housing policy, subletting and the urban poor

H O U S I N G POLICY, SUBLETTING A N D THE URBAN POOR

Evidence from Cape Town

VANESSA WATSON

Most formulators of housing policy will state as one of their central objectives the aim of providing adequate and appropriate shelter for the poor. The currently dominant Global Strategy for Shelter to the Year 2000 (GSS) - adopted by the United Nations General Assembly in 1988 - gives priority to the lower-income groups in the major urban areas of the developing world and to widening the range of housing choices available to all households. In South Africa the White Paper on the Reconstruction and Development Programme (RDP) contains similar sound goals for future housing policy: the national housing goal is 'to establish a sustainable housing process which enables all people to secure hous ing with secure tenure within a safe and healthy environment ..." (National Housing Forum 1994).

It is surely relevant, then, to consider how the unhoused poor, the future targets of housing policy, currently meet their shelter needs. This will give some indication of the problems and priorities of the inadequately housed, and allow the broad goals of national housing policies to be operationalised through specific and appropriate objectives~

In the urban areas of South Africa we know that many of those inadequately housed are to be found in the unserviced informal settlements, many are crowded into hostel accommodation, while some are on the pavements or in the night shelters. A very significant number, however, fall into the often neglected category of subletters and sharers: they find shelter with other families in backyard shacks, in garages, or in storerooms, or in part of a landlord's house. They choose to find accommodation in this way because it can offer some important advantages. There are important disadvantages, too, and not many renting families would claim that their form of accommodation met the housing goals of the RDP.

It is uncertain how many people in the urban areas of South Africa qualify as subletters and sharers. A recent survey by the Palmer Development Group (1993) in six formal African urban townships (in Cape Town, Johannesburg, Pretoria, Durban, Springs, and Welkom) estimated that 40 per cent of the surveyed populat ion lived in backyard shacks and a further 15 per cent were tenants within the formal dwellings; thus a possible 55 per cent of the surveyed population

28 URBAN FORUM 5:2, 1994

were renters and sharers. A survey in the African areas 1 of Cape Town (Watson 1994) indicated that in the older formal township of Guguletu, 100 per cent of properties were sublet; this fell to 86 per cent in the informal sett lement of KTC and to 68 per cent in the newer formal areas of Khayelitsha. Surveys in other parts of the developing world indicate equally high, and often higher, levels of subletting (see Edwards 1990; Gilbert 1991, 1993; Malpezzi and Ball 1991).

Despite the important role which subletting is playing in accommodat ing lower-income people, and the growing literature on this topic, its importance has still not been fully acknowledged in policy terms. This is true of the widely accepted GSS, p rompt ing the criticism from Habitat (UNCHS 1991:3) that the failure of most housing policies can be attributed to their failure to 'address the needs and interests of tenants, who form the majority of low-income households in many cities'. This is also true of the housing policy currently emerging in South Africa: while the need of some people to rent rather than to buy is acknowledged, it seems unlikely that the rental needs of the very poor will be addressed.

The aim of this paper is to draw attention to what is clearly a major hous ing sub-market in the urban areas, and to argue for the incorporation of sublett ing into future housing policy deliberations. The information contained in this paper was collected through a relatively small-scale survey 2 carried out in parts of Cape Town in early 1994; as such it is not necessarily valid for Cape Town as a whole, or for other parts of the country. However, significant parallels between the results of this survey and research on subletting elsewhere (see especially Gilbert [1993] on subletting in South America) suggest that the patterns and processes described here are not entirely off the mark.

This paper firstly identifies the issues and problems of subletting and sharing which are relevant to a policy perspective. The second section reviews the extent to which sublett ing has been incorporated into policy debates, and the third section puts forward a position on the role of subletting within a national housing policy.

SUBLETTING AND SHARING IN THE CONTEXT OF CAPE TOWN

Defining Subletting and Sharing

The focus, in this paper, is on those individuals or households which occupy dwell ing space which is under the 'control' of other private individuals. Within this broad definition, the nature of the occupancy and the rights and obligations which accompany it can take on a number of forms.

The nature of the relationship between the two parties is sometimes defined in legal terms, and the distinction is made between 'formal rental', where an individual or household rents an officially approved second dwell ing from a resident landlord, and ' informal rental', where the sublet space is a backyard shack or a room within a house. This may be technically illegal if the shack

HOUSING POLICY, SUBLETTING AND THE URBAN POOR 29

violates building regulations, if it involves the subletting of public rental housing, or if the sublet unit is unacceptably overcrowded. Sometimes, however, the tenure of the landlord may not be strictly 'approved ' either, as would be the case in some informal settlements.

The nature of the relationship may be defined in physical terms between subletters who occupy part of the landlord's dwelling and those who occupy relatively independent space (in the form of a backyard shack or outbuilding) on the property controlled by the landlord. While seemingly not a major issue in this country, in other parts of the world it has become meaningful to make the distinction between resident landlords and absentee landlords. Different physical forms of subletting give rise to different housing problems and policy issues.

The nature of the relationship is also sometimes defined in economic terms, and the distinction can be made between people who pay regular monthly sums of cash to a private landlord and those who do not. The latter category of people, frequently termed 'sharers' in the literature, may or may not be part of the extended family of the landlord. Significantly, however, sharers may still 'pay' for their accommodation in forms other than cash. A recent study of domestic relationships in an informal settlement in Cape Town (Ross 1993) showed that payment for access to a domestic unit frequently took forms other than cash. Thus services

- such as fuel collection, domestic labour, and social relationships - were often exchanged for other valued 'things' - such as food and shelter. It is important , therefore, not to view the sublet t ing relat ionship as a s imple commercia l transaction. It is highly interlinked with other forms of social and economic relationship which occur amongst low-income people, and needs to be viewed within the context of broader survival strategies.

In this paper, for convenience, the term 'subletting' will be used to refer to subletting and sharing in all its legal, physical, and socio-economic forms.

The Housing Context of Cape Town

The low-income housing profile of Cape Town differs in some respects from that in the other major urban centres of South Africa: Cape Town's status until the late 1980s as part of the 'Coloured Labour Preference Area' of the country, and its geographical location peripheral to the main 'homeland ' areas, have been important factors under lying this difference.

State provision of formal housing for African people was highly constrained in Cape Town (as it was in other urban centres) under the apartheid regime, and took the form of rental units in formal townships and in hostels. As population growth in the other larger centres of the country far outstripped available housing, extensive informal settlements grew up across nearby homeland boundaries. In the Western Cape, however, this was not a possibility: the nearest homeland boundary is some eight hundred kilometres away. The shortage of accommodation in Cape Town resulted in some informal settlement on vacant land within the


city, but many people preferred the less visible option of doubling up in existing accommodation and in backyard shacks. Historically, therefore, levels of subletting in Cape Town's African townships have probably been high.

The impact of more recent housing policy has been to focus delivery on the provision of serviced and partially serviced sites, under freehold, and usually on land on or near the urban periphery (in Khayelitsha in particular). The less desirable nature of this land in locational terms, and the lower levels of servicing and public facility provision, probably contribute to the lower levels of subletting in such areas.

By contrast, relatively large quantities of public funds were channelled into the provision of formal housing units for the numerically dominant 'coloured' people of Cape Town, prior to the mid-1980s. Coloured squatting, which had mushroomed dur ing the years of large-scale Group Areas removals, was almost eradicated. In recent years rates of construction of public housing for poorer coloured people have dropped dramatically, and there has been a greater focus on self-help and privately initiated housing developments, entirely for ownership. Most new developments have, in addition, been confined to the urban edge.

It is perhaps not surprising, therefore, that subletting appeared to be at a far lower level in all of the coloured areas surveyed (on average, 15 per cent of properties were found to be sublet, compared to the average of 86 per cent for the three African areas surveyed), but there are indications that sublett ing is increasing rapidly as unemployment intensifies and as population growth again runs ahead of hous ing provision.

An obvious influence on this situation would be the extent to which people in the previously classified coloured and African areas of Cape Town may have been decanting into 'white ' areas in recent years, as this could potentially have an effect on the differential demand for accommodation across the city. While there is no clear picture of the extent to which movement across previous racial boundaries has been taking place in Cape Town, it appears to be relatively limited at the present time. For a number of possible reasons not to be debated here, the poorer populat ion of Cape Town is currently continuing to accumulate in the historically lower- income residential areas rather than disperse to the less popula ted parts of the city.

The Nature of Subletting in Cape Town

Housing policy in South Africa is still in the process of being formulated and debated, but draft policy documents emanating from groupings such as the National Housing Forum give a good indication of the form this policy may take. While there is acceptance in principle that provision should be made for a range of tenure options and forms of accommodation, there is no doubt that the central element of the policy, in the urban areas at least, will be end-user capital subsidies for owner-occupiers of individual sites. The central assumption here is that the majority of, if not all, urban dwellers wish to own their own


home. It is this assumption which an understanding of the dynamics of subletting starts to question.

The Cape Town survey shows that while subletting has undoubtedly been fuelled by an overall lack of available and affordable accommodation, it none the less offers for many people a more convenient and flexible form of shelter than they might find elsewhere in the city. The fact that in the Cape Town survey 72 per cent of subletters in coloured areas and 84 per cent in African areas manifested a positive attitude towards their accommodation may be taken as support for this statement. Some of the relative advantages of subletting (under conditions of overall housing constraint) became clear during the course of the survey. These are discussed below.

Costs of Accommodation

The relatively low cost of subletting is clearly a major benefit. Significantly, subletting in Cape Town does not appear to take the form of large-scale, profit- seeking landlordism, as it does for example in urban centres in Kenya and Nigeria (see Amis 1984; UNCHS 1987). A relatively low percentage of landlords in the survey (24 per cent in the coloured areas and 31 per cent in the African areas) claimed that they sublet their properties in order to generate income. Rather, the form of subletting in Cape Town is far closer in form to that found in many South American cities, where landlords are no longer major property owners, but are more likely to be "humble citizens' renting out a portion of their property (Gilbert 1993; Edwards 1982). In the Cape Town survey, in cases where rental is paid in cash, the amount of the rent is broadly in line with the housing costs borne by the landlord: in the African areas rentals average R30 per month (ranging from R44 in the better located Guguletu to R25 in the unserviced KTC); in the coloured areas rents average R252 per month, reflecting the better level of housing and services in these areas and the higher quality of additional sublet units. In all areas the rental would usually also cover water, electricity, and other service charges.

In a high proportion of properties surveyed, however, subletters stated that they paid no regular cash contribution to the landlord at all: subletters on 62 per cent of properties in coloured areas and on 44 per cent of properties in African areas are in fact sharers in that they do not pay for their accommodation in cash. In some cases sharers are part of the extended family of the landlord, but this is not always the case. The point has been made above, however, that payment may take a number of forms, and may often be in kind or in the form of personal services rather than in monetary form.

Flexibility of Payments

A major advantage of subletting is that payment for accommodation can be negotiated with the landlord, and in the case of financial crisis can be deferred


or even substituted with payment in kind for a while. This overcomes a major drawback associated with formal ownership or rental: the need to make regular monthly payments when, for many poorer people, employment and income is highly irregular.

That payments may often be negotiated between landlord and subletters is supported by the evidence from the survey on the rate of evictions and the length of time for which subletters have stayed in one place.

In all the areas surveyed a surprisingly small proportion of subletters i n t e rv i ewed repor ted that they had ever been evicted from their past accommodation: 3 per cent of respondents in the coloured areas and 26 per cent in the African areas admitted to past evictions. This high level of stability is also reflected in the length of stay of subletters in their present accommodation. In the coloured areas surveyed 45 per cent of respondents, and in the African areas 31 per cent of respondents, had sublet on the same property for more than five years.

Long-term Commitment

For many people an advantage of subletting appears to be that it does not require long-term legal or financial commitment. Home-ownership, by contrast, usually requires a significant financial outlay, primarily for house construction (since the current capital subsidy scheme covers little more than the cost of land and services) and also for registration costs. Since there is currently almost no housing market in the African areas, and a very slow market in the coloured areas, it is also very difficult to realise an investment in housing or to dispose of a property at all should this become necessary.

In the African areas there are further processes at work which make subletting a more attractive tenure option for many people: it is now well documented that a significant proportion of the African population of Cape Town (and other South African urban areas) maintain a primary home base in the former homelands as well as a base (or bases) in the urban areas (see Dewar and Watson 1991; Mabin 1990). For these spatially divided households, the rural base plays an important "safety net' function (for times of illness, unemployment, retirement, rest, and accommodation of younger children) and is usually the focus of household investment.

For example, one mother living in a backyard shack in the formal township of Nyanga East told an interviewer:

I do not think I shall live in Cape Town for the rest of my life. As soon as my husband retires I think we shall pack our belongings and go back to the Transkei. We have a field to cultivate and my husband says he is going to save some money to add to the cattle back homeP

The implication of this kind of evidence is that some (and possibly many)


African families have no desire to make the commitment to home-ownership or housing upgrade in the urban areas since they do not view their stay there as permanent . Subletting is an appropriately cheap and flexible form of accommodation in the urban areas.

Access to a Good Location

Subletting is one way in which lower-income people can gain access to accommodation in better located parts of the city, and there are indications from the survey that the better located areas are generally more desirable from a subletting point of view.

Firstly, the better located areas in the survey have the highest proportion of subletters who are rent-paying tenants rather than sharers and it may be argued that this is an indication of a higher demand for accommodation and thus the ability of the landlord to demand a cash rental. In inner-city Woodstock 55 per cent of sublet properties in the area surveyed are sublet on a commercial basis; this proportion drops to 50 per cent in Athlone (compared to an average of 48 per cent for all the coloured areas surveyed). Similarly the better located African area in the survey, Guguletu, has 71 per cent of properties generating cash rentals (compared to 66 per cent for the three African areas surveyed).

Secondly, subletting is most intense in the better located areas: 18 per cent of properties surveyed in Woodstock are sublet (compared to the average of 13 per cent for all coloured areas surveyed), and 100 per cent of properties surveyed in Guguletu were sublet (compared to 87 per cent for the three African areas). Guguletu also has the highest number of households per property (an average of 2,4 compared to the average for the three areas of 2,2). In this instance the smaller than average number of households per property in Woodstock (2,3 compared to the higher 2,6 for Mitchell's Plain) is probably due to the very small size of properties in Woodstock.

Problems of Subletting

While there are clearly some important advantages to subletting, this form of accommodation should not be romanticised. There are also major problems and drawbacks in the subletting sector, both for the subletters themselves and for the authorities administering these areas.

Overcrowding In all the areas surveyed there are large numbers of people on sublet properties. In the African areas surveyed, 50 per cent of all sublet properties contain at least ten people, with this proportion ranging from 76 per cent of properties in Guguletu to 29 per cent in Khayelitsha. In the coloured areas su rveyed overcrowding, although not as high, is none the less an issue: roughly 25 per cent of all properties visited hold at least ten people.


There are also high levels of ove rc rowding wi th in l iving quar te rs . Approximately 28 per cent of all sleeping rooms in the coloured areas and 59 per cent in the African areas contained three or more people.

In some areas, moreover, the way in which the house is located on the plot is not conducive to backyard subletting. Where formal units take the form of row housing and additional accommodation has been provided in the backyard, the only access is through the landlord's house, which can become a thoroughfare for backyard residents and their visitors. This can sometimes become a source of conflict between landlords and subletters.

Poor Access to Services and Shelter Backyard accommodation provided for (or by) subletters is most often built from tin or wood. Thus in the coloured areas some 44 per cent of backyard units are built from these materials, and in the African areas this figure reaches 88 per cent (in these areas most units are made of corrugated iron. These materials are frequently inadequate for keeping out winter damp and wet in Cape Town.

Access to facilities such as toilets, taps, and electricity is often inadequate for subletters~ Across all survey areas there were more than ten people per on-site tap on 31 per cent of properties, and more than ten people per toilet on 37 per cent of properties. Physical access to taps and toilets is especially problematic where they are situated inside a formal unit.

Problems Experienced by Local Authorities

Local authorities, when questioned on their attitudes towards subletting, raised a number of problems.

The first is that landlords are intercepting income which should, they felt, be flowing to the local authorities for service provision and maintenance. Local authorities in the African areas claim that one reason for the high discrepancies between their expenditure on services and the income they receive in rents and service payments is due to subletting. A related problem is that central government bases its budgets for financial assistance to local authorities on the &jure population figures for these areas as opposed to the much higher de facto population.

The second problem is that despite many township bulk services having been overdesigned in the past, excessive population numbers caused by subletting have resulted in some services exceeding their capacities. Thus sewage systems are overloaded resulting in breakdowns, blockages and consequent seepages, certain services have had to be intensified (for example, the frequency of waste removal services has had to be increased), and water consumption has exceeded expectations.

Likewise, the shortage of public facilities in these areas has been exacerbated. This has particularly serious implications in the case of school facilities, where the 'platoon' system (classes given in two shifts per day) is frequently the norm.

The third problem raised by local authorities was that backyard shacking is


a fire hazard, as shacks are often built too close together and sometimes placed in such a way as to hinder the access of fire engines.

The fourth problem is that sublet units are sometimes built over service lines, making access to these for maint6nance or repair very difficult.

Finally, the impossibility of policing and controlling subletting and backyard shacking is a major problem in the eyes of the authorities. While regulations do exist which make much of the existing subletting technically illegal, authorities generally do not have the capacity to enforce controls. Significantly, not one of the landlords interviewed in this survey reported that they had ever been subject to official investigation.

P O L I C Y E X P E R I E N C E S O N S U B L E T T I N G

Gilbert (1993) makes the point that rental housing has generally been ignored in policy terms. Where governments have intervened in this sector of the housing market it has generally been with the aim of controlling two aspects perceived as negative: high rentals, and standards.

Rent controls were introduced by many governments, including that of South Africa, in the 1940s, but have since fallen from favour. The two major problems identified with this form of policy intervention have been, firstly, that rent control difficult to enforce where rental is informal, and, secondly, that it creates a disincentive to private investment in rental housing stock.

For example, in Caracas (Venezuela) legislation was introduced in 1947 to give tenants greater security of tenure (Gilbert 1993): tenants could only be evicted under certain specified conditions. Thereafter, in 1960 a Rental Directorate was created to set appropriate rent levels (10 per cent of the value of the bui lding for new rental housing). While the legislation seemingly served to keep formal rentals in Caracas low for some time, it never really had much influence in the informal hous ing areas. Further, it constrained investment in formal rental housing. In the 1980s government policy switched to an emphasis on home-ownership, and rent control was abolished in an attempt to stimulate greater private investment in rental housing.

Similarly, in Mexico City (Gilbert 1993) and Khartoum (Andreasen 1989) observers report that rent control legislation had little effect on informal rental, or, in the case of Khartoum, on rent levels or rental agreements. Overall, it is argued that landlord-tenant relations seem to function better when they are not regulated, and that in fact relations are often more conflictual in the very areas where contracts are common (Gilbert 1993).

Many Third World governments have, at various times, also a t tempted to regulate bui lding standards and levels of occupancy in lower-income hous ing areas (UNCHS 1991). These regulations have generally been difficult to enforce and have frequently been ignored. Shortages of trained state officials have resulted in slow approval and inspection procedures, and people have simply proceeded with bui lding without waiting for official sanction. Building s tandards and


planning regulations have often been imported unquestioningly from First World countries, and their inappropriateness in the local context has reinforced the tendency to disregard them.

Moreover, where the introduction of minimum standards and regulations is enforced, it can sometimes have unintended consequences. For example, in Mexico City in the 1940s the raising of building standards discouraged investment in formal rental housing stock and promoted the growth of informal rental (Gilbert 1993). In Bucaramanga, attempts to increase property taxes on legally constructed additional rooms also promoted a tendency to avoid formal approval processes (Edwards 1982). In Nairobi high construction standards for dwellings in site and service schemes had no effect on the extent of subletting, but simply resulted in rented rooms in the scheme being unaffordable for the poor (UNCHS 1987).

In South African cities the official approach has generally been a control- orientated one, although the extent to which it has been possible to control subletting has varied considerably across the urban areas. For example, in Cape Town the conservative white municipali ty of Parow has strictly prevented backyard shacking in its coloured residential area of Ravensmead (Du Toit, personal communication). But in the neighbouring coloured area of Bishop Lavis a local Member of Parliament recently established a moratorium on the demolition of backyard shacks (Holden, personal communication), and the Strand municipality changed its position from one of strict prevention to one of tolerance as it became apparent that the demolition of backyard shacks merely resulted in further overcrowding in the formal housing (Degenaar, personal communication). In areas previously classified as African under the jurisdiction of the Ikapa Town Council, eviction notices were used until the mid-1980s to prevent backyard shacking, but thereafter the authorities were prevented by a court order from evicting and removing people unless alternative accommodation was available. This made the control of backyard shacking a virtual impossibility.

Not all forms of state intervention into the subletting sector have been control- orientated, however, and there is currently some debate as to the best way in which to facilitate the operation of subletting.

The GSS sets out what it terms "enabling strategies and processes' in the field of shelter provision and improvement (UNCHS 1991). The GSS does not give prominence to the rental sector but acknowledges its importance within the framework of the enabling approach (UNCHS 1991).

In response to this position, the United Nations Centre for Human Settlement (UNCHS 1991:32-33) argues for a more explicit policy response to the rental sector. It generally supports the non-interventionist stance of the GSS, and holds that governments should not be directly involved in the provision of rental housing stock. Small-scale landlordism, it claims, is the best way to reduce supply constraints on the rental market. Thus the most useful form of state intervention is the promotion of self-help, home-ownership, housing, and removing (or at least managing) rent controls and building standards which discriminate against rental housing.


The Centre does concede, however, that in addit ion some specific actions can and should be taken to stimulate rental (UNCHS 1991:32), and it offers some examples of these. Financial assistance can be granted to those who wish to add to their homes for rental purposes. This measure has been in t roduced by the Central Mortgage Bank of Colombia, but it has been found that its impact on the supply of rental accommodation has been limited in quantitative terms. In Indonesia, the government has recently introduced the possibility of providing credit for investment in rental housing and of channelling this to small developers. It has been recognised, however, that this scheme would probably not be attractive to low-income home-owners, who often look at landlordism as a very informal and sporadic activity. Finally, the creation of tenant co-operatives, sometimes using upgraded city-centre rental tenements, has proved successful in Bombay and Mexico City.

Overall, the UNCHS (1991) document takes the posi t ion that the poor themselves are the most appropriate landlords of the future, and that ensur ing an adequate supply of land and finance directly to low-income people is the most effective way to maximise the range of housing choices available.

A Policy Approach to Subletting

Demands of a National Housing Policy Any housing policy must, as a starting point, recognise the value of achieving an adequately housed population: this has a positive effect on labour productivit~ promotes stable family life, promotes conditions for better health and learning, and is therefore an important pre-condition for economic growth. It is therefore ' in society's long term interests to ensure adequate housing for all' (Dewar and Watson 1982:25), and where individuals and the private sector cannot achieve this, then clearly the state needs to step in. Most importantly, it must be recognised that the aim of state assistance should not be simply to supply units, but rather to achieve access to units. There is clearly no point in increasing the supply of hous ing if it is either unaffordable to those who need it, if it demands legal and financial commitments which people cannot meet, or if it is in the wrong place. As the needs and priorities of those needing to be housed varies significantly, it must therefore be the long-term aim of any housing policy to ensure max imum housing choice: in terms of forms of housing, levels of finish, forms of tenure, and so on.

The clear implication of this, for housing tenure, is that the state needs to assist directly in the promotion of housing for ownership and for rental, as well as in providing assistance to co-operative and group housing initiatives. This position differs fundamentally from that of the GSS and UNCHS (and the position of some housing agencies in South Africa, such as the IDT), which wou ld argue for a primary focus on individual home-ownership. The assumption under ly ing this approach, that the demand for other forms of housing and tenure will be met by the private sector and individual efforts, has to be questioned.


The Role of Subletting

It is essential that subletting, as one common form of accommodation of the poor, is viewed within the context of the overall problem of ensuring adequate and affordable housing for all.

The information gleaned from the survey on subletting and sharing in Cape Town, as well as from other studies in South Africa, indicates that while subletting is playing a vital role in providing affordable, flexible, and often well-located accommodation for lower-income people, it cannot be considered an ideal form of housing. Thus the survey pointed to the high degrees of overcrowding, lack of privacy, poor quality shelter, and overload on services and facilities. There is a significant danger that, with the new policy interest worldwide in subletting, it will become to housing policy makers what the informal sector has become to economic policy makers: that is, it will be seen as offering a solution to housing the very poor which can absolve the state from responsibility in this area.

The position on subletting put forward here is as follows. Firstly, some people will always choose to sublet, for reasons of cost or

convenience, and as such subletting must be formally recognised in policy terms and be viewed as one element in a range of housing possibilities. Secondly, therefore, housing policy needs to include measures to facilitate this form of accommodation (as opposed to either ignoring or attempting to impose strict control over it, which have been the main features of past policies). Thirdly, state housing policy needs to address the problems of those who do not wish to sublet by focusing on the creation of maximum choice in the provision of subsidised housing forms. This will not only expand the total volume of accommodation which can be sublet, but will also allow subletting to become a matter of choice rather than of necessity. Fourthly, controls must be realistic and implementable, and aimed only at those impacts which are critical for health and safety.

To these ends, the following section proposes various areas of policy intervention.

Areas of Intervention

Financial Assistance to the Subletting Sector The De Loor Commission (1992:314) made the suggestion that capital subsidies could also be used for the subsidisation of rental housing, and that the income proceeds on the capital amount invested be used to rent accommodation in the formal housing sector. For example, a subsidy recipient could invest the capital subsidy in a special bank account (possibly with a national housing bank) and use a proportion of the monthly interest as a rental supplement. If 80 per cent of monthly interest on a R10 000 subsidy were drawn as a rental supplement, it would yield some R80 a month (assuming 12 per cent interest). This amount would be able to make only a small contribution towards paying for formal rental accommodation, but would be sufficient to cover the rental on an average


backyard shack. The capital subsidy would grow slowly, and could eventually be drawn to subsidise the purchase of a house if the recipient so desired.

State subsidisation of low-income private tenants is not a new idea; in Germany poorer people receive an income-related subsidy for private accommodation. Certainly, the concept meets an important subsidisation criterion of equity: households are not discriminated against because they do not wish to own or cannot meet the legal and financial commitments of ownership. However, the World Bank objection to this kind of proposal is that it would automatically lead to landlords increasing their rentals, and that it would result in the application of possibly several subsidies to one property.

If it is accepted that the pr imary purpose of a subsidy is to increase access to housing rather than simply to supply units, then the last-mentioned objection of the World Bank may be regarded as less relevant. The charge of exploitation is a more serious one, however, and one which may be very difficult to regulate in a Third World context. More pertinently, in the case of the African areas of Cape Town, this research indicates that the ability to pay rent is not a major obstacle to people finding accommodation to sublet. The major problem is that of f inding affordable shelter which is reasonably located.

This may suggest a phased approach: in the short term available national housing funds should be directed at expanding the supply of housing ( through state assistance to ownership, formal rental, and group forms of housing), and in the longer term consideration should be given to achieving equitable assistance across low-income households through the inclusion of subletting households in the subsidy scheme.

As part of the aim of expanding the supply of shelter, the policy of offering low-interest loans to home-owners or co-operatives who wish to add to their homes for rental purposes would be a useful one.

Conflict Resolution

Currently most subletters have no access to affordable legal channels which can deal with conflicts with landlords. One means of addressing this problem is to recognise subletting legally.

The relationship between landlords and tenants should ideally be governed by the common law regarding a lease agreement. This is an agreement whereby one party (the landlord) binds himself or herself to give another party (the tenant) the temporary use and enjoyment of a property, in return for the payment of rent. Legally, a lease agreement can be verbal, and there needs to be agreement only on the property to be let and the obligation to pay rent. The lease can be terminated by mutual agreement or by a material breach of contract by either party.

Proposals are currently being made for a more decentralised small court system. If such courts were able to enforce the common law obligations of landlords and tenants, they would provide an important channel for conflict mediation.


Physical Controls on Backyard Structures

In a context where administrat ive resources are l imited and populat ion g rowth rapid, it is extremely difficult to enforce any regulations which at tempt to control the bui ld ing of addit ional structures. This research has indicated the extent to which Cape Town local authorities have been forced to turn a blind eye to the bui ld ing of backyard structures, even in the formal, better regulated areas of the city. It is essential, therefore, that if any controls are in t roduced they are targeted only at the most critical problems arising from subletting. It is also recognised that the imposition of new regulations in areas which are current ly densely settled, wou ld be close to impossible. It is therefore p r o p o u n d e d that , in newly developing areas, measures should address the two critical issues of fire hazard and access for service maintenance.

Fire is probably the most serious danger in areas which are settled informally: loss of p roper ty and life is a frequent and tragic occurrence in most of these settlements.

In those areas where the form of development is to be site and service or self-help, the provision of a fire-wall (on at least one bounda ry of the proper ty) should be mandatory. This will not only reduce fire hazard, but will also provide at least one e lement of solid structure for a shelter, which may consist o therwise only of iron and wood or plastic.

The second critical issue which requires control is that of access to water and sewage networks which m a y run beneath a property. In new deve lopments a more realistic approach to this problem is not to locate service lines in the midblock. However , where this is not possible local authorities should have the power to require that structures which h inder access be moved.

Land, Infrastructure, and Service Provision

If the nature and scale of sublett ing in lower-income areas is accepted, then this has impor tant implications for the way in which infrastructure and services are provided, and for land subdivision.

Firstly, the provision of infrastructure, communi ty facilities, and maintenance services (such as refuse removal) must be based on more realistic assumpt ions regarding the existing or likely populat ion of an area. Current ly the provision of these services is based on an assumed population of two persons per bedroom (a National Building Regulations guideline); an estimate of five persons per plot is somet imes used as a rule of thumb. However, surveys have indicated that, in fact, occupation tends to be far higher than this; an average of ten persons per p roper ty was recorded in the Guguletu survey.

Secondly, the provision of service lines at the rear of sites, in the midblock, should be avoided, as access to them inevitably becomes difficult. Service lines should be located in the road reserve.

Thirdly, where housing del ivery takes the form of individual serviced sites,


toilets and taps should be positioned so as to be accessible to the rear half of the site: they should be located near the middle of the site rather than on the street front, which is often the case at present. If starter houses or wet cores are provided, then there should be an outside tap and external access to the toilet, to improve the access of backyard shackdwellers to water and sanitation.

Fourthly, if some form of formal or starter housing is being provided, semi- detached housing should be preferred to row housing (so that backyard dwellers have access to the street), or pedestrian access should be provided from the rear of sites via a midblock alley.

Fifthly, a range of plot sizes should be provided so that those people specifically wanting to sublet may have the opportunity of gaining access to larger parcels of land. Experience from Kenya shows that even plots of 200 square metres can accommodate subletting on a very large scale: rooming houses holding eight households are evidently not uncommon in the settlement of Thika, Nairobi (Housing Research and Development Unit 1980).

C O N C L U S I O N

There is evidence to suggest that subletting, as in other Third World cities, is playing a vitally important role in South African cities in accommodating people with tow and unpredictable incomes. While it does not always offer an entirely acceptable form of accommodation when judged against the standards set by South Africa's housing goals, subletting can none the less provide particular advantages for poorer people. Given that acute housing problems will undoubtedly persist in South African cities for the foreseeable future, it can be assumed with some confidence that the important role of subletting is set to continue.

Subletting, therefore, clearly demands recognition in policy terms. However, the danger now appears to exist that subletting, as was the case with the 'informal sector' of the economy, will be seized on by governments as offering a privatised solution to the particularly thorny problem of assistance for the urban poor. This certainly seems to be the approach of some currently dominant positions on housing policy. This paper puts forward a somewhat different policy position. Subletting must be formally recognised within a national housing policy as one of a range of housing options for the urban poor. As such, state intervention is required in order to contain particular negative aspects but primarily in order to facilitate the operation of subletting as a functional and flexible form of accommodation.

N O T E S

. While the categorisation of urban residential areas in racial terms is no longer legally relevant (and has never been morally justifiable) 'coloured' and 'African' areas have none the less been subject to very different forms


of state policy - policy which differs markedly in turn from that reserved for 'white' urban areas; thus all three areas, therefore, manifest very different social, economic, cultural, and physical attributes. For convenience, and until such time as more adequate descriptors become available, areas will be referred to by their previous legal categorisations.

2. The survey which informed this paper was carried out in early 1994 in eight lower-income residential areas of Cape Town. The areas were chosen to give a reasonable spread across the different types of lower-income living environments: five of the areas were previously classified 'coloured' areas, and three were previously classified 'African'. The areas varied in terms of age, location, dominant form of tenure, and level of servicing. The areas were Woodstock, Athlone, Hanover Park, Steenberg, Mitchell 's Plain (coloured), Guguletu, KTC, and Khayelitsha (African). Within each suburb a block of fifty properties was randomly chosen, and all properties in this block were visited. Where suble t t ing occurred, a ques t ionna i re was administered to both landlord and tenants.

3. This interview was conducted as part of a research project on migrat ion patterns of African people in Cape Town. The research project falls under the Urban Problems Research Unit of the University of Cape Town.

R E F E R E N C E S

Amis, P. 1984. Squatters or tenants: The commercialisation of unauthorised housing in Nairobi. World Development 12:87-96.

Andreasen, J. 1989. The poor don' t squat: The case of Thika, Kenya. Environment and Urbanization 1:16-26.

De Loor, J. 1992. Housing in South Africa: Proposals on a Policy and a Strategy. Government Printer: Pretoria.

Dewar, D., Rosmarin, T. and Watson V. 1991. Movement Patterns of the African Population in Cape Town: Some Policy Implications. Urban Problems Research Unit Working Paper No. 44: University of Cape Town.

Dewar, D. and Watson, V. 1982. Public Housing Rentals: Some Issues. Urban Problems Research Unit Working Paper No. 22: University of Cape Town.

Edwards, M. 1982. Cities of Tenants: Renting among the Urban Poor in Latin America. In A. Gilbert, J. Hardoy, and R. Ramirez (eds), Urbanization in Contemporary Latin America. John Wiley: Chichester.

1990. Rental Housing and the Urban Poor: Africa and Latin America Compared . In P. Amis and P. Lloyd (eds), Housing Africa's Urban Poor. International African Institute: Manchester University Press.

Gilbert, A. 1991. Renting and the transition to owner occupation in Latin American cities. Habitat International 15:87-99.

1993. In Search of a Home: Rental and Shared Housing in Latin America. UCL Press: London.


Housing Research and Development Unit. 1980. Case Studies of Site and Service Schemes in Kenya. University of Nairobi.

Kabagambe, D. and Moughtin, C. 1983. Housing the poor: A case study in Nairobi. Third World Planning Review 5(3).

Mabin, A. 1990. Limits of the urban transition models in understanding South African urbanisation. Development Southern Africa 7:311-22.

Malpezzi, S. and G. Ball 1991. Rent Control in Developing Countries. World Bank: Washington, D.C.

National Housing Forum 1994. Inputs for item 3.3 of the draft framework for the White Paper on the RDP.

Palmer Development Group 1993. Backyard Living in Inner City Townships: A Survey of On-Site Access to Water and Waste Services. Draft Report for the Water Research Commission, Cape Town.

Ross, F. 1993. Houses without Doors: Diffusing Domesticity in Die Bos. Master's dissertation, Department of Social Anthropology, University of Cape Town.

UNCHS 1987. Case Study of Sites and Services Schemes in Kenya: Lessons from Dandora and Thika. United Nations Centre for Human Settlements: Nairobi.

1991. Evaluation of Experience with Initiating Enabling Shelter Strategies. United Nations Centre for Human Settlements: Nairobi.

Watson, V. 1994. Subletting: Towards an Appropriate Policy Response. Urban Problems Research Unit Working Paper (forthcoming): University of Cape Town.

housing policy, subletting and the urban poor

Documents