Revisiting the capacity of social housing delivery in South Africa: an institutional and comparative analysis

1. Introduction

The question of whether non-profit institutions can deliver social housing at scale has long been contested in South Africa. Earlier critiques, including my 2006 paper, argued that Social Housing Institutions (SHIs) underperformed because they were insulated from ‘market discipline’, which in turn weakened accountability and reduced incentives for efficiency.¹Hendler, P 2006 Are NGOs Relevant to the Delivery of Social Housing on Scale?, INSITE.

This argument was both analytically coherent and empirically grounded in the performance of the sector at the time. I wrote this paper shortly after five years working for a for-profit company – Afhco, now with 5 500 units under management – renting accommodation in the Johannesburg inner city, as a marketing director and then also engaging with the Gauteng department of housing to secure institutional subsidies for the refurbishment of inner city buildings for rental. However, nearly two decades later, the evolution of the sector, together with international comparative evidence, necessitates a reconsideration of that position.

This is also the conclusion from our company INSITE’s 2022 report to the SHRA: Social Housing Base Line Report: socio-economic impact, achievements and limitations/challenges. Notwithstanding over R5 billion invested in social housing over 20 years, the INSITE report noted that social housing comprised a mere one per cent of the residential rental market.

2. Theoretical Framework

This analysis is grounded in a hybrid framework that integrates principal–agent theory, housing regime theory, and developmental state theory. Each of these perspectives illuminates a different dimension of the problem, and together they provide a coherent explanation of why social housing systems succeed or fail.

Principal–agent theory, rooted in institutional economics, examines the challenges that arise when one actor (the principal) delegates responsibility to another (the agent). In the context of social housing, the state delegates delivery to SHIs, expecting them to provide affordable rental housing efficiently and at scale. However, where monitoring is weak, information is asymmetric, and incentives are poorly aligned, agents may not perform optimally, giving rise to moral hazard and weak accountability.²Jensen, M & Meckling, W 1976 Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. See pp. 305–308 for definition of agency relationships and agency costs. My earlier critique of SHIs is best understood within this framework. However, because principal–agent theory operates primarily at the organisational level, it cannot fully explain system-wide constraints.

Housing regime theory shifts attention to the structure of housing systems. It distinguishes between unitary systems, where non-profit housing operates at scale and competes with private provision, and dual systems, where social housing remains marginal.³Kemeny, J 2001 Comparative housing and welfare: Theorising the relationship. See pp. 54–56 for unitary vs dual systems. The South African system exhibits many characteristics of a residual or dual system, which limits the ability of social housing to scale and to shape broader market outcomes.

Developmental state theory provides a macro-institutional lens, emphasising the role of the state in coordinating policy, mobilising resources, and shaping markets. A developmental state is characterised by policy coherence, strategic allocation of resources, effective land management, and strong inter-institutional coordination.⁴DNA Economics / DPME, 2017 Political-Administrative Interface DPME Research Unit 13 June 2019 Literature Review The Developmental State An exploration of the Concept in Relation to Chapter Thirteen of South Africa’s National Development Plan 2nd Draft. See pp. 9–10 for embedded autonomy and coordination. In the housing sector, this implies aligning subsidies, land, and municipal systems with affordability objectives.

Taken together, these frameworks support a central proposition: social housing delivery depends on the alignment of institutional incentives, system structure, and state capacity. The South African case demonstrates that while principal–agent problems persist, the dominant constraint lies in the incomplete realisation of the developmental state function.

3. International Counterfactuals

International experience demonstrates that non-profit housing is not inherently inefficient. In the Netherlands, housing associations manage approximately 2,3 million dwellings, representing about one-third of the national housing stock.⁵Van Deursen, H 2023 The People’s Housing: Woningcorporaties and the Dutch Social Housing System Part 1: The History, p. 4. States 2,3 million dwellings owned by housing associations. In Austria, limited-profit housing associations operate on a cost-rent basis and manage hundreds of thousands of units within a system designed to ensure long-term affordability.⁶Kössl, G 2022 The system of limited-profit housing in Austria: cost-rents, revolving funds, and economic impacts, pp. 3–4; 12–13. Defines cost-rent system and revolving funds. In the United Kingdom, private and local government housing associations together own 4,5 million homes within a regulated system of social housing provision.⁷GOV.UK – Regulator of social housing (website), 2025 RSH publishes annual statistics on stocks and rents in the social housing sector. Line 122 (footnote 6): private housing associations own 2,9 million homes while local authority housing associations own a further 1,6 million homes.

These systems share a common feature: non-profit provision is embedded within strong institutional, financial, and regulatory frameworks. Therefore, the shortcomings of the South African system cannot be attributed to the non-profit model itself.

4. Evolution of Social Housing in South Africa

The South African sector has evolved significantly. The early phase was marked by weak institutional capacity and poor delivery (as described in my 2006 article). The establishment of the Social Housing Regulatory Authority (SHRA) improved governance and introduced greater structure, but scaling remained limited.

More recent performance data indicate progress, with approximately 13,885 units delivered between 2019 and 2024 and 3,066 units completed in 2023/24,⁸ SHRA, 2024 Annual Report 2023/24 Financial Year, pp. 10–11. p.10 refers to affordability constraints, p.11 to delivery issues. according to SHRA. At the municipal level, Cape Town has delivered thousands of units across multiple projects.⁹City of Cape Town, undated Integrated Human Settlements Five-Year Sector Plan 2022/23 – 2026/27: 2025/26 Review, p. 94. However, these gains remain modest relative to need, and delivery is increasingly constrained by affordability pressures and rising service costs, according to the SHRA 2023/24 Annual Report.

5. Capacity Constraints

Capacity across the sector remains uneven. Business systems have improved but remain fragmented, while staffing and expertise are concentrated in a small number of institutions.¹⁰Department of Planning Monitoring and Evaluation (DPME), 2022 Impact and Implementation Evaluation of the Social Housing Programme, pp. 3–4. p.3 states that the programme has not been part of a coordinated restructuring framework because of inconsistent definitions, policies, plans and funding frameworks, and a lack of inter-governmental coordination. The same page also states there is ‘no formal strategy for growing the SH sector,’ and that there is a ‘lack of a rental housing policy and an incoherent subsidy support framework.’ Page 4 adds that there are very few capacitated SHIs and a lack of an agreed SHI growth strategy.

The most significant constraints, however, relate to funding and land. The sector relies heavily on capital subsidies that have not kept pace with rising development costs. At the same time, there is limited provision for ongoing affordability support. Evidence in the 2023/2024 SHRA Annual Report shows that affordability constraints—particularly those linked to municipal service charges—are now a major factor limiting tenant uptake.

This is compounded by systemic issues in municipal finance. Free basic services fall significantly below actual household needs, and many social housing tenants are unable to access indigent subsidies.¹¹Parliamentary Budget Office, 2023 Brief on 2023 Division of Revenue Bill and 2022 Second Adjustments Appropriations for Select Committee on Appropriation, p. 12. ¹²National Association of Social Housing Organisations (Nasho), 2024 Nasho Objection to Eskom’s Proposed (MYPD6) Revenue Application FY 2025/26 – 2027/28, p. 6. At the same time, municipalities face fiscal pressures. National Treasury documents widespread financial stress across local government; this suggests that municipalities may have limited fiscal room to expand subsidies for poor households, although the Treasury report itself does not directly state this conclusion.¹³National Treasury, 2024 The state of Local Government Finances and Financial Management as at 30 June 2024, pp. 6–8.

Land constraints further exacerbate these challenges. Well-located land is scarce and expensive, and public land release processes are slow and inconsistent. Studies indicate that land cost assumptions underpinning subsidy models are often unrealistic, making discounted or publicly provided land a decisive factor in project viability.¹⁴Turok, I et al, 2021 The role of social housing in reducing inequality in South African cities, final draft for AFD Research Papers, pp. 8, 33, 37, 55. P. 5 states that focusing on land already in public ownership avoids acquisition costs and improves project viability. Pp. 5 and 6 imply that restraining land costs is necessary to minimise rental levels. Pp. 6 and 44 imply that increasing property and building prices make it difficult to deliver social housing in well-located areas unless land is made available at discounted prices. Pp. 43 and 44 suggest that well-located public and state-owned enterprise (SOE)-owned land is often not prioritised for social housing.

This funding constraint becomes even more apparent when viewed against international experience. In countries such as the Netherlands and Austria, the viability of non-profit housing is not achieved through capital subsidies alone, but through a combination of deep upfront investment, access to low-cost finance, and system-wide cost moderation mechanisms (as explained in the referenced sources in endnotes 5 and 6). In these contexts, affordability is produced at a system level rather than at the level of individual projects. By contrast, the South African model relies heavily on once-off capital grants while expecting rental income to absorb operating costs and rising service charges. The absence of comparable operating support mechanisms therefore creates a structural affordability gap, undermining both tenant uptake and long-term sustainability.

6. Weighted Capacity Analysis

The relative importance of constraints reflects their systemic impact. Policy is the primary constraint because it shapes all other dimensions of capacity. Misaligned subsidy design, weak coordination, and the absence of an integrated rental strategy create structural limitations, as demonstrated by the Department of Planning Monitoring and Evaluation (DPME) report (referred to in endnote 10).

Funding and land are secondary but binding constraints. Without sufficient subsidy depth or access to affordable land, projects cannot proceed, regardless of institutional capacity.

Institutional capacity is a tertiary constraint. While uneven, it is not absent, and strong SHIs demonstrate that delivery is possible within existing structures, as noted by the DPME.

7. Conclusion

The evidence shows that social housing in South Africa has achieved institutional viability but not systemic scale. The primary constraint lies in the incomplete alignment of policy, funding, land, and institutional systems. To achieve this alignment, including sufficient quantums of subsidy as well as hectares of affordable land, requires a shift in thinking from market-led development to state intervention development, the home ownership market in Singapore being a recent historical example of this.

It was with this in mind that INSITE submitted a detailed proposal to the Department of Human Settlements, through the SHRA, in 2022 for a revised social housing policy. In terms of this proposal an additional 60 000 social housing units would be delivered over 10 years.

8. Implications for Policy Reform

Reform must focus on strengthening the developmental state function. This requires:

• Re-aligning subsidy systems with affordability realities
• Integrating housing and municipal finance
• Reforming land release systems
• Strengthening coordination across government
• Supporting institutional capacity

Final Insight

Social housing does not fail because it is non-profit. It struggles because it operates within a system that has not yet fully aligned policy, finance, land, and institutional capacity with its stated objectives. Therefore the inclusion of for-profit companies – so called other deliver agents (or ODAs) – while this might leverage off greater institutional delivery capacity, will not alter the limited output of social housing in the years ahead which is affordable for the stated target markets.

STATEMENT: During the preparation of this article the author used ChatGTP Plus in order to summarise his previous articles and to search for new information and data on the topic to update his analysis. After using this tool/service, the author reviewed and edited the content as needed, checked source references for authenticity and takes full responsibility for the content of the publication.

References

• 1
  Hendler, P 2006 Are NGOs Relevant to the Delivery of Social Housing on Scale?, INSITE.
• 2
  Jensen, M & Meckling, W 1976 Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. See pp. 305–308 for definition of agency relationships and agency costs.
• 3
  Kemeny, J 2001 Comparative housing and welfare: Theorising the relationship. See pp. 54–56 for unitary vs dual systems.
• 4
  DNA Economics / DPME, 2017 Political-Administrative Interface DPME Research Unit 13 June 2019 Literature Review The Developmental State An exploration of the Concept in Relation to Chapter Thirteen of South Africa’s National Development Plan 2nd Draft. See pp. 9–10 for embedded autonomy and coordination.
• 5
  Van Deursen, H 2023 The People’s Housing: Woningcorporaties and the Dutch Social Housing System Part 1: The History, p. 4. States 2,3 million dwellings owned by housing associations.
• 6
  Kössl, G 2022 The system of limited-profit housing in Austria: cost-rents, revolving funds, and economic impacts, pp. 3–4; 12–13. Defines cost-rent system and revolving funds.
• 7
  GOV.UK – Regulator of social housing (website), 2025 RSH publishes annual statistics on stocks and rents in the social housing sector. Line 122 (footnote 6): private housing associations own 2,9 million homes while local authority housing associations own a further 1,6 million homes.
• 8
  SHRA, 2024 Annual Report 2023/24 Financial Year, pp. 10–11. p.10 refers to affordability constraints, p.11 to delivery issues.
• 9
  City of Cape Town, undated Integrated Human Settlements Five-Year Sector Plan 2022/23 – 2026/27: 2025/26 Review, p. 94.
• 10
  Department of Planning Monitoring and Evaluation (DPME), 2022 Impact and Implementation Evaluation of the Social Housing Programme, pp. 3–4. p.3 states that the programme has not been part of a coordinated restructuring framework because of inconsistent definitions, policies, plans and funding frameworks, and a lack of inter-governmental coordination. The same page also states there is ‘no formal strategy for growing the SH sector,’ and that there is a ‘lack of a rental housing policy and an incoherent subsidy support framework.’ Page 4 adds that there are very few capacitated SHIs and a lack of an agreed SHI growth strategy.
• 11
  Parliamentary Budget Office, 2023 Brief on 2023 Division of Revenue Bill and 2022 Second Adjustments Appropriations for Select Committee on Appropriation, p. 12.
• 12
  National Association of Social Housing Organisations (Nasho), 2024 Nasho Objection to Eskom’s Proposed (MYPD6) Revenue Application FY 2025/26 – 2027/28, p. 6.
• 13
  National Treasury, 2024 The state of Local Government Finances and Financial Management as at 30 June 2024, pp. 6–8.
• 14
  Turok, I et al, 2021 The role of social housing in reducing inequality in South African cities, final draft for AFD Research Papers, pp. 8, 33, 37, 55. P. 5 states that focusing on land already in public ownership avoids acquisition costs and improves project viability. Pp. 5 and 6 imply that restraining land costs is necessary to minimise rental levels. Pp. 6 and 44 imply that increasing property and building prices make it difficult to deliver social housing in well-located areas unless land is made available at discounted prices. Pp. 43 and 44 suggest that well-located public and state-owned enterprise (SOE)-owned land is often not prioritised for social housing.