Research Advisor, United Kingdom Department for International Development
Land reform has been high on the development agenda for Uganda since the early 1990s. This paper examines, from an implementation perspective, the processes by which land issues have been integrated into the broader development agenda by first outlining the major linkages between land access, ownership and poverty, then chronicling the treatment of key policy issues in the making of the Constitution of 1995 and the Land Act of 1998. The paper illustrates how the parameters set down in the Constitution defined the scope of redistributive reforms on tenanted land and leasehold land, and the nature of rights to be determined for customary landholders. It then turns to a review of the emergence of discord between the objectives and the resource requirements of the Land Act and the imperatives of national macroeconomic management and national poverty reduction policy as enshrined in the Poverty Eradication Action Plan (PEAT). The emergence of new methods to bring about greater harmony between the land-reform process and the national poverty reduction strategy since 2000 is elaborated. Finally, the paper highlights some general lessons from the experience of implementation of Uganda's land reform.
Developments in public policy towards land in Uganda over the last 15 years have largely been pursued through two parallel processes, in different fora and with different lead actors. While both processes have at heart questions of land access and ownership, one has been dominated by the need to address political realities through the resolution of historical tenure problems, while the other has focused on issues of economic efficiency and eventually on poverty reduction. Integration of these two approaches into a coherent strategy placing land within the broader national development context has taken place only recently. While poverty reduction as a goal of land reform has been a common feature of debates, it has only recently become paramount through the debates around the contribution of land to achievement of national poverty reduction goals. These have highlighted the importance of land reform for enhancing landownership and productivity, safeguarding land access for the poor and pastoralists, and also for governance and domestic revenue generation.
THE RELATIONSHIP BETWEEN LAND AND POVERTY
There has been no comprehensive analysis of the relationship between land tenure, land use and poverty in Uganda. Information on land distribution, land tenure and land use in Uganda is poor and unreliable. However, an attempt can be made using available information to characterize the current socio-economic position concerning land and to review briefly possible links between the existing frameworks for land access, ownership and use and poverty.
Poverty in Uganda is generally declining. In 1997, 44 percent of the population was deemed unable to meet its basic requirements and to live below the official poverty line; by 2000 this figure was believed to have fallen to 35 percent (MoFPED, 2001). Poorer households are more likely to live in the north, reflecting poorer market access and the effects of prolonged conflict in this region. However, throughout Uganda, households with a non-working head are experiencing worsening poverty. The heads of such households are commonly the elderly, children, or the disabled, which is of particular significance in view of the social impact of the HIV/AIDS pandemic (McKinnon and Reinikka, 2000).
The agricultural sector is the most important source of income and livelihoods for Uganda's predominantly rural population. Agriculture contributes 43 percent of Gross Domestic Product (GDP), 85 percent of export earnings, and 80 percent of employment (Republic of Uganda, 2000). Although food crop production dominates, only one-third of the food crop produced is marketed. Major crops are bananas, cereals, root crops, pulses and oil seeds. Export crops, including coffee, tea, tobacco and cotton, account for only 8 percent of cultivated area.
Because of the significance of agriculture to rural livelihoods, land is the most important asset for many Ugandan households. Average land holdings are estimated at 2.2 ha per household, although there are inter- and intraregional inequalities in this distribution, and evidence suggests much of this land is not cultivated (McKinnon and Reinikka, 2000). Land is accessed both with and without the market. Clearance of land is uncommon as the majority of land has already been claimed. The bulk of evidence suggests that in aggregate land is more likely to be obtained through inheritance than through purchase. However, recent evidence points to a rise in the proportion of land obtained through purchase, such that acquisition through the market is more common than through inheritance or gift in all regions other than the north (MISR, 2000). In addition, around 5 - 10 percent of households access land predominantly through borrowing or renting, and although absolute landlessness (having no access to land at all) is uncommon, there is evidence to indicate the growing presence of a group of households whose access to land is restricted to land acquired by borrowing or renting, and who have no land of their own. Such households are often among the poorest and supplement their incomes by casual labour. For those renting land, rental rates frequently represent a substantial proportion of the value of crop produced and this may contribute to poverty. Nevertheless, in areas where land is in short supply, the availability of rental land provides an opportunity for some households to access land and can be an important part of a strategy for increasing income and acquiring assets. At the time of development of key elements of land reform, the extent and consequences of inequality were not well documented or understood, and it was believed that land distribution was relatively equitable.
A number of theoretical arguments are commonly advanced concerning links between landownership, land markets, security and production incentives. It is argued that:
1. the presence of an efficient land market will tend to result in a more efficient allocation of land between producers and that this will stimulate production increases;
2. secure land rights will stimulate the credit market through the use of land as collateral;
3. security of land rights is linked to increased incentives to invest in land and land-based production (Platteau, 2000).
Although empirical evidence for all of these propositions within Africa is weak, these perspectives have been influential in the debate on land reform in Uganda. Historically, land markets in rural Uganda have tended to operate within customary tenure systems. Thus markets, although present in most of Uganda today, tend to operate within a system of social control, privileging access of members of the dominant social group above that of "outsiders". Only in the Mailo areas of the central region and in urban and peri-urban areas has there been significant availability of marketed land to persons with different origins from the majority group. By and large, this situation persists. One of the advantages of a free market in land - that of permitting interregional migration - is therefore limited. Land sales, on the other hand, have tended to be associated with increasing poverty, with the majority of sales being prompted by "distress", and the associated loss of control of assets leading to increased household vulnerability.
Availability of credit in Uganda has tended to be relatively narrow, and uptake among rural households not high. The formal credit system operates on the basis of land as collateral, with a preference for urban and titled land that is more easily disposed of in cases of default. Savings and credit clubs and other informal sources of credit are known to accept rural land as collateral for small microfinance loans (Hunt and Eturu, 1999). However, with the recent expansion of microfinance institutions willing to accept many other assets as security, or to lend on the basis of peer group pressure, the necessity of land as collateral is much reduced. Availability of land for collateral is therefore unnecessary for much of the population, and is more relevant for large-scale investment.
Tenure insecurity has not been commonly identified as a problem in Uganda, even though the legal rights of the majority of the population are inherently insecure. Claims to land were traditionally secured through length of residence or through use, and through social links. Locally respected boundary demarcations are understood in most parts of the country. However, under the pressure of commercialization and population growth, there is evidence of increasing demand for more formal recognition of land claims, in particular in respect of land transacted through market mechanisms, for example for tenancies purchased in urban areas or for land purchased in rural areas (Leatherdale and Palmer, 1999). There is some evidence that tenure insecurity does have an impact on land use, e.g. in relation to rental land where some people feel they will not be able to enforce a short-term lease provision, and in respect of certain types of land use, e.g. permanent crops or tree crops.
In the case of women, however, land tenure insecurity is severe and is linked both to low incentives to produce, and to poverty and food insecurity. A frequently cited statistic is that women in Uganda provide 80 percent of farm labour, yet own only 7 percent of the land. The statistical validity of this claim is suspect, but its foundation is legitimate: the stake of Uganda's female farmers in this critical productive asset is severely limited and the manner of accessing land for women perpetuates social and economic inequality. Aside from the difficulties of divorced and widowed women in retaining access to marital land, a growing body of evidence points to the fact that married women are denied the opportunity to plant perennial cash crops or tree crops, and withdraw their labour from cash-crop production on fields they do not control. Direct links between women's land tenure insecurity and household food insecurity are becoming more evident. Food insecurity has not been of widespread concern in Uganda, although localized and seasonal food shortages do occur, and malnutrition affects significant numbers of children. Historically, food shortages coincide with periods of civil war, and food production has recovered since 1986, except for some areas in the north and northeast where disruptions to production systems from war persist. However, at the household level, there are growing indications that gender divisions of labour and control of income are tending to undermine food security where the expansion of cash crops is at the expense of food crop cultivation, and that this is having an effect on nutritional status of children. The gender distribution of land rights has been linked to this phenomenon, specifically because where women own land or have a stronger stake in decisionmaking on the family farm, they also have a higher degree of control over production and income from cash crops (Ovonji-Odida et al., 2000). The issue of women's land rights is therefore of critical importance not only to the land-reform process but also to wider poverty reduction strategies through both production and social welfare impacts.
PRIORITY ISSUES FOR UGANDAN LAND REFORM
Early priorities for the reform process tended to focus on issues of tenure security in relation to specific historical tenure problems. Reform of the structures for land management and land administration, and issues of land use and environmental sustainability assumed greater importance at a later stage in the process.
Land tenure issues
The current distribution of land rights, together with the systems of land administration and management, are rooted in the past, and any attempt to resolve Uganda's land problems requires an appreciation of past policies and their impacts. In common with most colonial regimes, the Protectorate Government in Uganda implemented a dualist system of land tenure, through which parts of the country were brought under a system of statutory controls while in the remainder traditional customary systems were maintained. Because Uganda was not a settler colony, the dualist system privileged certain Ugandan political and ethnic interests over others. Within the customary sector, regional and local variations existed related to ethnicity, farming systems and agro-climatic conditions. The impetus for resolution of these tenure issues was largely political and the connection with poverty reduction may not be immediately evident, although in large measure they deal with the problems arising from the need to strengthen the rights of occupiers vis-à-vis those of owners.
Rights of mailo tenants
In Buganda, the fertile central area of the country and home of the capital city, members of the royal family, county chiefs and other leaders granted official or private estates in land under a form of freehold known as "mailo" under an Agreement made between the British and the Baganda rulers in 1900. The occupants of these lands, previously engaged in a semi-feudal customary relationship with the chiefs, became tenants open to expulsion by the landowner. Throughout the twentieth century, through various legislative twists and turns, the security of tenants was first strengthened, then weakened, and finally eradicated. However, for the most part, the owners did not evict their tenants, and the majority of tenants continued to occupy the land without legal protection. It is estimated that this applies to around 19 percent of the population, occupying around 14 percent of the land area (Oates and Ecaat, 1999).
In other parts of the country, particularly in Bunyoro in Western Uganda, land had been gifted by the British to a class of absentee Baganda landlords, while the inhabitants had also been rendered tenants. The Banyoro never recognized the rights of the landowners, and many landowners never occupied the land. Despite a 1964 referendum resulting in the administrative return of the area to Bunyoro, the land was not returned, and the inhabitants of Kibale continued to farm their land as tenants of absentee landlords. Since the 1960s, governmentsponsored resettlement has brought settlers into Kibale from the densely populated southwest. Ethnic resentments have added a further dimension to the historical grievance of the Banyoro.
Customary land rights
Customary land rights were also vulnerable as a result of the declaration of Crown Land status by the colonial administration. Customary users were tenants of the Crown, and Crown Land could be alienated in freehold or leasehold, whereupon customary occupiers had either to leave the land or stay as tenants at the sufferance of the freeholder or leaseholder. These provisions were applied in many cases, linked to the expansion of the churches and the limited development of plantation agriculture. Despite a limited number of pilot schemes for adjudicated freehold in the 1950s, the attempt fell foul of farmers' fears of alienation of land to non-Africans in the wake of the Mau Mau rebellion in Kenya and was never widely extended. Around 75 percent of Uganda's population are believed to occupy land under customary tenure, either on an individual or a communal basis (Oates and Ecaat, 1999).
Post-Independence developments had also left a legacy of land tenure issues to be resolved. The vesting of public land (former Crown Land, i.e. customary land) in the state at Independence provided the basis for the allocation of land in leasehold through political patronage networks operating through District Land Boards and the Uganda Land Commission. There are thought to be around 68 000 leaseholds in Uganda, although it is not known what land area they cover (Leatherdale and Palmer, 1999). Requirements to demonstrate the land in question was either unoccupied or that customary users had been compensated prior to grant of lease were manipulated, and subsequent development conditions have rarely been enforced, so that much leased land has remained under- or unutilized.
In urban areas, particularly Kampala, informal trading and inheritance of land created a complex web under which original mailo owners, tenants, and informal "squatters" all claim rights. Similar situations exist in leasehold properties with sublease rights being traded for substantial fees.
Women's land rights, secondary and third-party rights
Because of the links between tenure insecurity and poverty for some groups, particularly women and pastoralists, the land-reform process also considered the status of a variety of secondary and thirdparty rights existing within the various tenure systems. Chief among these are the rights of family members, predominantly women and minors, whose claims to access and ownership derive primarily from social relationships. Ugandan societies are patrilineal and most women gain access to land through marriage. Despite inheritance laws, widows are frequently evicted, and children disinherited by family members. Similarly, when land transactions are made, spouses and children frequently are not party to the transaction. These occurrences are common to both registered and customary land. A variety of other land and resource rights, chiefly concerned with access to water and pasture, but varying around the country, are recognized but are under stress without legal recognition.
Land management and administration
In addition to tenure issues, the landreform process needed to consider the appropriateness and cost-effectiveness of existing systems of land administration and management. The principal features of these systems are centralized and inefficient land management and a costly and inaccessible system of land administration. Overall authority to allocate land was centralized, first in the Crown, and subsequently in the Ugandan state through the Uganda Land Commission. Land-use planning was also a central function, under the direction of the Town and Country
Legislation relating to registration, surveying, and land acquisition is outdated and not suited to the current needs of Uganda's landowners and users. The requirements for absolute guarantee of Certificates of Title by the state and a full cadastral survey have made the process of acquiring registered title and registering subsequent transactions and transfers difficult and costly, and with the addition of bureaucratic inefficiency and the rent-seeking activities of an underpaid civil service, well beyond the means or patience of the majority of land users. The Constitutional Commission reported that many people have given up trying to obtain titles to their land because of corruption in the land office (Republic of Uganda, 1993). Underfunding and underinvestment, together with lack of qualified personnel in some areas, hampers efficiency further, particularly due to the poor condition of records and maps and lack of equipment. Several thousand titles are reported to be missing or misplaced.
The inefficiencies and inequities of the judicial system regarding land were assuming greater importance as the numbers of disputes climbed in response to the increasing commercialization of land in some areas and the problems in the land registry. A dual system of justice based on Magistrates Courts and "popular courts", known as Resistance Committee (RC) Courts permitted the presentation of disputes before multiple fora. By the early 1990s it was evident that this was leading to an inequitable system. The majority of land disputes, relating to customary boundary and inheritance issues but frequently involving matters of customary marriage or paternity, would be heard by RC courts, but those with the means would pursue cases through magistrates courts, leading to the practice of "forum shopping". The advantages of RC Courts in terms of accessibility (operation at village level, use of local languages and local values, low cost) were acknowledged, but at the same time, there were concerns that, being elected bodies with legislative powers as well as judicial powers, RCs were in a position to abuse their powers and were vulnerable to influence and bias. More importantly, there were concerns that RCs operated without adequate knowledge of the law, and kept poor records, if any.
Land use and environmental issues
Although initially peripheral to the debate, land use and related environmental issues are increasing in importance, as evidence mounts of environmental stress owing to the competing demands on Uganda's land resources from agriculture and human settlement.
As Uganda sought to expand production during the 1990s, there was concern about the low rate of utilization of agricultural land. In 1993, it was estimated that only 5 million hectares out of a total 17 - 18 million available were under cultivation (World Bank, 1993). In the central region there was concern that the persisting land-tenure impasse between the (frequently absentee) mailo landowners and occupants was preventing significant amounts of land from being brought into production. Although the amount of land under cultivation is rising (estimated to have reached around 8.5 million ha in 1999), soil degradation, erosion and deforestation are also increasing, and this rate of growth without adequate soil conservation measures is considered by environmentalists to be unsustainable.
Urban and peri-urban areas have been growing rapidly in Uganda over the last 15 years. While still accounting for a tiny fraction of total land area (less than 0.5 percent), almost 10 percent of the population now resides there, and a far greater proportion depend on them for employment, income and amenities. In 1950, Uganda's population was 4.7 million, of whom 3.1 percent lived in urban areas. Recent estimates suggest that the population will grow to around 33 million by 2016, and that around 9 million will live in urban areas (Uganda Office of Statistics, 2001). Planning of urban and peri-urban settlements, provision of infrastructure, and avoidance of harm to fragile environments are emerging as vital issues for the landreform process.
Uganda's other vital natural resources, forests, wetlands and wildlife, are also threatened by human activity. Less than half of Uganda's forests (around 1.9 million ha) are protected and the remainder is threatened by demand for timber and timber products. On private land, tenure insecurity, particularly in mailo areas or in the case of female land users, has been identified as an obstacle to planting of tree species and the development of agro-forestry, as tree crops are regarded as conferring long-term rights to land, and on mailo land may be prohibited by landowners. Woodland in communal areas is threatened by inadequate management and regulatory provisions. Wetlands, accounting for 13 percent of Uganda's land area and providing a range of benefits, especially for poorer people, are under pressure from surrounding communities for cultivation or are threatened by pollution. Until recently, wetlands were unprotected and liable to privatization through extension of private cultivation and even the building of structures. Some have been alienated in leasehold. Protection of wildlife provides challenges in terms of balancing the needs of local populations and conservation, and is perceived to be unfair in some parts of Uganda, in particular in the Kasese district in Western Uganda, where 67 percent of the land area is within protected areas, and the pastoralist region of Karamoja, where 48 percent of the land is protected (Moyini, 2001).
Increasing competition between different land uses because of population growth, commercialization and inadequate or ineffective regulatory policies is generating intensified conflict over land use in many areas. At times conflicts have become violent, even recently threatening the stability of some districts in the northeast as clashes between pastoralists and settled cultivators were exacerbated by the presence of firearms. In general, structures for common management of resources have not been supported and in some cases have been undermined by the previous land tenure and management policies.
INTEGRATING LAND ISSUES INTO THE NATIONAL POVERTY REDUCTION STRATEGY
The integration of land issues into the national poverty reduction strategy has been to a large extent a story of two parallel processes, pursued in different fora and with different lead actors. Early moves towards addressing land issues during the late 1980s were driven by an economic agenda with its principal goal and output being a Bill aimed at removal of impediments to the land market and the introduction of uniform, individualized land tenure throughout the country (Tenure and Control of Land Bill, 1993). This process stalled before legislation could be brought before Parliament. During the mid-1990s two processes commenced, the first pursued predominantly in the political arena, and the second in that of economic management. The objectives and priorities of these processes were different, and their treatment of land issues diverged to a degree. In the first of these processes, land issues were consulted on and debated through the making of the Constitution in 1995, and thereafter a Land Act was drafted and passed in 1998. In the second, a comprehensive and participatory national strategy for poverty reduction was pursued through the formulation of Poverty Eradication Action Plans (the first in 1996 - 97, and the second in 2001, accepted as Uganda's poverty reduction strategy). Through an iterative process, these divergent views and priorities are now being brought closer together, and land issues and strategies are now being brought into line with the national poverty reduction strategy. However, there is still no comprehensive policy statement linking land and national development strategy.
Uganda's Constitution was developed through a consultative process in the early 1990s, and was intensely debated through an assembly of elected representatives known as the Constituent Assembly (CA). The most hotly debated land issues during the CA were the rights of tenants and owners over land in mailo areas, and the status of customary tenure.
Chapter 15 of the Constitution, together with a few articles scattered elsewhere, represents the principles of land policy as they can be discerned in Uganda. The principal constitutional provisions relating to land vested in the citizens of Uganda uphold private property rights against public acquisition, provide for customary, mailo, leasehold, and freehold, and for customary and leasehold tenure to be converted to freehold, and guarantee the security of "lawful and bona fide" tenants on mailo land. In addition, the Constitution required the establishment of District Land Boards and Land Tribunals, and provided for the state to protect and manage natural resources under public trust. The CA, being unable to agree on the respective rights of mailo owners and tenants, deferred this task to Parliament, which was required to enact a law regulating this relationship within two years. There were two key factors at play: the most significant apparently being the desire to preserve the status quo, but a second objective was to lay the ground for an expansion of individualized land rights and the land market. In opting to preserve the status quo and to support the rights of citizens against the state, the 1995 Constitution pre-empted a more comprehensive discussion about land policy, particularly about redistribution, and it laid down principles that have in some senses narrowed the debate concerning the role of land in national development.
The national poverty reduction strategy
Uganda's macroeconomic policy since 1986 has focused on market development and export-led growth. Key macroeconomic issues have been the maintenance of macroeconomic stability, price stabilization, and shifting production from untraded to traded goods. In an effort to combine these policies into a coherent national development strategy, Uganda was one of the first low-income countries to prepare a comprehensive and participatory national strategy for poverty reduction, through the formulation of the first Poverty Eradication Action Plan (PEAP) in 1996 - 97. The PEAP aimed to refocus and redirect government policy towards the eradication of poverty, and to state a high-level political commitment to this end. The formulation of the PEAP was spearheaded by the Ministry of Finance and Economic Planning.
In the first PEAP, developed in the period between the promulgation of the Constitution and the enactment of the Land Act, access to land and sustainable use of land were recognized as critical issues for enabling the poor to increase their incomes. The strategies included in the PEAP were not at odds with the provisions of the Constitution, and recognized tenure security, particularly that of smallholders, women, and tenant farmers, as a key objective of government policy towards land. The PEAP based its strategy on the argument that tenure insecurity was a barrier to increased production by smallholders. In addition, the lack of acceptability of rural land as collateral by the banking system was believed to be a problem.
At this point the two processes were relatively harmonious, if disjointed, and there was little hint of conflict between the goals of the Constitution and those of the national poverty reduction strategy. This situation was to change substantially with the subsequent preparation of the legislation required by the Constitution.
The Land Act
The Constitution, while providing the basic principles and framework for land legislation, had left many of the priority issues for land unanswered. Chief among these were the rights of tenants on mailo land, the status of certificates of customary ownership, and the procedures for conversion of leaseholds and customary holdings to freehold. In addition, the vexed issue of resolving the historical dispute between tenants and owners in Bunyoro had not been directly addressed. In drafting the Land Act, these issues assumed paramount importance. Additional concerns over the rights of women and rights over communal areas were added through the advocacy efforts of Non-Governmental Organizations (NGOs) and some donors. At this stage, there was no attempt to link land legislation to the overall macroeconomic framework and the goals of the PEAP. The treatment of the priority land issues outlined was as follows.
Rights of mailo tenants
The perpetual security of two categories of occupant is guaranteed provided they pay a nominal statutory rent (less than US$1) to the landowner. Those who had heritable rights based on the former laws governing landlord and tenants' rights (lawful), and those who had acquired occupancy rights through transfer, settlement or occupation for at least 12 years (bona fide) are also permitted to obtain a Certificate of Occupancy within defined boundaries, and to trade in and mortgage their tenancies. Limited market-based redistribution is provided for between tenants and owners, each being required to give the other the right of first refusal when land is offered for sale.
In Kibale, the Act had to provide a mechanism for government to play a stronger role in the redistribution of rights, but the scope for manoeuvre was complicated by the Constitutional principles of respect for established property rights and the limited grounds on which land could be acquired in the public interest. The Act provides for a fund to be used by the state to compensate the owners and redistribute the land to the occupants, known as the Land Fund. During the parliamentary debate, interests from other areas of the country successfully argued that Kibale was not the only case deserving of state assistance in redistribution, and in the final event the Land Fund provides for the state to compensate owners in a variety of circumstances, including where fragile environments are at issue, and also provides for tenants on mailo land to be assisted in buying out their tenancies through a loan fund.
Customary land rights
Measures for certification of customary holdings prompted significant advocacy efforts by civil society. Concerns focused on two issues: the potential negative effect of individualization of customary land on tenure security through the adjudication process or increased market transactions, and the need to provide for recognition of group rights to allow for communal certification of land. Both individual and group rights are provided for in the Act, but restrictions on the operation of the market on individualized holdings are limited to provisions relating to family control over transfers. The Act provides for the recognition of group rights over land used for communal purposes, through the registration of communal land associations with elected management committees. The Act also provides for procedures and conditions for the issuance of customary certificates and their conversion to freehold, including the restriction that recognition of individual rights must be acceptable in local custom, and that all those with an interest in the land must be party to the verification of rights.
Leasehold had been recognized by the Constitution without reference to the inequities of distribution of leaseholds and the right to convert leasehold to freehold enshrined under Article 237. Leaseholds can therefore only be freely converted to freehold where it can be verified that the interests of the customary tenants were compensated for at the time of granting of the lease, and the right of conversion is normally limited to 100 ha. Land over 100 ha which is converted must be paid for.
While the rights of "lawful and bona fide" tenants on mailo land were clarified by the Act, this still left a significant category of occupants of mailo land uncovered by the Act and unprotected. This group, known as "squatters", includes all those who occupied land after 1983, even if they had lived on and used the land undisturbed by the owners. Based on the law of limitation, which permits title to be claimed only after 12 years of undisturbed occupation, all those who occupied after 1983 are required to take steps to identify the owner and negotiate their occupation.
Women's land rights
The critical issue of women's land rights was raised during parliamentary debate with reference to customary tenure, generating the ongoing controversy around what is referred to as the "Lost Amendment". The Act contains provisions restricting the transfer of land to cases where the spouse and children with an inheritance right have consented through various processes. In the last days of the debate, champions of women's rights proposed an amendment seeking to extend the rights of married women under customary tenure to that of joint ownership over land acquired during the marriage for the purposes of family residence or subsistence, a matter of some importance for the national poverty strategy as has been discussed above. The House accepted the principle behind the amendment, but it was not included in the final version of the Act that was assented to by the President.
Land management and administration
The Land Act follows the Constitution in decentralizing control of land allocation to District Land Boards and establishing dedicated Land Tribunals for the resolution of land disputes. Following the policy of decentralization established in the Local Government Act of 1997, it also establishes a district-based system of land administration through District Land Offices, with provision for recording at subcounty (subdistrict) level through the new office of Recorder. In addition to these reforms, the drafters of the Act sought to bring land administration and dispute resolution closer to the citizens and to provide the means for verification and demarcation of land claims at local level through the establishment of land committees at parish level and land tribunals at subcounty level. Thus, the Act provides for a much-enlarged bureaucracy for the administration of land and the resolution of land disputes, for which controls and accountability are unclear.
Land use was not an explicit priority for the Land Act or, for that matter, for the Constitution. Nevertheless, the Act contains provisions that have profound implications for the control of land use. The Act restates the principle of public trusteeship over natural resources with all other state land being held under private tenures, restrictions placed on acquisition of land for planning purposes because of the requirement to pay compensation for land, and the requirement that all privately held lands be used in accordance with a range of environmental statutes.
In sum, the Land Act and the Constitutional dispensation steered a difficult course between the demands of "modernization" and support to market allocation of land and concerns to protect the rights of those likely to be rendered more vulnerable by the expansion of the market. Reflecting the PEAP objectives, the Act provided mechanisms for "entrenching" the rights of smallholders and ostensibly laid the ground for an expansion of landbased rural credit as both Certificates of Customary Ownership and Certificates of Occupancy were to be mortgageable. In practice, the likelihood of banks accepting either as collateral was limited because the banks' acceptance of land is based more on location and ease of disposal in the event of default than on tenure status. Although the Act went some way to strengthening the rights of women over land as envisaged by the PEAP, it did not establish a relationship between the registration of land and protection of women's rights.
The passing of the Land Act was greeted with a mixture of euphoria and dismay by the various sections of the public. There were many misconceptions about the intentions and the content of the law, but there were also high expectations relating to the Land Fund and to the ability to use land for credit. For the policy-makers, jubilation was short-lived as the enormity of the challenge of implementation rapidly became apparent.
CHALLENGES OF IMPLEMENTATION: THE LAND ACT MEETS THE POVERTY ERADICATION ACTION PLAN
The Land Act was enacted essentially without forethought concerning the funding and human resource requirements for executing the wide-ranging tenure and institutional reforms it proposed. The need for an implementation plan was quickly established, and, with donor support, an interim programme seeking to inform people of their new rights and establish a minimum institutional infrastructure while developing a longer-term strategy for full implementation was developed. With the commencement of implementation and longer-term planning, cracks began to appear between the land-reform process and national development goals.
The extent of the divide between the land reform proposals and national poverty reduction strategy became apparent when the Land Act Implementation Study (LAIS) presented its findings in August 1999. LAIS was conducted for the government by a team of 17 international and local consultants. It set out to assess the likely social, economic and environmental impacts of the Act, and examined the technical, legal, institutional, and financial requirements for its implementation. Broadly, it concluded that the cost of the Land Act was enormous while the social, economic and environmental benefits were likely to be negligible, and the bulk of the benefit would be for the wealthy. Although many of the study's assumptions and findings were contested, the fact that there was not adequate funding available for the Act as it stood was incontestable. Implementation stalled.
Four main points of contention between the Land Act (LA98) and the national poverty reduction strategy were highlighted by the study.
1. The overall estimated annual cost of implementing LA98 was six times what was available at that time to the land sector.
2. The demand for, and likely economic benefits from, providing certificates of ownership to customary landholders were judged to be uncertain, and not adequate to justify the cost of providing these services throughout the country.
3. Full implementation of the Land Fund (redistribution) component was estimated to require US h 700 billion (Uganda shillings: around US$500 million at the time). Aside from being unaffordable from the Ugandan Government budget, it was contended that the most obvious benefits would be windfall cash gains to the already wealthy landed class, loans for tenants would be likely to be unpopular since few tenants would wish to acquire a debt in order to obtain a registrable right when they already had security of tenure, and government intervention in the land market would have an undesirable impact on land prices.
4. LAIS highlighted the gap between the financial and human resources required for district provision of technical land services as provided for in the Act, the likely demand for these services in the short and medium term, and the cost recovery potential under the proposed system of fees and charges. LAIS concluded that such an arrangement would be unsustainable, and in an effort to provide an alternative framework, proposed the privatization, or agentization, of some aspects of technical land services within a fiscal framework that would enable them to cover the costs of some services.
The Ministry of Lands considered the study and sought to refute some of the arguments of LAIS with reference to mistaken assumptions on cost and on the implementation structure, and the political need to resolve the Kibale issue and to build capacity for decentralization. However, the basic argument that the Act as it stood was unaffordable could not be refuted. The LAIS study was circulated widely among stakeholders, and its conclusions on the likely contribution of the Land Act to poverty reduction negatively influenced perceptions in the Ministry of Finance and among the donor community. It was at this stage that the development of clear links between the land-reform process and national poverty goals became a necessity. The Ministry of Lands was directed by Cabinet to develop a plan for the land sector, including the implementation of the Land Act, taking into account available resources and capacity and matching land sector policies to poverty reduction.
THE LAND SECTOR STRATEGIC PLAN, THE PLAN FOR THE MODERNIZATION OF AGRICULTURE AND THE POVERTY ERADICATION PLANS
Renewed efforts were made to place land issues squarely within the framework of national poverty reduction and to develop realistic strategies for implementing reforms in the land sector. A draft sector plan, known as the Land Sector Strategic Plan (LSSP) was developed, which mapped out new strategies for addressing the difficulties highlighted by LAIS, anchored firmly within macroeconomic imperatives including the available resource envelope. The key issues for the sector plan were to develop:
a workable strategy for demarcation and certification of customary land and occupation rights;
an affordable and realistic means of providing decentralized land services;
an affordable structure for resolution of land disputes;
a strategy for utilizing the Land Fund to address the most pressing environmental and social problems.
The strategies of the LSSP were informed by the goals of national development as expressed in the revised PEAP, the Plan for the Modernisation of Agriculture (PMA), decentralization policy and the Medium Term Competitiveness Strategy (MTCS).
The PEAP was being revised as the LSSP was being prepared, but any synergy between the two processes was still limited. The current PEAP does not therefore reflect the LSSP framework, but links have been established between the goals of PEAP and the strategies of LSSP. Land reform broadly has been established as having potential impacts on all the four PEAP goals of creating an enabling environment for rapid and sustainable economic growth, good governance, direct income-enhancing actions, and direct enhancement of the quality of life of the poor. In relation to land, PEAP II has progressed from PEAP I and is less concerned with security of tenure (except for women's land rights, which remain a concern) and more concerned with redistribution to the land poor.
The principal national development arguments for the land-reform process are still couched in terms of increasing the ability of the poor to raise their incomes, and are linked closely to the recent formulation of the PMA. Although the focus of the PMA is agricultural development, its content is that of a wide-ranging strategy for rural development, incorporating a more holistic, "livelihoods"-based approach. For PMA, land reform is an important factor for two key ends: providing opportunities for the resource poor to access land via redistribution, resettlement, or rental and lease markets; and increasing and sustaining the productivity of land currently under agricultural use. Thus, both PEAP II and PMA focus on the need for land redistribution and providing opportunities for access to the resource poor, a feature that has scarcely been on the agenda of the reform process to date.
The LSSP is also concerned with linking the land sector to the other goals of PEAP II, particularly in terms of good governance and economic growth. Specifically, the LSSP links improvements in land administration and management to the potential to increase local government revenue and therefore contribute to improved capacity for local funding of basic services, and links rehabilitation of the land registry and existing land records to the encouragement of investment.
Content of the Land Sector Strategic Plan
The LSSP includes six broad themes: policy and legal framework; sustainability of land use; land tenure and protection of land rights; land administration and land information; decentralization; and sector efficiency and self-sufficiency. These are linked and backed up by a three-year budget plan linked to the national resource envelope for land. The plan explicitly acknowledges the ambiguity concerning the impact of customary registration and other aspects of land sector reform on poverty, and therefore adopts a pilot approach to many aspects of the programme, with a strong monitoring and learning component to feed into subsequent planning.
A number of important shifts of emphasis and priorities, together with modifications to the structure for land administration and dispute resolution, are contained in the LSSP.
(1) In relation to tenure insecurity in customary areas, the LSSP adopts a strategy of systematic demarcation on demand by communities. Adjudication and demarcation methods are to be piloted in a number of areas, with the eventual aim of providing this service to parishes in accordance with subcounty development plans and incorporating a community contribution to the process. The benefits of this approach over supporting individual certification on demand are twofold: there is less danger of the poor being dispossessed through a process in which all rights are adjudicated simultaneously, and providing certification services in this manner allows for significant cost savings (the cost of systematic demarcation is estimated at only 10 percent of the cost of providing the service on an individual basis).
(2) The tenure difficulties arising from colonial gifts, and the landlord - tenant relationship and informal settlement, will be addressed through the use of the Land Fund in two ways. The Land Fund will be utilized first to assist in the redistribution of land rights in Kibale district, whereby land will be purchased by government and redistributed. The criteria will be to target redistribution towards the goal of increasing access to land for the rural poor. A second use for the Land Fund will be to assist with the readjustment of rights in other mailo areas and peri-urban areas, again linked to criteria of poverty reduction. Land Fund money will be used to facilitate negotiations between landowners and tenants whereby land will be swapped between the two parties to permit each to obtain clean title and to facilitate proper planning.
(3) A link between the challenges of decentralizing land administration and increasing accessibility of registration directly to the potential economic benefits of improved land information. Systematic demarcation and certification/registration will provide spatial information with a variety of potential uses. This also requires investment in the existing survey and registration records systems, and links the decentralization of these records to the creation of locally accessible and accurate land information to be used for community land use planning and dispute resolution, for the planning of infrastructure and other services, and for enhancement of the local revenue base. A three-year strategy for increasing district capacity has been developed under which districts will continue to share services until such time as adequate capacity can be established countrywide. However, the option of privatization will also be explored during the initial three years of the plan.
As a result of these modifications, and in order to reduce costs, a number of changes to the implementation structure as envisaged by the Land Act will need to be made, reducing the number of new institutions to be created and maintained. Operating savings from these modifications are US h 12 billion. As a result of the development of LSSP and the development of a sector budget clearly related to the goals of the PEAP, there was a threefold increase in allocations of resources from the Treasury to the land sector under the Poverty Action Fund (PAF) in 2001 - 2002. This increase in allocation was assisted by the move towards budget support on the part of some donors, and in general the aim of the LSSP is to move towards a sectorwide approach to funding.
LESSONS FROM THE PROCESS
Uganda's experience in land reform has demonstrated bluntly the importance of integrating land sector issues and programmes into the overall framework for national development. As a result of the failure first to assess the resource implications of the Act at the drafting stage and second to link the potential impact of implementing the Act on poverty reduction goals, implementation in Uganda has been somewhat hampered by lack of resources. Related to this, and more important in many ways, is the fact that the Act was drafted without any plan for its implementation. The difficulties faced in, for example, establishing land tribunals or making the Land Fund operational were only partly financial. Rather, the delays have resulted from a combination of the inherent ambiguity of many of the provisions of the Act and a lack of a clear, prioritized and sequenced plan for their implementation. The delays in implementation have had some clear social and economic costs, principally through the absence of land dispute resolution mechanisms and the inability of those faced with eviction to obtain legal advice. The extent of unlawful evictions in the period since the enactment of the legislation is not known, but it is apparently a significant problem, as is the amount of disputed land effectively taken out of production pending proper resolution of land disputes. Larger conflicts are known to have caused loss of life. Another potential cost of the implementation delays is in the effect of unrealized expectations and the perpetuation of misinformation on the objectives of the law on the willingness of people to participate in land sector programmes. The poor perception of land policy and land sector agencies among sections of the public may yet affect the success of programmes under the LSSP.
Nevertheless, some positive lessons can be drawn. Although the land-reform process has not always been as participatory as some demand, both the implementation process to date and the development of the LSSP represent a significant broadening of stakeholder involvement in the land sector, from multistakeholder task forces to stronger links with district land agencies and local government in general. The formulation of the LSSP involved wide consultation and this should provide a more robust basis and greater legitimacy for its implementation. Through the involvement of these stakeholders, including civil society groups, technical land officials have been exposed to a range of perspectives and have been able to draw on a wider range of skills in the implementation and planning processes. For example, the public awareness campaign has been developed and implemented almost entirely by civil society organizations with input from media specialists.
The LSSP does not represent a resolution of all the priority land issues outlined at the start of this paper. It does, however, provide a framework for the resolution of many of them in line with the national poverty reduction strategy. Because of the existence of outstanding issues, there is still a need for development of a national land policy statement clearly linking land reform to poverty reduction strategies. Development of a land policy is one of the activities of the LSSP. In addition to the myriad challenges to be faced as Uganda attempts implementation of the LSSP, there are three principal unresolved issues at policy level: the land rights of women, redistribution, and the framework for urban and peri-urban land rights and management.
Much of the public debate about land in Uganda at the time of the drafting of the Constitution and the Land Act was heavily influenced by the issue of what to do about the past. Concerns about the likely distributive effect of expanding the market in land were very much secondary. One of the key pressure points for the kind of broad-based economic growth that will benefit the poor is likely to be the gender division of access to and control of key economic assets, including land. In Uganda this issue has been especially divisive, setting advocates of increased ownership rights for women against the entrenched political, economic and social superiority of men. This paper cannot go into the details of the debate as it has played out so far. However, it can be noted that there is already a substantial body of empirical evidence demonstrating the negative effects of women's tenure insecurity on their incentives to produce, especially for the market, and on the welfare of their families as increased market penetration leads to a reorientation towards cash crops. The original "Lost Amendment" has been redrafted to take account of many of the original objections, but the provision has now been "diverted" to domestic relations legislation, where its full impact is unlikely to be realized. At a policy level, there is need for clear and unequivocal recognition that access rights for women are not adequate to support poverty reduction or agricultural transformation, and that these rights need to be entrenched in the main body of land law.
Both the PEAP and the PMA, on the basis of analysis of land distribution, note the importance of increasing land access for resource-poor households. As has been discussed, redistribution was mainly debated in relation to specific historical circumstances in the making of the Land Act, and scope for redistribution is limited within the current legal framework. This issue should be re-examined in the land policy process to determine what possibilities are available for redistribution within the framework of the Constitution. For example, what should be the policy on existing leaseholds when the lease term expires, and what should be the policy on leasehold land over 100 ha that is not converted to freehold? How much such land is likely to become available? There is little information available currently on the amount of land under leasehold, its location, current use, market value, or likelihood of conversion to freehold. Clear information will be needed on this to inform the land policy process.
The land-reform process has focused mainly on rural land issues, as the vast majority of land in Uganda is rural. Urban land problems are significant and affect a substantial proportion of the population. The impact of the privatization of land through the Constitution on planning of urban areas is a critical policy issue, as is the effect of the regularization of tenure status on the large number of urban dwellers who have gained access to land through informal mechanisms. Land speculation is an increasing problem, fuelled in part by the sale of government properties. While some of these issues will be handled through the development of the land use policy, government should consider developing a full urbanization policy addressing not just land issues but also population issues, provision of services, employment, security, public health and the range of issues affecting urban development.
Although the process of land reform in Uganda has not followed an exemplary course, with a smooth flow from policy to legislation to implementation, the process of integration undergone since the passing of the Land Act makes it much more likely that land reforms will eventually be implemented in a relatively sustainable manner and with a high level of local ownership over the process.
Hunt, D. & Eturu. L. 1999. Economic appraisal, Land Act implementation study, Annex 1.
Leatherdale, J. & Palmer, D. 1999. Land services appraisal, Land Act Implementation Study, Annex 5.
McKinnon, J. & Reinikka, R. 2000. Lessons from Uganda on strategies to fight poverty. Washington, DC, World Bank.
Makerere Institute of Social Research (MISR). 2000. Land Act monitoring report 1. Kampala, Makerere Institute of Social Research.
Ministry of Finance, Planning and Economic Development (MoFPED). 2001. Uganda poverty status report. Kampala, Ministry of Finance, Planning and Economic Development.
Moyini, Y. 2001. National land use policy issues paper. Kampala, MWLE.
Oates, P. & Ecaat, J. 1999. Strategic Environmental Assessment, Land Act Implementation Study, Annex 4.
Ovonji-Odida, I., Muhereza, F., Eturu, L. & Wily, L. 2000. Land, gender and poverty eradication: is there a case for women's coownership of land? Kampala, Uganda, Ministry of Water, Lands and Environment.
Platteau, J.-P. 2000. Does Africa need land reform? In C. Toulmin and J. Quan, eds. Evolving land rights, policy and tenure in Africa. London, FID/IIED/NRI.
Republic of Uganda. 1993. The report of the Uganda Constitutional Commission. Kampala, Republic of Uganda.
Republic of Uganda. 1995. Constitution of the Republic of Uganda. Kampala, Republic of Uganda.
Republic of Uganda. 1997. Poverty eradication action plan: a national challenge for Uganda, 1997 - 2000. Kampala, Republic of Uganda.
Republic of Uganda. 2000. Plan for the modernisation of agriculture. Kampala, Republic of Uganda.
Republic of Uganda. 2001. Poverty eradication action plan, 2001 - 2004. Kampala, Republic of Uganda.
Uganda Office of Statistics. 2001. Population projections. Kampala, Republic of Uganda.
World Bank. 1993. Country study: Uganda agriculture. Kampala, Ministry of Finance, Planning and Economic Development.
 An earlier version of
this paper was presented to a World Bank Conference on Land Reform, Kampala,
April - May 2002|
 Landlessness does not equate to poverty for many households, which may have access to other forms of compensating income or wealth, for example.
 "Distress sales" would equate with sale of land to meet immediate household needs, and not necessarily those associated with a shift to longer-term investment, for example investment in education.
 The specific area under discussion is now known as Kibale district, sometimes referred to as the "Lost Counties".
 In 1988, the Resistance Committees (Judicial Powers) Statute made provision for elected committees of nine members (RCs) in all villages to resolve minor civil disputes and customary law matters including land disputes, operating under the supervision of the High Court and with appeals lying with the Chief Magistrates Court of a district.
 At the time of writing, the LSSP is still under discussion by Cabinet. It is possible that some of the strategies will be further modified during this process. The draft LSSP, however, already forms the basis of activity within the land sector.
 Sector-wide approach: there is no single agreed definition of a sector-wide approach, but common elements include an agreed sector policy and plan, agreed management frameworks, common financial accountability mechanisms and donor coordination.