1.1 – The aspirations for social housing

7Social housing is often held up as a remedy for many of the problems facing cities and towns in South Africa, sometimes even a cure-all. This is apparent from many recent policy documents and ministerial statements (e.g. DHS, 2015; SHRA, 2017; Mfeketo, 2018; National Treasury, 2018). For example, social housing is seen as a potent means of promoting economic inclusion and upward mobility by offering people a safe and supportive living environment close to jobs. It is seen as a stepping stone which families will move through and into the private housing market as their economic circumstances improve, thereby releasing space for other needy households to occupy and advance over time. This makes it a valuable public asset that needs to be managed for the long-term by robust organisations with an enlightened agenda of supporting personal development, building socially-mixed and vibrant communities, and facilitating social and economic progress. The government also believes that many social housing organisations should be black-owned and controlled, and that women and young people should benefit from the jobs and other opportunities afforded by the growth of the sector (e.g. in property maintenance, security, gardening and cleaning) (SHRA, 2017).

8Social housing is also intended to serve a broader purpose in restructuring the segregated and fragmented spatial form of cities and towns. Indeed, it is the only policy instrument that is explicitly designed to support higher density housing in well-located urban areas (NASHO, 2018; National Treasury, 2018). This makes it quite distinct from the government’s other housing programmes, which are mainly concerned with large scale provision at relatively low cost. The social housing capital subsidy is larger per dwelling unit than the other housing subsidies, and expected to be aligned with efforts by municipalities and other public bodies to release and accelerate the development of vacant and under-used land in conducive locations. This is intended to help knit-together the disjointed urban fabric and improve the viability of public transport services by increasing ridership on key corridors that connect central cities to outlying townships and suburbs. A more compact and consolidated urban form should reduce the costs of public infrastructure and service delivery and moderate traffic congestion (National Treasury, 2018). It should also yield more functional human settlements and more productive cities that attract additional private investment and create more jobs. The government has a further expectation that social housing organisations should be involved in rehabilitating run-down inner-city buildings, which will generate positive externalities in surrounding areas and create beneficial economic, social and environmental outcomes (SHRA, 2017; NASHO, 2018).

9The Minister for Human Settlements captured the breadth of these aspirations in her foreword to the recent State of the Sector report (SHRA, 2017). She also referred to increasing confidence in the sector, widening tenure choice for households, creating sustainable and integrated human settlements, and attracting external investment to fast-track production and meet the growing demand for affordable rental housing in good locations. There was no acknowledgement of any tensions or dilemmas facing the providers of social housing, nor much clarity about the government’s top priorities for the sector:

10

“Social housing is … a key component of government’s commitment to … address structural, economic, social and spatial dysfunctionalities in South African society … Social housing is pivotal for … spatial transformation, public transport efficiency and urban inclusivity … critical for urban consolidation and redevelopment … unlocking private sector investment … offering social services and security … (and building) social capital”.

1.2 – Core propositions for investigation

11Clarity of purpose is important for any policy to be effective and enduring. It is not credible for policy-makers to ‘have it all’ in practice. Loosely-specified policies with lengthy objectives tend to get diluted over time and they lose whatever distinctive contribution they might have made if they had adhered to their original purpose.

12The goal of this study is to interrogate two particular propositions that lie at the heart of prevailing ideas about social housing. These propositions can be seen as constituting part of a ‘theory of change’ underpinning social housing policy. They have not been articulated in this form before, although there have been other theories of change formulated for social housing (e.g. DPME, 2016; Genesis Analytics, 2019). The propositions outlined here draw different strands together into a coherent set of ideas that resonate with the conventional wisdom about the role of social housing. One centres on the development of urban land and the other relates to the development of people.

13First, social housing is widely seen as a way to reverse apartheid spatial planning and post-apartheid residential sprawl by enabling urban integration through higher density, well-located housing. The proposition is that a concerted intervention in the land market can undo historic patterns of race-based population settlement. Such action can also counter prevailing property market trends which tend to distribute low income households to the urban outskirts, where land is cheap and plentiful. The policy requires identifying and acquiring strategic land parcels in places that are reasonably close to economic centres, social amenities and connecting infrastructure. Once the affordable (subsidised) housing is developed it needs to be protected and maintained for poor households to avoid gentrification and displacement by higher income groups. This is why rental housing is so important, and why it needs to be owned by the state or a non-profit organisation with broader objectives than maximising income.

14The core assumption is that the cost to the state of acquiring this land and property is not prohibitive because it is a worthwhile long-term investment considering the wider benefits to be gained from creating mixed communities living at higher densities, instead of segregation and sprawl. A related assumption is that social housing projects will spur further changes in the surrounding built environment and local economy through demonstration effects and tangible spillover effects. The new investment in social housing will boost confidence in the locality and raise land values, attract investment in shops and consumer services, and encourage property renewal and redevelopment at higher densities. This might also require complementary investment in the area’s infrastructure by other government entities.

15The simplest approach to launch this process of land-use intensification and regeneration is to focus on land that is already in public ownership to avoid the cost of acquisition, although maintaining a pipeline of land supply will require periodic acquisitions of new sites. It may also be advantageous to target inner urban areas characterised by current uncertainty and risk, where property values are depressed but there is scope for uplift. This would include run-down areas with vacant infill sites and existing buildings which are old and somewhat neglected, such as former industrial and warehousing districts in the zone of transition surrounding many CBDs. The lower cost of property acquisition would improve the financial viability of social housing projects. Over time a valuable asset would be created that could be leveraged to raise additional funding for new projects.

16The second proposition is that social housing is a means of lifting people out of poverty by providing them with decent accommodation and support services in places that are accessible to economic opportunities and social facilities (such as good schools and hospitals). This kind of intervention in the labour market may improve people’s chances of obtaining and retaining employment in two ways: (i) By moving closer to economic opportunities they will be better placed to compete for available jobs. They should be more attractive to employers because they will be more punctual and reliable, and they will be more motivated because they will save considerable commuting costs and time. (ii) Safe and supportive living environments will enhance their skills, social networks and awareness of job opportunities and employer expectations. Job-search support, childcare, financial advice and other supplementary services provided by social housing organisations should help to provide a stable foundation for people’s economic circumstances to improve.

17In addition, some tenants may take advantage of their new living environment to start their own enterprises by providing goods and services to meet the demands of the expanding community and improving local economy in the surrounding area. Group of tenants may also come together to form community organisations of various kinds (covering cultural, religious, recreational, sports and other activities) which further raise morale, develop capabilities, strengthen social ties and give rise to other constructive, income-generating initiatives.

18The selection of tenants for social housing is important to this proposition. They should be able to afford the rent but not have sufficient income to afford accommodation in the private housing market. Their position in the labour market is also important. At least one member of each household should be economically active in order to benefit from the locational opportunities. Some might also have ambitions to enter self-employment or to initiate a small business. In the interests of fairness and creating a mixed or balanced community, a minority of tenants might be retired, disabled or focused on bringing up children. Projects should also comprise mixed income groups and people from diverse racial backgrounds.

19These features of social housing are very appealing in the context of high unemployment, persistent racialised poverty and entrenched inequality. They offer a vision of a more inclusive and prosperous urban future. Hence it is not surprising if social housing has considerable symbolic attraction and is sometimes portrayed as a panacea. However, there is little consideration of the necessary conditions for promoting social housing on a large scale, the various risks and dilemmas faced, and the inevitable trade-offs associated with this form of housing. For example, can one assume that accessible urban land is available to develop medium-density housing at a viable financial cost? Such land is often in high demand from multiple users, even if it is in public ownership. Public bodies will have to sacrifice a financial windfall if they are to transfer their surplus land to social housing organisations at low cost. There are sometimes legal reasons why this is difficult and they may not support the logic of regenerating run-down areas through social housing. They may not have the resources to maintain a pipeline of land supply for social housing. There may also be substantial economies of scale associated with building larger numbers of social housing units on cheap peripheral (greenfield) land. Experience suggests that building on brownfield sites is a slower process subject to more complex negotiations, regulatory hurdles and bureaucratic procedures.

20The second proposition plays down the general lack of demand for low-skilled labour and the intense competition for available jobs in economic nodes. Moving people closer to these opportunities will not necessarily transform their chances of obtaining work or better paid jobs, especially if they face other barriers to employment, such as limited skills, disabilities or onerous domestic responsibilities. In these cases, huge costs may be incurred in accommodating households in better living conditions, but without the economic spin-offs of upward mobility and recycling the rental stock. Many more households could perhaps have been offered better living conditions if the project had been located on cheaper land? This is a serious consideration for political decision-makers, given the social and political pressures arising from the large housing backlog and desperation of hundreds of thousands of people for improved accommodation.

1.3 – The underlying problems

21A simple elaboration of the scale and nature of the urban housing challenge in South Africa is helpful to understand what lies behind the social and political pressures on the state to accelerate delivery. The high level of need or demand for housing partly reflects the growing urban population, combined with a decline in the average size of households over time. The number of households requiring accommodation is growing faster in cities than elsewhere because of rural-urban migration. Between 2001 and 2016 there were an estimated 210,000 new urban households each year as a result of migration (National Treasury, 2018). People are moving to the cities in search of jobs and livelihoods because of the dearth of opportunities in the rural areas and small towns. Most of the individuals and families requiring housing in the cities have low incomes. They cannot afford private housing in the formal market and private developers cannot afford to supply these households with proper homes because the cost of provision exceeds what they could afford to pay. Consequently, many of these households find shelter in the informal sector or they move in with family or friends. Most informal dwellings are insecure, unsafe, overcrowded, unregulated (in terms of rentals) and lack basic services, such as decent sanitation, clean water, refuse collection and electricity. The unsatisfactory nature of these solutions explains why many people look desperately to the state for assistance in providing affordable (subsidised) housing.

22The issue of location lies at the heart of the urban housing challenge. Apart from the history apartheid spatial planning, this is partly because the majority of low cost housing built by the state since the 1990s has been on the periphery, far from economic opportunities and social facilities. This reflects the steep land gradient in the city and the attractiveness of cheap land on the outskirts. Many of these settlements have been built at densities that are too low to support public transport and other important amenities. The subsequent growth of informal settlements in and around these townships, and of backyard dwellings within the townships, has reinforced the problem of poor accessibility to jobs and livelihoods. The ‘spatial mismatch’ between these overcrowded and underserviced residential areas and the areas of economic opportunity has increased the burden on poor households forced to travel long distances to work or to seek work. Up to 40% of their disposable income is absorbed by transport and up to 3-4 hours a day may be diverted from more productive uses, such as education, parenting or additional income generation. This undermines self-reliance, inhibits upward social mobility and deepens household dependency.

23Selected statistics from the situation in Cape Town help to illustrate the housing problem in other major SA cities. To begin with, there are only 10 formal houses in the city for every 17 households, resulting in a housing backlog estimated to be about 350,000 (CAHF, 2017; City of Cape Town, 2018). This is made up of 144,000 households in informal settlements, 75,000 in backyard shacks (both based on 2011 statistics) and the remainder in overcrowded or otherwise unacceptable housing conditions. The city council estimates that 35,000 housing units need to be built each year for the next 20 years to eradicate this backlog and also to cater for new demand that will emerge over this period (CCT, 2018). Between 1996 and 2007, the average number of formal houses built each year was 16,000, i.e. almost half the number actually required. The private market supplied 60% of these and government programmes 40%. However, over the last decade only 7-10,000 formal houses have been built each year (ibid). This reflects the economic slowdown, higher unemployment, credit constraints and the government’s main housing programme running out of steam. At the current rate of delivery using conventional methods, the CCT estimates that it will take more than 70 years just to eradicate the current backlog.

24The gap between formal housing supply and demand has been filled by informal housing in freestanding shack settlements and in backyards. While ‘only’ a third of the new housing provided in Cape Town in the 2000s came from the informal sector, this has since risen to about half. Interestingly, many backyarders are employed and pay a reasonable rent (between R 500 and R 2,500 per month). They constitute a natural source of demand for social rented housing because they fall into a gap between eligibility for RDP/BNG housing and qualifying for a formal bond from private financial institutions to buy their own homes. Moving to a well-located and more supportive social rented unit would free up valuable time and resources and might provide a major stimulus to advance their position in the labour market and enable them to move further up the housing ladder.

1.4 – The core research questions

25The broad purpose of this study is assess the effectiveness of government efforts to reduce poverty and inequality in the major cities through the provision of well-located housing. The study can be divided into two core research themes at the programmatic and project levels:

2.1 – The importance of housing policy to poverty and inequality

31Governments around the world face many challenges of providing adequate housing that everyone can afford. The global urban affordable housing gap is currently estimated to be 330 million households, and it is expected to grow to 440 million, or 1.6 billion people, by 2025 (King et al., 2017). The housing crisis is particularly acute in much of the developing world, where about one third of the urban population lives in informal dwellings with limited access to basic services, insecure tenure and far from urban economic opportunities and social amenities (ibid.). Living in such precarious conditions is both a consequence of and a contributor to urban poverty and social inequality. Sub-standard housing affects personal health and well-being and constrains socio-economic development. The rising number of people living in poor conditions presents a serious challenge to governments in achieving the sustainable development goals (UN-Habitat, 2012).

32Most governments formally recognise citizen’s right to adequate housing, as part of ratifying the Universal Declaration of Human Rights and the International Covenant on Economic, Social and Cultural Rights. This political and legal commitment to reducing housing poverty and inequality places important responsibilities onto the state (Clapham, 2019). In addition, there are many social, economic and political benefits from improving people’s housing conditions. Clapham (2019) distinguishes between three overarching goals of public housing policy: (i) to reduce harm to households, (ii) to correct externalities where the costs and benefits of individual producers and consumers do not reflect societal interest and (iii) to accommodate particular groups of the population neglected by market processes. Governments devise housing policies, defined as “any action taken by any government or government agency to influence the processes or outcomes of housing” (Clapham, 2019, p. 11), to address one or more of these goals.

33While households with higher incomes usually access housing through the formal private market, which is reasonably responsive to household preferences, many low-income households depend on government assistance to obtain adequate accommodation. Housing policy gives governments important powers to mitigate urban poverty and reduce socio-economic inequality. States differ considerably in the type of housing assistance provided, which is informed by the political-economic and socio-historical context of the country. The most common interventions include the provision of social housing, housing subsidies, tax incentives and vouchers, low-interest financing, and land, infrastructure and planning concessions to affordable housing developers (Clapham, 2019; Lund, 2017).

34Several factors influence the impact of housing on poverty and inequality. Location is one of the most important (Chetty and Hendren, 2018; Ryan-Collins et al., 2017; Turok, 2016). Good public policy integrates housing into well-located, human settlements in order to promote economic development and upward mobility. On the one hand, successful government interventions are about place-making and delivering houses in areas with social and economic infrastructure. On the other hand, it is about supporting households with more than a roof over their heads, i.e. with skills development, job searching and other active labour market measures. Related to this is a life-cycle view of housing, which recognises the different housing needs of households during their lifetime, including the option to choose between rental and ownership tenure. From this developmental perspective, well-located adequate housing can lift people out of poverty and contribute to more prosperous societies (Turok, 2016; UN-Habitat, 2012). Given that housing is an important element of personal wealth, and a major driver of social inequality, interventions in the housing market can promote equality and social justice in the city (Madden and Marcuse, 2016). At the same time, housing policy that treats housing merely as shelter and fails to consider its multiple socio-spatial dimensions can have limited or even detrimental effects on poverty and inequality (Turok and Scheba, 2018). Although housing is important in its own right, for its benefits to improve health, dignity and social protection, the issues of location and the labour market are fundamental in its contribution to economic development (UN-Habitat, 2012). The rest of this section considers these issues in the particular context of social housing.

2.2 – Social and public housing

35Social housing is an important pillar of housing policy in many countries. However, the sector is enormously diverse. Its main objectives, institutional arrangements, tenure forms, beneficiary eligibility criteria and social impacts vary widely in different places. Accordingly, there is no single, simple definition of social housing. Social housing is generally defined as government-subsidized rental housing provided by local authorities or non-profit organisations with the rent set well below market levels. Dwellings are allocated to tenants on the basis of their relative need rather than their ability to pay. This gives low-income households access to better quality accommodation than they could otherwise afford, with consequent benefits for their general well-being, health, safety and productivity. Despite the enormous diversity of social housing in different places, there are at least two common elements across all contexts.

36First, social housing is not delivered by the market, but allocated based on eligibility criteria that take into account income and other household characteristics (e.g. age, disability, ethnic group). Social housing as a broad category is thus equivalent to non-market housing and performs an important welfare function in society. Hence, social housing is often promoted as a kind of safety net that provides poor households with a secure and dignified living environment. Social housing can be built and managed directly by the state or by non-governmental organisations or for-profit housing companies. In the case of direct state provision, it is often called public housing.

37Second, social housing benefits from supply side subsidies in the form of grants to public or private developers to provide housing at a reduced rent or purchase price to the dweller. Government subsidies make the construction and management of social housing viable and allow landlords to charge rents below market price. This ensures affordability and allows low- and middle-income households to access decent housing despite their limited spending power.

38In some countries, social housing includes ownership or the option of buying the housing unit after a certain period. Social housing is usually restricted to rental as this is considered cheaper for low-income households and keeps the asset in public hands. In most instances, social housing is fundamentally about providing a welfare service to low- and medium-income households, rather than promoting economic mobility or urban restructuring. As such, its key objective is to provide permanent affordable rental accommodation. Whether this welfare service is targeted at the most vulnerable or wider segments of the population varies between countries and over time (Clapham, 2019; Scanlon et al., 2015).

39In countries such as Germany, Austria and the Scandinavian nations, it caters for large parts of the population. Here social housing is viewed positively as a tool that can help to achieve a wide range of societal objectives: higher quality housing stock, more affordable housing, social integration and neighbourhood development, compact urban form, environmental sustainability and social justice. In other countries, like the UK or Australia, it has evolved into a small residual sector that functions as the ‘tenure of last resort’. Here it is often viewed negatively and associated with it an inefficient, wasteful and counterproductive way of spending government money that patronises the poor and concentrates them into segregated neighbourhoods. The marked differences in experiences and perceptions of social housing in diverse contexts make generalisation and objective evaluation difficult (Clapham, 2019; Scanlon et al., 2015; Forrest, 2014).

40Different forms of social housing entail different degrees of government influence, depending on the mode of delivery and governance arrangements. A recent study found that in countries with a small proportion of social housing stock (less than 10% of the total) the state tends to administer it, while in countries with large housing stocks, non-profit companies or housing associations are more dominant (Genesis Analytics, 2019). In cases where the state directly provides public housing, it has full control over the rate of house-building, the physical quality of the stock, the rent levels and rules governing the allocation of accommodation to households in need. Yet, public housing has tended to become somewhat ossified over time and unresponsive to household needs and aspirations. Second, public housing has become a bigger financial cost to governments over time, and a bigger managerial burden. In an era of fiscal restraint and pressure for a leaner public sector, many governments have reduced direct provision of public housing and supported other organisations to deliver social housing (Clapham, 2019; Scanlon et al., 2015).

41Social housing that is built by non-governmental organisations has emerged as the dominant alternative, with both positive and negative consequences. Governments still accept key responsibilities, such as to underwrite finance, provide a subsidy, release land and infrastructure, provide social amenities, expedite planning approvals and regulate landlord-tenant relations. Yet smaller and more flexible housing institutions can be more nimble and responsive to household needs and aspirations. They can also be more enterprising and package funding from different sources, thereby saving government funding. At the same time, there are concerns over their growing dependence on private commercial lending and the consequences this has on the nature of housing provision (Forrest, 2014). These issues are explored in more detail in the next section.

2.3 – Social housing in the Global North

42Social housing has a long history in Europe (Scanlon et al., 2015). Its roots date back to the latter half of the 19th century, when the English government enacted the Housing of the Working Class Act of 1885 to close down unhealthy, slum-like settlements, and replace them with better quality, serviced alternatives. The first large-scale social housing project was built in London in 1900, sparking growing interest and support in other city councils. During and after the First World War, health concerns dominated the construction of social housing in the UK. Thousands of social housing units were built as part of the “Homes Fit For Heroes” campaign, which specifically targeted the health of urban conscripts (Clapham, 2019). In France, the Siegfried Act of 1894 founded the social housing policy and created la Société Française des Habitations Bon Marché (HBM, “cheap housing”). However, the scale of social housing remained limited until the 1950s, after which it gained increased government attention that accelerated delivery (Lambert, 2019).

43Social housing underwent a massive expansion across the developed world after the Second World War. In Europe it became a crucial part of the welfare state and the social contract between government, citizens and business (Scanlon et al., 2015). Significant government resources were spent on the sector, and in some Northern European and Scandinavian countries, social housing accounts for a sizeable share of the total housing stock - 32% in Netherlands, 24% in Scotland and 24% in Austria in 2011. Its significance is even greater in particular cities, e.g. more than 60% of Vienna’s residents currently live in social housing (Forrest, 2018). In France, the social housing stock comprises 4.8 million dwellings in 2017, which is one in six dwellings. It accommodates 10.7 million people and consumes almost half of total housing subsidies (Lambert, 2019).

44In these countries social housing is a mainstream tenure catering for a wide spectrum of people from low- and middle-income backgrounds. It enjoys a strong popular support and there is a wide consensus among policy-makers regarding its necessity (Scanlon et al., 2015; Forrest 2014). Nevertheless, the sector is not spared from challenges. In France, demand far exceeds supply with 2.1 million applications outstanding. In addition, the sector is faced with declining state subsidies, rising construction costs, aging housing stock, increasing residential segregation, declining economic mobility and a growing migration crisis. Many initiatives have been taken over the past 30 years to address these shortcomings, but they have usually fallen short of solving the problems (Lambert, 2019).

45In some countries, social housing is reserved for the poorest groups, which has further contributed to their marginalization, segregation and stigmatization. This has been particularly the case in the UK, where privatization, marketization and liberalization of the social housing sector since the 1980s resulted in substantial downscaling, disinvestment and growing challenges with maintenance and management. Pressure to reduce public spending, restrict direct government support and promote home-ownership led to the large-scale sale of units through right to buy agreements as well as prohibiting new construction in the sector (Clapham, 2019). These policies dramatically changed the nature and composition of social housing. For instance, the share of households in the top decile income distribution living in social housing plummeted from 20% in 1979 to close to zero in 2004-05 (Hills 2007 in Ryan-Collins et al., 2017). Social housing has a similar residual character in the USA and Australia.

2.4 – Social housing in the Global South

46A more positive picture of public/social housing is emerging in the global South, where governments of several countries have expanded the sector to accommodate more people. One of the best known and most admired is the Singaporean public housing system, which accommodates some 90 per cent of the population, almost all now as homeowners. Hong Kong is another exceptional example of public housing provision. The country known for its ‘freewheeling capitalist image’ has some 30 per cent of its population living in public rental accommodation, and the sector continues to receive substantial government investment and popular support (Forrest 2014; Chiu 2010). Social and public housing have expanded in other parts of Asia including South Korea, Malaysia and Vietnam. In contrast to European societies, housing became a central element of an ‘asset’ or ‘property-based’ welfare state of some developed Asian economies (Ronald and Doling, 2010).

47In Latin America, social housing programmes are most prevalent in Brazil, Colombia and Mexico, where governments in collaboration with the private sector and financial institutions built millions of homes for low- to middle-income households with formal jobs and permanent income. While these programmes have been successful in supplying a substantial number of new homes, most of them have been in peripheral urban locations with inferior access to employment, social amenities and social networks (Molina et al., 2019; Monkkonnen, 2018). The peripheral location of social housing projects has contributed to rather than mitigated against the concentration of urban poverty and spatial inequality (Libertun de Duren, 2019). A recent study of three social housing projects in three cities in Brazil, Colombia and Mexico found that location had a significant impact on the benefits and burdens to beneficiaries. The units located in the periphery recorded 40% lower housing values and 50% lower rental values than similar central units. Peripherally located households spent about twice as much money, three times as much time commuting to jobs and had fewer opportunities to access social networks and services (such as childcare). In addition, peripheral social housing projects suffered from social isolation, segregation and a lack of diversity (Libertun de Duren, 2019). There are many similarities to state-subsidised (RDP/BNG) housing in South Africa.

2.5 – Current international developments

48In recent years, the social housing sector has undergone important changes in many countries. Central and local governments have increasingly moved away from directly provision of social housing units. Non-profit housing associations have become important deliverers and managers of social housing projects (Murphy, 2019; Fields and Hodkinson, 2018; Wetzstein, 2017; Aalbers et al., 2017). In some contexts, the regulatory conditions and growing reliance on private finance have encouraged these organisations to employ market-like practices (Murphy, 2019). This trend influences the management of the housing stock and landlord-tenant relationships, which to replace a social developmental approach with a more contractual approach with less support to tenants that are increasingly being cast as clients (ibid).

49Some analysts argue that the increasing marketization and financialisation of social housing creates tensions between meeting financial obligations and delivering on their social mandate, which requires them to “reconcile often incompatible sets of rules” (Jacobs and Manzi, 2019, p. 9). Fields and Hodkinson (2018, p. 2) argue that “Whereas social housing once counterbalanced the failings of the private market and effects of recession, its funding model now both subjects social housing to market volatility and makes it likelier to intensify such volatility”. Critics of this trend predict a growing policy-outcome gap between social housing policy objectives and the actual outcomes for society (Fields and Hodkinson, 2018; Wetzstein, 2017).

50To conclude, social housing remains one of the most common government interventions to mitigate housing challenges, urban poverty and inequality, but the extent to which it achieves these objectives varies across contexts and depends on various factors, including location. Social housing is essentially a non-market, subsidised form of housing that is delivered to segments of society based on eligibility criteria that vary in different contexts. Institutional designs, tenure forms and delivery mechanisms also vary greatly. Recent years have seen important changes to the sector (with exceptions), most importantly the decline of direct state provision, a growing role of non-profit social housing institutions and their increased reliance on private funding due to public funding constraints. Social housing is often considered to perform an important welfare function in society, by providing low and middle income households with secure and decent accommodation, but there is growing recognition of the potential economic benefits of well-located social rental housing. These may be undermined if housing projects are not planned and implemented sensitively. The next section discusses the contribution of social housing to poverty reduction and social inequality in the context of South Africa.

3.1 – Housing policy post-apartheid

52Housing was a priority of the first democratic government, which recognised its significant role in transforming a deeply divided society (Jones and Datta, 2000). The apartheid government segregated people according to race and confined the black population to substandard dwellings in overcrowded townships and informal settlements far from jobs. Meanwhile, whites were subsidised to live in well-serviced suburbs with generous space standards and good public facilities (Lanegran and Lanegran, 2001; Mackay, 1999; Wilkinson, 1998). Providing decent housing for the poor was a prominent election promise of the African National Congress (ANC) and became a central plank of its overarching Reconstruction and Development Programme (Jones and Datta, 2000). The challenge was formidable with a housing backlog of 1.5 million units, growing at 200,000 new households a year, plus 450,000 households in hostels that required upgrading. This was aggravated by the need to transform the unjust and inefficient policies and institutions that the government inherited (Department of Housing [DHS], 1994).

53The government’s first housing policy was released two months after the 1994 election. The Housing White Paper (DHS, 1994) followed two years of negotiations within a multi-party National Housing Forum. This brought together stakeholders from the public and private sectors and civil society to formulate solutions to the housing crisis. The new policy emphasised partnerships between spheres of government, the private sector and communities as a “fundamental prerequisite for the sustained delivery of housing at an unprecedented level” (DHS, 1994, p. 4). An enabling environment had to be created “conducive to attracting the necessary private investment, both of the household as well as that of the institutions” (ibid., p. 4). Recognising the difficult economic context (low growth and high unemployment) and high inequality, the policy called for ‘people-centred’ development with housing seen as a way to grow the economy and restructure society. Given the high levels of need and limited state resources, support would focus on the poorest households, and sizeable contributions would be expected from people themselves, the private sector and financial institutions.

54The White Paper recognised that housing is more than shelter: “A housing programme cannot be limited to housing, but needs to … give meaning to the goal of creating viable communities” (DHS, 1994, p. 10) and “government strives for the establishment of viable, socially and economically integrated communities, situated in areas allowing convenient access to economic opportunities as well as health, educational and social amenities” (ibid., p. 19). The policy also identified the important role of land and integrated planning in creating sustainable communities, arguing that a new approach was required to make this a reality: “Efficient assembly and release of appropriately located land for housing is critical to achieving the desired rate of delivery of housing” (ibid., p. 29).

55Specific interventions were proposed to stabilise the housing environment and support poor households to get better accommodation. The principal component was the introduction of a project-linked capital subsidy, which became the cornerstone of housing policy (Charlton and Kihato, 2006; Huchzermeyer, 2001). This was paid to private developers to build small housing units for people earning under R 3,500 per month. According to the Housing White Paper, the projected monthly household income distribution was as follows (DHS, 1994, p. 8):

56The subsidy was supposed to vary on a sliding scale depending on household income. Some argued that the objective was to increase property ownership among the black population, and create an asset-based society linked to the financial sector (Marais and Cloete, 2017; Mackay, 1999). Some National Housing Forum (NHF) participants argued for a mass rental housing programme as well, but this was rejected because of concerns over the administrative and financial burden on the state (Lanegran and Lanegran, 2001; Tomlinson, 1999; Watson and McCarthy, 1998). Local authorities were already struggling to manage their existing rental housing stock because of rent boycotts, capacity constraints and perceptions of unfair beneficiary selection (Tomlinson 1999). Rent boycotts had become an important resistance strategy under apartheid (Watson and McCarthy, 1998). There was also popular pressure for home ownership as a way to create wealth among blacks (ibid).

57The Housing Act was passed in 1997, bringing housing policy in line with the new Constitution, which enshrined access to adequate housing as a socio-economic right. Meanwhile, the ANC government set a bold target to build a million houses in its first five years. The resulting preoccupation with quantity had a profound effect on the way the new housing policy was implemented. The Housing Act recognised multiple dimensions of the right to housing and sustainable neighbourhoods, but the political pressure to deliver large numbers favoured mass provision of uniform units on cheap peripheral land, thereby relegating other forms of housing tenure and neglecting the importance of location. The consequences for entrenching urban inequality and creating new poverty enclaves have been profound (Turok and Scheba, 2018; Huchzermeyer and Karam, 2016; Charlton and Kihato, 2006).

58Between 1994 and 2003, more than 1.5 million housing units were constructed (Gardner, 2003 in Tissington, 2010). This rate of delivery was unprecedented internationally, but criticisms soon emerged over the quality, location and socio-economic impacts (Hendler, 1999; Rubenstein and Shubane, 1996). Instead of creating sustainable communities, the scheme created monolithic dormitory settlements on the outskirts, which reinforced the dysfunctional form of South African cities and exacerbated spatial inequalities (Turok, 2016b; SACN, 2016; NPC, 2012; Tissington, 2010). Community participation was limited and the houses contributed little to lifting people out of poverty beyond the shelter and security aspects (CDS, 2015). Very few people obtained mortgage finance to invest in improving their properties (Marais and Cloete, 2017). The unfavourable location and high transport costs also prevented people from gradually upgrading and consolidating their homes, and resulted in many selling or renting them out, thereby creating an informal secondary market where sale prices were often well below construction costs (Huchzermeyer and Karam, 2016; Charlton and Kihato, 2006; Jonas and Datta, 2000). The limitations of the Reconstruction and Development Program (RDP) programme led eventually to policy revisions, discussed in section 3.3.

3.2 – The institutional subsidy: the beginning of post-apartheid social housing

59While public debates and government policy squarely focused on promoting homeownership, a small group of housing experts began to advocate for rental as an important alternative form of tenure. A concept note was developed, which laid the basis of the institutional subsidy that was introduced together with the individual and consolidated subsidies in 1995. The institutional subsidy provided funding to institutions to develop and manage affordable housing with forms of tenure other than immediate ownership (i.e. rental, instalment sale, shareblock and cooperative) [1]. The subsidy was given to institutions, but it allowed people to take effective ownership of the unit after renting it for a minimum of four years. This marked the beginning of social rental housing after 1994, albeit the focus was still on promoting individual ownership (Rust, 2006; Wicht, 1999).

60The Housing White Paper developed in 1994 had acknowledged the importance of social housing: “Government believes that it has a significant role to play in actively developing and supporting new and innovative approaches to social housing, especially where these provide for the self-management of housing stock within the ambit of such institutions” (DHS, 1994, p. 40). Nevertheless, there was no effort made to formulate a social housing policy, and critical issues were not addressed, including defining the specific role of social housing, developing an institutional and regulatory framework, identifying ways of managing the housing stock and devising suitable funding instruments. There was also no strategy put in place to implement social housing projects (Cloete et al., 2009). The government introduced the institutional subsidy, due to pressure for some kind of rental support from housing experts, without establishing a strong institutional and policy framework regulating the social housing sector.

61Despite the policy vacuum, not-for-profit organisations began to use the institutional subsidy to develop social rental housing project-by-project. Many combined the subsidy with loan funding from the National Housing Finance Corporation (state-owned finance institution) and overseas donor funds (especially from the European Union; Dutch and Canadian governments) to ensure viability. Some schemes focused on regenerating inner-city neighbourhoods through affordable renting, especially in Johannesburg (Mackay, 1999). Other social housing institutions (SHIs) saw the subsidy as a stepping-stone to individual homeownership. Households were expected to obtain mortgages from banks after the first four years of renting. The SHIs saw this period as helping low-income households overcome the barriers to mortgage finance by demonstrating a track record of monthly payments (Scheba and Turok, 2018). However, many projects failed to link people to banks, so most of the stock remained in SHI ownership. They retained responsibility for property maintenance, ongoing management and rent collection. This created many problems in due course, resulting in bankrupt SHIs and decay of the housing stock (Scheba and Turok, 2018). For example, many projects encountered serious financial difficulties, caused by rent boycotts, poor governance and/or inadequate subsidy arrangements. Others suffered from inexperienced management, political interference or inadequate policy support. Research by the Social Housing Regulation Authority (SHRA) found that 12,168 of the 19,630 units built between 1996 and 2006 failed, were in distress or were not viable (they had to be supported by municipalities or donors). This was 62% of all the units financed through the institutional subsidy (Scheba and Turok, 2018).

Figure 3.1

Financial situation of institutional subsidy units produced between 1996-2006

Financial situation of institutional subsidy units produced between 1996-2006

62The need for a comprehensive social housing framework to address these shortcomings became increasingly apparent in the late-1990s (NASHO, 2016; Wicht, 1999). In 1997 the Social Housing Foundation was created out of the development arm of the National Housing Finance Corporation (NHFC) to build capacity and provide technical assistance to existing and emerging SHIs (NASHO, 2016). The Foundation also pushed for a coherent social housing policy. The formation in 2002 of the National Association of Social Housing Organisations (NASHO) as the representative body of many SHIs was also important. These and other bodies helped to develop a more serious social housing policy, which became a reality after national housing policy was overhauled in 2004.

3.3 – Breaking New Ground: A new era for social housing

63The government’s new housing policy was called Breaking New Ground (BNG). BNG reflected on the first decade of housing delivery, and supposedly drew upon extensive stakeholder consultations and research, although some have disputed the extent to which evidence actually influenced the new policy (Charlton and Kihato, 2006). Nevertheless, BNG recognised a slowdown in RDP house-building, the inadequate quality and location of the housing, limited impact of housing on poverty and inequality, the changing nature of housing demand and the limited contribution from the financial sector. The negative effects of peripheral location of subsidised housing were highlighted early on in the document:

64

“The lack of affordable well located land for low cost housing resulted in the housing programme largely extending existing areas, often on the urban periphery and achieving limited integration. Post-1994 extensions to settlements have generally lacked the qualities necessary to enable a decent quality of life”.

65In response to these shortcomings, the policy announced a more holistic, progressive and responsive approach to create ‘sustainable human settlements’. Housing was to have an explicit role in promoting more compact, efficient and equitable cities and towns. BNG stressed the importance of:

• housing quality over quantity,
• a variety of tenure options,
• community participation in the design and implementation of projects,
• a greater role for local government and the private sector,
• the need for social and economic infrastructure linked to housing, and
• the role of housing in social integration and urban restructuring.

66It stated:

67

“The new human settlements plan moves away from the current commoditised focus of housing delivery towards more responsive mechanisms which addresses the multi-dimensional needs of sustainable human settlements. This approach is intended to provide maximum flexibility and will ultimately enhance the mobility of households”.

68The new policy aimed to expand government’s mandate to cover the entire residential housing market to support poor and lower middle-income households. It called for new approaches to spatial planning, land-use management and private-sector contributions to deliver sustainable human settlements. Integrating previously excluded groups into the city was viewed as essential to develop sustainable human settlements. The policy encouraged densification coupled with fiscal incentives and residential development permits (basically an inclusionary housing policy), alongside more decisive intervention in land markets and the creation of a single overarching planning authority. The policy recognised the “slow and complex process” of identifying, acquiring, assembling and releasing state-owned and private land for housing. It announced the development of a new strategy to fast-track the release of well-located public and private land for municipalities at no or reduced costs, funded through a separate funding mechanism. BNG also gave more emphasis to upgrading informal settlements and integrated residential development projects. However, many of the policy’s recommendations have not been followed through in practice, which means that well-known problems continue to persist (Turok and Scheba, 2018; Huchzermeyer and Karam, 2016).

69BNG highlighted the critical role of social housing in achieving sustainable and integrated human settlements. The policy envisaged a strong role for social housing in contributing to urban restructuring and supporting inner-city regeneration. According to the policy: “Social Housing … may make a strong contribution to urban renewal and integration. Social housing interventions may also be used to facilitate the acquisition, rehabilitation and conversion of vacant office blocks and other vacant/dilapidated buildings as part of a broader urban renewal strategy” (DHS BNG, 2004, p. 15). BNG recognized the potential of social housing to reducing poverty and inequality by “enhancing flexibility and mobility” of people and creating a “non-racial, integrated society” (DHS BNG, 2004, p. 19). The policy acknowledged the need for strong social housing institutions that hold social rental accommodation in public hands so that more than one generation can benefit. It announced the creation of a new funding instrument and an intention to scale up social housing delivery to 110,000 units over the following four years. It also promised measures to build capacity in the sector and establish a regulating authority to administer and monitor accreditation of SHIs.

70The Social Housing Policy was announced in 2005, providing a dedicated institutional and regulatory framework for the sector for the first time. The Housing Code was passed in 2007 (revised in 2009) and the Social Housing Act in 2008. This made the new national subsidy programmes compliant with BNG. These laws formed the new social housing policy framework, which aimed to fill the policy vacuum and address the many shortcomings that previously existed in the sector. The new policy supposedly presented “an aggressive and bold indication of government’s commitment to making the social housing sector work because of the benefits that it brings to the country” (DHS, 2008 p. 7). In developing this framework, the government drew on international funding and expertise, e.g. from Europe and Canada (NASHO, 2016). There were many laudable intentions and progressive aspirations expressed, but the devil was in the detail of its content and practical implementation, as we shall see. A visual overview of the social housing institutional environment is provided in Appendix 1.

3.4 – Social housing policy and programme

3.5 – Recent policy reforms and sector developments

83Several important policy and legislative changes have affected the social housing sector in recent years, including the Presidency’s Outcome 8 delivery targets (2010), the National Development Plan (NDP) (2012) and the Medium Term Strategic Framework (MTSF) (2014). The MTSF set an ambitious target to deliver 27,000 social housing units between 2014-2019. This became a policy imperative driving accelerated delivery. Coincidentally, an evaluation of the Social Housing Programme carried out in 2014-2015 concluded that nearly 10,000 rental housing units had been built and that this represented value for money (DPME, 2016). The SHRA’s State of the Sector Report (2016) also stated that “the sector is meeting its core critical objectives under the Social Housing Act”. However, it also acknowledged that it is:

“falling short of the scale of delivery required to truly consolidate its gains and make a substantial and lasting contribution to the restructuring of the country’s urban landscape, and to the economic revitalisation of inner cities and urban nodes”.

Figure 3.2

The economics of operating a new social housing unit: 2008 and 2015 compared

The economics of operating a new social housing unit: 2008 and 2015 compared

84There were also serious problems with the limited oversight and support of SHIs provided by SHRA (DPME, 2016; NASHO, 2016). Furthermore, the Department of Planning, Monitoring and Evaluation (DPME) evaluation recommended that the restructuring zones be defined more narrowly so that the housing projects would have a more concentrated and therefore catalytic impact (DPME, 2016).

85During 2017, the Department of Human Settlements made four significant changes to the social housing programme to address the delivery crisis. First, they revised the income bands for beneficiaries to allow for inflation. The new upper income limits of the primary and secondary markets were raised to R 5,500 and R 15,000 respectively. So the social housing programme could now support households with incomes between R 1,500 and R 15,000 per month. Second, the RCG grant was raised from R 124,000 to R 155,000 per unit.

86The Department also combined the institutional subsidy and RCG into a single funding stream – the ‘consolidated capital grant’ (CCC). The responsibility for administering the CCC was transferred to the SHRA. In announcing the revised income bands, the minister reiterated the supreme importance of ensuring low income households could access urban opportunities. She stated that the social housing programme should “continue its focus on including the primary target market in well-located projects, so that the poor are integrated into cities” (SA News, 2017).

87Third, they added 138 additional restructuring zones (located within six provinces and 38 municipalities) to the 127 existing ones. This extended their coverage into intermediate cities and smaller towns (SHRA, 2017). The effects of this are discussed further below. Fourth, the SHRA underwent major restructuring, appointed new leadership and implemented far-reaching organisational improvements. It introduced new rules regarding the accreditation of SHIs and entered new partnerships with government bodies (e.g. Housing Development Agency, City of Johannesburg) and other sector stakeholders (e.g. NASHO and Canadian government) to improve the performance of the sector (SHRA, 2017).

88The social housing sector also benefited from wider legislative reforms and policy changes that aimed to promote urban restructuring. The Spatial Planning and Land Use Management Act (SPLUMA) (2013) requires municipalities to promote the principles of spatial justice, inclusion and sustainability in planning and land-use decisions. The National Treasury introduced Built Environment Performance Plans (BEPP) as a requirement for metropolitan municipalities to align their capital budgets with their plans for urban integration and spatial transformation. The metros designate Integration Zones and Transit Oriented Development precincts in their BEPPs, which create opportunities to align with social housing restructuring zones and private sector investment. The metros are also obliged to identify catalytic land development programmes in their BEPPs. Meanwhile, the Department of Human Settlements required municipalities to identify social housing as part of new catalytic human settlements projects (also known as mega-projects).

89Beyond this, the NDP and the Integrated Urban Development Framework (IUDF) identified social housing as an important mechanism to create more compact, efficient and equitable cities. At international level, the New Urban Agenda added political support to the provision of social housing. According to SHRA’s State of the Sector Report: “In arguing for greater and integrated investment in spatially well-located areas to increase urban amenity access by low and moderate-income households, the NDP and IUDF highlight the importance of rental housing and specifically, social housing” (SHRA, 2017, p. 32). Social housing is clearly attracting considerable support as a way to achieve important urban policy objectives. Many observers expect social housing to achieve even greater prominence in current government deliberations about the future direction of housing policy (SHRA, 2017).

90In parallel to these policy developments, there has been growing civil society pressure in cities to deliver social housing, along with heightened media interest. Some of the media attention has been sparked by the resistance SHIs have faced from tenants in the context of rising rent levels and occasional evictions for repeated non-payment. Other interest has been sparked by the broader issues of social justice and spatial transformation. In Cape Town, for example, the Reclaim the City movement has transformed public awareness and support for social housing in the inner city (NASHO, 2018). It has used all sorts of methods to advocate greater public investment in well-located affordable housing, including research, formal engagement with city officials and the private sector, symbolic protest action, occupation of vacant and under-used land and buildings, commentary in the mainstream media and social media, and legal action against public authorities and private developers. These pressures have had important political and practical consequences as well, including rifts and disagreements within the administration and delays in the commencement of new property developments because of physical and legal obstruction.

3.6 – Key challenges and tensions

91The social housing sector faces many challenges and dilemmas in expanding its contribution to tackling the housing crisis. Five of the most important are outlined below.

3.6.1 – Scaling-up delivery vs budgetary constraints

92One of the greatest challenges is to accelerate the delivery of housing units, specifically in areas of greatest demand. Over the MTSF period, 13,958 units have been delivered (including in 2018-2019), which falls far short of the target of 27,000 units (SHRA, 2019). It is also much lower than initial policy expectations of 50,000 units within the first five years of the programme. Some experts actually suggested that between 100,000 and 150,000 units were needed by 2020 to make a substantial impact (NASHO and AFD, 2011). While government implemented important legislative changes and sector reforms to boost delivery (including the increase of subsidy quantum, adjustment of income bands and reorganisation of SHRA), these actions appear to have been insufficient to create the conditions for sustainable growth of the sector. Even though the subsidy was increased to R 155,000 in 2017, the additional amount was below actually incurred inflationary increases and, more importantly, was again not indexed to future inflation (SHRA, 2017). With construction costs increasing above inflation, the current benefits will have diminished in a few years’ time, which is very likely to stop delivery again. Meanwhile, the allocated budget is not enough to fund the 27,000 units planned for this MTSF period.

93Financial constraints are the main reason for government’s hesitation to link subsidies to inflation and periodically adjust the income bands. Social housing is but one government mechanism aimed at mitigating the housing crisis in the country. The emphasis is still overwhelmingly on promoting homeownership through integrated residential development projects and mega-projects. Another increasingly important policy focus is informal settlements upgrading. The MTSF target of 27,000 social housing units is very small and only makes up 3.6% of the overall goal of creating 745,000 housing opportunities in this period. Similarly, social housing receives only 2% of the total human settlements budget. Worsening fiscal constraints and a reduced human settlements budget threaten the sustainability and growth of the sector (SHRA, 2017).

Figure 3.3

Social housing as percentage of total human settlements budget

Social housing as percentage of total human settlements budget

94National government is hoping for the private sector to take a leading role in unlocking investment. Social housing institutions and other delivery agents are expected to develop new innovative public-private partnership models and enter into financing arrangements with commercial lenders to increase delivery. Mega-projects (catalytic human settlements projects) are increasingly being pushed to deliver social housing on a large-scale. The National Department of Human Settlements requires municipalities to identify catalytic projects that accommodate a mix of tenure forms, including social housing. Using different tenure options the department seeks to deliver “mega, high-impact integrated and sustainable human settlements that clearly demonstrate spatial, social and economic integration in close partnership/collaboration with provinces, municipalities and private sector partners” (SHRA, 2016, p. 39). Another prominent suggestion is to develop inclusionary housing policies to set aside affordable rental in new private-sector led developments. Evidence suggests that other delivery agents from the private sector are indeed becoming increasingly important in the delivery of social housing units (SHRA, 2017).

95Until now, SHIs have been dependent on government grants and financing. Government subsidies in the form of RCG and IS (nowadays CCG) cover the majority of capital costs, on average 70%, and are thus vital for the viability of projects. Most SHIs do not have large equity to invest in the creation of new stock. Own equity contribution by SHIs is limited to 3% of total capital costs on average across the sector (SHRA, 2017). In addition, with a few exceptions, SHIs struggle to obtain loan finance from commercial banks (SHRA, 2017; DPME, 2016). Emerging SHIs find it extremely difficult to obtain commercial loans due to their inadequate track record and limited collateral (SHRA, 2017). Commercial lenders remain reluctant to invest in the sector due to following concerns: (i) no reliable funding commitment by government, (ii) a negative perception of the regulation of the sector; (iii) different sources of funding complicate managing the different funding streams; (iv) legislated limitations on transfer and disposal of social housing stock in terms of default; (v) rentals revert to the initial levels when tenants vacate the units eroding the real value of social housing rental income streams over time, and (vi) some SHIs are not deemed financially sound due to weak financial management controls (SHRA, 2017).

96The bulk of debt finance is therefore provided by government-owned banks such as the NHFC and Gauteng Partnership Fund (SHRA, 2017). Growing the social housing sector requires government support in the form of grants, loan finance and capacity building. While fiscal constraints and budgetary realities are acknowledged, scaling up delivery to levels where it can make a larger-scale impact will require more substantial funding and commitment to periodically adjust subsidy amounts and income levels to inflation. Aside from the national department, provincial and local governments are called on to support SHIs by providing incentives and reductions in the cost of municipal services. Without public financial support, they will not be able to manage units viably, which is a key expectation outlined in the policy.

4.1 – Data and methods

111There are two main challenges in reviewing the impact of the social housing programme to date. First, there is conceptual complexity contained within the three core objectives: spatial transformation (i.e. urban restructuring through compaction, integration and/or connectivity), social advancement (i.e. progress through upliftment, employment, access to services and/or racially-mixed communities) and economic revitalisation (i.e. urban regeneration through attracting private investment, stimulating small business development and reversing urban decay). There are differences of emphasis and subtle trade-offs between these objectives that impinge on which groups of tenants should be targeted, where projects should to be located, and the scale of delivery that is feasible.

112The second challenge is one of practical measurement of the impact, given the imprecise and inconsistent nature of these objectives, and the need to collect and collate a wide range of socio-economic indicators at the level of the individual tenant, the housing project, the neighbourhood and perhaps even the city as a whole. This task is hindered by the uneven availability and quality of data across different metros, as well as major gaps in data for some core indicators. For example, the lack of systematic information on the spatial distribution of firms in SA cities makes it difficult to assess the proximity of social housing projects to jobs. Yet this is a fundamental objective of social housing policy.

113A comprehensive evaluation of the nature and impact of social housing is beyond the scope of this paper. Instead, we provide a review of programme outputs and the location of social housing projects, drawing on available secondary sources. We begin by analysing programmatic information collected by the SHRA covering the last decade, which splits into two electoral terms (i) from 2008-2009 – 2013-2014 (linked to the strategic plan known as the Medium Term Strategic Framework (MTSF)) and (ii) from 2014-2015 – 2018-2019 (excluding the last two quarters of 2018-2019). This involves a review of some of the key features of projects delivered to date in terms of their size, budget, location and institution.

114We then undertake a more detailed assessment of the spatial distribution of social housing projects across the metropolitan municipalities of Cape Town, Johannesburg, Ekurhuleni, Tshwane, eThekwini, Nelson Mandela Bay and Buffalo City. The first task was to compile a comprehensive database of relevant projects. We began with a curated dataset from the HDA/NASHO covering the period until 2016 and then updated and triangulated it against records reported by the SHRA. The database only contains social housing projects geared towards rental tenure, in accordance with the objectives of the social housing policy. It may be misleading to report on social housing stock which has long since been sold or otherwise transferred into the hands of private households. This means that even fairly recent rent-to-own schemes such as Morgans Village or Harmony Village as managed by the Cape Town Community Housing Company are not included in the database.

115The final database we produced was reviewed by NASHO in consultation with their membership, which provided an additional layer of audit and quality control. Overall, we believe that the database is up-to-date and reliable although there may be a few discrepancies in geo-coding for smaller SHIs. In the interests of transparency, the database is included in Appendix 3. The database contains records of 185 projects in the metros as well as a few secondary cities which have also benefited from social housing, such as Sol Plaatje (Kimberley) and Msunduzi (Pietermaritzburg). This is a huge improvement on the database produced by the HDA/NASHO in 2013 which had geo-coded information on only 32 projects. NASHO is working with the SHRA to see how this database could get institutionalised in order that the records are continually updated.

116A methodological issue is how to identify whether individual projects have contributed to the objectives of spatial transformation, social advancement and economic revitalisation, at least in terms of their geographical location. We build upon the classification systems used in previous studies (see NASHO and HDA 2013; Genesis Analytics, 2019) to allocate projects in the geo-coded database into seven location types:

• Core city: The area within approximately 5km of the city centre. It maximises access to employment and related facilities such as training colleges. It includes older and somewhat neglected industrial and warehousing districts surrounding many CBDs, where property is cheaper. Social housing has the potential to help revitalise these areas through infill development on brownfield sites and rehabilitation of vacant, derelict and ageing buildings (especially in and around Johannesburg CBD). Major outlying business nodes are also included in this category if they are long-established and showing signs of significant decay.
• Inner suburbs: Predominately residential districts in the historic (and formerly white) parts of the city with good access to public services and public transport nodes in areas typically surrounding the core city. Inner suburbs are well-established areas of relative prosperity. They are potentially open to spatial transformation and a change in the social composition through social housing infill development (and perhaps more intensive land-uses, such as apartment blocks replacing free-standing houses) that provides greater tenure diversity and housing choice for low- and middle-income households.
• Outer suburbs: Predominately residential districts further away from the city centre with fewer public services and poorer public transport connections. They have limited potential for spatial transformation and social advancement because of their peripheral location, high cost of land and poor connectivity. Any projects are likely to be on greenfield sites.
• Grey zone: These are greenfield areas left vacant under apartheid or with marginal industrial development typically situated between historic ‘white’ suburbs and outlying ‘black’ townships. Their potential for large-scale development depends on concurrent investment in bulk infrastructure and transport connectivity to enable major ‘new townships’, often comprising mixed tenures and mixed income - everything from RDP/BNG units to rental apartments and freestanding open market bonded housing (e.g. Fleurhof in Johannesburg).
• Greenfield mega-project: These are undeveloped areas on the periphery requiring large-scale public investment in bulk infrastructure and transport connectivity. Cheap land is a key attraction, along with the ability to generate economies of scale in the production of mass housing. Some may be accessible to outlying industrial areas (e.g. Cosmo City in Johannesburg and Cornubia in eThekwini). There is little potential for spatial transformation and economic development may take many years to materialise.
• Established township: Former ‘black’ settlements generally located on the outskirts of the ‘white’ cities. They are generally far from economic nodes and their bulk infrastructure and other public services are often inadequate to cope with fast-growing populations, many of which are in crowded backyard structures. Social housing may meet purely housing objectives, rather than spatial or social transformation or economic revitalisation goals.

117The results of applying this typology to the social housing projects are shown in section 4.3 below. Observations are also made about the Restructuring Zones that municipalities have designated for social housing projects to qualify for Restructuring Capital Grants. The coding of locational typologies is open to interpretation and in some instances would require additional information about precise locations and boundaries. For example, some of the peripheral projects could just be an incremental extension to an existing township, or they could be part of a new mega-project, or even in a grey zone. However, we also provide maps which give a good impression of where social housing projects are spread across each metro.

118The findings are useful, although they also reveal certain limitations of this approach. The categories are broad and it is not possible to compare the merits of different projects within the same category. A more sophisticated approach would be to develop criteria to enable each project to be assessed more systematically. This seems to be the direction preferred by the SHRA in advocating a movement away from predetermined restructuring zones towards site-specific assessments.

4.2 – An overview of delivery

119A decade has passed since social housing was institutionalised in SA and the SHRA was established in 2008. This is an opportune time to reflect on what has been achieved so far, and to identify patterns and emerging trends in the scale, cost and spatial distribution of social housing units. The annual SHRA performance and financial reports provide a regular source of information on the development of the sector. The SHRA’s self-reporting has improved steadily over the time and it has also added periodic ‘State of the Social Housing Sector’ reports to offer a wider account on the sector. These are positive signs of growing institutional capacity.

120We offer an independent review of information released in such reports. Figure 4.1 shows that the construction of new social housing units during the 2008-2009 to 2013-2014 MTSF period initially fell well below expectations and achieved less than 2,500 units against a target of 12,800 units [3]. Delivery has since increased with more than 12,800 units constructed to date (end of 2018) in the latest MTSF (2014-2015 to 2018-2019). This is a considerable improvement, although the number of housing units delivered still falls below the target of 27,000. Although there are currently more than 18,000 units in the construction or planning phase, it is unlikely that these will be delivered by the end of the 2018-2019 financial year.

Fig 4.1

Scale of delivery

Scale of delivery

Note: *this target is estimated from the 2014/15 SHRA Annual Report which did not report on the 2008/2009 financial year

Fig 4.2

Allocation across provinces (no of units)

Allocation across provinces (no of units)

Fig 4.3

Allocation across agents (no of units)

Allocation across agents (no of units)

Figure 4.4

SHRA grant disbursements (both RCG & CCG) by province (value in millions of ZAR)

SHRA grant disbursements (both RCG & CCG) by province (value in millions of ZAR)

121The SHRA annual report for 2018-2019 states that only 51% of the social housing units targeted in the 5-year MTSF were delivered (SHRA 2019, see table 2). The report does not explain this under-performance, but instead focuses on the incremental improvements in units delivered compared to previous periods.

122The 2017 State of the Sector Report provides more information on the challenges. The slow pace of delivery is mainly attributed to the failure to update the state subsidy in line with inflation. The subsidy levels were based on benchmarks from 2007 and were only revised in 2017. The income bands were also updated, making it easier for SHIs to achieve the required tenant mix whilst remaining financially viable. The observed upsurge in planned projects in figure 4.1 suggests greater optimism as a result of the updating.

123Figure 4.4 shows that capital grants were unevenly distributed across the country, with funding concentrated in Gauteng (Johannesburg, Ekurhuleni, Tshwane and Rand West), KwaZulu-Natal (Durban and Pietermaritzburg) and Western Cape (Cape Town). The latest tranche of funding for new building was also skewed towards Gauteng, although the Eastern Cape also received a sizable portion of new grant allocations. The construction of social housing units over time has followed this pattern (see figure 4.2), with Gauteng steadily increasing its share of units from 24% to 42% between the previous and current MTSF, and 65% of the units to be delivered in the future. Interestingly, KwaZulu-Natal’s share of units has fallen over time with very few projects approved for the future.

124There are many reasons for the differences between provinces. One is the level of capacity of the small pool of SHIs nationally. Another is the support from other stakeholders, particularly local and provincial government. Gauteng has benefited from preferential funding through the Gauteng Partnership Fund: “This confers a substantial advantage on eligible projects in Gauteng” (SHRA, 2017: 101). The unbalanced distribution of funding supports the argument that the SHRA needs to play a stronger developmental role in capacitating SHIs, municipalities and provinces throughout the country – moving beyond institutional oversight and programmatic reform. Nevertheless, the dominant position of Gauteng is also justified by the extreme pressures on housing in the province, as South Africa’s fastest growing and largest city-region. Gauteng nearly doubled in size between 1996 and 2019 with a total population size of 15.2 million compared to 6.8 million in the Western Cape (StatsSA, 2019). Of course, the demand for housing also outstrips supply in all the metros.

125Table 4.1 shows the 10 largest SHIs involved in producing social housing in the current MTSF, summing up all units delivered and planned for the future. The top ten organisations together accounted for 44% of the total grant allocation and 45% of the total rental stock out of a total of 54 delivery agents in the SHRA’s records. The Johannesburg Social Housing Company was the largest and six of the top ten had operations in Gauteng. JOSHCO is one of the longest-established and most capacitated SHIs in the country.

Table 4.1

Top ranking SHI agents, MTSF 2014-2015 / 2017-2018

Rank SHI/ODA Province Units (planned and delivered) % share total Total Grant (millions) % share total 1 JOSHCO SHI Gauteng 2 949 9.5 430.9 8.5 2 Innovative Strategic Investments Property ODA Gauteng/North West 1 883 6.1 236.5 4.7 3 Arrow Creek ODA Gauteng 1 590 5.1 199.7 3.9 4 The Housing Hub ODA Gauteng 1 480 4.8 337.5 6.7 5 DCI Holdings SHI Western Cape 1 235 4.0 328.4 6.5 6 Msunduzi SHI KwaZulu-Natal 952 3.1 119.6 2.4 7 Imizi SHI Eastern Cape 929 3.0 171.7 3.4 8 FRESCHO SHI Free State 917 3.0 87.5 1.7 9 Salamax SHI Gauteng 903 2.9 113.8 2.2 10 Instratin Properties SHI Gauteng 888 2.9 236.2 4.7 Top 10 Combined Totals 13 726 44.4 2 261.8 44.6

Top ranking SHI agents, MTSF 2014-2015 / 2017-2018

126Table 4.1 indicates a growing reliance on Other Delivery Agents (ODAs) (such as private companies) in the expansion of social housing projects. This is a deliberate strategy by the SHRA, which set a target for ODA participation of at least 15% of the total CCG in order to broaden involvement (SHRA, 2017). Three of the top four organisations were ODAs. This may reflect the importance of scale in attracting commercial interest. Figure 4.3 shows a steady increase in ODA participation over time – roughly a third of units planned in the next term will be delivered by ODAs. Their involvement deserves close monitoring to avoid dilution of the wider social objectives of the programme, including the intention to target well-located areas.

4.3 – Analysis of spatial impacts

127A neglected dimension of the regular reporting on social housing is location. This is surprising considering that the location of social housing is fundamental to achieving the explicit policy objectives of economic inclusion, spatial transformation and urban revitalisation. The 2017 Sector Report mentions the goal of “spatial restructuring” or “spatial transformation” more than a dozen times, but provides no evidence at all on whether this is taking place.

128An important study by NASHO and the HDA in 2013 was the first to indicate that the programme may be compromising on its original intentions to promote urban regeneration. The study sought to analyse where the bulk of funding was allocated using a spatial typology similar to our own (see section 4.1). Although the sample was limited to 32 projects, it was supplemented with key insights from qualitative interviews. The evidence suggested that social housing was moving away from the ‘urban core’ towards ‘outer suburbs’ and ‘grey zones’ because of the availability of land at the right price. The trend was also seen to be the result of the inconsistent application of restructuring zones, which were “becoming increasingly loose in its application by detracting from the original urban restructuring intent of the Social Housing Policy and national strategic programmes” (NASHO and HDA, 2013: 5).

129A more recent evaluation of the social housing programme in 2019 undertaken by Genesis Analytics and commissioned by the SHRA also included a review of the spatial and socioeconomic objectives. A major strength of the study was the large survey of 1 636 tenants over three provinces in 10 social housing projects. This showed measured but positive outcomes for tenants in terms of perceptions of affordability and safety. The outcomes for education and employment were also reported as positive although there are some problems with the way this information was captured and analysed. Evidence of urban regeneration looked at the impact of social housing on property data in the area. The result was fairly weak with “…no discernible trend in property prices and therefore is no observable impact of property prices” (Genesis Analytics, 2019: 55). The study was limited to 10 projects and did not discuss the challenges of finding well-located land for the social housing programme as a whole.

130We provide a more comprehensive account of the way social housing policy has translated into locational outcomes by drawing upon our geo-coded dataset that includes 185 projects across the country. Figures 4.5 to 4.16 capture the distribution of social housing projects in the seven largest metros, along with the Restructuring Zone boundaries. A brief description of the evolution of social housing projects in each metro follows:

• Johannesburg and Ekurhuleni: The Restructuring Zones (RZs) in Johannesburg and Ekurhuleni are far too large to guide project location decisions in any purposeful manner. The zones cover as much as 70% of both metro territories. They are clearly designed to ensure that as much land as possible is included for eligibility. Ekurhuleni also extended its RZ further in 2017 to include an additional area around Boksburg, south of the airport.
  Johannesburg has the largest stock of social housing, with 37% of all units in the country. Ekurhuleni is gradually building up its stock. Social housing in Johannesburg was initially concentrated in the inner city or in adjacent suburbs. Since the mid-2000s, projects have moved away from the urban core and into townships such as Soweto and Protea Glen. Inner-city developments have slowed considerably, whilst there has been a notable increase in large ‘mega-projects’ on the periphery.
  The locational pattern of development in Johannesburg is linked to the type of entity responsible for the project. With a few exceptions, SHI’s are much more likely to develop projects in the inner city, whereas municipal entities (in this case JOSHCO) and ODAs are behind most of the projects in outlying areas. The reasons for this warrant further investigation. The price of land is clearly a major factor, especially as the ODAs are unlikely to be able to access subsidised public land and the social housing funding formula hinders development in well-located areas.
  In Ekurhuleni there is no clear spatial pattern of development, apart from a cluster of projects around Germiston. There is a very large project of approximately 900 units under construction near Carnival City Casino.
• Tshwane: The RZs in Tshwane also cover an extensive area of the metro, including locations outside the core, such as Soshanguve township in the North. Many of these areas do not seem very suitable for promoting access to economic opportunities or for spatial transformation. In practice, the actual location of social housing projects in Tshwane is quite different in that they remain firmly centred on the CBD. There is only one development (near the entrance to Atteridgeville) out of 22 projects which is located away from the CBD. Tshwane is unusual among the metros in this respect. One of the reasons for this has been the availability of relatively low cost vacant buildings for conversion to social housing, as in Johannesburg.
• Cape Town: The logic behind the RZs in Cape Town is also not immediately apparent. They cover quite a small area of the city, which assists targeting. However, they incorporate a range of different types of neighbourhood, including some inner and outer suburbs as well as townships on the periphery, such as Mitchells Plain. Some are well-located on rail transport corridors, but not those in the north-east of the city. The RZs exclude the CBD, even though there is some vacant public property there. The area to the east includes Salt River and Woodstock, and is probably more viable for social housing because property prices are lower, although this is changing (NASHO, 2018). The RZs in Cape Town were not updated in 2017. This was apparently because the City of Cape Town wanted to designate the entire metro as an RZ and had therefore abandoned the idea of zoning altogether.
  Cape Town has 22 social housing projects altogether, 12% of the national stock, despite being the second largest metro. Social housing has been almost completely provided by SHIs with no role for ODAs and no municipal housing entity. The reasons for this need further analysis, including the role of the municipality and province. The Cape Town Community Housing Company was formed by the municipality in the late-1990s. It was very active early on, and focused on rent-to-buy rather than pure rental. However, it incurred sizeable losses and may have fallen out of favour subsequently.
  There is little sign of any clustering of social housing projects in Cape Town apart from a handful of small, historic projects in and around the inner suburb of Brooklyn. A very large project of over 1000 units is planned for Goodwood Station, but this has stalled. It would be a promising location aligned to local development plans along the Voortrekker Road corridor. In general, the location of projects in Cape Town is scattered and does not conform to any obvious spatial logic. However, in recent years there has been a concerted attempt by a group of officials in the municipality, NASHO and several NGOs to develop social housing in the Salt River/Woodstock locality. The idea is to go beyond piecemeal projects by formulating a coordinated, programmatic approach (NASHO, 2018). A variety of suitable land parcels in municipal ownership have been identified and technical studies are underway to explore the feasibility of building social housing on them.
• eThekwini: The 2011 RZs in eThekwini cover a moderate area of the city including the CBD, surrounding core, coastal belts and an industrial node around Pinetown. Some of these seem to be sensible locations to promote spatial inclusion and transformation. However, the 2017 extensions include mainly peripheral locations to the south and north, such as Chatsworth, Phoenix and KwaMashu. Hence, the potential for spatial restructuring in eThekwini seems to have been diluted in the latest round of zoning.
  eThekwini has a fair number of social housing projects spread across a range of different location types and one very large project (more than 1000 units) planned for Bridge City. This is quite a good location in terms of public transport, amenities and its situation between several different kinds of neighbourhood. eThekwini’s social housing is not concentrated in the CBD, despite the opportunities arising from vacant and derelict buildings and the need for regeneration. Projects appear to spread further away from the city’s core over time, as in Johannesburg. There has been little participation from ODAs or municipal entities, with the exception of the Bridge City project which is being developed by a private company. In recent years there has been an attempt by a group of officials in the municipality and NASHO to develop social housing in and around the CBD. The more coordinated approach that is being formulated has learnt from Cape Town’s experience, with an emphasis on developing on subsidised municipal land (NASHO, 2018).
• Nelson Mandela Bay and Buffalo City: Nelson Mandela Bay and Buffalo City have smaller urban footprints and therefore shorter distances between places within each city. This should mean that the RZs need to be more tightly defined to maintain focus. The 2011 RZ in Nelson Mandela Bay is reasonably focused on suburbs in the urban core. However, the 2017 extensions have led to a complete loss of focus with various peripheral zones included along the N2 corridor to the South as well as the township of Motherwell to the North.
  Buffalo City’s selection of zones in 2011 was not very coherent, with a mixture of areas including the CBD but also the peripheral location of Reeston. The 2017 RZ extension for Buffalo City included Bhisho to the North of King William’s Town (outside the visual scope of the map) which is another lagging and peripheral location, rather than a zone of economic opportunities.
  There were only 6 and 14 projects in Nelson Mandela Bay and Buffalo City respectively. This makes it difficult to observe clear trends over time. Most social housing is located outside the CBD, although with reasonable transport connections to economic nodes. All future investments in both metros are planned for greenfield locations.

Figure 4.5

Social housing by year in Johannesburg and Ekurhuleni

Social housing by year in Johannesburg and Ekurhuleni

Figure 4.6

Social housing by entity type in Jo’burg and Ekurhuleni

Social housing by entity type in Jo’burg and Ekurhuleni

Figure 4.7

Social housing by year in Tshwane

Social housing by year in Tshwane

Figure 4.8

Social housing by entity type in Tshwane

Social housing by entity type in Tshwane

Figure 4.9

Social housing by year in Cape Town

Social housing by year in Cape Town

Figure 4.10

Social housing by entity type in Cape Town

Social housing by entity type in Cape Town

Figure 4.11

Social housing by year in eThekwini

Social housing by year in eThekwini

Figure 4.12

Social housing by entity type in eThekwini

Social housing by entity type in eThekwini

Figure 4.13

Social housing by year in Nelson Mandela Bay

Social housing by year in Nelson Mandela Bay

Figure 4.14

Social housing by entity type in Nelson Mandela Bay

Social housing by entity type in Nelson Mandela Bay

Figure 4.15

Social housing by year in Buffalo City

Social housing by year in Buffalo City

Figure 4.16

Social housing by entity type in Buffalo City

Social housing by entity type in Buffalo City

131Drawing these threads together, there is little sense of a consistent spatial logic in the way social housing projects have evolved in most cities. Figure 4.17 shows that the majority of projects are located in places classified as ‘outer suburbs’, ‘greenfield sites’ or ‘townships’ – with the exception of Tshwane which has maintained its focus on the CBD. Peripheral location types make up as much 77% of all units in Nelson Mandela Bay, 75% in Cape Town and 61% in eThekwini. It is difficult to envisage how ambitions for economic integration or spatial restructuring can be achieved if the bulk of social housing projects are not located within the urban core or along key public transport corridors.

Figure 4.17

Social housing by location type across metros (current & planned units)

Social housing by location type across metros (current & planned units)

132Weak spatial targeting is perhaps unsurprising bearing in mind the indiscriminate nature of most RZs, alongside the piecemeal, project-based approach to implementation. The RZs seem unlikely to be helpful in guiding locational choices because the metros either designated too many zones or ignored the aim of spatial transformation when selecting zones. Almost all the new extensions in 2017 are isolated from the urban core and many are in townships. Social housing projects have clearly been vulnerable to ‘spatial drift’ in the absence of any countervailing mechanism to steer them towards well-located areas. A good place to start to rectify this would be for SHRA to create a geo-coded database to ensure that location features more prominently in future project appraisal decisions.

133The tendency for project locations to disperse further from the established urban cores over time is disconcerting. Decentralised locations are less suited to economic inclusion, spatial restructuring and urban revitalisation. It is also harder for scattered projects to build upon each other in a cumulative manner that is more likely to achieve transformative impacts. Figure 4.18 shows how the pattern of site selection has evolved since the 1990s. In the early 1990s and 2000s more than 80% of investments were either in the core city or in the surrounding inner suburbs. This share has fallen steadily since the mid-2000s, giving way to more investment in the outer city or in townships. More than a third of future projects will be constructed in townships and another 13% in greenfield mega-projects.

Figure 4.18

Social housing by location type across years (current and planned units)

Social housing by location type across years (current and planned units)

134The importance of accelerating the overall level of social housing development and restraining the cost of land in projects in order to minimise rent levels is well understood. However, neglecting the policy’s spatial objectives could undermine the fundamental rationale for the social housing programme and jeopardise its future existence. It is not as if there are no opportunities to develop projects in and around the core urban areas. Most of the metros have vacant and under-used well-located land together with vacant and derelict buildings that would be suitable for development or conversion to social housing. Many of these properties are in public ownership. Greater political will and determination are required to make them available for social housing. Alternatively, the value of the subsidy earmarked for land in the social housing funding formula would need to be increased beyond the current R 30,000 per unit.

135The 2017 State of the Sector report admitted that:

136

“The delineation of RZs is influenced by available land rather than spatial restructuring objectives. The large sizes of some RZs have made them irrelevant in terms of focused restructuring … The 2017 SoSR proposes that RZ delineations are removed and that the focus is on the objectives of the RZs and that greater alignment between social housing investment and other municipal spatial investment targeting instruments is created.”

137Improving the links between social housing and spatial transformation requires more than a simple review (or abandonment) of restructuring zones. The 2013 NASHO/HDA report was the first to highlight the issue of spatial drift. It highlighted several blockages to prioritising location including: (i) the availability of suitably priced land and buildings, (ii) loose implementation of RZs, (iii) ad hoc project evaluation by the SHRA and (iv) poor integration with municipal plans and other regeneration initiatives. Stronger leadership is required from public authorities if well-located state-owned land and redundant buildings are to be prioritised for social housing. Encouraging participation by private developers will increase efficiency, but it also means a stronger preference for greenfield sites in order to economise on the land and development costs. Our locational analysis suggests that the social housing programme is at risk of ignoring location in pursuit of quantitative delivery, but in the process compromising on its core spatial objectives, including providing low income residents with better access to opportunities.

4.4 – The impact on neighbourhoods and households

138There is surprisingly little evidence of the impact of social housing projects on the trajectory of their immediate neighbourhoods or the socio-economic circumstances of tenants, despite these being crucial policy objectives. The 2017 State of the Sector Report also neglects these issues. In recognition of this gap in knowledge, the SHRA commissioned a study of the socio-economic and spatial impacts of social housing that offers useful insights. The study collected original data by surveying 1636 tenants in 10 projects in three provinces (Western Cape, Gauteng and KwaZulu-Natal). It also conducted 21 key informant interviews and 14 focus group discussions (Genesis, 2019). Of all the surveyed tenants, 71% were African, 24% Coloured, 4% Indian and 1% White. There were more women (57%) than men (43%) among the surveyed population. Their ages ranged from 18 to 83 years old, with an average of 39.

139Considering the neighbourhood dimension first, the study found that some projects contributed to social mixing and racial integration, although the extent to which this occurred varied greatly, depending on the project location. Social housing projects attracted tenants from both within and outside the local community. In six out of 10 projects, black Africans made up over 95% of tenants. The other four projects were more mixed. Two of them were in Cape Town, which has a much larger coloured population than elsewhere. Most projects mirrored the racial profile of the ward they were located in, although a few may have helped to contribute to social integration. With regard to income, the study suggests that social housing contributes to a more diverse income mix – although ward level data is arguably not disaggregated enough to capture neighbourhood-level dynamics. Summing up, the extent of social integration was varied. Unfortunately, the study does not explain whether some projects tried harder than others to achieve this, and what the main factors underlying the variability were.

140The study found little evidence that social housing projects attracted economic investment or promoted economic development in their areas, with the exception of tenants spending money on local consumption and limited small business development. While some public and private investments occurred in the locality, the small scale of social housing projects made it impossible to attribute them to the social housing programme. There was no evidence of social housing projects having a discernible impact on the prices of surrounding properties. The job creation potential of social housing projects was also limited, usually focussing on tenant management and maintenance activities. Therefore, it seems that social housing did little to regenerate or upgrade their localities.

141Turning to the household level, the study found evidence in the tenant survey of people feeling more secure and safer in their units, but this did not apply to the areas surrounding the project. The projects also contributed to better access to health care and educational facilities, although many pupils did not change schools and continued to travel long distances. Most projects were situated near to education and health facilities. The average distance to the nearest healthcare facility was 2.7 km and to the nearest primary school was 1 km. Some social housing providers had community development programmes that improved their tenants’ health. Social housing seemed to make a strong positive contribution to pupil’s learning outcomes for those who switched schools.

142Whether tenants’ access to regular public transport improved depended on the location of the project in relation to existing transport routes. Inner-city projects improved people’s access to regular public transport, whereas projects in suburbs with limited public transport created challenges for tenants to find employment or to get to school on time. (This confirms the importance of a good location). Tenants across the ten case study sites spent an average of R217 per month more on transport compared to their previous residences although this statistic could be due to inflation. Real increases in transport could be the result of longer distances travelled to work as well as children changing from public transport to private lift-clubs, which were prevalent where public transport was not readily available. Another reason for the increased transport costs could be higher disposable incomes and/or the higher price of petrol.

143There were only two households out of the 1 636 surveyed where no-one was employed (probably due to a recent job loss). This is partly to be expected since formal employment is an important entry requirement. Nevertheless, it is still an important finding. Overall, 96% of tenants were permanently employed, 20% temporarily employed and 7% self-employed. These numbers add up to more than 100% because some tenants had multiple jobs (e.g. temporarily employed and self-employed). The study found that 79% of economically active men and 61% of economically active women were employed, another notable finding since these proportions are much higher than in the general population. Unfortunately, the study did not consider the turnover of tenants, and how many may have exited or been evicted after losing their jobs.

144In our next study we intend to pursue these insights from the Genesis study further by including a clearer focus on household-level mobility. We hope to capture information on the employment status and income of tenants at the time of entry into the social housing and compare that with their income at present to assess whether they have become better off as a result. We also intend to compare projects in different locations within the same cities to try and gauge the significance of location on the employment status and income of tenants. We will explore the feasibility of conducting exit interviews with former tenants to understand the reasons for leaving and their housing pathways after departure (did they progress into better housing or fall back after losing their jobs?). Another study objective is to assess the impacts of community development programmes on tenants and explore ways of linking them better to existing development initiatives in the neighbourhood.