COMPARATIVE ANALYSIS OF LAND ADMINISTRATION SYSTEMS:

AFRICAN REVIEW

With special reference to Mozambique, Uganda, Namibia, Ghana, South Africa

By Clarissa Augustinus

WORK UNDERTAKEN FOR THE WORLD BANK, FUNDED BY DFID

January, 2003.


1. Background

This review is based on five case studies of countries in Africa, namely Mozambique, Uganda, Namibia, Ghana and South Africa. These countries were chosen because it was considered that they had undertaken innovative land administration exercises and/or there had been important project interventions. The five case studies were undertaken using existing material and no new research was done specifically for the study, which has meant that there are numerous gaps in the data. However, five case studies were done, more than in the other regions, to try and fill these gaps and create an overall picture. The Africa review had to dovetail with the other regions as much as possible. This was somewhat problematic as the majority of tenure in Africa is outside of the conventional land titling system and the focus of new land laws has been on new forms of title and tenure security, rather than on extending the conventional titling system. This review therefore describes the conventional systems, the new tenure forms under development and that the majority of rights are currently outside of these systems.

1.1 Factors that Distinguish Land Administration in the Region

Land administration in Africa consists of the conventional land administration systems based on land registration and; customary and/or informal land administration systems. As Table 1. illustrates, the vast majority of African countries’ populations use the customary land administration system, especially in rural areas. As most African cities are 30-80 percent informal (UNCHS:1990:4, 1991:3), most urban residents are also not using the formal land registration and administration system in urban areas.

Table 1: Formal and customary coverage
Title and customary / Uganda / Namibia / Mozambique / Ghana / South Africa
Customary coverage / 62% of surface but about 68% of population / Majority of population / 90% of transactions / 78% of total area / 10-13% of area but about 25-30% of rural former homeland population
Title/deed coverage / 12-15 % / Majority of surface area but not majority of population / 1-15% / Unknown / 80-90% of area but excludes at least 25-30% of population
No. of registered titles/deeds / 700,000 titles / Unknown / Unknown / 11,383 titles /deeds unknown / 6,996,658 deeds

Due to the complex nature of the cadastre and property rights, colonial land administration laws and regulations remain entrenched in many countries still to this day in Africa (United Nations:1997:4). In a number of countries, such as Uganda, Ghana, Namibia, Mozambique and South Africa, new land registration laws have been or are being introduced and discussed. These laws are an attempt to move away from colonial forms of land administration on the one hand, but also to develop land administration systems and laws that more closely reflect the social land tenures on the ground (customary and/or informal).

A review of the land registration, cadastral and land information management systems in Africa indicates that:-

·  Less than 1 per cent of sub-Saharan Africa is covered by any kind of cadastral survey (UNCHS:1991:3, 1990:4) and the case studies show that the vast majority of people in a country are instead using customary tenure forms (Uganda, Namibia, Mozambique, Ghana case studies).

·  Most African countries do not have a national land information management system using LIS/GIS (UNCHS:1991,1993, Uganda, Namibia, Mozambique, Ghana case studies).

·  Cadastral systems, generally in manual form, and with incomplete coverage, are supplying most of the available land information. (Okpala:1992; UNCHS:1991, Uganda, Mozambique, Ghana, Namibia case studies).

Some of the major problems with land registration systems in Africa are:-

·  There is a general lack of financial, technical and human capacity throughout Africa. Because the systems are under-resourced many of them are out of date, expensive to maintain and inefficient. (Durand Lasserve: 1997:4-5,12,16; Ezigbalike:1996:350; Okpala:1992:93-4, Uganda, Namibia, Ghana, Mozambique case studies).

·  “A World Bank study on Africa.. (showed that) ..if no dispute occurs, the process of land registration takes an average of 15 to 18 months, and that normally, a period of two to seven years is not uncommon. This lengthy and costly procedure.. (means that).. tens of thousands of land titles.. (are usually).. pending.” (UNCHS:1991:5, Uganda, Ghana, Mozambique case studies).

·  Most of the systems are centralized (UNCHS:1993, Uganda, Ghana, South African case studies).

·  Available information often relates only to the part of the city or rural area where formal legal procedures were used for planning (UNCHS:1998:4). Yet most decisions need to be made about the non formal and/or customary parts of the country, which are not covered by the cadastre (Okpala:1992:94, South Africa case study relating to former homeland areas, Uganda case study).

·  “Despite numerous initiatives during the last decade in sub Saharan Africa to set up new land information systems or to modernize existing ones, limited results have been achieved.” (Durand Lasserve:1997:12, Uganda, Namibia, Ghana case studies).

·  Where information exists, it is often spread among several government departments and accessing it is difficult. These institutions are usually public, highly centralized, not well managed, and have over-lapping responsibilities. (Durand Lasserve:1997:12, UNCHS:1998:4, Uganda, South Africa, Ghana case studies).

·  Many of the parcels in the land registration systems are cloudy and hold ambiguous information despite attempts to create land registration systems with certain, highly accurate information (UNCHS:1996:11; Durand Lasserve:1997:2; UNCHS:1993, 1995,1991:3;1990:4; UNECA:1996, Uganda, Mozambique, Ghana case studies).

To contextualise the information presented above and below further, a number of Africa’s characteristics that affect land administration are identified.

  1. Customary tenure is generally secure and titling is unnecessary. However, customary tenure also transforms under certain conditions and becomes less secure for customary residents –these conditions include, urbanization, ribbon development, cash crops. Also, the formal land registration system is not neutral and where titling is implemented often customary tenure people lose their rights (women and overlapping rights holders are very vulnerable).
  1. Land markets exist all over Africa, both in rural and urban areas. They are not a recent phenomena. They are not free land markets and the sale of land is limited to relatives (by blood and/or marriage), and/or ethnic/national groups, and/or religion in certain areas and/or to men. All of these sales generally take place outside of the formal titling system. There is an active rental market both in rural and urban areas, also to strangers who do not fit into the above categories.
  1. Due to the general weakness of the central state in Africa, it is not possible to implement at scale a centralised titling programme, or land use controls, or enforcement. Most implementation is only in the cities, especially in the capital city, and ad hoc.
  1. The patron-client relationship in relation to land can be very strong at national, ethnic, family, local government/politician levels.
  1. Women’s rights to land are often nested in that of the family. In countries where there have been wars, genocides etc. women often struggle to obtain land rights when the men in their families are deceased, because society understands their rights in terms of family/men’s rights.
  1. There are large scale problems around the flow of spatial information for land administration purposes both within government, between departments at national level, between national and lower level tiers, and between government and the private sector and users. Coordination is a critical issue. There are few comprehensive national digital systems in operation that contain land information for land administration purposes. Where they do exist they only include that part of the country covered by the cadastre, typically formal urban areas.
  1. Most countries have colonial forms of legal evidence requiring high standards and professional inputs. There are few registered professionals –many countries have less than 30.
  1. Many countries have started law reform to change the situation and introduce new forms of evidence and approaches, but the scale and comprehensiveness of change needed is huge and has not reached implementation at scale level. Systematic titling is not considered an option for most of Africa for a range of reasons.
  1. Many of the existing titles are cloudy and require legal processes rather than just administrative processes to transfer. In defending their rights people will refer to both the paper and to customary evidence to protect their rights, sometimes this is legal, occurs frequently in the legal system, or forms part of land reform procedures.

2. Comparative Analysis

The five countries reviewed have a range of tenure types, both colonial and introduced under new law.

2.1 Land Tenure Systems

Table 2: Tenure types –formal and informal

Tenure types / Uganda / Namibia / Mozambique / Ghana / South Africa

Freehold

/ Yes / Yes / No / Yes / Yes
Registered leases / Yes / Yes / Yes / Yes / Yes
Customary (not necessarily legal) / Yes / Yes / Yes / Yes / Yes
Occupancy rights / Yes / No / Yes / Unknown / Yes
Anti-eviction rights / Yes / No / No / Unknown / Yes
Group/family titles / Yes / Unknown / Yes in law / Yes / Yes
Modern ‘starter’ /provisional type titles / Yes -pilot / Not yet / No / Yes / No –planned
Adverse possession/’squatters rights’ / Unknown / Unknown / Unknown / No / Yes
State land ownership / Yes / Yes / Yes / Yes / Yes
Informal settlement / Yes / Yes –at least 30,000 individuals / Yes / A few / Yes
Tenants / Yes / Unknown / Unknown / Yes / Yes

The new tenure types which are considered Best Practice include occupancy rights (generally critical for customary tenure rights if they are not protected specifically), anti-eviction rights, adverse possession, family/group rights, and modern ‘starter’/provisional titles, which can be upgraded.


2.2 Legal Framework

Table 3: Legal systems

Legal systems / Uganda / Namibia / Mozambique / Ghana / South Africa
Legal legacy / British / South African/Roman Dutch / Portuguese
& socialist / English / Roman-Dutch
Title system / Yes / No / No freehold / Yes / No
Deed system / No / Yes / No freehold / Yes / Yes
Legal liability / Yes –state / Yes –private sector / None / Yes –title system only (yet to be implemented) / Yes –private sector
Colonial forms of legal evidence used / Yes / Yes / Yes / Yes / Yes
New forms of legal evidence also used / Yes –pilot / Not yet / Yes / Unknown / No
Legal pluralism / Yes / Unknown / Yes / Yes / Yes –former homelands

2.3 Technical arrangements and indications of functionality

Table 4: State of conventional systems

State of conventional systems / Uganda / Namibia / Mozambique / Ghana / South Africa
Time taken for new registrations / Years –large backlog for first titling / 3 years for first titling / 2-10 years (trying to implement 90 days) / 6 months if no dispute / Unknown
Time taken for registration of transfers / Unknown / Unknown / Unknown / Unknown / 10 days once in the registry
Geodetic / Poor / Fair / Poor / Fair / Good
Records / Poor / Fair / Poor / Poor / Good
Mapping / 30 years old / Unknown / 30 years old / 30 years old -majority / Good in 87-90%
Coverage / 12-15 % / Majority of surface area but not majority of population / 1-15% / Unknown / 80-90% of area but excludes at least 25-30% of population
No. of registered titles/deeds / 700,000 titles / Unknown / Unknown / 11,383 titles/ deeds unknown / 6,996,658 deeds
No. of land parcels still to be registered (if desired) / 5 million / Unknown / Millions / 15,000,000 / Millions
Spatial data infrastructure for country / No –under development / No / No / No / Yes
Technical staffing / Insufficient –less than 20 professional land surveyors / Insufficient- less than 20 professional land surveyors / Insufficient –less than 20 professional land surveyors / Insufficient / Good
Modern mortgage law / No / Partially / No / Yes –title system only / Yes
Donor funding for conventional system / Small amount / Small amount / Yes / Yes / None
Systematic titling schemes / No –systematic demarcation for spatial information (piloted) / No –will be done per informal settlement / No –sporadic titling and sporadic demarcation of communities / Yes / No –only for upgrading informal settlement projects

Wherever possible the costs related to the existing land registration systems and the new systems were identified and these are presented in Table 5 below. In general these figures were not available because some of the process is undertaken by the private sector, or public sector officials operated as ‘consultants’ undertaking private jobs, or new forms of titling had not yet started as a routine operation, or the accounting systems of the agencies were not yet capable of producing such information, and there are no systematic titling projects.

Table 5: Costs

USD / Uganda / Namibia / Mozambique / Ghana / South Africa
Total cost per parcel / Unknown / Unknown / Unknown / Base US$ 4 + ad valorem scale / Unknown
Cost per parcel for state / Unknown / Unknown / Unknown / Unknown / USD 4
Cost of registered transfer to user / Unknown / USD 2.87 / Unknown / Base USD 13 + ad valorem scale / Between USD 518 for USD 7,940 –USD 4,538 for USD 57,537 per property
Cost per survey / Unknown / Unknown / Unknown / Scale regulated -based on size and location of parcel. / Unknown
Costs for systematic titling unit / Not being undertaken / Not being undertaken / Doing sporadic only /

Unknown

/ Not done
Percentage of national budget / Less than 2% for whole department / Unknown / Unknown / Unknown / Less than 1% for whole department
Running costs of registry / Unknown / Unknown / Unknown / About USD 300,000 p.a. / USD 19,907,940 p.a.
Running costs of Surveyor General / Unknown / Unknown / Unknown / About USD 1 million p.a. / About USD 8 million
Running costs of land titling at local government level / Unknown / Unknown / Unknown / Not applicable / None
Annual funding by donors for land administration / Unknown / Unknown / About USD 2 million / Varies / None

2.4 Land market information