MIGRATION INTO DFR INFORMAL SETTLEMENTS: AN OVERVIEW OF TRENDS

President's Council (1991) 'Report of the Three Committees of the President's Council on a National Environmental Management System'. Government Printer, Cape Town. Quinlan T (1993a) 'Environmental Impact Assessment in South Africa: Good in Principle, Poor in Practice?' South Afican Journal of Science, Vol 89. Quinlan T (1993b) 'The Educational Agenda of Sustainable Development and the Potential of Video in Environmental Education'. Perspectives in Education, Vol 14 No 1. Ramphele M and McDowell C (1991) Restoring the Land: Environment and Change in Post-Apartheid South Afica. Panos, London. Rogerson C (1992) 'Sustainable Urban Development in South Africa - Issues and Problems'. Regional Development Dialogue, Vol 13, No 4. Thornton R and Ramphele M (1988) 'the Quest for Community', in Boonzaier E and Sharp J (eds) South Afican Keywords. David Philip, Cape Town. Zimbabwe Trust (1990) 'People, Wildlife and Natural Resources - The CAMPFIRE Approach to Rural Development in Zimbabwe'. the Zimbabwe Trust, Harare. CHAPTER SEVEN MIGRATION INTO DFR INFORMAL SETTLEMENTS: AN OVERVIEW OF TRENDS By Catherine Cross, Simon Bekker and Craig Clark The success of South Africa's transition to democracy depends upon the success with which the demands of the black urban population for amenities to which they have previously had limited access can be realised. This urban need is now largely focused on informal settlements, which have become home to a large and increasing population. To satisfy the aspirations of informally housed people, state and private sector planning and development agencies will require a subtle appreciation of the dynamics which exist on the ground in such areas. Information is lacking regarding the character of the informal population. Surveys indicate that the informal population is heterogeneous, underscoring the importance of understanding the factors which determine needs in areas against the overall context of the urbanisation process. This report draws upon five recent attempts to contribute to this important process of inquiry. These are four studies of migration into informal settlements undertaken by Cross et a1 in the western, northern, central and southern zones of the Durban Functional Region (DFR), and a study which synthesised the data from these four investigations into a general ove~ew of migration. CHAPTER 6 i E CHAPTER 7

The sample includes 18 settlements and 1 057 households in separate freestanding and infill informal settlements as well as backyard shacks and infill in formal townships, and rental tenancy in row housing in Clermont township. A township closely associated with its surrounding settlements, Klaanvater in Mariannhill, was also compared. In the western zone 263 households were surveyed in Tshelimnyama, Mpola, Dassenhoek and St Wendolins; in the northern zone 264 households in Lindelani, Amatikwe, Bhambayi and Nhlungwane; in the central zone 230 families in Cato Manor, Kennedy Road, Briardene, Block AK, Umlazi backyard shacks and Clermont rental; and in the southern zone 300 families in Bottlebrush, Umgababa, Geza, Mgaga and Umlazi infill. MIGRATION TRENDS: A THEORETICAL FRAMEWORK It is often mistakenly assumed that the increase in the urban population of the DFR is a consequence of a decrease in the rural population of KwaZulu-Natal. This is not completely true: natural increase in urban areas contributes substantially to the growth of the urban population, while the rural population continues to grow. But it is also true, as Rogerson notes, that urban populations are growing faster than rural ones, and the proportion of the total population living in urban areas is increasing. For this reason, rural to urban migration remains a major component of urban population growth. Several different models of rural to urban migration have been proposed. The classical model is Todaro's, which is based on the premise that migration to urban areas is a rational decision arising from a perception that there are more and better paying opportunities for economic activity in urban areas than in rural ones. For Todaro, urbanisation is a one way flow: individuals and household migrants remain in urban areas. This approach is also evident in models that emphasise the gravity flow principle based on distance and trip time in relation to the attraction of the destination, and tend to assume that the population is in stable equilibrium after arrival. The Todaro model and the simple gravity flow concept have been challenged in the developing world by extensive empirical evidence which suggests that urban populations are not in fact spatially stable: rural to urban migrants move around within urban areas and may even move back to rural areas. Two alternative models for urbanisation have been proposed for the developing world: @ Oscillatory migration, in terms of which rural to urban migrants are unable to acquire the security sought after in the urban area, and so continually move back and forth between an urban base and a rural one @ Circulatory migration, in terms of which the migrant continues to live in the urban area, but engages in constant on-migration between different urban locations. Oscillatory migration is widely reported throughout the developing world, and extensive work on the phenomenon has been done in Kenya. Apart from the insecurity of the urban base, an important factor explaining oscillatory migration is status in the rural base. According to Stark: "... for quite a long period of time after moving to the urban sector, migrants retain a strong degree of identification, allegiance and social connectedness with their village of origin. Their social status and prestige depend on their standing in their home village...". Circulatory migration is more complex, reflecting the degree to which local conditions govern access to settlement opportunities. Few migrants move directly from a rural home into an urban area and thereafter remain in place. Instead, many moves take place after the household's arrival in the urban area. These moves count as intra-urban, as Mabin points out, although those involved may be mostly rural born. The importance of understanding which form applies to urban areas is self evident: different types of migration result in different expectations and needs among new arrivals in an area. The results of the present series of migration studies undertaken by the Rural-Urban Studies Unit at the University of Natal suggest that continued on-migration is one of the hallmarks of migration into settlements in the DFR. The data suggests that the most important migration process in the DFR is intra-urban migration. It also appears that the informal population is spatially unstable, and that rural to urban migrant households continue to migrate within the urban area after their arrival. An important point needs to be made about the character of rural to urban migration in the DFR right from the outset. The rate of arrival in CHAPTER 7 k - CHAPTER 7

the DFR appears to have reached a maximum, and may even be tapering off. For this reason, the most pressing problem confronting city management is not an anticipated wave of influx, but continuing in-migration from rural areas together with the size and needs of the existing informally housed population. MIGRATION PROCESSES IN DFR SETTLEMENTS Migration into DFR informal settlements appears to be a complex process. The migration process seems to be opportunistic, with migrants taking advantage of opportunities as they arise. Rural individuals and households arrive in metropolitan informal settlements by complicated routes. Survey results indicate that not all settlements are equally acceptable or open as destinations for rural to urban or intra-urban migrants. The age and size of the settlement may channel different prospective entrants into settlements of given types. The possibility of onward migration by a household in possession of a site is related to a number of other factors, most significantly the location of the other site, the availability and levels of service delivery and infrastructure, and the perceived likelihood of violence in the new area. O Arrival of Rural Migrants It is widely assumed that the high proportion of informal settlers in the DFR represents a recent wave of rural to urban migration after the scrapping of influx control in early 1986. However, by far the majority of the present informal population - 83% - was in the DFR before 1986. Early entry into the DFR is not evenly distributed across all the informal settlement complexes. Younger settlements in the central DFR have the lowest percentage of households who arrived in Durban before influx control was lifted, while the settlements of the western periphery at Mariannhill have the highest, at nearly 97%. Around one in five households in the southern DFR arrived after the lifting of influx control, and 8% in the northern periphery. Only one settlement, Block AK in central Durban, had a majority of relatively recent arrivals. This is not to assume that the informally housed have been established in their present settlements since prior to 1986. There appears to have been a considerable spatial displacement of population inside the DFR since that date. This movement represents on-migration. The long delay involved in urban on-migration can be seen when one compares the date of arrival of households and individuals in the DFR with the date of arrival in their current settlements. In the western DFR settlements, according to Cross, only one in five households had arrived in the settlement since 1986, but this is in an established and relatively stable area which exerts firm control over the influx of outsiders. In Inanda in the north, some 54% of people have entered their settlements since 1986. The figures for the southern and central zones were 78% and 89% respectively. In the central and some southern settlements, most of the population has arrived since 1989. It is clear that the easing of the state's attitude to land occupation and invasion is largely responsible for the younger settlements of the central DFR. O Rates of Entry into the DFR Entry into the DFR appears to have taken place on a relatively smooth, 0 slightly rising curve up to 1986. It is difficult to tell whether or not a significant further rise in rural to urban migration is taking place, and G it is likely that the 1960s and 1970s are underestimated. In-migrants oe from this period will have been thinned by natural attrition and by c!jreturn migration, leaving behind children who now count as urban born. If assessed at the point of first entry to the DFR, there appears to be an increase in rural to urban migration since 1986, but a very slight one. It is also unclear whether or not this rising trend has flattened out (Table 1). It appears that the anticipated massive increase in rural to urban migration following the abolition of influx control has not eventuated. Instead the rate of migration may be fluctuating over time. Rural to urban migration may have climbed gradually until very recently, or may be continuing to climb. It appears to have increased slowly through the 1960s from a base level of about 1% a year to around 4% by 1970. The recorded migration rates indicate that migration was high at some 8% in 1980, but then declined to 6% from 1984 to 1985. Thereafter recorded in-migration climbed again to around 9% in 1988 and 1989. It appears migration has declined since then: Cross et al estimate a figure of 6% for 199 1. z - Zi

Table 1 Year of Arrival for DFR In-Migrant Informal Population now in Place. / Year entered DFR % Year entered DFR % Year entered DFR % 1 Pre F946 3 1946-47 * 1948-49 1 I Table 2 Origins of the DFR Informal Population. Area where Respondents were Born Percentage distribution (percentages rounded: do not add to 100%) * denotes percentage less than 1 Oj0 1990-91 6 (1 991 data incomplete) URBAN AND RURAL SOURCES OF MIGRATION The sources of the survey population are shown in Table 2. Only 0,3% of the informally housed DFR population were born in the areas where they resided at the time of the survey. Thus by far the majority of them are migrants, but not all are of rural origin. A substantial percentage in fact appear to be of urban origin. For the most part these households represent people who have been unable to secure formal accommodation in the townships. Most of the adult populations of settlements - 65% - appear to be non-urban in origin. Nearly half of the remaining 35% who are urban in origin are found in settlements in the western Mariannhill area. For the northern settlements at Inanda, the younger central settlements and the southern area, the adult informal population is 76% non-urban in origin. The dynamics of the urbanisation process are importantly influenced by these differences between settlement zones. The most important areas of origin for the DFR informal population are the rural districts of former Natal, which contribute some 46% of the informal population. At some 5%, the peri-urban districts are relatively less significant sources of inflow, as is the Transkei at 8%. The contribution from other former homelands and countries outside South Africa represent Dbn central 2 3 2 2 2 2 * 1 * * Outer Dbn Southern urban 1 2 5 13 7 5 Northern urban * 21 10 5 12 9 * 6 Western urban 58 3 3 19 All Urban 61 26 \r 23 24 24 35 Southern peri-urban 4 4 4 8 5 5 * * * Northern peri-urban * 2 * Western ~eri-urban 1 3 1 All peri-urban 5 7 5 8 7; 5 1 Western urban perph 7 4 3 3 3 1 South coast North coast South nearer rural North nearer rural South rural interior North rural interior Transkei Outside NatalITkei TOTAL 99 102 101 99 100 " 99 N= 302 264 230 300. 794 1096 (Source: Cross et al 1994 in press) a negligible 2%, but these may be under reported. The informal population is thus predominantly from KwaZulu-Natal or Transkei, and is mostly of rural origin, although there is a substantial urban origin component. O The Urban Component A very small percentage of the DFR's informal population originates in the formerly white central city itself. This is likely to undergo profound change over the next 25 years: it is in this area that settlement is consolidating, and challenging the urban spatial morphology of the classical 'apartheid city' first described by Davies.

The balance of the urban origin informal population come from large formal townships in the northern, southern and western urban districts. Urban origin is concentrated in the western zone, where 58% of informal settlers are drawn from the townships of the western periphery and Pietermaritzburg. If one excludes Marianhill, the contribution of the urban townships to the population of the DFR settlements declines. Overall, some 24% of the adult population of settlements excluding the western zone may be said to be of urban origin. According to Cross, the outlying peri-urban areas of the DFR, which are mostly located in the former KwaZulu, contribute a total of 7%, mostly from the southern peri-urban region. The origins and migration behaviour of the population of the Pinetown and Mariannhill area suggest that this is a separate and semi-sealed population which is predominantly local and urban. Clearly, it is the large Durban settlements which are receiving the bulk of the region's rural to urban migration. O The Rural Component The southern Natal region contributes some 22% of the informally housed population of the DFR, a significant source of population. Most of this comes from southern rural districts on the borders of KwaZulu and nearest to the city, which provide 11% of the DFR informal population. Some 2% come from the southern interior rural districts, and the remaining 9% from the south coast. The northern Natal region is somewhat more significant, at 24%. Some 14% of the DFR informal population originates in the northern nearer rural districts, with 5% each coming from the northern rural interior and the north coast districts. O Overview of Origin Areas The most important population sources for informal settlements of the DFR, contributing more than 7% of the total population, are the peripheral urban areas in the north, comprising the northern townships and informal areas, the south coast, the nearer rural areas to the north and south, and the Transkei. Outflows of between 5% and 7% come from the southern urban areas, the southern peri-urban districts, the north coast and the northern outer rural areas. Outflows of less than 5% characterise the central city districts, the western urban area centred on Pinetown, the western urban periphery centred on Pietermaritzburg, the northern and western peri-urban districts, the southern outer rural districts, and all areas outside Natal other than Transkei, according to Cross et al. The informal population thus has tended to migrate into the DFR from areas closely linked to Durban and within reach of transport connections which lead into the city. What also characterises the sending areas is a more mobilised economy and a relatively higher population. Of far smaller significance than expected were the densely populated KwaZulu peri-urban districts surrounding Durban. The last 10 to 15 years have seen population transfers out of remote rural KwaZulu into these peri-urban areas, which offer access to the job market and amenities of Durban, without involving the important adjustments to urban lifestyle. It would appear that KwaZulu's peri-urban periphery has so far provided acceptable access to urban goods. This may change as serviced land in and near the city centre becomes more available, and becomes known z to be more available. These areas are embraced by urban information 0 networks, and it is possible that part of the peri-urban population may migrate into the urban core in the near to medium future. 5 OT (3 DIRECT AND INDIRECT MIGRATION Rural to urban migration into informal areas includes both indirect and direct population flows. Rural to urban migrants may follow a sequence of indirect moves in a scouting process until a relatively satisfactory place is located. Later direct migration by a pipeline process from the rural area into the new locality will depend on connections established by the initial indirect scouting and staging process, resulting in local colonies of in-migrants from the same home areas. Indirect migration appears to take place by several moves. Once migrants have arrived in an area, many or most moves seem to be staged from nearby. With pressure building up in the freestanding settlements and sites more difficult to obtain, an increasing percentage of new entrants appear to stage their moves from areas immediately adjacent to the new site. Around 20% to 25% of people may now be arriving in settlements from directly adjacent townships or other residential areas. - E

Only 29% of the total in-migration from rural areas was direct: the remaining 71% was indirect. To transfer people on any scale, the pipeline formation process seems to depend on the location of the urban base in a well known, and therefore both larger and older, urban area. Direct migration into the younger central settlements, which are comparatively small and only locally known, is very rare indeed. These rural pipeline connections are most prevalent in Inanda, which receives far more direct rural migration than any other zone, with many people arriving from northern inner rural districts as well as from the north and south coast. The most striking of these was a Transkeian pipeline into Bhambayi, which accounted for about 20% of the population. In the southern informal settlements, which paradoxically also receive important migration flows from the northern inner rural districts, only 6% of the recorded migration paths were direct. The Umlazi interior infill areas, however, received substantial direct flows from the adjacent townships, as did Umgababa, where many new arrivals seemed to be urban residents fleeing political violence. Little rural to urban migration of either direct or indirect type was reported in the western zone, where the entry of outsiders is fairly formally controlled. There is evidence to suggest a degree of self selection in the direct migration process from rural to urban areas. In-migrating households moving to the city by a pipeline connection can obtain substantial information on possible destinations prior to deciding to move. Relatively large numbers of female headed households were found in migration streams to areas such as Cato Manor, which were accepting women heads of household. Direct migrants, Cross et al note, also tend to be slightly younger than indirect migrants, and to have fewer and younger children. In other areas male headed households tend to predominate. Adult educational levels decline slightly with the distance of the area of origin from Durban, while the age of the household head rises. Larger households are clearly linked to the western areas, probably because they are older and more differentiated than those in the Durban shack population. Urban direct migrants appear mostly to represent population overspill from the townships. Their moves are made over relatively short distances, particularly into township infill settlements. The migrants are younger and better educated than is the norm, and the infill areas are accordingly better off socio-economically than the large free-standing settlements. 1 Table 3 Dates of Activity for Zones Sending and Receiving Population! I lp7' 1991 DFR North Northern townships coast Southern townships Central DFR 1989 DFR South Southern Northern townships coast and Northern border areas South interior 1985 Pietermaritzburg North West interior DFR Northern townships RECENT AND ESTABLISHED MIGRANT POPULATIONS Transkei There are substantial differences between sub-regional sending areas in 0 regard to whether migrants arrived before or after the lifting of influx control, Cross et al have found. These differences are reflected in Table 3. The western zone of Mariannhill is the earliest recent centre of migration in the DFR, with one in every three households moving into their present settlements before 197 1. Migration activity into settlements from the northern urban areas accelerated after 1971, reaching 30% between 1986 and 1989 and 31% in 1990 and 1991. More recently, the southern urban and central Durban populations have become active. Up to 50% of the present informal population of these zones have moved into their present settlements since 1989. The period of greatest activity in terms of sending migrants into the DFR for the southern coastal and southern and northern interior rural districts was 1986 to 1989. Before 1985 was the period of greatest migration activity for the northern nearer rural population. The north coast, southern nearer rural districts, the Transkei and areas outside Natal became active after 1990, and have subsequently been sending the most migrants. -

Households which settled after 1990 show a significantly higher percentage of household heads aged below 30 years. This is most noticeable in the 20 to 24 years cohort, which shows a change from 15% in 1989 to 62% after 1990. The magnitude of the shift suggests it is probably related to the setting up of younger central settlements and to greater opportunity to obtain land in the DFR, which have allowed young people to move out of home earlier than previously. Aside from violence driven refugee processes, on-migration appears to be driven by a desire to improve the household's situation in regard to urban access and transport, security and autonomy. Life in a settlement offers the household relatively good urban access and full autonomy at a low price. Further on-migration might be occasioned by the levels of service delivery in other areas, on network connections offering alternatives, and on outside factors, most importantly, levels of violence. O Overview of Intra-Urban Migration A significant part of migration appears to be between urban areas. If the immediate source areas of the population are considered, the predominance of migration in short hops, from adjacent areas, is clear. It appears that the informal population locates accommodation predominantly through a process of urban on-migration which is affected by linkages between specific established residential areas - formal or informal - and specific informal settlements. Most of these relationships involve population transfers across short distances, but some resemble pipeline connections and may operate over considerable distances. Cross et a1 note that accommodation is sought and found through relatives, home people, friends and co-workers, while the African independent churches may also play a central role. These networks then structure opportunities for further on-migration. In addition, they constitute a nucleus of what will become civil society in the new informal settlements. O Urban On-Migration The DFR informal population appears to be extremely unstable, with nearly 90% having moved at least once since 1986. In-migrants may work their way inwards towards the urban core, staging in through temporary accommodation. Several moves are common: a small number of respondents reported up to 10 moves. Individuals who arrive in town alone often form households on the way inward. However, some individuals obtain sites as single person households. On-migration includes two forms of moves: initially, migrants who do not enter an informal settlement as direct migrants move between forms of temporary accommodation, in townships, at work, in hostels or otherwise. Several stops may be involved. Once a site is obtained, fewer households, perhaps about one third overall, move on again. SUMMARY AND CONCLUSIONS The first requirement for promoting a viable settlement process appears to be stabilising the high level of on-migration. The second is likely to be establishing a viable relationship between informal settlement and the urban planning process. Two forms of urban on-migration appear to be common: firstly, through temporary accommodation prior to obtaining a site and, after that, between informal settlements. The first type of continued on-migration is a natural consequence of the rural to urban migration process, and cannot be eliminated. But the provision of attractive sites may reduce it. The easing of restrictions on settlement may already be reducing the incidence of this form of migration. Well designed development interventions may also reduce spatial instability if they reduce the length of the in-migration pathway. The second type of continued on-migration may be the result of violence or attempts to improve the situation of the household. Suppression of violence, the single most substantial cause of continued on-migration, will significantly reduce residential instability. Development interventions are more problematic since they may provoke violence and competition as well as attract households to move in from more disadvantaged areas. Making upgrading and service delivery as widely available as possible may reduce differentials in levels of services between areas, and be the most effective means of stabilising this migration stream. Wide scale provision may require a reorientation in methods for delivery of se~ces. Either form of on-migration may prove an obstacle to promoting stable communities in informal settlements. Changing political conditions, and the increased accessibility of land and services, make it uncertain whether there will be a rise or fall in the medium term in the total volume of in-migration moving through the DFRs shack communities.

Mabin A (1991) 'The Dynamics of Urbanisation Since 1960', in Swilling M, Humphries R and Shubane K (eds): Apartheid City in Transition. Oxford University Press, Cape Town. Mabin A (1992) 'Dispossession, Exploitation and Struggle: An Historical Overview of South African Urbanization', in Smith DM (ed) The Apartheid City and Beyond. Routledge, London and New York, and Witwatersrand University Press, Johannesburg. Mprller V and Schlemmer L (1985) 'Migrant Workers: A Profile of Their Rural Resources', in Giliomee H and Schlemmer L (eds.) Up Against The Fences: Poverty, Passes and Privilege in South Afnca. David Phillips, Cape Town. O'Connor A (1983) The Afncan City. Hutchinson Group, London. Oosthuizen G and Clark C (1994) 'The Development Activities of African Independent Churches'. Paper for the Informal Settlements Programme. Rogerson C (1993) 'Managing Urban Growth: The International Experience'. Urban Foundation, Johannesburg. Stark 0 and Lucas R (1988) 'Migration, Remitttances, and the Family', in Economic Development and Cultural Change, Vol 36 No 3. Todaro MP (1982) Economics for a Developing World, second edition. Longham, Harlow CHAPTER EIGHT CONFLICT AND MIGRATION IN KWAZULU-NATAL By Simon Bekker and Antoinette Louw Research into informal settlements of the Durban Functional Region has revealed a very high population exchange. The high incidence of 'moving on', reported by Cross and Clarke in their research for the Rural Urban Studies Unit at the University of Natal, is explained in a complex manner. But one important element is the role that both perceived and experienced violence play in decisions to quit one residential area and seek out another in which to settle. This research also reveals that most recent migration streams in informal settlements are intra-metropolitan: rural to urban migration comprises a small proportion of the volume and, once having entered the Durban Functional Region (DFR), households tend to 'move on' within the area. A different research programme, Conflict Trends in Natal, has as its aim to develop a data base of conflict events in the KwaZulu-Natal region from 1986 to the present. The project, which defines conflict as collective violent behaviour, uses multiple sources and has captured some 10 000 conflict events. Analyses based on the data, by Bekker and Louw, have shown not only that regional conflict is sustained, but that its occurrence in urban informal settlements is pronounced. Since this data base enables analyses which are sensitive to place as well as time of conflict event occurrence, it is possible to describe, i k CHAPTER 8