The limitations of state regulation of land delivery processes in Gaborone, Botswana (original) (raw)

The limitations of state regulation of land delivery processes in Gaborone, Botswana

Abstract

This paper assesses the strengths and weaknesses of a state-led land delivery process in Gaborone, Botswana, in particular the extent to which such a process enables the poor to access land with secure tenure. The paper observes that, despite its efforts, the government has been unable to supply sufficient urban land to satisfy demand, largely because of three interrelated factors: (i) the changing and speculative nature of demands made by middle- and high-income beneficiaries; (ii) the evolution of a ‘culture of entitlement’; and (iii) government reluctance to address the land and housing needs of the poor. Consequently, while the unattainable land demands made by the rich have resulted in a sprawling albeit wellplanned city, the poor have utilised their collective agency to stake their claims by ‘illegally’ occupying, first, state land within the town and, later, customary land in peri-urban areas. These contradictions and contestations have, in the long run, forced the government to rethink, revise and rewrite its policies and approaches to urban land supply and development processes, although poor people continue to lose out in the struggles over policy, state land allocation and increasingly commercialised processes of informal peri-urban land subdivision.

Botswana is one of the few countries in sub-Saharan Africa in which the government has steadfastly attempted to provide adequate land for its burgeoning urban population. Enabled by windfall revenues from mineral resources, the Botswana government has sought to develop orderly, well-planned and adequately serviced towns and cities free from shacks, slums and ‘squatter’ houses. Although the state has generally achieved its objectives, this success has been attained at the expense of the poor, who have been excluded from mainstream urban development and housing processes. This paper is based on qualitative and quantitative data collected during a study of informal land delivery processes and access to land for the poor in Greater Gaborone, undertaken by Kalabamu and Morolong (2004). 1{ }^{1} The aims of the study were, among other things, to assess the strengths and weaknesses of formal and informal land delivery mechanisms and how each of these enables the poor and other vulnerable groups to access land with secure tenure. The poor are here defined as individuals and households who, in comparison to the rest of Botswana’s population, are unable to afford facilities and services offered by either the private or the public sector. They include the destitute

[1]

1. Faustin Tirwirukwa Kalabamu is senior lecturer in the Department of Architecture and Planning at the University of Botswana, Private Bag ozz, Gaborone, Botswana; email: kalabamu@mopipi.ub.bw or kalabamu@hotmail.com
   1 The research was undertaken as part of a comparative study funded by the UK Department for International Development. DFID supports policies, programmes and projects to promote international development. It provided funds for this study as part of that objective but the views and opinions expressed are those of the author alone. ↩︎

and individuals who undertake irregular casual work (e.g. car washing, scavenging and sex work), as well as domestic workers and labourers earning the stipulated minimum wage or less.

The paper is divided into four main parts. This introduction is followed by a quick review of ‘land control’ as a source of economic and political power. The next section presents a historical perspective on land tenure and administrative institutions in Botswana, as a context for understanding the strengths and weaknesses of postcolonial urban land delivery mechanisms. This is followed by an examination and assessment of land delivery processes in Greater Gaborone.

Land control, power and exclusion

As its point of departure this paper goes beyond the definition of land tenure as a ‘bundle of rights’ over a piece of land enjoyed by an individual or group to examine the power relationships between land administrators on the one hand and land users on the other. Land administrators are here defined to include individuals, agencies and institutions that manage and make day-to-day decisions on access, allocation, use and transferability of publicly or privately owned land. Land users are then defined as households, men and women who occupy, use or work the land for their everyday needs and survival. Thus land tenure is viewed as the written and unwritten rules, norms and practices that govern and regulate the relationships between land administrators and land users. As argued by Rakodi and Leduka (2003), in a power relationship, the dominant actors (e.g. land administrators) have power because the dominated actors (e.g. land users) accept the situation as legitimate, although the latter also command resources that may enable them to influence the actions of the dominant actors. Land tenure rules, norms and practices prevailing at a given moment in time and in a particular society, therefore, reflect the outcomes of historical claims, counterclaims, conflicts, contestations, negotiations, exchange and appropriations, while present acts help to define and legitimate future claims and relationships (Berry, 200I). Changes in political, cultural and economic dispensations provide opportunities for renegotiating rules and redefining relationships between land administrators and users.

Powelson notes that, throughout the history of humankind, different ‘people were allowed to do different things on the same land, and for the privilege of feeding themselves they performed obligations to superiors who held proprietary rights’ (Powelson, 1987, III). However, in the feudal system in Europe, for example, peasants and serfs could gain better land rights by allying with one rather than another ruler, thereby levering their power upwards. Although the supported ruler often pleaded concern for the landless, ‘invariably he was himself the principal beneficiary’ of such power struggles (Powelson, 1987, 115). Berry likewise observes that struggles 'over land

[in sub-Saharan Africa] have been as much about power and the control of people as about access to land as a factor of production’ (Berry, 200I, xix). As in the rest of the world, she argues that long before colonisation, land allocation and administration in Africa served as an important avenue of political competition. Although chiefs in sub-Saharan Africa did not receive tribute for administering tribal or customary land, a chief’s ability to avail and guarantee adequate land to his subjects translated into support against his rivals: land control connected chiefs to their male warriors. Berry (200I) further contends that, during the colonial era, the value of land to European settlers also lay in their ability to exercise power over African labour and African societies by controlling access to land. Consequently, European settlers amassed more land unto themselves than they could effectively use on their own.

Powelson (1987), Berry (200I) and Daley and Hobley (2005), among others, note that although post-colonial land reforms ostensibly sought to redress the land inequalities caused by colonial regimes, these reforms have in practice been utilised to consolidate the political power of new elites: ‘land reform and its encumbrances became instruments of political control: peasants had to support the government or they would lose the sources of credit and inputs and have nowhere to sell their products’ (Powelson, 1987, II8). As power relations shifted, ‘land distribution was seen as part and parcel of the broader post-colonial redistribution of power’ (Daley and Hobley, 2005, 9) to the extent that several governments in sub-Saharan Africa (including Tanzania, Mozambique and Ethiopia) ‘nationalised’ all the land within their countries.

Buttressed and sometimes ‘forced’ by the World Bank and other international agencies, many governments in sub-Saharan Africa (e.g. Kenya and Malawi) have opted for registration and titling of individualised land rights, which in effect secures the rights of one person to the exclusion of other family members (Daley and Hobley, 2005). Whatever approach has been adopted, post-colonial land reforms in sub-Saharan Africa do not appear to have benefited the majority of land users. In the meantime, politicians and state bureaucrats have often been accused of corrupt practices and turning publicly owned land into personal estates (Powelson, 1987). Men and women disadvantaged or excluded by post-colonial land reforms have reacted by, inter alia, negotiating and devising new rules and relationships for accessing and utilising land outside the official structures. These are, in the eyes of the state, ‘informal’ or ‘illegal’. Although they are detested by the powers that be, they have delivered much of the land needed for housing in urban areas and are recognised by the actors involved (Rakodi, 2000).

The increase in ‘informal’ or ‘illegal’ land transactions in cities and their surrounding settlements may partly be attributed to the ‘policy ambiguity, procedural complexity and prohibitive costs involved in obtaining documents (titles) which legalise ownership of urban land …’ (Kombe and Kreibich, 2000: 6), and partly to the economic opportunities in peri-urban areas (e.g. poultry farming and rental housing) (Mbiba and

Huchzermeyer, 2002). Leduka (2000; 2004), Kombe and Kreibich (2000) and Nkurunziza (2004) indicate that ‘informal’ land transactions draw from both customary norms and statutory practices, while the actors rely on traditional leaders as well as state agencies, suggesting that power relationships in land matters are not only complex but also multi-layered. The dependence on both customary and statutory provisions facilitates the manipulation of events and reinterpretation of the rules by various actors at each level so as to obtain power, land and influence (Fourie, 2004).

Highlighting power relationships, the remainder of this paper will explore the rules and processes for accessing, owning and disposing of land in Botswana and Greater Gaborone in particular. It will also examine the opportunities and constraints posed by some of the land reforms introduced by both colonial and post-colonial governments.

Botswana: colonialism, land tenure and urbanisation

Although Botswana is a large country ( 582,000 km2582,000 \mathrm{~km}^{2} ) with a small population ( 1.7 million inhabitants in 2001), it experiences acute land shortages because more than two thirds of its land is desert. Consequently, more than 80 per cent of the country’s population, settlements and economic activities are concentrated along a 20okm-wide strip on its eastern border. This strip, which is endowed with better soils and surface water, and also receives more reliable rainfall than the Kalahari Desert, has been occupied by agricultural-pastoralists for over I,500 years (Campbell, 1982). These farmers ‘lived in large settlements of mud and pole houses’ (Ngcongco, 1982, 23). However, due to droughts, famines, pestilence and dynastic squabbles and rivalries among their leaders, as well as searches for greener pastures and better living conditions, the settlements split and shifted quite often (Ngcongco, 1982). Since the late nineteenth-century colonisation of Africa, the settlement patterns and livelihoods of people living in Botswana have been altered irreparably.

Colonisation and administrative changes

After much hesitation and reluctance, Britain declared Botswana a protectorate in 1885. However, it ‘pursued a policy of indirect rule that involved minimal interference in the internal governance and customary law’ (Adams et al., 2003, 55). Despite the declared ‘indirect rule policy’, the British did in practice introduce new administrative, judicial, economic and settlement patterns, all of which had profound effects on land rights and tenure systems. ‘Indirect rule’ initially meant two parallel systems one that regulated the affairs of the indigenous population and the other that applied largely to European settlers. However, over time the colonial administration introduced regulations that sought to restrain and control the powers of chiefs and the

settlement patterns and economic activities of the indigenous populations. In order to finance the costs of administering the protectorate, as well as to integrate men into the capitalist cash economy, the colonial government levied taxes on all adult men. To pay the taxes, cattle owners sold some of their cattle, while those who did not own cattle sought paid employment in South African mines, farms and towns, since wage employment was non-existent in Botswana (Colclough and McCarthy, 1980; Good, 1992). The British colonial administration further outlawed the traditional practice whereby villages shifted regularly in search of better soils, water sources and pastures (Schapera, 1994) and established several townships: Francistown, Lobatse, Kasane and Ghanzi. Additional far-reaching effects emanated from the various land tenure systems introduced during the colonial and post-colonial eras.

Changing land tenure systems

Through a number of proclamations, orders and concessions, the colonial administration divided Botswana’s land into three categories - native reserve land or tribal territories, crown land and freehold land. As a result, Botswana currently has three main categories of land tenure: tribal or customary land, state land (formerly crown land), and freehold or privately owned land. Most of the fertile areas along the country’s eastern border are freehold land, including farms acquired by European settlers through concessions made by chiefs (Colclough and McCarthy, 1980) and the Ghanzi Block, which was allocated to a column of Boers to act as a buffer against German expansion from South West Africa (now Namibia) (Ng’ong’ola, 1997). While freehold and crown land were framed around Anglocentric conceptions of land rights and ownership, ‘tribal land’ tenure systems were initially based on unwritten indigenous norms and practices that have been partly ‘modernised’ and written down by the post-colonial government.

Tribal/customary land

The most outstanding feature of the customary tenure systems in their original formulation was the free right of avail - a right that was uniformly applied to all adult men and automatically shared by all men belonging to each tribe (Schapera, 1943; Jeppe, 1980; Kalabamu, 2000). It was from this right of avail that other rights - individual or communal - were deduced. Male siblings had equal rights to be allocated land from their fathers’ holdings or the tribal reserve without having to make any payment or tribute to the chief or headman. Allegiance to the chief was the only requirement for acquiring and maintaining land rights. Women accessed land only through their husbands or male relatives in their fathers’ lineage.

All allocated or inherited land remained the exclusive property of the concerned man and his family. Land rights could only be extinguished if the holder ceased to

belong to the tribe. However, in times of need (e.g. congestion or shifting of a village), land rights could be reallocated as part of the ward or village re-organisation process. Contrary to common Eurocentric misconceptions at the time of colonisation, chiefs and headmen did not own tribal land - they only administered its use, allocation and transfer. Ownership of tribal land was vested in the entire tribe, while chiefs acted as trustees or custodians (Schapera, 1994).

Within two years of attaining independence, the government passed the Tribal Land Act (1968) which, among other things, transferred the administrative power to allocate land rights from chiefs and headmen to ‘land boards’ established under the same act. Chiefs and headmen were at that time increasingly accused of favouritism and misuse of their powers to administer and allocate land. To the government (GOB, 1970; 1983a) and Mathuba (1989), the Tribal Land Act was passed in order to (i) provide a written law that could easily be referred to; (ii) improve livestock production; and (iii) provide people with land rights that are both saleable and bankable. However, to Ng’ong’ola (1997) the primary aim of this Act was to provide the government with greater political control over land administration.

The Tribal Land Act (1968) was extensively amended in 1993 in order to remove ambiguities and contradictions identified during its implementation (Kalabamu, 2000). The major ambiguity related to sections Io(I) and Io(2). While section Io(I) appeared to vest ‘ownership’ of all tribal land in the relevant land board, section Io(2) appeared to vest the ownership of some pieces of land or rights to water in individuals in their personal and private capacities. The ambiguity led to contradictory interpretations and numerous court cases (Kalabamu and Morolong, 2004; Kalabamu, 2000; and Ng’ong’ola, 1993). In 1993 Section 10(2) was repealed and section Io(I) revised to provide that land in each tribal area belongs and is accessible to all citizens of Botswana. The latter amendment has led to increased demand for land in the periurban areas surrounding Gaborone and other urban centres, which are under the administration of land boards.

Section 2.4.I of the Tribal Land Act empowers land boards to repossess land for the purpose of redistribution, involving the cancellation of customary rights and rezoning of agricultural land for residential, commercial or industrial uses. While the government has tended to pay handsomely for freehold land acquired by the state, the compensation offered by land boards for repossessed tribal land has been inadequate (GOB, 1992a; Kalabamu, 2000; Natural Resource Services, 2003). Land boards have argued that, since tribal land is ‘free’, it is impossible to quantify, in monetary terms, loss of rights to use a particular piece of land beyond the unexhausted improvements on it (e.g. standing crops, boreholes, fences, buildings, ploughing). Thus, to land boards, compensation does not need to reflect the development value of land, even in peri-urban areas.

Crown/state land

The colonial administration considered the Kalahari Desert unutilised, declared it crown land and placed it under the authority of the British monarchy. The High Commissioner was given powers to make grants or leases on crown lands on the terms and conditions he deemed fit. The State Land Act (1966) transferred those powers to the President or his nominee. At present state land is leased to individuals and private companies under Fixed Period State Grants (FPSG) or Certificates of Rights (COR) (Kalabamu, 2000). The FPSG is a 50-99 years capitalised lease, at the end of which the land, together with all improvements thereon, reverts to the state without compensation. The leaseholder may sell or transfer to any person what remains of his or her grant period (Dickson, 1990). Introduced in the 1970s to meet the needs of the urban poor, the COR is a lease system which confers on the holder perpetual usufruct rights for the purpose of erecting an owner-occupied house. It is unregistered title that provides holders with secure tenure while avoiding high cadastral costs. COR plots are provided with basic services (e.g. gravel roads, communal water taps, pit latrines).

Contemporary land tenure systems

The post-colonial government has maintained the three land tenure systems it inherited from the colonial administration. However, as noted earlier, it has progressively changed the rules and procedures on access, use, transfers and entitlements on state and tribal land, as well as placing a moratorium on the creation of new freehold land from state land. In addition, the government has converted large portions of state land into tribal land and some freehold farms into state land. As a result, the proportion of tribal land increased from about 49 per cent in 1966 to 71 per cent by 1998, while state land decreased from 47 per cent to 25 per cent, as shown in Table I. At 4 per cent, Botswana has one of the lowest proportions of freehold land in the region. In 1999 freehold/privately owned land accounted for 72 per cent of the total land mass in South Africa, 44 per cent in Namibia and 38 per cent in Swaziland (UNDPA, 2003, 13).

Any adult citizen of Botswana is, in theory, eligible to acquire an unlimited number of pieces of land for residential, commercial, agricultural or industrial use on tribal

Table 1: Land tenure categories in Botswana, 1966-1998

Source: Adams et al. (2003,56)(2003,56)

land anywhere in the country and one residential plot on state land in each township. However, married couples are eligible for only one residential property on state land in each township, although there are no limitations on the number of properties they may own on freehold land (GOB, 1990). As will be discussed later, these rules and entitlements have tended to increase the demand for urban and peri-urban land.

Rapid urbanisation and socio-economic development

Although at the time of independence in 1966 Botswana was a predominantly rural society, it has since urbanised rapidly. The population living in urban areas 2{ }^{2} rose from about 4 per cent in 1964 to 54 per cent by 2001 (GOB, 2003). Gaborone and, to a lesser extent, Francistown have been the centres of the rapid urbanisation process, which may be attributed to the discovery and exploitation of mineral resources. Mineral resources have also led to the transformation of the country’s economy. Between 1966 and 2001, the contribution of agriculture to the GDP fell from 43 per cent to 3 per cent, while that of the mineral sector increased from zero to 37 per cent (GOB, 2003). The GDP per capita rose (at constant market prices) from Pula 1,683 in 1966 to Pula 9,793 in 2001 (Pula 5.00=5.00= US$ i.00). This economic transformation has translated into people migrating from rural areas and entering the urban labour market, which has increased demand for urban land and housing.

State-led land delivery in Greater Gaborone

At the time of its designation as the capital of independent Botswana in the early ig6os, all that existed in what is now Gaborone was a few government offices and houses, a church, railway station and a few shops (Dickson, 1990). The settlement has grown into a commercial, industrial and financial centre and the administrative headquarters of the Botswana government and the Southern Africa Development Community (SADC). Gaborone’s population increased from 3,855 in 1964 to 185,891 in 2001. Although the populations of the surrounding villages of Tlokweng, Mogoditshane and Gabane initially declined, they started to increase rapidly in the 1970 (Table 2). The fastest growing village has been Mogoditshane, which since 1971 has consistently recorded more than twice the national population growth rate. Population increases in Gaborone and its peripheral settlements have had a profound influence on the demand for land and housing.

Land delivery in Greater Gaborone (and Botswana in general) is dominated, in theory and practice, by state structures and institutions. The government owns and

[1]

1. 2 Defined as settlements with at least 5,000 inhabitants and in which over 75%75 \% of the labour force is engaged in non-agricultural activities. ↩︎

Table 2 Population growth in Gaborone and surrounding settlements

Figures in parentheses indicate annual growth rates from the previous census year.
Source: GOB (1972; 1983c; 1992c; 2002)
controls about 95 per cent of the land within Gaborone Township and, through land boards, indirectly administers and controls all of the tribal land in the surrounding settlements. It has achieved this control through purchases of freehold land on the one hand and amendments of the Tribal Land Act on the other.

Supply of and demand for state land plots

The post-colonial government inherited and retained the responsibility for supplying land for urban development and building houses for its employees. During the 1960s, the government supplied about 2,800 surveyed and serviced plots, most of which were developed into houses for senior civil servants. Only 35 per cent of the plots were designated for low-income houses, of which 515 were built by the government and the rest by individuals on a self-help basis.

The government, like its predecessor, expected most labourers and casual workers to reside in the surrounding villages and commute to work in Gaborone. Contrary to these expectations, many labourers chose to build houses for themselves on a piece of state land within the township they nicknamed Naledi (‘star’). The govern-

Figure 1: Gaborone: residential suburbs
Source: Author
ment reacted to the emergence of Naledi by servicing about 700 residential plots at Bontfeng, to which it expected the ‘squatters’ to relocate and build ‘traditional’ houses for themselves. It also built several hundred mass-produced houses at New Naledi, with the hope of demolishing Old Naledi after relocating its residents. However, the ‘squatters’ who were resettled in New Naledi quickly returned to their original plots

in Old Naledi (van Nostrand, 1982). The government adopted a ‘squatter’ upgrading policy in the early i97os following the receipt of advice from the United Nations Economic Commission for Africa and after realising that government departments and agencies (e.g. the Botswana Housing Corporation) were unable to offer a range of houses affordable by different income groups (GOB, 1983b). Old Naledi was then recognised as a legal settlement and its residents granted secure land titles.

Having abandoned its ‘squatter’ demolition policies, the government undertook to ‘encourage the building of new urban housing for all income levels at a pace which will ensure that no citizen of an urban area is forced to reside in an unauthorised settlement’ (GOB, 1982, 3). This undertaking motivated it to purchase freehold farms for the development of Gaborone (Figure I). Initially, the government purchased Broadhurst and Tsholofelo farms and expanded the township northwards (Dickson, 1990). These developments provided about 6,000 plots for low-, medium- and highincome housing, industries and shopping facilities and doubled the physical size of the town by the mid-I98os. About 70 per cent of all plots in Broadhurst and Tsholofelo were reserved for self-help housing (Dickson, 1990). Faced with further demand for land, the government purchased three more freehold farms (Bonington, Glen Valley and part of Content Farm) and expanded the township farther to the north and west. Glen Valley was put to military use, Content Farm became an agricultural college and research centre, and Bonnington Farm, which measured about 3,000 hectares, was subdivided into residential neighbourhoods referred to as Gaborone West Phases I-IV (Figure I). The four phases generated about 15,000 mixed-income residential plots.

Despite these efforts, the demand for more land and houses was unsatisfied. For example, in 1987 the Botswana Housing Corporation had a total waiting list of 20,000 applicants for its houses in Gaborone. By 1990, the waiting list had increased to 22,000 (MLGL, 1990). To provide more land, the government launched the Accelerated Land Servicing Programme (ALSP) in the late i98os and supplied over io,210 plots in Gaborone within a period of less than seven years. All ALSP plots are title surveyed and fully serviced with tarred roads and provisions for private water and sewer connections but lack provision for communal stand pipes and do not permit pit latrines.

Parallel to the ALSP initiative, the private sector, which had hitherto been prohibited from participating in the formal land delivery process, was able to deliver about I,500 serviced plots. By 1997 Gaborone Township had a total of 33,343 serviced residential plots (including regularised plots in Old Naledi), of which 23,294 (almost 70 per cent) were fully or partially developed; and io,049 were yet to be allocated or had been allocated to people who had not fully paid for them (DTRP, 1998). Yet by January 2001, the waiting list for plots stood at over 23,500 (MLH, 200I). Thus despite the ALSP, demand for residential plots has persistently outstripped supply.

Accessing state land plots

Land in Gaborone has historically been allocated for free or ‘sold’ to those who satisfy various criteria set by the state. Until the early 1970s, land and housing were allocated to civil servants (mostly senior and a few junior). Private-sector employees and labourers were expected to fend for themselves or reside in surrounding rural settlements. Consequently, they decided to ‘squat’ on state land at Old Naledi, which made the plight of the poor visible to the state and forced the government to revisit its land and housing polices. During the upgrading of Old Naledi, both ‘squatters’ and tenants of rooms were allocated land. House owners were issued with title to whatever land they occupied, while tenants and those owners whose houses were demolished to make way for roads were allocated land in Broadhurst and Tsholofelo.

To reach out to sections of the low-income group, the government introduced the Self-Help Housing Agency (SHHA) scheme in the mid-r97os. This scheme was responsible for (i) upgrading ‘unauthorised’ settlements in various cities, including Old Naledi in Gaborone; and (ii) allocating, at no cost to the beneficiary, minimally serviced land to the ‘poor’. To qualify for land under this scheme, the applicant had to be a citizen of Botswana aged 2I years and above with a verifiable, regular income. Furthermore, the applicant was required to have lived in the respective township for at least six months prior to submitting his or her application. These criteria discriminated against new arrivals (rural-urban migrants), casual and self-employed workers, and youths who were parents and workers but aged under 2I. Although the criteria were ‘gender neutral’, in practice they tended to discriminate against women, who constituted the majority of teenage parents as well as casual and self-employed workers (Kalabamu, 1998; Musyoki, 1998).

With the introduction of the ALSP, income has become the most important criterion for allocating state land. Under this programme, land is allocated at ‘affordable cost’ to low-income earners, at ‘cost recovery’ to middle-income earners, and at ‘market prices’ to high-income earners. ‘Affordable cost’ is defined as being no more than 25 per cent of total earnings for a low-income household, while ‘cost recovery’ is defined as the price that will enable the government to recoup all of the cost of servicing the land (GOB, 1985; MLGL, 1990). The minimum age limit was reduced to 18. In addition, applicants must be formally employed or legitimately self-employed. The applicant’s name is retained on the waiting list if he or she is transferred to a workplace outside Gaborone. Only one application per couple is entertained. An individual who marries a person who already owns a plot on state land is required to withdraw his or her name from the list.

The above criteria tend to disadvantage some. First, they exclude people who are unemployed but capable of building houses using their own labour and funds from the sale of other household resources such as cattle and rural produce. Secondly, they exclude people (notably women) with partners working outside Gaborone or

the country (e.g. in South Africa). Thirdly, the ‘affordable price’ excludes the most needy and vulnerable households. Fourthly, by restricting land holdings to one plot per couple, married women are disadvantaged because traditional norms and practices dictate that land belongs to husbands, who are assumed to be household heads. Fifthly, given that more women than men are unemployed or work as housewives, the criteria tend to disadvantage more women than men.

Delivery of plots on tribal land

Prior to 1970, chiefs and headmen were responsible for allocating tribal land in Mogoditshane, Tlokweng and other peri-urban villages. An applicant, together with the headman, would identify the piece of land to be allocated. If there were no objections from neighbours or any person within the ward, then the headman would formally allocate the plot to the applicant. This procedure was retained by the land boards, which replaced the chiefs as land administrators in 1970. The land boards relied on headmen for identification of vacant residential sites. When headmen faced increased demand in the late 1970s, they started to allocate residential plots in areas previously reserved for cultivation (Sithole, 1995). The demand came from people who worked in Gaborone but were forced to live in surrounding villages (especially Mogoditshane) because it was easier, cheaper, quicker and more certain to obtain land in these villages than in Gaborone (Kweneng District Administrator, 1982). The upsurge in demand for land in peri-urban villages may also be attributed to the upgrading and formalisation of Old Naledi and the strict ‘anti-squatting’ policies pursued by the government in Gaborone Township.

Concerned about rapid sprawl in Mogoditshane, Kweneng Land Board directed headmen and land applicants not to identify plots beyond a certain limit defined by the physical planners. The area beyond the limiting line was made subject to ‘modern’ town planning procedures whereby plots would be demarcated before allocation. However, attempts to demarcate plots in the designated areas were abandoned following objections from field owners. The latter denied that they had been consulted on the land board’s intention to repossess their fields and queried the inadequate compensation offered by the board (Sithole, 1995). Subsequently, dissatisfied field owners (including ward headmen) subdivided their fields into residential plots and sold them without consulting the board. Field owners, however, expected that the board would later formalise their allocations, as had been the case with plots allocated by ward headmen (Sithole, 1995).

The situation worsened with the passage of time, forcing the government to set up a Presidential Commission of Inquiry into land problems in Mogoditshane and other peri-urban settlements in 1991. The commission’s findings confirmed, inter alia, the mushrooming of an ‘informal land market’ in these areas (GOB, 1992a, b). It

noted that arable land was being subdivided and sold for residential use without the authority of either the District Administration or the responsible land board. Furthermore, some sellers had issued fictitious ‘certificates of customary land grants’ stamped with fictitious names such as ‘Mogoditshane Land Board’ and had asked buyers to develop the plots as soon as possible. Cabinet Ministers and high-ranking government officials were among the people who had ‘illegally’ acquired land in Mogoditshane.

Following the commission’s report, the government decided that all citizens who had ‘illegally’ acquired and developed land in Mogoditshane and other peri-urban areas should pay a penalty proportionate to the size of plot occupied (GOB, 1992b). One of the ministers implicated was forced to resign. In addition, a subordinate land board for Mogoditshane was established and, as in Old Naledi, the land rights of the occupiers who paid the penalty fee were regularised. However, the properties of those who failed to pay the penalty and subsequent ‘squatters’ were demolished between 2000 and 2002 by the land board with the support of the President and under the supervision of the armed forces. It is estimated that over 500 houses were demolished. In the meantime, between 1992 and 2000 the land board received over 35,000 applications but succeeded in demarcating only 3,570 plots in Mogoditshane.

Limitations of state-led land delivery mechanisms

The duality of ‘formal’ and ‘informal’ land delivery processes in Gaborone and surrounding settlements highlights various power struggles over access to and control of land. As in the past, when chiefs and colonial administrators determined where people should reside and work, the post-colonial government has been anxious to regulate where urban residents live. This has been contested by disadvantaged and sidelined people, and even some beneficiaries of formal policies appear to be dissatisfied. Clearly state-led land delivery mechanisms have a number of limitations not only for those excluded from urban land ownership but also for the government itself, in its quest to ensure orderly urban development. The main limitations and some reactions to them are identified below.

A well-planned city or urban sprawl?

The government has been able to create and maintain Gaborone as a city with relatively well-planned and serviced residential suburbs. The land delivery process is also relatively orderly, with the exception of property demolitions witnessed during the last few years. Invasion and squatting appear to be rare. Of the 282 plot owners surveyed in 2002/3 (Table 3) 42 per cent reported that they had obtained their plots through allocation by a public sector body (SHHA or the land board), followed by

Table 3: Method of plot acquisition in Mogoditshane and Old Naledi, Greater Gaborone (2003)

Source: Field Surveys, December 2002
purchase ( 18 per cent), allocation by a chief ( 16 per cent) or inheritance ( 16 per cent). 3{ }^{3} Only 6 per cent reported that they had acquired land ‘illegally’. However, people who had initially acquired land ‘illegally’ but later formalised their rights considered themselves to have been ‘allocated’ the land by the responsible authority.

Nevertheless, the government’s efforts to satisfy the apparent demand for land, coupled with the sale of fields in tribal land areas, have resulted in urban sprawl. As shown in Table 4 and Figure 2, Gaborone and its surrounding settlements have encroached on formerly agricultural land. The built-up area of Gaborone Township has, on average, doubled every ten or so years. However, Mogoditshane has experienced the most dramatic lateral growth, especially since the mid-1970s. The growth of Tlokweng, Gabane and Mmopane was initially less dramatic but appears to have accelerated after the mid-198os when land transactions in Mogoditshane started to draw the attention of central government land administrators. Sprawl results not only from high demand for residential land but also the prevalent patterns of development. Gaborone is characterised by low-rise single-storey detached houses with low plot coverage rates. The average plot sizes specified for high-, medium- and low-density housing are 300−375 m2,375−400 m2300-375 \mathrm{~m}^{2}, 375-400 \mathrm{~m}^{2} and 400−1000 m2400-1000 \mathrm{~m}^{2} respectively. Moreover, in both low-income high-density and high-income low-density areas maximum plot coverage

3 A survey of plot owners in the peri-urban area of Mogoditshane was carried out in 2002/3. 123 of 1,617 plots in Old and 128 of 1,909 in New Mogoditshane were sampled, yielding 97 and 94 completed questionnaires respectively. 19 per cent of plots were undeveloped or occupied only by tenants. The survey was complemented by semi-structured interviews with key informants and land owners, two focus group discussions in each area, one with adults and one with young people. A similar survey of plot owners in the regularised ‘squatter’ area of Old Naledi was carried out at the same time, as well as FGDs with young women and adults (Kalabamu and Morolong, 2004).

Figure 2 The spatial growth of Gaborone and its surrounding settlements (1966-1999)

Table 4 The growth of built-up areas in Greater Gaborone (1964-99) (hectares)

Source: Calculated from topographical maps supplied by the Government of Botswana, Department of Surveys and Mapping
has traditionally been set at 50 per cent. The demonstration effect of these formal standards and the apparent availability of land have encouraged similar densities in peripheral settlements.

Urban sprawl has increased servicing and infrastructural maintenance costs for both central and local governments and city inhabitants. Although Gaborone (with about 200,000 inhabitants) is by international standards a small city, it exhibits the characteristics of a larger city in terms of traffic jams, travelling and commuting distances and problems related to infrastructure networks.

Speculative demand for land

Despite the government’s undertaking that it would supply sufficient land and housing it has persistently failed to satisfy demand. This may be attributed to many factors, of which the ‘culture of entitlement’ and speculation are the most important. The culture of entitlement is connected to both the ‘customary right of avail’ and the post-I98o policies that guaranteed to provide land or a house to whosoever placed his or her name on the government’s waiting list. To try their luck, many people register their own names and those of their spouses and children on waiting lists in several locations. This game of chance is fuelled by fears that the government may eventually abolish the entitlement or run out of land to distribute.

As noted earlier, many people allocated plots do not immediately build houses on them - by 2001, 48 per cent of the plots serviced and allocated under the ALSP were undeveloped and/or not paid for (MLH, 2001, 15-16). It appears that this speculative demand is exacerbated by the ‘culture of entitlement’, the subsidised prices at which public land is ‘sold’ to the rich and poor alike, and the allocation of land at no cost to beneficiaries in tribal areas. The ‘demand’ for land in Gaborone has not been satis-

fied because a substantial proportion of plots are allocated to people who have no immediate desire to develop them.

Exclusion amidst plenty

Despite the large volume of serviced land produced by the government, the allocation criteria have persistently excluded the ‘very poor’ - that is, casual workers, the unemployed and people earning below the minimum wage. They are excluded on the assumption that they are incapable of building ‘houses’ that comply with urban planning standards and regulations. Theoretically land is available for ‘free’ in periurban settlements (e.g. Tlokweng and Mogoditshane). However, according to focus group informants, land in these areas has become inaccessible to the poor due to the long waiting lists and ‘bribes’ demanded by land board officials. Claims of bribery, nepotism and corruption among such officials were aired in all the focus group discussions held in Old Naledi and Mogoditshane: ‘we are told by SHHA office that the available spaces are reserved only for businesses, churches, recreational places and others excluding residential. But we see councillors and rich people being allocated such plots…’ said one youth in Old Naledi. Participants in Mogoditshane claimed that, ‘The Mogoditshane land board believes in bribery… They [land board officials] tell us that if you give me money I will allocate you a plot quickly.’

Indeed our study revealed that poor people have not been able to acquire plots in Greater Gaborone. Three quarters of the 285 plot owners interviewed in Old Naledi and Mogoditshane had acquired their residential plots when the head was working. In New Mogoditshane 79 per cent of the respondents had been working at the time of plot acquisition, compared to 74 per cent and 70 per cent in Old Naledi and Old Mogoditshane respectively. Of the employed household heads, 83 per cent had been in full-time wage employment, 6 per cent in part-time wage employment, 7 per cent had casual jobs and only 4 per cent were self-employed.

Youths were particularly concerned about the present land delivery processes and criteria. They feel ‘discouraged to apply for land because those who applied many years ago are still waiting’. ‘We, as the youth, are left out in land allocation… even if we apply our applications are never considered… if I apply now for a plot I will wait for about five or twelve years without any reply and if you ask you are told to go and wait, as they will write you a letter’, said one youth in Mogoditshane. A young woman in Old Naledi noted that young people are further disadvantaged by their employment status: ‘We are not employed so we think of how we will develop those plots in case we get them’, she said.

In Mogoditshane, youths were opposed to land board repossession of arable land:
The land boards are very unfair. If in the family there are five or more children they deny us our rights to inherit our parents’ property. This is because when they take our

Figure 3 Several houses containing rental rooms on one plot
Source: Author’s photo, 2003
parents’ fields for allocating to people as residential plots, they say that in the family they only allocate plots to two children or dependants of the field owner. This discriminates against other children… We end up looking for plots and renting accommodation although our father kept land for us…

To redress their disadvantaged position, youths made two proposals. First, they want to be permitted to inherit their parents’ position on the waiting list. Secondly, they want to be given preferential treatment instead of being ‘made to wait in a queue with people who already have plots’. A few youths suggested that land boards should be disbanded and the duty of administering tribal land be reassigned to chiefs and headmen.

Unable to acquire land and build their own houses, the majority of the poor (especially the young) satisfy their shelter needs through ‘rooming’ - that is, several households renting one or two rooms in one house or homestead. According to the 1991 census, 44 per cent of all households living in rented accommodation were ‘rooming’. Our study revealed that the older the area, the larger the proportion of plot owners renting rooms to tenants: 72 per cent in Old Naledi, 52 per cent in Old Mogoditshane and 42 per cent in New Mogoditshane. Instead of subdividing their plots, owners erect rooms for rent, sometimes several dwellings consisting of two or more rooms, each with an external access door (see Figure 3). Some plots may contain up to six houses separated from each other by narrow passages 1.5 to 2 metres wide. All the dwellings on a plot share facilities such as a toilet and tap. On average 2.7 households were living on the plots surveyed in our study, while the maximum was fifteen households.

All focus group participants concurred with the perception that land administrators do not discriminate against women and this is borne out by the figures in Table 3, which demonstrate that 38 per cent of plot owners in Old Naledi and Mogodit-

shane are women household heads, an identical proportion to Gaborone as a whole (GOB, 1994). Like male heads, female heads accessed their plots through purchase, allocation by state or traditional authorities, or inheritance. ‘There is no gender bias in land allocation. The biasness is found between the rich and the poor. The rich get plots because they manage to pay bribes’, said one woman. However, it has been reported elsewhere that some land boards do not allocate land to married women on the grounds that ‘traditionally a woman could not request a plot independently from her husband, and to allocate her one would be seen as divorcing the couple …’ (Natural Resources Services, 2003, 163).

Contesting exclusion

The emergence of Old Naledi on state land reserved for industrial use led to an outcry by labourers, casual workers and all those whose land and housing needs were ignored by colonial and post-colonial administrators. Their collective resistance to demolition and relocation constituted their bargaining power, which eventually forced the government to concede to their preferred area of residence and way of life. As a compromise, a few residents agreed to relocate in order to facilitate provision of basic services such as piped water, tarred roads, social facilities and, above all, secure land titles.

While Old Naledi epitomised struggles over land between the ‘poor’ and the state, unauthorised land transactions in Mogoditshane and other peri-urban villages represent various forms of power struggle, in the course of which the need to avail tribal land for urban development has been established and customary land tenure rules redefined. The first struggle is between the state and those people excluded by the land delivery system in Gaborone; the second is between headmen and field owners on the one hand and the state and land boards on the other; and the third between land speculators and the state. Headmen want to retain the land allocation responsibilities that are the source of their local political power, while field owners want to acquire economic power through the sale of their land rights. Speculators too want to make money from land transactions. The state and land boards want to establish and retain ‘formal’ delivery of ‘free’ land, so as to legitimate their existence and political power. The collaboration between headmen and land rights holders in selling arable land may be seen as a way of protecting the power base of traditional leaders (headmen in particular) and protesting against disenfranchisement on the part of land rights holders. Headmen would have been sidelined in the allocation of plots demarcated according to physical planners’ layouts. At the same time, land rights holders were generally dissatisfied with the amount of compensation paid by land boards for the loss of user rights. The Presidential Commission of Inquiry revealed that by the early 1990 land speculators were collaborating with land rights holders and

headmen in the sale of tribal land in peri-urban settlements. The District Administration, although not tasked with the land administration portfolio, got itself involved as an agent and representative of the central government.

Another symptom of the contested relationships between land rights holders and the land boards was the increased frequency of disputes. Our study revealed that land-related disputes were not common in Greater Gaborone - only 25 ( 9 per cent) of the 285 respondent owners reported that they had experienced any disputes. About half the reported disputes related to plot boundaries, 36 per cent to ownership, 8 per cent to inheritance and 4 per cent to other issues. Most were reported to have been amicably resolved with the assistance of the relevant land administering authority. Many more disputes arise between land rights holders and those trying to acquire plots on the one hand, and land boards on the other. Between 1998 and 2002, the Land Tribunal registered I,OI4 complaints against land boards, of which 37 per cent were against the Kweneng Land Board whose jurisdiction covers Mogoditshane. The disputes related to boundary extensions, multiple claims, inheritance rights, the land board reversing its decisions, and interpretation of the Tribal Land Act (1968) and the I993 amendments. A review of cases brought before it showed that the Land Tribunal has generally adopted approaches based on mediation and reconciliation to resolve matters.

Conclusion

The paper has revealed power struggles between land administrators and users and the influence of state political structures on urban land policies and processes of delivery. Policies have largely been defined to meet the interests and priorities of those who hold political and economic power (chiefs, colonialists and senior civil servants) rather than ordinary owners of fields in peri-urban areas or low-income urban households. The latter groups have, however, persistently utilised their agency and the resources at their disposal to create spaces and rules for transacting in land rights outside state structures, which has in turn forced the government to occasionally revise its policies.

Because of its wealth and policies that favour the better off, Botswana has performed exceptionally well in ensuring that its towns and cities develop in an orderly and wellplanned manner. Using revenue from its mineral resources, the government has striven to ensure that its urban populations have access to generously sized, well-planned and adequately serviced residential plots. This commitment has, however, created high expectations on the part of urban residents, to the extent that the government has found itself unable to satisfy the ever-increasing demand for land and houses. The demand for land appears to have been fuelled by the relatively low costs at which state land plots are offered to rich and poor residents alike and the ‘culture of entitlement’

inherited from pre-colonial and colonial experiences but condoned and nurtured by the post-colonial state. The problem appears to have been exacerbated by ‘speculative demand’ and uncertainty about where, when and whether additional land supplies will be made available. This uncertainty, coupled with speculation, appear to have created room for corrupt practices among land administrators.

In desperation, the poor in Gaborone have erected unauthorised shelters, initially on a piece of state land reserved for industrial use (Naledi), and later on tribal land in peri-urban settlements. In a way, the unauthorised occupation of state and tribal land has made the plight of the (relatively) poor visible to the government. That it did finally regularise much unauthorised settlement and also revised state policies and laws on land and housing demonstrates that they have had some influence on revising regulations and procedures for accessing, controlling and transferring land rights. By collaborating with embattled headmen and arable land rights holders, low-income people have successfully utilised the plurality of land administrators (in ministries, land boards and sub-land boards), as well as ambiguities in the Tribal Land Act (1968), to negotiate new terms for accessing and utilising communal land. Our study revealed that the ‘illegal’ transfer and sale of tribal and state land continue unabated using inheritance provisions and other loopholes in the current legislation. As long as land and housing policies benefit those endowed with political and economic power and ignore the needs of the poor, informal land subdivision will persist. Indeed, the history of state-regulated land delivery processes in Botswana is consistent with experiences observed elsewhere: the elites that control access, use and transfer of land are invariably the principal beneficiaries.

References

ADAMS, M., KALABAMU, F. T., and WHITE, R. (2003), ‘Land tenure and practice in Botswana: governance lessons for Southern Africa’, Austrian Journal of Development Studies, XIX, 55-74.
BERRY, S. s. (2001), Chiefs Know Their Boundaries, Cape Town, David Philip.
CAMPBELL, A. C. (1982), ‘Notes on the prehistoric background to I840’, in R. R. Hitchcock and M. R. Smith (eds), Settlement in Botswana, Gaborone, Botswana Society and Heinemann, I3-22.
colclough, c. and mCCarthy, s. (ig8o), The Political Economy of Botswana: A Study of Growth and Distribution, Oxford, Oxford University Press.
daley, e. and hobby, m. (2005), ‘Land: changing contexts, changing relationships, changing rights’ (paper commissioned by the Urban-Rural Change Team, DFID).
duckson, w. L. (1990), Land Tenure and Management in a Developing Country, Gaborone, University of Botswana, NIR Working Paper No. 54 .
DTRP (Department of town and regional planning) (1998), ‘Gaborone City Development Plan’, Gaborone: Department of Town and Regional Planning, Final Draft, unpublished.

FOURIE, C. (2004), ‘Land readjustment for peri-urban customary tenure: the example of Botswana’, in R. Home and H. Lim (eds), Demystifying the Mystery of Capital: Land Tenure and Poverty in Africa and the Caribbean, London, Glasshouse Press, 3I-49.
GOOD, K. (1992), ‘Interpreting the exceptionality of Botswana’, The Fournal of Modern African Studies, 30, 69-95.
gob (government of botswana) (1970), National Development Plan: 1970-75, Gaborone, Government Printer.
gob (government of botswana) (1972), Report on the Population Census 1971, Gaborone, Government Printer.
gob (government of botswana) (1982), National Policy on Housing, Gaborone, Government Printer.
gob (government of botswana) (ig83a), Report of the Presidential Commission on Land Tenure, Gaborone, Government Printer.
gob (government of botswana) (ig83b), An Evaluation of the Self-help Housing Agencies, Gaborone, Ministry of Local Government and Lands.
gob (government of botswana) (ig83c), 1981 Population and Housing Census: Guide to Villages and Towns of Botswana, Gaborone, Government Printer.
gob (government of botswana) (ig85), National Development Plan 6: 1985-1991, Gaborone, Ministry of Finance and Development Planning.
gob (government of botswana) (1990), Allocation of State land: New Policy, Gaborone, Ministry of Local Government, Lands and Housing.
gob (government of botswana) (1992a), Report of the Presidential Commission of Inquiry into Land Problems in Mogoditshane and other Peri-Urban Villages, Gaborone, Government Printer.
gob (government of botswana) (1992b), Land Problems in Mogoditshane and other Peri-Urban Villages, Gaborone, Government Paper No.i of 1992, Government Printer.
gob (government of botswana) (1992c), Population of Towns, Villages and Associated Localities, Gaborone, Government Printer.
gob (government of botswana) (1994), 1991 Population and Housing Census: Summary Statistics on Small Areas, Volume I & 2, Gaborone, Government Printer.
gob (government of botswana) (2002), 2001 Population and Housing Census: Population of Towns, Villages and Associated Localities, Gaborone, Government Printer.
gob (government of botswana) (2003), National Development Plan 9: 2003/04-2008/09, Gaborone, Ministry of Finance and Development Planning.
JEPPE, w. J. o. (1980), Bophuthatswana: Land Tenure and Development, Cape Town, Maskew Miller.
kalabamu, F. т. (2000), ‘Land tenure and management reforms in East and Southern Africa - the case of Botswana’, Land Use Policy, 17, 305-19.
kalabamu, F. т. (1998), ‘Gender neutral housing delivery systems: a critical analysis of housing in Botswana’, in A. Larsson et al. (eds), Changing Gender Relations in Southern Africa: Issues of Urban Life, Roma, National University of Lesotho, 56 -76.
kalabamu, F. t and morolong, s. (2004), Informal Land Delivery Processes and Access to Land for the Poor in Greater Gaborone, University of Birmingham, School of Public Policy, International Development Department, Informal Land Delivery Processes in African Cities, Working Paper 3 .

kombe, w. J and kreibich, v. (2000), Informal Land Management in Tanzania, University of Dortmund, Faculty of Spatial Planning, SPRING Research Series No. 29.
kweneng district administrator (ig82), ‘Mogoditshane: Gaborone’s refuge or traditional village’ (unpublished memo).
LEDUKA, R. C. (2000), ‘The role of the state law and urban social actors in illegal urban development in Maseru, Lesotho’ (unpublished PhD thesis, University of Wales, Cardiff).
leduka, R. C. (2004), Informal Land Delivery Processes and Access to Land for the Poor in Maseru, Lesotho, University of Birmingham, School of Public Policy, International Development Department, Informal Land Delivery Processes in African Cities Working Paper 5.
mathuma, в. м. (ig89), Report on the Review of the Tribal Land Act, Land Policies and Related Issues, Gaborone, Government Printer.
mbiba, b. and huchzermeyer, m. (2002), ‘Contentious development: peri-urban studies in subSaharan Africa’, Progress in Development Studies, 2, i13-31.
mlgi (ministry of local government and lands) (ig90), ‘An analysis of providing serviced land and housing in Botswana’s seven centres’ (unpublished report, Gaborone, Ministry of Local Government and Lands).
mleh (ministry of lands and housing) (2001), ‘Review of the Accelerated Land Serving Programme’ (unpublished report, Gaborone, Ministry of Lands and Housing).
musyokI, A. (1998), ‘Perceptions on gender and access to housing in Botswana’, in A. Larsson et al. (eds), Changing Gender Relations in Southern Africa: Issues of Urban Life, Roma, National University of Lesotho, 266-85.
natural resource services (2003), ‘Review of Botswana National Land Policy’, Gaborone, Department of Lands (2 volumes), unpublished.
ngcongco, L. (ig82), ‘Prehistoric migration in southeast Botswana’, in R. R. Hitchcock and M. R. Smith (eds), Settlement in Botswana, Gaborone, Botswana Society and Heinemann, 23−3023-30.
ng’ong’ola, c. (1997), ‘Land rights for marginalized ethnic groups in Botswana, with special reference to the Basarwa’, Fournal of African Law, 41, i-26.
ng’ong’ola, c. (1996), ‘Land tenure reform in Botswana: post-colonial developments and future prospects’, SARP/PiI
ng’ong’ola, c. (1993), ‘Ownership of tribal land in Botswana: Kweneng Land Board versus Kabelo Matlho and Pheto Motlhabane’, Fournal of African Law, 37, 193-98.
nkurunziza, e. (2004), Informal Land Delivery Processes and Access to Land for the Poor in Kampala, Uganda, University of Birmingham, School of Public Policy, International Development Department, Informal Land Delivery Processes in African Cities, Working Paper 6.
powelson, J. P. (1987), ‘Land tenure and land reform: past and present’, Land Use Policy, 4, III-20.
rakodi, c. (2000), ‘Preface’, in W. Kombe and V. Kreibich (eds), Informal Land Management in Tanzania, University of Dortmund, Faculty of Spatial Planning, SPRING Research Series, No. 29, iii-iv
rakodi, c. and Leduka, R. C. (2003), Informal Land Delivery Processes and Access to Land for the Poor in Six African Cities: Towards a Conceptual Framework, University of Birmingham, School of Public Policy, International Development Department, Informal Land Delivery Processes

in African Cities, Working Paper I.
schapera, i. (1943), Native Land Tenure in the Bechuanaland Protectorate, Cape Town, Lovedale. schapera, i. (1994), A Handbook of Tswana Law and Custom, London, International African Institute and Gaborone, The Botswana Society.
stihole, т. r. (1995), ‘An evaluation of urban planning in peri-urban settlements of Botswana: a case study of Mogoditshane’ (unpublished MURP thesis, Heriot-Watt University and Edinburgh College of Art).
undpa (united nations department of political affairs) (2003), Land Reform in Southern Africa: Case Studies and Recommendations, United Nations Department of Political Affairs, first draft van nostrand, J. (1982), Old Naledi: The Village Becomes a Town, Toronto, James Lorrimer and Company.

Forthcoming in IDPR 28/3

Viewpoint

Urban-rural linkages: towards a more balanced approach to development of sustainable human settlements
Don Okpala

Articles

Patterns of population and employment change in Shanghai in the ig9os: a zonal analysis with international comparisons
Yu Zhu
Tourism planning and regional instability: the consequences of rapid expansion of the tourism sector in Jordan
Petra Doan
Housing and development strategies in Ghana, 1945-2000
Godwin Arku
Valuation of environmental resources for tourism in small island developing states: implications for planning in Jamaica
Elizabeth Thomas-Hope and Adonna Jardine-Comrie
Participatory force? Youth and the making of urban places in Zimbabwe Amin Y. Kamete

Determinants of success in community self-help projects: the case of the Kumbo water supply scheme in Cameroon
Ambe J. Njoh