Informal Land DeliveryProcesses and Access to Landfor the Poor: A ComparativeStudy of Six African Cities

Carole Rakodi and Clement R. Leduka

International Development DepartmentSchool of Public PolicyThe University of BirminghamEngland

Department of GeographyNational University of LesothoRomaLesotho

Info

rmal

Lan

d D

eliv

ery

Pro

cess

es i

n A

fric

an C

itie

s

Pol

icy

Bri

ef -

6

Summary 1

Introduction 3- Why research informal land delivery processes?- Aims of the research- Starting points- The comparative study

Delivery channels : Land for residential use in African cities 13- Allocation of public land- Purchase of land through the market- Delivery of customary land through state-sanctioned channels- Delivery of land through customary channels to members of the group- Allocation by officials- Purchase of customary land- Self allocation

Conflict or harmony? Relations between informal and formal 24institutions and actors- Conflicts- The tactics of non-state actors- State tactics

Conclusion 37- Main conclusions- Policy implications

Footnotes 43

Contents

International Development DepartmentSchool of Public Policy, J G Smith Building,

The University of Birmingham,, Edgbaston, Birmingham B15 2TT, UK

Website: www.idd.bham.ac.uk

The research was funded by the UK Department for International Development. DFID supportspolicies, programmes and projects to promote international development. It provided funds forthis study as part of that objective but the views and opinions expressed are those of the authorsalone.

Summary

1

To improve policy and practice, a betterunderstanding is needed of how the informal systemsthrough which half or more of the residential land inAfrican cities is delivered operate, are evolving andinteract with formal land administration systems. Thisstudy analysed the characteristics of informal landdelivery systems in six medium-sized cities inAnglophone Africa: Eldoret (Kenya), Enugu(Nigeria), Gaborone (Botswana), Kampala(Uganda), Lusaka (Zambia) and Maseru (Lesotho).Its aims were to

assess the strengths and weaknesses of alternativeland delivery mechanisms, especially with respectto the extent to which they enable the poor andwomen to access land with secure tenureincrease understanding of the institutions thatunderpin and regulate transactions and are usedto resolve disputes in land

The main conclusions were that

i. Informal land delivery systems are in part acontinuation of earlier land administrationpractices and in part a response to the failures ofthe formal tenure and land administration systems.These failures include the low levels ofcompensation paid by government when itexpropriates land, as well as cumbersome andcostly regulatory procedures.

ii. Informal systems of land delivery are the mainchannels of housing land supply. In the past, inmany cities, they enabled all but the poorest toaccess land for self-managed house construction.Today, non-commercial channels for obtainingland are restricted and the vast majority ofhouseholds who obtain land through informalchannels purchase it. The plots are suppliedthrough subdivision and sale of land held undercustomary tenure (Enugu, Gaborone, Maseru),by owners and tenants of mailo land (Kampala),by the shareholders in landbuying companies(Eldoret), and semi-officially by party and localgovernment officials in and adjacent to regularisedinformal settlements (Lusaka). It is no longerpossible for poor households to access land fornew residential building, with a few, often minor,exceptions:

membership of an indigenous landowningcommunity (e.g. Enugu, Maseru)settlement in hazardous areas (e.g. Kampala)allocation of customary land or a serviced plotin Gaboronepayment for informally subdivided land ininstalmentspooling resources to purchase a plot, forexample in areas subdivided by landbuyingcompanies in Eldoretsome local residents in regularised informalsettlements in Lusaka.

Women generally get access to land throughmarriage. In older-established areas women plotholders are mostly widows. Women with meansincreasingly purchase land through informal channels.

iii. For many new households in contemporaryAfrican cities, especially the poor, the only wayof becoming the owner of a plot on which tobuild a house is through subdivision or inheritanceof a parent’s plot. In practice, most poorhouseholds are tenants.

iv. Informal land delivery processes are ofteneffective in delivering land for housing, becauseof their user-friendly characteristics and sociallegitimacy. This legitimacy derives from the widelyunderstood and accepted social institutions thatregulate transactions in these informal systems.These institutions are generally derived fromcustomary institutions, but have evolved overtime. In particular, in urban contexts, they haveoften borrowed from and mimic formal rules andprocedures, or take advantage of ambiguities andinconsistencies in formal rules.

v. Urban growth and development increase thepressure on such social institutions, and in somecases, they weaken and break down, leading toincreased tenure insecurity. In such situations,actors in land transactions and disputes seek touse formal institutions to protect their rights andinvestments. The extent to which formal legal andadministrative systems recognise and work withor resist informal practices varies.

2

Policy implications

i. Informal land delivery systems play a significantand effective role in urban residential land supplyin African cities and so should be tolerated andaccommodated. Their strengths should berecognised and built on. However, theirshortcomings should also be identified andaddressed.

ii. To encourage investment in both owner-occupied and rental housing, the tenure securityavailable to those who access land throughinformal delivery channels should be enhanced.In some circumstances, this may imply individualtitling, but wholesale titling is often notappropriate, for three main reasons:

Titling massively increases the value of urbanland, making it even less accessible to low-income groupsThere is rarely capacity in the formal regulatorysystem to adjudicate, survey and register largenumbers of individual titles. Rather than issuingtitles to a minority of landholders, theresources available should be used morestrategically, to guide urban developmentOwner-occupiers are unlikely to mortgagetheir homes in order to release the capital tiedup in property for other purposes. The reasonsinclude the absence of developed financialsystems, as well as the priority householdsgive to livelihood security and their desire tobequeath urban property to their successors.

In many countries, alternatives to universalindividual plot titles are available under existinglegislation.

In addition, threats to security often arise fromgovernment action, especially evictions.

Governments should provide basic short-termsecurity to residents in informal settlementsand, in the vast majority of cases, should ceaseto evict settlers and demolish houses

Security can be enhanced by public sectoragencies accepting innovations in proceduresand documentation that have emerged ininformal systems, because these are popularlyunderstood, widely accepted, cheap andprocedurally simple.Threats to women’s security of tenure shouldbe tackled through matrimonial as well asproperty law.

iii. The poor layouts and inadequate services thatoften characterise informal settlements can beaddressed by recognising such areas, paving theway for working with subdividers and sellers toimprove layouts and enabling the early provisionof basic services.

iv. As well as making it possible to charge users forservices, the registration of occupiers enableslocal governments to generate tax revenue.

v. To build on the strengths and address theweaknesses of informal delivery systems invarying local contexts, the formal landadministration system should be decentralised,in particular to provide for local registration ofland rights and transactions.

vi. Revised compensation provisions are needed,requiring governments to pay adequate and faircompensation when they expropriate land forpublic purposes from private or customary rightsholders. This would

Deter premature informal subdivision intendedto preempt arbitrary and under-compensatedexpropriation (e.g. Enugu and Maseru)Improve the operation of some state-ledsubdivision and allocation processes (e.g.tribal Land Boards in Gaborone)Increase the ability of governments to provideland for infrastructure or industry withoutantagonising local land rights holders (e.g.Enugu)Enable governments to increase the supplyof serviced land for low income housing.

A F R I C A N

cIT

ES

3

IntroductionWhy research informal landdelivery processes?

The colonial powers in Africa introduced urban landadministration systems that were modelled on thesystems of their home countries. The extent to whichindigenous tenure systems were understood,recognised and incorporated varied from colony tocolony, but it was generally believed that only aformal system based on a European model couldprovide a framework for urban development andprotect the rights of urban property owners (who atthat time were expatriates). These landadministration systems, which were inherited atindependence, are governed by formal rules set outin legislation and administrative procedures.However, the legislative provisions and theadministrative systems that were established toimplement them proved quite unable to cope withthe rapid urban growth that occurred afterindependence.

The state-led approaches to development favouredin the 1960s and 1970s were associated with large-scale public intervention in urban land deliverysystems. However, the cost of implementation andcompliance has been too high for low-incomecountries, cities and inhabitants. At their extreme,land and property markets were perceived asineffective or exploitative. These views weretranslated into attempts to de-marketise land bynationalisation and/or government control over landmarket transactions. Whether or not the conceptson which such land policies were based were sound,limited capacity at national and municipal levelsensured their failure. The subdivision and allocationof residential plots by governments has very rarelymet demand and attempts to regulate and registerall transactions in land and property have beenuniversally unsuccessful. As a result, most land forurban development has been supplied throughalternative channels1.

In the early years of rapid rural-urban migrationmany households, including poor households, wereable to get access to land to manage the constructionof their own houses for little or no payment, through‘squatting’ or similar arrangements. Followingresearch in the 1960s and 1970s, there was a feelingthat the processes of ‘squatting’ and the allocationof customary land by legitimate rights holders werefairly well understood. Upgrading projects of the1970s and 1980s were designed and implementedon this basis.

Most countries have now reversed some of the mostextreme versions of state intervention, but othercomponents remain despite serious implementationfailures. There is considerable doubt about whetherrecent attempts to improve land management willbe any more successful than previous approaches.In part, pessimism about the prospects for efficientand equitable urban land management arises fromthe continued lack of resources and capacity ingovernment, but it also stems from doubts aboutthe appropriateness of the principles and conceptson which recent urban land policies have been based.

Much research on land and property in Africantowns and cities assumes that the state has both theduty and the capacity to take on a majorinterventionist role in land management. Itconcentrates on documenting and explaining thefailures (and more rarely successes) of stateinterventions. Despite their significant role inproviding land for urban development, there hasbeen relatively little recent in-depth research onprocesses of informal land delivery or the institutions(rules and norms of behaviour) that enable them tooperate and that govern the relationships betweenthe actors involved.

Informal systems for delivering and accessing landoperate according to a variety of social rules thatare understood and complied with by actors in thesystem. These can be conceptualised as institutionsthat enable transactions to occur and regulaterelations between actors. They have been

4

documented for indigenous land tenure systems,generally with reference to the rural areas. However,not only do indigenous land tenure arrangementschange over time in rural areas due to changingsocial, political and economic circumstances, theyalso change when they come face to face with thedemands of urban growth. In part the changes arereactions to the demands for small plots forresidential development by both indigenes andstrangers. However, they are also adaptations to theintroduction by colonial and post-colonialgovernments of land tenure changes and formal rulesintended to regulate urban land development. Thusin urban areas, the social institutions that regulatetransactions in land and relations between the actorsinvolved are hybrids of formal and informal rules.

A F R I C A N

cIT

ES

However, the ways in which informal land deliveryprocesses are evolving in contemporary cities havebeen little studied. As a result, the land policy andadministration reforms on which many countries haveembarked since the 1970s have not onlyconcentrated on rural land but have also often beenill-informed and ineffective with respect to urbanland. In addition, they have lacked legitimacy, givingrise to difficulties of compliance and enforcement,in part because they are not based on anunderstanding of the social rules governing howpeople act in partly commercialised informal landsystems. To improve policy and practice, a betterunderstanding is needed of how both formal andinformal systems operate, interact and are evolving.

Maseru

Aims of the research

The aim of the project was to improve understandingof contemporary informal land delivery processesin six African cities and their relationships with formalland administrative systems. It analysed thecharacteristics of informal land markets and deliverysystems

to assess the strengths and weaknesses ofalternative land delivery mechanisms, especiallywith respect to the extent to which they enable

the poor and other vulnerable groups (especiallywomen) to access land with secure tenureto increase understanding of the institutions thatunderpin and regulate transactions and disputesin land, andto identify and explore implications for policy.

Six medium sized cities in Anglophane Africa werestudied: Eldoret in Kenya, Enugu in Nigeria,Gaborone in Botswana, Kampala in Uganda,Lusaka in Zambia and Maseru in Lesotho. In all ofthese cities informal land delivery systems are

5

A F R I C A N

cIT

ES

Starting points

Much research on urban land and property in Africantowns and cities assumes that the state has both theobligation and the capacity to intervene in landmanagement. As a result, analyses often focus onfailures to comply with formal state law andregulations, thereby elevating formal state law overother socially embedded rules on which actors drawto regulate transactions in land. In this research, weattempted to give non-statutory law and custom, aswell as social relations between actors in informalland delivery systems, equal status with state lawand organisations, and to focus on the predominantcontemporary processes of land delivery and notmerely the weaknesses of public land administration.To understand how the processes of land deliveryoperate and how transactions and disputes areregulated, we drew on insights from three types ofsocial analysis.

Social structure and the agency of socialactors

Many studies assume that the state plays a dominantrole in society and analyse urban development interms of actors’ compliance with formal state rules.It is assumed that the state has access to resourcesthat enable it to exercise power over other socialactors, who respond passively2. In practice,although power is unequally distributed, no actorsare powerless. Even dominated social groupspossess resources that enable them to activelyengage with other actors. These resources give themagency and therefore power, perhaps enabling themto influence the actions of dominant groups. Forexample, apparently powerless actors can re-interpret, use or challenge the formal rules specifiedby the state, creating opportunities for changes bothto the rules themselves and to the relationshipsbetween state structures and non-state actors, in thisinstance with respect to urban land and property3.If we are concerned to understand not just theexercise of power by the state and elite social groupsbut also what happens when apparently less powerfulgroups take action themselves, we need tounderstand the nature of relationships betweenvarious social actors. Institutional analysis can assistin developing such an understanding.

Analysing social institutions or rules

Social institutions govern the social, economic andpolitical relations between individual actors and,together with resources of various types, make itpossible for social systems to exist and function.They may be divided into formal and informalinstitutions.

Formal institutions are rules of the game that areexplicitly drawn up and defined, in particular statelaw. They specify rules to be followed in performingcertain activities or fulfilling obligations, in theoryenforceable because of state access to legal andcoercive power.

important. They also typify different colonial andpost-colonial policies, legal frameworks, governancearrangements and urban management experiences.

In this summary paper, the empirical and theoreticalstarting points for the research are first outlined,followed by the analytical and methodologicalapproach adopted. Each of the main channels ofland delivery is then discussed in turn, and theirstrengths and weaknesses identified, with referenceto a common set of criteria. The paper draws onselected empirical examples from the case studycities to illustrate ways in which the social institutionsgoverning relations between actors in informal landdelivery systems have evolved, as a result of conflictor accommodation. Finally, the main conclusionsand policy implications are identified.

6

Informal institutions, on the other hand, areembedded social norms and practices, includingcustomary rules4. They are revealed through actors’everyday practices, transactions and disputes5. Thisimplies a pluralist view of law, which sees non-statelegal forms, whether or not recognised in state law,as relevant to understanding practices and relationsrelating to land and does not give one system of lawgreater standing than another6.

Institutional analysis evolved to explain the operationof markets. It stresses the importance of institutionsor rules because of their roles in minimising the costof transactions or proscribing certain actions andbehaviour, thereby enabling markets to workeffectively7. It has been used to explain land disputesin informal settlements in Jordan8; land and housingmarkets and informal credit in Trinidad9; theprovision of infrastructure services to informalsettlements in Turkey and Egypt10; and the strengthsand weaknesses of informal land managementsystems in Tanzania and Lesotho11.

In urban studies, institutional analysis has primarilyfocused on the efficiency of rules in enhancing orconstraining market transactions, but has been rathersilent on issues of equity and power relationsbetween market actors. While sociological theoryacknowledges, as suggested above, that the pooractively engage with the better off and with the state,we need a way of understanding how such groupsorganise themselves and exert influence on morepowerful social actors.

Understanding the actions of social groups

To analyse how apparently powerless social groupscan organise themselves and assert their owninterests against more powerful groups, including thestate, it is helpful to conceive of a social group withfew formal resources as nevertheless capable ofgenerating informal norms and rules of behaviour,with which its members can induce or enforce

compliance12. This does not mean that such groupsare autonomous with respect to society as a whole– each is also embedded “in a larger social matrixwhich can, and does, affect and invade it..”13.Members of such a group may, however, jointly failto comply with rules that do not suit their interests,in particular state or formal rules.

The idea of societal non-compliance arises fromacknowledging the agency of individual social actors.Sometimes termed a ‘weapon of the weak’, non-compliance may be exercised by those without overtpower in order to subtly challenge the actions ofthose with formal political or organisational power14.If non-compliance becomes sufficiently widespread,it may even produce changes to government policyand practice, as Tripp argues it did in Tanzania withrespect to government attitudes towards informalsector activity15.

The interaction between land market actors may takea variety of forms: direct confrontation betweensocial groups and state agencies, attempts by thedisadvantaged to capture property rights or powerfrom the rich, or ‘covert’ non-compliance with formalrules. Direct challenges or protests by the poor, forexample through land invasions, are risky. Analternative may be to seek political patronage, inorder to access state resources, obtain redress forgrievances or provide backing for challenges toprivate property rights. Probably more often, quietindividualistic actions, with no discernibleorganisation or planning, are taken, in which actorsaim to avoid direct confrontation with the authorities.Informal settlements in the cities of developingcountries are often the outcomes of a combinationof tactics. To Razzaz, the most compelling attributeof those involved in gaining access to land throughinformal channels is their ability to take advantageof inconsistencies in state rules and enforcementstrategies16.

A F R I C A N

cIT

ES

7

Hypotheses

Based on insights drawn from this wider body ofliterature, the starting hypotheses of this researchwere, first, that the success of informal land deliverysystems in supplying between a half and seventy percent of all land for urban residential development,including land for the poor, can be attributed to theirpractical attributes and their social legitimacy17. Thepractical attributes, it was suggested, make thearrangements more suited to the needs of urban landrights holders and those seeking land for housing,including the poor, while wide understanding andacceptance of the social rules governing relationsbetween actors in the system serve to secure widercompliance than is common for formal landregulation.

Second, it was hypothesised that, as urbandevelopment proceeds, the informal institutions thatregulate land transactions and use change over time,vary between residential areas and sometimes breakdown. The pressures generated by urban propertymarkets and increased demand result in changes intraditional social institutions in order to make themmore suited to the circumstances of urban areas. Innewly urbanising areas, as shown by research in Dares Salaam, modified versions of traditional socialinstitutions underlie quite smooth processes ofconversion of agricultural land for urban uses.However, this research also showed that, as areasconsolidate and the density of developmentincreases, the informal social rules used to regulatetransactions and resolve disputes becomeincreasingly strained, and may eventually breakdown altogether18.

Third, the relations between disadvantaged groupsseeking land and state agencies are likely to involveboth conflict and accommodation. A distinctionneeds to be made between cases where non-compliance leads to conflict and where it leads toaccommodation, identifying the reasons for thesedifferent reactions. The outcomes of conflicts are

uncertain – in some instances, state agencies canenforce compliance, for example, through evictionand demolition, but in other cases, informal actorsmay prevail, perhaps by sheer weight of numbers,leading to recognition and upgrading of informalsettlements.

In their interactions, both state and non-state actorsmay adhere to the formal rules of state law andregulation, but may also be influenced by informalsocial norms and ways of regulating transactions.The relations between state structures and informalactors are, therefore, complex and dynamic,because, although formal state rules are in theoryomnipotent, in practice they are negotiable19.Understanding the circumstances in whichaccommodation rather than conflict occurs ispotentially important in identifying workableimprovements to urban land delivery systems.

A F R I C A N

cIT

ES

Busega, Kampala

8

The comparative study

Coordinated by Carole Rakodi of the University ofBirmingham and Clement Leduka of the NationalUniversity of Lesotho, the research was comprisedof in-depth studies of six cities. The criteria forselection of the cities are explained first, followedby a summary of the methodological approachadopted. Some basic information on each city is alsogiven.

Even within countries, the circumstances in differentcities vary. International comparisons are even morecomplex, because of the enormous variation in thecharacteristics of cities, their land markets and theiradministrative arrangements. In order to make itpossible to draw conclusions from comparativeanalysis, it is desirable to hold some factors relativelyconstant. In this research, it was considered that onesuch critical factor was the principles on which theformal legal system is based and it was, therefore,decided to select cities from Anglophone Africa.These include cities from eastern, southern and westAfrica, but exclude cities where recent or currentresearch on related issues is already under way20.

The cities are located in countries with differentcolonial policies with respect to land and urbandevelopment, arising in part from whether a systemof direct or indirect rule was adopted. They alsorepresent countries with very different post-colonialeconomic and land policies. These vary from free-market oriented Kenya to heavily state-ledBotswana and Zambia, and include countries whichhave been subject to military or single party rule inthe period since independence (Nigeria, Zambia,Kenya, Lesotho) as well as a country which hasbeen a multi-party democracy throughout(Botswana).

Some had attempted to nationalise land and introduceother reforms in the 1970s (although many of thesereforms had subsequently been reversed) and somehad not. The governance arrangements at bothnational and local level, including the role of

traditional authorities, vary between the countriesand the responsibilities for urban land delivery,regulation and tenure registration are differentlyallocated between government levels and agencies.

It was decided to avoid the largest cities, partlybecause they have already been relatively wellstudied, partly because their very active propertymarkets are not necessarily typical of cities in general,and partly because researching land in situationswhere land issues are highly politicised is very difficult.Thus the research focused on six medium-sizedcities, some capitals of relatively small countries, andothers secondary cities. The cities and the localresearchers are shown in the box opposite.

The aims of the project and the methodologicalapproach were jointly developed by the researchers.The local teams studied

a) The context for urban development, includingsocio-political and economic conditions, the legaland administrative framework for landadministration and the actors in land development

b) Land delivery systems, including the volume,types and regulation of land transactions in eachdelivery channel21. Although the researchfocussed on informal land delivery, formalsystems were also investigated (mainly usingsecondary sources). It is well known that,although the performance of formal deliverysystems is variable, they have often not deliveredland with appropriate characteristics and insufficient volumes to meet the demand for housingland. A review of their effectiveness was,however, felt to be necessary as a basis forcomparison with informal delivery channels.Moreover, there is often no clear divisionbetween formal and informal systems.

c) The characteristics of those accessing landthrough the alternative delivery systems, in orderto assess whether the channels identified did inthe past and continue today to supply land to the

A F R I C A N

cIT

ES

9

urban poor and other disadvantaged groups,especially women.

d) The authority structures, both those vested in thestate and those associated with other socialinstitutions, and their roles in land transactionsand dispute resolution

e) The institutions regulating transactions anddisputes in land, including an assessment of theextent to which these function well and arelegitimate in the eyes of actors in land delivery

f) The effects of urbanisation pressures on landmarket operation and regulation, and the extentto which the channels through which most land isdelivered for residential use can be relied on tocontinue to meet the demand for secure tenurein appropriate locations as urban growthproceeds.

In each city, a city-level analysis was complementedby detailed studies in three informal settlements – aperipheral developing area, a partly consolidatedarea in which active subdivision and developmentwas still under way, and a consolidated inner cityarea with a relatively high density, where pressureson land might be expected to produce a higher levelof problems and disputes. A combination ofquantitative and qualitative methods was used,

A F R I C A N

cIT

ES

Eldoret, Kenya Rose Musyoka, Department of Physical Planning, Government of Kenya

Enugu, Nigeria Cosmas Uche Ikejiofor, Federal Ministry of Works and Housing, Gusau,Zamfara State, Nigeria

Gaborone, Botswana Faustin Kalabamu, Department of Architecture and Planning, andSiamsang Morolong, Department of Law, University of Botswana

Kampala, Uganda Emmanuel Nkurunziza, Department of Surveying, Makerere University

Lusaka, Zamba Leonard Chileshe Mulenga, Institute for Social and Economic Research,University of Zambia

Maseru, Lesotho Clement Leduka, Department of Geography, National University ofLesotho

drawing on both secondary sources and primarydata collection. In each of the case studysettlements, a sample survey of plotholders wascarried out using a structured questionnaire. Thesurvey was complemented by key informantinterviews and a series of focus group discussions.

Finally, each team enlisted the services of a lawyerto provide background on law and court casesrelevant to understanding urban land issues in generaland the resolution of disputes in particular, includingcases initially dealt with through informal orcustomary mechanisms.

Findings and policy issues were discussed atworkshops in each of the cities, to obtain feedbackfrom relevant stakeholders and make a contributionto current debates about land policy andadministration in each of the countries studied. Theresearch teams generally identified some of the policyimplications of their findings rather than makingdetailed recommendations, because the researchersall believe that policy formulation and legislativechange should be negotiated processes involving allthe stakeholders in land management. Followingthe workshops, full reports of the studies and policybriefs were published and are listed on p. 45.

10

A F R I C A N

cIT

ES

Eldoret, Kenya

With a population of about 197,000 in 1999,Eldoret is the fifth largest town in Kenya anda major regional administrative, commercial,educational and industrial centre. It is thecentre of Uasin Gishu District and isadministered by an elected Municipal Council.Located on poor quality land in an otherwisehigh potential agricultural region in the so-called ‘white highlands’, it developed as anagricultural service and agro-processingcentre for the surrounding Europeancommercial farming region. Selected as agrowth centre and politically favoured in the1980s, it attracted public investment ininfrastructure and industries.

Although the publicly owned land on whichthe town initially developed has mostly beenbuilt on, the urban boundary has beenextended to incorporate privately ownedfarmland. Today, there is limited public landfor subdivision and a cumbersome formal landdelivery system.

The availability of formerly European-ownedfarms in and around the urban boundaries andpost-independence encouragement toKenyans to purchase private landholdingshave led to the purchase of farms bylandbuying companies formed for thepurpose. Transfer of freehold title is followedby subdivision and sale: formal in the case ofhigh-income developments, or informal formiddle and low-income purchasers. Largenumbers of plots have been provided in thisway.

Enugu, NigeriaOriginally a coal mining town, Enugu laterbecame more important as a railway andadministrative centre. Its population in 1991was just under half a million and today maybe between 800,000 and 1,000,000. Capitalsince 1991 of Enugu State, today the city hasadministrative, educational, industrial andcommercial functions. The city is currently splitbetween three Local Government Councils,each with a directly elected Chairman andcouncillors elected on a ward basis. However,many important land-related and utilityservices are provided by State agencies.

Indigenous landholding groups ceded someland for mining, railway and housingdevelopment in the early years of Enugu’sdevelopment. However, the colonial systemof indirect rule left most land in the hands ofthe indigenous groups. Today, there is littleundeveloped land in public ownership andpublic agencies use expropriation powersunder the 1978 Land Use Decree to obtainland for public purposes, including industrialestates and major infrastructure.

The indigenous groups and families haveformally subdivided and leased large tractsof their land. Nevertheless, they retainownership of much land both within the builtup area and on the outskirts of the city. Familyheads and traditional rulers of the variouslandowning communities have to secureagreement of the family or group to thedisposal of farmland, while homestead landis retained for use by the family and itsdescendants. Farmland is subdivided and soldto individuals or speculators. A large volumeof informally subdivided land for residentialdevelopment is thus provided for groupmembers as well as middle and upper incomepurchasers.

11

A F R I C A N

cIT

ES

Gaborone, Botswana

Gaborone is the capital of Botswana and hada population of about 186,000 in 2001. Ithas mainly developed on state land, enablingrapid and large-scale subdivision of publiclyowned land for housing for all income groups,assisted by buoyant government revenuesderived from Botswana’s mineral wealth.Surrounded by privately owned commercialfarms, when additional land has beenrequired, the government has been able topurchase a large farm and has alsooccasionally acquired areas of tribal land.

With the exception of Old Naledi, an earlylabour camp, which has since been regularisedand upgraded, informal settlements within theurban administrative boundary have not beentolerated. The obstacles faced by householdswishing to obtain a residential plot areconsiderable. They include a ban on newconstruction between 1982-7 because ofwater shortages, long waiting lists andinfrastructure costs.

As a result, in recent years there has beenrapid subdivision and development in areasof tribal land outside the administrativeboundary to the west and east of the city(Mogoditshane and Tlokweng respectively).In theory this development is under the controlof Tribal Land Boards established by thegovernment for this purpose, although thesystem does not operate smoothly. Theelected Gaborone City Council and the twodistrict councils within whose boundaries themain areas of informal settlement lie(Kweneng and Southeast) have limitedresources. As a result, central governmentretains the main policy and administrative rolesrelated to land.

Kampala, UgandaKampala is the capital of Uganda, with a 2002population of 1.2 million. Capital of theBuganda kingdom since the 1700s, colonialrule transformed it into a divided city, partgoverned by the Kabaka (king) and part bya local council established by the colonialgovernment to administer the Europeansection of the city. Much land was ceded tothe Crown, while the Baganda chiefs weretransformed into a ruling landed oligarchy,exercising extensive control over mailo land,at the expense of both the Kabaka and thepeasantry. While colonial indirect rule initiallydepended on the chiefs for localadministration, by 1920 they had become lessnecessary and their power declined, althoughthey retained their status as big landlords.

The Kabaka’s attempts to retain control overthe kingdom’s land and administrative power,the chiefs’ determination to protect their owninterests, and the peasantry’s struggles torestore the land rights eroded by colonialadvancement of the chiefly class have beenimportant in both the pre- and post-independence periods. The destructive ruleof Idi Amin from 1971 to 1979 led to a rapidexpansion of the informal economy, urban-rural migration and near-collapse of the civilservice. The current regime has struggled torestore economic growth and state institutions.

A unified administration for the city ofKampala was established in 1966/7, andlower layers of local government put in placeby the current regime in 1986. However,parallel systems of land tenure andadministration persist and little progress hasbeen made with implementing the 1998 LandAct provisions intended to regulate theownership and use of land and simplifyownership and occupancy systems.

12

A F R I C A N

cIT

ES

Maseru, Lesotho

With a 1996 population of 140,000, Maseruis the capital of Lesotho. The innermost partof the city, within the original 1905administrative boundary, developed oncolonial government reserved land. This wasinherited as public land at independence.

The city is surrounded by villages andextensive informal settlements have developedon former agricultural land held undercustomary tenure arrangements. Familiesretain ownership and use of their homesteadland, while masimo (fields) are subdividedfor sale. Approximately 70 per cent of all landdemand is met outside the formal state deliverysystem. The government’s reaction hasgenerally been benign neglect, punctuated withinstances of intolerance marked by evictionsand demolitions. These have usually occurredwhen the government has had financialresources for land servicing and development,mainly from donor funds.

In order to curb the process of informal landdevelopment and loss of agricultural land,especially in the peri-urban areas, newlegislation was put in place in 1980, the LandAct 1979. This Act effectively nationalised allland, with rights to be leased from the state.It also extended the urban boundary toincorporate large areas of informal settlement,with the intention of establishing controls overfurther subdivision. Dogged by ambiguitiesand implementation problems, few of its aimshave been achieved. A new land bill has beendrafted, but has yet to be enacted into law.

Lusaka, ZambiaThe capital of Zambia, Lusaka originated asan agricultural service centre on the railwayline leading to the Copperbelt. Along therailway, a belt of land was taken into Crownownership and subdivided into commercialfarms for Europeans. Only at some distancefrom the urban centre was there any land inindigenous ownership, and much land in andaround the town was retained in publicownership. In addition, much of the land nearthe central business district was retained inpublic ownership because it was unsuitablefor farming and costly to develop. Asindependence neared, some Europeanfarmers permitted shack development on theirland, while many European farms wereabandoned.

The relaxation of migration controls and post-independence growth in wage employmentled to rapid population growth. Many of thenew households were able to obtain free landand to manage the construction of their ownhouses, although the provision of infrastructurein the informal settlements had to awaitregularisation and upgrading, starting in the1970s.

It was important to the rival political partiesof the 1960s and 1990s (and to UNIP underthe one-party state of the 1970s and 1980s)to secure the support of residents in informalsettlements. Thus party and City Counciladministrative structures developed side byside in these areas. Despite their designationas Housing Improvement Areas, Zambia’seconomic and political difficulties and the lackof local government financial andadministrative capacity have hindered theachievement of improved living conditions fortheir residents.

13

Delivery channelsLand for residential use in African cities

The studies identified the different channels throughwhich land for new residential development is madeavailable in each of the cities and assessed thestrengths and weaknesses of these alternativechannels. This assessment used a number of criteriasuggested, first, by the research questions andhypotheses, and also by the research participants’responses with respect to the attributes that theyvalue in urban residential land. These criteria were:

i. Scale: has the channel delivered land in sufficientvolume (and in appropriate locations) to meetthe demand for housing land from a rapidlygrowing urban population, is it continuing to doso today, and what are the prospects of itcontinuing to do so in future?

ii. Cost: has the channel delivered housing plots ata cost that can be afforded by people seekingland for housing, especially those with middle orlow incomes? Is it continuing to do so?

iii.Security of tenure: has the channel deliveredhousing plots with sufficient security of tenure toencourage owners to invest in housing? What arethe threats to security and can owners deal withthese threats and retain their rights?

iv. Access to disadvantaged groups: has thechannel in the past and today delivered residentialplots to disadvantaged groups, especially poorhouseholds and women (whether they are headsof household or not)?

v. Service provision: has the delivery of landthrough each channel been accompanied by theprovision of infrastructure and services, either inadvance, on subdivision or subsequently?

vi. Dispute resolution: are there widely availableand socially legitimate means of dispute resolutionavailable to those accessing land through eachchannel?

The research bore out the general view that mostlandholders in Africa’s cities both in the past andtoday obtain access to housing plots through informalchannels of land supply. In the past, many were able

to obtain plots through non-commercial channels,through squatting on publicly owned land or landabandoned by private owners, or membership ofindigenous rights-holding groups. The availability ofsuch channels enabled many relatively poorhouseholds to obtain a plot (although not necessarilythe poorest). Today, the research showed that it isno longer possible for poor households to accessland for new residential building in urban areas, withsome relatively minor exceptions:

Members of indigenous land owning families andcommunities in EnuguIndividuals in Kampala who claim wetland areas,initially for cultivation and then for building, atconsiderable risk to themselves and theirinvestmentsPeople who pool their resources to buy a shareor part of a share in a land buying company inEldoretPoor households who are allocated land for freeby the tribal Land Boards on the periphery ofGaborone, although the process operates veryslowly because the Boards lack the resources tospeed it upThose allocated a plot in a public-privatepartnership serviced plot programme inGaborone, although they are only able to accesssuch a plot after a long waitLocally resident households with good partypolitical or official connections in areas in Lusakawhere party or local government officials aresemi-officially subdividing publicly or privatelyowned land in or adjacent to informal settlementsdesignated as Housing Improvement Areas.

The vast majority of households in contemporarycities obtain land for residential use through purchase.The sources of plots include

Sales of customary land in Maseru, Enugu andGaboroneInformal subdivision by land buying companiesin Eldoret

14

Informal subdivision by mailo owners and tenantsin KampalaInformal subdivision and ‘sale’ by party or localgovernment officials within or adjacent tosettlements designated for regularisation inLusakaPurchase of undeveloped land from existingplotholders.

Access to land, therefore, is restricted very largelyto households with the necessary financial means topurchase it. These are primarily middle and upperincome households. Low-income households thathave some income (not the poorest) can oftennegotiate flexible methods of payment for landdelivered through informal channels, includinginstalments. In some cases, the payments are token‘fees’ to party or government officials rather than amarket ‘price’ for the land. Nevertheless, for manynewly formed households, especially the poor, theonly way of accessing land is through

plot sharing, either from the outset (e.g. by buyinghalf a share and thus half a plot in an informalsubdivision in Eldoret) or through the subdivisionof a plot by a parent for a child (Eldoret,Kampala, Maseru).inheritance, at least until the plots are too smallfor further subdivision and sharing amongstchildren, at which point the prospect of beingable to inherit a plot will decrease.

In practice, most poor households are tenants.

Many households who are unable or do not wish toacquire land to construct their own houses purchaseexisting properties, often in informal settlements. Therelative importance of markets in secondhand housesincreases as cities and settlements age, but thesemarkets were not the main focus of this research.

The grounds for the main conclusions will now beelaborated by examining the alternative channels forland delivery in turn.

Allocation of public land

The allocation of public land for housing is of varyingsignificance. In Gaborone, it always has been andcontinues to be the main source of residential landwithin the urban boundary. There have been servicedplot programmes in Maseru, Eldoret, Lusaka andKampala, often when external funding is available,but their scale is limited. In Kampala, customaryoccupiers of land taken over by the Crown underthe 1900 Buganda Agreement retained their usufructrights and subsequent legislation has reinforcedprotection of their rights. Today, land administeredby the Kampala District Land Board (municipal land– 15 percent of the city’s area) and the Uganda LandCommission (government land – 20 percent) is fullysubdivided and occupied.

Because of the historical development of Enugu, thepublic sector has never owned much land in the cityand land for public purposes has to be purchasedfrom private rights holders, curtailing thegovernment’s ability to subdivide land for housing.In addition, the charges for allocation of plots includean application fee, survey fee, development premium(supposedly to cover the cost of infrastructureinstallation) and fees for the preparation of thecertification of occupancy. The total amount payableplaces such plots, even if available, beyond the reachof low-income households.

The reasons for the large contribution of this channelin Gaborone include government capacity, the natureof the political regime and the patterns of landownership on the outskirts of the urban area. Thegovernment of Botswana (and, by extension,Gaborone City Council) has been relatively wellendowed financially because of the discovery ofdiamonds. The country has had a relatively stableregime, based on multi-party democracy, peacefultransitions of power and elite consensus. This hasenabled it to manage its economic resources relativelywell, buy in expertise when required and adoptdevelopmental policies that have spread the benefits

A F R I C A N

cIT

ES

15

of revenue from diamond exports broadly withinsociety. In addition, much of the land on the outskirtsof the city is divided into large commercial farms inprivate ownership, which has enabled the governmentto purchase a farm when land is required for urbanextension. One or more of these conditions are absentin all the other cities studied.

Thus in Gaborone, public land allocation has resultedin a significant volume of land for residentialdevelopment, but this does not apply in the othercities. When available, subdivided public landprovides a plot with security of tenure (formal freeholdor leasehold title), but the process of allocation issusceptible to the discretion of those responsible andtherefore sometimes to favouritism and corruption(see box). Land provided by the government is morelikely to be serviced, but the standards adopted andtherefore the cost is likely to reduce access to suchland by low income groups and never provides accessfor the poorest (for example in Maseru). A furtherproblem is that the system of public land supply isoften not well known and transparent, so that peoplewith low levels of literacy and little knowledge of thebureaucracy are marginalised. This is likely to includepoor people and many women. Women heads of

household qualify for publicly provided plots in allthe cities studied, although they are often under-represented in actual allocations because they findit more difficult to satisfy eligibility criteria such aswage employment or a minimum income. In somecircumstances, the regulations governing allocationallow women who are not heads of household toobtain land in their own names, although there isconsiderable social resistance on the part of bothofficials and women and their husbands to exercisethis right, since for a woman to acquire land in herown name is considered to indicate a weakmarriage. However, women married in communityof property (especially in Lesotho) cannot registerland in their own names without the permission oftheir husbands.

There are normally provisions for individuals withcomplaints against the system of public landallocation to appeal and for disputes to be resolvedor redress sought through the formal court system.However, bureaucratic unresponsiveness, arcanelanguage, lack of court capacity and the cost meanthat the official channels for obtaining redress aredifficult to access, slow and costly.

A F R I C A N

cIT

ES

Ha Matala in Maseru

16

Purchase of land through themarket

In two of the cities studied, purchase of land throughthe market was the most important means ofaccessing land for housing for all income groups. InEldoret, informal (as well as some formal and semi-formal) subdivision by land buying companies has,since independence, provided plots for both initialshareholders in the companies and subsequentpurchasers of plots.

In Kampala, by far the largest source of land is mailo– the freehold land conferred on the King ofBuganda and his chiefs under the 1900 BugandaAgreement, in which allotments were measured inEnglish square miles. The tenure rights of pre-colonial occupiers of this land were downgraded totenancies-at-will and the rights of clan heads to holdand allocate it eroded. Political turbulence, conflictand population movements in the 1970s and 1980s,as well as inconsistencies in the land law and statecollapse, enabled people to acquire or sell land inwhich the ownership of rights was unclear. Landnationalisation in 1975 resulted in further confusionbefore it was reversed in the 1995 constitution and1998 Land Act, which attempted to improve thesecurity of occupiers of various types of land. Mailotenure was again defined as freehold, subject onlyto the rights of ‘lawful or bonafide’ occupants of theland, including longstanding tenants, purchasers andpeople who had occupied land for at least twelveyears prior to 1995 without challenge by the owner.In Kampala, the tenure security of many in the lattercategory is uncertain because a large proportion of

plotholders occupied their plots more recently andthe status of those who have inherited from ‘lawfulor bonafide’ occupants is unclear. Nevertheless, the45 percent of Kampala’s land under mailo tenureis today the main source of new housing land supply.

In both Eldoret and Kampala, private sales of landto individual purchasers provide a significant volumeof housing plots, but in other cities it is unimportant(see box). It affords some access to land by thepoor, for example, through purchasers combiningto buy a single share in a land buying company inEldoret and subsequently subdividing the plot, orthrough arrangements to pay in instalments.However, in no cases did this channel enable thepoorest households to access land and most of thepurchasers are those with middle level incomes.

Although the initial owners of the land may haveformal rights of ownership, they rarely have individualtitle. In Kampala, given the contested tenure rightsof many occupiers and the complex landadministration system, title registration is slow anddifficult. In Eldoret, the title is generally held jointly

A F R I C A N

cIT

ES

Strengths Weaknesses Significant land supply (Gaborone) Insignificant land supply (other cities) Security of tenure Susceptible to discretion/corruption More likely to be serviced Standards/cost reduce access by the poor Mechanisms for dispute resolution/complaints Opaque processes Access to women Redress hard to obtain Married women discriminated against (Maseru)

Mbuya, Kampala

17

by all the shareholders of a land buying company.However, the processes of obtaining officialpermission for subdivision and house constructionare slow and the standards required are oftenconsidered unsuitable by the actors involved, so thesubdivisions do not comply with formal subdivisionand development regulations.

In these circumstances, title cannot be transferredor registered following a transaction and typically aletter of agreement is used instead. Such letters areusually witnessed (by local leaders, neighbours etc).Their validity is generally respected by other actorsin the land delivery process, including the formal landregistration agency if title is applied for and the courtsif a dispute is taken to court. However, it is possiblefor an owner to sell a plot more than once, using adifferent set of witnesses to a subsequent sale.

Women can purchase land through this channel, ifthey have means. However, married women areconstrained from purchasing land in their own nameby its social unacceptability, as noted above.

Disputes over transactions are rare, but when theydo occur, they are often resolved by local leaders –the elected chairmen of the lowest level of localgovernment (effectively neighbourhood councils) inKampala and the village elders (leaders usually

identified by approbation rather than a formalelection) at the neighbourhood level in Eldoret.However, both Local Council chairmen and villageelders may be partial, favouring one party overanother for reasons to do with politics, ethnicity orother factors, or they may be corrupt and susceptibleto bribery.

Land buying companies formed by shareholders ofmixed income groups appear to be a uniquelyKenyan phenomenon, which emerged in the periodaround independence, when President Kenyattastressed that Kenyans could not expect to get landfor free, and the purchase of land from departingsettler farmers was encouraged. The ease withwhich a company can be established in Kenyafacilitates the mechanism, and high levels of trustbetween shareholders in the land buying companiesin Eldoret are based on shared ethnicity – the vastmajority of companies are formed by members froma single ethnic group. Some low-income peopleare able to access land by the purchase of a part-share or subdivided plot. However, membershipbased on shared ethnicity by definition excludesmembers of other ethnic groups from becoming initialshareholders of plots in purchased farms (althoughnot from subsequently buying land from the originalshareholders)22.

A F R I C A N

cIT

ES

Strengths Weaknesses Significant supply (Eldoret & Kampala) Insignificant supply in some cities Affords some access to land for the poor Does not provide access to land for the poorest Letters of agreement generally witnessed, respected and recognised in applications for title Access to land for women with means

Possibility for multiple sales using different witnesses Record-keeping by lowest level of government poor (Kampala)

Disputes often resolved by Local Councils (Kampala) or elders (Eldoret)

Married women’s access to land constrained by social rules and customs

Trust between shareholders in landbuying companies based on shared ethnic origin

LCs/elders may be partial or corrupt Excludes other ethnic groups from membership of landbuying companies

18

Delivery of customary landthrough state-sanctioned channels

The extent to which customary systems of tenureand land administration were formally recognised inlaw and policy has varied between countries both incolonial times and since independence. In Botswana,one of the channels by which land is made availablefor housing can be termed state-sanctionedcustomary land delivery23. As noted above, someof the land on the outskirts of the city of Gaboronewas subdivided into privately owned commercialfarms. However, in other directions, land under theadministration of customary authorities adjoins theurban boundary. Under current government policy,this land has been vested in Land Boards (one foreach tribal area) on behalf of all citizens of Botswana.The Boards acquire land from individual rightsholders, subdivide it and allocate plots to individualsfor an indefinite period, with a customary landcertificate24.

In Gaborone, although there has been a significantsupply of publicly subdivided and serviced landavailable for both high and low income householdswithin the city boundary, the eligibility criteria andallocation process have resulted in the formation ofvery long waiting lists. In addition, large numbers ofplots have been allocated but remain undeveloped,and the adoption of higher standards in recent yearshave made the serviced plots unaffordable to lowincome households, even though no charge is madefor the land itself. As a result, much new housingland is being made available, acquired and developedin a few areas under the administration of tribal LandBoards outside the city boundary.

The study concentrated on land delivery inMogoditshane to the west (see also p.24-25). Insuch areas, a significant volume of land for housinghas been made available in recent years. However,acquisition of land by the Boards is hindered bydisputes over the level of compensation payable to

customary rights holders. In addition, plots aresupposed to be surveyed and serviced prior toallocation, but the Boards lack the financial andtechnical resources needed to achieve this.Therefore demand exceeds supply, encouragingcustomary rights holders and prospective acquirersto seek ways around the system, and temptingBoard members to favour those with something tooffer.

If the Board officially allocates a plot with acertificate, then the acquirer has security of tenure,including the right to pass on the land to heirs and tomortgage it. Although in principle the Board isentitled to demand the return of land, the legislationdoes not spell out the timing or conditions of anysuch return, and in practice it has not occurred andis not likely to occur. Women heads of householdcan obtain land through this system, with a certificatein their own name, and in theory married womencould do so also. However, in practice social normsheld by both purchasers and Board officials meanthat the latter would not do so in practice withoutthe explicit permission of their husbands.

As in all the other channels of land delivery studied,disputes are uncommon and are normally resolvedat local level between neighbours or families. When,relatively rarely, they cannot be easily settled,disputants appeal to the local Customary Court.Many of the disputes are between customaryowners or plot acquirers and the Land Board overownership rights or compensation. While sometimesresolved by an approach to the Board, few trustthis organisation and most litigants prefer to taketheir grievances to the Customary Court. In the late1990s, a Land Tribunal was established to hearappeals from or against the Land Boards. Generally,both the Customary Courts and the Tribunal adopta more conciliatory approach than the formal courtsystem, favour customary owners and those towhom they have sold land rights over the Boards,and are trusted by land market actors.

A F R I C A N

cIT

ES

19

Strengths Weaknesses Significant land supply (Gaborone) Security of tenure Women household heads entitled to land, and in theory married women also Dispute resolution mechanism available (Board, Customary Court, Tribunal)

Acquisition of land hindered by disputes over compensation Requirements for survey and servicing prior to allocation restricts supply Excess demand strains capacity and encourages rent-seeking

Boards can demand return of the land In practice, married women are not allocated

land without their husbands’ permission

Delivery of land throughcustomary channels to membersof the group

Only in some of the case study cities was there landin customary ownership on the outskirts of the urbanarea in the past or today, and today, even in thesecities, supply of free land through this channel isincreasingly limited. It primarily occurs in Enugu,where land is effectively ‘owned’ by families (whohave often registered their title), and members ofthe family are often still able to obtain land for newhousing in family homestead areas. It also occurs inMaseru, although the supplies of new land availablethrough this channel are increasingly limited.Nevertheless, inheritance is a significant way ofobtaining family land for new households.

This land is often supplied free (or in exchange for atoken of appreciation), and so it is one of the onlyways in which poor households can obtain access

A F R I C A N

cIT

ES

Strengths Weaknesses Often free Access to unmarried men restricted (Maseru) Access to poor members of the group Access to all/majority of women only through

men Security of tenure for men Fast

Limited supply for groups in built-up area (Enugu)

Institutions widely understood and generally respected within the group

Vulnerable to government intervention to acquire land for public purposes (Enugu)

Dispute resolution mechanisms effective and respected, with some exceptions

Restrictions on the sale of customary land inhibit choice

to land in contemporary cities. Men allocated plotsthrough this channel have security of tenure, althoughthere is some vulnerability to government action. Forexample, the Government of Nigeria has powersunder the 1978 Land Use Decree to expropriateland required for public purposes, which it has nothesitated to use on quite a large scale (for examplefor industrial areas and the airport in Enugu).However, access to land in customary tenure isrestricted almost entirely to men, and women canonly gain access to such land through theirrelationships with men (normally their husbands). Itis a straightforward way of obtaining land, since theeligibility criteria are well known and the processessimple.

The social institutions governing land transactionsand dispute resolution are widely understood andgenerally respected within the group in both Enuguand Maseru. However, the amount of land availablethrough this channel is shrinking. In inner city Enugu,

20

for example, the land belonging to some communitieshas long been built up and they no longer haveundeveloped land to allocate to new householdsformed within the group. In addition, there areoccasional threats to the security of tenure of thoseto whom land is allocated through this system andexamples of the customary dispute resolutionprocedures being challenged.

Purchase of customary land

In both Maseru and Enugu, the sale of land held bycustomary rights holders is the predominant meansof delivering land for new residential development.In Enugu, this channel can be subdivided intoinformal subdivisions and planned layouts, in whichthe landowning community decides to formallysubdivide and sell part of its farmland. Both theseprocesses have been central to the development ofthe city for decades and continue today (see alsop. 26). In Kampala, the subdivision and sale ofcustomary land is also important, but in terms ofvolume contributes less than subdivision of mailoland (see also p. 16-17). In Gaborone, it is alsosignificant in the periphery of the built-up area.However, it is officially prohibited and so sellers andbuyers disguise it as ‘inheritance’.

The sale of customary land thus contributes asignificant volume of plots for housing developmentin many cities, and those acquiring plots through thischannel are confident that they have de facto securityof tenure. However, sellers do occasionally sell thesame plot to more than one buyer if the first buyerhas not developed it, and government interventionrestricts supply in Gaborone. The sale of customaryland provides access to non-members of the group,in other words those who would normally not beentitled to free allocation, but may also make itpossible for members of indigenous groups that haveno remaining undeveloped land to obtain a housingplot. The prices are lower than for plots with titlepurchased through the formal private market, but

they are nevertheless market prices (except formembers of the group), which precludes poor peoplefrom accessing land through this channel. However,it does facilitate access to land for women, especiallywhere customary rules exclude them.

Markets in this type of land may become moreefficient if institutions emerge to improve informationflows. For example, brokers have started to operatein Kampala and Enugu. In addition, whereas in thepast transactions were verbal, increasingly, writtenevidence of a transaction is secured. Initially suchagreements involve lay witnesses such as seniorfamily members, neighbours and local officials. Later,lawyers are often employed to draft and witness anagreement of sale.

Although letters of agreement are exchanged, tenuremay be relatively insecure if sales have to beconcealed or if evictions are in progress anywherein the urban area. For example, evictions ofpurchasers in one neighbourhood in Maseru (on thegrounds that the subdivision and sale is illegal) makepurchasers in other areas jittery. In Enugu, althoughin many cases consultations within the family orcommunity precede the sale of land, where thesehave not occurred (and sometimes even if theyhave), challenges from family members may arise ata later date. In addition, systems of keeping recordsare undeveloped, which tends to cause moreproblems as time passes.

A F R I C A N

cIT

ES

Ogui Nike, Enugu

21

The social institutions underpinning this system arewidely understood and generally respected,including by the formal legal system, which oftenaccepts letters of sale and written agreements asvalid evidence of transactions. However, thedocuments produced are not always valid or trusted,leading some purchasers to attempt to upgrade theirsale agreements to state titles, especially in Enugu.

Allocation by officials

In Lusaka, precedents for the operation of landdelivery mechanisms today were set by widespreadsquatting on publicly and privately owned land,informal subdivision by occupiers given permissionto build a house on land by its owner (often theirformer employer), the issue of land cards by theMinister of Lands when areas were outside the urbanadministrative boundary and the exercise of manygovernance functions in informal settlements by thelocal party political organisation. In addition,legislation in the mid-1970s provided for thedesignation of informal settlements as HousingImprovement Areas. This entitled house owners toOccupancy Licences (30 year renewable use rights)and provided for the extension of infrastructure

services. By 2002, 22 of the 37 settlements in thecity had been regularised. Today in such areas,despite the establishment of non-party politicalResidents’ Development Committees, the dominantpolitical party still exercises considerable power,although service improvements generally depend onthe availability of external funding.

Informal subdivision and allocation of plots occurson both publicly and, in some circumstances,privately owned land. In the early years of asettlement’s life, newcomers are encouraged by bothparty officials and the original settlers, since increasednumbers reduce the likelihood that the governmentwill attempt to demolish houses and evict theresidents. Once an area has de facto security becauseof its large population or following its designation asa Housing Improvement Area, the process ofinformal subdivision becomes more commercialised.Thus newcomers seeking land for houseconstruction obtain plots from early settlers, partyleaders or local government officials, both within andadjacent to the boundaries of the designated area.Payment is often expected, but is generally not theequivalent of a market price, since the officials arenot entitled to sell the land. Some (e.g. local residents

A F R I C A N

cIT

ES

Strengths Weaknesses Significant supply Providers access to land to non-members of the group as well as members Relatively cheap Facilitates access to land for women with means Efficiency of the market increases as institutions emerge to improve information flows (brokers in Kampala, Enugu) Written evidence of transactions

Possibilities for multiple sales of the same plot Market price restricts access by the poor Insecure, especially if sales have to be concealed Government intervention may restrict supply Systems of keeping records undeveloped Documents not always valid/trusted – owners try to upgrade to state-sanctioned titles Insecurity leads purchasers to seek legal title (mainly Enugu)

Formal legal system accepts these types of written evidence

Despite obtaining family/group agreement, some sales challenged by family members

Institutions supporting system widely understood

Often family/group agreement precedes sale

22

Strengths Weaknesses Significant land supply (Lusaka) Relatively cheap (and free to some) Women can obtain access to land Written evidence of transactions Tenure generally secure following regularisaction or because of semi-official sanctions Local dispute reduction

Subdivision may encroach on land in private ownership, designated for non-housing uses or unsuitable for building In desirable regularised areas, payments tend to exclude the poor Most plot-owning women are widows Issue of occupancy Licences constrained by limited local government capacity Infrastructure provision inadequate Households without party political connections disadvantaged

A F R I C A N

cIT

ES

who are supporters of the political party in controlof the settlement) may be allocated plots free or fora reduced payment.

Today, in practice, especially in the older-establishedareas, the majority of plot holders have purchasedor inherited a house. Residents and prospective plotowners have become so used to the major roleplayed by party organisations that they acceptcontrols exercised by the party controlling an informalsettlement, even though the one party system of the1970s and 1980s has been succeeded by a multi-party system.

Designation as a Housing Improvement Areaprovides security of tenure, through registration ofplot holders in local government records, eventhough many occupiers never complete theadministrative procedures required for issue of anOccupancy Licence. Nevertheless, significantminorities claim that they have an OccupancyLicence or registered title. The obstacles to universalissue of Licences are less to do with inappropriate

procedures than with limited Council administrativecapacity and lack of enforcement mechanisms(unless an owner wishes to sell the plot). However,most of those without Licences have documentaryevidence of ownership, generally a letter ofagreement between seller and buyer, and many investconsiderable amounts in house construction andimprovements.

Disputes, most of which appear to be overboundaries or encroachment on adjacent plots oraccess ways, are generally resolved betweenneighbours, with appeal to the party branchchairperson or the Local Court. However, anaggrieved owner is constrained from taking anunresolved dispute further because a premium isplaced on neighbourliness by households consciousof their dependence on others during times ofdifficulty. In addition, households with little politicalinfluence may not be able to secure intervention bya party official to resolve a dispute.

23

Self allocation

There is little opportunity in contemporary cities forpeople to obtain land through non-commercialchannels by their own actions. It occurs throughdifferent mechanisms in three of the case study cities,but only on a small scale.

In Kampala, some households settle on wetlands invalley bottoms. This strategy provides one of thefew ways in which poor households can obtainaccess to land for free, but it has problems from thepoint of view of both settlers and the government.Settlers face extreme insecurity, since building in theseareas is prohibited and houses may be demolished(especially if they have not sought the permission ofthe Local Council officials). In addition, in the earlyyears, until further infill and drainage has occurred,their living conditions are very unhealthy and theirhouses liable to flooding. The land is officially ingovernment ownership so not only is settlementforbidden for environmental reasons, but also if theinitial settled area is later subdivided and sold, suchsales are illegal. Moreover, even if the settlementbecomes extensive and semi-permanent, the

topographical conditions make the sites difficult toservice.

In Maseru, women may allocate themselves plotsof family farmland without the permission of theirfamilies. As noted above, under customary rules ofaccess, women are not normally allocated land intheir own names and divorced or never marriedwomen may not be able to obtain family land.However, the number of women taking matters intotheir own hands is small.

In Gaborone also, family members may occupy anarea of family land to which they consider themselvesentitled without obtaining the express permission ofthose with decision-making power. This process islabelled ‘squatting’ by the government and made outto be widespread, although in practice it appears tooccur only on a very small scale. By using thepejorative term, the government is probablyexpressing its disapproval of the more widespreadsubdivision and sale of land by customary rightsholders, which is informal in the sense that it is donewithout obtaining permission from the Land Board(see above).

A F R I C A N

cIT

ES

Ngombe, Lusaka

24

Conflict or harmony?Relations between informal and formal institutions and actors

The purpose of this section is to provide furtherinsights into the ways in which formal and informalinstitutions and actors relate to each other in termsof land delivery. The examples deepen ourunderstanding of how the social rules governingtransactions in land operate and are understood bythe actors involved, how and why conflicts occur,and how day-to-day practices evolve as a result ofboth conflict and accommodation. They providepointers to promising innovations and trends, onwhich future policy and practice can build, as wellas problems that need to be addressed. Selectedexamples of interactions between formal and informalinstitutions and actors from each of the cities arepresented in the boxes on from p. 24 to 36.

Conflicts

Attempts by the state to enforce formal rules byeviction and demolition of houses built by settlersthey regard as illegal result in overt conflicts. Givenstate monopoly of coercive power, informalsubdividers and settlers generally cannot win in suchconflicts, as demonstrated by recent evictions inGaborone (see box p. 24-25) and Maseru. Legalprovisions that enable public agencies to appropriateland from customary owners with compensationlevels equivalent to the value of unexhaustedimprovements on the land also give rise to conflictswith landowners, especially when the land soacquired is not, in the eyes of its former owners,used appropriately. Again, owners cannot win inovert conflicts with the state, as illustrated by theexperience in Enugu (see box p.26), Gaborone andMaseru. Elsewhere (for example, in Kampala), theproblem for those who believe themselves to haveownership rights is that tenure complexity and failureto satisfy planning standards makes it difficult forthem to prove and register their rights.

Gaborone: land rights in astate-sanctioned customarytenure system

Before 1970 chiefs, through ward headmen,allocated tribal land in the areas surroundingGaborone. Since 1970, control over landsubdivision and allocation in tribal areas hasbeen vested in a hierarchy of tribal LandBoards, which are, as described on p. 17,supposed to re-acquire land from fieldowners, plan its subdivision and provideinfrastructure before the plots are allocated.Until the late 1980s, however, the land boardin Mogoditshane, on the outskirts of the city,merely rubber-stamped allocations of arableland agreed by headmen, who were paid‘thank you fees’ for their services.

Alarm at the rapid subdivision and sale of plotsin the older parts of Mogoditshane,increasingly to non-tribespeople, led to theboard demarcating a semi-circular road as aboundary beyond which further subdivisionwould not be permitted until it had re-acquiredthe land and subdivided and serviced it.Following a Presidential Commission ofInquiry in 1991, detailed layout plans wereprepared for these areas. However, existingplot boundaries and buildings were ignoredand no local consultations undertaken.Between 2000 and 2002, a number of housesthat did not comply with the new plans weredemolished.

The result has been silent confrontationbetween the state and tribal land (masimo)owners in peri-urban areas. Masimo ownershave continued to privately subdivide theirland into residential plots for sale, with thecollusion of headmen and purchasers, often

25

A F R I C A N

cIT

ES

disguising the transfer as ‘inheritance’. Theconflicts between the Mogoditshane LandBoard and customary owners or plotacquirers, as well as the tactics used to evadeBoard controls and requirements, areoutcomes of unclear, even unfair, state landacquisition rules and lack of land boardcapacity. While the state pays handsomecompensation when it acquires freehold land,it pays virtually nothing for the acquisition oftribal land, save for unexhaustedimprovements, such as standing crops,boreholes and buildings.

Land Boards argue that tribal land is a freegood, the market value of which is difficult todetermine beyond the loss of unexhaustedimprovements. They are reluctant to use theunit price of freehold land as a guide to theamount of compensation to be paid.Inadequate compensation has encouragedpre-emptive action by masimo owners andheadmen and fuelled the informal subdivisionand sale of masimo land in Mogoditshane andother peri-urban areas.

The land boards are also unwilling to paycompensation because they claim that,according to the 1970 Tribal Land Act, triballand rights are vested in them and masimoowners have only usufruct rights. Thesedisappear upon change of use by the tribalholder or when the land is no longer suitableor required for the purpose for which it wasinitially granted i.e. agriculture. However, onthe basis of their pre-1970 rights, individualmasimo owners assert ownership as well asuse rights. These conflicting claims have ledto numerous court challenges between triballand boards and masimo owners. In themajority of cases, the courts and land tribunalshave decided in favour of masimo owners.Kalabamu and Morolong argue that these

decisions seem to confirm that informalconversion of masimo land to predominantlyresidential uses without land boardauthorisation is lawful, negating the landboards’ claim that masimo owners are todayonly entitled to usufruct rights25.

In addition, the Mogoditshane board’sinability to issue Certificates of Rights to allthose granted plots by headmen has resultedin lack of clarity over rights to individual plots.The demolitions of 2000-2 temporarily haltednew construction. However, the board’sinability to acquire, subdivide and servicesufficient land to keep pace with demand hasencouraged continued subdivision by masimoowners. Further plans for demolitions, longwaiting lists and alleged land board favouritismand corruption encourage owners andpurchasers to find ways around the board’sattempts to enforce the legislation.

In particular, Kalabamu and Morolong arguethat inheritance claims have been used to justifyoccupation of land without due authorisationor to resist eviction by the land boards. Theformal rules (the 1970 Act) recognise landacquisition through inheritance, and also seemto put such inherited land outside thejurisdiction of the land boards. Kalabamu andMorolong cite numerous court cases wherethe land tribunals have ruled in favour of peoplewho the land boards claimed to be unlawfullyoccupying and using land in peri-urban areas.All these cases rested on the counterclaimsof the accused that they were the lawful heirsof the individuals from whom they hadinformally acquired the land. The success ofsuch cases, according to Kalabamu andMorolong, has encouraged people to claimfictitious blood relationships in order tocircumvent the rules of the land boards.Amendments to the law have apparently donelittle to change the situation.

26

A F R I C A N

cIT

ES

Enugu: the interface betweenindigenous, market andgovernment systems

The 1978 Land Use Decree provided for thegovernment to appropriate land fromcommunal owners for public purposes,including subdivision and supply for lowincome housing, on payment of compensationequivalent to the value of the unexhaustedimprovements on the land. Not only are thelevels of compensation regarded asinadequate, but also payment is often delayed.In addition, communities resent the fact thatmuch of the land is not used for the statedpurposes but is parcelled out to individualusers (‘big men’) for private rather than publicuse and that the land seized pays no attentionto the location of ancestral homes.Resentment of the powers the Act confers ongovernment and of the limited compensationpaid has fuelled informal subdivision and saleof community land, especially in Emene, inthe east of the city, where large amounts ofland have been appropriated in the past forinfrastructure and industry. Nevertheless,indigenous communities feel powerless toresist such compulsory acquisition.

Many sales by indigenous communities inEnugu are semi-formal. For example, inAchara in the early 1960s, indigenouslandowning families requested the EnuguTown Planning authority to prepare a planningscheme for areas of their farmland, inaccordance with which individual plots(typically 20x30m) have been demarcated andsold freehold, often for the construction ofblocks of apartments for rent as well asowner-occupation. The extension in whichactive subdivision and construction is currently

occurring was not part of the original layoutbut has since been subdivided and surveyedby its indigenous owners – here plots are soldleasehold. Direct sales to individuals arepreferred. Although land brokers increasinglyoperate as go-betweens, they are perceivedas driving up land prices and are regarded asa necessary evil rather than a desirableinformation channel. Documentation oftransfers takes the form of letters of agreementaccompanied by receipts, legalmemorandums of agreement or titles.Disputes are rare, in part because transactionsare accompanied by elaborate rituals.

In theory purchasers of land should obtain abuilding permit from the local government andpay a once-off development levy and annualproperty tax. In practice, as soon as a buyerbegins to develop a purchased plot, officialsof the local planning authority arrive at the sitewith an order to stop work. The developerpays the required levies in order to obtain abuilding permit and resume work. Seemingly,local governments regard this processprimarily as a way of generating revenuerather than regulating development. However,owners have little success in holding them toaccount for their failure to provide theinfrastructure ostensibly funded from theproceeds of the development levy. Instead theindigenous selling community considers buyersresponsible for obtaining utility connectionsand extending road access26 .

27

The tactics of non-state actors

In reaction to the exercise of coercive power by thestate, non-state actors have devised a variety ofstrategies, including evasion, non-compliance andthe development of clientelist relationships withpolitical and government officials. The illustrationsgiven in this section show that land rights holdersdevise tactics based on false compliance, appeal tosympathetic components of the state and preemptiveaction.

False compliance

The use of backdated Form Cs in Maseru (see boxp. 27) and inheritance claims in Gaborone (see p.25) are examples of sellers and purchasers of plots,with the collusion of chiefs and headmen, usingdissimulation to increase the apparent legality oftransactions that are, strictly, illegal.

Maseru: documentingtransactions and de factorights through ‘falsecompliance’

Under Lesotho law, a certificate of ownershipof customary use rights to rural land (a FormC) was issued by the customary chief (inconsultation with his advisory committee, butoften alone) after allocating land to a subject.Introduced in 1965, this provision was carriedthrough to subsequent Acts. On June 16th

1980, the LA 1979 came into force. AllForm Cs issued in respect of allocations madeprior to this date were considered lawful, andthose that were issued for allocations after thisdate were considered illegal27.

The LA 1979 also extended the boundariesof Maseru town to cover many peri-urbantraditional villages, most of which still hadextensive areas of masimo land (fields), whichwere quickly being informally subdivided intoresidential plots for sale. As indicated earlier,the government had hoped that, by extendingthe town boundaries, the informal subdivisionof masimo land would cease. In practice,local customary chiefs and masimo ownerscontinued to subdivide and sell masimo landas if no law existed. However, to maintain ashow of compliance with the legislation, albeitfalse compliance, Form C certificates were(and still are) routinely dated prior to June16th 1980. A primary reason for this practiceis that, until 1986, no compensation waspayable by the state for the acquisition of landheld under customary use rights, save forcrops and other improvements. Amendmentsto the LA 1979 in 1986 and 1992 resulted invirtually no change to the status quo. A newland bill is now in place, but is yet to beenacted into law.

Appeal to sympathetic components of the state

When in dispute with an official body, land marketactors appeal to the mechanisms most likely to decidein their favour, including local informal mechanisms(administrative or political officials, customaryauthorities), customary courts or formal tribunals.In, Kampala, LCI officials, and in Lusaka, local partyofficials, play an important intermediary role. InGaborone a Land Tribunal has been established toresolve disputes between plot holders and tribalLand Boards. (see p. 24-25)

A F R I C A N

cIT

ES

28

A F R I C A N

cIT

ES

Preemptive action

To prevent compulsory acquisition of land withinadequate compensation, customary owners, withthe collusion of the relevant traditional authorities,subdivide and sell their farm (not homestead) land.Such sales are generally informal (as in Maseru –see box, Mogoditshane in Gaborone, and Emene inEnugu, and on mailo land in Kampala). Occasionallythey are formal (as in the layouts prepared byindigenous communities in Enugu – see box p. 26).Sometimes, such actions seem reasonable andchecks and balances on abuses, such as multiplesales or chiefs profiting at the expense of theirsubjects, still appear to operate. However,elsewhere, as in Ha Mabote in Maseru, the actionsof customary owners are designed to exploit theopportunities for excess profits arising from stateactions (see box p. 28).

Maseru: expropriation offarmland for public purposesIn order to curb the process of informaldevelopment of land, especially in the peri-urban agricultural areas around Maseru, a newland law was put in place on the eve of June16th 1980, the Land Act of 1979 (LA 1979).This act effectively nationalised all land, withrights to be leased from the state. In order tofacilitate the acquisition of land for publicpurposes, the LA 1979 made provision forthe designation of areas for development asSelected Development Areas (SDAs). Theeffect of SDA declaration is to extinguishexisting interests in land, pending the grant ofsubstitute rights by the minister responsiblefor lands.

One of the first areas to be declared an SDAin 1980 for the purposes of a World Bankfinanced low-income housing project was

Khubetsoana on the periphery of the city. Thestate successfully appropriated the landwithout compensation, which was one of theobjects of the SDA declaration. However,although the land acquisition process inKhubetsoana was uneventful, the experiencesof masimo owners had telling repercussionsin adjacent areas, which in turn frustratedfurther state land appropriation efforts.

One such area was Ha Mabote, which isseparated from Khubetsoana only by a mainroad. Ha Mabote was declared an SDA in1984 for the purpose of a mixed income site-and-service project co-sponsored by the thenBritish Overseas Development Administration(ODA) and the government of Lesotho.Private (informal) subdivision in this area wasencouraged by the appropriation of landwithout compensation in neighbouringKhubetsoana.

In Ha Mabote, the state attempted to evictso-called ‘illegal’ occupants using policeforce, but called off the attempt when ittranspired that one of the local customarychiefs had solicited assistance from the armedforces, members of which had acquired plotsfrom him at concessionary prices. As a result,the enforcement strategy changed fromeviction to accommodation, which entailed thestate undertaking to recognise existing (albeitillegal) occupation in the process ofimplementing the project. The change ofenforcement strategy by the governmentopened up an opportunity for private(informal) subdividers to compile fictitious listsof plot buyers, burdening the project withaccommodating the excess demand for plots.In the end, many field owners were able tomake significant profits, as were their localcustomary chiefs, who charged a fee for every

29

false Form C certificate that they issued.When the ensuing chaos could not beresolved, the ODA withdrew its support andthe project collapsed.

State tactics

As noted above, sometimes the state attempts toenforce formal rules through the exercise ofadministrative or coercive power. More commonly,the large scale of non-compliance with formal rulesand regulations, in the face of limited public sectorcapacity and the need to secure political support,leads the state to adopt a strategy of accommodationrather than conflict. Some parts of the stateapparatus may refuse to compromise, hoping thatone day illegal land transactions will be halted andonly those subdivisions and buildings that complywith formal requirements will be regularised.However, many other parts of the state apparatusare prepared to tolerate, compromise with and evenrecognise land subdivision and transactions that donot comply with one or more legal requirements.

Recognition and reform of customarypractices

Sometimes, colonial practices accompanyingindirect rule resulted in the recognition of customarytenure and land administration practices. Toleratedsince independence, in some cases post-independence governments have, for political orpragmatic reasons, formally recognised the rights ofcustomary owners and the arrangements used toadminister land in customary tenure, even formalisingsuch arrangements by legal and administrativechanges. Examples in the case study cities includethe Buganda Land Board in Kampala and the tribalLand Boards in Botswana. However, the experiencein Gaborone demonstrates that successfulintervention is hard to achieve (see box p. 24-25).

Toleration of informal subdivision

In many cities, the activities of informal subdividersare tolerated, especially if the legal provisions areambiguous. Toleration is also common where theactivities of those engaged in informal land deliveryare directly or indirectly sanctioned by the politicallyinfluential, local officials sympathise with them, orgovernment employees (such as surveyors) areprepared to act in a private capacity to facilitatetransactions. In Kampala, complex and uncertaintenure rights and political imperatives have led towidespread tolerance of informal settlements (seebox p. 30).

A F R I C A N

cIT

ES

Farmstead in Mogoditshane, Gaborone

30

A F R I C A N

cIT

ES

Kampala: formal andinformal actors colluding incompliance and non-compliance

Subdivision and building consents depend onthe production of a registered title, as well ascompliance with complex bureaucraticrequirements and high standards. As a result,regulations are unknown, ignored or evaded.Either evasion or obtaining permission is likelyto entail the payment of bribes to officials atboth senior and street levels. Not surprisingly,the majority of households who hadsubdivided their plots had neither soughtpermission from the lower level councils (LC1s) nor the Kampala City Council (KCC),as required by law.

Nkurunziza argues that, despite elaboratelegal powers at the disposal of the KCC,planning officers confessed powerlessness incontrolling the development and use of privateland, particularly mailo land. Thispowerlessness was partly due to limitedmaterial and human resources and partly tolack of political muscle to confront the vestedinterests of different actors. Moreover, strongevidence exists that KCC planners andsurveyors undertake subdivision and surveywork privately and, because of theirknowledge of how the system operates,approval of their subdivision plans is virtuallyguaranteed. Therefore, it is not only the poorwho are implicated in informality, but also statebureaucrats and the well-connected, who arewell able to stand astride the divide betweenstate rules and everyday practices outsidestate rules28.

Compromise

The state may compromise with informal actors to

secure political support (e.g. Lusaka – see boxp. 30)provide services to prevent epidemicsobtain external finance which is conditional onthe regularisation of informal areas (e.g. in Lusakaand Eldoret)raise revenue from registration of plot holders insuch settlements (e.g. in Eldoret – see box p.31).

Lusaka: conflict andaccommodation with privatelandowners

Originally a compound for workers in thepaddock of a ranch, new occupiers werepermitted to join the original inhabitants ofNg’ombe when ranching declined in the1960s. Gradually, settlement encroached onmore of the land belonging to the ranch ownerand an adjacent church. In the mid-1990s anindigenous NGO started to supportinfrastructure installation and a Residents’Development Committee was established.Eventually, in 1999, the owners surrenderedpart of their land, in order to prevent furtherencroachment.

The area taken into public ownership wasdeclared a Housing Improvement Area. Thispaved the way for the issue of occupancylicences and service improvements. It alsofuelled further subdivision and plot allocationby party and local government officials in theundeveloped parts of the designated area. Inaddition, the area’s desirable location andrecognised status attracted further settlers,leading to further encroachmention to

31

A F R I C A N

cIT

ES

adjacent areas still in private ownership. Thistime, however, the owners felt that theiragreement with the Lusaka City Council gavethem sufficient legal and political standing toevict the settlers and demolish their houses.

Eldoret: recognising letters ofagreement for property taxpurposes

Officially, the Eldoret Municipal Council canonly collect property tax on properties withregistered title. However, a large proportionof all housing is in informally subdivided areasbelonging to landbuying companies, wherethere are significant constraints on tenureregularisation. As a result, Musyoka argues,an unwritten understanding has evolvedbetween the municipality and plot-holders inunregistered areas to use letters ofagreement as evidence of ownership.

Most owners of subdivided plots in Eldoret,especially where the original company title isnot disputed, consider that they have securetenure and many invest considerableresources in house construction. Their senseof security is buttressed by the widespreadadoption of written documentation oftransactions in land. Most of those who havepurchased land have letters of agreement,typically witnessed by neighbours, familymembers and/or local elders. While the courtsaccept such letters of agreement as evidenceof transactions when settling disputes overland, the Eldoret Municipal Land ControlBoard, which is responsible for givingpermission to subdivide farmland, refuses to

accept such letters for the purpose of grantingconsent to subdivide. The Board’s inflexibilityhas encouraged landowners to follow theinformal route.

However, because the Municipal Council hasbeen willing to compromise, it raises revenuefrom property taxes set at rates related to thelevel of services it provides in particular areas.This has enabled it to incrementally improveservices, exercise a limited degree of controlover development and negotiate withoccupiers to reserve land for improvingaccess or providing social facilities.

In one large informal settlement (Langas) alongstanding dispute over the original title hasprevented regularisation of title and hinderedinfrastructure upgrading. Today, plot-holdersin this area are trying to use the MunicipalCouncil’s recognition of their ownership forproperty tax purposes as leverage to resolvethis problem. They are threatening to withholdrates payments, demanding that titles to theirplots are issued and municipal servicesimproved29.

Such compromises may take the form of

de facto recognition of informal settlements,regularisation despite non-compliance withformal standards (e.g. in Eldoret – see box p.32), or by the Buganda Land Board in Kampala– see box p.33),the introduction of legislation containing specificand appropriate provision for regularisation (e.g.in Zambia)acceptance by the formal (state or customary)court system of oral traditions and testimony and/or informal letters of agreement as evidence ofland transactions and ownership rights (e.g. inEnugu – see box p.34).

32

A F R I C A N

cIT

ES

Eldoret: compromising overplanning standards toregularise informalsubdivisions

In Eldoret, formal land delivery has beenhampered by the limited supply of land in publicownership, high planning standards, andcorrupt and costly formal procedures fordelivering and registering land. As a result,urban land supplied through formal channelsis beyond the reach of the majority of the poorbecause access requires economic means andsocio-political patronage. Lack of capacityto provide adequate land through the formalsystem has been compensated for bytolerance of informal delivery systems, whichco-exists with elements of exploitation and/or manipulation.

The formal title for land purchased bylandbuying companies is transferred jointly tothe shareholders in the company, whosubdivide the land in proportion to theiroriginal shareholding, for their own occupation

and/or for further subdivision and sale. Theslow speed and cost of obtaining subdivisionand development permission, together withplanning standards that are regarded asinappropriate, mean that all such subdivisionis undertaken outside the formal rules,although it is not necessarily unplanned. Someshareholders employ professionals to preparelayouts, while others attempt to adopt goodpractices such as leaving land for access toeach plot and sites for community services,such as markets and schools. In suchcircumstances, title to individual plots can onlybe registered if the plot is surveyed. Not onlyis this costly, but it is also impossible if theplot does not satisfy the planning standards.

In some areas, negotiations between theshareholders, plot purchasers and the officialbodies have eventually resulted incompromise, followed by the issue of titlesfor individual plots.

The original residential portion of Kamukunji, Eldoret

33

Kampala: impediments andincentives to regularisation oftenure

Uncertain tenure rights and complex titlingprocedures mean that plot occupiers andpurchasers rarely comply with full proceduralrequirements. In colonial times, the Kabakaof Buganda ceded large areas of land tovarious churches and missionary groups.Today, for an occupant of church land, thefirst step in obtaining title is to obtain proof ofuser rights from the Local Council and manydo not proceed beyond this stage becausethey feel that LC consent guarantees themtenure security. Few mailo owners take thecomplicated and costly steps needed to obtaintitle to their land and so subdivision and saleproceeds informally, subject only to informalwritten agreements or endagaano. These aretypically witnessed by members of the localLC 1 executive, as well as elders, religiousleaders, relatives or neighbours, and areaccompanied by a sketch map.

A F R I C A N

cIT

ES

Purchasers of subdivided plots (especiallythose that satisfy the official minimum plot sizestandards) may further increase their securityby obtaining a cadastral survey of the plot –the first step in registering a title (they mayeven instal beacons without a survey).However, many owners stop once a plot hasbeen surveyed and beacons installed, sincethe presence of a surveyor and beaconsprovides public evidence of their claim, andpotential challengers need never know thatthey have failed to pursue the subsequentsteps necessary to obtain a registered title.The main motivation for those who do takethe process further is to protect the claims oftheir descendants.

In contrast, following the restoration ofBuganda kingdom properties and theKabaka’s private assets in 1993, the BugandaLand Board has, albeit without direct legalsanction, devised relatively simple and userfriendly procedures for occupiers andpurchasers from existing occupiers toregularise their tenure by obtaining leases,although relatively few occupiers (especiallyBaganda) feel the need to do so.

Apartment blocks in Achara, informal subdivision, Enugu

34

A F R I C A N

cIT

ES

Enugu: legitimising informaltransactionsIn Emene, the Amechi, Oguru and Otukucommunities were allocated land in perpetuityby the traditional ruler of Nike. In the 1940s,each of these communities arranged for aboundary survey of their land. From longpractice, these communities have developedskills in laying out and measuring plots, andthey try to maintain registers of land sales.

Every letter of agreement (so-called leasecertificate) must carry the signature of thelessee and ten members of the land sellingcommunity, including its ruler. Today they areoften drawn up by lawyers, and are acceptedby government as evidence of ownership if abuyer wishes to obtain a certificate ofoccupancy or formal leasehold title. If a buyerwishes to re-sell, consent must be obtainedfrom the community that originally sold theland. Nevertheless, differences of opinion dosurface within landowning communities,especially between those who have benefitedfrom land sales in the past and young menwho fear being deprived of their land rights,and the actions of the latter may threatenparticular transactions. Reduced lease periodsin recent years have emerged as a strategy toreduce family members’ opposition to landsales, although they are unpopular amongstbuyers.

While levels of trust in the initial transactionsbetween indigenous communities andpurchasers of subdivided plots are high,subsequent transactions in the purchased plotsare purely market transactions, and are notgoverned by social obligations and processes.Actors involved in such transactions thereforefeel obliged to use the provisions of state lawto protect their interests and register

transactions. Only if they have obtained asurvey and title to the plot being bought andsold do they feel confident that their claimscan be defended in a court of law. However,state institutions accept the letters ofagreement associated with initial transactionsas evidence of valid ownership.

Supporting informal actors

Sometimes, governments have supported actors ininformal land delivery systems even when not legallyempowered to do so. Examples include theregistration of property owners in order to chargerates (property taxes) in Eldoret (see p. ?) anddispute resolution mechanisms in which formalsystems recognise the functions and decisions ofinformal dispute resolution mechanisms, including theevidence they use and the decisions they reach.Examples from Kampala, Eldoret and Enugu aregiven in the boxes.

Kampala: linking formal andinformal rules and systems fordispute resolutionIn Kampala, disputes were more common inthe highest density settlement studied, where22 percent of landowners had experienced adispute, compared to 9 percent in the newestand least densely settled area. Increaseddispute occurrence seems to be associatedwith a period of intense activity in the landmarket in the 1990s. Dispute’s mostly concernownership or boundaries, with inheritancedisputes more important in the oldest area.

Until 1998, LC 1 chairmen and their courtswere entitled to hear and resolve land disputes,referring them to Magistrates’ Courts if

35

necessary. The Land Act withdrewjurisdiction from both LC and Magistrates’Courts and bestowed it on land tribunals.However, the latter have not been establishedand in 2000 the powers of LC andMagistrates’ Courts were restored to enablethem to clear the backlog of cases.

LCs have continued to perform this function,treating fresh disputes as though they wereoutstanding in 2000. In practice, LCs continueto deal with land matters, some in ignoranceof their legal position, and most LCchairpersons regard the settlement of landdisputes as part of their administrative rolesand community leadership obligations. Thesystem operates at a local level and is relativelyquick and informal, generally using the locallanguage. Although the LCs are supposed tokeep written records, they often do not doso – this speeds up the process but does posedifficulties if a litigant appeals against adecision. Before 1998 appeals moved up theladder from LC1 to LC2 and LC3 levels,before reaching the Magistrate’s Court. Withthe removal of these provisions, appeals haveto be referred directly to the High Court,although to avoid this, some disputants solicitassistance from the Resident DistrictCommissioner (a central governmentappointee) instead.

Specific examples illustrate the ways in whichLC dispute resolution links formal andinformal rules and administrative systems. Forexample, in one case an official valuationsurveyor was employed in a private capacityto arrive at a compensation figure foroccupants who had not been notified of a landsale by an owner.

In a boundary dispute, members of the LC 1court were able to combine the roles of jurists,land surveyors and community leaders. Theprocedure adopted to resolve the boundarydispute mimicked the formal approach of‘boundary opening’ undertaken by qualifiedsurveyors. Using tape measures borrowedfrom local builders and the sketchaccompanying the plaintiff’s original writtenagreement, the LC members were able toretrace the original boundary. In passing theirjudgement, they adopted a ‘traditional’approach by emphasising reconciliation andimposing a ‘beer’ fine (a symbolic cash fine).

In another boundary dispute, the LC courtadopted a similar procedure to confirm aboundary, but also observed that thedefendant had failed to leave a setbackbetween the plot boundary and hisdevelopment for access by his tenants. In thisexample, the LC mimicked formal planningapproaches by devising local ‘byelaws’,although the setback standard used was moreaffordable than the official one (0.9mcompared to 3m in the Public Health Act).

Since legislative changes have removed thepowers of LC courts to enforce theirdecisions, LC arbitration now often aims toobtain agreement and conciliation betweendisputants. However, their lack of legalstanding also makes it more difficult for theLCs to charge court costs or for the aggrievedto hold LC court members to account.

A F R I C A N

cIT

ES

36

A F R I C A N

cIT

ES

Eldoret: linking informal andformal dispute resolutionmechanisms

Disputes over land are relatively rare inEldoret, with only one in ten plot owners ininformal settlements reporting a dispute. Mostdisputes had been settled by local elders -unpaid leaders identified at community level,usually acceptable to the local administration,and generally having social legitimacy in theeyes of local residents. Although they lack legalor coercive powers, in taking on disputeresolution roles, they draw on understandingof the traditional role of elders in village society.Where the elders are unable to resolve adispute, it is referred first to the local chief orsub-chief (administrative officers appointed bythe national government). Only then isrecourse had to the District Officer (the chief’ssuperior) or the Land Tribunal, from wherethe dispute may be referred to the courts.

Formal and informal rules coincide in theresolution of disputes over ownership,conferring it on the first buyer of a subdividedplot, especially where that person hasconstructed a building. Owners’ knowledgeof this rule and their confidence that it will beupheld by both the formal and informalsystems, together with the written agreementsthat are commonly used, provide them with arelatively high degree of perceived de factosecurity of tenure.

Enugu: customary and statemechanisms for disputeresolutionIn Enugu, disputes are, if possible, resolvedthrough customary arbitration, relying onfamilies, elders and traditional rulers, who areconsidered to be knowledgeable with respectto the customs and traditions regulating landtenure in a community. A litigant may take acase to the Customary (Native) Court (ShariaCourt in northern Nigeria), from where appealis possible to the Customary Court of Appeal.Sitting in judgement in a Customary Court isa government-appointed president, who isusually an indigene of the area where the courthas jurisdiction and is versed in local customsand traditions. Even purchasers of land fromindigenous communities prefer to use thesechannels where possible, rather than resortto the formal court system, because of thehigh cost of litigation and delays. In practice,disputes are rare – no plot owners in Emene(a developing peripheral area) reported adispute and one in ten or fewer in the otherareas studied.

Both oral and documentary evidence is usedin dispute resolution, as well as oath taking.The formal court system admits oral evidencebased on family history or communal tradition,the decisions of Native Courts, anddocumentary evidence such as letters ofagreement. However, juju oaths are notaccepted unless there is sufficient supportingevidence.

37

ConclusionsConclusions and policy implications

Main conclusions

Five main conclusions emerge from this overviewof detailed studies of housing land delivery in sixcities. Inevitably, conclusions drawn from acomparative international study emphasise commonfeatures of the case study cities and some of thecontrasts are downplayed in this summary.

Informal land delivery systems build onearlier practices and respond to state failures

Informal land delivery systems are in part acontinuation of earlier land administration practices,especially those associated with customary tenuresystems, whether or not these were recognised andcodified by colonial or post-colonial governments.They directly involve indigenous communities in citieswhere such groups hold customary tenure rightsover land ripe for urban development. The continuedsocial legitimacy and organisational capacity of suchsocial groups enables them to exercise considerableagency, and to enforce social institutions amongstboth their own members and many of the non-members purchasing their land. Even wheresubdivision of land by groups with customary tenurerights is not possible, the practices associated withcustomary tenure systems influence the assumptionsand actions of many actors in both the formal andinformal land administration systems.

Informal land delivery is also a response to failuresof the formal tenure and land administration systems.These include the low levels of compensation paidby government when it expropriates land, which leadlandowners and customary rights holders to resistsuch acquisitions. They also include the cumbersomeand costly procedures that prevent many aspiringlandowners from complying with formal regulationsfor subdivision, registration and construction.

The main channels of housing land supply

Informal delivery systems are the main channels ofhousing land supply. Widespread non-compliancehas, in the face of limited government capacity forenforcement of regulations and the need to securepolitical support, generally led to accommodationrather than overt conflict.

In the past, in many cities, informal deliverymechanisms enabled all but the poorest to accessland for self-managed house construction. Today,non-commercial channels for obtaining land arerestricted and the vast majority of households whoobtain land through informal channels purchase it.The plots are supplied through subdivision and saleof land held under customary tenure (Enugu,Gaborone, Maseru), by owners and tenants of mailoland (Kampala) and by the shareholders of landbuying companies (Eldoret). In Lusaka they are alsosold by original settlers and allocated semi-officiallyby party and local government officials, who mayor may not expect payment.

It is no longer possible for poor households to accessland for new residential building, with a few, oftenminor, exceptions. These include

membership of an indigenous landowningcommunity – some male members of indigenouslandowning communities in Enugu and Maserucan still claim their entitlement to a plot of familyland

settlement in marginal or hazardous areas, suchas in wetland areas in Kampala

allocation of customary land or a serviced plot inGaborone, the former by tribal Land Boards andthe latter by government in serviced plotprogrammes

payment for informally subdivided land ininstalments

pooling resources to purchase a plot, forexample, in areas subdivided by land buyingcompanies in Eldoret

38

allocation of free land to some local residentswith good political or official connections in oradjacent to regularised informal settlements inLusaka

Women generally obtain access to land through theirhusbands, and increasingly by inheritance. Mostbelieve that informal social roles provide them withsufficient rights and protection. Legal provisions forjoint registration of ownership, consultation beforematrimonial property is sold, and inheritance of thematrimonial home are quite common (though notuniversal). However, social attitudes deter womenfrom exercising their rights. Increasingly, women withmeans purchase land through informal channels.

New and poor households face severedifficulties in accessing housing land

However, for many newly formed households inurban areas, especially the poor, the only way inwhich they can access a plot or house today isthrough their parents. This may be through a processof plot sharing, in which parents allow a son ordaughter to build a house on part of the parents’own plot, or through inheritance of the parents’ plotor house. Scope for the former will decrease infuture, as plots become too small for furthersubdivision, a situation that has already been reachedin some densely settled areas in Kampala and Enugu.In practice, most poor households are tenants.

Informal systems are effective in deliveringland for housing

Bearing out the first hypothesis stated in theintroduction, it is clear from our detailed empiricalevidence that informal systems are often effective indelivering land for housing, because of their user-friendly characteristics and social legitimacy.

User friendly features include simple locally-witnessed written agreements between sellers andbuyers, informal recognition of transactions fromneighbourhood power holders, use of customaryand local mechanisms for dispute resolution,standards adjusted to suit local realities andaffordability, and use of customary materials fordemarcating plot boundaries.

Their legitimacy derives from the widelyunderstood and accepted social institutions thatregulate transactions. These tend to be derivedfrom customary institutions, but have, ashypothesised in the introduction, evolved overtime, and often are very different from those thatoperated in pre-colonial times in rural areas. Inparticular, in the urban context, they have oftenborrowed from and mimic formal rules andprocedures. In addition, they may take advantageof formal rules, especially where these areambiguous or inconsistent.

Institutional analysis has provided a potent tool forunderstanding the widely understood informal socialrules with which actors engaged in informal landtransactions comply, their interpretations of formalrules, and the ways in which they exercise agencyto use and adapt formal rules when appropriate.

Informal social institutions are under pressure

Many of the social institutions revealed are resilient.However, as hypothesised at the outset, urbangrowth and development increase the pressure onthem, and in some cases they weaken and breakdown, leading to more frequent disputes andincreased tenure insecurity. In such situations, actors

Kamwokya, Kampala

A F R I C A N

cIT

ES

39

A F R I C A N

cIT

ES

in land transactions seek to use formal institutions toprotect their rights and investments, for exampleregularising their ownership of a plot by titleregistration. As shown in the previous section, theextent to which formal legal and administrativesystems recognise and work with or resist informalpractices varies between cities and over time.

Informal land delivery systems have both strengthsand weaknesses. Their strengths include their abilityto provide land in significant volumes to meet thehousing needs of various socio-economic groups thatsometimes include the relatively poor and women.Their weaknesses include the inappropriate locationsin which settlements are occasionally located, thepoor layouts that sometimes emerge and the almostuniversal infrastructure and service deficiencies.Arguably, however, these weaknesses stem as muchfrom their relationships with inappropriate formaltenure and land administration systems andineffective government agencies as from their ownshortcomings.

Policy implicationsThe findings of the study were fed back to the localcommunities studied, to validate the findings andobtain their comments on some of the policy issues.The conclusions of the study were also discussed atpolicy workshops in each of the case studycountries, and at an international comparativeworkshop. The conclusions and policy implicationshave, therefore, drawn on feedback from both keyinformants at the local level and those in the publicand private sectors with the relevant expertise andresponsibilities. Although not all those attending theseevents would agree with all the policy implicationsidentified below, all have support from some actorsin the case study cities and countries. The land policyand legislative reviews that are under way in all thecountries studied will provide opportunities forfurther assessment of their relevance andappropriateness to local circumstances.

Informal land delivery systems play a larger role inresidential land delivery in most African cities thanformal and public sector led systems. The deliveryof over half the land needed for middle and lowincome housing by informal systems has not resultedin chaos. Such systems play a significant andeffective role in housing land delivery systems. Theyshould, therefore, be tolerated and accommodated.Their strengths should be recognised and some waysin which their positive features can inform futureaction are explored below. However, theirshortcomings should also be identified and policyshould concentrate on addressing these weaknesseswithout compromising the positive contribution theymake to housing land supply.

Informal land delivery systems should betolerated and accommodated, building ontheir strengths but also identifying andaddressing their weaknesses.

Shared Water supply in Ogui Nike, Enugu

40

Urban residents and house builders seek security oftenure – even without well-developed housingfinance systems, this often leads to substantialinvestment in housing for both owner occupation andrental. One priority should therefore be to improvethe tenure security available to those accessing landthrough informal delivery channels. In somecircumstances, this may imply individual titling, butwholesale titling is often not appropriate, for threemain reasons:

Titling massively increases the value of urbanland, making it even less accessible to low-income groups

There is rarely capacity in the formal regulatorysystem to adjudicate, survey and register largenumbers of individual titles. Rather than issuingtitles to a minority of landholders, the resourcesavailable should be used more strategically, toguide urban development. This may includeinvestment in main access roads and water supply,and surveying the boundaries of large landparcels, within which non-state actors cansubdivide and sell land.

Owner-occupiers are unlikely to mortgage theirhomes in order to release the capital tied up inproperty for other purposes, such as businessinvestment. One reason is the absence ofdeveloped financial systems that might enableowners with title to release equity in theirproperties . In addition, middle and low-incomehouseholds in African cities give high priority tolivelihood security and are reluctant to jeopardisethis by becoming indebted. Further, urbanhouseholds attach high importance to ensuringthe future wellbeing of their children and regardtheir ability to bequeath urban property asimportant to achieving this goal.

In many countries, alternatives to universal individualplot titles are already available under existinglegislation. Experience with implementing alternativeforms of tenure, usually based on usufruct, varies.However, the legislative base and implementation

experience of countries such as Zambia (blocksurveys, occupancy licences) can inform futuredevelopment of policy and practice.

One of the main threats to tenure security is oftenthe action of governments themselves, particularlyevictions. Thus one of the most obvious ways ofimproving tenure security is for governments, in thevast majority of cases, to cease to evict settlers anddemolish their houses.

Governments should provide basic short-termsecurity to residents in informal settlements,perhaps by a simple statement that residents willnot be evicted. In some instances, (constructionon privately owned land, land needed for otherpublic purposes, or land that is completelyunsuitable for residential use), it will not bepossible to tolerate informal settlements. In suchcases, eviction (accompanied by suitablearrangements for resettlement) may be the onlyoption. However, generally governments shouldcease to evict settlers and demolish houses.

Security can be enhanced by public sectoragencies accepting innovations in procedures anddocumentation that have emerged in informalsystems, because these are popularly understood,widely accepted, cheap and procedurally simple.They include witnessed letters of agreement andlocal registers of transactions, as well as ways inwhich formal and informal systems work togetherto solve problems and resolve disputes.

Threats to women’s security of tenure should betackled through matrimonial as well as property law.Because there is considerable resistance to increasingthe rights of married women in practice, advocacyand changes in social attitudes will also be needed.

A F R I C A N

cIT

ES

41

To encourage investment in both owner-occupied and rental housing, the tenuresecurity available to those who access landthrough informal delivery channels should beenhanced. In most cases, this impliesrecognition of informal settlements and ceasingto evict settlers. In some circumstances themost appropriate way of increasing tenuresecurity is to extend individual titling.However, given limited state capacity and thethreat posed to middle and low-incomehouseholds’ ability to access land by priceincreases and speculation, alternatives toindividual titles may be more appropriate.

Recognition of areas in the process of being settledthrough processes of informal subdivision and salecan pave the way for working with subdividers andsellers to improve layouts, ensure the reservation ofaccess ways and sites for social facilities, and makeit possible for the early provision of a basic level ofservices. Local initiatives have instituted such aflexible approach in Eldoret, even when externalfunding for regularisation and upgrading has not beenavailable. Recognition can also contribute toincremental improvements in service provision, sinceonce areas are de jure or de facto recognised, utilitiescan be provided on a full or partial cost recoverybasis, as demonstrated by some utility providers inthe case study cities (commonly electricity, oftenwater, more rarely other infrastructure).

The poor layouts and inadequate services thatoften characterise informal settlements can beaddressed by recognising such areas, pavingthe way for working with subdividers andsellers to improve layouts and enabling theearly provision of basic services.

As well as generating user charges for services, theregistration of occupiers makes it possible for localgovernment to generate tax revenue. However,recognition of informally settled areas and acceptanceof their occupants should be designed in such a waythat, wherever possible, the poor are not furtherdisadvantaged by the imposition of unaffordablecosts or gentrification. The difficulty of ensuring thisshould not be under-estimated. Earlier sites andservices programmes were often small-scale andunaffordable by low income households, and somelocal governments’ ability to provide serviced plotsis constrained by the limited land remaining in publicsector ownership. Nevertheless, strategies to reducepoverty and improve livelihoods in urban areas mustinclude ways of ensuring access to adequate shelter.This entails providing land for low-income housing,as well as giving policy attention to rental housing.As private land delivery becomes increasinglycommercialised, it may be necessary forgovernments to revisit sites and services approaches.

A F R I C A N

cIT

ES

Informal settlement on the outskirts of Maseru

Local governments should register occupiersin informal settlements. As well as making itpossible to charge users for services, this canenable local governments to generate taxrevenue.

Laws relevant to the administration of urban landinclude tenure laws, planning and developmentcontrol legislation and provisions for land taxationand dispute resolution. In each case, the basiclegislation is accompanied by regulations andprocedures that govern administration and practice,sometimes decided centrally and sometimesspecified in local bye-laws. Much of the legislativeframework for the administration of urban land hasbeen inherited from laws introduced by Britishcolonial governments, influenced by the system ofrule adopted and the colonial administration’srelationships with the indigenous authorities. Muchof this legislation is inappropriate and outdated.Post-independence land reforms have oftenneglected the realities of land markets, limitedgovernment capacity and political interests, especiallyin the large constituencies constituted by informalsettlements. In addition, they have oftenconcentrated on rural issues and neglected urbanland. As a result, they have also often beeninappropriate and ineffective. Nevertheless, despiteimplementation problems, not all the existinglegislation is inappropriate.

Current reviews of land policy should considerwhether the existing legislation, including laws,regulations and administrative arrangements,provides an adequate framework for guiding urbangrowth and development. To build on the strengthsand address the weaknesses of informal deliverysystems in the local context, much of the relevantlegislation is likely to need revision. In addition, asagreed in most of the country policy workshops,there is a need for the formal land administrationsystem to be decentralised, in particular to providefor local registration of land rights and transactions.

To build on the strengths and address theweaknesses of informal delivery systems invarying local contexts, the formal landadministration system should be decentralised,in particular to provide for local registrationof land rights and transactions.

Legislation providing for compulsory acquisition ofland to which indigenous groups claim tenure rightsgives rise to many problems, exacerbated byprovisions for the compensation payable to be basedon the value of improvements on land rather than itsmarket value. Not only does this seem unfair torights holders, particularly in the case of peri-urbanland, but also delays in payments and governmentfailure to use the land for the public purposes forwhich it was ostensibly acquired increase theirresistance to such acquisitions, especially inGaborone, Enugu and Maseru. Althoughgovernments claim that their resources are insufficientto pay higher rates of compensation, the currentprovisions encourage owners to take matters intotheir own hands, with adverse results for the patternof urban development.

Revised compensation provisions are needed,requiring governments to pay adequate and faircompensation when they expropriate land for publicpurposes. This would

Deter premature informal subdivision intendedto preempt arbitrary and under-compensatedexpropriation (e.g. Enugu and Maseru)

Improve the operation of some state-ledsubdivision and allocation processes (e.g. theoperations of Land Boards in Gaborone)

Increase the ability of governments to provideland for infrastructure or industry withoutantagonising local land rights holders (e.g. Enugu)

Enable governments to increase the supply ofserviced land for low-income housing.

A F R I C A N

cIT

ES

42

Footnotes1 Rakodi, C. (1997) “Residential property markets

in African cities” in Rakodi, C. (ed) The UrbanChallenge in Africa, Tokyo: United NationsUniversity Press, p. 371-410.

2 Giddens, A. (1984) The Structure of Society,Cambridge: Polity Press

3 Healey, P. and Barrett, S.M. (1990) “Structureand agency in land and property developmentprocesses: some ideas for research”, Journal ofPlanning Education and Research, 27(1), p. 89-104; Tripp, A.M. (1997) Changing the Rules:The Political Economy of Liberalisation andthe Urban Informal Economy in Tanzania,London and Los Angeles: University of CaliforniaPress; Mbiba, B. and M. Huchzermeyer (2002).“Contentious development: peri-urban studies insub-Saharan Africa.” Progress in DevelopmentStudies 2(2), p. 113-31.

4 Pamuk, A. (2000) “Informal institutionalarrangements in credit, land markets andinfrastructure delivery in Trinidad”, InternationalJournal of Urban and Regional Research, 24(2)p. 379-96; Van Horen, B. (1999) “The de factorules: the growth and change of an informalsettlement in Durban, South Africa”, Third WorldPlanning Review, 21(3), p. 261-82.

5 Razzaz, O.M. (1998) “Land disputes in theabsence of ownership rights: insights from Jordan”,in Fernandes, E. and Varley, A. (eds) IllegalCities, London and New York: Zed Books, p.69-88.

6 Rakodi, C. and Leduka, C. (2003) InformalLand Delivery Processes and Access to Landfor the Poor in Six African Cities: Towards aConceptual Framework, Birmingham:International Development Department, School ofPublic Policy, Informal Land Delivery Processesin African Cities, WP 1; Benda-Beckman, F. von(2001) “Legal pluralism and social justice ineconomic and political development”, IDSBulletin, 32(1), p. 46-56; Benton, L. (1994)“Beyond legal pluralism: towards a new approachto law in the informal sector”, Social and LegalStudies, 3, p. 223-42.

7 North, D.C. (1989) “Institutions and economic

growth: an historical introduction”, WorldDevelopment, 17(9), p. 1319-1332.

8 Razzaz, O.M. (1994) “Contestation and mutualadjustment: the process of controlling land inYajouz, Jordan”, Law and Society Review, 28(1),p. 7-39.

9 Pamuk, 2000, op.cit.10 Leitmann, J. and D. Baharoglu (1999). “Reaching

Turkey’s spontaneous settlements: the institutionaldimension of infrastructure.” InternationalPlanning Studies 4(2), p. 195-212; Assaad, R.(1996). “Formalising the informal? Thetransformation of Cairo’s refuse collectionsystem.” Journal of Planning Education andResearch 16(2), p. 115-26.

11Kombe, W. (2000) “Regularising housing landdevelopment during the transition to market-ledsupply in Tanzania”, Habitat International,24(2), p.167-84; Kombe, W. J. and V. Kreibich(2000). Informal Land Management inTanzania. Dortmund, University of Dortmund,SPRING Research Series No. 29; Leduka, R.C.(2000) The Role of the State, Law, and SocialActors in Illegal Urban Development inMaseru, Lesotho, Unpublished PhD dissertation,Cardiff: University of Wales, Department of Cityand Regional Planning.

12Moore, S. (1973). “Law and social change: thesemi-autonomous social field as an appropriatesubject of study.” Law and Society Review 7, p.719-46.

13Moore, 1973, ibid, p. 720.14Razzaz, 1994, op.cit.; Scott, J.C. (1985)

Weapons of the Weak: Everyday Forms ofPeasant Resistance, New Haven and London:Yale University Press.

15Tripp, 1997, op.cit.16Razzaz, 1994, op.cit.17Rakodi and Leduka, 2003, op.cit.18Kombe, 1994, 2000, op. cit.; Kombe & Kreibich,

2000, op.cit.19Rakodi and Leduka, 2003, op.cit.20Maputo and Luanda: Jenkins, P. and H. Smith

(2004). “Collaborative approaches to knowledgedevelopment: action research in urban land issues

43

in Mozambique and Angola.” InternationalDevelopment Planning Review 26(1), p. 25-46Dar es Salaam: Kironde, J. M. L. (2000).“Understanding land markets in African urbanareas: the case of Dar es Salaam, Tanzania.”Habitat International 24(2), p. 151-66;Kombe, 2000, op.cit.; Kombe, W. J. (2005).“Land use dynamics in peri-urban areas and theirimplications on the urban growth and form: thecase of Dar es Salaam, Tanzania.” HabitatInternational 29(1), p. 113-36; Kombe andKreibich, 2003, op. cit; Lupala, A. (2001). Peri-urban Land Management for RapidUrbanisation - The Case of Dar es Salaam.Dortmund, University of Dortmund, Faculty ofSpatial Planning, SPRING Research Series 32.Accra: Gough, K. and P. W. K. Yankson (2000).“Land markets in African cities: the case of peri-urban Accra.” Urban Studies 37(13), p. 2485-500; Larbi, W. O., A. Antwi, P. Olomolaiyi (2003).“Land valorisation processes and state interventionin land management in peri-urban Accra, Ghana.”International Development Planning Review25(4), p. 355-72.South Africa: Royston, L. (2002). Security ofurban tenure in South Africa: overview of policyand practice. In Durand-Lasserve, A. and Royston,L. (eds) Holding Their Ground: Secure LandTenure for the Urban Poor in DevelopingCountries. London, Earthscan, p. 165-81; vanHoren, 1999, op.cit.

21The small proportion of land delivered throughformal channels and officially registered, as wellas the general inadequacy of data on population,land transactions and construction meant that itwas not possible to undertake quantitative analysesof the land markets in these cities.

22There is a longstanding dispute between originalshareholders in one company. This concerns thetitle to the farm concerned. Disputes over the titleheld by shareholders in other companies in Eldoretappear to be rare, as are disputes over subdividedplots. By all accounts, the operations of land buyingcompanies give rise to more conflicts elsewhere

in Kenya. The smooth process in Eldoret may bedue to the ability of plot supply to keep pace withdemand because of the ready availability of farmholdings for purchase, compared to the pressureon land in and around other Kenyan cities.

23The situation in Ghana, especially in Kumasi, issomewhat similar.

24In Kampala, the Buganda Land Board, which wasestablished by the Buganda Kingdom, also has astatutory right to allocate land to individuals onbehalf of the Kingdom. However, today itregularises land occupied by mailo tenants or thoseto whom they sell, rather than allocating new plots.

25Kalabamu, F. and Morolong, S. (2004) InformalLand Delivery Processes and Access to Landfor the Poor in Gaborone, Botswana,Birmingham: International DevelopmentDepartment, School of Public Policy, Universityof Birmingham, Informal Land Delivery Processesin African Cities, WP 3.

26Ikejiofor, C.U. with input from Nwogu, K.C. andNwanunobi, C.O. (2004) Informal LandDelivery Processes and Access to Land for thePoor in Enugu, Nigeria, Birmingham:International Development Department, School ofPublic Policy, University of Birmingham, InformalLand Delivery Processes in African Cities, WP 2.

27Leduka, 2000, op. cit.; Leduka, R. C. (2004).Informal Land Delivery Processes and Accessto Land for the Poor in Maseru, Lesotho.Birmingham, International DevelopmentDepartment, School of Public Policy, InformalLand Delivery Processes in African Cities, WP 5.

28Nkurunziza, E. (2004). Informal Land DeliveryProcesses and Access to Land for the Poor inKampala, Uganda. Birmingham, InternationalDevelopment Department, School of Public Policy,University of Birmingham, Informal Land DeliveryProcesses in African Cities, WP 6.

29Musyoka, R. (2004). Informal Land DeliveryProcesses and Access to Land for the Poor inEldoret, Kenya. Birmingham, InternationalDevelopment Department, School of Public Policy,University of Birmingham, Informal Land DeliveryProcesses in African Cities, WP 4.

A F R I C A N

cIT

ES

44

PublicationsInformal Land Delivery Processes in African Cities

Working papers

1. Rakodi, C. and Leduka, C. (2003) Informal Land Delivery Processes and Access to Land for thePoor in Six African Cities: Towards a Conceptual Framework.

2. Ikejiofor, C.U. (2004) Informal Land Delivery Processes and Access to Land for the Poor in Enugu,Nigeria

3. Kalabamu, F.T. and Morolong, S. (2004) Informal Land Delivery Processes and Access to Land forthe Poor in Greater Gaborone, Botswana

4. Musyoka, R. (2004) Informal Land Delivery Processes and Access to Land for the Poor in Eldoret,Kenya

5. Leduka, R.C. (2004) Informal Land Delivery Processes and Access to Land for the Poor in Maseru,Lesotho

6. Nkurunziza, E. (2004) Informal Land Delivery Processes and Access to Land for the Poor in Kampala,Uganda

7. Mulenga, L.C. and Rakodi, C. (2004) Informal Land Delivery Processes and Access to Land for thePoor in Lusaka, Zambia

Policy briefs

1. Ikejiofor, C.U. (2004) Informal Land Delivery in Enugu, Nigeria: Summary of Findings and PolicyImplications

2. Kalabamu, F.T. (2004) Land Delivery Processes in Greater Gaborone, Botswana: Constraints,Opportunities and Policy Implications

3. Musyoka, R. (2004) Informal Land Delivery in Eldoret, Kenya: Summary of Findings and PolicyImplications

4. Leduka, R.C. (2004) Informal Land Delivery Processes and Access to Land for the Poor in Maseru,Lesotho

5. Nkurunziza, E. (2004) Informal Land Delivery Processes in Kampala, Uganda: Summary of Findingsand Policy Implications

6. Rakodi, C. and Leduka, R.C. (2004) Informal Land Delivery Processes and Access to Land for thePoor: A Comparative Study of Six African Cities

Publications can be obtained by either telephoning Carol Fowler on 44 (0) 121 414 4986or Email: c.a.fowler@bham.ac.uk and also downloaded as a PDF file from http://www.idd.bham.ac.uk/research/researchprojs.htm

45