Land grabbing in Botswana: Modern era dispossession

Land grab refers to the formal transfers of large tracts of communal land to foreign or locally based investors for carrying out activities associated with livestock rearing, carbon trading and commercial food production. With the acquiescence of host governments, transnational and multinational corporations are viewed as key players in land grabs. Among the major consequences of land grab is the involuntary loss of land by the rural poor. While the presence of external players in land grab is portrayed as dominant, this article introduces a land grab model where the dominant players are the host country’s ruling elite. Using case examples from Botswana, the article aims to expose the land-grabbing tendencies of the country’s land-tenure reforms, as well as document sites and spaces of resistance available for local communities to curtail land grabbing. It is contended that, in a bid to curb state-sponsored land grabbing, social justice activists in Botswana can draw lessons from the community initiatives discussed in this article.


INTRODUCTION
Land grab features as one of the topical subjects in the global South and, according to Alden Wily (2012), discussions on the subject have intensified since 2008. Zoomers (2010: 229) defines land grabbing as large-scale cross-border land deals or transactions carried out by transnational corporations or initiated by foreign governments. Borras and Franco (2010) understand land grabbing to be a catch-all phrase that describes (trans)national commercial land transactions occurring in recent years. Alden Wily (2012: 751) defines land grabbing as involving the formal transfer of land from peasants' farming and pastoral activities to large-scale land producers and/ or speculators. A commonly held concept of land grabbing places the phenomenon within the context of an ever-expanding globalised capitalist transformation that leaves dispossession in its wake (Alden Wily, 2012: 752). Similarly, Zoomers (2010) views land grabbing as driven, among other factors, by the liberalisation of land markets, globalisation, and the quest for Foreign Direct Investment (FDI).
In terms of magnitude, geographical areas targeted for land grabs include sub-Saharan Africa, South America, and Asia. As aptly captured by Dlamini (2019: 387), these areas are viewed as having an abundance of arable and unexplored land. Alden Wily's definition draws attention to two important features of land grabbing that will be explored further in this article: its legal nature and the resultant land dispossession of peasant farming communities. In the case of Botswana, the legality is provided for by the land-tenure reforms that culminate in the de facto privatisation of communally owned land.
In analysing land grabbing, it is important to understand how the role played by the domestic elites and government is often interpreted. Most common definitions of land grabbing view external playerstransnational and/or multinational corporations -as the key players; hence, the constant reference to 'sale and leasing of land to foreign companies and individuals'. This is evident, for example, in the call for papers for the first Africa Conference on Land Grab held in 2014, where it is stated that "[s]peakers, panellists and papers will look at country case studies, challenges, advantages and disadvantages of the sale and lease of large swathes of arable land to powerful foreign individuals, multinationals and governments" (CIFOR, 2014: online (2012) views governments as the primary agents in land transfers, as they formulate and legitimise the policies and laws that promote land grabbing.
According to Peters (2004: 271), land grabbing as appropriation by the elites must be situated within broader processes of social inequality and class formation as well as within what commentators call new forms of governing. The acquiescence of the host government in land grabbing highlights the predatory nature of the national elite. This is aptly captured by the former South African president, Thabo Mbeki (2014: 20), when he referred to "the undeniable reality that African independence has resulted in the emergence in our countries of a dominant predatory political and economic class as a rent-seeking elite which pursues selfish interests that have nothing to do with the fundamental aspirations of the generality of the African masses". Under this interpretation, national governments facilitate land grabbing, while the 'grabbers' are external players. While the involvement of external players is dominant in other parts of Africa, the case of Botswana shows that it is not so much transnational players who are at the centre of land grabbing; instead, the land grabbers are the national elite. The identification of land grabbers is significant particularly when it comes to alternatives regarding how land grabbing can be curbed.
Although the term 'land grabbing' is not commonly used in Botswana, the process is implied in deliberations on the land question, especially in analyses that warn against land dispossessions of the San peoples in Botswana (Ng'ong'ola & Moeletsi, 1996;Good, 2001;Manatsha & Marhajan, 2010;Sapignoli & Hitchcock, 2013). This article argues that land grabbing is driven by the national elite in search of rent-seeking opportunities in commercial opportunities such as tourism, mining, cattle ranching, real estate and, in some instances, land hording for speculation purposes.
Using case examples from different parts of Botswana, this article introduces a land-grab model where the key players are the country's ruling elite. Through successive land-tenure and policy reforms, Botswana's ruling elite has managed to privatise communal land and rendered such areas inaccessible to sections of the rural communities. In addition, the article documents sites and spaces of resistance through which communities have started to challenge the elite-aligned land-tenure and policy reforms. It is contended that, in a bid to curb statesponsored land grabbing, progressive actors and social justice advocates in Botswana can draw lessons from these community initiatives.

Land grabbing through land reforms: The policy context
Land grabbing in Botswana should be understood within the wider context of the country's political economy.
Botswana is often described in the most contrasting views that ranging from those that portray it as Africa's 'great success story' to those that project a not so positive view of the country. According to Good (2008), the view of Botswana as African 'miracle' growth, governance, stability and democracy is projected by the country's leadership and some Western commentators. Contrary to the 'success' story cited above, Botswana has been described as exhibiting authoritarian liberalism (Good, 1993;Taylor, 2003). According to Taylor (2003: 216), "Botswana has combined high growth rates and visible development with structured autocracy that belies the benign image internationally." To support the above assertion, analysts often cite the country's poor record on poverty and human rights, particularly the issue of the San people (Saugestad, 2001;Good, 2008). Questions have been raised concerning what is viewed as lack of accountability and transparency in the manner in which government carries out its business. The contention is that Botswana's success narrative masks major inequalities in resource distribution and disguised land dispossessions of rural communities.
Land grabbing in Botswana results from successive land-tenure and policy reforms initially introduced in the colonial period and further entrenched in the post-independence period. The story of land-tenure and policy reforms in Botswana has been one of relentless appropriation of the communal land by the country's elite. The elitist land-tenure reforms, carried out in the name of development, involve the gradual replacement of the endogenous communal land-holding system with individualistic land-holding models. The reigning logic that drives these reforms is founded on economic models that portray communal land holding as inhibiting investment in agriculture. The state and its administrative structures in Botswana project themselves as promoting "national interests", as opposed to the interests of any particular class. Rakodi (1986) rightfully argued that this conception of the state informs technocratic and rationalistic views of policy formulation in developing countries. This conception of the state is further promoted through what has been referred to as the ideology of developmentalism, in which the elites, forming a government in league with those manning the state institutions, view themselves as modernising agents (Charlton 1992: 274). Under this thesis, class issues implicit in the policies being pursued are obscured as everything and everybody are viewed as united in seeking "development for the good of all". In the process, the development process is depoliticised, as policies contained in development plans are usually couched in a neutral language. This view of policy and policy formulation pervades the development discourse within which land grabbing occurs in Botswana.

Land grabbing in the colonial period
Land grabbing in Botswana can be traced back to the land reforms initiated by the colonial administration through the introduction of a new land-tenure system. Prior to colonisation in 1885, the nationalities that later constituted Botswana lived as autonomous entities consisting of Bantu and non-Bantu groups. Among the Bantu groups, land was communally owned and administered on behalf of the morafe (nation) by the kgosi (chief) (Schapera, 1994). A tripartite land-use system was common among the Bantu groups and each member of the community had access to land for residential, arable and livestockgrazing purposes (Kalabamu, 2012;Schapera, 1994). Land rights for the above uses were inheritable and perpetual (Kalabamu, 2012). An important attribute of this indigenous tenure system, which was to change later in the post-colonial period, was that every member of the morafe had access to natural resources in both unallocated and fallow land (Schapera, 1994).
The aim of colonial land tenure was primarily to further the economic and political objectives of colonisation and differentiated between native reserve (communal land), freehold and Crown land which was renamed state land at independence. Freehold land was created specifically for the settler communities engaged in pastoral farming. Crown land was reserved mainly for game and nature reserves as well as for some of the emerging urban centres. Central to the colonial land-tenure system was the preference for private land ownership over communal as the latter was viewed as inappropriate for investment in agriculture.
The view that the colonial landtenure system promoted land grabbing stems mainly from the land dispossessions that accompanied the carving out of both freehold and state land. Local communities were evicted without compensation to give way to freehold farms and game reserves. In the case of the presentday North East District, for example, activities of the Tati Company led to the eviction of local communities to give way to farms for settler farmers. Similar evictions and relocations took place in the Tuli Block and Gantsi Block Farms were John Cecil Rhodes' British South African Company was active in settling European farming communities (Molebatsi, 1994). Freehold land was carved out of the country's most fertile areas and, once evicted from those areas, the native population either enlisted in labour migration system for South African mining economy or became labourers on the farms owned by Europeans. In the case of the Gantsi Block Farms, for example, displaced San communities became labourers on the freehold farms where they worked under extremely exploitative conditions (see, for example Mogalakwe, 1986).
An important development in the creation of freehold farms was the fencing off of land held under freehold. Fencing became a defining characteristic of freehold farms in the Gantsi Block, Lobatse Blocks, Gaborone Block, the Tuli Block and the Tati District. It was generally believed that fencing would lead to better land management of freehold farms. It has been argued that most of the freehold farms were indeed well managed and, to an extent, provided range management practices to be emulated (see Chambers & Feldman, 1973;Tsimako, 1991). While fencing of land had some advantages regarding land management, its downside is that, in communities where land provided more resources than merely livestock grazing, fencing effectively excluded communities from accessing other sources of livelihood found within the fenced areas. As shown in the next section, fencing of communal land and accompanying land dispossession of rural communities were at the centre of two landreform policies that were adopted in the post-independence period.

Land grabbing in the postindependence period
During the post-independence period, the colonial land-tenure system was retained with changes mainly in the nomenclature. While the category 'freehold' was retained, native reserves and Crown land became tribal land and state land, respectively (Malatsi & Finnstrom, 2011). While there has been an increase in the proportion of tribal land and freehold, state land has significantly declined. This has led Adams et al. (2003) to conclude that the policy in Botswana was to increase the area of tribal land at the expense of both state land and freehold. While the above observation is true, it should be stated that, through the issuance of leasehold and various land management practices, a significant proportion of tribal land is held under leaseholds that have led to the de facto privatisation of such land. This is the case with the land reforms ushered in by the Tribal Grazing Land Policy (TGLP) (1975) and its successor -the fencing component of the National Policy on Agricultural Development (NPAD), introduced in 1991.
Several authorities have debated the circumstances surrounding Botswana's post-independence land-policy reforms ushered in by the TGLP and the NPAD (RoB, 1975;Sandford, 1980;Tsimako, 1991;Sapignoli & Hitchcock, 2013). According to Cullis and Watson (2005), the changes in land policy were founded on Garret Hardin's 'tragedy of the commons' thesis, which projects a negative view on communal land. Within this thinking, communal land is viewed as implying open access and as such lacking in control mechanisms. It is, therefore, highly susceptible to degradation, due to lack of control mechanisms on the number of animals that could be allowed into such areas. Individuals within the communal areas are viewed as pursuing personal interests with minimal regard for consequences on the environment. The contention was that rangeland degradation could be addressed by replacing communal ownership with individual allocation of land.
The policies sought to address the perceived overgrazing in the country's rangelands by creating commercial leasehold ranches. It was envisaged that this would enable large livestock owners to move out of the overstocked and overgrazed communal areas. Furthermore, the new ranches, each measuring 8 km x 8 km, would also provide an opportunity for the farmers to adopt an improved livestock management system. The shift of livestock by large cattle owners from communal areas to leasehold ranches was also viewed as promoting social equity, in that it would create room for expansion for small farmers. Evaluation of the Tribal Grazing Policy suggests that, instead of promoting social equity, the policy led to the problem of dual grazing rights. Research has shown that, by moving their cattle to the TGLP ranches, ranchers do not relinquish their rights in the communal areas. Quite often, they continue to enjoy exclusive grazing rights in their ranches, while simultaneously maintaining their access to communal areas. The tendency has been that farmers, who were allocated ranches, would release their livestock into the communal areas during drought periods and move them back into the ranches as and when grazing conditions improved. The argument maintained by perpetrators of dual grazing is that owning a ranch does not strip them of their rights as members of a community. According to Machacha (1996), attempts by some Land Boards to curb dual grazing rights were unsuccessful, as legal advice indicated that the perpetrators had constitutional rights regarding access to communal grazing areas.
The NPAD was an extension of the TGLP and called for the expansion of leasehold ranches in communal areas. Fencing was to take place "where possible'', a phrase that was interpreted by the land authorities to mean communal areas with minimal clustering of livestock watering points. It is important to note that, like the freehold farms before them, once fenced, these former communal areas become semi-private areas accessible only with express permission from the leaseholders. Between the two land-reform policies, roughly 45 347 km 2 or 8% of Botswana's land mass was allocated to individuals for ranching purposes.

Impacts of land grabbing on rural communities
Land grabbing in Botswana has led to land dispossession of the vast majority of rural communities who depend on the communal land for their livelihoods. These include subsistence farmers and those who depend on veld (natural) products gathering 1 as a source of livelihoods.
Over two-thirds of the country's livestock cattle depend on communal grazing areas. The creation of fenced ranches within the commons has, 1 Veld products gathering refers to the subsistence harvesting and collection of natural resources products for food, construction or medicinal purposes. The term 'veld' (also spelt 'veldt') is commonly used in Southern Africa in reference to natural, open lands. therefore, effectively reduced the amount of land available for grazing and resulted in overstocking and, ultimately, rangeland degradation in communal areas. The losers, in this instance, are the small-scale or subsistence farmers who cannot afford the leasehold ranches.
In addition, the procedures followed and the requirements or qualification for ranch allocation are generally beyond the reach of subsistence farmers, as illustrated by the case cited below. In 2010, the Ngwato Land Board advertised commercial ranches in the Bangwato Tribal Area (Central District). This followed the identification and demarcation of 30 ranches in the Kaka/Western Sandveld Areas. Each ranch measured 6 km x 6 km and was advertised in local and national newspapers. Application forms were obtained from Serowe, the district headquarters, and were obtained at a fee of P500. According to the application package, the P500 was for 'cost recovery'. Applicants were also required to include a Management and Development Plan for the ranch. Under finances, applicants were required to supply financial statements for the past three months, available plant and machinery, and proof of financial support for ranch development. This could come from financial institutions or be based on the applicant's current livestock numbers certified by the Animal Health and Production (Ngwato Land Board, 2010). Further, it is obligatory that the ranches be developed within three years of allocation. Recipients are expected to undertake the following development activities on the ranches: erection of a perimeter fence around the farm and subsequent paddocking; water reticulation, and stocking. It has been argued that the allocation of the ranches is shrouded in secrecy, resulting in allegations of favouritism towards political elite (Sunday Standard, 2012). The above requirements are certainly above the means of small-scale farmers.
Fencing of communal land has also led to loss of additional sources of income for rural communities. A common attribute found in both policies is their failure to recognise the multiple use nature of rangelands (Rampete, 1996;RoB, 2003;Cullis & Watson, 2005). Apart from livestock grazing, communal rangelands are also used for gathering and foraging by local communities. The collection of veld products or gathering forms an important component of livelihoods for the rural poor in Botswana. Despite public sector pronouncements on the significance of gathering to rural communities, gathering is not yet a recognised activity for which land has to be set aside.
The non-recognition of veld products gathering as a land use in its own right is closely linked to the sociopolitical and economic marginalisation of groups historically associated with hunting and gathering, namely the San communities. Marginalisation of gathering as a land use and the subsequent classification of communal rangelands as either commercial farming or communal grazing areas means that those who use these areas for purposes of gathering or foraging often lose their land-use rights to cattle owners. Under the existing set-up, expansion of agricultural land -be it grazing or arable farming -cannot be curbed or halted on grounds that such expansion threatens the viability of gathering by local communities.
Resource depletion, due to increased livestock population, was widely reported in selected remote settlements in Kgalagadi and Ghanzi Districts (RoB, 2003). Similar problems were reported in the settlements of Khwee and Mmiya (Ndozi & Toteng, 1989). Studies have shown that failure to recognise veld products gathering as a bona fide land use has effectively rendered the San landless (Moeletsi, 1993). Quoting the work of Campbell, Main and Associates, Good (2001: 22) concluded that, in the case of the Western Sand Veld, the demarcation of ranches proceeded without regard for the San who could be told to move by anyone who had been awarded land rights by the land board (Good, 2001: 22). In a study of Western Ngwaketse, Childers (1976) revealed the negative impacts that TGLP ranches had on people whose livelihoods were partly based on gathering. Commentators on land reforms in Botswana have raised equity issues in relation to the de facto privatisation of communal land. Perkins (1996) refers to 'fencing off equity issues' -an attempt to describe the de facto privatisation of communal land under the NPAD. Perkins' point emanates from the fact that often the subdivided land is regarded as 'unoccupied', since, for both TGLP and NPAD, 'occupation' seems to be confined to recognised land uses, as shown above, excluding veld products gathering.
Land grabbing in Botswana is accompanied by the displacement and relocation of rural communities and, as in the case of threatened livelihoods discussed earlier, San communities are at the centre of such forced relocations. Geographically, the majority of the TGLP and NPAD ranches are located in the western parts of Botswana. This was largely due to the non-recognition of San land-use systems that differed significantly from those of the dominant Tswana/Bantu groups. Relocations were instituted to pave the way for the establishment of ranches and other land uses such as tourism and mining. The most publicised case was the relocation of San communities from the Central Kalahari Game Reserve (CKGR) in 1997 (see, for example, Good, 2008). Despite initial denials by the government that the relocations were motivated by mining and tourism interests, by 2013 there was active diamond mining in the CKGR (Motlogelwa, 2013: online; Survival International, n.d.).

Spaces and sites of resistance to land grabbing
The preceding discussions clearly demonstrate that state-sponsored land grabbing is widespread in Botswana and has resulted in land dispossession among rural communities. This section documents sites and spaces of resistance considered critical in providing entry points, through which local communities and activists could address land grabbing. The concept of sites and spaces of resistance has been employed to describe opportunities and possibilities, through which the marginalised or those hard done by the development project can begin to challenge the status quo. According to Cornwall and Coelho (2007: 1), these spaces may be provided by the state, and they may also be spaces conquered by civil society demands for inclusion. The contention is that such spaces could be innovatively used by the marginalised to call for meaningful changes in policy formulation. Miraftab (2009) views these spaces as potential bases for what she dubbed 'counter-hegemonic' planning practices. While such spaces might be created by the state, they can be creatively used to mobilise for social transformation. The following are regarded as offering opportunities for curbing land grabbing in Botswana: using the country's sensitivity to international pressure; using existing legislation and structures, and supporting land claims activities by civil society organisations. Attempts to have one of the farms withdrawn from the communities and allocated to a syndicate were abandoned following protest from NORAD (see Saugestad, 2001).

Sensitivity of the Botswana Government to international pressure
The protest campaign waged by the London-based Survival International against the 1997 relocation of San communities from the CKGR also demonstrates the country's susceptibility to external pressure. The strategy adopted by Survival International, in particular, linking the relocation to diamond mining in the CKRG and in the process equating Botswana's diamonds to blood diamonds, prompted the Government to change its strategy towards the organisation. At a press conference convened on 7 November 2002, the then President of Botswana Festus Mogae declared openly that "Survival International is now shifting its campaign away from government and is now targeting consumers in an attempt to kill our diamonds. Now we are going to respond to every allegation that Survival makes because that is serious sabotage to our economy" (Botswana Daily News, 2002a).
Allegations by Survival International and counter-allegations by the Government of Botswana triggered numerous visits by delegates from outside Botswana. These were mainly political figures from Britain, including Glennys Kinnock and a delegation led by Richard Howitt who were obviously taken seriously by the Government (Botswana Daily News, 2002b). Although the Government of Botswana did not rescind the CKGR relocations, international pressure led to what could be described as overprovision of infrastructural services in the new settlements. The aim of this move was to show the international community that the services provided in the new settlements were indeed of high quality. In providing services to these settlements, planners worked, in some instances, with presidential directives such as, for example, in the establishment of the settlement of Xere. Between 1997 and 2002, roughly P44 million was spent on the three settlements on physical infrastructure such as water, schools and health facilities as well as the issuance of livestock and the construction of factory shells.
In suggesting the use of international pressure to urge the Government to address land grabbing, it is important to acknowledge possible hurdles inherent in this strategy.
If not carefully thought out, external pressure can be counter-productive not only in diverting attention away from the case in hand, but also by galvanising support for the Government. The contention is that, in seeking external and/ or international support, lobbyists from within the country should guard against possible backlash and accusations of being unpatriotic.

Working within existing legislation and structures
Rural communities in Botswana have used existing legislation to increase their access to land and the same strategy could be used to guard against land grabbing. Existing legislation and structures have successfully been used by the San communities in challenging relocations. In one of the longest trials ever conducted in Botswana, San communities successfully challenged relocations from the CKGR. The Government of Botswana lost a similar case, in which residents of Ranyane were to be relocated from their ancestral land (Open Society Initiative, 2013: online).
The use of existing legislation is also proffered by Ngo'ngo'la and Moeletsi (1996) and Moeletsi (1993) as possible ways whereby the state could be pressurised into improving San communities' land rights and access to natural resources. Opportunities for addressing land grabbing also exist in the Constitution of Botswana. Ngo'ng'ola and Moeletsi (1996) suggest invoking Section 8(1) of the Botswana Constitution as a possible strategy, whereby rural communities could be compensated for loss of land-use rights, due to the implementation of TGLP. Still on the question of loss of land-use rights, the fencing of communal grazing areas under the NADP is supposed to take place "where possible". However, the phrase "where possible" has not been defined in any specific manner. It can be argued that, where fencing has been proposed, despite obvious negative impacts on rural communities, the "where possible" clause could be invoked in challenging the move. This will obviously require vigilance on the part of those who intend challenging such moves.  KFO, 2010). As in the case of TXY, the two organisations used existing structures to facilitate land acquisition by San groups. One of the strategies employed to assist the San to acquire land was through the establishment and registration of syndicates. An evaluation exercise of KFO activities revealed that TOCaDI managed to assist 10 groups to register borehole syndicates. For their part, between 2009 and 2010, Komku increased the registration of cattle syndicates from 5 to10. This is a major achievement, given the role that boreholes play in facilitating or limiting access to grazing land in Botswana. Water rights conferred on individuals or groups automatically give them de facto grazing rights over an area that could range from 3 600 ha to 6 400 ha. This section submits that several options could be used to guard against land grabbing by the country's ruling elite.

Supporting land-claims activities by civil society
San communities have also embarked on activities aimed at reclaiming land that the lost to other groups in Botswana. These efforts are at the heart of Participatory Geographical Information System mapping (PGIS) exercises facilitated by the KFO. According to Taylor and Murphy (2006), PGIS has been used in Community-Based Natural Resource Management (CBNRM) projects in southern Africa to empower local communities to 'articulate their spatial thinking'. Nuulimba (2003)

Building on existing efforts to curb land grabbing
The main suggestion advanced in this section revolves around the sensitization of the general public on the magnitude and effects of land grabbing and resultant land dispossession among rural communities. It is argued that such a strategy could build on efforts by groups and individuals both inside and outside the public sector. It is quite evident that, within the public sector, as much as there are those who are obviously content with the status quo, there are officers and activists who are committed to social justice and as such opposed to statesponsored land grabbing. Traces of such opportunities, though scanty, are nevertheless evident in decisions made by some district authorities.
The fencing of communal grazing areas ushered in by the NDPA has prompted some districts and actors within the public sector to argue fairly strongly and successfully for the reservation of some areas for the rural communities, in particular the San. In the Boteti sub-district, for example, it has been suggested that the area known as 4B (between Orapa and Mokoboxane) be reserved for remote area communities in the area. In 1995, a study of the Kaka area by the Central District Technical Committee (on the Fencing Component of the NPAD) demonstrates the desire to guard against land dispossessions among the rural poor. The study followed a decision to use Kaka as a pilot area for the fencing of communal rangelands. Extensive socioeconomic studies have revealed the presence of San communities in the area. Most of these people were settled around boreholes and were regarded by the borehole owners as squatters. The Technical Committee recommended that at least some of the ranches should be reserved for the local San communities. On the basis of these recommendations, approximately five ranches (each measuring 6 km x 6 km) were reserved for remote area communities. While it is evident that these efforts are not adequate, they nevertheless present a conducive environment for mobilising support for curbing land grabbing in Botswana.
The preparation of the CKGR management plan also provides an example that researchers and activists can adopt in a bid to draw Government into dialogue with representatives of rural communities. One of the aims of the European Commission-funded project (entitled Wildlife Conservation and Utilisation in Central and Southern Botswana) was the production and implementation of a management plan for the CKGR and Khutse Game Reserve. Although the original plan was later abandoned, the concept that initially informed the preparation of the plan is of interest to the argument presented in this instance. The preparation of the plan involved extensive consultations and dialogue with local communities. Central to the plan was the belief in the co-existence of people and conservation of wildlife. The exercise shows possibilities for constructive engagement between Government and other stakeholders. Instead of relocation and exclusion, the plan sought to promote accommodation and co-existence. Events that led to the abandonment of the proposals contained in the Third Draft also provide useful lessons to those going into dialogue with the Government.

CONCLUSION
This article attempted to retell the story of land grabbing in Botswana. Despite the developmentalist language, in which the country's policy processes are couched, the fact remains that there is relentless land dispossession among the rural communities. Apart from documenting the magnitude and nature of land grabbing, the article also attempted to point towards opportunities found within the existing legislation and administrative processes which, it is argued, provide entry points for transformative agendas that could curb land grabbing sponsored by the country's land-tenure and policy reforms. As shown in the article, these opportunities have successfully been utilised by land activists among San communities in Botswana. Through constant engagement, both researchers and activists can begin to expose the darker side of land-reform policies in Botswana. Currently, in a bid to expand some of the urban areas, the Government purchases freehold land which is then changed into state land and allocated for various urban land uses. Still, in the case of some districts such as the North East, freehold land has been purchased and converted into tribal land. Activists working with local community groups could possibly begin to explore this path more rigorously.