Submission on land purchases by non-South Africans

03 January 2005
Views 215

Free Market Foundation


Submission on the subject of land purchases by non-South Africans


The debate over land purchases by non-South Africans arises because of the frustrations of previously disadvantaged people that their land and housing needs have not been met.  The highly visible purchase of very expensive properties by non-South Africans presents an opportunity for the landless to give vent to their dissatisfaction and attract attention to their plight. This submission does not therefore confine itself to dealing with the issue of foreign buyers – it deals with the land issue in its entirety and makes far-reaching practical proposals for the rapid provision of land ownership to disadvantaged people so as to remove the frustrations they now suffer. If the proposals in this submission are adopted, the land, housing and poverty problems of low-income South Africans will be rapidly alleviated.


Land and Housing Reform – Problems and Proposals






Many black South Africans have suffered dispossession of their land, denial of adequate access to land, inadequate tribal or apartheid forms of title to accessible land, and restrictions on their use of accessible land. These historical disadvantages have severely handicapped them, especially regarding land reform, small business development, small-scale agriculture and housing. Other factors, including unrealistic land survey, conveyancing and deeds registration standards – a hangover from an archaic past – impede land reform, investment in land and economic growth.


The world over “the land question” is characterised by needless controversy, conflict and confusion. Notwithstanding the importance of historical, cultural and deep psychological reasons for intense land-related emotion, governments need to be informed by a clear understanding of basic legal and economic principles.


Fundamentally, land is like any other asset. Policies should encourage investment in land, as they should in job-creation, housing, education capital formation and enterprise – and for the same reasons.


The concern about whether foreign investors drive land prices up unduly in South Africa, thus making it unaffordable for citizens, especially historically disadvantaged black South Africans, is profoundly misplaced. Restricting and discouraging foreign investors will harm intended beneficiaries directly and indirectly in many ways, and contribute almost nothing to their ability to access land affordably.


It has become customary for government to ask for alternatives when there is resistance to proposed policies. The first, yet surprisingly least obvious and often most desirable alternative, is to leave the status quo. If nothing else is done, that would be preferable.


However, in this case there is a great deal else that should be done, not merely as an alternative, but as a matter of the highest order of urgency and importance for the nation, namely, as elaborated below, tenure upgrade for millions of people still living under the yoke of apartheid forms of land title, and the redistribution of huge amounts of superfluous state-owned land to the landless masses.


Regarding South African citizens and foreigners, especially historically disadvantaged people, land, like other assets, should be the subject of unambiguously secure and freely tradable rights, which is the only known way for it generate prosperity for all.


There is no real difference between local and foreign investment: laws and policies that encourage or discourage local investors also do so for foreigners, and vice versa. Fortuitously, the laws of economics and international experience mean there is no win-lose trade-off. There is a harmony of interests between South Africans and foreign investors.


The lack of progress with land reform and empowerment of the masses is frustrating for all pro-transformation South Africans. But the government must guard against allowing this to result in counter-productive palliatives. It should remain focussed on land reforms that will make a real difference in the real world for millions of historically disadvantaged South Africans. To that end it needs far-reaching reforms to pre-transition laws, procedures and institutions .


The gradualist approaches thus far adopted to address the huge need for land and housing have hardly made a dent in the problem. The longing for property and shelter cries out for a swifter and more purposeful approach. Many of the measures recommended in this submission involve the simple suspension or abolition of obstructive laws and regulations. The measures can be implemented at the stroke of a pen and will have a dramatic and virtually immediate effect.


Recommendations (summary)


  1. Property and rights in land
  1. Amend the Constitution to prevent future legislation from placing obstacles in the way of obtaining secure, tradable title.
  2. Grant all South African landholders security of tenure.
  3. Give people the right to deal with or dispose of their land without undue restriction.
  4. Ensure free transferability of traditional land rights subject only to the right of the tribe to confine transfer to its members if desired.
  5. Respect tenant/owner agreements relating to land.


  1. Superfluous state land
  1. Transfer superfluous state land to the homeless.

  1. The challenge of secure title
  1. Allocate new land by a simple low-cost form of registration.
  2. Exempt unsurveyed land from the Land Survey Act for conversions to freehold for existing occupiers.
  3. Make the Deeds Office registration requirement optional.
  4. Scrap the requirement to use a conveyancer for routine property transactions.
  5. Upgrade existing magisterial registries to provide accurate and legally adequate records of land rights.
  6. Convert all “apartheid” forms of title to full ownership.


  1. The challenge of financing
  1. Repeal usury law ceilings.


  1. Removal of restrictions on land use
  1. Scrap prescriptive land-use controls.
  2. Simplify requirements for the establishment of formal and less-formal settlements.
  3. Create Special Housing Zones exempt from building codes, interest-rate ceilings and costly formalities for mortgage or non-mortgage finance.
  4. Reduce minimum housing standards to realistic levels or, preferably, scrap them in favour of common law relating to safety and neighbourhood effects.
  5. Amend town planning schemes and municipal by-laws, replacing zoning laws with nuisance and neighbourhood law.


  1. The National Home Builders Registration Council Act
  1. Scrap theNational Home Builders Registration Council Act.


7               Removal of stamp duties and transfer duties

  1. Abolish imposts on land transfer and mortgage (transfer duties and stamp duties) – especially for low-income communities.


General considerations

Relaxed standards and formalities will probably result in slightly increased uncertainty. There is a trade-off between reduced costs, speed of delivery and more land reform, on the one hand, and administrative imperfections, objections from vested interests, and less orderly development on the other.


The question is whether net benefits exceed net disadvantages.


(Note: The established convention in the discourse on land tenure and reform is to refer to the need for “security of tenure”. To minimise confusion, this paper reflects that convention. However, it should be stressed that “security” of tenure is a misnomer. The essential problem is not that diverse “apartheid” forms of tenure – PTOs, township leases, quitrents, etc – are insecure. On the contrary, apart from apartheid-specific removals (see below), they have been and jurisprudentially are as secure as “freehold”. In practise, it is extremely rare for any form of “apartheid” tenure to be repealed or cancelled. Township leases and tribal allotments are essentially in perpetuity.


Under apartheid, forced removals amounted to relocating blacks to “homelands” and whites away from “consolidation” land. Legally, the process whereby white landowners and black occupiers were dispossessed was similar. There were, for political reasons, huge disparities in the numbers of people affected and the compensation granted.


The notion that “freehold” is secure tends to be exaggerated. Government “expropriates” freehold land with impunity for multifarious purposes ranging from roads to wilderness areas, and rubbish dumps to post offices. The point is that, apart from forced removals under apartheid, “black” land tenure enjoyed and continues to enjoy a degree of security comparable with “white” forms of tenure such as short and long leases, and freehold.


What is needed is not an improvement in security of tenure for existing lawful occupiers, but tradability of rights in land i.e. the right to sell, let, sub-divide, consolidate, mortgage and otherwise deal in or encumber land.


Accordingly, in this paper, and, we submit, in land literature generally, references to “security of tenure” should be understood to mean – unless the context requires otherwise – “tradability of rights”.)





The Free Market Foundation and its associate organisation, the Law Review Project (LRP), have worked on land reform law for 15 years. One of the LRP’s projects, in the Mathanjana district of Mpumalanga, is referred to in this paper for case-study purposes.


The Mathanjana project, known officially as the Bakgatla ba Mocha (Mathanjana) Land Tenure Project, is a comprehensive development project funded by the Department of Land Affairs. The most germane components concern tenure upgrade, land claims and land redistribution.


The salient facts are that it took several years of intensive and costly effort by a number of consultants to reach completion.


The conclusion to be drawn from this experience is that the prevailing laws and institutions governing land reform are, for practical purposes, prohibitively complex and costly. It is inconceivable that land reform on the scale and with the expedition demanded by transition from apartheid will be achieved unless there is bold reform which does not regard pre-democracy practices and procedures as sacrosanct.


Under the project:

  1. Earthspace Development Planners conducted a comprehensive door-to-door survey to establish community priorities and needs.
  2. A mechanism was instituted for the identity and title of all plot-holders to be established, and for disputes to be resolved.
  3. Approximately 1300 properties were surveyed and pegged.
  4. An additional 500 plots were planned and pegged and allocated.
  5. To these ends, numerous tribal, community, local government and other meetings were held.
  6. Throughout the process government officials and consultants frequently travelled to Mathanjana.
  7. Total expenditure amounted to at least R2,000,000-00 which is the equivalent of more than R1,000-00 per plot.
  8. In other words, it took several years (at least five years in all) and over R1,000-00 per property to upgrade tenure and provide plots for the homeless on land already in the possession of the relevant authorities.


Estimates of the number of lawfully-held properties in South Africa scheduled for tenure upgrade (in urban and traditional areas), excluding the number of additional portions of land to be provided for landless households, and excluding informal settlements, vary widely: from 5 to 20 million.


It should be borne in mind that the human capacity – planners, lawyers, surveyors and others – does not allow for simultaneous processing.


The Mathanjana experience suggests that it would take many years, perhaps a generation or more, and from R60bn to R240bn (based on R1,200-00 per plot).


Clearly, land reform on an adequate scale in an acceptable time-frame at an affordable price cannot occur without far-reaching reforms. This paper recommends reforms which could enable the country to achieve otherwise unattainable objectives of “land reform”.


1               Property and Rights in Land


Recommendation 1

Amend the property clause in the Constitution to include a requirement to prevent future legislation from placing any obstacle in the way of obtaining secure, tradable title.


The Constitution’s Bill of Rights property clause protects those who already hold secure property rights from being deprived of those rights. It charges the state to foster conditions that will enable citizens to gain access to land. It recognises that those with legally insecure land tenure are entitled to legally secure tenure or to comparable redress.


What the Constitution does not do is help the historically disadvantaged to gain access to land. Government’s commitment to help is needed, and it is equally important to avoid creating further obstacles in future. When amending the Constitution it may be desirable to extend the property clause to include a requirement to prevent future legislation from placing any obstacle in the way of exercising the right of access to land. This would mean, for instance, that unduly cumbersome and costly land development, registration, use and survey laws would be unconstitutional, as would perpetuation of restrictive forms of “apartheid” title such as quitrent, deeds of grant, PTOs, township leases, et al.


Recommendation 2

Grant all South Africans security of tenure, including the right to hold, use and enjoy, sell, let, and mortgage land and rights in land.


It should hardly be necessary, in the light of the world’s experience, to present the case for security of rights, and particularly land rights. Ordinary people need the right to hold, use and enjoy, sell, let and mortgage land and rights in land. Secure land rights promote optimal land use, provide access to finance, and encourage investment. People who feel secure and can trade their land rights freely, are more likely to look after what they hold, to optimise its short-term use, to enhance its long-term value, and to seek the ideal balance between long-term and short-term trade-offs. They will be less inclined, for instance, to cause long-term damage such as soil erosion for short-term gain if this will be detrimental to their ultimate interests, or those of their heirs.


Historically “black” tenure under a so-called “township leases”, “permissions to occupy”, “quitrents” and countless other restrictive rights of occupation under apartheid, are still subject to legal insecurity and, more seriously, prohibition or severe curtailment of tradability. Although this was one of the highest priorities for past-apartheid reform (tenure upgrade legislation was passed during the first session), most black South Africans do still not own the land they occupy lawfully. Instead of unwarranted concerns about foreign investment in land, attention should be turned to the much more substantial question of how to empower blacks South Africans without further delay by the summary conversion of all relics of apartheid tenure to proper tradable ownership.



Whilst tenure upgrade is an important and neglected priority, it should recognise and preserve legitimate traditional and community rights as envisaged in the Communal Land Rights Act, and by way of appropriate Conditions of Title.


Considerations (regarding Recommendations 1 to 2)

Greater security and tradability of tenure will meet resistance from a minority of traditional leaders, politicians and officials who do not want to relinquish the power they exercise over disadvantaged people in historically black areas, or who are conservative reactionaries for other reasons.


Recommendation 3

Give people the right to deal with or dispose of their land without undue restriction.


Whites have always been relatively free to deal with their land as they wish. They have been allowed to contract freely between “willing buyers and a willing sellers”, inherit, donate, lease or mortgage. Under apartheid, blacks had no such freedom. Apart from outright prohibition in “white” areas, they could (with rare exceptions) occupy but not deal in land in black “homelands”, “townships”, “locations”, “black spots” and the like.

Occupants of tribal land under “permission to occupy” (PTO) are, to this day, subject to severe restrictions on their capacity to transfer their rights to any other person. These are in the form of legislation, conditions of title, or traditional law. In all situations where the rights are held under forms of title – there are about ten – and not under traditional law, there is no good reason not to repeal all restrictions on tradability. They are an unacceptable legacy of apartheid not yet excised from the new South Africa.


The Communal Land Rights Act has introduced limited rights for communities to deal in their land. The act, like preceding post-apartheid land reforms, will not “deliver” unless supplemented by substantial reforms of the kind described in this submission to facilitate tenure upgrade and access to additional land (under proper tradable tenure).


The right to sell is often curtailed where land is acquired with state assistance. The reason is the fear that people will sell the land and squander the cash. The evidence to date does not support this fear. A glance at any predominantly “black” newspaper or a visit to any “black” area makes it clear that we are far from a world in which black South Africans trade in land rights. Quite the opposite of restrictions on alienation by choice is needed. If anything, black South Africans with newly-acquired land rights should be encouraged to regard land rights as a legitimately-traded part of the economy.


In any case, it is racist and patronising to restrict black people’s right to dispose of their own land. It also reveals an unwarranted lack of confidence in the common sense of ordinary people. Government policy in an emancipated society should assume that individuals are capable of looking after themselves – that they are “rational profit-maximisers” as economists would put it.


Recommendation 4

Promote free transferability of communal land rights, confined to the members of the community if that is the community’s preference.


Communal land rights often appear to be more complex, covering for instance, familial rights and the “land ethic”. But whatever the rights are, including familial rights, there seems to be no good reason why they should not be freely tradable, though confined to members of the community if that is the community’s preference.


Considerations (regarding Recommendations 3 to 4)

The same resistance from a few reactionary community leaders mentioned in respect of security of tenure might arise. The solution may be to require, by regulation or condition of title, that the property cannot be transferred to a person who is not a member of the same community without the approval of the community.


Recommendation 5

Base arrangements such as labour tenancy solely on agreements between owner and tenant.


An effective way for low-income earners, most of whom are black, to gain access to land is through lease agreements such as labour tenancy.


People who cannot afford to buy, or who prefer not to do so, should have the option to rent for cash or kind. If they have no money, the law should permit them to gain access to land by offering their services in exchange. The Land Reform (Labour Tenants) Act, 1996 entitles labour tenants to land owned by farmers, and this discourages farmers from taking on or keeping labour tenants. Instead, the law should uphold whatever agreement a farmer and tenant choose to make.


Holders of rights should be able to use them as they see fit, not confined to personal use or outright sale. More fundamentally, the law should be amended to allow farmers to sell to labour tenants. This would be a more efficient and more moral solution.


Considerations (regarding Recommendation 5)

The Land Reform (Labour Tenants) Act, 1996, was intended to grant benefits to labour tenants by providing them with greater security of tenure and the opportunity to purchase land, and ensuring that they could not be summarily evicted. Although existing labour tenants may have secured benefits under the Act, the concern is that no landowners will enter into new agreements with labour tenants whilst the Act remains in place. Because of the failure to adopt a dynamic labour tenancy policy, many disadvantaged people will remain or become landless.



  1. Superfluous Government Land


Government owns vast tracts of South Africa’s land through the Defence Force, the Departments of Land Affairs, Water Affairs and Forestry, Transport, and Works, former homeland governments, wilderness areas, and provincial and local governments. Much of this land is defined by the Department of Land Affairs as “superfluous”, and is readily disposable.


Recommendation 6

Transfer superfluous state land to the homeless free of charge.


Some of this superfluous state land could be used dramatically and immediately to empower every homeless and landless household with a free plot of land. A “one-household-one-plot” approach could proceed without financing or housing construction, and without costly and time-consuming surveys and deeds registration. Land could be transferred into full and immediate ownership under secure and unambiguous title that can be freely sold, mortgaged or let.


Currently, land is made available only after infrastructure or basic housing can be provided. So most landless people are destined to stay that way indefinitely. A national site-without-service approach (such as the Mayibuye programme in Gauteng) would establish people with undeveloped, securely-held land which they can develop. Site-without-service is a proven approach, as it has always been the basis of traditional land allocation.


A short-term benefit of this approach would be an end to land invasions and sprawling, uncontrolled squatter settlements. Large numbers of plots could be provided at very low cost. The approach can dovetail easily with existing housing subsidies, and there would be no need to wait for housing subsidies to be processed.


There is at least 32 million hectares of state land, the real figure probably being closer to 40 million hectares. Of eight million households with an average family size of five, up to one-third or perhaps three million families are homeless or landless and need land or housing urgently.


This estimate may be too high, as many of the nominally homeless are tenants or communal land residents or live in accommodation supplied by employers. But even on this estimate the state could supply land easily to all landless South Africans, using less than six per cent of its land. One-third of these households might prefer rural land. They could receive one rural hectare each, while the remaining two million households might receive urban plots of 200 square metres each. This would require less than two million hectares. So the government could provide all homeless households with unencumbered title to land and still retain at least 30 million hectares.


Alternatively, rural families could be given up to five hectares each, with one million hectares going to urban households. This would still leave the state with well over 25 million hectares – nearly a quarter of all South African land.


All South Africans would then at least have a place of their own, and be able to build a house, even if the initial structures are rudimentary. This is better than nothing and would launch a dynamic and dramatic process of empowerment and housing improvement.


What is envisaged is not a repeat of homeland “dumping grounds” far from employment opportunities or of poor farming potential. State land is of great variety, and common sense should be used to allocate appropriate land to meet the needs of each target homeless population.


Considerations (regarding Recommendation 6)

Even though most government land is idle, and the economy is presently denied any benefit from it, there is a reluctance on the part of governments to part with government land.


On purely economic grounds and without regard to historical realities and the need for redress, it might be considered more “efficient” to sell superfluous government land to the highest bidder. There is, however, good economic evidence that if the land is distributed as suggested, the distribution will result in efficient outcomes, provided that initial recipients are not prevented from dealing freely with the land. In addition, it could be argued that the symbolic benefit of a massive land redistribution from the state to the historically disadvantaged would be so enormous as to far outweigh any concerns about economic optimality.



  1. The Challenge of Secure Title


The challenge is to transfer government land into private hands with secure title, and to convert inadequate private or communal title to secure title, without costly and unnecessary survey and deeds registration.


The legal requirements of land survey and deeds registration impede rapid large-scale provision of cheap land and housing. Systems of registration and survey historically used for “white” land are far too costly and slow, for too little gain, to use in the short term, if our aim is to facilitate the swift creation of secure title.


Recommendation 7

Allocate new land under a simple low-cost form of registration, defining land boundaries by description.


Frequently, survey and transfer costs exceed the value of the land in rural areas. If there were a market (and some informal markets do exist), an undeveloped plot in a rural area might be worth around R500. A cadastral survey of this plot would cost anything from R200 to R5 000, depending on the number and size of plots surveyed. Conveyancer’s costs to register such a newly subdivided plot in the Deeds Office would most likely also exceed its market value.


Thus, existing “apartheid” titles should be upgraded to full ownership under existing forms of registration and demarcation. New land should be defined and allocated under a simpler form of registration, like vehicle ownership registration, with land boundaries defined inexpensively by description.


People who oppose the suspension of cadastral survey and Deeds Office registration requirements argue that banks need title to be secured by survey and registration before granting mortgages. In fact, the banks need to comply with the law. The law currently requires that they lend only against title secured in this way, but the law can and should be changed. Banks and other lenders will then decide for themselves.


Land survey is not necessary


Recommendation 8

Suspend the Land Survey Act wherever it applies to conversion to freehold and transfer to existing occupiers.


“White” boundaries in South Africa are “defined” by costly land survey. In the 1960s South Africa adopted an even more accurate, and therefore even more costly and inappropriate, system of cadastral land surveying, potentially the most accurate in the world.


The Deeds Registries Act requires a Surveyor-General (SG) diagram to be attached to a deed of transfer. The Land Survey Act requires an SG diagram to comply with cadastral standards for a detailed survey showing X,Y co-ordinates. (An SG diagram’s unique positional (X,Y) coordinates relate to the country’s network of trig beacons and hence to planetary latitude and longitude.)


But there can be secure title with boundaries perhaps less accurately defined by means other than surveys. Title in more mature democracies such as the USA and the UK commonly exists without survey. Deeds of transfer existed in Holland since the 12th century containing descriptions but no diagrams. In the past, our courts have also said that a survey diagram is not absolutely necessary to own land. In the Transvaal, title deeds of farms were issued to burghers with a written description of the boundaries, without a survey. Our judges have said that the boundaries of a piece of land can be fixed monuments or natural objects. There is no magic in a diagram. Clear boundary descriptions in title deeds take precedence over a diagram.


The Land Survey Act states that if a surveyor surveys land for which a title deed has been issued without a diagram, he must ensure that neighbouring landowners agree to the boundaries that he is adopting in the survey. In other words, the factual position accepted by the neighbours is more important than the survey.


The Act also says that an arbitrator deciding a boundary dispute where a diagram exists must be guided by the principle that the original boundary markers (beacons) of a piece of land define the true boundaries even if they do not correspond to the diagram.         


It would therefore not violate legal principles and tradition in this country to once again permit ownership to be acquired before a survey.


Nevertheless, as stated above, the law presently requires a survey, done to the highest standards, before transfer of ownership. This imposes huge costs and delays. It has been estimated that all existing South African land surveyors (if they were available) would need ten to fifteen years to survey only existing unsurveyed properties. Black South Africans cannot wait this long to acquire secure title to their properties, especially when the cause of the delay, the survey of existing boundaries, is of negligible legal value. Furthermore, a cadastral standard survey (R200 - R5 000) of each such property could cost a few billion Rand and such money is clearly not available. For converting existing black plots and allotments to freehold and transferring them to lawful occupiers, therefore, the Act should be suspended. Survey would then be optional for owners and mortgagees, and conversion and transfer could proceed immediately.


Most ownership disputes cannot be blamed on a lack of surveyed, registered title deeds. The disputes have arisen because, for instance, a number of farmers sold land unlawfully to blacks more than once in the same land’s history without transferring, and later acknowledging, real ownership. The proposals in this document will not result in such disputes.


To mass-produce new properties, where survey is appropriate, the government might choose to adopt less accurate forms of survey than cadastral land survey.


Whether or not banks will give mortgages is no reason to withhold security of title from the victims of apartheid. If blacks get unencumbered land, without diagrams, there is no fear that they would be unable to use it as security for borrowing. Where the property is of high value, it would be within the interest and means of the owner or potential purchaser to have it surveyed by choice. If, say, an expensive casino or shopping centre is to be built in a rural area, it would be up to the developer concerned to have the property accurately surveyed and registered.


Considerations (regarding Recommendations 7 to 8)

There is a danger, though it has been somewhat exaggerated, of an increased number of boundary disputes if the procedures relating to land survey are changed. It must be borne in mind, however, that this recommendation does not suggest that existing boundary definitions should be degraded. It merely suggests that high-standard and costly land surveys should not be a precondition for the granting of more secure ownership rights to the millions of people whose rights are presently insecure.


The interests of the people who presently suffer as a consequence of insecure title should take precedence over vested-interest arguments against the introduction of a more rapid and less costly means of identifying properties and their boundaries.


Deeds registration is not necessary

Conveyancing is the branch of law dealing with the transfer of ownership of property, after the document (conveyance) which transfers legal title. It is in the interests of conveyancers and deeds registrars that upgraded apartheid titles should have to be registered in the existing deeds registry system or some slightly amended form of it. Although this faith in deeds registry is without practical substance, three years into post-apartheid South Africa it still denies black South Africans the enjoyment of immediate, secure land title.


Recommendation 9

Make the Deeds Office registration requirement optional.


Formal registration under the existing deeds registry system should not be allowed to delay upgrading titles and normalising the position for millions of South African blacks. It is not possible without intolerable delay and cost to bring all upgradeable titles into the deeds registry system.


There is a great deal of obstructive mythology surrounding the current deeds registry system, characterised by a mind-set that needs to change fundamentally if the land question is to be addressed expeditiously and effectively. Myths include the mistaken view that:

  • the existing deeds registry system is near-perfect – almost sacred – and shouldn’t be tampered with.
  • all land should or must be registered in a one-size-fits-all single registry.
  • the current system provides and guarantees security of title – that registration is proof of title.
  • deeds registration in its present form, or something like it, is a pre-condition for ownership.
  • ownership is a pre-condition for land to be tradable.
  • registration must be accompanied by an S-G diagram.
  • a diagram is necessary for boundaries to be identified accurately or adequately.
  • a diagram provides proof of boundaries.


There are other related myths, but these are the most serious. They account for the fact that so little progress has been made. There is unfortunately a disguised form of racism that informs the esteem in which the existing system is held. It assumes that what whites created and reserved for themselves was ideal – regardless of cost. It ignores systems that worked alongside the “white” system for blacks – under which a greater number of parcels of land have been held for many decades.


This is not the place to address each in detail – which we are happy to do if required. Instead we confine ourselves to core issues that should be addressed as matters of the highest national priority. There is a need for a fundamental mind-set change on these matters.


Deeds registration does not guarantee title in law. It is evidence at of ownership, but that there is provision for the rectification of titles and diagrams is necessary precisely because they do not guarantee title or boundaries respectively. Registration and boundaries will, for instance, be inaccurate in the event of fraud, or bona fide error, or unregistered unsurveyed servitudes arising from long-term usage. Where diagrams create boundaries (new townships, subdivisions, servitudes etc) they define them, and diagrams are legally binding conclusive proof of boundaries. However, where diagrams describe “what is” – where they come after the fact describing pre-existing boundaries, they are of severely limited legal significance. In the first case the surveyor creates the boundary; in the second s/he does his/her best to describe boundaries created by others. This second situation applies to all unsurveyed unregistered land (eg “communal” or “township” land).


For many decades traditional and “township” land has been held under other forms of registration (magistrate’s courts, local governments etc), or none at all, and has not presented significantly more problems than registration of historically “white” land in the deeds registry. Title and boundary disputes have been extremely rare – that they are common is an urban legend – and, when they arise, they are almost always easily expeditiously and cheaply resolved.


The proverbial bottom line is that, for practical purposes, the system must be reformed if there is to be land reform, and if land is to available to black South Africans under secure tradable title in significant quantities at affordable prices.


Recommendation 10

Suspend the requirement to use a conveyancer for routine property transactions involving simple transfer of ownership or the registration or cancellation of a mortgage bond.


There is no reason why title of land should not be established in the same way that other forms of title are. There is also no reason why there should be only one form of registration of title in a country. For land that is already surveyed and registered, the existing “white” system can be maintained, and co-exist with a parallel system or systems. Transfers and mortgages should take place in much the same way as vehicles and shares are transferred or pledged, or as land registers already operate in most areas of the world.


All forms of “black” land title have worked well for centuries without being registered in a deeds office. Customary law provides mechanisms to allot and later transfer land. Title disputes are extremely rare and when they arise they are usually readily resolved by courts or traditional authorities. Land title in black areas is recorded in various ways, from formal sealed contracts to a handshake or even tacit non-verbal agreement like a mere nod to secure a binding legal contract of sale. A voetstoots sale was secured, for instance, by the action, before witnesses, of placing the object being sold on the ground and pushing it by foot across to the buyer. Such visual memories in the minds of witnesses and community elders are a proven centuries-old method of recording title. Title might be secured similarly by ceremonially breaking off a branch to hand it to the new holder, imprinting this visual record indelibly on people’s minds. By the time they die, title is so adequately secured as to need no further evidence of this kind.


The great historical episodes of prosperity in the first world occurred when there was no land registration. In some of the world’s richest countries, land registration is not compulsory to this day, and some of the most valuable properties in Europe’s greatest cities are unregistered.


In the USA, title may be secured, if the owner chooses, by insurance rather than registration. Insurance costs are lower and delays briefer than those incurred by our existing registry system.


Recommendation 11

Upgrade existing magisterial registries to provide accurate records of land rights.


All existing unregistered titles could become the lawful property of the existing lawful holder “at the stroke of a (statutory) pen”. Where registration already exists, it can be used and upgraded. Permission to occupy (PTO) is usually registered in the local magistrate’s office. To provide secure tenure in the short term, it is much more realistic to upgrade existing magisterial registries than to transfer all properties to the deeds office. In the old Transvaal republic, before the deeds registry was established in 1867, it was legal and customary for land transactions to be executed at the office of a notary and then registered by the local landdrost (magistrate).


Land reform must not be held up by requiring conveyancers to process all acquired land at huge cost. Almost everything that conveyancers do has become an anachronism. No routine property transactions involving simple transfer of ownership or registration or cancellation of a mortgage bond should require a conveyancer; costly property (for example, jewellery, boats, cars, planes, oil rigs and machinery) is transferred daily throughout the country without them. Conveyancers should be used only for unusual transactions, for example, where a servitude, usufruct, subdivision, or consolidation is concerned.


Registration in a Deeds Office is costly and cumbersome. It is not needed to establish adequate legal ownership of land, just as it is not needed to establish vehicle or machinery ownership (where the value often far exceeds that of plots of land). Local and international history shows that land titles and land boundaries will be secure under most methods of tribal land allocation and demarcation. The costs and delays of Deeds Office registration should not be allowed to impede the immediate upgrading of title, in traditional community areas and elsewhere.


Considerations (regarding Recommendations 9 to 11)

The conveyancing and deeds registry communities are likely to resist these proposals, which argue that increased security should be extended to historically disadvantaged people without having to wait for the “Rolls Royce model” of property registration. Government will have to choose whose interests should have precedence.


It is simply impossible to survey or register all the land presently held by black South Africans, and land which will be acquired by them, for many years. We really have no choice. Either we permit ownership under systems other than those that exist at present, or the deprived majority of South Africans will continue to be deprived for many decades.


If these proposals are implemented, there would continue to be two systems of land ownership and land registration. This recommendation proposes that the second and less costly system be used to record ownership as well as historical forms of tenure.


Recommendation 12

Convert all forms of apartheid title such as permissions to occupy and all forms of lease to freehold title and ownership by the existing lawful occupant, after suitable local advertising of property-holder lists where these exist.


It should prove simple to convert all existing tenancies (formerly residential permits) in established townships such as Soweto to secure title automatically by operation of law. Initial local advertising of lists of tenants will reveal if there are competing claimants. The tenancies should thereafter be summarily converted to ownership without depending on prior deeds registration.


It should also prove simple to convert all informal titles to rural parcels of land held under some form of individual or non-“communal” arrangement, whether residential or non-residential, into secure title automatically by operation of law, bypassing the costly and cumbersome “perfect system” of both cadastral survey and deeds office registration.


All this summarily converted urban land and non-”communal” rural land held under some form of individual title should automatically become the property of the existing lawful holder. Legal provisions for rectification of title can handle any boundary problems which may emerge later.


Considerations (regarding Recommendation 12)

Existing records in many parts of the country are notoriously inaccurate or derelict. Predictably there will be disputes regarding the rightful beneficiaries. In traditional areas, the rights of relatives will be problematic. Additional complaints may come from people who purchased neighbouring properties at great personal cost and who may feel that they have been unfairly treated.


Problems of this kind explain the current inertia. Government has the unenviable choice of being accused of non-delivery, on the one hand, or of dealing with myriad problems relating to rapid property-rights recognition on the other. Clearly the potential benefits far outweigh the risks.



  1. The Challenge of Financing


Recommendation 13

Repeal usury law ceilings on effective interest rate to enable banks to lend against low-value property to those perceived as high-risk borrowers.


Banks would be more willing to lend, against the security of low-value property such as small plots of land, to those perceived as high-risk borrowers if there were no usury law to limit interest rates and related charges. Moreover, they would not necessarily need to raise interest rates if they were allowed to adjust their administration fees to compensate for risk and transaction size.


Housing finance comes mainly from government subsidies and private financial institutions. Usury laws and other laws relating to financing prevent poor people, who are perceived as high-risk borrowers without either security or a credit record, from borrowing lawfully for housing. Even when they have title which can be mortgaged, formal-sector lenders usually will not provide finance because the permissible return is too low and the perceived risk too high.


A possible solution to this problem would be to establish Special Housing Zones which would be exempted from interest-rate ceilings and costly formalities for mortgage or non-mortgage finance. Low-income people in these areas would then be able to obtain housing finance formally and lawfully. Though the bonds would be at higher interest rates than those available to lower-risk borrowers, they would nevertheless be at more favourable rates, and have less onerous repayment terms, than those offered by informal and other underground lenders to whom the poor must currently go.


A further issue needs to be addressed, and that is the likelihood of payment. Even if banks are satisfied that Deeds Office registration is not needed for security, and they are allowed to charge higher interest rates to high-risk borrowers, they will still reject applications for bonds unless they are reasonably certain that they will be repaid. Lenders need the likelihood of payment even more than security, and it is admittedly not clear whether rural borrowers will pay dependably. As soon as legal barriers are removed, banks can move into historically closed areas and this market issue can be addressed. This may not pose a serious problem for progressive and innovative banking. One possible solution may be for a loan applicant’s stokvel to pledge collective security for repayment.


As far as government financing of land and housing is concerned, it is important not to repeat the error of soft loans to farmers through the Land Bank or some equivalent. Anyone who doubts the unsuitability and counter-productivity of soft loans should read Symond Fiske’s excellent book Damage by Debt (FMF Books, 1995) in which he explains why debt subsidies waste capital and hurt the people who use them.


If the government has a million Rand to help aspirant landholders, it can maximise the benefit of the money, not by lending it to a few borrowers at unjustifiably low rates, but by using it to underwrite numerous loans at market rates of interest. For example, it could enter into a contract with commercial bankers to underwrite, say, a fifth of their loans to historically disadvantaged people.




Considerations (regarding Recommendation 13)

Unsophisticated people may possibly over-extend themselves but this problem has to be seen in the context of overall policies during the transition. Government has tough choices: it can strive for a perfect risk-free world for its citizens, or it can have a dynamic prosperous economy. Unfortunately it cannot have both.



  1. Removal of Restrictions on Land Use


Until the 1960s, as in many other countries, it took about four months from start to finish to develop new residential land into “townships”. The conditions of establishment were seldom onerous; often the only requirement was for graded roads. Residents provided their own boreholes, septic tanks and other infrastructure.


Thereafter, township development procedures became ever more cumbersome and costly, and today only wealthy developers can afford them. Under township development laws the government, through township development boards, requires costly infrastructure including long-life tarred roads, storm-water drains, water-borne sewerage, water and electricity. Formal approval from countless government departments and organisations can take a few years.


If laws for establishing “formal” settlements reverted to the 1960s versions, this would lead, as it did then, to empowerment of lower-income developers and an explosion of property development and rapid delivery of affordable land and housing to large numbers of people.


Acknowledging that matters got out of hand, the Development Facilitation Act was adopted by the apartheid government during its twilight years in an endeavor to “facilitate” development; to make it easier, faster and cheaper to provide enough affordable land for black South Africans. Three serious problems presented themselves as a result of the dirigiste mind-set that would not let go of excessive regulation and central planning: (a) very little use was made of the Act – virtually none for the first few years; (b) it entailed too much regulation and delay to solve the problem. and (c) it created confusion as to how land should be developed, especially for low-income communities.


Recommendation 14

Scrap prescriptive land-use controls.


The right to hold and the right to use tend to be confused. It is possible to render the right to hold property meaningless by curtailing the right to use or enjoy it. If an inner-city office block were rezoned as a wilderness area, for example, the owners’ or tenants’ rights would be rendered valueless. In extreme cases the courts in some countries have considered the curtailment of the right to use and enjoy to be tantamount to confiscation or expropriation. Normally, however, the curtailment of use rights is less obviously confiscatory and holders suffer losses without compensation or the right to due process. As a matter of principle, therefore, land policy should regard any new restriction on usage as a usurpation of property rights for which there should be just compensation.


In practical terms, land rights should include the reasonable right of use and enjoyment. Many existing restrictions, especially conditions of title and zoning, are excessively restrictive and should be reviewed critically.


Considerations (regarding Recommendation 14)

This recommendation is controversial because of South Africa’s tenacious planning mentality, which is a legacy of our authoritarian past. It is very difficult for most South Africans to imagine a voluntary spontaneous order. Their immediate assumption is that the absence of coercive central planning will lead to chaos or anarchy. Experience around the world proves that this is not so, but often perceptions are more important than realities. There will accordingly be considerable resistance from vested planning interests in the private and government sectors to increased land use freedom.


These fears are not entirely unfounded. It is true that a controlled society is often more orderly. But it is also more stifling, and inhibits the cultivation of a national enterprise culture. If this recommendation is implemented, there will be undesirable neighbourhood effects, but again, the question is whether benefits outweigh advantages, and whether alternative methods of nuisance control would be more effective.


Recommendation 15

Drop most approval requirements for the establishment of formal settlements and make costly infrastructure (roads, drains, electricity, etc.) optional.


Formalities and compliance costs when developing property for residential purposes are huge obstacles to housing delivery, as is the need for approval of building plans for housing. Township development laws and building codes were recently relaxed, but the urgent plight of the homeless cries out for a more radical interim solution. As explained above, the problem has not been solved by the Development Facilitation Act.


Recommendation 16

Establish new Special Housing Zones in low-income areas, in which township-development laws and slum laws are suspended. Exempt these zones from building codes (reduce minimum housing standards to realistic levels or, preferably, scrap them in favour of common law relating to safety and neighbourhood effects), interest-rate ceilings and costly formalities for mortgage or non-mortgage finance.


A low-income housing option would be to suspend or relax housing, township development and slum laws within new Special Housing Zones (SHZs). There homeless people would settle on land in which they have security and mortgageable wealth and where they can build immediately, confident that their buildings will not be condemned. This would rapidly improve the present plight of squatter settlements and obviate the need for land invasions. SHZs should be located with regard to neighbourhood effects and other externalities. In due course social and physical infrastructure and housing subsidies would lead to further improvements.


Housing of a low standard is preferable to no housing, and lawful housing is better than unlawful housing. To address the urgent needs of the homeless, with or without Special Housing Zones, low-cost housing needs a blanket exemption from expensive legal requirements.


Recommendation 17

Reduce minimum housing standards to realistic levels.


South Africa’s housing backlog cannot be alleviated for decades at current rates of delivery or within the existing policy framework. Housing delivery is far below government targets. Adjusted housing policy might accelerate housing provision. Housing standards should be drastically reduced, then raised by increments later if necessary. Only after most people are housed does it make sense to impose high minimum standards.


For example, fuses and innovative earthing systems could be permitted in place of circuit breakers and earth leakage protection. The latter provide a desirable safety net to cover other electrical shortcomings, accidents and abuse, but are hardly used elsewhere in Africa.


Recommendation 18

Amend town planning schemes and municipal by-laws, replace zoning laws with nuisance and neighbourhood law (common law). Enable the courts, especially small claims courts, to enforce these and to issue interdicts.


Zoning laws should be relaxed or replaced with “neighbourhood laws”. Instead of trying to dictate to people how they should use their property, the state should let the market decide, while protecting the rights of neighbours against land uses that cause a nuisance. Nuisance and neighbourhood law is common law that has evolved over the centuries as a perfectly adequate method of protecting people against socially harmful land use.


What is lacking is not law, but access to redress. The courts, especially the small claims courts, should be able to enforce neighbourhood and nuisance law, and to issue interdicts. This would require an amendment to town planning schemes and zoning laws as well as extension of the powers of the small claims courts.


Around the world the poor typically live on the most valuable inner-city land and drive the rich out to the periphery of cities. They out-compete the rich by living either in apartment blocks or in high-density housing developments. In South Africa, however, Group Areas Act zoning has created an anomalous situation in which the poor live in peripheral “townships” while the rich live in suburbs close to city centres, such as Houghton in Johannesburg.


Zoning laws place a density restriction on land use; their effect is to make valuable land available at artificially low prices to high-income people. The laws and practices should be liberalised to allow market-driven land use and development. This will densify land use close to the CBD, making more land available for industrial and commercial purposes and (in optimal localities) for lower-income people and emerging entrepreneurs. These suggested reforms should apply to all existing zoning laws, and not just to the Development Facilitation Act and other zoning laws made after 1994.


As with all reforms, there will predictably be enormous resistance from those who benefit from the status quo. They will raise all the stale arguments about “regressive” reforms “lowering standards”. But the economic consequence of pandering to the elite is an inefficient use of land and other resources. Inefficiencies are by now absolutely pervasive, and the economic, environmental and human impact is hard to quantify. The fact that most people are forced to live far from places of employment and from commercial and entertainment centres not only increases travelling costs, but also unduly burdens infrastructure. An enormous amount of potentially productive time is wasted and millions of Rands have been invested in transport and other facilities which otherwise would not be necessary.


In many lower, and to a lesser extent, middle-income areas there are either no zoning provisions or the law is ignored. Areas in which strict zoning cannot be enforced in practice should be formally exempted so that the law coincides with reality and common sense.


Land use should be neighbourhood-focused and government’s concern should shift from grand town-planning schemes to what communities want for themselves. In particular, zoning laws that are hostile to business should be amended to promote an enterprise culture and international competitiveness. Presently the law restricts the use of urban land for business purposes, rather than preventing the disturbance of neighbours by the manner in which the land is used.


If someone makes furniture for themselves in their garage, it is lawful under zoning laws. However, if the same person makes the same furniture, but for sale, it would be unlawful. There is clearly no logic in this position.


Considerations (regarding Recommendations 15 to 18)

Relaxed zoning laws are not problem-free. Yet methods can be found of allowing people to make optimum use of their land without being a nuisance to their neighbours.






  1. The National Home Builders Registration Council Act


Recommendation 19

Scrap the National Home Builders Registration Council Act or exempt all small-scale builders.


Current laws impose entry barriers into the building industry and minimum standards for building materials and labour, thus raising costs and curtailing the supply of low-cost housing. The National Home Builders Registration Council Act imposes further costs and entry barriers. It creates an unnecessary, undesirable and potentially harmful bureaucracy.


The National Home Builders Registration Council (NHBRC) is a Section 21 non-profit company which seeks to provide warranties on the building of new homes. Funds for the operation of the Council are obtained from home builders who pay a registration fee, an annual subscription, and a levy on all homes built.


Notwithstanding their opposition to the NHBRC scheme, home builders find themselves in a trap. The members of the Association of Mortgage Lenders refuse to provide mortgage finance to purchasers of newly built homes which sell at between R25 000 and R250 000 unless the homes have been built by home builders who are registered with the NHBRC and the individual homes have been “enrolled” with the NHBRC. In effect, this means that mortgage finance is not available at all for homes in that price range built by builders not registered with the NHBRC, as all the large providers of mortgage finance on homes are members of the Association of Mortgage Lenders.


Some of the criticisms of the scheme from members of the building industry are:


  1. The warranty which forms the basis of the scheme, and without which it would have no function, is an unenforceable contract. The home buyer is required to sign an undertaking that he/she will have no legal claim against the NHBRC: the NHBRC says only that it will “endeavour” to support the consumer in the event that the developer fails to comply with NHBRC requirements.
  2. Land value and sectional title infrastructure are included in the amount upon which the levy is calculated, although the scheme’s purported warranty applies only to the building.
  3. The buyers of homes in the lower income bracket, who have the greatest need for help in assuring quality, are excluded from the scheme, whilst home buyers who are capable of protecting themselves against poor workmanship are included.
  4. The levy adds R3 250 to the cost of a R250 000 home.
  5. Much better and more flexible warranty schemes would probably be offered by the country’s highly competitive insurance companies.
  6. Homes are already subject to inspection by local authorities and mortgage lenders. A third inspection agency is unlikely to detect additional examples of defective workmanship that escaped the notice of the existing agencies.


Despite all the controversy surrounding the NHBRC, the National Home Builders Registration Council Act gives statutory authority to this NGO. The Council not only has the coercive support of the Association of Mortgage Lenders but is also be in a position to apply for interdicts against home builders and levy fines for non-compliance.


The NHBRC Act does not improve building quality. There will always be building failures, because of inherent, irreducible risk and where these occur the Act tries to apportion liability for failure among owner, developer, builder and NHBRC, but with little clarity or certainty. The NHBRC itself has discretionary liability, and if it rejects liability in a particular case, the homeowner has no right of claim against it, even if the home was enrolled and built by a registered builder, and even if there was a clear breach of the standard warranty, and even if the builder is no longer in business or is unable to meet his obligations. Builders and developers have to cover this risk by insurance or other means, and are then subjected to the additional cost of NHBRC levies and fees.


The Act states that all contracts for building homes must be in writing, with plans, specifications, price and payment terms. This section is presumably intended to protect home-buyers, but will have the consequence of disempowering illiterate and monolingual builders. It also empowers the NHBRC to prescribe a so-called “standard” warranty, which the NHBRC may nevertheless amend at any time. The Act states that the NHBRC may, as part of this warranty, impose any obligation on a home builder. This NHBRC warranty can require a home builder to rectify, at his cost, any major structural defect in the home during any period, however long, that the NHBRC might fix. These ambitious attempts to obtain perfection in home-building will, in practice, be passed on by builders to purchasers in the form of higher prices.


Together with “registration” and “warranty”, there is a third line of bureaucratic control, namely “enrolment” of buildings. Mortgage lenders “shall” not lend to owners unless builders are registered, and conveyancers “shall” ensure that builders are registered. This is a deliberate attempt to reduce the possibility of competition from new entrants into the home loan market who may, in the absence of the proposed NHBRC bureaucracy, cut the cost of lending, establish their own method of ensuring the soundness of their security, and out-compete the other lenders.


Inspectors cover the whole country including the loneliest rural areas. A small builder in a village, who is unaware of the existence of the NHBRC, and builds a house which falls within its ambit, is guilty of an offence and liable to a fine or imprisonment. If the government wishes to promote small business, it should scrap laws that have these implications. Honest, hard-working people should not have to worry constantly that they may be committing an offence that carries a harsh penalty.


Withholding information is an offence, and courts are given interdict powers to stop a construction project. Penalties of R25 000 and/or one year of imprisonment are set for not registering or “enrolling”, for lending money for a non-registered or “non-enrolled” project, or for withholding information, as if house-building were like some heinous crime, to be prevented at all costs!


The NHBRC Act should be withdrawn in toto. There is no justification for the registration of builders, let alone the “enrolment” of houses. This Act ensures that we get less housing, rather than better housing.



7               Removal of Stamp Duties and Transfer Duties


Recommendation 20

Abolish imposts on land transfer (stamp duties and transfer duties).


In a country where land redistribution is a priority, it is anomalous that the state intervenes to make it difficult and costly. Land transactions should not be a source of general revenue, so stamp duties and transfer duties attached to land transactions should be scrapped for all first home-buyers.


Transfer duties originate in the medieval notion that the land belongs to the Lord of the Manor. In the modern world there should be no transfer cost, and imposts on land transfer should be abolished forthwith. The state should stop prohibiting and inhibiting land redistribution, and facilitate it instead.


Considerations (regarding Recommendation 20)

The obvious consequence of this recommendation would be a loss of revenue. This will be offset by the economic benefits of removing a restriction on market-driven land redistribution, as well as the benefit of removing an obstacle in the path of landless people.

Comments on Submission on land purchases by non-South Africans