Social Housing and Upward Mobility in South Africa: an Assessment of Household Outcomes

Abstract

Social housing offers a way of promoting upward mobility and urban integration by improving people’s access to economic and social opportunities. These are prominent goals of South Africa’s social housing programme. In practice, household advancement tends to be taken for granted rather than deliberately pursued or monitored. We review survey data for 1632 households across ten social housing projects and find modest evidence of social mobility and racial integration. One reason for this is that other considerations have taken precedence during implementation. We conclude by identifying various ways of reaffirming the transformative goals of social housing.

Introduction

The high and rising levels of income inequality and residential segregation in cities of the South are serious public concerns (van Ham et al., 2021). Polarised cities aggravate poverty, hinder human development, foster mutual suspicion and fuel social discontent and instability (United Nations, 2020). The causes of growing urban divides are multi-faceted and include changes in the labour market, such as declining industrial employment and increasing casual and informal work. They are compounded by disparities in neighbourhood amenities, public services, housing conditions, crime and the quality of life (Chetty & Hendren, 2018; Massey et al., 1987). Urban divisions are also amplified by public and private organisations discriminating against particular race, gender or income groups, or stigmatising specific areas and communities, such as ‘red-lining’ by financial institutions (Mazzucato & Farha, 2023).

Social housing is a potential mechanism for countering inequality and segregation by enabling low- and moderate-income households to live in homes and neighbourhoods they could not normally afford. It may be possible to break the cycle of poverty, social disadvantage and spatial marginalisation by giving people access to resources and opportunities to enhance their skills, well-being and economic circumstances (OECD, 2020). The rationale is that individuals and families who are provided with a secure and supportive living environment that is close to jobs, public facilities and social amenities will gain greater control over their lives and be empowered with the capabilities and connections to transform their life chances and thrive (Best & Shimili, 2012; Mazzucato & Farha, 2023).

These issues are particularly important in South Africa because of the extreme levels of racialised inequality and entrenched urban segregation (Schotte et al., 2018, 2022; World Bank, 2018). The political transition to democracy has done little to reverse the economic deprivation and exclusion of the black majority of the population, although the main source of social cleavage has shifted from race to class with the emergence of a modest black middle class (Chatterjee et al., 2020; Pabon et al., 2021; Visagie, 2015). South Africa’s two largest cities were recently ranked among the most segregated in the world (Turok et al., 2021a; van Ham et al., 2021). Socio-spatial polarisation remains closely associated with the inherited apartheid geography, whereby the vast majority of poor black households occupy overcrowded and under-serviced townships and informal settlements on the urban periphery (Hamann & Horn, 2015; Maharaj, 2020; Lemon et al., 2021).

One of the government’s top priorities since 1994 has been to reduce the housing backlog that soared during the colonial and apartheid eras. The democratic state has conceived of housing as a human right which requires giving social security and dignity to poor households living in dire conditions (Presidency, 2019). Consequently, the main housing programme has constructed between three and four million fully subsidised dwellings, which is by far the biggest public housing scheme on the African continent (Bah et al., 2018). The emphasis has been on large-scale delivery of small units, each on its own plot of land. The unintended consequence has been to create low-density dormitory settlements on large parcels of cheap peripheral land (Bradlow et al., 2023; Buckley et al., 2016). Being far from jobs and amenities places a heavy transport burden on these communities and worsens social isolation. Consequently, the housing programme has been criticised for its siloed, standalone approach which perpetuates the sprawling, segregated form of the apartheid city and aggravates poverty and exclusion (Presidency, 2019; National Planning Commission, 2023).

The government’s social housing programme emerged partly as a response to these criticisms (Turok et al., 2022). It is based on a broader and more dynamic conception of housing as a means of household advancement through improved access to economic and social opportunities. It is also a mechanism to help integrate different communities and densify well-located urban areas, with the aim of ‘restructuring South African society to address structural, economic, social and spatial dysfunctionalities’ (Parliament, 2008). This means a focus on rental housing rather than ownership, and actively managed blocks of flats in precincts offering childcare and other support facilities, rather than freestanding homes. The property is retained in collective ownership and protected for successive cohorts of low-income tenants. Renting confers flexibility and enables multiple households to benefit from the same stock and amenities. As their economic circumstances improve, they are expected to move out into the private market and make room for others.

The overall budget for social housing has always been much smaller than for the main housing programme, partly because there is more emphasis on these qualitative features and less on quantitative delivery. Multi-storey flats are more expensive to construct than single-storey units, and well-located land is more costly to acquire than marginal land. Social housing is also more complicated and time-consuming to plan because it requires proximity to other land uses (which risks neighbour objections) and a coordinated, precinct approach to development, rather than piecemeal projects. In addition, such projects necessitate ongoing management and maintenance by social housing organisations (SHOs).

In practice, the goals of household progression and spatial transformation have received less attention in the programme’s design than one might have anticipated. The broad assumption was that if the homes were centrally located and the rents were modest, then upward mobility and urban restructuring would automatically ensue. The actual mechanisms through which these changes would occur were not spelt out in any detail. SHOs face many tensions and trade-offs in creating viable projects, but without much guidance from the centre about household development and social mobility. The dilemmas include balancing deeper social needs against financial viability, and larger scale delivery against superior locations (Turok et al., 2022). SHOs have also had to formulate their own criteria for tenant selection, choose what kinds of support services to offer, and work out how their residents might in due course move into home ownership or private rental housing. There was also no analysis done centrally of the gap in affordability between social housing and the private housing market that households would have to bridge across in order to progress, and how realistic this was.

The broad objective of this paper is to assess whether occupying social housing has led to enduring household outcomes. Has it given tenants a helping hand to get out of poverty and live more secure and happier lives? Has it done even more than this by serving as a springboard to transform their living standards, move up the housing ladder and into their own homes? Conversely, has it prevented people from losing their homes (e.g. through eviction) and sliding back into poverty and insanitary living conditions? And by assisting household progression and creating pathways to a more prosperous future, has it altered the existing geography of inequality in urban areas?

The original contribution of the paper is two-fold. First, we identify the multiple ways in which social housing could in theory make a measurable difference to the lives and livelihoods of tenants. Second, we analyse survey evidence of actual household outcomes. These findings suggest that the positive effects are modest. We also offer some tentative suggestions for why this might be the case, including the lack of policy focus on this goal and the objective difficulties for achieving this. To our knowledge, this is the first time that the impact of social housing on household mobility has been analysed in South Africa, although the issue has been mentioned in previous evaluations of the programme (DPME, 2016; Genesis Analytics, 2019).

The article proceeds as follows. First, we review the international evidence linking social housing and upward mobility. This is followed by a discussion of the mechanisms by which social housing could promote household progression, i.e. the theory of change. Next, we discuss the methods and data used for the analysis, including a large tenant survey covering ten social housing projects. Three dimensions of ‘transformation’ are evaluated in the results section: racial and social integration, livelihoods and mobility, and tenant progression from entry to exit. The paper ends with concluding remarks and policy recommendations.

The International Evidence

There is surprisingly little systematic analysis of the impact of social housing on upward mobility. This may reflect the fact that more emphasis has conventionally been attached to the role of education and the labour market in shaping people’s ability to get ahead and improve their living standards. Furthermore, research on the relationship between housing and life chances tends to consider housing in general rather than social housing in particular. This research (discussed below) also devotes more attention to the neighbourhood dimension than to the individual dwelling unit or residential complex. Neighbourhoods encompass schools, healthcare, amenities, crime and other attributes that may have a bigger influence on human capabilities and upward mobility than the character of the individual home or its tenure.

There is increasing research on the relationship between neighbourhoods and mobility, especially in the USA, where this has become a popular research topic. The emphasis is on the impact of residential segregation—according to race or income—on constraining disadvantaged households because of their inferior access to opportunities and their exposure to violent crime, drug dealing and other social and environmental problems. The most prominent research has been led by the economist Raj Chetty (see Chetty et al., 2014, 2016; Chetty & Hendren, 2018). He has pioneered the use of ‘big data’ that tracks 20 million Americans from childhood to their mid-30 s in order to assess the effects of living in different districts on their chances of escaping poverty. The analysis focuses on household incomes, although other indicators are also used, such as college attendance, fertility, marriage patterns and whether people end up in prison.

An important finding is that small area variations matter greatly for people’s life chances. Just shifting focus one or 2 km from a ‘low opportunity’ to a ‘high opportunity’ neighbourhood transforms outcomes. Chetty has also studied policy initiatives that assist poor families to move from areas of concentrated poverty to better districts, concluding that: ‘Of the various things I’ve studied over many, many years, this turns out to be one of the highest-impact, most successful things I’ve seen’ (quoted in Pinsker, 2019). Furthermore, young children experience the biggest benefits from growing up in a better locality. Chetty acknowledges that the reasons for this are unclear, although it appears to have something to do with being exposed to decent schools, less poverty, more fathers present in families and positive social norms.

This research tends to favour policies such as housing vouchers, rather than traditional public housing projects located in deprived areas. This means people-oriented (demand-side) rather than place-based (supply-side) solutions. Yet, Chetty recognises that moving deprived families out of poor neighbourhoods is not a scalable or sustainable solution in and of itself. This is even more apparent in South African cities, given the much larger scale of poor areas compared with affluent and middle-income neighbourhoods. With half the population trapped in chronic poverty (Schotte, 2022), there are far too many poor households to be absorbed into advantaged localities. Ultimately, investments are required to make all communities areas of opportunity, through improved schools, more facilities, better transport connections and less stigma. This requires additional investment in poor districts, not abandonment. It should be accompanied by support for affordable housing in well-located areas, wherever opportunities exist. In other words, a multi-pronged approach is necessary.

There has been more detailed research on social housing and upward mobility in the UK than elsewhere. This research suggests that there is little upward progression for tenants living in social housing because most of them remain in chronic poverty (Stephens et al., 2014; APPG, 2020). This is linked to the fact that many of these households are persistently unemployed or economically inactive. In other words, social housing in the UK does not generally function as an escalator, at least not anymore. It operates more like a safety net than a trampoline that bounces people back into the labour market and out of poverty. This is related to the disposal of the best social housing stock in recent decades. The remainder has become more like a residual, enclave sector occupied by households who are particularly vulnerable to poverty and worklessness. Many other countries have also experienced a growing concentration of low-income and disadvantaged households in social housing (OECD, 2020).

The problem also stems from the fact that much of UK social housing is located in lagging regions and districts with weak labour markets (APPG, 2020). Tenants have relatively low skills, limited social connections to find vacancies, and travel constraints to access jobs. Social housing rent levels also determine people’s disposable incomes and chances of moving above the poverty threshold. Nevertheless, there are actions that SHOs can take to help tenants to improve their living standards by tackling the specific barriers they face, providing training and job search support, creating jobs themselves (e.g. in cleaning, catering and building maintenance) and locating future housing projects in areas with more employment opportunities (Gallagher et al., 2018).

Although equivalent longitudinal analysis does not appear to have been undertaken elsewhere, there is evidence of similar tendencies at work in other advanced economies, especially in large, globally connected cities that have prospered in recent years (Madden & Marcuse, 2016; United Nations, 2020). Commercial pressures in their real estate markets have raised property prices across the board, encouraging speculation and redevelopment of older buildings and under-used land. For example, an Australian study found that tenants struggle to graduate out of social housing programmes because of a lack of affordable alternatives in the private rental market (Wiesel & Pawson, 2015). All forms of public and social housing have come under threat of being sold off or dismantled in the face of the financialisation, privatisation and commodification of housing and the built environment (Mazzucato & Farha, 2023; OECD, 2020). Rising rents and evictions have caused displacement and homelessness as the exchange value of housing trumps its use value. In short, housing is being decoupled from the purpose of meeting social needs to serving as an investment vehicle or liquid asset for global finance (Madden & Marcuse, 2016).

Summing up, the international evidence suggests that the position occupied by social housing within the wider housing system has a big influence on whether it enhances upward mobility. If social housing is reduced to functioning as a last resort for people with little choice, it may end up as a safety net offering basic protection and security for households at the bottom of the social hierarchy, but not a springboard for people to get ahead through employment and higher incomes. Considerable effort and investment will be required to transform this kind of residual social housing into a vehicle for upward mobility.

In contrast, if social housing serves a broader purpose, it is better placed to enable human progress. This will require higher quality stock located in areas accessible to jobs, decent schools and other opportunities. Tenants will be more diverse, including younger adults and people with education and skills. These housing complexes will be able to charge higher rents to cover their higher costs because more residents will be employed. This will also narrow the affordability gap to the private housing market. The potential for social housing to perform this transformative role will be enhanced if it is complemented by other social infrastructure, such as decent childcare, effective policing, safe public transport and appealing recreational facilities.

How Could Social Housing Promote Upward Mobility?

The mechanisms by which social housing could promote household progression have nowhere been spelt out in any detail. Drawing on previous research and our own ideas, we seek to tease out the various ways in which social housing can support upward mobility, both directly and indirectly. We describe this as a theory of change. The point of departure is that social housing enables low and moderate-income households to live in better accommodation and/or neighbourhoods than they could normally afford. The benefits go beyond lower costs and improved quality of accommodation to include a more advantageous living environment with more resources and opportunities available that have the potential to influence household prospects through improved capabilities and other channels (Chetty et al., 2016; Genesis Analytics, 2019; Baffoe, 2019).

Table 1 sets out various pathways that support household advancement, including the likely outcomes and indicative time horizons. The first proposition is that social housing provides improved access to economic nodes in the city for people who previously may have faced lengthy and costly commutes, which inhibited job search and made it less worthwhile to work. Proximity to workplaces reduces travel costs, increases the chances of securing employment and opens up a wider choice of jobs for future career progression. Local consumer demand is also stronger in better-off neighbourhoods, which improves livelihood opportunities for people starting their own enterprises, such as informal traders, tradesmen and mechanics.

Table 1 Mechanisms linking social housing and upward mobility

Second, social housing can improve access to good education through proximity to superior local schools. Children with better qualifications are likely to climb higher up the occupational ladder and raise their future earnings potential. Third, proximity to better healthcare facilities can also improve well-being and workforce productivity in due course. While these benefits may take time to be realised, households can also benefit from the immediate convenience and savings in transport costs to schools, clinics and hospitals. Fourth, social housing can benefit households by improving access to local amenities, such as libraries, parks and leisure facilities. These improve the quality of life and enhance people’s attachment to their areas, which can help them to live more fulfilling and rewarding lives. Fifth, living in socially mixed communities can widen people’s social networks and improve their connections, with various benefits including up-to-date information about job vacancies.

SHOs can also contribute directly to social mobility amongst residents and their children. Well-managed precincts create a safe and stable living environment which promotes peace of mind and personal advancement, without the risks, fears and disruptions caused by crime and violence. Socially responsible SHOs can also protect households from periods of joblessness and financial distress by offering temporary rent relief. Commercial landlords are more likely to evict people from their homes, thereby undermining their health and well-being. Enlightened SHOs provide a variety of support programmes to give households greater control over their lives and help them to thrive, including training and job search assistance. Childcare schemes can enable tenants to obtain employment, especially women. SHOs may even create jobs themselves to empower people and give them work experience and skills through building maintenance, cleaning and security.

None of these propositions is guaranteed to hold in practice since they are likely to depend upon various contingencies. At least four factors stand out. First, household characteristics (initial education, skills, employability, age, family size and aspirations) are bound to affect their prospects of advancement. For example, older people with limited education will be more constrained than skilled young adults. Second, the housing quality and supplementary services available are likely to influence tenant progress—presumably the more support the better. Against this, the more the rent is subsidised, the bigger is likely to be the step up to enter the first rung on the private housing market ladder. Third, the location of the housing complex and its proximity to jobs and other facilities is also bound to affect people’s chances of success. This includes the neighbourhood character and the positive or negative external effects on household prospects. Finally, the state of the wider labour market and the strength of demand for the kinds of skills possessed by tenants is likely to be extremely important in influencing their chances of escaping poverty. Where unemployment is high and job vacancies are scarce, it is bound to be more difficult for households to improve their economic circumstances.

The benefits are also bound to take effect over different timescales, so sensitivity to timing is important in assessing their impact. Some may be once-off and fairly immediate, such as for households moving from substandard shelter on the urban periphery to better housing in more accessible locations. Alternatively, the benefits may be prolonged if households are exposed to supportive living environments that help them to acquire new skills and competencies, develop new social networks and pursue new initiatives. Households may make crucial savings from the subsidised rents to spend on food, education, childcare, healthcare or insurance against some future risk or disaster. Some of these benefits may only accrue to the children, thereby contributing to inter-generational mobility.

A desirable scenario is where the benefits of social housing keep recurring as successive families move through the housing complex and into the private market as their circumstances improve, thereby releasing the dwellings for others to occupy and advance over time. This enables social housing to function as a ‘social escalator’ and helps to justify the higher public investment. It is a far cry from private housing that exists as a commodity to be bought and sold, and from which families may be evicted if they cannot afford the rent or to repay their loans. It is also quite different from the kind of informal, makeshift accommodation that is occupied as a last resort and is little more than physical shelter—a roof over people’s heads. Here, people are vulnerable to all kinds of hazards, and the threats of eviction and extortion can also loom large.

Methods and Data Sources

The multiple, interacting pathways to progress, coupled with the variety of contingent conditions, make the measurement of social housing’s contribution to upward mobility very difficult. There are also likely to be differences in how household outcomes (e.g. earnings, employment, education and health) interact with geographic scales (e.g. the home, complex, neighbourhood, region). The different timescales by which these effects come about (i.e. immediate, enduring, inter-generational) further add to the complications. Further methodological challenges arise in accurately pinpointing the impact, including identifying a valid counterfactual, disentangling cause and effect, and selecting appropriate time horizons.

A comprehensive investigation into social housing and mobility would rely on longitudinal data to track individuals over space and time. This would be a lengthy and expensive exercise beyond the capabilities of most researchers. The approach followed in this paper is exploratory and pragmatic in making the best use of existing surveys and administrative data. It falls short of a definitive study. Consequently, the results should be interpreted as preliminary and laying some foundations for further research.

The backbone is a tenant survey commissioned by the Social Housing Regulatory Authority (SHRA) in 2019, which collected a wide range of socio-economic information for tenants and other members of their households.¹ The survey gathered responses from a sample of 1636 leaseholders from 10 social housing projects in four metros (eThekwini, Cape Town, Johannesburg and Tshwane) across three provinces (KwaZulu-Natal, Western Cape and Gauteng). The response rates were fairly high with roughly 40% of leaseholders from each project completing the survey. Therefore, the data is robust at the project level, although it was never designed to be representative of all social housing projects nationally. The projects cover a range of location types, from the inner core of the cities (BG Alexander, Tau Village and Thembelihle in the CBD and City Deep nearby), inner suburbs (Roodepoort, Belhar) and outer suburbs (Valley View, Avoca Hills, Lakehaven and Scottsdene). As discussed in the ‘How Could Social Housing Promote Upward Mobility?’ section, location is expected to play an important role in upward mobility, so we are careful to report findings at the project level to show where differences emerge.

We also assembled a geo-coded database of each tenant’s residence prior to moving into social housing from their responses to the same survey. This is important for analysing where people had moved from in order to contrast conditions between their current and former residence. If the positive outcomes of social housing are premised on people moving from ‘inferior’ to ‘better’ neighbourhoods, then this is a critical element to explore. We managed to successfully geo-code the vast majority (91%) of the original sample which should ensure that each project population is highly representative.²

In addition, we draw on data provided to us by Madulammoho SHO that offers basic descriptive information on four of their social housing projects: Fleurhof Views and BG Alexander in Johannesburg and Scottsdene and Belhar in Cape Town.³ Three of these projects were selected as case studies in the Genesis survey, which provided a useful quality check. Madulammoho also provided information on tenant turnover and exits for the period November 2019–February 2020 covering their full suite of ten projects (across Cape Town and Johannesburg), including reasons why tenants left their homes. These exit surveys are an unexplored source of information about the social housing programme and potentially very useful in analysing mobility among tenants. The data provides insights into whether people move out for positive reasons, such as buying their own home, or negative reasons, such as unaffordability or job loss.

Findings

Racial and Social Integration

Spatial transformation is an important objective of social housing policy, although the actual mechanics have not been analysed at the household level before. Social housing could reduce urban divides and contribute to racial integration and economic mobility by accommodating low- and moderate-income black households within the urban core and other high-opportunity areas. For the policy to be realised—as experienced in the everyday lives of beneficiaries—social housing needs to (i) target previously disadvantaged households living in marginal areas and (ii) enable them to move into well-located areas (i.e. close to jobs and amenities). These conditions have not been tested before, despite their obvious salience.

Table 2 shows the results of geo-coding tenants’ previous place of residence. It reveals quite a mixed picture in which most moves into social housing were essentially local, but not exclusively so, since a fair proportion of tenants also came from further afield. In seven of the ten projects, more than half of the tenants moved less than 5 km into their social housing unit. In all ten projects, more than half of the tenants moved less than 10 km. However, every project also recruited residents from further afield. At a minimum, one in eight tenants came from more 10 km away. In five of the projects, between a third and a half of all tenants moved from more 10 km away. The median distance moved for each project ranged from just over 1 km in the case of BG Alexander to 8 km in the case of Lakehaven.

Table 2 Distance moved from previous residence

Projects in the inner city were most likely to recruit locally. More than two-thirds (71%) of BG Alexander tenants moved within the same neighbourhood (less than 2 km). This reflects the strong local demand and shortage of affordable accommodation within Johannesburg’s inner city. Tau Village and Thembelihle also experienced strong demand from people already living in the Pretoria CBD. Some projects in suburban and peripheral locations, such as Belhar and Scottsdene in Cape Town, also had many tenants moving within the vicinity. The evidence suggests that SHOs do not give much priority to recruiting tenants from outlying districts. Insights from key informant interviews suggest that SHOs are often encouraged by local politicians to select households from the surrounding area. This may reinforce a natural tendency of many people to prefer moving locally for familiarity and social networks.

The fact that most beneficiaries were living in the same district before moving to the social housing seems to challenge a core policy assumption. It means that looking for causal links between social housing and spatial transformation outcomes—such as more social inclusion or better access to economic opportunities—may be misplaced for most households. Most moves seem too short to make a major difference in these respects. Even so, such housing can still support social upliftment. For instance, urban decay and depressed property prices within most inner cities have enabled many poor households to move into these areas in recent years, but they often occupy rundown buildings without basic services (Turok et al., 2021b). Creating decent and secure accommodation for these people could well improve their life chances even if they were already living locally. Social housing that benefits families previously occupying rudimentary backyard structures in surrounding neighbourhoods could also have a profound effect. In other words, long-distance relocation from an outlying township is not a necessary condition for a positive outcome.

Nevertheless, the choice of location for social housing remains vital for changing entrenched spatial structures, even if projects do not draw most of their tenants from far-flung areas. Other studies suggest that central locations have not been given sufficient priority to date because of the prohibitive cost of private land and because public bodies have been slow to release any of their surplus land to SHOs. Indeed, there has been a disconcerting tendency for the share of projects located in central areas to decline over time, in favour of peripheral locations (Turok et al., 2022; SHRA, 2018; NASHO and HDA, 2013).

The location of the ten project sites included in the tenant survey shows wide variation in desirability and proximity to the city core. BG Alexander (Johannesburg CBD), Tau Village and Thembelihle (both Pretoria CBD) are located in highly suitable areas for social housing. City Deep is also situated close to Johannesburg CBD, although technically in a former industrial area. Roodepoort and Belhar are situated in secondary urban nodes and also appear to be reasonable locations for access to opportunities. The remaining four projects are in more questionable locations. The three projects in eThekwini (Avoca Hills, Valley View and Lakehaven) are each in areas that could be classified as ‘outer’ suburbs. Similarly, Scottsdene is in Kraaifontein on the outskirts of Cape Town and far from economic centres.

The choice of location along with the share of residents moving from outside the neighbourhood affects whether social housing supports racial integration. Figure 1 compares the racial mix of each social housing complex with its surrounding area and the neighbourhood from which tenants originally moved.⁴ There is considerable similarity between all three patterns. In other words, few people were moving into neighbourhoods with a different racial composition. Moreover, the racial mix within most social housing projects was very homogenous, so there was little diversity within each complex. There were few white households in any of the projects, although this is unsurprising considering that white households are a relatively small and affluent population group.

Fig. 1

Racial composition of social housing projects and neighbourhoods (%). Source: SHRA 2019 tenant survey; Census 2011

The racial composition of each project’s surrounding neighbourhood is clearly important. Only Valley View, in Hillary, was located in a district that is patently diverse. This is because of its position between eThekwini’s inner core and periphery with Cato Manor to the North (former black area), Umbilo to the East (former White area) and Chatsworth to the South (former Indian area). More than two-thirds of the Valley View tenants moved over 5 km to the project (see Table 2) and had moved from neighbourhoods that were almost two-thirds (61%) African. Therefore, this is an uncommon case of a project outcome that involved considerable racial and locational change.

Lakehaven (also in eThekwini) is also interesting, albeit for different reasons. The racial composition of this project was unusually diverse—46% of tenants were African, 28% Coloured and 25% Indian. Meanwhile, the local neighbourhood (Newlands East) was largely Indian (69%). Therefore, this project made a unique contribution to enhancing the racial diversity of its neighbourhood. None of the 10 projects was located in former White suburbs, where public facilities and amenities tend to be superior. The CBDs of Johannesburg and Pretoria have long since transformed and are now predominately African.

Promoting racial integration within social housing complexes does not appear to be a prominent objective across all projects. Siting such complexes within well-located and/or racially diverse neighbourhoods is also unusual. It seems that SHOs need clearer guidance and greater government support to prioritise high-opportunity areas, such as releasing unused state-owned land or subsidising development costs.

Livelihoods and Mobility

Social housing could still enable social and economic transformation if the beneficiaries are found to have improved their living standards. This could be linked to increases in disposable income arising from monthly savings on rent or transport costs (for those moving closer to workplaces). There could also be long-lasting improvements in personal circumstances where people gain better access to economic opportunities, decent schools, healthcare and a safe and secure home environment. Some SHOs offer employment and skills training for their tenants, although this is not a policy requirement.

For social housing to act as a social elevator, it needs to target relatively poor households. Reaching those in need largely comes down to the criteria for tenant selection defined by the SHRA. The programme targets low- to middle-income households who do not qualify for free RDP/BNG housing (because they have a basic income) and who find it difficult to obtain a commercial mortgage (because their income is too low or unreliable). The primary target market covers households earning between R1500 and R5500 per month. These income levels are low considering that the national minimum wage is around R3350 per month (R20 per hour, depending on the sector). The secondary target market covers households earning between R5500 and R15,000 per month which is still fairly low for a household of 3–4 people.

Monthly rentals are regulated by the SHRA and set at 25–27% of income for the primary market (which works out between R375 and R1485 per month) and 32–37% of income for the secondary market (R1760–R5550 per month). There is some cross-subsidisation between the primary and secondary markets because rentals are higher in absolute terms and as a proportion of income as incomes increase. These rent levels are generally quite a lot lower than the formal private rental market, although the difference depends on the location. Our interviews with SHOs suggest that they have many people on their waiting lists, which reinforces the sense of strong demand.

To illustrate this, Fig. 2 shows the distribution of tenant incomes for projects managed by Madulammoho SHO.⁵ A quarter (25%) of tenants had a per capita income of less than R2,500 per month. This is only slightly above a basic poverty line of R1427 which is widely used in the literature (see Budlender et al., 2015). Incomes are consistently low for most tenants, and below R4000 per month up to the 90th percentile. BG Alexander tenants appear to have had higher incomes across the board. Closer inspection of the data suggests that this is somewhat artificial. Tenants in BG Alexander actually had similar income levels to other projects, but there were more single and communal bedroom units for one-person households, which skewed the average per capita income distribution upwards.⁶

Fig. 2

Distribution of income per capita (rands per month). Source: Madulammoho (2020). Notes: historic income data adjusted for inflation; household size based upon dwelling size

This evidence suggests that social housing is well targeted towards low-income households. The strict focus on this group is commendable, although it complicates the task of financial sustainability for SHOs because they have to work within tight margins. Anecdotal evidence suggests that SHOs make a loss from their lowest income units because the rentals do not cover the costs of routine maintenance and everyday management. There is a risk that projects fall into disrepair and households end up receiving sub-standard services. Some SHOs argue that the SHRA should relax their strict rules regarding the incomes of tenants to allow projects to accommodate more middle-income households to help cross-subsidise the others. It would also be simpler for these tenants to graduate into the private housing market afterwards.

Targeting poor households is a good start, although there is no guarantee they will climb the income ladder. Figure 3 shows how tenants perceived their disposable incomes had changed since moving into social housing. On the whole, tenants were positive (25–63%) or neutral (23–53%) about these changes, with a minority (6–33%) indicating that their income had declined. The positive responses are reassuring, although there is no information about how much incomes had increased. Interestingly, the variation between projects suggests that additional factors besides location are involved. For instance, 62% of people in Tau Village believed that their disposable income had increased, compared to only 25% of tenants in Thembelihle (just a few kilometres away). People in Scottsdene were the most positive about their incomes, even though tenants have to pay high transport costs in this peripheral area. There is no obvious link between the project location and the change tenants reported in their incomes. Further research is required to explain the varied responses.

Fig. 3

Change in disposable income since moving to social housing. Source: SHRA 2019 impact study database

Distance to the workplace is presumably a key factor in determining whether tenants could save on transport costs by moving to social housing. Tenants in projects in and around the CBD (BG Alexander, City Deep, Tau Village and Thembelihle) did report shorter distances travelling to and from work (70% work within 15 km of their residence) compared to other projects. However, we also know from Table 2 that most beneficiaries did not end up moving very far to take up their residence, which must have limited the savings they experienced.

Another channel for sustaining upward mobility is improved access to employment. SHOs normally require applicants to have a secure job (in order to afford the rent), so it would be misleading to consider the rate of employment amongst tenants as a sign of progress. Yet, other adult household members (besides the main leaseholder) who were previously unemployed could have improved their job status through social housing. This could have occurred as a result of advice or training they received from the project or better access to job vacancies in the surrounding area.

Figure 4 shows the rate of unemployment for adult household members in each of the 10 projects. Unemployment was a serious problem in all of them, except Belhar and Scottsdene where it was below 10%. The unemployment rate was over 20% in most projects and over 27% in Valley View, City Deep and Roodepoort. This is surprisingly high considering that the main tenant should have had a job at the outset. Despite not knowing exactly what the unemployment rate was before families moved into the social housing, it seems hard to imagine that their labour market position has improved since then.

Fig. 4

Unemployment rate. Source: SHRA 2019 impact study database

Education is another potential mechanism for promoting upward mobility. Data to examine the effects of social housing on education are unavailable, but we can at least consider whether social housing is linked to more investment in education, which could pay dividends in due course. In the survey, tenants were generally positive about the quality of education at new schools with 63–88% of them reporting an increase in grades. Yet, closer inspection of the data reveals that only 23% of children actually changed schools when moving to the social housing. In Scottsdene, only 6% of children changed schools, and in Tau Village, there were none. The fact that most children did not change schools seems to imply that access to schooling was not a serious motivation for moving into social housing. The reasons why most learners did not change schools need further exploration.

One last piece of evidence looks at how social housing contributed to education and capacity development in a broader sense. Figure 5 shows the extent to which tenants were aware of any community development programme in their complex. Education and training were first on a long list of potential capacity building activities (including health services and sports activities).

Fig. 5

Awareness of a community development programme in complex. Source: SHRA 2019 impact study database

The results show that awareness of support services was very mixed, ranging from very low in some projects (like Valley View, Lakehaven and Avoca Hills, all in Durban) to very high in other projects located in the Johannesburg and Pretoria CBDs (i.e. BG Alexander, Tau Village and Thembelihle). The main message is that promoting personal development of tenants did not seem to be a priority in many social housing projects, resulting in low levels of awareness. The SHRA does not monitor household outcomes systematically, which may reduce the incentive for SHOs to take on the additional costs involved in providing community development programmes.

From Entry to Exit

A third perspective on economic mobility emerges from evaluating tenants’ reasons for moving into and out of social housing. People may apply for many different reasons, although the policy could legitimately assume that their main intention is to improve their personal and family circumstances and not just to find a cheap place to stay. One would also expect SHOs to select tenants with such hopes and desires. A clear indication of upward mobility would be where households exit social housing to purchase their own home or because they can afford to rent something closer to their aspirations in the private market.

Figure 6 shows the reasons tenants gave for originally moving into social housing. Affordability dominates the responses in every project (between 39 and 64% of responses). This is perhaps unsurprising considering the shortage of decent affordable housing in all cities and the attractiveness of the subsidised rents for people on low incomes. SHOs know their rents are below market prices, although they have been rising in recent years because of higher maintenance costs and municipal service charges. They also have waiting lists for many of their projects, as mentioned earlier, which enables them to pick and choose their tenants to some extent.

Fig. 6

Reasons for moving into social housing. Source: SHRA 2019 impact study database

Figure 6 also suggests that many tenants appreciate the wider benefits afforded by social housing. Proximity to work was the second most common reason. This is somewhat reassuring, considering the earlier questions raised about the location of certain projects, although one might have expected a stronger response if these projects were better located. Security comes next, followed by the quality of the accommodation. These responses varied greatly in importance across the projects. For instance, security was most significant in Cape Town (Belhar and Scottsdene). These projects are both in secure complexes and surrounded by tough neighbourhood environments with high levels of crime and gangsterism. Mention of ‘independence/place of my own’ suggests that some people were previously living in overcrowded conditions or with their extended families. Interestingly, social ties or schooling did not feature prominently.

The reasons why people leave social housing provide some insights into where they may end up. Figure 7 presents data from the social housing projects managed by Madulammoho (10 projects across Johannesburg and Cape Town, including BG Alexander, Belhar and Scottsdene) for the period November 2019–February 2020. The story that emerges is certainly not one of consistent upward mobility. In fact, the main reason people left social housing was because they could no longer afford the rent (33%). This included eviction in some instances. This is somewhat surprising considering the social purpose of the programme, although SHOs work on tight financial margins and cannot sustain non-payment by tenants for long. The fact that households in financial distress are obliged to leave is a concern, and more information is needed on how strictly or leniently their cases are managed. It may be necessary for the SHRA to consider some form of safety net to help those who experience a temporary loss of income to get back on their feet and start paying their rent again. It is not difficult to envisage how eviction could be a serious setback for families.

Fig. 7

Exit reasons. Source: Madulammoho (2020)

Most other reasons cited by tenants leaving Madulammoho were neutral. Some were ‘right sizing’ in moving to another social housing unit (18%), others were relocating to another city (11%), and a third group was leaving for family reasons (3%). There were also some notable positive cases. We assume that all work-related cases (14%) were positive although they were not necessarily linked to a promotion and/or increase in earnings. Only 7.5% of tenants left the social housing specifically because they were moving into home ownership. Further research is needed to understand the detailed circumstances that enabled this to occur, and whether living in social housing had played a part in laying the foundations.

Considered all together, the evidence related to social mobility is patchy. Social housing did seem to make a once-off improvement to many people’s lives, mainly because the rent was more affordable. Yet, there was little sign of more profound or abiding improvements in life chances, whether through better access to economic opportunities, improvements in the quality of schooling, or the SHO’s personal and community development activities. Part of the explanation could be that many households were already living in the same neighbourhood. One might expect stronger evidence of upward mobility if there was greater focus on household upliftment, hand-in-hand with the core function of delivering decent, affordable housing.

Conclusions and Recommendations

South Africa’s social housing policy has ambitious and important goals to narrow socio-economic and spatial inequalities. Social housing could play a valuable role in helping to build an integrated, non-racial and prosperous society. Improving the well-being and life chances of individual households is an essential and integral part of achieving wider social and spatial changes. By providing a decent and secure home base with improved access to economic opportunities, public facilities and social amenities, social housing can help families to get on in life and thrive. The paper has elaborated multiple ways in which such accommodation can bolster household capabilities and connections to achieve enduring social outcomes. This amounts to a compelling rationale for social housing to help overcome the entrenched spatial and social divides in South African cities and promote shared prosperity.

Yet, social housing policy has been couched in the somewhat abstract and rarefied language of social and spatial transformation. Policy-makers have not translated these lofty goals into specific objectives and procedures for social housing projects and practices. The implications of social transformation for practical actions towards individual households have not been spelt out. Consequently, the policy does not indicate the particular ways in which progress could be achieved at the household level. There is no specific guidance given to SHOs indicating how they should promote household advancement and what pathways and support measures are most likely to produce sustained upward mobility. In addition, SHOs are not monitored or evaluated on the basis of the socio-economic status of households and their progression over time, so they are not held accountable for this goal. Even when tenants secure a sizeable increase in their disposable income, this is ignored by SHOs and the SHRA because of the sensitivity of collecting information on income.

Not surprisingly, the focus of policy and practice is on delivering the core product of subsidised housing, namely liveable and affordable dwelling units. It is entirely appropriate that national government, the SHRA and SHOs treat this as the primary function of social housing. A safe and sound home base is a precondition for people to live more stable and happier lives. It is also understandable that the national government and the SHRA have placed great emphasis on the sheer quantity of housing delivery, because the sector’s track record has not been strong in this respect and budgets have been under-spent (SHRA, 2018; Scheba & Turok, 2023). In addition, long waiting lists of households eager to access social housing put pressure on the sector to accelerate production.

Yet, some of the central goals of social housing seem to have been neglected in the drive to hasten delivery. One of the objectives that appears to have been undervalued is the enduring developmental outcome for tenants. Insufficient attention has been paid to ways of accelerating household upliftment and preventing people from sliding back into poverty. In the absence of policy guidance and ongoing monitoring, SHOs have been left to decide for themselves how much importance to attach to two interrelated issues that underpin upward mobility: (i) tenant development and (ii) the location of social housing projects. The result is that practice is very uneven between projects—some SHOs are committed to household improvement and take these matters seriously, while others are more likely to set them aside.

We believe that the lack of a concerted policy towards location and tenant development helps to explain the very diverse outcomes apparent in the analysis. It seems that a range of other factors have intervened to the extent that many social housing projects are not very well located, and many tenants have not improved their circumstances to the extent that one might have anticipated. Depressed economic conditions and chronic unemployment have no doubt also contributed to the general lack of mobility. In addition, the affordability gap between social housing and the first rung on the private housing ladder may just be too large for many households to bridge across. Having said all this, social housing does not appear to have done all it could to lift people out of poverty by maximising the opportunities for them to progress.

Looking to the future, it seems that for social housing to be more effective as a springboard to alter people’s life chances, more emphasis will need to be given by the SHRA and SHOs to mechanisms to facilitate upward mobility and ensure enduring household progress. This will also require more practical experimentation by different SHOs and more investment in research and data collection to improve understanding of how social housing affects pathways to household prosperity, and what kinds of actions and interventions are most effective at strengthening positive outcomes.