Why can't we clear the housing backlog? - IRR

Mary Tomlinson writes on why, despite the enormous delivery that has taken place, more progress hasn't been made

South Africa’s Housing Conundrum

Since 1994 the Government has provided more than 2.5 million houses and another 1.2 million serviced sites. Over this period, the housing backlog has nevertheless increased from 1.5 million to 2.1 million units, while the number of informal settlements has gone up from 300 to 2 225, an increase of 650%.

At the same time, the housing subsidy has gone up from R12 500 per household to some R160 500 today, while state spending on housing and community amenities will soon have risen from 1% to 3.7% of GDP. Yet the quality of the houses being delivered is often poor. Many people have long been urging the State to transfer the housing subsidy directly to them, as they could build far better houses for themselves. In this article, housing expert Mary R Tomlinson seeks to explain the housing conundrum and offers some thoughts on how it might be resolved.


South Africa’s housing policy has now been in place for some 20 years. It arose from debates at the National Housing Forum (NHF), a multi-party organisation established in 1992 to negotiate a new non-racial housing policy and strategy. These debates focused, in particular, around such issues as:

- whether housing should be provided by the private or the public sector;

- what standard of housing should be provided: a completed four-room house or “progressive” (incremental) housing; and

- how rapidly the housing backlog should be eliminated.

The National Housing Policy framework that arose from this process put forward a vision and goal for housing delivery. It also set out the principles and strategies to be used in fulfilling these aims. Through a new “housing subsidy scheme”, qualifying households were to have access on a “progressive” basis to:

- a permanent residential structure with secure tenure, ensuring internal and external privacy and providing adequate protection against the elements; and

- potable water, adequate sanitary facilities, and domestic energy supply.

The delivery goal was set at 350 000 units per annum until such time as the housing backlog had been removed, though the policy also noted that meeting this goal would have to be subject to government fiscal affordability. In 1994, the backlog was put at 1.5 million units for the 1.5m households then “inadequately housed” because they lived in informal settlements, backyard shacks, or hostels. The policy also noted that a further 200 000 units a year would be needed to keep up with new family formation.

Since 1994, more than 3 million housing units (including both subsidised and rental housing) have been delivered to poor and low-income households. South Africa nevertheless sits today with a housing backlog of 2.1 million units. How can this be? More precisely, how can South Africa have spent R125 billion, in 2010 prices, over 20 years, delivered more than 3 million units, and yet have a larger housing backlog than when it began?

The aim of this paper is to put forward some possible explanations for this conundrum. It will thus re-visit some of the negotiated agreements that underpin housing policy, identify some of the significant policy shifts and tensions that have arisen in implementing policy, and evaluate whether these factors have contributed to the unintended consequence of a seemingly bottomless backlog. Some ideas on how to resolve these tensions in a way that would result in both delivery at scale and satisfied beneficiaries are also put forward.

Increasing housing backlog figures and the constitutional mandate

In reviewing South Africa’s housing policy, it is useful to begin by examining housing delivery statistics. As earlier noted, the Government since 1994 has delivered more than 2.5 million completed houses and another 1.2 million “housing opportunities” (where housing subsidies have been approved and water and sanitation services have been installed, but construction has not yet been finished).

As impressive as these figures sound, they becomes less so when stacked up against the housing backlog. This amounted to 1.5 million units in 1994, was put at 2.7 million units in 2012, and now stands at some 2.1 million units. This last figure – the one most recently computed by the Department of Human Settlements – is based on the number of people with their names on the national housing waiting list. (The reasons for these differences lie partly in the fact that backlog statistics are neither systematically compiled nor located in a single place, which leaves them open to varying interpretations.)

In addition, in 1993 there were an estimated 300 informal settlements across the country, whereas by the 2009/10 financial year the number had risen to some 2 700, an increase of almost 800%. More recent figures provided by the Department of Human Settlements put the total number of informal settlements at 2 225, of which some 1 260 are located in KwaZulu-Natal, Gauteng, and the Western Cape.

This large increase in the number of informal settlements since 1994 shows that seemingly impressive delivery figures have in fact been insufficient either to clear the backlog or to keep up with new family formation. Why has this occurred? Did policy-makers overlook the likelihood that rural households would continuously move into urban areas over the next two decades? Did they also overlook the fact that households residing in urban areas would continually outgrow their crowded homes and so spill over into informal settlements?

Neither of these explanations is convincing. At the time that housing policy was being formulated, policy-makers were well aware of these potential trends and took them into account in their forecasts of what annual new family formation was likely to be. Today’s backlog numbers are thus more likely the result of how policy has in practice been interpreted over the past two decades, rather than how it was intended to be implemented.

One key issue is that the housing policy adopted in 1994 has no “grandfather” clause limiting the provision of housing to those who lived and suffered under apartheid. There is also no time limit to the constitutional clause (Section 26) stating that “everyone has the right to have access to adequate housing”. South Africans born both before, as well as after, the end of apartheid thus take the view that “they have a right to a free house” if their monthly income is below R3 500 per month.

Accordingly, households continually break themselves up into smaller units in the expectation that each new unit will become entitled to a housing subsidy. National, provincial and local government housing officials (interviewed by the author in 2014) all refer to this as the growth of an “entitlement syndrome” and say it has made the goal of eliminating the housing backlog simply unattainable.

In October 2014 the minister of human settlements, Lindiwe Sisulu, tried to break the logjam by announcing that no one under the age of 40 would receive a free house. Said Ms Sisulu: “Our intention in giving free houses was to right the wrongs of the past and make sure that we can give our people dignity. And that group of people is not the people below the age of 40.” Young people, she went on, needed to clearly understand that they would not receive free housing. “You the young people have lost nothing to apartheid... None of you are ever going to get a house free from me while I live.”

Some commentators applauded her for attempting to encourage self-reliance and to limit the housing demand on the public purse. Others said her comment was “premised on the false belief that the past does not affect someone under the age of 40”. Moreover, it was the Government itself which had fostered a culture of entitlement and raised “false expectations”. It was also the Government that had failed to “generate enough opportunities for young people”, who were often unskilled and unemployed and would find it difficult to buy houses for themselves.

Decreasing housing delivery figures and the role of politicians

Alongside the backlog figure is the worrying fact that delivery figures have been decreasing over time. In the early years of the policy’s implementation, approximately 200 000 housing units per annum were being constructed. Delivery figures show that, from a peak of 235 600 units in the 1998/99 financial year, housing delivery has decreased to some 106 000 units in 2013/14.

The national housing department attributes declining delivery in the first ten years after 1994 to the slow release of land for housing and often unsettling policy shifts, among other things. However, to understand the sharp decrease in delivery, it is also necessary to revisit the difference between policy intentions and the way these have been interpreted in practice.

Part of the problem, as earlier noted, is that poor households have interpreted the constitution’s housing mandate as giving them “a right to a free house”. This interpretation has never been in keeping with the agreement reached at the NHF, which was to deliver “progressive” or incremental housing. In addition, the relevant constitutional clause is carefully phrased, stating that people have a right of “access to adequate housing”, the key words here being “access” and “adequate”. The clause also gives the State the task of “achieving the progressive realisation” of this right, while recognising that it can do so only within the limits of “its available resources”.

It is also worth recalling the terms of the agreement, hammered out in the NHF, that the Government’s delivery approach would focus on “breadth” rather than “depth”. This meant, in essence, that the Government would provide a lesser standard of housing to as many people as possible, rather than a higher standard of housing to fewer people.

This agreement was intended to help realise the goal of eliminating the housing backlog within a five-year period. In line with this objective, the “national housing subsidy” was intended to provide beneficiaries with their first step on to the housing ladder. Beneficiaries would thus be given a serviced site, together with a rudimentary structure and secure tenure. The intention was that, over time, households would consolidate and extend what they had received from the Government. By using their own resources and gaining access to housing finance, if this was available, they would eventually end up with the desired four-roomed house.

However, after the 1994 election that brought the Government of National Unity to power, the first group of provincial housing MECs rejected this approach and began exerting pressure on the Government to deliver formal houses. Ever since, both politicians and communities have persistently demanded increases in the subsidy amount, so that this would become big enough to build a four-room house. Over time, the housing subsidy has thus grown from its original R12 500 per household to R160 500 per household in 2014, an overall increase of some 1 200%. Today, moreover, this increased housing subsidy is intended to cover only the construction of a house, with land and service costs coming out of provincial and local government budgets.

In practice, thus, South Africa’s housing policy has seen its focus turn from “breadth” to “depth”. This has mixed ramifications. On the one hand, the standard of housing delivered today has improved significantly from what was provided in the early years. On the other hand, this improvement has come at the cost of shrinking delivery figures. It has also seen expenditure on housing grow faster than any other budget item, including social grants.

In 1994 government spending on housing and community amenities accounted for 1% of gross domestic product (GDP). In 2015/16, according to the current budget, expenditure in these spheres will have risen to 3.7% of GDP, making for an overall increase of 270%. This is a faster rise even than expenditure on social grants and other forms of social protection, which is expected to rise from 2.8% of GDP in 1994/95 to 3.8% of GDP in 2015/2016, an increase of 36%. Already, annual spending on housing and community amenities consumes 11.4% of government expenditure, which is almost as much as health.

In seeking to understand what has happened over the past 20 years, it is often easy to overlook the significance of the political dispensation agreed in constitutional negotiations. The African National Congress (ANC) was keen on a unitary system of government, with a strong national government and weak provinces.

Housing delivery would then, it says, have been directed by a national housing department with clear authority to decide on the standard of housing to be delivered. Instead, a more federal system was adopted, in which provincial administrations have “concurrent” legislative jurisdiction with the national government in various spheres, including housing.

Though the national government still has the power to override provincial decisions and require uniform “norms and standards” in wide-ranging circumstances, in practice the provinces have assumed an authority to decide for themselves what standard of housing they wish to provide. Provincial administrations also face unrelenting political pressure to deliver a house, rather than a serviced site. The majority of resources allocated to housing delivery have thus ended up being concentrated on the few rather than the many.

Conflicts over powers and functions: local versus provincial government

The Constitution divides governmental functions between the national, provincial, and local tiers. When South Africa’s housing policy was being formulated, local government was seen as playing little part in the delivery of housing and services. Such delivery was instead to be driven by private sector developers, as this seemed the best way of bypassing the chaos likely to result from the thorough-going “transformation” of local government then being planned.

However, early beneficiaries of the rudimentary RDP (Reconstruction and Development Programme) houses delivered by the private sector soon expressed great unhappiness over the size, quality, and location of their new homes. (This was also hardly surprising when these houses were being largely financed by a R12 500 per household subsidy which was supposed to cover land and services as well as the house itself.) The national government’s solution was to transfer responsibility for managing and implementing housing developments to the public sector.

This shift is reflected in the National Housing Act of 1997, which makes it clear that the private sector is no longer to drive delivery but rather to act as contractors to the public sector. The statute also states that municipalities, in preparing their “integrated development plans (IDPs)” – which are supposed to reflect a holistic approach to housing and service delivery – must take all reasonable steps to ensure that people in their areas of jurisdiction have access to adequate services and housing. The Act also envisages a potential role for local government to participate in housing development.

However, this shift towards public sector responsibility for housing and service delivery, including at the local level, overlooks the parlous state of local government capacity and finances. In particular, it brushed over the fact that South Africa’s 278 municipalities have vastly different capabilities – and that only about 50 of them have dedicated housing departments. Capacity therefore differs widely from well-staffed departments in metropolitan areas to poorly resourced secondary cities. Rural small towns often have only a single housing official, which makes it impossible for them to deliver on the housing mandate.

Under the Housing Act, the provinces have largely assumed responsibility for managing housing development. Provincial departments approve projects, allocate the necessary revenue and other resources, hire and manage building contractors, and administer the subsidy scheme. Local authorities may have these powers “delegated” to them, but only if they have already been “accredited”’. A framework for accrediting local authorities was finally put in place in 2006, but to date only a handful of local authorities, the metros, have been granted a limited form of accreditation, as further described below.

Conflicts regularly arise in situations where a local authority lacks “accreditation”, for the province then retains authority to manage housing developments within the local authority’s area of jurisdiction. This often puts municipalities in an invidious position as regards their constituencies.

Households within their areas of jurisdiction naturally view the local authority as responsible for deciding what type of housing is delivered, who does the construction, and which households are fortunate enough to be granted housing subsidies. Because of this blurring of responsibilities, communities often accuse local officials of non-delivery and corruption – when in fact their powers over housing are very limited.

From RDP housing to informal settlement upgrading

Criticism of housing delivery has generally revolved around what is in fact “adequate”. Communities and housing actors generally agree that houses are often badly built, poorly located, and too small. Even after the passage of the Housing Act in 1997, delivery continued to follow the same approach.

People living in informal settlements were thus removed to new “green-field” sites, located on the peripheries of cities and towns, where they were supplied with a uniform and limited product. Beneficiaries were dissatisfied with how little input they could make into the delivery process. They often said that they could build bigger and better houses for themselves if they were simply given direct access to the housing subsidy.

This persistent dissatisfaction with housing quality prompted the national department to embark on a major re-think of policy. So too did the department’s exploration of Indian and Latin American experience in dealing with shacks and slums, and its engagement with a group of NGOs active within informal settlements. Particularly important here were members of the Urban Sector Network and representatives of the Indian Shack/Slum Dwellers International (SDI) alliance.

These factors led to the adoption in 2004 of a revised housing policy, called Breaking New Ground: a comprehensive plan for the development of sustainable human settlements (BNG). BNG rejected the idea of small RDP houses on poorly located land and instead proposed the delivery of “sustainable human settlements”. This dramatic about-turn was accompanied by an equally notable shift: a new-found acceptance of informal settlement upgrading as an acceptable delivery option.

Upgrading of informal settlements

The BNG policy shift has led to the establishment of a number of programmes intended to support in-situ upgrading. These include the Upgrading of Informal Settlement Programme (UISP); the National Upgrading Support Programme (NUSP), which seeks to build capacity among provincial and local officials; and an Extended People’s Housing Programme, as further described in an explanatory article. However, it is difficult to assess how much progress has in fact been made in upgrading informal settlements, as the BNG document urges.

In 2009/10 the national department said it had identified “at least 2 700 informal settlements” across the country, of which most (635) were found in KwaZulu-Natal, followed by Gauteng (with 489) and the Western Cape (with 445). It was busy giving formal recognition to these settlements, and then re-establishing them as serviced sites with water and sanitation infrastructure provided by municipalities. This was mostly being done via the Upgrading of Informal Settlements Programme, under which the State provides serviced sites to residents to improve their quality of life but does not guarantee the provision of an accompanying top structure. (This is because the people living in informal settlements may not qualify for the housing subsidy: for example, because they are not married, or lack South African citizenship or permanent residence.)

More recent figures from the national department put the number of informal settlements, in March 2014, at 2 225 – a significant (475) decrease from the 2 700 such settlements earlier identified. Decreases are most marked in KwaZulu-Natal (511 in 2014, versus 635 before); the Western Cape (350 in 2014, from 445 before); and Gauteng (399 in 2014, from 489 before). According to the national department, it has delivered more than 322 000 serviced sites over the six years to March 2015, which may have further reduced the number of informal settlements. The department adds that it encourages people to invest in their own housing, and that a “significant” number of homes have been built by people provided with serviced sites. However, it has no statistics on this readily available.

In-situ upgrading is still the “Cinderella” element in the overall housing programme, even though the Government has recognised since the BNG policy shift in 2004 that it cannot provide a new formal house for everyone. In-situ upgrading is also essential in an era of rapid urbanisation and significant internal migration, mostly to Gauteng and Western Cape.

Why has in-situ upgrading progressed so slowly? This is partly because provinces prefer to concentrate on formal housing, while municipalities generally still lack accreditation for housing development. Local authorities may also be reluctant to extend water and sanitation services to areas where residents are unlikely to be able to pay for them, as this will simply add to existing pressures on municipal budgets. (As the BNG document noted in 2004, municipalities saw RDP housing projects as liabilities rather than assets, because of “the inability of recipients of subsidy-housing to pay for municipal services”; and the affordability problem could well be worse in informal settlements.) In addition, the upgrading process is often technically complex and fraught with intra-community tensions and conflicts.

Much of the difficulty stems from a lack of clear leadership structures within informal settlements. In many of these areas, shacks are also built cheek-by-jowl, with little space between them for access roads or community facilities. Bitter conflicts can easily arise over the location of roads, schools, clinics, and the like. Often, moreover, no development is possible without taking down some of the shacks and at least temporarily relocating the people living there, while the affected residents often resist being moved in this way. To help resolve these issues, the BNG document puts great emphasis on community participation, but progress here has also been slow and uncertain.

Community participation and engagement

The People’s Housing Process (PHP) aims to help households take the initiative in organising the planning, design, and building of their own homes. In 1998 the first policy to facilitate this process was drawn up by the Government, which in the same year decided to establish formal links with a grassroots fund, the uTshani Fund. This Fund was based on the idea that savings groups linked to the South African Homeless People’s Federation (SAHPF) would be able to make deposits into it and also access loans for housing from it. In 1998 the Government “accredited” the Fund, so as to facilitate the channelling of housing subsidies to these savings groups. It later also provided the Fund with R10 million to help it grant credit for housing to these groups.

But the PHP has nevertheless struggled to gain traction, while the uTshani Fund collapsed in 2001. Critics argue that this is largely because the Government continued to see informal settlements “only in negative terms” and had no real interest in their incremental upgrading. It also showed “very little support for low-income households to build their own homes”. Hence, the State gave little space to the poor to participate in either the formulation or implementation of housing policy, while its overwhelming focus remained on the delivery of “formal” housing at scale.

In 2006 the Government tried to rectify this situation by instructing the national department to give more emphasis to the PHP. The Government also began to engage with the recently established Federation of the Urban Poor (FEDUP), which had risen from the ashes of the SAHPF, managed to salvage the uTshani Fund, and established a new Community Organisation Resource Centre (CORC).

FEDUP had also developed a participation model that encouraged communities to clarify their wishes, improve their leadership skills, and organise themselves for effective engagement in development processes. FEDUP had also developed an “empowerment through engagement” model, which aimed to change the way in which solutions were defined, negotiated, and implemented.

In 2007 Ms Sisulu (the then national housing minister) promised the FEDUP-CORC-uTshani alliance 6 000 housing subsidies to be used in upgrading informal settlements and building houses on the PHP approach. But the 1997 National Housing Act gives responsibility for allocating subsidies to the provinces, which failed to make the allocation pledged. These provincial MECs no doubt wanted to see their housing budgets spent in a timely manner, whereas upgrading informal settlements is an extremely time-consuming process with many moving parts that can easily break down, thereby stalling or shutting down delivery altogether.

It is also difficult for NGOs and CBOs to help overcome these problems and become more involved in settlement upgrading. Though many officials would welcome their assistance in strengthening community participation, few of these organisations have the formal legal structures and controls required under the Municipal Finance Management Act of 2003 and other legislation. With public concern about financial mismanagement growing, provincial housing departments find it easier to contract with private sector firms able to meet these compliance requirements. Hence, if CBOs and NGOs are to play a greater part in upgrading, they need to structure themselves into formal entities with the ability to tender and contract in accordance with the Government’s regulatory framework.

Communities are often unhappy with this situation and argue that the Government should simply hand over the subsidy (or, at least, part of it) to them directly, so leaving them free to manage their own housing delivery. This demand forms part of an on-going debate as to whether households should be viewed as “beneficiaries” of government resources, or as “consumers” with their own preferences and capacity to “buy”. Communities generally take the view that they should be treated as “consumers”.

Some issues to consider

In looking back over twenty years of housing policy and its implementation, it is notable how much effort has been made to engage with critics, improve delivery, and satisfy beneficiaries. But policy still confronts the conundrum of a seemingly intractable housing backlog, notwithstanding the enormous delivery that has taken place. The debate about the need to increase community participation is also on-going, as is the tension between formal statutory processes and more fluid community-driven ones. To re-cap:

First, the “right to housing”, as provided for in the Constitution, is the cornerstone of housing policy and is likely to remain so. However, this “right” has been re-interpreted from its original intention, from the delivery of “progressive” housing to the delivery of formal homes. Over time, this shift has meant that an approach that was initially intended to deliver “breadth” over “depth” has shifted to one today that focuses on “depth” over “breadth”. This situation makes it impossible to remove the backlog with the resources available. It may therefore be time to re-visit this change in interpretation.

Second, housing policy has become much more complicated over time and so requires significant capacity to implement, which many local authorities lack. In addition, for local authorities to become housing developers, they must first be fully “accredited” by national government. To date, this has not occurred. Why this is so deserves further scrutiny, especially as provincial resistance to losing control over housing resources and subsidy decisions is likely to be a key factor. Local authorities also need more resources and skills if they are to take on the task of housing development within their areas of jurisdiction.

Third, the need to upgrade informal settlements has been recognised and some progress has been made. But there are still some 2 225 informal settlements across the country, roughly 2 000 more than there were in 1994, and the upgrading process remains complex and time-consuming. The need for increased NGO/CBO participation in the delivery process has been recognised, but many of these organisations lack the legal structures required by the Government’s regulatory processes.

A final point worth mentioning is that South Africans have been free since 1994 to elect their own government. If voters are unhappy with housing delivery, they have the freedom periodically to replace their elected representatives with people they believe will better represent their interests.

Mary R Tomlinson currently lives in Australia but has researched and published on South Africa’s housing policy for more than 25 years. This paper was written following a visit to South Africa in March and April 2014, during which a number of interviews were conducted with housing officials.

This article first appeared in @Liberty, an occasional publication of the Institute of Race Relations.