Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Botswana's land tenure: Institutional reform and policy formulation

B. Machacha
Lands Division, Ministry of Local Government and Lands, Gaborone, Botswana

Historical Background

Up to the 1850s, the only major native settlements in Botswana were in those regions generally east of where the railway line now runs. Expansion into other areas of presentday Botswana occurred as groups broke off from the main tribal settlements and as a result of pressure from Boer and Zulu expansions in the east and south. The country became a British Protectorate in 1885, and in 1895 a railway line was built from c to Bulawayo as a link to the north. In exchange for this the chiefs gave the colonialists grants of land around Lobatse, and Gaborone, and along the Limpopo. The first white settlers settled in Ghanzi in 1898 and later were followed by others around Lobatse, Gaborone, and Tati, and along the Limpopo. These areas thenceforth became known as freehold land.

In the 1890s, the colonial administration requested the chiefs of the five principal Tswana tribes (Ngwato, Ngwaketse, Kwena, Kgatla, and Tswana) to identify the boundaries of their tribal territories. When these tribal boundaries had been defined, the remaining land within the Protectorate which had not been claimed (excluding freehold land) was declared Crown land to be under the jurisdiction of the colonial administration. The tribal territories became known as "native reserves." Within the native reserves the chiefs retained autonomous rule and controlled all aspects of tribal life.

At independence, in 1966, Botswana inherited three types of tenure:

Since independence, significant amounts of State land have been converted into tribal land to relieve the congestion in tribal areas. Today, only 23 per cent of Botswana's land is under State control, but this figure may be lowered further as more State land is tribalized.

In tribal areas, the chiefs were the overall authority in all aspects of tribal life. There were, however, many checks on their power, most important being the ability of their people to break away under a preferred brother or son to settle elsewhere or unite with another Tswana group. For example, even though the chief controlled tribal land, this control did not result in absolute power because he could not alienate land without the consent of his tribe. Once tribesmen acquired land, the land could not be repossessed unless it was needed for a public purpose, and even then an alternative piece of land would be given. For as long as the land was used, security of tenure was assured. Access to land was by membership of a given tribe. Land was apportioned to ward heads and subsequently to families and individuals. Tribal land was divided into three use categories: residential, arable lands, and grazing areas. These classifications can roughly be conceptualized as a series of concentric circles, in the centre of which was the village, where individuals obtained plots for residential sites. Land surrounding the village was allocated for arable production, and land further from the arable lands was designated as communal grazing areas. Access to grazing land was open to everyone in a given ward, although historically some individualization of tenure occurred as tribesmen moved to remoter areas. This has accelerated more recently as private water sources have been developed.

Even though, today, Botswana is still characterized by three types of tenure (tribal, State, and freehold), 80 per cent of the population (both human and livestock) is residing in the tribal areas. This paper consequently focuses on tribal tenure rather than the two other types of tenure. As tribal land caters for the majority of the people, most tenurial reforms have been aimed at it; hence, the interest in looking into this system of tenure.

Prior to independence a growing concern about the traditional tribal land tenure began to manifest itself. Primarily, this concern had two major components: the feeling of insecurity of tenure resulting from the arbitrariness of chieftaincy allocations and reallocations of land, and the recognition of the inability of existing customary land law to provide the security of tenure necessary for capital investments in the land for commercial purposes.

In 1968, the Botswana Government passed the Tribal Land Act. This act marked a major change in the institutional framework that handled the country's land matters. It established land boards, vested tribal land in these boards, and defined their powers and duties. With the enactment of the Tribal Land Act the authority of the chiefs in relation to tribal land was transferred to the land boards, which now hold land in trust for the tribe and continue to allocate land for customary usage. Section 10 (i) of the Tribal Land Act sets out this transfer as follows:

All the right and title of the chief and the tribe in each tribal area ... shall vest in the Land Board ... in trust for the benefit and advantage of the tribesmen of that area and for the purpose of promoting economic and social development of all the peoples of Botswana.

The Tribal Land Act also spells out the procedures for allocating land under common law land grants, which provide security of tenure in the form of leasehold land rights as well as access by non-tribesmen to tribal land. The Act further states the grounds on which customary and common law land rights may be cancelled, and establishes appeals procedures and the mechanism for State acquisition of tribal land and compensation.

The Land Boards as an Institution

With the enactment of the 1968 act, 12 main land boards were established in the country and started operation in 1970.

Structure of Main Land Boards

Land boards are non-political bodies composed of members elected and nominated in the following manner:

The membership varies from eight to 10 depending on the size of the district. Over and above this, the land boards may, with the consent of the minister, co-opt not more than two people with suitable expert knowledge or experience to assist them in the discharge of their functions.

Other people working with land boards are district agricultural officers, ward heads, who sign the letters of no objection to show that the plot applied for is available and has no conflicting use, district officers (lands), and the district commissioners, who sit in land board meetings as ex officio members.

The land board offices are staffed by land board secretaries who are also district council secretaries, administrative secretaries who work directly with land boards, administrative assistants, clerks, and land board technical officers.

The distribution of the above personnel is very skewed, with some districts having the required number of staff while others are grossly understaffed. The situation is much worse with subordinate land boards, which have only one clerk to do all the administration (handling correspondence, recording land board meetings, filing, typing, giving advice to the land board on the Tribal Land Act, etc.) and field-work (going on land allocation tours and investigating disputes).

Duties of Main Land Boards

The duties of land boards are as follows:

  1. formulation of land policies-When land boards were established, they were given the power to identify land for particular uses (e.g. grazing, arable, residential), and later were to prepare land-use plans.
  2. consultation -Before any policy can be put into practice, people have to be consulted and their responses solicited. Land boards go around notifying people of such policies. People's contributions have to be considered in the final implementation of these policies.
  3. implementation of land reform policies -This is the actual practice of what has been planned. At the moment, land boards are implementing the Tribal Grazing Land Policy and Arable Land Development Policy.
  4. imposition of restrictions on the use of land - Land, once acquired, has to be used for the purpose for which it was acquired. If a plot is used for a different purpose other than that for which it was acquired, the land board repossesses it. To ensure that land is used efficiently, the land board can repossess it if land allocated for agricultural purposes is not used for five years. Land boards also stipulate numbers of cattle to be watered at one bore-hole in an effort to conserve the grazing around water sources.
  5. collection of rentals that accrue for the use of land held under leasehold -Rentals are charged for the use of land for commercial purposes. Rentals are only paid for commercial sites because this type of land tribe has to be compensated for giving away its land for commercial usage and foregoing its traditional use.

Subordinate Land Boards

When land boards started operation in 1970, their major area of concentration was allocation of land.

The major change in the institutional framework led to doubts in many people's minds. While some thought that the new institution was going to deny them their rights to free use of land (these were mainly people who had enjoyed extensive use of land during the chieftainship era either because they were the chief's relatives or favourite friends), some felt that the new institution was a saviour in disguise. They saw the land board as the only institution through which they could gain access to land and thus establish permanent rights to the land. (These were people who for personal reasons and because of their status in the tribe had not been able to acquire land which they could refer to as theirs; they had the use of some land, but it was their master's land.) As a result, many people made applications to the land boards and they were soon overwhelmed, that is, a lot more people applied than could be allocated land. Consequently, in 1973, a network of subordinate land boards was established in bigger tribal areas to relieve the backlog in land board work. Each subordinate land board was given a working area and boundaries between them were defined. Each has five members nominated in a similar way to main land boards.

The main duties of subordinate land boards are:

Main land boards with subordinate land boards under them were left to deal with policy matters and common law allocations.

Problems of Land Allocation

In their main duty of land allocation, land boards encounter problems. The following is a breakdown of major problems encountered.

Self -Allocations

As previously mentioned, land boards found that far more people were applying than land could be allocated for. Even with the establishment of subordinate land boards, there was still a backlog of land board work. This problem was aggravated by the fact that there was inadequate transport to enable land boards to effect allocations. People resorted to selfallocations, since they could not wait for months or even years before an allocation could be effected. Some cases of self-allocation were discovered and, in most cases, the land board was blamed for taking too long to effect allocations. Most land boards were sympathetic to such problems and would formalize allocations by issuing certificates to the people involved.

Old Allocations

When land boards were established, they were told to recognize old allocations. Since in the past no register existed to show ownership of land, land boards found themselves with the problem of not knowing who owned what land. Anybody could lay claim on any piece of land on the pretext that it belonged to their parents and therefore to them under traditional inheritance laws, or that they were allocated the land by a deceased chief.

This type of claim became very difficult to disprove and, as a result, land boards found that people were occupying large pieces of unused land. Several studies on land boards have revealed this as a major obstacle in land board work. The Inter-ministerial Committee on Land Board Operations (Botswana 1977) described the problem thus: "Administrative inadequacy is particularly apparent when Land Boards are asked to confirm land rights claimed prior to their creation. No adequate registers of such claims exist. Therefore no Land Board can with assurance declare what land has or has not been allocated ... this weakness continues because there does not exist an adequate register to systematically record allocations, relate them to a particular piece of land and to a national map." Comaroff (1977) also brought this problem up in his study of the Barolong Land Board: "There was no way the Land Board could cope with pre-allocations; people applied for large tracts of land and when refused they would justify their claim on the basis that they already owned the land from an old allocation." Today, most land boards are still faced with problems of claims on land which are said to be from old allocations. Where there are alternative vacant plots, land boards normally avoid conflicts by asking applicants to seek other plots elsewhere. By and large, these plots are not yet allocated or have been reclaimed.


When land boards started operation, they were expected to keep records of the land allocations they made. Each land board was left to devise its own record system which it understood. Recently, it was noted with concen by the higher authorities that, although there are records in land board offices, these records do not in any way help land boards because they cannot be referred to easily. The main problem is that of record fragmentation, so that when one traces the history of a single plot, one may have to go through more than 10 sets of records. The tendency is for officers to be reluctant to refer to such records and, as a result, some cases which have been dealt with are replicated over time. Most people have come to realize this handicap and take advantage of it. For example, if a person appealed and lost a case, he waits for three to five years and resubmits the appeal, well knowing that the land board will not be in a position to know of its former decision. This happens frequently when there is change of office by members or land board personnel. Sometimes the appeals are ruled otherwise and this is an embarassment for the land board if later on they find that they have changed their first decision.

With the workloads that land boards handle and the records that are kept, land boards find it difficult or impossible to check on allocations they have already made. As a result, people can themselves extend or change plots without the land boards finding out. Sometimes land boards reject the suggestion that they ever allocated such plots until a certificate issued by them is produced. Again, this can cause people to doubt that land boards know of their own allocations. In the case of appeals, if land boards hand down decisions, people may choose not to heed such decisions, and unless the rights of other people are infringed upon, the land boards never know whether their decisions have been implemented or not.

Enforcement of Land Board Decisions

The Tribal Land Act gives land boards sweeping powers, including:

But the act does not provide for the enforcement of these powers. Even if land boards hand down decisions on certain issues, there is no enforcement capacity to make sure that those decisions are observed. The only way land boards get decisions implemented is when they take defaulters to court and a court eviction is made. This is a very expensive and time-consuming exercise, which most land boards are reluctant to undertake. Thus, in most cases, land boards never even bother to make decisions, or if such decisions are made and people do not observe them, they prefer to ignore the problem and let it pass.

Solutions to Problems of Land Allocation

For land boards to work efficiently, the problems outlined above should be addressed. Some of the possible measures that can be applied are discussed below.

Strengthening Land Board Enforcement Powers

The lack of enforcement powers as a major handicap to land board work has not been noted by land boards alone. The Interministerial Committee on Land Board Operations (Botswana 1977) also indicated this as a major problem and recommended that land disputes be transferred to customary courts if they dealt with customary land grants and to subordinate courts if they concerned common law land grants. These courts will in turn be responsible to enforce the decisions reached. The transfer of disputes will not require those institutions to create any new structures, as they already have enforcement powers.

The Local Government Structure Commission (Botswana 1979) also endorsed this recommendation and further recommended that, since land boards will still be responsible for a wide variety of rulemaking and legal functions, such as: preparing and presenting land cases at the customary courts and magistrate's courts; executing lease agreements with common law leases; adjudicating and monitoring land reform policies and programmes; monitoring land-use plants; and formulating and implementing new policies, they need to be strengthened in their enforcement powers. The Commission recommended that this be done by establishing a Land Board Enforcement Unit inside the Attorney General's Office. In the meantime, before this unit is operational, funds should be made available for land boards to hire private lawyers. Presently, land boards seek legal advice from the office of the Attorney General and, if they have cases to take to court, the Attorney General takes up these cases.

Provision of Adequate Law

In 1968, when the Tribal Land Act was drafted, it was mainly based on the traditional tenure system that existed at that time. With the operation of land boards for over 10 years now, experience has been gained, new policies have emerged, and new reforms have been adopted and implemented that changed the tenure sysem. New laws need to be enacted to accommodate all these changes. There is also a need to amend the 1968 Tribal Land Act. It is understandable that the act has tended to be vague and never addressed certain issues, since at the time of drafting there was nothing to base it upon. Now, with new experience of what actually takes place as regards the use of tribal land, we are in a position to prepare a better document. The amended act should not only delineate what the duties of land boards are but should also outline how land boards are to go about carrying them out and what they should do if certain problems arise (a point which is lacking in the current land act). The new act should also update Botswana land law to accommodate and facilitate the new land reforms that have been undertaken.

Proper Record-keeping and Registration of Land

The record system of any institution is the central system that controls everything. It would be useless to correct other land board deficiencies and not the core one, that of records. A proper registry system should be designed which could be used by all land boards. This system should be easy to use and refer to. The Ministry of Local Government and Lands has now designed such a system and land boards have been asked to comment on it before it is implemented. This system will make supervision by the Government easier as it will be uniform; each land board uses its own method at present and the supervisors have to learn each method before they can comprehend what is being done in each office.

Once records are straightened up, the next step will be to devise a land registry which will relate all land allocation to a particular location and then to a national map. Only in the case of commercial plots have land boards, after employing land board technical officers, been able to draw sketch plans for such plots. It is only after land registration that land boards will be able to know what land has been allocated, where the allocation has been effected, and who owns the plot. This information will in turn alert land boards as to who owns how much land, whether that land is effectively used, (and, if not, whether they can repossess some land for reallocation), and how much land is left for allocation. Land boards are at the moment unable to make such decisions because they are ignorant of these facts.

Land Board Training

To ensure that the above corrective measures are properly implemented, land boards personnel should be given proper training. A training unit has been established in the Ministry of Local Government and Lands and training programmes for land board members and staff are presently under way.

Policy Perspective

Section lI (i) of the Tribal Land Act stipulates that land boards are responsible for policy formulation and implementation in respect of land use in their areas. During their term of office major policies have been formulated, some of which have involved major land policy changes. Some of these policies have demanded technical expertise and sophisticated management, which land boards lacked drastically at the time of implementation. In some cases, technical staff had to be hired to handle all the technical issues that needed attention. Land boards have since been involved with taking technical decisions, such as approving land-use plans. Examples of major policies under the administration of land boards are briefly discussed below.

Tribal Grazing Land Policy

This policy was initiated in 1975 to curb the widespread overgrazing that was evident in the country and to improve the cattle industry by the use of simple farm management skills. The introduction of this policy marked a sharp diversion in land board work and the traditional tenure system of the use of grazing land. Traditionally, grazing was communally shared, but the policy called for individualization of grazing rights, allowing people exclusive rights to certain grazing areas. The Tribal Grazing Land Policy provided for the zoning of tribal land into the following three categories:

  1. communal-in which the traditional existing tenure system would continue to be used with a few modifications (like stock limitation and fencing of grazing areas) if the community consents;
  2. commercial-which would be areas designated as ranches. Groups of people or individuals with more than 400 livestock units would be encouraged to move into those ranches and practice simple farm management like paddocking, stall feeding, daily watering, early weaning, breeding, etc. People in ranches would be given exclusive rights and they would hold land under a leasehold system. For this privilege they would be required to pay a minimum rental of P 0.04 per hectare (to be reviewed every five years);
  3. reserved land-which would be areas set aside for future expansion and for those who do not own cattle at present.

The zoning exercise culminated in the preparation of land-use plans in the whole country. Six districts (Central, Kweneng, Southern, Kgalagadi, Ngamiland, and Ghanzi) identified commercial zones for ranching. The implementation of the Tribal Grazing Land Policy started in 1977 and 1978, and the following processes began to be administered by land boards:

To date, all the six districts have at least allocated the Tribal Grazing Land Policy ranches in their areas and, in all, 302 ranches have been demarcased, 171 ranches have been advertised, 122 ranches have been allocated, and 52 leases have been signed (or are in the final stages of signing).

Emergence of Problems

With the implementation of the policy there emerged numerous problems, some of which are discussed below.

The assumption made at the planning stages of this policy that there were vast areas of open unused land in Botswana was proven wrong during implementation. To date, substantial numbers of ranches have been dezoned to give room to communal rights, upon which they had encroached. Compenstion pressures have mounted and this is another indication that there are people with established rights in the commercial zone.

Another assumption made by the policy that large herds will be moved to commercial areas to leave room for small holders has also not been realized. In most cases, these ranches have been zoned in areas that have already been utilized, which means people who have water rights in those areas will be given the ranches. In actual fact, what has happened is movement of people from commercial areas back to communal areas (small holders, hunters and gatherers).

Major decisions have to be made by land boards about to whom they should allocate the ranches, how many ranches to allocate to one person, whom to move out of the commercial areas, where to place them, how much compensation has to be paid and what constitutes adequate compensation. These are major questions, which, if not seriously addressed, could hamper the progress of the policy.

The tenurial implications of the Tribal Grazing Land Policy as a national policy are far reaching. While the initial intent of the policy was to benefit all, recently the policy has been labelled as one for the rich. This is because poorer people are being forced out of their traditional areas to make room for richer commercial ranchers. The first experiences with the policy were that compensation amounts were assessed and people who accepted moving were paid. But it turned out that ranchers delayed payment of the compensation and that boards could not provide immediate services to areas that the people moved into. The situation in Service Centres (i.e., areas in which displaced people are put) has become controversial. These displaced people were suddenly exposed to a different type of environment, for example, the hunters and gatherers suddenly found themselves in a limited area of 6,400 hectares to do their hunting and gathering. With commercial areas now identified in most districts, opposition to the allocation of these ranches is intensifying. People now realize how these ranches will affect them and they are saying they originally accepted the Tribal Grazing Land Policy because they thought the ranches were going to be far from existing settlements. There is widespread fear that land will soon be taken up by ranches, and this further intensifies resistance to the policy.

Land boards have been very careful in the allocation of ranches and where there has been adequate proof that people's rights have been infringed upon they have willingly dezoned the ranches. The problem is that as pressure grows against allocation of the Tribal Grazing Land Policy ranches, land boards will have difficulties even in allocating ranches which genuinely do not affect the rights of other people.

The problems of land boards are aggravated further by the farmers who have been allocated the Tribal Grazing Land Policy ranches but have since misused their ranches and are now returning to communal areas. It will now be impossible to convince small farmers that they will have more land to themselves with the movement of larger herds to commercial areas. The success of the Tribal Grazing Land Policy lies in the very strict application of regulations to ranchers: to fence, observe stocking rates, and apply simple farm management skills. Failure of responsible bodies to enforce those regulations will destroy the policy and discredit land boards as the administering body, and this can complicate the tenure problems of this country even more.

Arable Lands Development Programme

The Arable Lands Development Programme developed as a counterpart of the Tribal Grazing Land Policy, to address the problems of arable farmers rather than cattle farmers. It should be noted that in Botswana 45 per cent of the people do not own livestock at all and that 50 per cent of the national herd is owned by only 3 per cent of the population. The major components of the policy are to give subsidies for fencing, building small dams, and buying draught power (donkeys and oxen) and other crop growing implements. Under this policy, land boards are required to provide information as regards their policy on arable land allocation. In 1980, the Ministry of Local Government and Lands requested each land board to forward its policy on arable land allocation.

The Arable Lands Development Programme has several landrelated implications. As people make capital improvements to their land by way of fencing and establishing reliable water sources, security of tenure will increase. With increased security, people will want to increase their land holdings and consequently the land base will become exhausted, that is, land will become more scarce. Further, the provision of subsidies for draught power will make it easy for people to own more animals, which will in turn compete with arable uses as they have to be provided with grazing. Lastly, land boards may be forced to reallocate unused land, some of which may be fenced. This brings in compensation issues, which have proven to be very difficult to handle under the Tribal Grazing Land Policy.

The tenure implications of the Arable Lands Development Programme as a policy aimed at the majority of Botswana (people from Botswana) needs close monitoring. A quick survey should be carried out to help land boards formulate proper policies on the size of plots to be allocated in their districts. If haphazard allocations continue for too long, the political and social problems that may arise when a different trend is later adopted may be considerable.

Land Inventory and Registration Programmes

In the discussion of land allocation as the land boards main duty, the chief problems outlined were later suggested to be due to the fact that there does not exist a register to which land boards can refer and see who has rights to how much and where. The solution to this problem was indicated as the establishment and use of a land register providing such information.

To establish such a register would require technical inputs, which Botswana's land boards do not have at the moment, that is, qualified staff to draw maps, land boards personnel who can read and understand these maps, and improved record-keeping for the administration of the system. Consequently, before a land register can be used, all these should be available. The need to prepare our land boards and supporting staff to undertake these types of programmes has already led to the recruitment of land board technical staff and the offering of courses to land boards on map-reading and use. In some districts, this has been taken further by the introduction of land inventory programmes, although they are presently carried out on a pilot project basis. The land inventory programmes were introduced as a result of the acute shortage of land in some districts and the realization by some land boards that investories may improve their land allocation records, and that this will help them to determine what land is available for allocation (unused or allocated but Iying unused for some time and now available for reallocation). Presently, land boards, once they have allocated a plot, have difficulties in knowing that the allocation was done by them. If the owners of such plots cannot produce certificates, land boards can easily deny the allocations.

With the information obtained from land inventories (who owns the plot, use of plot, size of plot, when allocated, etc.) land boards will have more complete land records.

With pilot projects now at their completion stage, the next step is to test the administrative capacity of land boards to administer these inventories, that is, to keep the necessary records, utilize them, and update such records as and when new allocations or transfers of plots are effected. Without proper record-keeping, these land inventories can be rendered useless within a short time. Thus, it is the wish of the central Government to build up the administrative capacity of the land boards before these inventories are undertaken at a national level.

It is rather early to envisage what implications land inventories will have on the tenure system of the country, but certain assumptions can be made. Once plots have been identified and their uses and owners known, land boards will have information as to whether there is multiple ownership of plots and whether the plots are efficiently used. Considering the land pressures in some districts, some land boards, on the basis of this information, will be forced to reallocate unused plots or to cut down the acreage per person. Further, the registration of plots may result in other members of the family being left out. Traditionally, the man, as the head of the family, applies for land to be used by the family. If the plot is registered in the name of the head of the family, legally the plot belongs solely to him and, in the case of divorce, the other members of the family may have no claims on such land. This means that the registration has worked against their customary interests and rights. This may also mean that there may be multiple ownership of land if one person registers all the plots in his name while, in actual fact, the plots belong to other members of the family who are his dependents. If land boards reallocate on the basis of number of plots held by the individuals, some big families may be disadvantaged as they will be forced to lose some of their land.

Land boards have to be very careful in the use of information obtained from land inventories; the traditional lineage structure should be considered before any concrete policies can be formulated based on information in the inventories. Since no studies have been done yet to evaluate these inventories, it is hard to tell what people think of these programmes. Studies of this nature are now required.


This chapter has indicated the type, status, and volume of work that land boards, as land management institutions, have to handle. Most studies on land boards have, on the one hand, revealed that land boards are hardworking, conscientious, and honest, but, on the other hand, that they have not been doing their work efficiently. After summing up their work, we can appreciate their problems. Actually, for new institutions of their calibre, land boards have functioned fairly well, considering that major policies like the Tribal Grazing Land Policy have got off the ground under their administration. Given proper training, proper supporting staff, and proper policy guidance, land boards can perform even better. The future of Botswana's land tenure and the success of land reform policies rest with the development of these institutions.


Botswana. 1977. Report of the Interministerial Committee on Land Board Operations. Government Printer, Gaborone.

________. 1979. Report of the Presidential Commission on Local Government Structure in Botswana. Government Printer, Gaborone.

Comaroff, J.L. 1977. The Structure of Agricultural Transformation in Barolong: Towards an Integrated Development Plan. Good Hope, Johannesburg.

Contents - Previous - Next