Women's participation in housing delivery: the case of Ezilweleni, South Africa

Abstract

As part of a study on housing delivery in post-apartheid South Africa, research was conducted to explore women's participation. The interest in women was based on the assumption that their involvement would contribute to greater equity in the residential construction sector. Focus group discussions and a household survey were conducted. Participants were asked about their income sources and their involvement in various phases of housing delivery such as securing tenure, planning, procuring materials and building. The findings indicate that women influenced decisions about the level of infrastructural services in Ezilweleni. They participated in housing delivery but reverted to informal sector activities despite training in building. Both formal and informal education are necessary if greater equity is to be achieved in the construction sector.

1. Introduction

This paper discusses research findings from Ezilweleni, a community in KwaZulu-Natal. The aim of the study was to understand women's participation in housing delivery in post-apartheid urban KwaZulu-Natal. The topic of women's participation was of interest because women's involvement in housing development in South Africa has for a long time been minimal. The paper explores women's position in relation to housing in apartheid and post-apartheid South Africa and then elaborates on the methodology that was used to gather the data presented. It discusses the community of Ezilweleni and gendered perceptions of participation in housing delivery, and concludes by recapitulating the key ideas arising from the study.

2. From Exclusion to Gender Equality in Access to Housing in South Africa

During the apartheid era, women's involvement as beneficiaries of housing was minimal, as they were prohibited by law from owning houses except through their spouses. The enactment of the Group Areas Act in 1950 led to forced removals in areas which were subsequently declared for whites only, for example Cato Manor in Durban, District Six in Cape Town and Sophiatown in Johannesburg (Morris, 1981: 43). For a period of almost 40 years (1930 to 1970), the only public housing available to Africans in the city of Durban was in the townships of Umlazi and Kwa Mashu, both of which had about 34 100 housing units (Maasdorp & Haarhoff, 1984; Maylam & Edwards, 1996). Women in general and African women in particular had few or no rights to housing. Women's living conditions were worsened by subsequent housing regulations (Morris, 1981).

In 1967 the department of Bantu Administration and Development (formerly Native Affairs) passed a regulation for African women and non-productive Africans. According to the circular, African women were not to be placed on waiting lists for family housing in urban areas. Those who qualified for accommodation had to seek it as tenants in registered households, which put women in the position of minors under the state regulations. A divorced woman could stay in her home only if she was not the guilty party and if she had been granted custody of the children. All those who were considered non-productive Africans (widows, the handicapped and the elderly) living in white areas were to be resettled in the homelands (Morris, 1981). African women's housing problems were further aggravated by regulations passed in 1968, which stipulated that housing permits could be provided only to qualified African men over the age of 21 who were employed and had dependants legally living with them (Morris, 1981). According to a circular by the Department of Bantu Administration and Development (1969), proposed to local authorities, where a white city was close to a homeland, the Africans employed in the city would be required to live in the homeland. However, if the distance between the town and the homeland was too great, hostel accommodation was provided for the workers with a view to having them visit their families periodically. In Durban there was only one women's hostel, Thokoza.

An examination of the Women in Development (WID) approaches is useful for understanding women's position in housing and their subsequent participation in housing delivery in post-apartheid South Africa. As the theme of the women's decade (1975–85), the equity approach recognised that women were active participants in the development process and through both their productive and reproductive roles made an important contribution to economic growth, which was not always acknowledged. While accepting women's basic need to earn a livelihood, the equity approach was more concerned with issues of inequality between men and women in both the public and private spheres of life and across socio-economic groups. It identified the origins of women's subordination in the family and also in the relations between men and women in the public domain (Kabeer, 1994; Young, 1997).

Central to the equity approach was the assumption that women had lost ground to men in the development process. Redistributive measures therefore had to include a process in which men shared in ways that would enable women from all socio-economic classes to gain and men from all classes to lose (or gain less). State intervention through affirmative action policies was seen as the way to meet women's strategic needs for economic autonomy and equality with men. Such measures would challenge women's subordinate position in society. Despite the criticisms levelled at the equity approach, it has continued to provide a framework for governments to improve the status of women through official legislation on issues such as property rights. Resistance to the demand for equality led to a policy shift among feminists so that their demands coincided with the general direction of development, which emphasised the goal of meeting basic needs (Kabeer, 1994; Visvanathan et al., 1997; Young, 1997). In housing, the equity approach focused more on the impoverished sections of the population, and arguments were put forward for participation.

Because of the political situation during the first two international women's conferences in Mexico (1975) and Nairobi (1985), the policy debates arising from these conferences cannot be said to have greatly shaped development thinking among women in South Africa. In terms of actual involvement in women and development policies, South Africa appears to have leapfrogged from ‘women in development’ to ‘gender and development’. However, because the ideas of WID are still relevant in analysing the position of women with regard to housing in South Africa, this paper draws on the equity approach as a framework for understanding women's participation in housing delivery. When the equity approach was being advocated elsewhere in the world, the structure of the South African society was based on apartheid, with racist and segregationist policies, and therefore notions of equality, let alone gender equity, were out of context in such a society. The housing policies developed during this period in South Africa tended to reflect the broader socio-economic and political context of the country and the state policies remained oppressive to all non-whites and all women regardless of race. During the 1985 Women's Conference in Nairobi, the South African women in exile forged links with women from other developing countries and began to advocate for the recognition of gender rights as a political and economic issue and not just as a social matter (Ginwalla, 1999). The influence of these ideas on women's rights and equality found expression in the demand for the inclusion of the equality clause in the post-apartheid constitution (LRC, 1998).

Alongside the right to equality is the right to adequate housing, which was first inscribed in the White Paper on Housing (RSA, 1994) and was included in the socio-economic rights in the 1996 constitution that was finalised in 1997 (LRC, 1998). These constitutional provisions were followed up by the creation of the Commission on Gender Equality, whose role was to act as a government watchdog to ensure that gender equality was achieved in all sectors of the economy. Equality in the current housing policies is illustrated in the design of the housing subsidy scheme that grants the poor a housing opportunity through a one-off grant. The White Paper recognises the need not only to build capacity in housing delivery but also to address gender discrimination in skills training by stating that

it is essential that new policies and legislative action by the state should be particularly sensitive to the removal of entrenched discriminatory mechanisms and conventions in respect of gender, race, religion and creed. Government has particularly identified the need to support the role of women in the housing delivery process. (RSA, 1994: 23)

However, although the policy guidelines in support of women's participation in shelter development are in place, the implementation of these varies from area to area even within the same province (Ndinda, 2001).

The primacy of participation by the various stakeholders in housing delivery is underscored by the statement that

Government is committed to a process driven from within communities. Through its policies and strategies it will encourage and support local initiatives emerging from communities or broader local social compacts aimed at equipping and empowering people to drive their own economic empowerment, the development of their physical environment and the satisfaction of their basic needs. Policies must recognise and give effect to this approach … this will include not only financial resources but the creation of appropriate institutional frameworks and support structures. (RSA, 1994: 23)

Embodied in the post-apartheid policy are the twin objectives of participation and gender equality in shelter delivery, which form the core of this paper. Various studies suggest that the involvement of women in the planning, design and implementation of housing projects results in a higher success rate of projects than where women are excluded (Vance, 1985: 33; Nimpuno-Parente, 1987: 71). The concepts of gender and participation when interrogated in the light of shelter development raise important questions that lead to debates on shelter delivery in post-apartheid South Africa.

3. Methodology

The data presented in this paper were collected as part of a larger study conducted by the author in KwaZulu-Natal in 2000 but this paper focuses on one settlement, Ezilweleni, in the Southern Pinetown area of Durban now known as Ethekwini Municipality (see Figure 1). Ezilweleni borders another community known as Luganda, which also formed part of the study.

Figure 1: Location of Ezilweleni in the eThekwini municipality

The multidisciplinary nature of the research, that is, housing development, which is empirical, and gender studies, which is phenomenological, made a combination of research paradigms inevitable. The research used qualitative tools to examine theoretical issues arising from the survey of literature on women's involvement in housing delivery in South Africa, but to capture the level of involvement it was necessary to complement the qualitative responses with quantitative data. When the study was conducted there were 160 households in Ezilweleni, each occupying a serviced site. Household questionnaires were administered to 10 per cent of the household heads and an in-depth interview was conducted with a key informant. In this area two focus group discussions were conducted, one with ten women and another with 12 men. The household questionnaires were computed using a statistical package for social science and descriptive statistics and were compiled to highlight the demographic, socio-economic and housing satisfaction levels. Notes were taken from the in-depth interview and the focus group discussions and the content analysed to reveal the key themes: participation in housing delivery, organisation, levels of empowerment, planning, and the successes and problems of community involvement in housing delivery.

4. The Process of Securing Tenure in Ezilweleni

Ezilweleni was initially an area where people had settled and were paying rent to Indian landlords. At the time of writing, in 2006, this semi-urban area in Marianhill on the outskirts of Greater Durban is a land-owning community, but before 1989 the residents were tenants. Land ownership came in two phases. In 1989 the residents were threatened with eviction by the local authority and the landlords. This motivated them to work together to resist eviction by forming a civic committee, its members drawn largely from the residents' association, an earlier organisation. This committee's role was to lead the residents of Ezilweleni in the fight against forced removal from the land. It organised meetings with the residents and mobilised them to demonstrate against the landlords and the local authority. In this campaign it worked with the Greater Marianhill Coordination Committee (GMCC), based at the Marianhill Mission Institute. The GMCC brought together stakeholders in the sub-region and coordinated the activities of all the groups that were fighting for security of tenure. The civic committee also made contact with the Built Environment Support Group (BESG) that was supporting the campaign against eviction through the GMCC. On one side were members of civil society represented by civic organisations and non-governmental organisations united under the umbrella organisation, GMCC, and on the other were the landlords and the local authority, eager to carry out the removals in the whole Marianhill area (Ndinda, 2002).

The civic committee in Ezilweleni made residents aware of their rights and of the issues that were at stake in the face of threats of removal. Mobilisation of the community was done through ‘prayer’ meetings where tenants narrated their ordeals at the hands of landlords. These so-called prayer meetings were actually political meetings aimed at mobilising the community and raising awareness. Demonstrations were organised against the local authority that was perceived as supporting the landlords. The role of the Ezilweleni civic committee was also to represent the community's interests in negotiations with the local council and to provide feedback to the community through mass meetings. When an eviction notice was issued to Ezilweleni residents, the civic committee took up the issue with the landlords and the local council. It was also discovered that some of the landlords in Ezilweleni were harassing individual households and had gone to the length of demolishing some of the shacks, while negotiations about the land were going on. As a result of this action by the landlords, one of whom was also a shopkeeper in the area, the committee organised a consumer boycott that lasted for three days, forcing the landlord to concede to the residents' demands. The committee also demanded that the demolished shacks be rebuilt and that the residents be allowed to return to their land. Since their shacks had been demolished, residents demanded a reduction in the rents when they moved back. It was at the ‘prayer’ meetings that the residents planned the necessary mass action to have their demands met by the landlords and the local authority, and through this mass action they succeeded in bringing them to the negotiating table (Ndinda, 2002). Women were involved as part of the civic movement even at this stage of fighting for land tenure.

The ensuing negotiations resulted in the landlords agreeing to sell the land to the tenants. By working with BESG, the residents secured funding from the Independent Development Trust (IDT). In 1992 they used this funding to establish the Zilweleni Development Trust (ZDT) ¹ because they were unhappy with the way the local authority implemented projects in the area, which they considered top down and not consultative (Ndinda, 2002). The community felt it could do better if it formed its own trust to oversee development in the area. The objectives of the Trust included encouraging community participation in the upgrading of Ezilweleni, providing security of tenure and helping members acquire housing, raising funds for the Trust, and managing the funds in accordance with the objectives of the Trust. The Trust was also tasked with implementing development in the area. Using the funds from the IDT, the ZDT bought the land for R199 000 and subdivided it into 187 plots. It is noteworthy that women in Ezilweleni represented their own interests and those of their households even before the South African Constitution allowed them to do so.

The role of the civic committee after acquiring the land was to act as the ZDT's watchdog by overseeing the Trust Funds and ensuring that obstacles to the development of the area were cleared. For instance, in the infrastructure planning process, if the proposed road passed through people's houses, it was the role of the civic committee to speak to them and provide alternative plots to relocate those who were affected. This approach was rather top-down as there seems to have been no involvement of the residents in planning the settlement layout and design. The civic committee also resolved disputes among the residents.

4.1 The phases of housing development in Ezilweleni

4.1.1 Phase 1

The settlement was initially developed as a site and service scheme. Land tenure was secured by 1991 when the apartheid laws under which women did not have the right to own public housing were still being enforced. The community therefore debated the issue of allocations and came up with an allocations policy based on the categorisation of residents into four groups:

A	=	Residents with structures
B	=	Mature residents, i.e. children of residents who had lived there for many years
C	=	Sub-tenants
D	=	Residents from the sub-region (loosely defined, with shifting boundaries, and referring to residents in the immediate neighbourhood or to those in the greater Marianhill area).

Category A residents had the choice of continuing to live on their sites or moving to any other vacant site. Category B, C, and D residents were allocated their sites through a lottery system and after that were shown the sites by the surveyors in the presence of other residents. Allocating the plots using the lottery system enhanced the credibility, transparency and accountability of the leaders. As will be shown in this paper, women were part of the leadership in this community.

It will be noted that the A, B, C, D criteria are apparently gender-neutral – though, as I shall show below, this was not completely the case. Initially, some residents argued that men over 25 and women over 30 who lived in the settlement should be allocated sites. The argument for the differentiation in gender terms was based on the notion that men over 25 would remain in Ezilweleni whereas women of the same age were likely to get married and leave the area. However, it was believed that women of 30 or over were unlikely to get married. The residents' association that pre-dated the civic organisation (referred to above) insisted on gender equality and it was agreed both men and women over 25 would qualify for the sites.

The committee, with the assistance of BESG, brought in the technical team which surveyed the land, subdivided it into plots and put in basic infrastructure such as gravel roads, storm water drainage, ventilated improved latrines, and electricity for each household. They also supplied street lighting and provided water in the form of communal standpipes. The team was paid from the Trust funds for the subdivision and putting in the infrastructure. The local authority later installed telephone lines. From 1993 residents started building their own houses slowly with their own savings. It can be seen from this brief historical account that community participation in decision making related to housing, including the participation of women, dates back to 1989.

4.1.2 Phase 2

The second phase came about in 1994 when, under the new housing policy, the community applied for the consolidation subsidy to build the houses. The 1994 government-implemented subsidy scheme was based on the principle of capital subsidy (Adler & Oelofse, 1996: 116). When the subsidy was designed and implemented in 1994, it used a strong targeting mechanism in which subsidy allocations were based on the household income, with the lowest earners accessing the full grant and those earning close to R3500 accessing less. Since 2004, the housing policy has shifted to include households earning R3501 to R7000 in the credit-linked subsidies (DoH, 2004). In the current design of the housing subsidy, all eligible households earning below R3500 qualify for the full subsidy that is meant to provide a 45 m² starter unit.

Although the subsidy programme has been expanded to include more categories of households, the eligibility criteria have remained the same since 1994 (see Table 1). People married in terms of civil or customary law and couples living together but not married qualify. The beneficiaries must be lawful South Africa citizens. Recipients of the subsidy must be over 21 years and legally competent to contract, and have financial dependants – a view that drew widespread criticism from various analysts and organisations. The subsidies are aimed at helping households acquire ownership of fixed residential properties for the first time. The different types of subsidy are individual, project-linked and relocation assistance, institutional, and consolidation. The subsidy is linked to the household income, as shown in Table 1.

Table 1: Subsidy quantum for 30 m² houses in the 2006/2007 financial year

Income category	Previous subsidy	New subsidy	Contribution	Product price
Individual, project-linked and relocation assistance subsidies
R0 to R1500	R31 929.00	R36 528.00	None	R36 528.00
R1501 to R3500	R29 450.00	R34 049.00	R2479.00	R36 528.00
Aged, disabled or health stricken R1501 to R3500	R31 929.00	R36 528.00	None	R36 528.00
Institutional subsidies
R0 to R3500	R29 450.00	R34 049.00	Institution must add capital	At least R36 528.00
Consolidation subsidies
R0 to R1500	R18 792.00	R21 499.00	None	R21 499.00
R1501 to R3500	R16 313.00	R19 020.00	R2479.00	R21 499.00
Aged, disabled or health stricken R1501 to R3500	R18 792.00	R21 499.00	None	R21 499.00
Source: Adapted from DoH, 2006.

People who had received housing assistance from the state in the form of serviced sites (including sites serviced by the IDT) before the new subsidy scheme could apply for the consolidation subsidy if their income was below the stipulated maximum of R3500, as was the case with all the residents of Ezilweleni. This subsidy was provided for upgrading the structure on a serviced site. It was sufficient for a core or starter structure that could be incrementally built. As the matter was understood by both the men and women's focus groups, everyone in the community was poor and there was no difference between male-headed and female-headed households. However, as Figure 2 shows, no women earned more than R1500, whereas there were men who earned more than R1500. All the 160 households in Ezilweleni qualified for the consolidation subsidy. They were involved in mutual self-help housing and not in the people's housing process which became policy only in 1998.

Figure 2: Income level by household head in Ezilweleni

4.2 Organising, and women's participation

After the acquisition of land tenure in 1991, the civic committee became the housing development committee. During the struggle for land tenure, the Ezilweleni residents had divided themselves into seven blocks so that all residents would be represented in decision making, and each block was expected to elect two members to the development committee. However, the level of participation waned after land tenure and subsidy funding for house construction had been secured, with many residents opting to employ their own builders instead of participating in self-build together with other members of the community. Some of the committee members opted to step down from the committee, arguing that they were not getting paid for their involvement although they spent a lot of time in meetings, with the result that instead of the committee having its full 14 members, with all the blocks represented, there were now ten: six men and four women.

The elections were supposed to be held every two years but often the community asked the existing committee to continue with its work. The committee called for elections because they had insisted on operating on democratic principles, but elections were not held regularly since members were obliged to remain in their positions owing to community unwillingness to join the committee because there was no remuneration.

The community was involved in the planning phase of the Ezilweleni project. The leaders raised the issues and the heads of households would be called to mass meetings to say whether or not they agreed with the decisions, a fact reflected in the questionnaire in which 100 per cent of the respondents indicated that the community made the decisions about the development of the settlement. In all the seven blocks, committee members were elected at meetings consisting of household heads, where the name of a potential representative, a household head, would be suggested and supporters would give a show of hands.

Elected members allocated the development committee positions among themselves based on their skills as they understood them. The men held the key decision-making positions of chairperson, vice-chairperson and treasurer. At the time that the study was conducted, women held the positions of secretary and three additional positions of member without portfolio (see Figure 3). Evidence from the planning process showed that women actively participated in decision making in the community. Although most decisions were taken by the community at mass meetings, women ensured a democratic process at the committee level and in the community during the allocation of sites. Patriarchal control in Ezilweleni was nevertheless evident in the fact that although there were women in the committee none held a key position. The key decision-making positions (chairperson and deputy chairperson), where the incumbents had the right to use their own discretion, were held by men.

Figure 3: Structure of development committee in Ezilweleni

Neither the proportion of women leaders to men in the development committee in Ezilweleni nor the capacity in which the women leaders served (as secretary and additional members, which reflect their traditional roles in the home and in the labour market) sufficiently account for their success in influencing key decisions. This paper argues that this success was the result of the strategies the women employed, for example in lobbying for decisions on site allocation to be taken in mass meetings, and in convincing the progressive men in the community to support gender equality in the site allocations policy.

4.3 The building process

End-user participation was limited in this community. When people were fighting eviction, there was unity and the community worked together. They had clear objectives and this contributed to group cohesiveness until they secured land tenure. Up to the implementation phase, community participation was beneficiary driven, with the motive of addressing the needs of the participants.

The project was implemented jointly with Luganda (a neighbouring community, see Figure 1) through the Sibambisene ² Joint Venture, formed in 1996 (personal communication, key informant, 21 March 2000) The aim of the joint venture was to save on the subsidy by purchasing building materials in bulk and by allowing builders from one area to cross into the other when required. The role of the local authority was to administer the funds for the joint venture as they had the capacity to do so. Later the two communities decided to engage a private consultant to do the accounting and administration of the funds as the rates offered were lower than those the local authority was charging. The committees of the two areas involved in the joint venture met fortnightly to share ideas and lessons from the implementation process. It was in these meetings that building problems were raised and solutions reached.

As the project facilitator, BESG assessed the residents working as builders and grouped them into those who had some building skills, and those who had extensive experience in construction work. Those who needed additional skills were enrolled in a two-day induction programme and then in a two-week training programme with Khuphuka (personal communication, Mr Campbell, Training Manager, Albert Luthuli Development Trust (ALDT), Stanger, Durban, 28 March 2000). Those who had limited experience but were interested in construction were put on a full training programme of nine weeks by the KwaZulu-Natal Training Trust. Later in the implementation process it was noted that most builders were not setting up the houses correctly and this necessitated further training. The proportion of men and women trained is discussed below.

About 18.7 per cent of the household heads had been trained in construction. Of these, women comprised 33 per cent. From the discussions, it became clear that the criteria used to recruit builders focused on those who had prior experience in construction work and these were mainly men. Thus from the start the training had a strong male bias. The women's focus group reported that when the training began many women joined, but in the process most dropped out and of the original 33 per cent only one completed the training. Later, seven more women were trained in bricklaying by Kuphuka. The women who had been trained in construction said that they could not keep up with the training because of their domestic workload. Other women were simply not interested in learning to build because they would only use the skills in building their houses and that would be the end of it. They did not see themselves applying the skills beyond the project implementation phase.

The training manager at ALDT said that women were more interested in plastering than in bricklaying or any of their other courses (personal communication, Mr Campbell, 28 March 2000). He attributed this to their Zulu culture where the women's role in building the kraal is to plaster, and said that women who had been trained in plastering excelled in the trade to the extent that some advanced to become instructors. The women's lack of interest in bricklaying and general construction and their specific interest in plastering was commensurate with their role in building their cultural shelter. This points to a top-down approach in assessing the training needs of women from the communities during the delivery process, where their interests are not taken into consideration. Instead the primary objective appeared to be to impart the necessary skills for constructing the target number of houses. The comparison between women's and men's interest in continuing with construction trades seems to suggest that women are more interested in trades in which they have cultural experience and competence and which do not challenge the gender division of labour in either formal or traditional housing. Their gendered sense of self appears to influence their participation in training for construction work, as indicated by their lack of interest in construction tasks generally perceived to be men's work. The level of women's training in construction has an impact on the consolidation and maintenance of shelter and the long-term employment of women. With the right skills, women, like men, can compete for jobs in residential construction.

Women's training is closely related to their access to employment opportunities in the labour market. Their economic activities in Ezilweleni suggest that most worked in the informal sector (Table 2), selling in tuck shops, tailoring, hawking foodstuffs in the nearby schools and market, working as shop assistants, cleaners or machinists in the factories around Pinetown, working as domestics, operating shebeens and hair salons, and selling fruit and vegetables. Men's economic activities included working in factories around Pinetown, assisting in shops, and doing carpentry, metal work, building, painting, plastering, plumbing and electrical fitting work. These economic activities point to a distinct division of labour along gender lines. Table 2 shows that more men than women worked in the formal sector. Location in the labour market greatly influences the level of income and subsequently access to housing finance. The men's focus group, despite its expressed belief that men and women were equally poor, emphasised that women were more frequently poor because they had very few avenues for earning income. The general unemployment rate was about 25 per cent (18.7 per cent among men and 6.25 per cent among women) and most households were partly supported by incomes earned in the informal sector, remittances from children, or elderly relatives' pension grants.

Table 2: Type of occupation by gender in Ezilweleni

Gender	Formal employment	Informally employed	Unemployed not seeking work	Unemployed seeking work	Pensioner	Total
Male	31% (5)		6.25% (1)	18.75% (3)	6.25% (1)	62.5% (10)
Female	12.5% (2)	12.5%		6.25% (1)	6.25% (1)	37.5% (6)
Total	43.5% (7)	12.5% (2)	6.25% (1)	25% (4)	12.% (2)	100% (16)
Source: Ndinda, 2002.

Women's participation in the Ezilweleni project was confined to providing unskilled labour in the construction of their own houses, carrying the water, mixing the cement and clearing the sites before and after construction. It was in the construction of the crèche that those trained in bricklaying played a role. The untrained women were particularly committed to building the crèche, arguing it was for their children that they were doing the work, and therefore they were eager to build a good facility. Men were involved in their capacity as skilled workers and the few who did not have training learnt trades on the job, for example plastering and painting alongside the women. However, there was no differentiation in the construction of the community hall, with which everyone in the community identified.

With the subsidy, households were able to consolidate their shelter and to acquire additional facilities such as postboxes and a multipurpose community centre, funded by the remainder of the subsidy, after each household head had drawn the amount necessary for a core structure of two rooms (built by residents whose subsidies were approved between 1994 and 1996) and the residual from the IDT funds. The delivery system employed was a mixture of mutual self-help and contractor-built housing. The self-help part consisted of the subsidy beneficiaries providing their ‘sweat equity’, i.e. the household contribution to the labour in constructing their own house.

That women were worse off in Ezilweleni can be seen by comparing the level of housing consolidation among male- and female-headed households. Of the households sampled, 37.5 per cent were female-headed and the rest (62.5 per cent) were male-headed. As Figure 4 shows, about half of the households headed by women were ‘extended’ (that is, the households comprised the women, their children and other relatives), pointing to the diversity even among female-headed households in Ezilweleni. Whereas 40 per cent of the male-headed households had been able to improve their starter units from two rooms to five, only 16 per cent of the female-headed households had achieved this. Most significant is the fact that while in at least 10 per cent of the male-headed households sampled a member of the household was involved in the construction of the house, no members of the female-headed households were thus involved. It is therefore likely that female-headed households had to spend more on their labour costs since none of their members provided ‘sweat equity’.

Figure 4: Household types in Ezilweleni

The liberal feminist welfare perspective advocates women's equality of access to opportunities through state intervention (Tong, 1989: 12). Instruments such as affirmative action policies and equality legislation are aimed at achieving equity in training as well as in employment. Such enabling measures are in place in South Africa but, as the findings of this research suggest, equality in training has not been achieved at the grassroots level. It appears that societal attitudes about what is men's work and what is women's have to be addressed too, as well as the criteria used in recruiting candidates for training in construction work. Inequality in training results in a discrepancy between housing policy and practice in South Africa. Whereas the housing policy states that training in building is part of the housing delivery drive, women were not adequately targeted as a specific category, despite the overwhelming evidence that they were underrepresented in paid construction work in South Africa (CSS, 1998: 20; Edigheji & Maharaj, 1999: 86; Kallman, 1999: 90–1). Training in construction could provide women with useful skills for employment in the construction sector, which is better remunerated than the services sector where most women work. The sector needs to be gender-sensitive to embrace the diversity that women bring.

4.4 Between entrenched social practices and progressive gender policies

Through involvement in the project from its inception to the planning and execution phase, the beneficiaries were able to identify with and gain a sense of ownership of the project. The community confirmed this view, asserting that ‘It is the people who started the project and they saw it as their own. Since it was their own project they have continued to take a keen interest in the affairs of the Trust’ (Key informant, men's focus group, Ezilweleni, 21 March 2000).

Community involvement in Ezilweleni had arisen out of the people's own initiative. They had identified their problem as the lack of legal ownership of the land and set out to achieve this goal by forming a committee that reported directly to the community on how matters were proceeding. The need to address a community-wide problem created interest and the participants strongly identified with the cause. Yet this sense of ownership was not strong enough to override the gender divisions which prevented women from participating fully and men from enabling them to do so.

It was noted that the subsidy was insufficient to build the level of starter housing that the residents desired, a finding reflected in both the household questionnaires and the women's focus group discussions:

The subsidy is not enough. The community was only able to build starter houses – two-roomed houses, then one is supposed to increase on their own. The houses are about 30 square metres. Everyone chose their own design of house. (Single woman hawker from Pietermaritzburg, women's focus group discussion, 21 March 2000)

Although the women were dissatisfied because the subsidy could only provide a small starter unit, the men thought that it was sufficient. According to the women, households were supposed to extend the starter house with their own funds and this appeared to be a problem, particularly among the female-headed households. The battle to survive on low-status jobs with low incomes made increasing the size and improving the quality of their houses a challenge they were ill-equipped to meet (see Table 2 and Figure 2). The notion of women battling to survive is not new and has been highlighted in the literature, which suggests that in many cases where women are expected to participate in incremental housing most of them drop out of the projects because of inability to meet the financial demands (Machado, 1985: 15; UNCHS, 1985: 4–6; Vance, 1985: 24; Mjoli-Mncube, 1998: 118). The size of the houses in Ezilweleni contrasts with the size of those in projects such as Kutlwanong, developed in 1996 in Kimberley, where the same level of consolidation subsidy in a site and service scheme provided a 50 m² house comprising two bedrooms, a living room, a kitchen and a bathroom and toilet (Peer Africa, 1997: 42). While the department of housing has been allocating an extra 15 per cent of the subsidy to deal with topographical issues around the country, the larger size of the houses in Kutlwanong was the result of the community's innovative ways of cutting costs and sourcing extra financing in order to get a larger starter unit. This was not the case in Ezilweleni.

About 44 per cent of the residents in Ezilweleni said they had benefited from their participation in housing delivery and 43 per cent of these were women. The size of the houses varied according to household but most residents received a two-room starter unit. In discussing the outcome of their participation in the delivery process, Ezilweleni male residents argued that

Our houses are better and stronger because we ensure we have an inspection team to ensure that the contractors do a good job. The contractors know that if they do not do a good job they will not get paid. Other houses built by contractors from outside the community are very small. (Additional committee member, men's focus group discussion, Ezilweleni, 21 March 2000)

This observation is consistent with the literature, which suggests that when beneficiaries are involved in constructing their own housing the product is often of a better quality (Steinberg, 1992: 367). When the community participates, there is a sense of ownership and an assurance of good quality workmanship, because should a contractor from the community do a shoddy job he would face the consequences from the community, his own neighbours and relatives. The inspection team in Ezilweleni, which had been trained in construction through the training programmes mentioned above, ensured that the design, the strength of the structure and the quantity of materials were in accordance with the specifications in the building plan – playing the roles of architect, engineer and quantity surveyor.

5. Conclusion

Despite the theoretical recognition of equality of the sexes in the constitution, it is clear that women in Ezilweleni did not take advantage of this law to improve their status, even though some had training in construction. It is argued here that even with the legal provisions for gender equality, training and the opportunity to move to better remunerated occupations, the old ideology of male supremacy and female subordination remained intact, at least in this community and probably in many others, and was not addressed as an issue in the project design and implementation. This explains why some women, despite being trained, still reverted to poorly remunerated occupations in the informal sector and others dropped out because they could not cope with construction training and work.

While a radical transformation of the community may not be a viable short-term strategy for improving women's status in Ezilweleni, gradual reform of the leadership structures might do so. One strategy would be to have a rotating system whereby men and women would alternate in the role of chair, thus ensuring that both have an equal chance of making key decisions and having their views represented at the highest level of decision making in the community. The same strategy could be applied to the other leadership positions. This is because even where democracy exists it does not always result in equity. For equity to occur there is need for affirmative action as enacted in post-apartheid policies.

The tendency among women in Ezilweleni to revert to their gendered roles even after receiving training in fields that are dominated by men points to various problems. One is that people continue to perceive different kinds of work in gendered terms. It is essential that both men and women be educated about how the roles that they perform place them either in the formal or informal sector and the overall impact of this on their income levels. The ‘feminine’ gendered economic activities of women in Ezilweleni at present tend to take place in the informal sector and to be extensions of women's domestic roles: dressmaking, sub-letting rooms, selling in tuck shops and selling processed foods – all of these within their own impoverished community and therefore unable to generate substantial income. While the attitudes of employers in the formal construction sector may well need to be changed before women are employed in large numbers, no such change can take place so long as women fail to offer themselves for training. To change attitudes relating to gender requires the help of influential people and organisations in the area: civic leaders, churches, youth groups and women's groups. This education, of course, need not be formal – it can be informal, for instance adult education via print media and radio.

Additional information

Notes on contributors

Catherine Ndinda∗

∗ ∗Research Associate, School of Politics, University of KwaZulu-Natal,Pietermaritzburg.

Notes

∗Research Associate, School of Politics, University of KwaZulu-Natal,Pietermaritzburg.

¹Source: Trust Deed, Zilweleni Development Trust. ‘Zilweleni’ is the English version of ‘Ezilweleni’. The Development Trust was registered using the English version, but this study uses ‘Ezilweleni’, as that is how the residents referred to their area.

²A Zulu term that means ‘holding together’.