land for low-income housing in south africa

L A N D FOR LOW-INCOME H O U S I N G IN S O U T H AFRICA

Lessons from the Delhi Development Authority and a Proposal for a Land Development Corporation

RICHARD TOMLINSON

This paper 1 considers, and attempts to encourage debate surrounding the potential role of a land development corporation (Landcorp) in addressing the shelter needs of South Africa's urban poor. The position advanced is that there are many potential shortcomings of a Landcorp, a number of which are revealed in the experience of the Delhi Development Authority (DDA), but also that Landcorps are a valuable means of addressing land shortages and that the DDA provides a lesson on how South Africa might itself attempt to structure Landcorps.

The paper has four parts.

�9 The first describes the economic and demographic context of urban policy in South Africa. It will be seen that declining per capita economic growth, coupled with the rapid growth of the labour force, is leading to deteriorating incomes and, especially among blacks, often an inability to afford even the most rudimentary services.

�9 The second refers to land and housing conditions, more specifically to housing need, affordability constraints and economic capacity, and urges that the country needs a programme directed more to the provision of sites than to the supply of dwelling units.

�9 The third part summarises the experience of the DDA and of land banking, the latter since it is this concept which underlies the DDA. The literature on the DDA is not always very flattering, but it seems that the DDA is not exceptional in this respect. Land banking does not have a good record.

�9 Last, a Landcorp is proposed which, following on the experience of the DDA, is more a land developer than a land banker.

84 URBAN FORUM

One introductory caveat is that it is disconcerting that the following material should focus on blacks since, ordinarily, one would want to use the term 'black' to include coloureds 2 and Asians. However, as is evident in Table 1, the material circumstances of the different groups differ so markedly that the narrower use of the term is more accurate when discussing urbanisation policy.

Table 1 South Africa: Population Size, Growth Rate, Urbanisation Level, Racial Proportions and Income Distribution

Asian Black Coloured White Total

PopuhtionTotaland Exponent~l Growth Rate 1990: 36 481 000 2010: 54 230 000

1,58 2,59 1,86 1,53 2,32 1,02 2,00 1,22 1,11 1,80

Proportion of Total Population 1990: 2,6 74,8 8,5 14,3 2010: 2,2 77,8 7,6 12,4

Urbanisation Level 1990: 90,6 50,0? 80,6 89,4

Income Group per Month Before Tax in 1985 R 1- 99 0,3 22,5 7,2 0,2 R 100- 199 2,5 16,8 13,4 1,1 R 200- 299 5,4 17,1 10,5 1,1 R 500- 999 34,4 19,6 32,1 13,8 R1000-1999 31,1 3,5 17,9 R2000+ 11,9 0,7 3,1

37,0 43,0

Sources: Calculated from Johnson and Campbell (1982), Simkins (1983) and De Vos (1987)

The Table shows that future urban growth will predominantly comprise the urbanisation of a low-income black population - they are the poorest group in the country, constitute about seventy-five per cent of the population, have the highest population growth rate, and have an urbanisation level which is well below that of the other groups. A number of specific problems follow: most blacks cannot afford formal housing; throughout the country, areas zoned for

LAND FOR LOW-INCOME HOUSING IN SOUTH AFRICA 85

whites, coloureds and Asians have a 'housing shortage' of about 55 000 units, whereas in 'black' areas the 'shortage' is reported to be about 1,8 million units (Weekly Mail 21-28 Oct. 1988); it is blacks who most suffer from inequitable access to land; and so on. Accordingly, it is not out of choice or any desire to speak in racial terms that this paper addresses the shelter needs of the black population.

SOUTH AFRICA: DEMOGRAPHIC A N D ECONOMIC CONTEXT

Inequality, unemployment and poverty underlie the difficulties confronting effective shelter policy.

In respect of inequality, the white share of the total population, about 12,4 per cent in 2010, is in marked contrast to the white share of the national income, which is expected to decline from 61,5 per cent in 1980 to 54 per cent in 2000. These figures point to a level of income inequality which apparently is higher than any other country for which the data exist (Wilson and Ramaphele 1989) and which, understandably, exacerbates political divisions. However, inequality was dropping rapidly in the first half of the 1970s during a period of high economic growth and, indeed, since then black incomes have been growing ten times faster than those of whites (Knight 1986). The result is that the Gini coefficient 3 of income inequality for employed persons in 1980 was 0,57 (Devereux 1983) - better than Brazil and slightly worse than Mexico.

The problem with the above data is that unemployment conditions have been deteriorating. In a context of political strife, foreign debt obligations, and financial and trade sanctions, the country's potential rate of economic growth lies between 0,5 and 2,5 per cent. The result is that employment growth in the formal sector has declined from 2,5 per cent per annum between 1946 and 1975, to 1,4 per cent between 1976 and 1982, and zero between 1982 and 1986 (Spies, nd). In contrast, World Bank estimates (cited by Knight 1986) have it that the black labour force grew by 2,9 per cent per annum between 1970 and 1982 but, due to demographic lags, will grow by 3,3 per cent per annum between 1980 and 2000. The implications for black unemployment are that formal black unemployment is expected to increase dramatically from about 12 per cent in the early 1970s to 55 per cent in the year 2000 and to continue to increase thereafter (Roukens de Lange 1984). The outcome is mass urban poverty, with dire implications for shelter policy.

Would the advent of democracy enable a future South African government to undertake rapid improvements in the allocation of resources? In 1984 Simkins showed that if one assumed about a one per cent per annum real growth rate in per capita incomes between 1976 and 2000 and continued high tax rates, and interpreted the consequences of this slow (but seemingly optimistic) growth for the budgetary capacity of the government, then racial equality in social expenditure in 2000 implies a per capita expenditure about equal to that of coloureds and Asians in 1975 / 76. In other words, blacks will not

86 URBAN FOR UM

see the rise they desire and whites will see a considerable drop.While the example is best viewed as illustrative - actual conditions may differ - it is clear that there can be no simplistic assumptions regarding subsidies. A democratic government can only hope to achieve results at the margin of pervasive welfare problems. Unhappily, it is in this light that shelter policy options should be examined.

THE LAND AND HOUSING PROBLEM

Reports in the press have it that in order to overcome the housing backlog South Africa needs to build a thousand houses a day: this despite the economy's supposed maximum and unrealised capacity of producing seventy thousand 4 houses a year. Here the discrepancy between housing need, economic capacity and effective demand are considered. The need identified is not for a greater supply of houses, but for an increased rate of supply of serviced sites.

The Housing Backlog Calculations of the housing backlog have very limited value. They generally start with an assumption that, in the absence of effective demand, every household, however it is defined, is entitled to a house. This method ignores or disqualifies many of the housing arrangements common in middle-income countries like South Africa - squatter shanties, backyard shacks and families living in rented rooms - and proceeds to estimates of housing requirements which neither the supposed occupants nor the government can afford. De Vos's (1987) estimate of the housing backlog, for example, is based on an uncertain population estimate. He then divides this estimate by an average family size of 5,94 and subtracts the existing housing stock from the result. The backlog so calculated draws considerable attention but is, in fact, not very significant. For illustrative purposes only, the backlog suggested for the country has been reported at 1,8 million units (WeeklyMai121-28 Oct. 1988), and to be increasing at an average 127 000 units per annum for the next twenty

years (Weekly Mail 6-12 Oct. 1989; the latter figures exclude the 'homelands'.)

Affordability Estimates regarding the ability of the poor to afford houses generally assume that they can afford to spend a certain proportion of their income on housing. A common assumption is that a household can allocate a quarter of its income to housing. The following data assume that a household can allocate money to housing once the the household income exceeds the Household Subsistence Level (HSL). 5 Neither assumption is likely to be correct. For example, Mayo (1987, pp.65) has shown that 'low-income households [are] willing to spend higher fractions of income for housing than are higher-income households [and that as] economic development proceeds the share of household budgets allocated to housing increases among households at all income levels'. He therefore holds that this issue is a matter for empirical enquiry.


The following should therefore be viewed as somewhat wanting and is only intended to be illustrative. The earlier references to the number of blacks who could afford houses was based on De Vos (1986). As noted, he first deducted the HSL from household incomes and assumed that the balance was available for housing. The HSL refers to the cost for a household of maintaining, over the short term, 'a defined minimum level of health and decency', and in 1985 the measure for a black household stood at R313 per month, excluding accommodation. Thus it was that in 1985 sixty-seven per cent of the country's black population could not afford a serviced site which nowadays costs about R6 000 (including the land) and that approximately eighty per cent could not afford a minimal formal house costing around R20 000 (including the land). (The calculation is based the amortisation of a 25-year loan at a low 13,5 per cent per annum with equal monthly payments.)

Economic Capacity The supposed capacity constraint of seventy thousand units per annum (in 1986) has already been mentioned. How does this compare with international norms for housing expenditure? The share of investment in residential buildings in South Africa, as a percentage of gross domestic fixed investment, declined from 17,0 per cent to 10,77 per cent between 1960 and 1980, but then increased to 15,1 per cent in 1988. When compared with the norm for less developed and middle-income countries of between 15 and 30 per cent (Renaud 1987), it is apparent that restrictions both on black urbanisation and on the construction of housing have severely contributed to the current housing crisis, and that the economy can be expected to make a much greater contribution in the future. For example, prior to the Black Community Development Amendment Act of 1986, the private sector was discouraged from serving the black market. Now, owing to the 1986 Act which encourages private enterprise in the black land and housing markets, the situation has reversed markedly - in 1984/85 the private sector built 1 503 units for blacks; in 1987/88 the figure was about 45 000 units - but the improvement was largely for relatively high priced housing units costing R35 000 and more and must be seen against the state's withdrawal from producing housing. No doubt the private sector will reach down market

- the Loan Guarantee Fund orchestrated by the Urban Foundation facilitates this process - and, owing to political pressure, it is also likely that the public sector will again construct low-income housing. However, it seems that public and private sector activity in the low-income housing market tend to displace one another rather than be cumulative (Mayo 1987) and, furthermore, it appears that the public sector, even in socialist countries, has allocated housing inequitably (Tomlinson 1990).

Serviced Sites It is clear that the foremost shelter need of the poor is access to sites. To say this, though, is not to assert the need for large-scale developments, outside the city, on green field sites. This:

88 URBAN FORUM

�9 was proposed by the Venter Commission (South Africa 1984) which favoured the New Town experience of England and France;

�9 describes the Mitchell's Plain development outside Cape Town which consists of a large, dormitory suburb for about 280 000 people, about 26 kilometres outside the city; and

�9 parallels the activities of a number of utility companies working in close consultation with black local authorities.

In fact, there are at least three potential sources of sites - new developments, whether formal or informal, urban infill and increasing urban density. With regard to urban infill, Soweto, for example, has 118 300 formal dwelling units and about 80 000 backyard shacks. Infill is proceeding apace. In respect of density, the extent to which urban density increases further reduces the need for greenfield sites. The Johannesburg example is illustrative. Residential density has two components - built density and the density at which the buildings are occupied. 'White' areas of Johannesburg have a low density in both respects with the result that the city has a population density of about 1 245 persons per square kilometre - below that of Los Angeles (Senior 1984). In comparison, densities of Third World cities range from about three or four thousand to nine thousand per square kilometre.

Johannesburg can expect an increased population density, and the existing infrastructure network of the metropolitan area can apparently cope with approximately 1,4 million more residents. 6 As it turns out, urban densities are already increasing in some residential areas where integration is occurring, for example in Joubert Park, where it is not uncommon to find three or four families sharing a three-bedroomed apartment.

The balance of this paper focuses on new developments and the role of a Landcorp in the provision of serviced sites.

THE DELHI DEVELOPMENT AUTHORITY

The context for the DDA is one of a rapidly urbanising India. 7 Delhi's population increased from 2,3 million to 5,7 million between 1961 and 1981, and is expected to grow to 14,3 million by 2001. In 1961 the city's squatter population constituted 8 per cent of the whole. In 1981 it totalled 24,8 per cent of Delhi's households.

The DDA was formed in 1957 and was empowered to acquire land for public use and to act as a land bank, its function being to service, develop and allocate that land and to fund itself by selling leases and using the funds so generated to enable it to acquire and develop additional land. The underlying principle was that the DDA would acquire cheap, agricultural land on the urban periphery and be in a position to prevent speculation on land prices. It was anticipated that this would allow increases in value to accrue to the DDA, that it would enable the city to direct future development through its control of


fringe land and that it would ensure equitable outcomes. For example, it was intended that serviced plots be allocated on a ratio of 50:30:20 to low-, middle- and high-income groups. (The actual population distribution according to income is 76:21:3.)

Most of the land the DDA allocated was to be in the form of a 99-year leasehold. Leaseholders would pay a ground rent equal to 2,5 per cent of the land's value plus a surcharge, with a view to capturing part of the gain in value resulting from the land's development. Some of the developed land, however, could be auctioned in order to regain the full value of development and to be in a position to subsidise the cost of land and housing for the poor.

By all appearances, the DDA was the kind of vehicle necessary to service the land requirements of the poor. The criticisms that follow are chastening and have directed the following proposals.

Land acquisition occurs as a result of a freeze in land value which is imposed when the DDA notifies the owner that the land is needed for public purposes. The landowner is paid the market value for the land plus a 15 per cent (later 30 per cent) consolation fee and is also provided with one free serviced site. The DDA placed such a freeze over 27 520 hectares, but by 1982 it had only acquired 18185 hectares. The point here is that, irrespective of when the land was frozen, the owners are only paid when the land is actually acquired - more than a decade may lapse between these two actions. Since land prices increased fifty to sixty times between 1961 and 1981, since the price paid to owners was the value when the land was first frozen, and since the land was largely allocated to more wealthy urban groups, the effect was a redistribution of wealth from poor farmer to relatively wealthy urbanite.

Another problem was that the decision to lease rather than sell land gave rise to a large bureaucracy which was necessary in order to check on illegal transfers of the land and violations of the lease. This occasioned high overhead costs and apparently also led to corruption, the lesson being that leasehold can only be superior to freehold when there is an effective administration. The government is now considering selling the land.

As far as the land which was to be sold in order to allow for subsidies for the poor was concerned, the DDA in fact sold considerably more land than was necessary for this purpose. It appears that the DDA was to a significant extent motivated by the desire to increase its own revenue. It exploited its monopolistic control of the land market by speculating in land and, through staged releases of land, actually drove the price of land up.Indeed, this is a problem which is intrinsic to land banking since, once the bank has acquired a portfolio of distant and proximate land, it is actually in the bank's interests that the latter land have a high price which enables the bank to acquire more capital and so to obtain more land. In other words, a monopolistic public sector authority is led to speculate with greater market powers than individual, private landowners.

This is not to say, though, that there was no provision for the poor. About 198 000 plots covering 1 692 hectares were provided for the poor but usually in

90 URBAN FOR UM

'far-flung and locationally poor' areas. Much of this almost free land and services was also subject to ineligible uptake and downward raiding. Most of the developed residential land - 52 per cent- was sold to the relatively rich, also occasionally at subsidised prices. In the mean time Delhi's squatter population increased to 24,8 per cent of the city's population. Confronted by land frozen against private sector development and the slow pace of public sector development, about a million squatters were forced into 'unauthorised settlements'. The squatters, however, were not always in a hopeless situation - through gaining the patronage of local politicians they were often able to obtain not only official recognition of their occupation of the land but also services for their settlements.

Land Banking 'Land banking usually refers either to advance acquisition of sites for government use or to large-scale public ownership of undeveloped land for future urban use' (Shoup 1983:146). The reasons for land banking include:

�9 avoiding the premature commitment of too large a proportion of urban land for private use- land for schools, clinics and parks is obviously much cheaper if it is obtained in advance of urban development;

�9 enabling the government to recoup the costs of installing services (in effect, a form of betterment tax);

�9 stockpiling public land which may be selectively released with a view to restraining price increases;

�9 using public ownership of land as a means of determining the location and form of future development; and

�9 obtaining land for urban growth, most often by getting it at its agricultural value in order to facilitate public sector housing efforts.

These are all worthy goals. However, there are a number of problems which intercede between land banking and an effective housing policy.

1. Planning and policy The DDA example demonstrates the obvious point that public acquisition of the land is no solution where subsequent policies and practices are lacking. The 'experience of countries that have nationalised all land demonstrates the essential weakness of public land acquisitions - it is not the acquisition per se that accomplishes useful purposes, but rather, the policies and procedures for land use that follow acquisition' (Kitay 1985:35). It seems that 'there are two essential factors for a successful programme of land banking: close connection with the planning and land-allocating authorities and coordination between the various agencies involved in development plans' (United Nations 1983:42). This is difficult when:


�9 there are bureaucratic delays in preparing a site plan, undertaking the necessary investments in infrastructure and allocating the land;

�9 the land banking institution develops interests contrary to its mandate; and

�9 when, as is often the case, it is difficult to achieve an honest appreciation of deserving locations and the maintenance of secrecy during the process of acquisition.

It is no surprise, therefore, that 'Most forms of expropriation.., directly benefit landowners' (Kitay 1985:223) (although it seems not in the Delhi instance). Unless public acquisition of land is associated with a commitment to providing serviced land for the poor, little purpose is served by public intervention or control of the land market.

2. Cost of land The cost of the land is critical for any public agency undertaking land banking and, in order to avoid capital outlays which, by definition, will see a return only years hence, it is necessary to obtain land as cheaply as possible. The ability to obtain land cheaply is a reflection of the relative balance of power between landowners and other interest groups. For instance, it was reported in 1983 (United Nations 1983:141) that the Burmese authorities were empowered to purchase land at prices prevailing in 1948, and that Swedish authorities could pay compensation on prices prevailing ten years earlier!

The demand for land will be close to sites of employment, but this is where land will be especially expensive. The agency might therefore be forced to acquire land further out and to hold on to it in anticipation of urban growth. But the longer the agency holds on to the land, the more expensive it becomes; that is, while the price originally paid for the land may seem small, in reality the same money, wisely invested, might be worth more than the land would cost were it to be acquired at a later date. Moreover, ironically, "Inasmuch as large- scale acquisition of public land increases demand and restricts supply, land prices tend to be forced up, hurting the urban poor in particular' (United Nations 1983:42). Indeed, as was the case with the DDA, once it had acquired a portfolio of both distant and proximate land, it was in its interests that the latter land have a high price, thereby enabling it to acquire more capital and so to obtain more land.

3. Squatter invasions It is a world-wide phenomenon that public land has been the primary target of squatter invasions, and public authorities may be unwilling or unable to regain control of their land. Where the government fails to recapture some or all of the costs of acquiring the land, then the ability to sustain a programme of land banking is obviously impaired. Furthermore, unplanned squatter settlement on public land will considerably increase the cost of any subsequent attempts to service the settlement.

92 URBAN FORUM

It is somewhat disappointing, therefore, to have to refer to the recommend- ations of a recent United Nations panel of experts (Rodwin and Sanyal 1987:21):

A number of difficulties.., make our group reluctant to recommend this policy except when the circumstances are clearly exceptional. To be successful, land banking requires administrative capabilities which many municipal governments in the [Third World] do not have . . . Land purchases also presuppose ample funds at low interest rates. Just the opposite is now the case. There are also the uncertainties of advance planning, land management, and price setting, not to mention the increased risk of squatter invasion which can be resisted only at some political cost.

CONCLUSION: A LAND DEVELOPMENT CORPORATION

On the one hand, there are evidently tremendous problems associated with land banking. On the other hand, it is the view of the United Nations (1983:35), for example, that there is a 'fundamental n e e d . . , for Governments to obtain a sufficient degree of command of the land-supply market to ensure a continuous throughput of well-located land accessible to the poor'. The following proposal attempts to reconcile these views and to propose a Landcorp which is constituted in such a manner that it avoids many of the problems described in respect of the DDA.

The Landcorp suggested would operate as the development arm of city governments and be subject to the direction of those governments. The reason for this institutional location is that higher levels of government, throughout the world, have a poor reputation for serving the needs of the urban poor. Thus one should strive to locate institutional control over the Landcorp in a context where the poor matter. In addition, of course, local governments have a predisposition against accommodating the poor - they do not offer much potential for rates and taxes and, instead, they appear to represent a drain on resources. But, on the other hand, it is a common thesis that the poor are more likely to gain influence in local rather than in regional or national government (McCarthy and Smit 1984). This may occur through urban social movements, which are well developed in South Africa, or through a democratic struggle where the votes of the poor are significant to the outcome. Thus, as decision making in South Africa becomes more democratic and city governments become unified, public intervention in the supply of serviced sites is best located at this level of government.

This begins to look as if all the eggs are being put in one basket. What if city governments remain unresponsive to the poor? One is reminded of the position of blacks in Alabama during the period when Kennedy initiated many human rights reforms. Had it not been for deliberate federal intrusion into

LAND FOR LO W-INCOME HOUSING IN SOUTH AFRICA 93

Alabama's affairs, many reforms would have been lost. The lesson is that, despite hopes that they would represent the poor among their constituency, city governments should not be left unattended.

Probably the best means of influencing the form of development undertaken by city government is found in control of the finances. The Landcorp proposed would therefore not be self-financing - it should have no interest in holding on to land (land banking) or in speculating in the land market. It would acquire land for the purpose of specific development projects and, due to the debt incurred when acquiring the land, would turn the land over as rapidly as possible. The Landcorp would obtain the necessary capital from a central development bank and not commercial banks, since the majority of projects would not be commercially viable, and the income and debts arising from projects would accrue to and/or be the responsibility of the local authority.

A central development bank would also help to avoid projects which are targeted in favour of people who can afford private sector housing. This means that the projects have to pass certain viability tests, including that of public participation. (The World Bank has found that the participation of project recipients in decision making is especially important in housing projects, for otherwise cost recovery is seriously impaired - Baum and Tolbert 1985).

What this means for city governments is that, in the hope of democratic outcomes they would be left to set building standards and to spend the money they raise in a politically expedient way; but the additional capital they need in order to undertake large shelter and infrastructure projects should be obtained through an institution such as a public development bank. The location of Landcorps within the ambit of a local authority obviates an additional problem found with the DDA, namely that the Landcorp acts as an independent and self-interested entity.

Under these circumstances, also, city governments will have an incentive to speed up administrative procedures and reduce standards - the former in order to minimise capital turn-around and the latter in order to lower the cost of serviced sites. This is, anyway, a sine qua non for an effective housing policy. If, as is to be hoped, the same reforms are applied to the private sector, then the private sector will itself be able to reach further down-market.

The actual product the Landcorps should provide is subsidised, serviced on many sites - large and small - within the city. This location is in contrast to large, greenfield sites outside the city which, if developed for the poor, will reinforce the form of the 'apartheid city' - indeed, access to finance should be contingent on equitable locations.

The difficulty this proposal creates is that a subsidy rendered to the poor in this fashion may not reflect their priorities. Surveys reveal that as the household income declines, so the relative focus on, for example, cheap food, free schooling and health services increases. Thus, if sites are provided at below market rates, and if the site allocation is to the poor, the sites would rapidly be reallocated through a process of downward raiding. The only way that the aid

94 URBAN FORUM

which the government offers the really poor will stay in their hands is to supply them with conditions which will not prompt downward raiding. It follows that probably the best way of reaching the poor is to deliver facilities which, on the whole, are unattractive to higher income groups and will not prompt downward raiding. This may seem like a cynical justification of doing very little for the poor but, in fact, the reverse is true. While one provides minimal help per household, this costs less and the government is able to reach more households. What it leads to is an argument for a subsidy for the supply of land with a low level of services.

The city government vehicle which could be used to enhance, not monopolise, the supply of land, would be a Landcorp which would be set up with a view to attempting an adequate and timeous supply of reasonably-priced land. It would have the right to acquire, develop and sell land, and through cheap capital and a low level of services be able to serve the needs of the poor.

No legislation, however, has any chance of success if its beneficiaries are solely the poor - a wider constituency has to be sought if the Landcorp is to get off the ground. In this regard, the majority of the black population face financial problems, and the Landcorp should be presented as a state-directed attempt to increase the supply of land and to restrain prices. This is exactly what it should do. The Landcorp might also be articulated as a government attempt to liaise with the private sector, and co-ordinated development projects may well be worth considering. An obvious example of this would occur when the Landcorp provides the land and the private sector develops part of that land for housing.

At no stage, however, should the Landcorp attempt to monopolise the land market. Instead, the private sector, including co-operative groups and employer concerns, should be urged to participate in the housing market. The greater the number of actors, and the more diverse their goals, the more likely it is that people's needs will be served. Any attempt to deny profits to the private sector will ensure that the rate of housing construction is diminished. Self-initiated activity in the middle and upper income housing market will serve a number of needs- the supply of housing, employment creation, a market for backward- linked industries - and set in place a filtering process where the existing housing stock is made available to the less wealthy.

NOTES

1. This paper is based on the author's recent book and utilises case study material excluded from the book (Tomlinson 1990).

2. The term "coloured' refers to people of mixed race. It is widely rejected and the common designation 'black' is frequently employed. The reasons for employing the expression are contained in the text.

3. The Gini coefficient takes values ranging from 0 to 1. A perfectly equal income distribution has a value of 0, conversely the closer to 1 the more unequal the income distribution.


4. The figure excludes house-building capacity in the 'independent' homelands, but this will not be substantial.

5. The Household Subsistence Level is an estimate of the minimum income needed by a household to survive over the short term. It includes food, fuel, clothing, lighting, washing and cleaning materials, rent and workers' transport. It excludes health care, recreation, other transport, the purchase of household equipment and other essentials.

6. This estimate was provided by Roger Boaden of the Department of Town and Regional Planning, University of the Witwatersrand.

7. The material for this section comes from Gupta (1985), Misra (1986) and Acharya (1987).

BIBLIOGRAPHY

Acharya, B.B. 1987. Policy of land acquisition and development: analysis of Indian experience. Third World Planning Review 9 (2), 99-116.

Baum, W.C. and Tolbert, S.M. 1985. Investing in Development: Lessons of World Bank Experience. Oxford University Press, published for the World Bank.

Devereux, S. 1983. Sou th African Income Distribution, 1900-1980. Saldru Working Paper No. 51, University of Cape Town.

De Vos, T.J. 1987. The Complexity of Housing in South Africa. Paper presented at the Second Techno-Economics Symposium, Pretoria, Council for Scientific and Industrial Research, 28 Jan.

Gupta, D.B. 1985. Urban Housing in India. World Bank Staff Working Paper No. 730, Washinton DC.

Johnson, P.D. and Campbell, P.R. 1982. Detailed Statistics on the Population of South Africa by Race and Urban~Rural Residence: 1950-2010. International Demographic Data Center, United States Bureau of the Census, Washington DC.

Kitay, M.G. 1985. Financing Land Acquisition. In Kitay, M.G. (ed.), Land Acquisition in Developing Countries: Policies and Procedures of the Public Sector. Oelgeschlager, Gunn and Hain, in association with the Lincoln Institute of Land Policy, Cambridge, Mass., pp.87-95.

Knight, J.B. 1986. A Comparative Analysis of South Africa as a Semi- industrialised Country. Paper presented at the Conference: The Southern African Economy After Apartheid. Centre for Southern African Studies, University of York, 29 Sept.-2 Oct.

Lim, G.C. 1985. Korea: Land-acquisition Policies and Procedures of the Korean Land Development Corporation. In Kitay (ed.), Land Acquisition in Developing Countries, Appendix A, pp.131-145.

Mayo, S.K. 1987. Household Preferences and Expenditure. In Rodwin, L. (ed.), Shelter, Settlement, and Development. Boston, Allen and Unwin, pp.60-72.

McCarthy, J.J. and Smit, D.P. 1984. South African City: Theory in Analysis and Planning. Cape Town, Juta.

96 URBAN FOR UM

Misra, B. 1986. Public intervention and urban land management. Habitat International 10 (1/2), 65-71.

Rodwin, L. and Sanyal, B. 1987. Shelter, Settlement and Development. In Rodwin, L.(ed.), Shelter, Settlement, and Development. Boston, Allen and Unwin, pp.3-31.

Renaud, B. 1987. Financing Shelter. In Rodwin (ed.), Shelter, Settlement and Development, pp.179-203.

Roukens de Lange, A. 1984. Demographic Tendencies, Technological Develop- ment and the Future of Apartheid. Paper delivered at the Conference: Economic Development and Racial Domination. University of the Western Cape, 8-10 Oct.

Shoup, D.C. 1983. The Rationale for Government Intervention. In Dunkerley, H.B. (ed.), Urban Land Policy: Issues and Opportunities. Oxford University Press, published for the World Bank, pp.132-152.

Simkins, C.E.W. 1984. Public Expenditure and the Poor: Political and Economic Constraints on Policy Choices up to the year 2000. Paper delivered at the Second Carnegie Inquiry into Poverty and Development in Southern Africa (Paper No. 253). University of Cape Town, 13-19 April.

1983. Four Essays on the Past, Present and Possible Future of the Distribution of the Black Population of South Africa. Southern Africa Labour and Development Research Unit, University of Cape Town.

South Africa. RP20/1984. First Report of the Commission of Inquiry into Township Establishment and Related Matters (Venter Commission). Pretoria: Government Printer.

Spies, P.H. nd. The Impact of Dominant SocietaI Forces on South African Agriculture. Institute for Futures Research, University of Stellenbosch.

Tomlinson, R. 1990. Urbanisation in Post-Apartheid South Africa. London, Unwin Hyman.

United Nations. 1983. Report of the United Nations Seminar of Experts on Land for Housing the Poor, Tallberg and Stockholm, March. Stockholm: Swedish Council for Building Research, pp.5-47.

Wilson, F. and Ramphele, M. 1989. Uprooting Poverty: The South African Challenge. Cape Town, David Philip.

land for low-income housing in south africa

Documents