by this date, 2 February 1990 remains the date referred to by most residents as the turning point away from Inkatha control, to 'democratic' control of settlements.  108  This post-1990 period represented a period of social reorganization, accompanied by the partial breakdown of institutions around which the resident population historically cohered, and the associated the breakdown of the order and discipline imposed by those institutions. In this regard, the so-called 'culture of violence' endemic in South Africa in the late 1980s (Byerley 1989; Bank 1991), the inversion ofjudicial and family authority hierarchy via the emergence of'people's courts' (Seekings 1992), the breakdown of respect for the elderly (Lobo 1982, 86), and the dominance and control of settlements by 'the youth', respectively, are examples of the emergence of such new and alternative forms of order in the settlements.  Symbolic changes made at this time included the dropping of the name "Ekuthuleni" due to it having been named by the Inkatha warlord, and its replacement with the name "Besters". "Besters Camp" - then comprising Besters, Nlungwane, and Ezimangweni (see Figure 3.7 below) - emerged as the more commonly used name of the overall area (Bester Homes was the name of a housing construction company that had, since the 1970s, been based in the construction yard at the entrance to the settlement). Throughout Besters, Nhlungwane and Ezimangweni, old Inkatha 'wards' were renamed 'areas' by the UDF supporters.  Symbolic changes aside, the transitionfromInkatha warlord control to one described in the rhetoric of the time as 'democratic', was an uneasy one. In the absence of adequate organizational structures, or reporting lines and lines of accountability, power cliques operating under the banner of the ANC or civics were, in many instances, no less guilty of violent intolerance of dissenting viewpoints, than their Inkatha predecessors had been. Each sub-area had an 'elected' area representative, who reported to an 'elected' Disciplinary Committee, colloquially referred to as the T)C'. These 'dc's' initial activities focused on rooting out people who remained sympathetic  109  Bhambayi  Dube's  Shembe  /  p  \  Ezimangweni /  ;*v  /  Newtown  / . Nhlungwane  Ntuzuma Hospital site  Phoenix  >> A Besters  /  /  N.  (  \ \ V \  ^—— Figure 3.7  /\  Besters, Nhlungwane, and Ezimangweni in 1990  KwaMashu  \  \  \  110  to Inkatha, and had not 'turned to UDF/ANC support. The administering ofjustice took place 1  entirely outside of the statutory judicial and policing system - police did not enter the settlements, and justice and 'discipline' was administered via 'peoples courts' which, as described in Seekings (1992), were also common in other townships and settlements across the country.  3.2.7 Organizational and leadership forms The euphoria in Besters Camp following the lifting of the State of Emergency and the unbanning of the ANC on February 2, 1990, distracted from fundamental underlying weaknesses that had been inheritedfromthe pre-1990 period. In the first instance, an organizational vacuum existed within settlements given not only their exclusionfromthe mainstream of urban civic politics, but also given that they had just emergedfromwarlord control in which organization beyond the tight hierarchy around warlords was non-existent. Having been denied of experience in making decisions at a settlement-wide level about the most basic issues affecting their lives had a serious impact in a situation where absolute local-level decision making power within settlements suddenly existed.  One of the major mobilizers of settlement residents internationally is the extreme threat of eviction or settlement demolition, although this was not a concern of residents in Inanda. Of greater concern in the case of Besters Camp, was the threat of attackfrompolitical adversaries which in the immediate post-1990 period, referred to attacks by Inkatha (Weekly Mail newspaper 2 November 1990,3; Daily News newspaper 3 November 1990,7). Importantly, while there was a strong sense of identity or 'collective consciousness', this consciousness often existed in relation to other settlements or external interest groups, rather than in terms of particularly strong internal organization and cohesion. In the absence or easing of such extreme threats, levels of  Ill  organization often diminished and residents operated as a disparate aggregation of individuals whose social relationship to the rest of the resident population of which they are part was defined by day-to-day basic needs. To the extent that organization existed, it was thus typically short-term and transitory, rather than continuing as a sustained longer-term process.  During this immediate post-1990 period, the emergent ANC and civics focused their attempts at organizing primarily in two areas. Afirstand obvious concern given the political history of the settlements, was to assert effective control. Whilst many settlements in Inanda were controlled by the 'comrades' by this time, Inkatha remained a powerful force in the adjacent formal township of Ntuzuma, in KwaMashu's hostels, and in the rural hinterland. A second focus was on making promises - not unlike the warlordsfiveyears earlier - to facilitate the delivery of basic services and infrastructure. Expectations of resource delivery that were initiated by warlords for the previousfiveyears, were thus fueled by their successors to power both as part of the process of undermining Inkatha warlords and the state, and then as a process of credibility building once in control. These expectations were further reinforced by the unbanning of the ANC and the lifting of the State of Emergency.  The concept of'going with the flow' introduced earlier was applicable both to leadership and ordinary settlement residents in Besters Camp. In addition to the changed 'allegiance' of the entire settlement from Inkatha to the UDF support in the space of a few months, was the pragmatism of numerous individuals occupying positions as senior 'comrades' in the Besters Camp settlements, who would go through the motions of pledging allegiance to the political adversary when visiting family in Inkatha-dominated rural areas on weekends. One notable case in Nhlungwane is an individual who survived as an active and prominent leader in all three  112  successive regimes - after 1990 he was a senior member of the Civic, notwithstanding having been one of the Inkatha 'councillors' in the settlement between 1985 and 1990, and having been a rent collector for the landowning Ramdarie family during the pre-1985 shackfarming period.  Regarding popular support for Inkatha versus the ANC, factors such as the respective ideological positions, organizational forms, and style of operation appear to have been of less importance than people's perception of each groups ability to facilitate the delivery of basic services and housing. For ordinary settlement residents who exist at the bottom of the power hierarchy (and therefore material and social well-being hierarchy), daily existence is driven by imperatives focused on meeting basic survival needs. As people are located further up the hierarchy, their activities also reflect attempts to improve material and social well-being, and they are more successful given the access to greater economic wealth, and political connections and patronage.  In a review of leadership types and activities in settlements in other Third World countries, Ward and Chant (1987) found that the motivations for taking on leadership positions, whilst sometimes including altruistic reasons, was generally weighted toward the advancement of personal prestige and/or personal political or economic gain. This is not inconsistent with the leadership type and style of operation evident in Besters Camp throughout the 1980s and early 1990s. The propensity to participate in organization was often guided by considerations of individualistic expediency and rationality, rather than by 'settlement-wide' interests. Actions and organization oriented toward furthering the 'greater social good' are thus the exception rather than the norm given that there are "...few issues which bear equally on everyone to make them matters, strictly speaking, of 'community interest'". In the La Laja barrio in Venezuela, for  113  example, "...water came closest to being such an issue." (Peattie 1968, 67).  Notably, and also consistent with experience in other Third World countries (Moser 1993a; 1993b), in areas such as water, health, and education (which are areas of settlement-wide interest) it was usually the women of Besters Camp who initiated and pursued activities oriented toward wider benefit, despite their marginalized position in the settlement. In addition to women's 'triple role' - which includes reproductive work (childbearing and rearing responsibilities), productive work (as primary or secondary income-earners), and community managing (organizing at the locality level to make demands for collective consumption items such as basic urban services) a further sexual division of labour is indicated by the markedly different activities and tasks undertaken by men and women in the settlement. In this regard, "while women have a community managing role based on the provision of items of collective consumption, men have a community leadership role, in which they organize at the formal political level generally within the framework of national politics" (Moser 1993a, 28, emphasis added).  In spite of their marginalizationfromeffective control over decision making power, women's 'community managing' role extended beyond organizing attempts to ensure the provision and maintenance of scarce resources of collective consumption such as water, health care, and education. For example, before Besters Camp was provided with basic services such as water, sanitation or refuse removal, women's triple role was compounded since it was generally they who took responsibility for cleaning and maintaining the environment.  3.2.8 External influences and interventions  Regarding the political relationship between Besters Camp and the immediate urban context,  114  the period between December 1989 and February 1990 was characterized by the emergence of embryonic civic organization within informal settlements, and the involvement of civics from adjacent formal townships (particularly KwaMashu, but also Newtown), as well as the Inanda Civic Association. The involvement of outside civics was ostensibly oriented toward the building up of emergent civics in the informal settlements, although power struggles between the KwaMashu Civic and the Inanda Civic Association - both attempting to extend spheres of control - were thinly disguised. These power plays were subsequently to become increasingly pronounced as the development process was initiated, generating tension and conflict around a process that implied an increased power base for the perceived 'deliverer' of development resources.  The relationships between the ANC and the civics in the first instance, and between these organizations and the development process in the second were not without their difficulties. In addition to the lack of political experience of people who assumed leadership positions in these informal settlements, attempts to establish organizational structures in informal settlements was not followed up by any organizational training or skills development programmes. Consequently, not only was emergent settlement leadership poorly equipped to engage their own rank-and-file constituencies, but they were even less able to counter the incursionsfromthe more experienced activists and civicsfromadjacent townships vying for control of the informal settlements.  Noteworthy at this time was the disjunction between the level of operation of leadership cliques in thefirstinstance, and their 'constituencies' in the second. Further complicating matters was an uneasy relationship between the ANC and civics. This was largely underpinned by a weak understanding that the relationship between the ANC as a political movement (or liberation  115  movement as the ANC would have it), and civics as CBOs could be most effective as a complementary rather than a competitive one, particularly until the ANC assumed control of government - at which time it would move out of the realm of civil society that it had shared with civics.  Furthermore, Besters Camp was characterized by the additional complicating factor of KwaZulu being the legal jurisdictional authority for the portion of the settlement within KwaMashu and Ntuzuma - the Inkatha spokesperson for the area was also the KwaZulu government appointed councillor for KwaMashu. Therefore, overlaying the de facto ANC control within the settlements was de jure Inkatha control of the portion of the settlement within proclaimed KwaZulu townships. This relationship represented an example of the tension between the de jure and de facto, most clearly manifested in the intense violence between Inkatha and the ANC which was the primary contributor to the Natal-KwaZulu region accounting for 91 per cent of the 1,403 deaths in political violence throughout the country in 1989 (SAIRR 1990, 235).  The policy context at that time was such that the expressed positions of the various authorities ranged from KwaZulu and the Durban City Council's passive acceptance of these informal settlements without any positive development intervention, through to Provincial and Central Government's hostility and non-acceptance of the permanence of informal settlements. In terms of constructive development intervention, the position in practice remained one of'hands off.  By the end of 1989 Besters Camp comprised over 6,500 shacks and was characterized by low levels of services and infrastructure. There were 4 water taps in Nhlungwane and Ezimangweni' serving the 50,000 people living in Besters Camp. There were no formally provided internal  116  roads, storm water control, sullage control, or electricity. Regarding social facilities, there were no formally provided facilities such as schools, creches, clinics, meeting halls or churches. Furthermore, none of the residents were in possession of legal tenure to the land they occupied. The housing stock was informally constructed, at high densities, on what was generally very steep land.  The picture of Besters Camp that has been described thus far comprises a number of layers. At the base were the masses of ordinary settlement residents whose resourcefulness is evidenced in their planning, construction, and management of the ongoing affairs of the settlements. However, notions of'community' that imply internal coherence and unity were contradicted by the settlements at the most basic level being comprised of an aggregation of individuals and households pursuing imperatives associated with their basic survival needs, while in the case of many leadership individuals, the accumulation of power and individual material self-gain ranked high as motivating forces. An exception were women's groups who generally acted autonomously from political and civic organization.  Impinging upon this base were interventions by civics and political organizationsfromadjacent townships vying for control of the settlements, resulting in competitive power struggles between organizations in the respective areas. These power plays were further complicated by the overt and violent conflict between Inkatha and the ANC, as well as the gangsterism conducted under the guise of organizational politics. All of these layers and relationships existed outside the bounds of legality or any direct control of government. In this regard, the role of the state in Besters Camp was historically non-interventionist in terms of direct involvement, while simultaneously providing tacit support to shackfarmers and warlords who imposed regimes that  117  prevented any form of civic organization emerging and organizing in the settlement, and thereby served a function of political control until the end of the 1980s. Following the upheaval and instability associated with the transition to popular control of settlements, 1990 marked the emergence of embryonic local level forms of civic organization that existed in a conflictual relationship with legal authorities.  3.3 THE DE FACTO AND THE DE JURE 'RULES OF THE GAME' Having constructed a picture of the nature of informal settlements both with reference to international literature and informal settlements in Besters Camp, Durban, and South Africa, it is possible to synthesize some of the key themes emerging from the discussion thus far.  The rules of the game in informal settlements and the rules of the game in the formal city overlap in some instances, while in others, they are fundamentally different. The de facto rules in the informal, while the de jure rules in the formal city. Within informal settlements, the rules of the game are defined primarily by settlement residents themselves, according to the circumstances of practice - or, as described by Perdomo et al. (1982), according to "spontaneous law" - and which have no status in statutory written law. These rules are informed by a 'reality' that is defined by peoples perceptions, rather than any notions of'scientific facts and rationality' that are dominant in the formal city.  This implies a fundamental logic quite different to that which informs the rules that are determined by the state or regulated markets, and which are accorded status in statutory law, as is the case in the formal city. The regulatory systems (the de facto 'rules of the game') in informal settlements are primarily informed by nature of the social fabric of informal settlements in the first  118  instance, and the relationship of those settlements to the remainder of the urban system in the second.  The usefulness of de facto as an organizing concept is that it makes the distinction between legality according to written law and according to the circumstances of practice. In this regard, all settlements, whether or not they are formally encompassed by the purview of the state, must follow some form of'law' which is understood and enforceablefromthe perspectives of all parties to a relationship (Leaf 1994). Given that a de facto instance is defined in terms of the simple 'fact' of its existence over time, criteria such as peoples' perceptions of reality are central to establishing that existence. For instance, an extensive range of (continually changing) political organization, leadership, and power cliques exist in the lives of informal settlements residents to the extent that they shape and control the social and physical make-up of settlements. This de facto political organization however does not 'officially' exist in the eyes of de jure authorities. For example, the Master and Structure Plans of many urban areas in developing countries only indicate formally planned development that has taken place within the parameters defined by statutory planning regulations. Informal settlements are not indicated and 'do not exist' according to those official plans.  Being located furthestfromthe locus of power and thereforefromdecision-making over urban resource allocation, informal settlement residents' access to information about such issues is limited, severely constraining their ability to interact and effectively engage interest groups beyond the local settlement level. Within informal settlements, residents' activities are primarily short-term in focus, particularly insofar as day-to-day basic needs are a priority. This conception of the immediate is however mediated by the value of time being defined by social value, rather  119  than economic value as in the formal city, an implication of which is the pace at which activities are conducted in the respective environments being vastly different. Implicit in the ability of informal settlements to rapidly change the rules of the game in relation to a changing urban context, is a significant degree of adaptability and flexibility. Furthermore, the fluid nature of social organization and leadership is an important characteristic of this constantly changing social fabric which functions according to a logic that is considerably more flexible than that which applies to the rigid, relatively permanent world of City Halls and planning statutes. INFORMAL S E T T L E M E N T S  ' F O R M A L ' CITY  R U L E S OF T H E G A M E  De facto rules of the game  De jure rules of the game  R U L E S DEFINED B Y  Settlement residents at local level  State at level of nation-state  S T A T U S OF L E G A L I T Y  'Spontaneous law'  Statutory law  ' R E A L I T Y ' DEFINED B Y  Perceptions  (Ostensibly) by 'facts' and 'scientific rationality'  POWER  Locus of own local level power system, furthest from state power  Locus in state  W O R L D VIEW A C C E S S T O INFORMATION  Parochial  Local through to global  Confined primarily to settlement locality  Well connected to other  limited access to city-level information  cities internationally  COMMUNICATION M O D E  Face-to-face  Hi-tech information technology  V A L U E OF TIME DEFINED BY  Social value  Economic (time value of money)  IMPERATIVE INFORMING ACTIONS  Day-to-day basic needs  Day-to-day criteria ostensibly guided by longterm strategic objectives  ORGANIZATIONAL STRUCTURES  Fluid, flexible, issue-based, short-term  Rigid, statutory, longterm  LEADERSHIP INDIVIDUALS  Patronage dispensers/facilitators activists  Appointed officials/ politicians  POLITICAL ORIENTATION  Overt patronage  Covert patronage  INFORMED B Y  Table 3.2 Overlaps and differences in the de jure and de facto 'rules of the game'  120  At the bottom line, there is a divergence between reality defined in terms of the law, and reality defined in terms of peoples perceptions (which includes a vast range of'realities' which can be represented along a continuum). Given that within informal settlements people conduct themselves in terms of their perceived reality, the de facto is of fundamental importance. In these terms, an instance being examined can be represented at points along a multi-layered continuum (some of the extremes of which are indicated in Table 3.2), rather than as a simple dichotomy.  Both formal and informal are therefore regulated: the formal city regulated by statutory governmentalframeworks,and informal settlement by de facto rules of the game. The continual interaction between informal settlements and urban systems is however characterized by an asymmetrical integration, with the formal being dominant over society as a whole. This dominancefindsexpression in political and economic terms (most obviously in terms of decision making over resource allocation and distribution in favour of the formal), in ideological terms (the informal as the 'other of society), and in spatial terms (the formally planned and regulated city as 1  opposed to informal settlements).  As is seen in the next chapter, there is a complex tension given the dominant de jure's attempts to extend control via formalizing (ie. regulating in its own terms) the informal, while the autonomy of the informal increases as it expands in absolute and relative terms. Simultaneously, a loss of connection (and power) is experienced by the informal as rules, norms, and institutions increasingly diverge between the formal city and informal settlements, and as informal settlements access to, and control over, urban resources consequently decreases over time.  121  In highlighting problems with conventional approaches to planning in informal settlements, the perspective to be presented in Chapter 4 is that planning in Third World cities - and the theory which informs it - needs to engage not only the specific de facto institutions of informal settlements, but also the tension between the de jure and the de facto 'rules of the game' as they apply in the respective, but integrated, worlds. It is to an examination of the insertion of the planning of upgrading interventions into this articulation between informal settlements and the formal city, that attention in the next chapter is directed.  122  CHAPTER FOUR T H E PLANNING OF INFORMAL SETTLEMENT IN-SITU UPGRADING  Having examined the nature of the environment in which planning is to be conducted, it now becomes possible to consider planning interventions themselves. This chapter examines the policy and theoretical underpinnings of common planning approaches to informal settlement upgrading, as well as some of the problems that are associated with these planning approaches. An alternative approach to planning informal settlement upgrading is then considered with reference to some of the key elements that need to comprise a framework to guide informal settlement planning. Since there was no South African precedent for large scale in-situ upgrading at the time of the initiation of the Besters Camp upgrade, this Chapter draws primarily from international experience.  4.1 PARADIGM SHUTS IN POLICY AND PRACTICE 4.1.1  Settlement bulldozing and public housing construction  The conventional 'wisdoms' about informal settlements, more dominant in the past, but still prevalent today, are that they are "septic fringes" and "cancerous growths" which the authorities would do well to completely eliminate. These perceptions which are described, for example, in Fanon (1963) and Dwyer (1974, 1975) portray informal settlements as being inhabited by people unable to adapt to the urban setting, associated with crime and prostitution, social disorganization, maladaptive rural behaviours, and being parasites on the economy, draining it of its vital resources and contributing nothing in return. In addition, they are presumed to pose a threat of violence and radicalism as their frustration at the wealth around them has the potential to turn to violence against the political system.  123  Rooted in earlier arguments that settlements comprise "human flotsam and jetsam", who inhabit the '"...'tin can cities' that infest metropolitan centres of every developing country from Cairo to Manila" (Lerner 1967, 24), are current perspectives that are still reflected in the approaches of some governments, for which the bulldozing of settlements remains an integral part of policy and practice.  In the post-war period, initial responses to informal settlements were thus oriented toward the prevention of their growth, most commonly via their bulldozing by the authorities and the attempted relocation of shack dwellers to their places of rural origin. However, given the limited success of most governments at restricting the influx of people into urban areas, and as many governments in developing countries acknowledged the inevitability of the urbanization process, there was a concomitant increase in pressure on the authorities to make provision for rapidly expanding urban populations. While not all developing countries have passed through all the stages described below, there are nonetheless a number of general trends that can be identified (following Permian 1987; Palmer and Patton 1988; Hardoy and Satterthwaite 1993) which, at different times, and to a greater or lesser extent, have been part of or have informed urban housing policy in Third World cities.  The move beyond attempts to eliminate informal settlements without providing any urban accommodation alternatives, was most clearly reflected in the implementation of numerous mass public housing projects in the 1950s and 1960s - albeit still linked to the bulldozing of'illegal' informal settlements. Literature focusing on public housing has noted both the successful implementation of a limited number of public housing programmes such as in Singapore (Goh 1988; Lee et al. 1993), but also on the criticism that public housing seldom accommodates the  124  lower-income people for whom it was originally intended. Gilbert and Gugler (1992, 137-139) point out that due to public housing generally being unaffordable to lower-income people, they are generally displaced from locations close to their place of employment, to more remote locations.  4.1.2 The Turnerist concept of 'self-help' The concept of 'self-help' as articulated by John Turner was reflective of an evolving conceptualization of the housing problem where informal settlements became seen more as part of the solution than as the problem. Turner's argument (1976, 1982, 1983, Turner and Fichter 1972) was that housing should be viewed as a process rather than in product terms, or " a verb, rather than as a noun", and that policy approaches should emphasize user autonomy, smallscale interventions using appropriate technology, and user control of decision-making in housing production. According to this perspective, "...when dwellers control the major decisions and arefreeto make their own contributions in the design, construction, or management of their housing, both this process and the environment produced stimulate individual and social well-being. When people have no control over nor responsibility for key decisions in the housing process, on the other hand, the dwelling environment may instead become a barrier to personal fulfillment and a burden on the economy" (Turner and Fichter 1972, 241). Regarding the role of the state, public housing policies and sites-and-services schemes which seek to improve the urban environment by removing and rehousing squatters are argued to have the effect of disrupting social networks and destroying real material and social resources which could otherwise be harnessed to improve the physical and social environment in which the urban poor live. From this perspective, the appropriate role of government is not to withdraw from the housing arena entirely, but to play a supportive, enabling and a facilitating role in respect of user-autonomy, decision-making, and 'freedom to build'.  125  Responses to the Turnerist conception of self-help have been addressed at two levels. First, at the level of theoretical debate, and second, at the level of policy and practice. Firstly, then, the Turner-Burgess debate - or as seen by others, the Turner-Burgess non-debate in which each argument is seen as making valid points, but no real debate took place because of a lack of agreement on mutually recognized terms of reference (Nientied and van der Linden 1985) - is taken as a useful starting point for an examination of the theoretical debates around self-help housing. Critics of the 'Turnerist' position argued that it pays insufficient attention to the broader economic context, as well as to the political dimensions of self-help housing within the context of capitalist social relations (Burgess 1978, 1982).  The issue of commodification was a central element in the debate. Whereas Turner focused on the use value of informal housing (rather than exchange value), Burgess's argument focused on the way in which the various aspects of the housing process were given value (valorized) as part of the process of capital accumulation. Without entering into the details of the debate (summarized in Mathey (1992, 380-385), Ward and Macoloo (1992, 61-62)), the structuralist view argued that given that the role of the state was to support the conditions necessary for capital accumulation, it would be unlikely to take any action to prejudice those interests. Given the nature of the commodification process within the broader process of capital accumulation, the appropriate role of the state in housing provision is of less interest for neomarxists than the argument that it is the root structural causes (i.e. capitalist social relations) that need to be transformed in order to address the housing issue which is but a symptom of the problem. This is however not particularly useful in providing insights as to how the immediate problems of informal housing are to be engaged. To the extent that the immediate is considered, increased state intervention via direct help in completing the 'use value' of housing has been  126  proposed (Ramirez et al. 1992).  The 'emancipatory' tone of the Turaerist position — 'freedom to build', 'self-help', 'dweller control' — has also been argued to be open to manipulation by governments using 'selfhelp' language as a justification for conservative deregulation and privatization policies, in which the poor are to be left to do more for themselves, with minimal state intervention. Another dimension of this argument is that self-help, when practiced as an individual rather than a collective effort, also serves to reinforce the status quo (Fiori and Ramirez 1992; Marcuse 1992).  4.1.3 Sites-and-services, and in-situ upgrading  The second level of responses to Turners contributions were located at the level of policy and practice, and were most clearly seen in the sites-and-services approach. This approach is also commonly traced back to the 'sites-and-utilities'concept introduced by Charles Abrams (1964) in which he proposed the settlement of the poor on serviced sites, as an alternative to squatting, which he argued would otherwise occur. However, sites-and-services schemes were not conceived of in terms of a devolvement of decision-making to residents as suggested by Turner, but rather as an application of an incremental process as a means of reducing construction costs.  Notwithstanding the popularity of sites-and-services schemes with the World Bank and other donor agencies, particularly throughout the 1970s and 1980s, these schemes have increasingly come under critical scrutiny. Among the primary criticisms that have been raised (Laquian 1977, 1983b; Peattie 1982; Bamberger et al. 1982; Swan et al. 1983; van der Linden 1986) include evidence that they are generally situated in peripheral locations with limited access to the urban economy, have high development costs, are often unaffordable to the lowest income  127  groups below the 20th percentile, have highly standardized building and materials requirements, they impose limitations on income-generating opportunities for households, and often result in increased social segregation.  In parallel with sites-and-services schemes, the increased adoption of in-situ upgrading as a strategy for engaging informal settlements was reflective of a more accommodationist approach in terms of which informal settlements became accepted as permanent elements of the urban fabric in developing countries. One of the major positives of in-situ upgrading is that such interventions are able to reach below the 20th income percentile. In addition, social networks within settlements are retained, few people are displaced, and the potential for participation of settlement residents in the development decision-making and implementation process is high.  Positive impacts of upgrading on income and expenditure patterns have generally been limited (Raj and Nientied 1990; Viloria and Williams 1987). Impacts of upgrading on gentrification have also been varied: in some cases there has been limited displacement of poorer groups in favour of more affluent groups (Robben 1987; Nientied and van der Linden 1985; Taylor 1987); whereas in other instances (and linked to steeper rent increases in upgraded areas), higher levels of displacement of the poor has taken place (Keare and Parris 1982; Moitra and Samajdar 1987). Although high degrees of public participation have been noted in many upgrading projects, this has often been used as a means to the end of service delivery, and has not always facilitated 'empowerment' objectives of self-reliance among ordinary informal settlement residents (Moser 1983; Rakodi 1989; Skinner etal. 1987).  Cost recovery remains problematic both in terms of a willingness and ability of settlement  128  residents to contribute to the capital costs of infrastructure as well as to the running costs of water, sanitation, and refuse removal systems (Rakodi 1992). In many cases this creates a dependence on subsidies (Agbola 1990). Linked to this problem, and in terms of creating the capacity to sustain the momentum for ongoing upgrading and development within settlements, upgrading projects have generally been limited to the confines of project boundaries and have not had a city-wide impact in terms of policy or institutional reform (Payne 1984; Rakodi 1991; World Bank 1991). The Kampung Improvement Program in Indonesia (Silas 1984, 1992; Taylor 1987; Baross 1989) represents an exception to this general experience.  4.1.4 Legalization and formalization  With many governments in developing countries having accepted upgrading as a component of housing policy, a diversity of approaches to upgrading have emerged, the most common of which include attempts to partly control the 'illegal' systems and curb their negative aspects, through to attempts to formalize them by incorporating them into statutory systems of land management.  Currently, the dominant planning orientation is toward informal settlements being endowed with a status implying legitimacy in terms of statutory law - whether by non-  interference, or by explicit legitimization such as via the delivery of legal tenure. In practice, the mechanisms via which this legalization takes place are primarily through the application of rigid planning tools which may include, inter alia, the delivery of legal tenure, political structures of government and administration, legal arrangements and Development Agreements in the form of legal contracts - all of which, as noted earlier, have their origins in, and were designed for the formal city. Put differently, the dominant orientation in planning practice in informal settlements  129  is to attempt to 'formalize the informal'.  Much of the focus on legalization was rooted in the argument that the provision of legal tenure is one of the most important components of upgrading on the grounds that it significantly increases the likelihood of housing investment and consolidation (UNCHS 1982). This thinking is reflected, for example, in the planning and tenure delivery component of major projects such as the Tondo and Lusaka upgrades.  One of the criticisms of current approaches to upgrading has been argued to be that the preoccupation with legality and illegality has given rise to a vogue of legalization programmes that have distractedfromwhat is after all the issue of primary importance: finding ways of improving the quality of life of people living in informal settlements (Varley 1989; Farvacque and McAuslan 1992).  Although consolidation and upgrading have been linked to increases in squatter incomes and security of tenure, land tenure legalization should not be seen as the sine qua non of housing improvements and settlement consolidation, but should be assessed in context with other policy instruments such as the installation of infrastructure and levy of charges (Angel 1983; Varley 1987). Experience on upgrades in the 1970s and 1980s indicate that it is de facto security, rather than legal title per se, that is most important in housing investment decisions (Angel et al. 1983; Skinner 1987). Furthermore, on the basis of experience in Mexico city, Varley (1987) argues that even without legal tenure delivery, changes in residents perception of security can also be produced by means such as the installation of services which people take as an indication of official acceptance of the continued existence of their settlement.  130  4.2 SOME UNDERPINNINGS OF THE DOMINANT PLANNING ORIENTATION IN INFORMAL SETTLEMENTS  Elaborating further on the notion of formalizing the informal, there are a number of immediately obvious reasons why this remains common practice. At a superficial level, there are factors such as the simple convenience of urban planners using procedural planning tools such as Master and Structure planning with which they are most familiar, rather than having to enter the somewhat messier interface between formal and informal which is less obedient to the dictates of planning statutes. Additional factors centre on the nature of the policy formulation process which is invariably such that policy is formulated by actors who are quite detachedfromthe onthe-ground realities of informal settlements, and whose political and economic interests may be served by policy geared to serve the formal city, rather than the informal. There are however some more fundamental reasons which explain the mindset which underpins formalization approaches.  4.2.1 An ideological and political dimension: the informal 'other'  If planning theory is about explaining, predicting and providingframeworksfor future action, the value laden question arises: 'action' with a view to achieving what, and in whose interests, and why? As pointed out by Corbridge (1992, 586), "it's not just a question of what development is and how it might be measured and promoted. It's more a question of who says so and in what terms and for whom?"  It is noteworthy then, that most of the planning theory applied in Third World cities is typically Western and First World in focus and origin, often with a colonial legacy (Mabogunje  131  1990). In this regard, much has been written on the phenomenon described as ethnocentrism which, in broad terms, refers to prejudicial attitudes between in-groups and out-groups by which 'our' attitudes, customs and behaviour are unquestionably and uncritically treated as superior to 'their' social arrangements. In the context of Third World development, this 'us' and 'them' mind  set finds expression in theory which McGee (1991) describes as theory grounded in the assumption that the theoretical tools that were developedfromthe historical experience and values of the Western world, can be generalized into an account of development applicable to other societies or on a global scale. As argued by authors such as Berger (1994), one dimension of this is that western values and theoretical tools become hegemonic, and non-western societies are encouraged to mimic them, perhaps with a little 'help', toward this goal of 'modernization' already achieved by the 'mature, industrialized, democratic market economies of the First World'.  One dimension of this argument relates to discourses, in respect to which it has also been argued that all development discourses establish an 'other' which is seen as lagging behind the dominant power on the evolutionary path towards development. In this regard, writers such as Escobar (1992,413-417) argue that the discourse of'Development' has been part of the mapping of the Third World, and the production of its specific economic and social configurations commonly referred to as 'underdevelopment' - all of which has required the underdeveloped Third World to follow the prescriptions laid down for them by those already 'developed'.  Picking up on these two themes, and of primary interest here, informal settlements also exist as an 'other' of society - different, and lagging behind the formal. As argued earlier, the 'informal' encompasses the sets of physical, economic, and social relations that exist outside of governmental regulatoryframeworks.The informal operates outside of aframeworkwhere  132  legality has anyfixedmeaning and is "the other of the grand society. This is why the informal has incontestably to do with legality, but a legality understood in a wider sense than the purely juridical" (Latouche 1993,131). The informal exists as the 'other' not only at an ideological level in terms of which an ethnocentric logic has relegated it to a subordinate, inferior status and as representing some transitory stage that precedes the 'achievement' of formality, but also in terms of which it represents a politically sanctioned form of social marginalization.  4.2.2 Rational comprehensive planning theory and Master Planning  Academic planning literature has provided a forum for the articulation of arichdiversity of approaches to planning theory. Of particular interest here is the Synoptic, or Rational Comprehensive approach which has historically been the dominant tradition in Western planning thought and, to varying degrees, informs most mainstream planning approaches which either build upon, or attempt to react against it. As articulated by authors such as Meyerson and Banfield (1955) and Herbert Simon (1965), this approach views planning as a goal-oriented and rational approach to the guidance of future action, according to a set of clearly defined steps and technical rules. Although the rational-comprehensive model has undergone numerous reformulations, including increased attention to social considerations, as well as contributionsfromapproaches such as systems analysis, it remains the subject of heavy criticism in planning literature.  Among the major criticisms are that it fails to adequately contextualize the planning process in the political and economic context within which planning is conducted, and presumes a general public interest rather than pluralist or conflicting interests in society. Furthermore, it provides a rationale for a priori goal setting and an autocratic top-down decision making style on the grounds of efficiency, whilst failing to recognize the cognitive limits of decision makers  133  (Goodman 1972; Faludi 1973; Lindblom 1973; Hudson 1979; Weaver et al. 1985). In addition, the approach remains within the positivist framework of the natural scientific method, and therefore remains vulnerable to criticisms that the methods of the natural sciences are inappropriate and inapplicable to the social sciences (Denzin and Lincoln 1994).  Notwithstanding these criticisms, the rational comprehensive approach remains the dominant theoretical underpinning of urban planning, both in Western, and in Third World cities. At the level of planning practice, one of the major contributions of the rational comprehensive approach - the Master Plan - remains a favoured planning tool of urban planners in many developing countries (Devas 1993). This remains the case notwithstanding the numerous problems with Master Planning, in respect of which among the more commonly cited problems include the following: • the concern with the preparation of a plan document rather than a focus on achieving an impact on the ground; • the attempt to be too comprehensive rather than focusing on key issues; • the detailed nature of rationalistic planning is a time-consuming, costly activity that frequently entails long delays in translating policy into action; • the preference for quantitative methods of feasibility analysis, appraisal and evaluation, leads to a bias toward large scale, high-technology capital construction projects most easily expressed in spatial and land-use terms, and a neglect of social, economic, and environmental dimensions of development; • the detail of plans do not ensure successful implementation given not only the unrealistic projections of public investment requirements, but also the limited administrative capacity in developing countries to implement such plans;  134  • the separation of plan-makingfromdecision-making about budgets, infrastructure development and service provision; • the absence of effective mechanisms for controlling land development whether because of legislative weakness, bureaucratic failures, corruption, or the absence of an adequate mechanism of relating the plan to the development control system; • the rigidity of detailed plans which are not flexible enough to be able to accommodate to the fluidity of the economic, social and political forces which shape Third World cities (Devas 1993, 72-3; Rondinelli 1993, 90-117).  Overlaid upon these general problems is an additional factor which emerges in the planning of informal settlements. Here, Master Planning is the mechanism via which informal settlement planning is conducted according to 'formal city' imperatives: the politics prevailing at City Hall, the economics of formalfinancialmarkets, the tenure of formally regulated land markets, the developmentfinanceof commercial lenders and of local government, and the planning approval steps stipulated by local government Planning and Engineering Departments. Importantly, the planning policy, principles, substantive content, and the procedures underpinning such a planning approach, are designed for, and by, planners in the formal city - often with a limited understanding, and with even less integration of the complexities of informal settlements in the various aspects of that process.  4.2.3 Links between knowledge and action A third underpinning of the dominant current approach to planning in informal settlements relates to the relationship between knowledge and action. One of the fundamental reasons for the ineffectiveness of the volumes of theory and policy prescriptions ostensibly generated with  135  a view to reversing the decline of increasing numbers into urban poverty, is that their formulation takes place at levels that are detached from the practical on-the-ground realities of the subject matter of the theory in question. In this regard, the "absence of strong links between understanding and action" (Edwards 1989, 11.7) has been noted by numerous writers in which  there is an observed tendency that research "scholars rarely appear to talk to (or write for) those practitioners who are engaged in planning and housing policy" (Ward and Macoloo 1992, 60), and even less to those who are after all the primary focus of those writings in thefirstinstance the urban poor.  A key issue here is that given their different vantage points and agendas, a planning theorist's 'theory' and a planning practitioners 'theory' are likely to be quite different - both in descriptive and normative respects - even when articulating those theories in respect of the same phenomenon and from the same ideological viewpoint. Thus the appeal for a "higher degree of interaction between active researchers in the urban researcher community and the national and international project managers who are dealing directly with 'urban management' projects" (Stren 1993,138). This interaction should however not be unidirectional - but rather an iterative process as suggested by Friedman's metaphorical description of the social learning approach to planning which he describes as beginning and ending with action (1987, 181). This need for planning theory and policy that is more strongly grounded in practice will be returned to later.  A second point in respect of the knowledge-action relationship is the importance of power as a mediating influence in that relationship. At the time of the widespread and increasing popularity of the so-called Alternative Development paradigm, Roweis (1981) made the point which is no less valid today - that planners working within the establishment whose objective is  136  system change are faced with the reality that whereas planning theory enthusiastically engages the urban 'ills' resultingfromparticular urban development trajectories, as well as what 'ought' to be done to rectify these ills, planning practice is such that if planners are to be effective at all, they are forced to become skilled at the 'art of the possible'. The implications of the power relations within which planners operate remain such that "as the gulf widens between plarining theory and practice, the choices open to planners begin to crystallize. Either they must stick to a 'tough line' of critique of given urban realities and of advocacy of a futuristic and more socially rational urban arrangements and forego any serious hopes for practical effectiveness; or, they must develop a 'softer' pragmatic, technicist and less theoretical line, more in tune with what is feasible in practice, and forego any serious hopes of producing significant structural impacts on urban development and growth." (Roweis 1981, 172). These tradeoffs apply as much to non-establishment planners such as those working for NGO's and CBO's, as they do to government planners when attempting to pursue a planning approach that runs counter to the mainstream. Therefore, while theoretical positions located within radical theory and the social mobilization tradition which are oriented toward major social transformation have usefulness at the level of the 'tough line' of critique of the status quo, power relations are invariably such that the limited likelihood of translating them into practically effective action abbreviates their usefulness in guiding policy and practice. Related to this, and as Sanyal (1994) points out, it was inter alia a failure to accurately assess and respond to power relations in a politically pragmatic manner that has contributed to the weakness of the so-called Alternative Development paradigm. Beyond the level of critique then, the importance of an understanding of power relations lies in informing planning strategies that are practicable and are designed not only to facilitate the translation of knowledge into action, but also the production of knowledge from action - and thereby to provide a more realistic prospect of moving beyond formalizing the informal.  137  Having noted some of the underpinnings of current approaches to planning in informal settlements, the next section will consider in some more detail the implications of planning approaches that have the formalization of the informal 'other' as their objective.  4.3 SO WHY NOT FORMALIZE T H E INFORMAL?  So why, then, should planners not continue to pursue conventional Master Planning in informal settlements, particularly given that the aspirations of informal settlement residents themselves are often expressed in terms of calls for achievement of one or another aspect of material or social life that is characteristic of the formal city, but absentfromthe informal? Simply put, such demands of settlement residents for delivery of one or another aspect of the features of the formal city reveal less an aspiration for formality and de jure regulation, than firstly a demand for improvement in material conditions to a standard that is at least equivalent to that enjoyed by those living in the formal city, and secondly a demand for increased access to power specifically in respect of access to and control over urban resources - the control of which not incidentally also resides in the formal city. Numerous sociological, anthropological, and economic studies of informal settlements following the oft-cited work of John Turner and William Mangin have demonstrated quite clearly that most informal settlement residents would prefer to retain their geographic place of residence, their social and economic networks, while improving their material quality of life via access to security of land tenure, improved quality of shelter, infrastructure and services.  4.3.1 Formal and informal planning processes, and different rules of the game  It will be useful to begin by distinguishing between the contexts in which formal Master and Structure planning may be appropriate, and contexts where it is not. In conventional  138  greenfields developments such as sites-and-services schemes, one of the primary purposes of planning is to predetermine and guide the nature offuture settlement and urban development via a set of procedures as prescribed by various statutorily defined planning tools. In such instances, the production of the Master or Structure plan documents which precedes human settlement, are framed within the rules of the formal city. In these terms, Conyers (1982) notes that the 'planning process' is focused on the production of a plan document, which implies limited involvement of planners in the formulation of policy strategy, and even less direct involvement in the implementation of plans.  In the context of informal settlement in-situ upgrading however, planning interventions need to be concerned with how to most effectively address situations in which settlement has already occurred. In such contexts, not only does the planning process need to be continuous and  involve (1) policy formulation, (2) planning design, and (3) implementation, but it also needs to be able to integrate the complex social, economic and political imperatives already existing within informal settlements. Logically then, the planning processes that apply to greenfields sites-andservices developments and those that apply to informal settlement in-situ upgrading are fundamentally different. The former create legal and regulatory frameworks into which people subsequently move, whereas the latter requires the integration of the formally planned city with the de facto social fabric of informal settlements.  As described by Baross (1990), conventional urban planning is thus pursued according to a sequence in which local site planning is pursued according to a (P) planning, (S) servicing, (B) building, (O) occupation sequence (PSBO). This formal planning process is, however, contradicted by the reality of Third World cities in which the expansion of city growth via the  139  construction of informal settlements is most typically based on a reversal of this conventional framework - in which 'formal' planning becomes the last step in the (O) occupation, (B) building, (S) servicing, (P) planning sequence of development (OBSP).  An important point in the context of informal settlements however, is that a separate and quite different planning process, is overlaid upon this OBSP sequence. In this regard, the informal planning process of settlers themselves not only precedes occupation, building, and initial servicing but also continues iteratively throughout the growth and longer-term existence of the settlement. It is this informal planning processes that comes into contact (and conflict) with formal planning at thefinalstage of the (OBSP) sequence, which is the point at which formal planning interventions arrive on the scene for thefirsttime. A central difficulty for planners working in informal settlements is that in urban systems where the formal and informal are highly integrated, planners effectively straddle the two 'worlds' but have generally only been educated and equipped with conceptual and practical tools to function in the statutorily regulated city in which legality is paramount.  Contradictions between formal and informal planning processes arisefromthe application of formal planning tools such as planning Statutes and Master Plans, notwithstanding the seemingly trite assertion that planning in a greenfields context in which settlement has not yet occurred (PSBO) needs to be fundamentally differentfromplanning in situations where settlement has taken place and in which an informal planning process is already underway (OBSP). Contradictions exist both at the procedural level noted above, as well as in terms of these formal planning tools' legalistic orientation. An implication of this legalistic approach is the attempt to 'fix' or formalize the (highly fluid) constituent parts of informal settlements.  140  The point arising is that development processes need to be structured in order to accommodate the messiness of the informal, rather than attempting to apply processes designed for the formal city on the assumption that those same processes can or should also apply to the informal. It is necessary to move awayfroma 'fit-your-(informal settlement)-problem-to-our(formal city)-solution' mentality, and rather attempt to construct conceptual tools and practical solutions around the specifics of the informal situation at hand. Here, the de facto social norms, institutions, and conventions applicable in informal settlements dictate the 'rules of the game', and planning in the informal setting needs to anticipate and accommodate continual changes in these rules.  Further implications of attempting to apply formal planning procedures to the informal are at least threefold:first,the marginalization of informal settlements is reinforced, second, it implies a breakdown of the social fabric of the informal, and third, it forgoes the opportunity to build on the many positive aspects of informal settlements.  4.3.2 Reinforced marginalization of informal settlements  Without entering into the extensive debate on problems surrounding the fonnal:informal dualism (reviewed, for example, in Peattie 1987), an important point here is that the conceptual separation between formal and informal, when translated to the level of practical intervention in informal areas - whether in terms of development intervention, or authority regulation and controls - manifests itself in a pattern where one set of practices and 'rules of the game' are applied in informal settlements, while another is applied in the formal. Alternatively, and where no policy exists which explicitly and comprehensively addresses the development of informal settlements then planning regulations, standards, and procedures that were designed for higher-  141  income formal areas are often applied uncritically to informal settlements. Either way, it is the interests of the formal city that are primarily served by such policies. For example, as Leaf (1991, 28) points out in a study of Jakarta, it was the affluent and powerful members of that society, through their political voice, who participated in the setting of policy; the 'have-not's, on the other hand, were excluded. While this is a phenomenon that is, to varying degrees, widely generalizable, the important point is that informal residents probably represent the most marginalized category of'have-not's' in Third World cities.  4.3.3 Breakdown of the social fabric The rules of the game in the formal and informal are fundamentally different: the informal society aims at effectiveness, not at efficiency as in the formal. Within informal settlements, "because the economic is set within the social, the coherence of the whole - or social reason - takes priority over market reason. Putting such a system into competition with the techno-economic model poses a severe challenge. Also, although informal social structures are almost indestructible on their own ground, they are very weak and vulnerable when engaging with outside territory" (Latouche 1993, 154). Formalization of the informal tends to destroy the social ties in informal settlements, and instead introduces a different set of norms and institutions grounded in rules defined by the state and regulated economic markets - which are given formal status in law and de jure regulation. When the informal starts to become formalized, it becomes more fragile in that social networks and rationality give way to the political and economic rationality of the formal city.  While empirical evidence provides the basis for consensus across the political spectrum that some breakdown in the coherence of settlements' internal social fabric does indeed occur during the process of urbanization and settlement growth, it is noted that the reasons for such social reorganization in informal settlements are the subject of much controversy, in respect of  142  which contending explanations were noted earlier. Importantly, it is not the relationship between formal and informal per se that contributes to breakdown of the social fabric of the informal. Rather, it is the nature of that interaction, in which the formal attempts to shape and regulate the informal in its own terms, that is central to that breakdown and the replacement of informality with formality.  4.3.4 Missed opportunities: building on the positives  On the other side of the proverbial coin, formalizing the informal forgoes the opportunity to build on the positive aspects of informal settlements, not only in terms of their internal functioning, but also in terms of the contribution they could make to more effective management of the broader urban system. As noted earlier, many of the positive aspects of informal settlements were clearly articulated in the tradition following the early writings of John Turner, and anthropologist William Mangin who argued in support of facilitating the development of the creative potential of shack dwellers. Leaving the ensuing self-help debate aside for the moment, the point to note is that many so-called 'bottom-up' approaches to urban development are rooted in conceptions of informal settlers which argue that "the true planners and builders of Third World cities are the poor. They organize, plan and build with no helpfromprofessionals. Drawing on their own skills, making the best use of limited resources and forming their own community organizations, they account for most new city housing"; indeed, as Hardoy and Satterthwaite (1989, 12) continue, between 70 and 90 per cent of new housing in Third World cities is built 'illegally'.  Whilst recognizing that although the quality of the physical housing stock often fails to meet the aspirations of their occupants, the informal housing delivery process is vastly more  143  affordable and flexible than the formal delivery process, and its incremental nature is something that the formal delivery process cannot match (Dowall 1991). Over and above the very substantial quantitative contribution to the overall housing stock, are some qualitative aspects that are worthy of note. The informal housing delivery process is a process that settlement residents can control themselves -fromthe planning of initial settlement, through shelter construction, and including the ongoing informal regulation of day-to-day functioning of the settlement. This is however not to over-emphasize the real 'empowerment' potential of such 'bottom-up' processes about which more later.  4.3.5 The challenge: a paradox Given these positives and negatives, a paradox to be engaged in the formulation of a planningframeworkfor informal settlement contexts, is what Latouche (1993) describes as the "logical paradox of the normalization of the informal'' in which the pressure on the informal to 'normalize' and 'modernize' is quite perverse, given that the informal is already a synthesis of modernity and tradition. The paradox, or contradictory double imperative that may be posed, thus involves "a question of somehow conserving the dynamic and original quality of this creative activity while simultaneously coming to take the form of a mimetic development" (Latouche 1993, 151-152).  Having thus far (1) described the dominant approach to planning in informal settlements, (2) why this approach continues to prevail, and (3) why such an approach is problematic, the discussion will now move on to a consideration of factors that may be useful in addressing the problem.  144  4.4 REGULARIZATION AND THE CONTINUITY OF THE UPGRADING PROCESS 4.4.1 Sustainability: an indulgence?  A concern with 'sustainability' issues - and preceded by literature dating back to the 'Limits to Growth' school (Meadows, et al. 1974) - is synthesised in the oft-quoted statement of the 1987 Brundtland Report that we must meet the "needs of the present generation without compromising the ability of future generations to meet their own needs". More recently, debates following the United Nations Conference on Environment and Development (UNCED) in Rio de Janeiro in 1992 (for example, Costanza 1992; Jannson et al. 1994), have reflected a number of strands within which social, economic, political, and ecological 'sustainability' issues are variously argued to be central.  However, within informal settlements where most people are primarily concerned with their immediate livelihoods and day-to-day survival issues, it is necessary to address the widespread perception by the poor that sustainability issues are a fashionable indulgence of academics, the affluent, or the middle-classes. It therefore becomes necessary to make explicit the link between concerns with 'sustainability' and the immediate life circumstances of the poor. Translating trade-offs in achieving social, economic, political and ecological goals into a practical strategy, means that 'development' goals first and foremost need to meet people's basic needs via the provision of access to adequate shelter and a healthy environment. By extension, access to basic physical and social services implies the right to participate in national and local politics, and access to adequate livelihood - which itself often implies access to natural resources.  In this regard it is useful to distinguish between 'growth' and 'development', arguing that "...growth is quantitative increase in physical scale, while development is qualitative improvement  145  or unfolding of potentialities" (Daly 1990, 1). In these terms, sustainable development is indistinguishablefromthe total development of society, given that 'sustainability' depends on the interaction of economic changes with social, cultural, and ecological transformations. Consequently, it is not easily subject to measurement by any concept of direct and measurable economic gain. According to this definition, (1) its quantitative dimension is associated with increases in the material means available to those living in poverty, so as to provide for adequate physical and social well-being and security; and (2) its qualitative dimension is multifaceted and is associated with ensuring the long-term ecological, social, and cultural potential for supporting economic activity and structural change. The primary objective is to provide lasting and secure livelihoods that minimize resource depletion, environmental degradation, cultural disruption, and social instability (Barbier 1987).  Moving down a level of abstraction, and applying these concepts to the project level, there are two central levels of concern. In order for the impacts of any upgrading approach to be sustainable, it is necessary both to achieve an immediate improvement of settlement residents quality of life, as well as the longer-term requirement of facilitating the continuity of the upgrading process. A necessary prerequisite to achieving this end is to build the capacity of institutions that will sustain the upgrading process. Since institutions comprise both structures and norms, this requires that planning interventions put in place (1) the organizational entities and relationships between them and (2) methods and processes that will sustain project impacts into the longer term. In this context, and drawingfromthe literature consulted thusfar, it becomes important then to outline at the level of principle, some of the prerequisites or necessary ingredients for building such institutional capacity in informal settlement upgrading and thereby facilitating the continuity of the improvement process.  146  4.4.2 Integration or disintegration of the de facto informal?  In attempting to plan toward the more efficient and equitable integration of informal with formal settlements, important questions arise. First, should an integrative planning process attempt to retain the social fabric (or at least aspects of it) of informal settlements? If so, what principles or reference points should inform the identification of compatible positives and negatives of the respective 'worlds' which could contribute to more effectively functioning urban systems?  It has been argued that a narrow conception of legality is problematic. This is however not to argue that legal frameworks and regulation be disregarded entirely. On the contrary, many of the de jure rules may ultimately be amended as direct result of'success' in upgrading processes anyway. Indeed, McAuslan (1992;1993) makes a strong argument in favour of a strong appropriate legal basis to urban development planning. He argues, furthermore, that some of the  principles that could usefully inform the construction of such a framework to facilitate the integration of formal and informal imperatives are that it should be equitable, flexible, environmentally conscious, participative and easily manageable, simple to understand and use, efficient and administratively fair.  With these principles as a starting point, it becomes more feasible to take the de facto rules of the game that are characteristic of informal settlements and integrate them with the rules according to which the 'formal city' operates. Importantly, integration is the operative word here. In that integration process, some of those de facto norms and structures will be eliminated completely, some will be amended and formalized, and those which contribute to a more efficient urban governance system will be retained and integrated into the overall metropolitan governance  147  There are two primary levels of importance here, each of which needs to be integrated with the other. Planning needs to build upon local informal settlement-level administration systems (Montgomery 1988) which are grounded in an understanding of the complex de facto social make-up of those settlements in the first instance, and the strengthening of de jure local government in the second (Baross and van der Linden 1990). This needs to be executed both at the local settlement level, and also in relation to the development of frameworks for the management and governance of metropolitan regions (Clarke 1991).  In order to achieve these objectives, assistance is provided by the broader concept of informal settlement 'regularization' - which can be defined to "include all measures (political, legal, administrative, etc.) and practices relating to already occupied urban areas, which help to improve or guarantee security of occupation. Thus, the legalization of individual occupancy by providing title deeds is only one form of regularization, albeit the most formal and definitive" (DurandLasserve and Pajoni 1993, 3). Regularization represents a more flexible approach to upgrading, allowing a gradation of strategies designed to facilitate the more equitable and efficient integration of informal settlements into their urban contexts, and thereby also contributing to an amended legal regulatory framework that is more appropriate in informal settlements - and which, as van der Linden (1994) points out, may be an emergent direction in which low-income housing solutions are evolving.  4.4.3  Balance between economic, social, and physical development  If planning is to more creatively engage the irregularity of informal processes, this requires a shift awayfromplanners' traditional orientation of attempting to strengthen formalized control, toward one that is, in addition, equipped to deal with complex and fluid sociological  148  processes. Although the importance of the social development aspects of upgrading have long been recognized in housing literature, upgrading initiatives have nonetheless historically been primarily focused on the delivery of physical product, whether in the form of physical services, infrastructure, or legal tenure. By contrast, the social aspects of development have been 'tacked on' as a secondary event, rather than as an integral part of the upgrading process.  The relationship between the social development and physical/technical development components of upgrading need to move awayfroman emphasis on physical infrastructural inputs as the sole or primary activity, to an integrated and more balanced relationship between both the physical and the social. This requires consideration of prioritization in terms of timing, and (financial and human) resource allocation in which the social development components need to be linked to physical infrastructural inputs, as an integral part of the upgrading process. Thus, rather than development debates reflecting an either/or conflict between placing politics or technical issues at centre stage, such an approach would recognize that upgrading processes are simultaneously political and technical. Rather than the problem being posed as product versus process, the focus should, as a principle, be on an integration of both: the most appropriate process via which to deliver product.  4.4.4  Impact on policy: linking the local and city levels  In order to avoid the benefits of upgrading being limited to the confined locality of specific settlements, it is important that local informal settlement-level interventions are linked to policy change at a city-wide level. A number of points arise here. Afirstimportant element of the local/city-level link concerns the scale of development activity - which raises the depth versus breadth trade-off. In terms of this trade-off, strategic policy needs to balance potential impact  149  in respect of day-to-day needs of ordinary informal settlement residents at the local level, with regional and national level imperatives.  One approach could be to initiate upgrading processes at a small (project-level) scale, and attempt a progressive build-up in the scale of development activity, so that the development process itself can be the stimulus for the building of capacity (Davidson 1991), and for broader urban change. Ultimately though, the scale of upgrading initiatives need to reflect a balance between (1) small-scale projects which achieve the benefits associated with depth, high levels of resident participation, greater likelihood of building strong social organization, effective ongoing settlement management, and successful cost recovery, and (2) large-scale programmes such as the KIP which achieve greater lateral coverage in terms of dispersal of benefits.  Second, when referring to the local informal settlement-level, the distinction needs to be made between settlement leadership and ordinary settlement residents. Unless this is done, the involvement of CBOs and their leadership is easily misinterpreted as reflecting the interests of the 'grassroots' masses - when, in fact, the leadership of CBOs are often somewhat detachedfromthe ground as they become increasingly involved in city-level and even provincial-level party politics, while the latter comprise the rank-and-file, ordinary settlement residents who often have little or no direct involvement in civic politics and whose lives are oriented primarily around day-to-day basic needs. This has implications for capacity building strategies targeted at CBOs - and which have the potential to reach and build the capacity of the leadership of these CBO structures. However, capacity building objectives within upgrading programmes need to be explicitly clear that improvements in the quality of ordinary peoples lives - not only of settlements political leadership - are paramount. Failing this, the impact of such capacity building strategies will be  150  to increase the gap between leadership and ordinary residents in terms of skills, access to the technology and resources of the formal city, decision-making power, as well as political and patronage connections.  Third, and on a more practical note, while ethnographic methods may be used to justify anthropological research objectives, a similar logic applies to planners working in informal settlements on development projects. The most useful, irrformative, and accurate sources of information on the perceptions and views of the mass of settlement residents vis a vis any development intervention are the informal, day-to-day, communications with, and stories of, ordinary people living in the locality (Peattie 1983, 1990). It is based upon an understanding of informal settlements at this level - which extends below the level of political leadership and of organized Committees - that the incorporation of indigenous knowledge, skills, and resources in the planning process becomes most feasible.  4.4.5 Economic replicability and feasibility At an economic level, the importance of the linkages between macroeconomic policies that are largely managed at the national level and urban economic policies that are managed at the city level has been extensively argued for example in World Bank (1991b). A similar argument can be put forward regarding fiscal, financial, and real sector linkages between the urban metropolitan economy and the local settlement level. Given the widespread withdrawal of national governments from city-level development initiatives, of increasing importance is the strength of the urban economy and the consequent availability offinancialresources as a base for economically sustainable development initiatives. In the case of upgrading in which substantial portions of the installation of infrastructure and services need to be provided on a non-cost  151  recoverable basis given the low incomes of target groups typical of upgrades, an economic resource base needs to exist such that funding availability can either be replicated on a grant or heavily subsidized basis, or via cross-subsidy mechanisms which may be via higher income users within projects, between projects, or through fiscal transfers at a metropolitan level.  Internationally, one of the major shortcomings of upgrading has been the limited ability to recover the running costs of ongoing service provision. International experience indicates that the recovery of these costs is less a function of narrow ability-to-pay criteria, than of the ineffectiveness of political and development management systems (Williams 1984; Laquian 1983 a; Keare and Parris 1982; Pugh 1988; Agbola 1990). Factors found to be of central importance in cases of successful cost recovery are the meaningful participation of settlement residents in upgrading decision-making (Bamberger and Deneke 1984), affordable standards of service provision (Pugh 1988), and effective development management systems (van der Linden 1986; Devas 1993).  Also, at the project or programme level, development finance parameters need to facilitate the achievement of development objectives, not hinder or constrain them. In the case of informal settlement in-situ upgrading, these financial parameters ideally need to be able to accommodate the flexibility and fluidity that is characteristic of informal settlements, without losing the economic rigour necessary for efficient development funding mechanisms.  4.4.6 Political feasibility Drawing from international experience, a critique of the so-called Alternative Development paradigm provides insights that could usefully guide future planning strategies in  152  the South African context. In this regard, the following passage is worth quoting at some length given its pertinence in the current South African context of political and institutional transition: "...the primary lesson is that one needs to be institutionally knowledgeable when proposing a new idea. A new idea should lend itself to being implemented by gradual modification of new ways of doing things. One should also realize, when proposing a new idea, that it must be able to generate institutional consensus^ and this requires that the new idea not be proposed in terms too antagonistic to old ways of doing things. One must also acknowledge that a new idea will need an institutional basefromwhich to be implemented. Institutions usually prefer clearly defined and uncomplicated ideas which are easy to implement. In case no existing institutions are able to absorb the new idea, however, its proponents may push for the creation of a new institution; but to be successful in that effort, they must be in close connection with top-level policymakers. [Furthermore], independence from powerful institutions — particularly the government — is not a precondition for success. Conversely, embeddedness in the existing institutional network is essential for the successful implementation of new ideas" (Sanyal 1994, 33). Unless practicable, all the principles identified thus far are of no more than academic interest and do nothing to bridge the disjunction between theory, policy and practice. Also, to argue in favour of practice-based theory and policy formulation is of limited use unless the lessons of practice have some realistic prospect of acceptance by policy makers. Of central importance therefore is the political feasibility of any approach being proposed.  The fact that the Besters Camp upgrade traversed contrasting political dispensations ranging from the State of Emergency under Nationalist Party government of the late 1980s through to the transitional national, provincial and local government arrangements of the 1990s further complicated attempts to formulate a politically viable upgrading strategy in Besters Camp. A first challenge was the need to gain acceptance of the concept of settlement upgrading from government officials whose approach to settlements had historically been to remove them. However, more important in terms of achieving longer term impacts, was that once initiated, the upgrading process would need to be have a realistic prospect of acceptance by new policy makers  153  in transitional government structures.  Having outlined a number of factors pointed to in the literature as needing to be considered in the formulation of an approach to plarining in informal settlements, it now becomes possible to examine the approach adopted in planning of the Besters Camp project and the extent to which it contributed to the building of institutional capacity, and thereby, to the continuity of the upgrading process.  154  CHAPTER FIVE INITIATION AND CONCEPTUALIZATION OF THE BESTERS CAMP UPGRADE  The continuity of any project's upgrading impact is affected by four broad groups of factors: (1) how the project is designed and implemented; (2) how the project is organized; (3) external factors operating at the local, national, and international levels; (4) and the responses of actual and intended beneficiaries (Valadez and Bamberger 1994, 188-191). Making reference to the evaluative criteria outlined in Chapter 4, the following chapters examine the Besters Camp upgrade with a view tofirstestablishing whether the upgrade made an impact in improving the lives of settlement residents. The second, and key, question that follows is whether the project built the institutions - meaning norms and structures - to sustain the improvement process into the longer-term. Answering these questions will involve consideration of what the upgrade delivered, how the delivery took place, and why delivery took place the way that it did - all with reference to the theme of whether the delivery process either inhibited or contributed to the continuity of the improvement process.  Given the complexity of the project, the following chapters will be structured in such a way as to present the historical evolution, or unfolding, of events on the project. This chapter begins with a brief description of the profile of Besters Camp immediately prior to upgrading in 1990, following which the relationship between policy formulation and implementation that was applied on the upgrade, is elucidated. The shaping of the participation strategy and how this informed the initial conceptualization of the project is then considered. This provides the basis for Chapter 6's examination of the local settlement level impacts of project implementation, and the consideration of the relationship between the project and its external context in Chapter 7.  155  5.1 BESTERS CAMP SHACK SETTLEMENT IN 1989 5.1.1 Physical and social infrastructure In 1989, the 160 hectare Besters Camp informal settlement was bounded by formally planned townships to the south, west and east, while to the north west, dense informal settlements extended for another 10 kilometres as part of the greater Inanda area. As indicated in Figure 5.1 below, to the east and north-east was the Phoenix residential area, with KwaMashu to the south and south-west, and Ntuzuma to the west.  The only tarred roads bounding the settlement were two major arterial bus routes. First, the MR93 separated Besters Camp from Phoenix, an area designated for Indian residential occupation by the Group Areas Act. This road was in the process of being upgraded to a four lane divided expressway linking the greater Inanda area to Durban's N2 outer ring freeway. A second bus route - the Ntuzuma access road - with its distinctive S-shape alignment wound through Besters Campfromeast to west. The settlement thus had easy access to public transport routes on which both public buses and minibus taxis operated. There were no formal roads within Besters Camp and so no taxis entered the settlement. Paths and tracks had been cleared during the process of settlement, but in the absence of any stormwater management system, were regularly washed away in parts or rendered unusable after heavy rains.  In 1989, Besters Camp did not enjoy the provision of any formal physical or social service infrastructure. In the absence of municipal garbage removal, on-site pit burning and the dumping of garbage into pit latrines were the most common forms of refuse disposal. There were 4 communal water points within the settlement. Sewerage was disposed of via pit latrines which were generally poorly constructed, and also used as garbage disposal pits, accounting for their  156  Bhambayi  Shembe Phoenix  Ezimangweni Newtown  Nhlungwane  Besters  Ntuzuma Hospital site KwaMashu Figure 5.1  Besters Camp and adjacent townships and settlements in 1990  157  average lifespan of less than one year. Other than trenches and channels dug by residents, there was no stormwater management system within the settlement.  Gradients throughout the settlement were very steep, rangingfrom1:6 to as steep as 1:2.5 in some areas, while densities ranged between 40 and 90 dwellings per hectare throughout the settlement. A survey of Area 1 of Besters (UFISD 1991,1) indicated a median occupation density within shacks of 6.3 square metres per person, while a quarter of residents had less than 2 square metres of living space within their shacks.  There were also no electricity or telephone connections within the settlement. Very few people had generators, with the most common energy source for cooking and heating water being paraffin (94 per cent), while 66 per cent of households used candles and 32 per cent used paraffin for lighting. Gas was used for refrigeration by 88 per cent of households. There was no educational or health infrastructure within Besters Camp. The health service used by most Besters Camp households (89 per cent) was a clinic that was operatedfromthe KwaMashu township. There were also no community facilities, and meetings took place outdoors, or in private shacks for smaller meetings (UFISD 1990b; RSA 1989b).  5.1.2 Demographic characteristics  The average age of Besters Camp residents was low, with an average age of 22 for the whole population, and 36 when considering adults only (people over the age of 20). This can be accounted for by the large influx of young adults into the area during the post-1985 period. The high profile of'the youth' in the politics of the settlement is given weight when considering that only 12 per cent of the Besters Camp population was older than forty years of age. This skewed  158 distribution in favour of younger age cohorts is consistent with statistics for greater Inanda (DDA 1990,23) which indicate the bulk of the population (52 per cent) being between 16 and 50 years of age, and only 3.2 per cent being older than 50 years of age. The sex breakdown in Besters Camp was about even, with 49 per cent of Besters Camp population being male, and 51 per cent being female.  Almost two-thirds of Besters Camp residents (68 per cent) were born in rural tribal areas, but considering that only 16 per cent had moved into the settlementfromrural areas, it is clear that the primary migration pattern into Besters Camp was not a rural to urban one, but far more strongly an intra-urban movement. More than one-third (36 per cent) of all residents, and onequarter of adults (25 per cent) were born in the Durban metropolitan region, indicating a high proportion of Besters Camp residents being second and third generation urban dwellers. This pattern is consistent with other studies indicating circulatory migration patterns as being typical in South Africa generally (Mabin 1990, 1992), and in Durban's informal settlements specifically (Cross etal. 1992a, 1992b).  Shack occupancy in Besters Camp was lower than average occupancy rates for informal settlements across the Durban metropolitan region of 8.9 (Sutcliffe 1989, 2), and averages of 8.5 for greater Inanda (DDA 1990, 187; UND 1990, 7; RSA 1989a, 6). Surveys in Besters Camp indicated average occupancies of approximately 7-8 people per shack (UFISD 1990b, 2.3; CHESS 1993, 11). Whereas average households per shack for older settlements was 1.6, the lower occupancy rates in newer settlements such as Besters Camp may be accounted for by the limited extent of sub-letting.  159  Education profiles indicated very low levels of literacy. The mean years of formal schooling for adults having had the opportunity to attend schools in rural areas or in townships was five years. Given the non-existence of schools in the informal settlement, and that only a small number of children were able to register at schools in adjacent formal townships, average levels of schooling for children were also low, at only 4 years. Furthermore, ten per cent of adults had never received any formal education at all.  5.1.3 Economic profile Socio-economic survey statistics on employment and income need to treated with caution. While all households had some form of income, whetherfromformal employment, the informal economic sector, pension or welfare payments, or family and kinship networks, survey statistics reflected an underreporting of income and employment levels - which could be interpreted to imply lower than real affordability levels, low ability to pay for services, and thus higher levels of subsidy required. In addition, higher income generating activities within Besters Camp such as drug growing and dealing, trading in stolen merchandise, as well as 'chop shops' (which stripped and reconstructed stolen vehicles) were unlikely to be reported. With those qualifying statements in mind, it is noted that the data in the following two sections is drawnfromproject documents and socio-economic surveys commissioned by the UFISD (UFISD 1990a, 1990b).  These surveys indicated that 42 per cent of adults were in formal sector employment, while 14 per cent derived their incomefrominformal sector economic activity. Unemployment levels were high at 40 per cent, although many of these people were engaged in undisclosed informal income-earning activities.  160  Regarding the nature of economic activity in which people were engaged in 1990, the manufacturing sector was the largest employer of Besters Camp residents (35 per cent), while commercial (22 per cent), construction (13 per cent) and paid domestic labour (15 per cent) were the next largest employers of people. Importantly, place of employment and commuting patterns suggested a high degree of economic integration with the broader urban context. Almost three quarters of adults (73 per cent) commuted to their place of work in the Durban municipal area on a daily basis, while another 16 per cent travelled to the nearby centres of Pinetown, Phoenix, and KwaMashu. Only 7 per cent conducted all their economic activity within the Besters Camp settlement.  Whilst highly integrated economically, economic links to the metropolitan core appear to befluidand transitory, reflecting a high level of economic mobility. Almost one-third of adults (30 per cent) had been in their present employment for less than two years, while over half (52 per cent) had changed jobs in the last six years. Previous periods of employment reflected similar levels of mobility, with 30 per cent of adults having spent less than two years in their previous jobs, and 58 per cent having moved occupations within the previous four year period.  Mean household income in Besters Camp in 1989 was R787, which was lower than the mean for Durban's African townships of R830. Expenditure patterns indicate a focus on basic household items, with food (43 per cent) and clothing (23 per cent) accounting for the bulk of monthly expenses. Given the extremely low incidence of sub-tenancy, and the termination of rental payments during the post-warlord period, expenditure on accommodation accounted for an average of less than 1 per cent of overall monthly household expenses.  161  Private car ownership was low, with only 7 per cent of households in Besters Camp owning vehicles. In addition to the obvious income constraint, vehicle ownership was also constrained by the rugged topography and high densities that rendered many parts of the settlement inaccessible to vehicles. Transportation - mainly the costs of commuting to and from work - was the third highest monthly expense at 15 per cent. Limited savings were reported, with most households indicating no monthly savings, while total savingsfromall sources ranged between Rl,581 in Besters, through to a mean of Rl,846 in Nhlungwane. These low total savings reflecting a limited private financial resource basefromwhich to pay for services, or to invest in future home upgrading.  5.1.4 Movement patterns While circulatory migration patterns indicate movementfromrural to urban, as well as numerous intra-urban moves, short-term patterns also indicate a high degree of mobility, with many adults keeping two 'homes'- one in the urban informal settlement, and one in a rural area. While spending most of their time in Besters Camp, 41 per cent of adults returned to their "other home" on a weekly basis - usually on weekends, while 26 per cent visited home at least once a month. This is consistent with migration patterns described nationally and at the metropolitan levels in earlier chapters and confirms the thesis that many migrants and their descendants seldom entirely cut ties with their (primary or secondary) bases in rural areas.  Regarding the degree of stability of the Besters Camp population, the average period of residence in the settlement was over three years (42 months), although profiles varied within subareas of Besters Camp given the different settlement histories of the respective areas. In Besters, only 2 per cent of adults had been living in the area for four years or more, while in Nhlungwane  162  and Ezimangweni, 35 per cent of residents had been living there for more than four years. Exploring the movement patterns further, 60 per cent of Besters Camp residents' previous place of residence was a formal urban township, while 15 per cent had previously lived in another urban informal settlement. Only 16 per cent of Besters Camp residents were living in the area as their first entry point into an urban area, having previously lived in rural tribal areas. Thus while its population was highly mobile in terms of continually changing economic activities, Besters Camp settlements were nonetheless characterised by a relatively stable core population that retained their residential base in the settlement for medium to long-term periods of time.  5.1.5 Legal and jurisdictional complexity At the time of project initiation, Besters Camp was located partly within the Durban municipal area, partly within the KwaZulu local authorities of KwaMashu and Ntuzuma, and partly within Released Area 33 (Table 5.1 and Figure 5.2 below). Township establishment procedures within each of these four jurisdictions were different. In the portion of the settlement within Durban, procedures as contemplated by the Less Formal Township Establishment Act of 1991 (LFTE) were applicable. In the portions of the settlement within KwaZulu, planning was Jurisdiction  Local authority  extent (ha)  Land ownership  Legal tenure  Durban  Durban City Council  36.68  RSA govt. 33 ha private 3.68 ha  Certificate of Ownership  KwaZulu  KwaMashu  13.8  KZgovt.  Deed of Grant  KwaZulu  Ntuzuma  12.9  KZgovt.  Deed of Grant  RSA govt 87.91 ha private 10.63 ha  Certificate of ownership  Released Area Dept. Of Development 98.54 33 Aid/NPA  Table 5.1 Jurisdictional makeup of Besters Camp settlement. Sources (UFISD 1989a; DDA 1990).  163  to be conducted according to KwaZulu's Proclamation 293. In the portion of the settlement within Released Area 33, a separate legal provision (Proclamation 131/90) was applicable. Provision however existed within the LFTE Act for it to be made applicable in Released Area 33, although no precedent for such a designation existed. Consequent upon these legal provisions, the forms of tenure that were applicable in the respective jurisdictions were Certificate of Ownership in Durban, Deed of Grant in KwaMashu and Ntuzuma, and Certificate of Ownership in Released Area 33.  These legal frameworks according to which the respective authorities required the planning process to be conducted, thus divided Besters Camp into four segments each to be treated differently, notwithstanding that in functional and social terms, Besters Camp constituted an indivisible entity.  5.1.6 Political antagonisms The theme of the tension between de facto and de jure that was introduced in earlier chapters was clearly evident in Besters Camp's complex political configuration. Following the elimination of Inkatha warlords in late 1989 and early 1990, and the subsequent consolidation of control by the 'comrades', the settlements of Besters, Nhlungwane, and Ezimangweni were under the de facto control of the UDF, and subsequent to the unbanning of the ANC, were commonly identified as 'ANC settlements'. While this shorthand was a somewhat generous description of the extent of real organizational presence in the area, and it would be more accurate to describe the settlements as being under the control of loosely affiliated groupings of activist 'comrades', the important point from a political perspective was that these settlements became identified as strongholds of anti-apartheid social movements - thereby placing them in direct opposition to  164  Figure 5.2 Legal jurisdictional boundaries in relation to Besters Camp in 1990  165  governmental authority structures. Police did not enter the settlement, thereby leaving residents to assert their own policing and administration of justice via 'peoples courts'. What made this politically complicated for the proposed project was that the legal authorities for Besters Camp spanned the spectrumfromthe Inkatha oriented KwaZulu authorities, Nationalist Party provincial and central government, and the Democratic Party - Nationalist party dominated Durban City Council. The local councillor for KwaMashu, within whose legal jurisdiction part of Besters Camp was located, was an outspoken Inkatha spokesperson for the region, while the Ntuzuma mayor, although less vocal, was also a prominent Inkatha member. Their township councils, also Inkatha, were initially poorly disposed to the concept of an upgrade which would imply the delivery of services to settlements that had been perceived to have changed allegiance to Inkatha's most violent enemy.  The portions of the settlement within Durban were under the political authority of a City Council which, at the time, was an amalgum between councillors with a liberal Democratic Party orientation and those with a conservative Nationalist Party allegiance, tempered by a small number of extreme right-wing councillors. Fortunately however, councillors who were well disposed to an upgrade were strategically well placed, being Deputy Mayor and Chair of the Health and Housing Committee respectively. Portions of the settlement within Released Area 33 fell under the jurisdiction of the South African Development Trust. In practice however, responsibility for development initiatives in this area were dealt with by the Department of Development Aid and the Natal Provincial Administration - both ultimately answerable to Nationalist Party central government in Pretoria.  The crux of the challenge then, was to formulate an approach that would be accepted by  166  the range of de jure authority structures, and for that approach not only to also be acceptable to the politically adversarial de facto regime in Besters Camp, but also for settlement residents to be active participants in the formulation and implementation of those proposals.  5.2 PROJECT INITIATION 5.2.1 The policy formulation process At the time of the initiation of the Besters Camp project in 1989, there had never been a successfully implemented large scale informal settlement in-situ upgrade in South Africa. Experience of'self-help' housing even in the sites-and-services arena was also limited. In 1983 the Urban Foundation (UF) published Guidelines for self-help housing: principles, practical issues, and case studies, a document based on thirteen major self-help projects completed in  South Africa, and which represented the bulk of low-income housing provision outside the homelands during the 1970s and early 1980s. These projects focused on the creation of new lowincome housing stock in the context of planned interventions including sites-and-services, core housing and shell housing, but excluded informal settlements and their upgrading - which were still the subject of forced removals and demolition.  The UF'sfirstattempt at a large scale in-situ upgrade was for the upgrading of the Crossroads settlement in Cape Town. However, following the devastation of the settlement in 1986, the Urban Foundation withdrewfromthe area without implementation having commenced. Other proposed upgrades such as that of St. Wendolins in Durban (BESG 1986) also never progressed far beyond the stage of proposal formulation, partly given the hostility of the state to urban informal settlements per se, and partly also given the absence of a policy,financial,or legal framework that was supportive of upgrading projects. In addition, the delivery capacity of non-  167  governmental development agents targeting the very low-income informal settlements in South Africa was, until the end of the 1980s, extremely limited. The Urban Foundation's project activities in low-income housing had been confined to the formal housing arena, and until the end of the 1980s, NGOs such as Planact in Johannesburg, and the Built Environment Support Group (BESG) in Durban, while being prominent actors in townships and shack settlements, were primarily advocacy organizations with no delivery capacity, never having completed the implementation of an upgrading project.  Taking a proactive stance, and supported by an extensive commissioned research data base, the UF formulated an informal settlement policy position in respect of site-and-service and in-situ upgrading in 1989. In terms of this framework, the primary goal of the UF was to "institutionalize informal housing delivery as a major component of housing supply in South Africa" (UF 1989a, 11). In order to achieve this goal, the UF saw its role both as change agent and as development agent whereby it would bring about change in both the public and private policy environments, by marketing the concept of informal housing delivery so as to change negative attitudes to informal housing, and via in-situ upgrading projects, demonstrate the efficacy of informal delivery processes.  Changing the policy environments was to involve "the development of the content of a national informal housing policy and get it adopted by government" (ibid), establishing an appropriate legal framework for the implementation of such a policy, and eliminating existing public policy constraints in respect of financial subsidy policies, standards, land identification and supply mechanisms. At the level of demonstration projects, objectives were to show what had already been proven in international upgrading experience: ie. that by upgrading to minimum  168  health and safety standards, upgrading can reach the lowest income percentiles, and borrowing from Turnerist language, "that settlements can be upgraded into settlements of hope" (ibid, 12). It was intended thereby to contribute to the case against continued forced removals of informal settlements by the South African government.  The Urban Foundation's subsequently published policy statement on informal settlements (UF 1991a) - which at the time of its publication was the only substantial statement on informal settlement policy in South Africa - was formulated before even the UF or any agency in the country for that matter had acquired any substantial experience in informal settlement upgrading. This was however not atypical of housing policy formulation processes more generally (see for example Stren 1993), in which the responsibility for formulating policy was vested with research and policy specialists with limited practical experience of working in low-income settlements.  5.2.2 Project identification In 1989, the Urban Foundation identified two potential 'demonstration projects' in South Africa - one being the Soweto-on-Sea settlement in Port Elizabeth, and the other being the Besters settlement in Durban. The initial identification of Besters only comprised the portion of the settlement located within the Durban Municipal area. The Urban Foundation conceptualised the intervention as a 'demonstration project' which would demonstrate to the South African state in general and the developmentfraternityin particular, that the in-situ upgrading of urban informal settlements should be seen, in combination with other approaches, as an important component of national housing policy (UF 1989a, 11).  A number of reasons were put forward as motivating the selection of Besters as a  169  demonstration project. First, it was argued to be 'typical' of shack settlements around the Durban metropolitan area and was consequently highly representative. Second, it was located in a high profile locale being thefirstsettlement visible when entering the Inanda conurbation. Third, minimum health and safety requirements of settlement residents had not been met. Fourth, the area was well located in relation to major transport infrastructure and a major industrial park in Phoenix - both of which would impact positively on the settlement's longer term economic viability.  Also, Besters was located within the boundaries of the Durban municipal area which was seen as a positive element given both Durban's strongfinancialstatus, as well as its stated (although not yet acted upon) willingness to become involved in development initiatives in the informal settlements on its boundaries. In addition, other provincial and national government initiatives such as the Natal Provincial Administration's expression of interest in the area, as well as the RKDP's (RSA-KwaZulu-Development-Project) proposed sites-and-services and upgrading initiative adjacent to Besters, were seen as providing opportunities to influence these other major actors in the policy arena.  5.2.3 Principles underpinning the initial relationship between the UFISD and Besters residents  "If the communication gap between planner and client is to be closed, a continuing series of personal and primarily verbal transactions between them is needed, through which processed knowledge is fused with personal knowledge and both are fused with action" (Friedmann 1973, 177). Initial contact between the UFISD and Besters Camp residents was made in December 1989, at a time when political activity and therefore unhindered access to the resident population  170  was restricted under the State of Emergency regulations applicable at the time. Also, with the regime changefromInkatha warlordism to control by the 'comrades' still underway in some parts of Besters Camp, sporadic incidences of violent conflict were not uncommon. Contact was thus confined to low-profile, informal meetings with key political individuals within the settlement in an attempt to assess the political viability of a development initiative. By early 1990, the 'comrades' had established effective control of Besters, Nhlungwane and Ezimangweni.  The UFISD engaged in a nine-month dialogue during late 1989 and 1990 during which project principles were debated, workshopped and agreed upon, as were various technical options and tradeoffs. Initial contacts with leadership individuals then facilitated broader interaction with Area Committees, which in turn facilitated a series of mass public meetings variously attended by between 100 and 1000 people. This consultation provided the basis for the formulation of the initial project funding proposal.  The context for the relationship between the UFISD and Besters Camp residents in late 1989 and early 1990 was one where the internal self-identity of sub-community areas was extremely parochial, a consequence of which was that horizontal links between areas were very weak, and in many instances non-existent. Furthermore, given thefracturedand embryonic regional civic movement in the area, links between local structures in Besters and broader civic and political structures remained tentative.  The 'DCs' which were the dominant organizational structures at the time had little potential for growing into structures suitable to guide a development process. The strategy embarked upon by the UFISD was thus as follows. First, it continued to insist on the  171  establishment of 'representative local development committees' which - much like the Development Committees described in the Olaleye-Iponri upgrade in Lagos, Nigeria (MakinwaAdebusoye 1988, 165) - would serve as the liaison point for the UFISD, and which would be responsible for representing the views of Besters residents in the development. In this process, UFISDfieldworkerswho were experienced in civic organization in other Durban townships, became involved in facilitating the establishment of such committees.  By September 1990, Local Development Committees had been set up in the seven subareas within Besters: Area 1, Area 4 ( W  - named after the contiguous Unit M of KwaMashu),  Area 2, Mzomusha, Area 3 ("Etshenr - meaning 'place of the big rock'), Area 10 ("Dalmeny" named after the original 'Dalmeny Farm' in Inanda), Area 8 ("Nsimbinf - meaning 'place of steel or metal'). There were in addition, a number of functionally organized committees, organized around, for example, water, and health and welfare. These structures exhibited widely varying degrees of representativeness, and therefore of effectiveness.  Second, and notwithstanding strong pressurefrompolitical organizations to the contrary, a clear separation was made between party political decision-making structures in the first instance, and 'development' (or project) decision-making structures in the second. The reasons  underpinning this strategy was provided by the political context characterised by political fluidity and volatility at a national level, compounded by acute tension and continuing violent conflict between Inkatha and the ANC at the provincial and local settlement level during the 1989/1990 period. Furthermore, the organizational weakness, and the lack of representativeness and legitimacy of umbrella civic structures such as the Inanda Civic Association at that time were interpreted to support the decision to facilitate the building of organizational structures from-the-  172  ground up, rather than link into weak higher level structures.  Third, and alluded to above, the UFISD attempted to avoid the 'us' and 'them' mind set typical of the relationship between development agent and residents on many development projects, and rather to build a relationship with settlement residents that was more reflective of a 'partnership' arrangement. If any adversarial relationship was to exist, it was preferred that this be such that UFISD and Besters Development Committees were on one side, and that funding agencies and local authorities were on the 'other' side of the proverbial table. This distinguished Besters Camp from many other international upgrades that were implemented by government agencies (such as KIP, Chaani, Tondo, Lahore, Old Naledi) or special project agencies linked to government (eg. Ismalia, Lusaka), insofar as an attempt was made to retain control of the project within the boundaries of the settlement itself, and thereby to avoid the possibility of externally located power cliques or organisations exerting control from a distance. All external actors (indicated by shaded blocks in Figure 5.3 below), including local government, political organizations, civic organizations, and funding agencies were to be involved in the project, albeit via informal communication relationships that were separatefromthe project's development decision-making structures (unshaded blocks in Figure 5.3 below).  The participation process was thus structured to function both at the level of relatively formal interaction between organisational entities (UFISD and various Development Committee structures within the project) at one level, and informal day-to-day interaction between UFISD staff and settlement residents which required the integration of UFISD field staff into day to day settlement activities at a second level. This attempt to integrate local knowledge into the planning process involved six layers of interaction between the UFISD and Besters residents: informal  I  Z  05  z  vT LU  Zrv  tT0  IS  2 0  ANC  Sol  cfcz no:  \ J UJ  OCS:LU  %8  0  <D E 5 o  5oo E  si *8  (QSIdfl) NOISIAia SlN3l/M31ii3S  !VIAId<bdNI Local Development Committee  4  z  o  CO Local. Development Committee  Area 3 (Etsheni)  Local Development Committee  Area 4 (MM)  ill  Area 2  Area 10 (Dalmeny)  Local Development Committee  Mzomusha  Local Development Committee  si  Area 1  Local Development Committee  Area 8 (NslmblnD  173  ^e «s  LU  < O 0  ON ON  s  to  1  !  ••c  P  NOUVaNflOd NVadfl  z  391  1  -JOT  3  m  P  18  I  174  liaison with individuals and households, local sub-area Development Committees, the Coordinated Development Cornmittee, functional committees (Water, Health & Welfare, Employment), mass public meetings, and workshops.  5.2.4 Incorporation of Nlungwane and parochial autonomy These principles were followed through in Nhlungwane and Ezimangweni, although were adapted to suit the specifics of those settlements. Members of the Nhlungwane Civic attended the meetings between the UFISD and Besters Development Committees in 1990 in informal 'observer' capacities. During this period, numerous request were made by Nhlungwane residents for that settlement to be included in the upgrading project. The UFISD participated in a number of meetings with the Civic and mass public meetings at which the possibility of an upgrade was discussed.  Membership of the Nhlungwane Civic remained constant throughout the 1990 period, with a very low turnover of personnel. A wide range of interest groups had representation on the Civic, including women, youth groups, and seniors (both men and women). This Civic did not have any sub-committees responsible for functional areas such as water, health and welfare as in Besters, since responsibility and control of these specific functions remained vested in the Civic members themselves. Within the Civic were a number of very strong personalities which, while contributing to the strength of the Civic, also precluded any one individual or clique dominating the organization. Also, the Civic was able to coopt onto its membership residents who had construction backgrounds as workers, contractors and builders, thereby making technical discussions between the UFISD and the Civic a more equal exchange than was the case in Besters or Ezimangweni.  175  Throughout the project, there were attempts by members of the ANC and SACP (South African Communist Party) as well as of Civic organizations in neighbouring KwaMashu, to gain control of the Nhlungwane Civic. These struggles were however conducted on a sporadic and interpersonal level rather than on a formal organizational level, and in its resistance to these attempts at controlling it, the Nhlungwane Civic developed a modus operandi that was quite different to Besters and Ezimangweni, and which therefore implied a different relationship with the UFISD. For the first three years of the project, the Nhlungwane Civic had no formal relationship with the Inanda Civic Association, nor to any other umbrella civic structures. The very vocal resistance to interventions by organizationsfromoutside the settlement was frequently and explicitly articulated with reference to an aversion to being controlled by external power groupings.  This stance had a number of implications for the decision-making style in this portion of the settlement. First, because the primary direction of accountability of the Civic was claimed to be with the residents of the settlement rather than to any umbrella organization, the Nhlungwane Civic was able to remain more focused on the settlement locality level than on external politics. Second, this stance translated into a very parochial orientation which was even taken to the extreme of resisting any formal organizational relationship or structured communication with the adjacent settlements of Besters and Ezimangweni - which required that the UFISD engage each settlement separately. Third, the principle of autonomy was also reflected in a strong working relationship with the UFISD, but where sufficient distance was maintained so as not to place their primary accountability to Nhlungwane's residents in question.  Regarding the subject of parochialism, there was a pervasive tendency not only on the part  176  of Nhlungwane insisting that it be treated as a project in its own right (ie. as distinctfromthe Besters and Ezimangweni areas), but also of the fifteen sub-areas within the overall Besters Camp project consistently arguing that they be planned, negotiated with, and upgraded independently of other sub-areas. UFISD decisions as to the degree to which such securalism should be accommodated in the structuring of the project were informed by trade-offs associated in the first instance with the obvious inefficiency implications of sub-areas not allowing contractors from other areas to be involved in the installation of infrastructure in their areas, but in the second instance also by the longer-term opportunities for local settlement management that built on potentially strong local level organisation.  5.2.5 Ezimangweni and limited effective organization  Ezimangweni, by contrast, was considerably morefracturedin its social makeup. During 1990, the area was controlled by a successive cliques which were characterised more by their lack of accountability and other misdemeanours many of which were associated with a lack of sobriety, than by any constructive inputs into the upgrading process. Late in 1990, a group whose rhetoric suggested familiarity with the language and principles (although not always of the practices) of the 'Mass Democratic Movement' took control of the settlement. This change of control marked the emergence of the Ezimangweni Civic which, with the assistance of a Durbanbased NGO - the Centre for Community and Labour Studies (CCLS) - was 'launched' and subsequently affiliated to a national civic umbrella organization, the South African National Civic Congress (SANCO).  In contrast to Ezimangweni Civic and its formidable organizational affiliations that existed on paper, the organization in reality comprised no more than three resourceful activists who were  177  able to draw on the technological and human resources of their NGO partner, thereby also lending it (political) weight and connections. At this time, Ezimangweni was also not yet part of the upgrade discussion or proposal formulation. However, as with Nhlungwane, some residents of the area informally attended Besters Development Committee meetings as observers, following which the UFISD was invited to address meetings in the area. Following a number of meetings with the Ezimangweni Civic and public meetings, and broad agreement on the principles that would inform the upgrading initiative, Ezimangweni was also included in the Besters Camp funding application.  It was thus in the context of this very unequal relationship between interest groups that upgrading service options were discussed. Although having been identified as a problem, and partially addressed through an intensive participation process, the capacity mismatch between residents and UFISD was not adequately addressed given that no programmes explicitly designed to address the capacity deficits of residents or the embryonic organizational structures were put in place. Not only was the UFISD staffed by experienced professionals with extensive technical, financial and political/social skills, but it was also able to drawfromthe well-resourced national UF organization, as well as consultants when needed. By contrast, compounding the negligible technical and organizational capacity of settlement residents, was the limited external support base or resources upon which they could draw.  5.2.6 Separation of de jure authorities from project decision-making  The UFISD attempted to develop strong relationships with legal authorities, but for those relationships to be independentfromproject decision-making structures. Advantages to such an approach were that the project management team remained small and efficient, and it was possible  178  to thereby avoid over-bureaucratizing the project, while still being able to draw upon the resources of authorities and retain the potential for influencing city-level policy. Effective control over project decision-making thus remained autonomousfromgovernment, located at the level of the project. Also, by mamtaining constant contact, politicians and senior officials were satisfied that they were involved in the project sufficiently to vet or influence technical decisions, whilst neither having to allocate excessive time or resources, nor assume any responsibility for what many perceived to be a novel and risky project.  Regarding the KwaZulu government, Department of Development Aid (DDA) and the Natal Provincial Administration (NPA), the approach taken was to establish 'Steering Committees' that met on a regular (bi-monthly) basis, and served to update and monitor progress on the project, thereby satisfying those authorities that they were part of shaping the overall process. Also, for as long as politicians were satisfied that their technical staff were overseeing the process, their interventions remained negligible. These meetings occasionally addressed anticipated problems, although specific technical problems were ordinarily addressed through dayto-day contact between project staff and government officials. In establishing a separate layer of interaction with government and local government officials, project decision-making control remained vested with the UFISD and the Besters Development Committees.  5.2.7 Relationship with funding agencies: Durban City Council and the DDT  Project financing was secured in three increments. First, the Durban City Council allocated Rl 80,000 for a rudimentary water supply system within the KwaZulu portion of Besters in November 1989. Although application for this funding had been made independently of the UFISD initiative in mid-1989, its availability was coincidental with Development Committees  179  attempting to establish themselves in Besters.  This initial funding allocation provided a boost to embryonic Development Committee structures in that they were able to take control of aspects of decision-making around water services delivery which commenced in February 1990, and were able to build legitimacy by taking credit for 'delivering' water which was a priority service. Also, future funding applications were able to point to the demonstrated ability to deliver, following rapid installation of the infrastructure via a labour based process and in which Besters residents set up Water Committees which participated in decisions regarding water point location and management systems.  Second, in August 1990, the City of Durban allocated grant finance for the upgrading of Besters, at which time the IDT emerged as a possible joint funder. In anticipation of the formal launch of the non-partisan Independent Development Trust (IDT), and the search for a politically suitable project to present at the formal launch of the IDT, Besters emerged as a possible candidate for IDT funding. The possibility of the Besters upgrade was presented, separately, to all the major political interest groups including senior members of the ANC, Inkatha and the KwaZulu Government including Chief Buthulezi, King Zwelethini, and the KwaZulu Cabinet, senior Natal Provincial Administration officials, the Durban City Council's Management Committee, and the IDT's Trustees. On 24 August, the DCC announced that itself and the IDT were to jointly fund the project, with each contributing 50 per cent of what then amounted to an Rl 1.3 million budget.  This announcement by the City of Durban also marked a significant shift in their informal settlements policy. Whereas Durban's 1987 policy statement had been to implicitly accept the  180  permanence of some settlements given that 'suitable' land had not been found for their relocation, it had not taken any proactive or positive action either in terms of formulating or implementing a policy position with respect to development interventions. The 20 August 1990 resolution however indicated both policy and financial support for the upgrading of informal settlements both within, and outside its municipal boundaries. In addition to Besters, the Durban City Council indicated its intention to upgrade the Bottlebrush and Kennedy Road settlements, both of which were located within the municipal boundaries. Furthermore, the DFR (Durban Functional Region) Development Account was created. This fund was set up as an attempt by the Durban City Council to contribute to the economic upliftment of the metropolitan region, without imposing a direct cost on Durban's ratepayers and without eroding the existing base of the city's financial reserves (DCC 1991, 5). The funds were drawn solelyfromthe interest accruing on investments, which will be discussed in more detail later.  At that time, the proposed project boundaries included only the Besters portions of the settlement located within the Durban and KwaZulu boundaries. Although the finance was to be provided on a grant basis, it was agreed with the UFISD project management that settlement residents should make "afinancialcontribution" to the costs of service delivery, and that this amount could then be channelled back into the upgrading project. This principle was agreed upon at Development Committee meetings, with the precise amount and the future use of collected funds, still to be resolved. Having installed water points in Besters, the upgrade was able to continue with the planning and construction of roads, lanes and footpaths, as well as the stormwater drainage system in thefirstphases.  Following requestsfromthe adjacent settlements of Nhlungwane and Ezimangweni that  181  they be included in the upgrade, an expanded project funding proposal for Besters Camp was submitted to the IDT in December 1990, while service delivery continued in the first phases of the Besters sub-area. The expanded upgrade proposal thus included the sub-areas of Besters, Nhlungwane and Ezimangweni, which together accommodated approximately 50,000 people. In September 1991, the IDT approved R60.75 million for the 8,100 site upgrade. Importantly however, the conditions in terms of which the funding was to be made available changedfromthe grant finance mechanism in terms of which drawdowns had been made until then, to the IDT's new Capital Subsidy Scheme - a change which is seen to have made a significant impact on the ultimate project outcome.  5.3 T H E UPGRADE AS CONCEPTUALIZED AND PLANNED 5.3.1 Project components: basic health and safety, consolidation  The UFISDs 1990 upgrade proposals drew bothfromthe UFs (UF 1989a, 1989b) strategicframework,andfromthe UFISD's participation process with Besters Camp residents. Priorities as expressed by residents at public meetings were focused upon the provision of basic services such as water, improved shelter, electricity, sanitation, drainage and stormwater control, and garbage removal. Regarding public goods, paved roads consistently ranked highest in expressed need, followed by health and education facilities, and public lighting. These expressed priorities were confirmed in various survey responses (UFISD 1990b; RSA 1989b; CHESS 1993).  The Besters Camp projectframeworkthat was described in the project funding proposal thus had two components: first, the securing of Basic Health and Safety, and second, Consolidation. The Basic Health and Safety component focussed on the provision of rudimentary  182  infrastmcture - water, sanitation, access roads, drainage, refuse removal, and electricity - and was designed to arrest further downgrading of the settlement. Running costs were to be recovered from the ongoing consumption of services, while at least a portion of the capital costs were to be subsidized in line with residents' affordability levels.  'Consolidation' as it was defined in the UFISD'sframework,built upon and extended the conventional understanding of consolidation as used in housing literature which refers to the process by which households incrementally improve their homes over time. On the Besters Camp project, consolidation referred not only to activities that were designed to create a momentum whereby households would continue to improve their dwellings (private consolidation), but also whereby the overall settlement was to be improved over time (public consolidation). More specifically, it was intended that residents were to be assisted in improving their shelter as well as developing a full range of social services in the settlement. It was envisaged that consolidation would involve the provision of tenure, access to small loans for home improvement, the provision of building materials and advice, training programmes, the employment of local labour in the construction process, and the investment in schools, public halls and resource centres (UFISD 1990a).  As described by the UFISD (1990a, 12-13), the project principles underpinning these interventions were as follows: • permanence: given the hostility of the state to informal settlements, the project was intended to establish acceptance of the principle that informal settlements were not a temporary or transitory phenomenon, but were rather a permanent part of the Durban metropolitan region which, given appropriate interventions, could be developed to become 'liveable' environments  183  over time. • replicability: it was argued that the upgrading process initiated in Besters Camp should be replicable in informal settlements elsewhere. This was understood to refer particularly to economic viability and replicability. • upgradeability: initial interventions needed to be designed so as to be upgradeable in the future, with the ultimate objective of the informal settlements being integrated into the formally established townships surrounding Besters Camp. • sustainability: it was argued that the upgrading process needed to create sufficient capacity and a momentum such that upgrading would be ongoing subsequent to the disengagement of the UFISD and other development agents, and would continue into the longer-term. • empowerment: upgrading of the Besters Camp settlement was to be undertaken in a way that was to explicitly promote the capacity of the residents to improve their circumstances through involvement at all levels in the planning and implementation of the project and the processes in which they have a vested interest.  5.3.2 Planning principles  Concurrent with the initial contact with settlement residents, a planning consultancy was briefed by the UFISD to prepare preliminary planning proposals for the upgrading of Besters settlement, comprising both the portions within Durban and KwaZulu. The initial brief and the proposals that were formulated were noteworthy in their attempt to conform to legal jurisdictional boundaries, and planning regulations applicable within the respective jurisdictions. Given that each jurisdiction had its own planning statutes, approval procedures and standards, this would have implied significantly different approaches to project decision-making, public participation, layout plarining, and standards applied in each of the respective jurisdictions. Such an approach  184  would have been consistent with previous major planning initiatives in Inanda, none of which were ever fully implemented, but which included the Inanda Development Framework Report (DDA 1990) which was an extensive two year exercise by a consortium offiveplanning consultancies, as well as the earlier Structure Plan for Inanda (Home Glasson 1982), and a proposal commissioned by the UF in 1982 entitled Upgrading Inanda (UF 1982).  The initialfieldworkprocess indicated quite clearly that even the initially conceived project boundaries (based on cadastral boundaries) bore no resemblance to boundaries as defined by residents themselves. First, most residents neither knew nor cared where exactly Durban/ KwaZulu/ Released Area 33 boundaries were - a point clearly evidenced not only by resident defined boundaries of sub-areas criss-crossing over legal boundaries (see Table 5.2 and Figure 5.2), but also with numerous individual sites and even some shacks being bisected by legal jurisdictional boundaries. Furthermore, the informal land transfer market took no account of formal registration of title or transfer of ownership, and was based on de facto site occupation which was perceived to translate into de facto ownership for site occupants.  Evidence in Besters Camp was consistent with studies in a number of Third World cities where, as was noted earlier, "increased de facto tenure security appears to reduce the need of individual squatters for official legal titles. Once tenure at the community level appears to be secure, titles are not essential and the willingness to pay for them is considerably weakened" (Angel 1983,132; also Taylor 1987, 55). Moreover, the majority of households in Besters Camp (72 per cent), when questioned on their tenure status, indicated that they believed that they had legal tenure over the land that they occupied - this belief being grounded in the payments made to warlords for 'ownership' of sites during the post-1985 period. A similar percentage (73 per  185  cent) believed that their occupational rights were sufficient to allow them to remain on their sites (UFISD 1990b). jurisdiction  number of shacks  population  Area 1  419  3352  Durban  Area 2  767  6136  Durban/KwaMashu  Area 3 (Etsheni)  396  3168  KwaMashu/Ntuzuma  Area 4 (MM)  239  1912  KwaMashu  Mzomusha  1,265  10120  Dbn/RA33  Area 10 (Dalmeny)  458  3664  Ntuzuma/KwaMashu/RA3 3  Area 8 (Nsimbini)  63  504  Ntuzuma/KwaMashu  Area 1  351  2808  RA33  Area 2  135  1080  RA33  Area 3  136  1088  RA33  Area 4  186  1488  RA 33/Ntuzuma  Area 5 (Piesang)  260  2080  RA 33/Ntuzuma  Spola  33  264  Ntuzuma  Area 1  776  6208  RA33  Area 2  609  4872  RA33  Area 3  477  3816  RA33  6570  52560  Area BESTERS  NHLUNGWANE  EZIMANGWENI  TOTAL Table 5.2  Resident-defined areas in relation to legal jurisdictions  A strategic choice thus had to be made on the project: either to adhere to the legally  186  prescribed system, and thereby ignore the informal land market already existing in the settlement, or alternatively to take the settlement's de facto realities as the starting point, and to resolve legalities as the project progressed.  Following numerous public meetings and workshops discussing the subjects of planning, survey, and tenure delivery, the approach proposed was one which took the de facto characteristics of the settlement as starting point, and worked toward an end-state that was acceptable in legal terms. This was the reverse of the conventional route of starting with statutory planning frameworks and procedures and attempting to 'fit' the plarining problem into those frameworks.  As indicated in Figure 5.4 below, this meant using resident-defined boundaries as the primary organising mechanism, and then working toward a rationalisation of boundaries as well as formulating a uniform set of planning procedures and regulations across jurisdictions. Thus, whilst the delivery of security was considered an integral aspect of the upgrading process, an attempt was made to provide for the delivery of a form of title that took into account the existence of the informal market, as well as the positives of formal legal title. This was consistent with international experience which indicated that while legality is an important consideration, it is important that theform of legality be appropriate to the context in which it is to be applied.  The planning approach that emerged out of thefirsttwo years of the Besters Camp project was thus underpinned by the following principles: • planning needs literally to be conductedfrom-the-ground-up.The argument was that planning should not be a desk-top exercise which is then imposed upon the settlement. Given that the  187  situation on the ground is not fixed, but is constantly changing (both socially and physically), the planning process needs to be able to accommodate such changes on an ongoing basis; • planning needs to be conducted with settlement residents involvement at all levels ranging from making decisions on planning principles, through to fixing the alignments and locations of service infrastructure, so as to marry the technically optimal with the socially optimal solutions; • planning should pursued as an iterative process such that the 'ideal end state' evolvesfromthe process, rather than being an initial imposition on that process. Layout plans, for example, should be arrived at toward the later stages of implementation in the form of 'as-built' plans, rather than preceding the upgrading process as Master Planning would dictate; • planning is only one element of what needs to be a multi-disciplinary approach to the upgrading process. Furthermore, planning needs to be inseparablefrom,and continue throughout, the implementation process.  5.3.3 Planning procedures  So how were these principles to be translated into practice? The proposed planning and tenure delivery process proposed was underpinned by an acceptance that it would be necessary to provide a tenure option that had registerable legal status. In addition, it was argued that there should be sufficient flexibility to allow a transitionfromthe informal land market in which ownership was determined on a de facto site occupation basis, through to an 'ideal end state' where conventional title was possible. The planning process thus traversed the following stages:  Step one: Status quo household definition: the primary objective here was to record, in detail,  the de facto situation as it existed in the settlement. On an area by area basis, detailed site and household information was recorded on plan and then computerised, thereby providing a data  188  HOUSEHOLD DEFINITION  •  De facto settlement characteristics recorded on plan  •  On-site walking of proposed service locations and alignments with residents  •  Residents and local authority approval of detail design  •  Defines 'blocks'  •  Dedensification  •  Final layout  •  Township establishment  ACCESS PLAN  SERVICE CONSTRUCTION  SITE SUBDIVISION LAYOUT  TENURE DELIVERY  Figure 5.4  Conceptual description of planning procedure  base which could easily be updated on site. This included details of the house number, the name of household head, tenure status of dwelling occupants (owner, tenant), and the number of people living in each dwelling. A description of the type of dwelling construction, building materials, problems requiring attention (eg. damp penetration, cracking walls, sagging roof), and an assessment of the quality or condition of the shelter were recorded. Also, details marked on plan  189  included residents' own definition of their existing site boundaries, the orientation of dwelling and other structures on site, the alignment of roads, tracks, footpaths, and the alignment of existing stormwater routes. During this stage, the resident population had no legal title, but enjoyed de facto security on the basis of site occupation.  Step two: 'Blocking': The outer limits of blocks were defined by roads, lanes and footpaths similar to the principle employed in the blocking in the Hai El Salam part of the Ismalia upgrade in Egypt (Davidson and Payne 1983,133-5). Blocks in Besters Camp however accommodated a smaller number of structures - an average of between only 20 and 30 shacks, which was considered to be a workable number in administrative terms and in the negotiations to be conducted between block residents around site boundaries, tenure forms and area management.  The block definition process distinguished between 'fixed' and 'flexible' elements. The primary fixed elements were roads, lanes and footpaths in the first instance, and public facility sites in the second. The flexible elements, during this stage at least, were individual residential site boundaries. Considering each of these in turn:  •  The circulatory system. The alignments of roads, lanes and footpaths placed a strong  emphasis on pedestrianised access. Linked to these were service run alignments, the location of water points, and the location of garbage collection points. The pre-existing system of tracks and paths provided the basis for fixing the alignment of the circulatory infrastructure. Taking existing paths as the starting point, UFISD staff and Development Committee members designed the circulation layout by walking along proposed alignments.  190  On-site walking along alternative alignments with area residents provided a layout that was more sensitive to the existing shacks than would have been the case with an off-site design, particularly given the high densities and steep gradients. This circulation system provided the basis of the organizing concept of'blocks' which comprised, at a maximum, rows of shacks no more than four rows deep, which would enable a panhandling option for sites opting for individual site access at a later stage. The preliminary plan did not indicate individual site boundaries, but rather defined 'blocks' of shacks, the outer limits of which were defined by roads, lanes or footpaths.  Regarding individual site definition within the blocks, there were two main categories of sites: public facility sites, and residential sites. • Public facility sites. The number of sites was constrained by the extremely limited space availability both within and adjacent to Besters Camp. Given that available space for public facilities within the settlement was considerably less than required by conventional town planning standards, the proposed layout made provision for major facilities such as schools to be provided on adjacent floodplain land which was to be raised. Within the settlement, minor facilities were provided and were multi-functional in their design so as to increase usage efficiency. These included meeting halls, in addition to space with specific designated uses such as creches, and health care facilities. Since these minor facilities were located within the settlement, the number of sites and the site sizes were determined by space availability between existing shacks, given the aversion to displacing existing private dwellings over the creation of public facility sites. Once demarcated, these sites were also considered to be 'fixed' elements of the layout.  • Individual household site boundaries. Taking the site boundaries recorded in the status quo  191  plan as the starting point, the approach was taken not to specify minimum site sizes, but given the steepness of much of the land, rather to prescribe minimum usable site space. This was planned such that at the minimum, it would be feasible to construct a dwelling with a 40 square metre floor area and a VJJP on sites that conformed as closely as possible to original de facto site boundaries.  The fixing of site boundaries was avoided in the early stages of the upgrade. This allowed flexibility in terms of continued site transfers, boundary adjustments between neighbours, and allowed people to continue to gain access to their shacks by walking across neighbours' sites as in an omnibus servitude. Importantly, this also allowed for the complex negotiations concerning dedensification to be concluded before any site boundaries were fixed.  One of the most important criteria informing the planning of the blocking process in Besters Camp was the lack of available vacant land which offered the possibility of resettlement to alternative sites. Whereas the 'reblocking' in the Tondo upgrade in Manila implied the resettlement of approximately 25 per cent of the total number of settlement households to alternative greenfields sites (Laquian 1983a, 22-23), this was not an option in Besters Camp. Both the extent and the alignments offixedelements - the circulatory network and the public facilities - were therefore not limited only by budget constraints, but also by the space-constrained imperative to keep relocation within the limit of the 400 available overspill sites.  End state: legal title: As required by the Capital Subsidy Scheme, it was necessary to survey individual sites and deliver conventional legal title to households. At this point, individual site boundaries werefinalizedand legal tenure delivered as contemplated by the Less Formal  192  Township Establishment Act (LFTE) for the parts of the settlement within Durban and Released Area 33, and Proclamation 293 in KwaZulu.  Given the absence of direct 'intervention' by external agents and organizations in the settlement, the UFISD and residents were able to plan, make decisions, and implement those decisions within the broader principles already agreed upon, but without having to constantly make reference to legal procedures, and official documents or relationships. Planning was defined by the rules of the game as they were understood to obtain in the settlement, rather than the rules applicable in the formal city which were defined by conventional planning standards and procedures.  Importantly, the lack of local or national precedent provided Besters Camp with considerable room for manoeuvre in the formulation of the planning approach to be applied on the project. The Durban City Council, in whose jurisdiction the first phases of project implementation commenced, were supportive of this approach, concurring that the planning process needed to make sense on the ground, which, if necessary, could involve the incremental development of planning standards and regulations as the upgrade progressed. Whilst this process of making decisions by a combination of 'intuition, experience, rules of thumb, and an almost endless series of consultations' is typical of an incrementalist approach to planning as described by Lindblom (1973), the approach in Besters Camp but was also grounded in understanding of power relations characterising the settlement. Along these lines, and emerging out of the post-1990 participation process, sets of planning procedures and regulations were developed as alternatives to those prescribed in the City's legal regulations. What was at issue here was not legality per se, but rather an attempt to develop an appropriate and amended form  193  of legal regulation, which was able to integrate aspects of the pre-existing de facto system into the planing process.  194  CHAPTER SIX L O C A L SETTLEMENT-LEVEL PROJECT IMPACTS  Drawing from UFISD project data, chapter 6 considers local level impacts of what the project delivered. In addition, via an examination of the primary institutions involved in the upgrade and their relationships with each other, it examines how project decision-making evolved throughout the 1990 - 1995 period.  6.1  SERVICE DELIVERY  Thefirstcomponent of the Besters Camp upgrade including physical service installation designed to secure basic health and safety commenced in 1990, and was completed in October 1995. Consolidation activities such as the construction of a limited number of basic public facilities, the delivery of legal tenure, facilitating the involvement of external support agencies from both government and non-governmental sectors, and the formulation of local regulation and planning systems were also completed during the initial upgrade.  6.1.1 Physical infrastructure: securing the conditions for basic health and safety  The upgrade clearly resulted in improved access to physical services for settlement residents. While efficiency problems sometimes imply cost overruns and implementation delays, upgrades are generally effective in the delivery of the intended quantity and quality of physical product (Skinner et al. 1987). Besters Camp is seen to be typical in this regard.  At the most basic level, reflecting a similar service delivery concept to the services provided in the KIP described in Silas (1992), the Besters Camp upgrade delivered rudimentary  195  physical services including a water reticulation system, a pedestrianised circulation system with limited vehicular access, and a stormwater drainage system. In addition, a sanitation disposal system, garbage removal, public lighting, and private electricity connections were also provided (summarized in Table A3 in Appendix). As with any other upgrade, the extent of physical services delivered is a function of the extent and terms according to which project funding is made available. In the Besters Camp instance, the Capital Subsidy Scheme required the delivery of serviced sites within an average of R7,500 per site.  A circulation network was constructed, comprising roads, lanes, and footpaths. The 3.5 to 4.5 metre wide roads and the 2.5 metre wide lanes were located within variable road reserves so as to minimize displacement or disruption of existing shacks. Roads and lanes were designed so as to cater for low vehicular traffic volumes, primarily comprising service and emergency vehicles, as well as privately owned cars. Given that alignments closely followed original track alignments, the reduced horizontal and vertical curves had the effect of reducing traffic speeds within what was essentially a pedestrianised system. As the major form of public space within the settlement, the circulation system also served as playing space for children, and gathering, spaces for adults. High road grades were also employed so as to achieve access to steeply sloping sites Roads and lanes were asphalted and the one-metre wide footpaths were concreted so as to reduce longer-term maintenance costs.  Given the high density of settlement and the high stormwater runoff volumes, a stormwater drainage network was incorporated into the road, lane and footpath system. Erosion protection, comprising lining of a total of 12.3 kilometres of stormwater drains, were used to reduce long-term maintenance costs and avoid property damage. In addition to piping some  196  stormwater drains, major drainage routes for above surface stormwater flow were created so as to reduce storm damage in the event of pipe blockages given the likely low levels of future maintenance.  A total of 41 water points were evenly distributed throughout the settlement. The system comprised water kiosks and neighbourhood standpipes at an average of 1 water point per 170 sites. Washing slabs and basins incorporating sullage drains were designed on-site with users and were provided at all water kiosks.  At the outset of the project, massive accumulated piles of garbage existed throughout the settlement, in addition to an old 2 kilometre long irrigation canal, and two agricultural dams which were full of stagnant water and garbage. As part of the bulk earthworks operation, refuse was removed to the nearest solid waste site in Durban and the dams drained andfilledwith soil, to create useable spaces for future public facilities. Garbage skips to be emptied by the DCC were distributed evenly throughout the settlement, at an average ratio of one per 205 households.  Following a geotechnical assessment, and given both the absence of trunk sewers in the vicinity of the settlement as well as the limited bulk water supply, site specific Ventilated Improved Pit (VIP) latrine toilets with a design capacity that required emptying every 5-6 years, were constructed on each site. Also on an individual site basis, each shack was provided with an electricity connection, comprising plug points for appliances, and an electric hot plate for cooking.  Notable when compared to eighty-seven other IDT projects throughout South Africa, is  197  the lower proportion of the project budget spent on services in Besters Camp, and the higher portion on project management. This can at least partially be accounted for by the greater political and technical complexity of in-situ upgrading, relative to greenfields sites and services projects which comprised the majority of IDT projects.  87 IDT projects Besters % of project costs % of project costs  Item  5.4  Land  3.8  13.0 Consultants fees 3 . 6 Township establishment Conveyancing 3.2 Engineering 6.2  12.0  61.3  50.5  0  1.4  Service delivery Water Sanitation Roads Electricity Garbage removal  11.5 29.2 16.8 3.8  Public facilities Project management overheads  11.2  17.1  Sub-total  90.9  84.8  9.1  10.7 1.8  Value Added Tax (VAT) Finance charges  2.7  Escalation Total  100  100  Table 6.1 Comparison ofproject cost breakdown between Besters Camp, and 87 other Capital Subsidy Scheme projects (source Mabin et al. 1994, 30; UFISD project data).  Service delivery implied very limited disruption of the settlement. Only 3 shacks (0.04 per cent of the total households) were relocated as a direct consequence of the installation of  198 infrastructure, the preference being to plan roads, lanes, and footpaths along the winding alignments of existing tracks. This flexibility was in contrast to the rigidity of City Engineers Departments in other upgrades - such as on parts of the Lusaka upgrade who insisted on straight service runs (Pasteur 1979, 74) - an implication of which was more widespread disruption of shelters.  In instances where infrastructure installation unavoidably affected specific shacks, the preference was to negotiate on-site solutions with households affected by the construction process. This involved the construction of retaining walls, the removal of room(s) on one side of a house and the construction of replacement room(s) on another part of the site or shack, which is very similar to the process used in the Tondo upgrade.  The low level of resettlement in Besters Camp distinguishes it from most other upgrades recorded in the literature consulted. Van der Linden (1994) notes in a review of recent upgrading experience that one of the disappointing features of most upgrades remains the high level of displacement of households. While upgrades frequently state minimal displacement as a project objective, resettlement of original residents ranges along a continuum with the zero displacement of the KIP in Indonesia, through to the majority of projects with a much higher proportion of overspill sites-and-services components which, on average, range between one-quarter and onehalf of total project sites. For example, in the Tondo project in Manila, 75 per cent of households were part of the in-situ upgrade, while 25 per cent were relocated to sites and services developments (Laquian 1983 a, 22), while on the Chaani project in Mombasa, Kenya, the in-situ upgrading:sites and services ratio was 63:37 per cent respectively (Kayila 1986, 152). In Lusaka, the ratio between in-situ upgrading and greenfields overspill sites linked to each squatter complex  199  200  was 56:44 (Pasteur 1979, 162), and as high as 42:57 per cent respectively in the Lahore Walled City project in Lagos, Nigeria (Ahmed 1986, 49). By contrast, Besters Camp tends toward the end of the continuum with very limited resettlement, with only 3 out of the total 7000 households being relocated as a direct consequence of service provision.  Sites with poor drainage, in areas subject to flooding, or those on unstable land were identified, and where feasible, were protected. Although provision was made for a potential total of 408 overspill sites contiguous to Besters, only half of these were immediately useable, the remainder being below a l-in-50 year flood line. These overspill sites were thus designated for households living on technically problematic sites where protection or stabilization was not possible, and for residents of the smallest sites (in the 30- 40 sq metre range) who wished to move, thereby assisting in the dedensification process. All households were relocated by negotiation, and were provided with a pre-built wattle and daub shelter (approximately 28 sq m in extent), a VIP, and an electricity connection on the overspill site. Once overspill sites are settled, this will take the upgrade: new sites ratio to 97:3 per cent respectively.  Regarding the upgradeabiliry of services, VIPs were designed to be upgradeable to a lowflush waterborne system, even though full waterborne sanitation was an unlikely option given the inadequacy of the bulk water supply serving the greater Inanda area. Also, aside from the disruption that would be implied by installing pipes to each site given the high densities of Besters Camp, the internal water reticulation system within the settlement was inadequate to serve a full waterborne sewerage system, and would thus also need to be entirely replaced in order to provide water points to individual sites. Upgrading of the circulation system is constrained by the topography and shack density, and although technically feasible, would imply considerable  201  expense and disruption of shacks in the settlement, thereby rendering it an unlikely future option.  6.1.2 Private consolidation: tenure delivery  Having purchased land occupied by informal settlementfrompublic and private landowners, the project delivered formal legal tenure to a total of 6,832 sites in Besters Camp, of which 5,179 were Certificate of Occupation, and 1,653 were Deed of Grant. Whilst tenure is generally acknowledged to be important in the consolidation process, it has been noted that it is not formal tenure per se, but the form or 'perception' of security of tenure that is most important. In many settlements, possession by occupation represents a form of concrete 'ownership', suggesting that de facto security is a primary tenure consideration for settlement residents. An opportunity missed in the Besters Camp project, and as argued a general level by Farvacque and McAuslan (1992), and in the sub-Saharan African context by Mabogunje (1992), was for the development of effective cadastre, land titling, and land registration systems to involve strategies that build on the de facto institutions and processes with which target populations are farniliar, and with which they are therefore most likely to participate actively and constructively. In upgrading projects in which a transitionfrominformal land transfer markets to formal land markets is implied by the delivery of legal tenure, the regularization process needs to integrate elements of the informal mechanisms into the process.  Although the planning procedure was sufficiently flexible to accommodate alternative tenure forms, the Capital Subsidy Scheme (CSS) requirement of the delivery of conventional legal tenure in Besters Camp represents an instance of attempting to formalize the land market, with no consideration of the existing informal system. Upon the flexibility and continued functioning of the informal land transfer market, a formal tenure system was superimposed involving site  202  boundaries being demarcated by pegs fixed in the ground, the location of which are recorded with surveyed precision and documented on cadastral plans in Deeds Registry Offices.  Therefore even though legal tenure was delivered (the surveying, conveyancing and sales administration costs of which amounted to R3.7 million, or 6 per cent of the project budget) the system is vulnerable to potential collapse uponfirstpoint of ownership transfer, since future ownership transfers being recorded at the Deeds Registry 80 kilometres away in Pietermaritzburg is highly improbable. Aside from the logistical inconvenience associated with the inaccessibility of the Deeds Registry, the expensive and complex registration procedures and transfer fees associated with pursuing the legal route is a furtherfinancialdisincentive to take cognisance of the legal system in respect of future transfers of site ownership.  While it is premature to assess the sustainability of the form of tenure delivered given that tenure in the last phases of the project was only delivered in early 1996, evidence from three of the sub-areas to which tenure was deliveredfirstwas that limited sales of sites had continued without ownership transfers being registered at the Deeds Registry. In three sub-areas surveyed in Besters, 5 such unregistered transfers were found in one sub area (Area 1), 2 in another (Area 3), and 1 in another (Area 2). While this amounts to less than 2 per cent of total households in those areas, this poses a longer term administration problem where legal site ownership records are likely to increasingly diverge from de facto ownership.  Also important is that in the few instances where sites were sold, selling prices ranged between R800 and R3,500 which is marginally higher than average pre-upgrade house selling prices - hardly keeping up with inflation rates - and in any event considerably lower than the value  203  suggested by the average subsidy of R7,500 to each site as per the Capital Subsidy System. Although too early to discern any patterns or form generalizable conclusions, this experience is consistent with other recently completed CSS projects where sites had also been sold for a small fraction of their Capital Subsidy value (Mabin et al. 1994, 86).  6.1.3 Private consolidation: shelter upgrading  The upgrade was not structured to include shelter upgrading given that the available finance precluded formal housing delivery. However, spontaneous upgrading of shelters was evident throughout the duration of the project. A comparison of project recordsfromthe household definition stage of the planning process, and a follow-up inspection in 1995 indicated that spontaneous upgrading had taken place on approximately 30 per cent of shacks throughout the settlement. Of these, approximately 210 shacks, or 3 per cent of the total shacks in the settlement exhibited significant improvements such as the addition of one or more rooms, with an average increase of 8 square metres in floor space. The remainder of improved shacks reflected more superficial improvements, mainly reflected in a change in quality of building materials such as the replacement of wattle and daub, cardboard, and corrugated iron walls with concrete block walls.  That some shelter improvement was evident both during and immediately after service and tenure delivery is consistent with international experience, suggesting that the greater the perception of security, the greater the willingness to invest in shelter upgrading. The extent of improvement in Besters Camp however tends toward the lower end of the spectrum when compared with other upgrades. In Tondo, Jiminez (1983, 75) indicates an increase in housing quality of between 30 and 40 per cent as little as three months after reblocking took place, which  204  is considerably higher, for example, than the nominal increase of 8 per cent in house values found in Cartagena, Columbia, during a five year period in the 1970s. While the extremely rapid improvements in Tondo were an outcome of the release of the pent-up investment potential following the reblocking, the low savings base in Besters Camp prior to the upgrade would have precluded any such rapid and widespread investment in shelter upgrading following either the delivery of services, or the delivery of tenure. In Besters Camp,fiirthermore,a high proportion households indicated a very strong perception of security even before the upgrade commenced, and so the delivery of services and tenure should be interpreted to have reinforced a pre-existing perception of security, father than create it where none previously existed.  Households were very reticent about divulging specific information such as amounts expended, and reasons or factors stimulating home upgrading. One of the reasons for households reluctance to divulge details of the shelter upgrading process relates to the primary source for materials for shelter upgrading during the 1990-1995 period being pilferagefrommaterials supplies for infrastructure installation. Materials such as sand, stone and cement, as well as concrete blocks used in the construction of toilets, footpaths, and some stormwater drains, were easily removedfromunprotected on-site storage points in between shacks. Engineering project budgets indicate pilferage of approximately R300,000 worth of these materials, all of which was ideally suited to shelter upgrading.  The most significant factor inhibiting households willingness to invest in shelter upgrading was however the widespread expectation of mass housing delivery associated with new political dispensation - and specifically the 'million houses infiveyears' commitment of the ANC. In terms of this committment, settlement residents clearly perceived responsibility for housing  205  delivery to lie with the state, and not with individual households themselves.  6.1.4 Public consolidation The upgrade involved the construction and operationalization of a limited number of public facilities including multi-functional halls, health and education facilities throughout the settlement. In Besters, a clinic was constructed in Area 2 and was operated by nursing staff from Durban's City Health Department. As part of the same building, a day care centre attended by 80 children on a daily basis, was built. The centre functioned on a user-pays basis since its inception, which covered all overheads and running costs, making it one of the few financially viable resident-run public facilities in the settlement. Also in Besters, a substantial brick building in Area 10 that was occupied by landowners during the pre-1985 period but destroyed in the August 1985 violence, was renovated into a multi-functional hall. Its use was administered by the local Development Committee and subsequently by the Civic, and was used variously for public meetings, as a temporary clinic, a day care centre during mornings, and as a classroom in evenings for the 70 adults registered in an adult literacy programme.  In Nhlungwane, the old abandoned and partially damaged Ramdarie mansion was renovated into a substantial public facility comprising a large hall with a stage, and 10 adjoining rooms which were intended to be used on a rotating basis by paying users. In addition, the Bhekokuhle Creche was attended daily by 16 children and was accommodated in a temporary prefabricated structure. In Ezimangweni, Jeena's store, also abandoned after being burnt in August 1985, was the focal point for a similar exercise of renovation and construction of a public hall, meeting rooms and offices. The construction of these facilities was intended to decentralize the focus of public facilities awayfromthe Besters Yard construction site at the entrance to the  206  settlement, and rather develop nodal points within each sub-area.  Funding had been committed and construction had commenced on these facilities before the implementation of the Capital Subsidy Scheme. However, in accordance with the requirements of the Capital Subsidy Scheme that physical infrastructure and tenure delivery be completedfirst,and if any residual remained then this could be allocated to consolidation, no other public facilities were funded by the project budget. This will be discussed in more detail in Chapter 7. Attention thus shifted to facilitating the involvement of other governmental and nongovernmental agencies in the consolidation process. Governmental agencies involved in the project included The National Council for Child and Family Welfare (NCCFW), St. Johns Ambulance, and the City of Durban's Parks and Recreation Department. Non-governmental and private agencies included the Centre for Community and Labour Studies (CCLS), the University of Natal's Community Services Training Project, the Centre for Health and Social Studies (CHESS), Planned Route into Science and Mathematics (PRISM), Training and Resources for Early Education (TREE), Sunflower Projects, School Leaver Opportunity Training, Turn Table Trust, and Read Educational Trust (READ). The involvement of these agencies is important given the gearing effect in respect of funds generated independently of the project budget.  6.1.5 Economic development and labour based construction  Labour-based construction methods, when properly managed, have been shown to achieve outputs that are competitive with conventional mechanized methods (Still 1991). As distinct from labour intensive construction methods which aim at maximizing employment, sometimes at the expense of efficiency, labour-based construction methods also maximize employment creation subject to it being economically as well as technically efficient. In addition, the process of  207  infrastructure design and materials procurement in labour based construction is oriented so as to make component manufacture andfinalconstruction more labourfriendly.It was on this basis that Besters Camp's service construction was designed and implemented.  Selected roads, stormwater drains and canals, lanes, footpaths, VIPs and the precasting of concrete slabs for VIP toilets, and the laying of water pipes were undertaken using labour based methods. Only the bulk earthworks and major roads that were technically or economically unfeasible to be undertaken by labour-based methods, were constructed by small works civil engineering contractors. Also, the construction of the clinic and daycare centre in Besters, as well as the renovation of partially destroyed buildings as public facilities in Nhlungwane and Ezimangweni, were all completed using locally trained contractors and labour.  A problem throughout the project occurred in respect of outside contractors who undertook tasks such as bulk earthworks and higher order roadwork that could not be efficiently performed by locally trained labour. These outsiders were the subject of continual harassment, robberies, muggings, and theft of vehicles and equipment. The resolution of the problem was to establish a group of "local community monitors", who in turn employed selected residents to 'protect' the contractors. Local strongmen were prominent not only in this 'protection' operation, but also in their influence over the recruitment of workers for labour-based work. An 'Employment Committee' was responsible for the selection of workers, the resolution of disputes between contractors and workers, and as mediators between contractors and project management - an industrial relations mechanism of sorts.  In total, 172 teams comprising over 900 residents of Besters, Mzomusha, Nhlungwane,  208  and Ezimangweni were employed in the labour-based construction process. These contractors and workers were paid approximately R15 million in wages during the upgrading project, which constituted 30 per cent of the infrastmcture budget. Aside from the short-term injection of funds into the settlement, more important in terms of building longer-term capacity was the development of a skills base deriving from the construction process.  The project aimed at developing a skills base such that contractors trained and employed on the upgrade would be able to establish economically viable contracting enterprises, and thereby ensure an ongoing source of income as a result of their involvement on the project. Specifically, skills were developed in construction procurement, contract administration, construction organization and planning, construction technology, and in the measurement and payment process.  While a technical skills base was put in place, no contractors established enterprises of their own. Among the reasons for this was that upon completion of their work on the Besters Camp project, many of the more proficient contractors were employed by formal contracting companies such as road construction companies, building operations, and some consultancies that required people with technical skills and who were also familiar with the internal norms of informal settlements. Second, the skills that were developed were adequate to build simple structures, but not to undertake larger projects at independent risk. Third, the parochialism evident in Besters Camp and its sub-areas was equally strong in other settlements, and so when contractors attempted to gain work in other settlements, they were not permitted to operate.  209  6.1.6 Cost recovery International experience indicates that one of the major weaknesses of upgrading projects have been poor cost recovery records. Although there are exceptions, the general experience of sites and services projects is of high levels of cost recovery. For example, in the 1970s and 1980s, cost recovery levels on sites-and-services projects rangedfrom63 per cent in Africa, to over 95 per cent in India and Indonesia, and 98 per cent in El Salvador (Williams 1984, 179, Pugh 1988, 37). The cost recovery experience in informal settlement in-situ upgrading is considerably less favourable (Skinner 1987). Cost recovery levels rangefroma generalized average cited by Pugh (1988, 38) of approximately 50 per cent, to 46 per cent recovery rates in the Madras upgrades (Pugh 1989,40), through to complete breakdown of cost recovery as in the Lusaka upgrade (Rakodi 1987, 364).  In-situ upgrades' poor cost recovery levels can be explained by a number of factors that extend beyond poor affordability levels. First, the nature of in-situ upgrading is that people are already living on site, are paying little or nothing toward the costs of accommodation and service consumption, and are often able to mobilize in response to external interventions that threaten the status quo. In addition, and as noted earlier, evidencefroma wide range of projects indicates problems arisingfrominadequate attention in project design to effective collection mechanisms and procedures, and the role played by residents in the cost recovery process.  Lessons learnedfrominternational projects in the 1970s and 1980s both in respect of the reasons for poor cost recovery, as well as alternative mechanisms designed to ensure payment, were not effectively applied in the Besters Camp project. In Besters Camp, the capital costs of infrastructure were fully subsidized, the replicability implications of which will be returned to  210  later. Notvvithstanding that cost recovery was thus confined to the relatively small component of monthly running costs of services consumed, the inability to recover these costs is one of the major failures of the project.  Average household incomefiguresfor Besters Camp indicate that service provision was comfortably within affordability levels. Average total monthly costs for all services amounted to R73 (Table A3 in Appendix), which was less than 10 per cent of average monthly household income: Taking into account the high income inelasticity of basic services such as water and energy for low income households, this is less than half of the 15-20 per cent of monthly income that lowest income households are generally able to pay for basic service consumption. Considering income distribution within the settlement, and assuming expenditure on services not exceeding 20 per cent of household income, only 15 per cent of all households in Besters Camp would have been unable to afford monthly service charges and would have required some subsidization. Moreover, households in Besters Camp were typically not allocating expenditure to accommodation, given that rental payments were suspended during the post-warlord period, thereby 'freeing up' a further 25-35 per cent of monthly income ordinarily allocated to accommodation by low-income households. Affordability measured in terms of ability to pay was therefore less of a constraint than the lack of willingness to pay for services consumed.  The roots of the poor cost recovery record in Besters Camp date back to the 1980s, when rental and service charge boycotts were a powerful weapon employed in underrriining the politically illegitimate Black Local Authorities in townships across South Africa. However, by 1995, after almost a decade of non-, or at most, partial payment for services, it became very difficult for local authorities to break what had become standard practice in many low-income  211  areas. National campaigns aimed at increasing levels of service payment such as the "Masakhane" (meaning 'lets build together') campaign, had by 1995 only made a limited impact,  with non-payment in a sample of major townships as high as 78 per cent (Financial Mail 6 October 1995, 45). In many townships and informal settlements, Civics had entered into agreements with transitional non-racial local authorities whereby the quality of service provision would be improved, and consumption costs of services would be paid by residents. However, the inability of many Civics to translate agreements into practice were partially a reflection of the disjunction between Civic leadership and their constituencies on behalf of whom agreements were being entered into, and partially a reflection of the inadequacy of organizational machinery to enforce the unpopular task of collecting user fees.  While the cost recovery history in Besters Camp is highly problematic, and can partially be linked to the more widespread non-payment for services during the 1980s and early 1990s, the problem was exacerbated in some cases by the failure to set up adequate collection mechanisms prior to the delivery of services. Considering the various services in turn, it is seen that in  instances where costs were recovered, this was more by default than by design of project planners.  Looking first at water provision, an attempt was made to establish organizational structures as well as payment procedures prior to the installation of the water reticulation system. Water Committees were set up in Besters and Mzomusha during the initial stages of the project. During 1990 and 1991 payment was made in some sub-areas, albeit often only a partial payment of the monthly water account. These Water Committees also liaised with the UFISD regarding technical problems and improvements to the water system. However, by 1992 all of these  212  committees had collapsed, largely due to conflicts arising from the mismanagement of funds. In Nhlungwane, the Civic was responsible for payment of water accounts, and was the only area in Besters Camp that has regularly made payments, with four of itsfivesub-areas being up to date in account payments by 1995. Ezimangweni had the worst record of payment for water consumed, never having made any payments, nor having any system in place to manage the water points or make payments for water consumed.  The short existence and subsequent collapse of locally managed systems set up to collect user charges such as Water Committees can be attributed partially to the unpopularity of the task, without there being any associated benefits. In the absence of any sanction by the administering authority, there was no real incentive for residents to collect user charges, nor any disincentive associated with non-payment. Furthermore, one of the disadvantages of extreme autonomy of upgrading management and the limited direct involvement of local government in this process, was that the UFISD was not organized to perform a traditional local government function such as the collection of user fees.  Garbage removal was another area in which payment for the service was very problematic. At the level of principle, local (and particularly health) authorities are faced with the dilemma of not having any sanction over people who refuse to pay for garbage removal other than suspending the service, in which case it becomes a matter of time before health problems emerge in the settlement, and which can spread to adjacent areas. Recognizing the poor health conditions in the settlement prior to the upgrade, the City of Durban began removing the contents of garbage skips during the upgrade, with the intention that a system for payment would be put in place at a later stage. However, with the precedent of a service being provided without payment being  213  made, all subsequent attempts to secure payment for garbage removal were unsuccessful.  Cost recovery in electricity provision is the most successful of all services provided and, in principle, represents a mechanism for cross-subsidizing other services. The electricity system was designed on a pre-paid card system used by the City of Durban's Electricity Department, where consumers make cash purchases of electricityfromvending stations, in exchange for debits which are recorded electronically on a card. The card can then be inserted into a meter located within the shack, and the amount is debited as electricity is consumed. The 15-30 Amp supply provided to shacks is enough to supply to any electrical load including stoves, geysers and all small appliances. As a means of raising levels of consumption, each household was provided with a two-plate stove. The combination of the connection charge of R137 per unit and the tariff charged was calculated so as to cover not only the actual costs of electricity consumed, but also the full capital costs of installation amortized over a 20 year period. On a per unit basis, the tariff paid (23.55 cents/kWh) is thus higher than the rate paid by middle and upper income residents of affluent Durban neighbourhoods (18.59 cents/kWh) whose capital connection costs have already been paid off (interview 2, Table A2 in Appendix).  6.1.7  Maintenance and ongoing management of services  Both the cost recovery and maintenance problems were inherited by the DCC following the signing of The Development Services Agreement between the City of Durban and the ICDT on 9 June 1993. Although senior City of Durban housing officials indicated the intention of collecting ratesfrominformal settlement residents not only for services consumed but also for general service maintenance, policy had not progressed much beyond loose statements of intent by 1995 - which was two years after maintenance responsibility for the settlement had been taken  214  over by the City of Durban. Considering the lack of success in collecting user fees for the use or consumption of specific services, the likelihood of residents paying for general rates for the less visible services such as general maintenance and cleaning would appear to be highly unlikely.  The Services Agreement made provision for the effective transfer of maintenance responsibility of services and infrastructure in Besters Camp to the DCC, and ultimately, to the Durban Metropolitan Authority. With the transformation of local government authorities involving the promulgation of the Durban Metropolitan Area in 1995 and the associated replacement of the DCC with a Durban Metropolitan Authority, the Metropolitan Authority assumed maintenance responsibility for the whole of Besters Camp, including areas outside pre1995 municipal boundaries. This signalled the shiftfromthe UFISD performing many functions of afifefacto local authority, to the Metropolitan Authority as both de jure and de facto authority for Besters Camp.  Analysis of infrastructure design and provision in the Hyderabad Slum Improvement Programme in India indicates that centrally important ingredients for successful resident participation in ongoing maintenance arefirst,their involvement in infrastructure design and implementation, and second, the design and choice of technology for infrastructure being supportive of using the skilled and unskilled labour availablefromwithing the settlement (Cotton and Franceys 1988; Cotton and Skinner 1990). Besters Camp scores well on both counts (see Table 6.2 below). While resident involvement in design was confined to the level of workshopping the principles underpinning alternative technological options (for example, waterborne versus VTP sanitation options), the involvement in implementation was extensive, and the services ultimately delivered were supportive of resident involvement in ongoing maintenance.  215  Labour  Specialist equipment required  Suitability of resident involvement  Roads Asphalt roads Concrete footpaths  Skilled Unskilled  Yes No  Limited Yes  Drainage Open drains Covered drains  Unskilled Skilled  No No  Yes Limited  Sanitation VIPs  Unskilled  No  Yes  Skilled  Yes  Limited  Technology  Water supply Public standpipes/ kiosks Electricity supply Street lighting- bulbs Domestic connections  Unskilled Skilled  No • Yes  Yes No  General cleaning  Unskilled  No  Yes  Table 6.2  Infrastructure maintenance in Besters Camp  Although the design was supportive of resident involvement in ongoing maintenance, the potential for resident involvement was not integrated into arrangements for ongoing settlement management, as reflected either in the Services Agreement or in maintenance practice to date.  6.2 PARTICIPATION AND PROJECT DECISION-MAKING INSTITUTIONS  Some of the central principles informing participatory approaches to development can be traced back to the principles underpinning contributions such as 'The Pedagogy of the Oppressed' (Freire 1972) in which dialogue and the value of local knowledge and participation are argued to be critical to development. In-situ upgrading of informal settlements are logistically well suited to a high degree of public participation. Public participation in upgrading projects however  216  ranges along a continuumfromapproaches such as the Kampung Improvement Program in Indonesia which view participation pr