Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
11 Residential property markets in African cities
Carole Rakodi
Abstract
Ce chapitre porte sur le fonctionnement des marchés immobiliers et la façon dont ils vent modelés par les relations politiques et sociales entre les parties concernées, plutôt que par les caractéristiques physiques de l'environnement construit résultant. L'on y examine attentivement la disponibilité des terres indispensables du fait de la rapidité de la croissance des villes et des besoins de terres où bâtir des logements. Avant même l'époque coloniale, les régimes d'occupation des sols commençaient d'évoluer compte tenu des situations nouvelles. Mais la nécessité d'évoluer se fit plus pressante avec l'impulsion du colonialisme et la croissance urbaine rapide qui suivit les indépendances. Les régimes d'occupation de terres importés de diverges parties d'Europe ont coexisté tant bien que mal avec les systèmes autochtones d'administration des terres. Avec l'augmentation des pressions, ces derniers se vent commercialisés de plus en plus tout en posant des problèmes, là où ils subsistent, pour les responsables de la gestion urbaine. Les tentatives de répondre aux besoins de nouveaux logements en offrant de vastes terrains répartis en lots et en y pourvoyant des services vent restées insuffisantes. Près de la moitié des pays de l'Afrique subsaharienne ont nationalisé les terres et/ou remplacé la propriété foncière perpétuelle et libre par la tenure à bail, mais sans obtenir les résultats souhaités. Malgré d'apparentes "réformes" il n'y a eu que peu de changements apportés aux règles et procédures de l'époque coloniale. En pratique, la majeure partie de la demande de terre à bâtir dans les villes est satisfaite illégalement, soit avec des installations de squatters soit, de plus en plus, avec des lotissements illicites. Les terres et d'autres composantes nécessaires à la construction de logements vent assemblées en combinaisons variables par divers ensembles d'acteurs. Dans la plupart des villes d'Afrique, la construction de logements publics n'a jamais satisfait qu'une petite proportion des besoins et la majorité des maisons, dans les secteurs formels et informels, vent construites par des particuliers, la construction autogérée ayant depuis longtemps remplacé la construction auto-assistée. En plus des difficultés que rencontrent de plus en plus les propriétaires prospectifs à obtenir des terres, la construction des logements est aussi handicapée par l'absence de crédit et l'insuffisance des matériaux de construction. L'accès à la propriété immobilière étant devenu toujours plus difficile dans de nombreuses villes, la location est de plus en plus courante. Les transactions immobilières dans les zones bâties, l'évolution et le fonctionnement mêmes du système immobilier, les échanges de logements et l'aménagement des centres villes font que même une zone déjà urbanisée n'est pas statique mais dynamique. Les opinions varient quant à la meilleure façon pour les pouvoirs publics de surmonter les problèmes d'occupation, d'enregistrement et d'administration des sols; certaines vent incompatibles ou prêtent à controverses. Les recommandations suivantes vent avancées: favoriser la propriété individuelle, (re)-privatiser les terres, régulariser et enregistrer les occupations illicites, réformer les normes et les procédures et prendre des mesures pour surmonter les autres obstacles à la construction de logements.
Introduction
The urban built environment is the outcome of a series of transactions in land and property, together with a process of construction. Deals in property are not merely economic transactions, but are shaped by political and social relations between the parties involved, and it is on these market processes that this chapter will focus rather than on the resulting physical patterns of land use and building. Attention will be paid both to the land development process by which rural land is urbanized and to transactions in land and property within cities. The state, both central and local, has had an important role in urban development, both as a direct actor and as a result of attempts to guide and regulate the process. The outcomes over time of these attempts to intervene through the instruments of land and housing policy will be assessed.
The first and largest part of the chapter will focus on evolving systems of land supply for urban development. Because the current situation is characterized by the coexistence of different modes of supply originating at different stages in the development of cities, the historical origins of these systems will be examined, as well as their more recent evolution. Pre-colonial land tenure, the impact of colonialism, changes to indigenous land allocation systems since independence, the impact of attempted reforms based on nationalization of land, and the illegal supply of privately owned land will be discussed in turn. This extensive consideration of land supply is justified by rapid urban growth rates and the primary importance of land in the production of shelter. The assembly of land together with other components, in varying combinations and by varying sets of actors, to produce housing is the subject of the second main section. The pattern of urban development and the housing stock that are produced by the urbanization of agricultural land and the construction of dwellings are, however, not static. The third section will outline property markets within built-up areas: rental tenure, exchange of land and houses, and inner-city renewal. In conclusion, some of the policy and management implications of evolving land and property markets and development processes will be briefly explored. Contrasts between cities in different parts of the continent will be illustrated and explained, but restrictions on space preclude full recognition of the variety of urban conditions or detailed qualification of the broad trends and characteristics identified.
The supply of land for urban development
Pre-colonial land tenure
In most of sub-Saharan Africa and despite considerable variations, it is safe to say that pre-colonial land tenure was based on the concept that ownership of land was vested in a community (tribe, clan, lineage, family), the head of which held it in trust and administered it on behalf of the group's ancestors, its currently living members, and its members yet to be born. All adult members of the community had rights to use the land, although these use rights varied with status, gender, etc. The head of the group had the right and responsibility to allocate unused land and arbitrate in disputes, and the usufruct rights were inheritable. Variations in the system arose from political structures, types of agriculture, rules of succession, and so on. The spread of Islam, in which private individual ownership of land is part of sharia law, led to modifications in indigenous tenures, for example in northern Nigeria, the Sudan, and coastal Kenya. The land system in Mombasa, for example, where "freehold" land can be leased to a tenant who pays a goodwill fee (kilemba) for the right to build and an annual rent, differs from the rest of Kenya. As a result, whereas in Nairobi 23 per cent of dwellings are in "shanties" and only 14 per cent in rooming (Swahili) houses, in Mombasa the proportions are 5 per cent and 66 per cent, respectively (Ondiege, 1989). In Morocco, several types of ownership are established in sharia law, for example habous, melk, and gulch (Ameur, 1995).
In urban settlements that existed before colonialism, rights in land were held by the indigenous occupants, originating with the king or paramount chief, who held the absolute rights in the land in trust, and descending to individuals via intermediate groups (Bruce, 1988). Each member of a rural community had a claim to as much land as he (usually) could farm or, in urban areas, as was his due, depending on seniority. There were a variety of secondary tenures, such as sharecropping or women's rights in their husband's land. Land allocated to an individual and his (or occasionally, in matrilineal societies, her) descendants under such usufruct tenure would be forfeit only if not productively used. Rights of occupancy might be subject to the payment of tribute or the rendering of other services (Acquaye and Asiama, 1986). Arrangements for accommodating non-members of the group existed in rural areas, for example tenancy, but prior approval of the group (typically the chief and elders) was needed (Bruce, 1988). Land was not, therefore, private property; control over and access to land were inextricably linked to socio-political relationships and land was not just a physical entity, but had symbolic or spiritual significance as the embodiment of a link between the generations (Aquaye and Asiama, 1986; Frishman, 198X; van Westen, 1990).
The indigenous system was adapted for use in urban areas. It provided access to urban land for group members of widely varying economic status, a means of administering the allocation and occupation of land, and a deterrent to the entry of land into the open market. Typically, family "compounds" accommodated (in a patrilineal society) a man, his wife/wives, their children, and his adult sons and their families (Mabogunje, 1993). However, claims to land were unequal between and within families, with the more powerful and/or wealthy being more likely to have their claims granted (Bruce, 1988). Commonly, urban administration was the responsibility of a traditional ruler, and cities were divided into quarters under the jurisdiction of sub-rulers with administrative roles including land allocation (Mabogunje, 1993). Arrangements evolved to deal with non-indigenes, for example traders and other migrants. In northern Nigeria, areas for northern Islamic non-indigenes (Tudun Wada) and southern migrants (Sabon Gari) were allocated (O'Connor, 1983). Elsewhere, non-indigenous in-migrants were dealt with on an individual basis and greater ethnic mixing occurred as a result. Traditionally, a token payment of "drinks money" was made to the chief or other person responsible for land allocation. Land was not regarded as a commodity and this payment was not a "price" related to the value of the land. However, in places, for example Kano, land had become scarce before the end of the nineteenth century and its allocation without charge had been replaced by sales (Frishman, 1988).
Indigenous tenure was, therefore, evolving to deal with new developments even before the advent of colonial administration. However, pressures for change intensified firstly with the imposition of colonial rule and subsequently with the increased rates of population growth that have succeeded independence. These changes in circumstances and the way in which urban land supply has adapted to them will be examined in turn.
The impact of colonialism on urban land
The impact on settlement patterns and processes of urban development varied, as has been described in chapter 2, with the identity and purposes of the colonizing power and the nature of the existing political economy on which external rule was imposed. In much of coastal West Africa, strong existing states and peasant agriculture that could produce the crops of interest to colonial traders gave rise to the British system of indirect rule and French colonization of Senegal. French administration was established by military conquest in much of the interior. It was accepted that existing urban settlements would remain, and many flourished as colonial trade and communications were extended, although others were bypassed. The urban residence of their African populations was accepted and existing modes of administration and tenure were allowed to continue, although they were modified by the needs of the colonial administration for land for its own administrative and trade functions and personnel.
Land around the urban centre or in a parallel European settlement was appropriated and declared Crown or state land. In the settler and labour reserve colonies of east, central, and southern Africa, where few substantial urban settlements existed, and where indigenous inhabitants were dispossessed of large areas of rural land and confined to reserves, urban settlements were established on state-controlled land. Land titling was introduced for this land, which was alienated to individuals under freehold or leasehold tenure. Legislation based on that which had evolved in Europe was enacted and thus varied between colonial powers, but the procedures for survey and registration were typically cumbersome and land was generally taxed for local revenue. Large areas of land were sold at low prices to European (and sometimes Asian) immigrants, especially in the settler colonies of Kenya, Zimbabwe, South Africa, Angola, and Mozambique, and much remained vacant for years. Such land entered into a land market, transactions in which reflected the economic fortunes of both the colonial power and its colony. Periods of economic prosperity and immigration, for example after World War II, were marked by price increases, new development, and further subdivision, often on a highly speculative basis. In periods of political uncertainty and economic depression, especially in the run-up to independence, prices were typically depressed and the property market stagnant, for example in Zambia and Zimbabwe (Van den Berg, 1984; Rakodi, 1995).
In the early years of colonial administration, Africans were not, generally, permitted to own land where they did not already do so in established settlements. However, in the settlements of ax-slaves, settlers were allocated freehold plots on arrival, for example in Togo and Liberia (O'Connor, 1983; Mabogunje, 1993). Occasionally the imported tenure system was adapted to local practice. For example, in northern Nigeria a leasehold system was adopted as being closest to the indigenous form of tenure (O'Connor, 1983). Elsewhere, intermediate levels of rights were introduced that gave Africans use of land for house construction but not ownership. In the French colonies, three basic procedures were instituted (Durand-Lasserve, 1993)
- le permis d'occuper (occupation permit), which gave use rights for a limited period, with no provision for conversion into a higher-level right;
- le permis d'habiter (permission to live), which was conditional on the builder satisfying minimum regulations, revocable at any time by the state and neither inheritable nor saleable;
- la concession foncière urbaine (land concession). These were plots made available on a provisional basis following public land subdivision, which could be made definitive when the plot was developed in conformity with the relevant regulations, the property had been valued, and the formalities completed.
In the labour reserve colonies, control over the workforce was enforced by a variety of mechanisms, including prohibiting urban home ownership to ensure that migrants returned to their villages on completion of their contracts or retirement. In countries such as Zambia, therefore, only rented housing was provided, both for African workers (by employers or government) and for colonial administrators. In the settler economies, especially South Africa, whites and other immigrants took up permanent residence, and thus ownership of land and property was crucial to their prosperity and maintenance of control, whereas the indigenous labour force was mostly forced to rent (as described by Beavon for Johannesburg in chap. 5).
Where indigenous tenure in urban areas survived the imposition of colonial rule, it was sometimes drastically modified. In Uganda, for example, large areas of land were allocated by agreement with the British as virtual freehold in square mile units (mailo) to the Kabaka and the chiefs. Most farmers and urban residents were, therefore, tenants on this land, which could be inherited but not sold (Barrows and Roth, 1990; see also Gayiiya and Walaga, 1995). Even where such transformation did not occur, indigenous tenure systems continued to evolve as a result of changes stemming from colonial rule. In Nigeria, for example, the British co-opted certain chiefs (usually the paramount chief) through whom to exercise indirect rule, thereby eroding the role of other chiefs who had perhaps had important roles in urban administration or land allocation (Bruce, 1988; Mabogunje, 1993). Financial transactions increasingly supplanted social transactions in both rural areas (for example tenancy arrangements) and urban centres. In Kano, there was evidence of a land market, which progressively excluded low-income residents, in the 1940s (Frishman, 1988).
At independence, therefore, African countries inherited a "dual" system of land supply, in which European notions of tenure, systems of land administration, and policies that embodied colonial aims and social relations coexisted uneasily with indigenous tenure and land administration that reflected the social and political relations of tribal society. Neither had been static: European urban policies had been modified as the revenue implications of earlier policies were realized and difficulties experienced in their implementation, for example to permit urban land ownership by Africans even in some of the settler societies; the structure of tribal society had changed under colonial rule, with effects on both administrative systems and approaches to urban land, in which family interests were becoming relatively more important. The main changes brought by independence were a determination to give Africans access to economic opportunities, education, and property and the ending of such influx control as existed, leading to more rapid rural-urban migration and urban population growth. However, as will be seen in the next section, much of the inherited urban land-supply system continued virtually unchanged until the 1970s.
Post-independence urban land development
In recent decades the indigenous tenure system has come under increasing pressure. In many West African cities, for example in Ghana and Nigeria, strong group or family attachment to land has been retained since pre-colonial times, inhibiting individual sales. However, unwillingness to sell also poses obstacles to the acquisition of land for public purposes and to redevelopment of old and poor-quality inner-city housing; as well as giving rise to many disputes over claims and failure to secure contributions to funding for infrastructure installation. Often rights are not clearly defined, overlapping rights exist, and "contracts" are unenforceable (Durand-Lasserve, 1993). As a result, transaction costs are high, and the residual risk of other claims coming to light following a sale decreases the value of the land and deters investment. In the area covered by the Traditional Council in Kumasi, Ghana, for example, the rights of the paramount stool chief (the Asantahene) and 55 sub-stool chiefs, each with sub-sub-stools and families, have produced a very unclear set of claims to land and complex procedures for acquisition. These have given rise to extensive disputes and litigation, with an outstanding backlog of 16,000 cases in the High Court (Asiama, 1989, 1995; Acquaye and Associates, 1989).
Long-term vesting of land in families has both reduced the control of chiefs and given rise to a tenure form that is more accurately described as "customary freehold" (Mabogunje, 1993) than usufruct. Accommodation in the family houses that are built on such land means that rent-free (or low-rent) accommodation is available to poorer members of kin groups (Korboe, 1992). One- and two-storey compound houses eventually containing as many as 80 rooms accommodate three-quarters of households in Accra, of which 65 per cent live rent free (Malpezzi et al., 1990; Tipple and Willis, 1991; Korboe, 1994). Because the rent control legislation (see below) allows tenants to be evicted if a dwelling is needed for a family member, and because of declining real incomes and the return of Ghanaians working in other countries, the proportion of households living rent free in family houses has increased (Korboe, 1992). As better-off households move out to avoid increasing responsibility for maintenance and other housing-related costs, the condition of the houses deteriorates, while sale for redevelopment is opposed by those who benefit from the cheap housing provided, and who are able to justify such opposition by appealing to the inalienable status of family land. In Nigeria likewise, customary tenure has prevented non-indigenes from achieving house ownership in cities such as Calabar and Akure, and housing built on land held under this form of tenure, especially in inner-city areas, is old and deteriorated, leading to the concentration of low-income residents in the centre of cities (Sule, 1994). As in rural areas, the solution often advocated is an individualization and privatization of tenure in order to encourage redevelopment (Barrows and Roth, 1990).
Not only are group claims to land being supplanted by family rights, but within families an individualization of rights is occurring. Obtaining access to urban land by purchase rather than grant or inheritance is regarded as an achievement, offering the possibility of founding a segmentary lineage (Barnes, 1979) or passing the property on to a man's wife and children rather than his traditional successors (Dickerman, 1988). However, the situation is complex: although the increasing prevalence of individual sales may give women with independent means a chance to purchase property and individualization of ownership may increase the rights of widows and divorcees, individualization of tenure in the names of men may represent a narrowing of rights to family land, and recourse to traditional rules of succession on the death of a property owner may disadvantage women (Dickerman, 1988; CASS, 1991).
Because transactions in land held under indigenous tenure have been either unrecognized or only partly recognized by both colonial and post-independence governments, legislation governing the exchange process has often not been enacted. One of the clauses such legislation might contain is provision either for the profits from land sales to be taxed or for sellers to be obliged to service the land in advance of sales. In Ghana, for example, levels of service provision to land disposed of by stool chiefs, as well as areas occupied under customary tenure arrangements, are inadequate. As a result, infrastructure standards and land prices are lower in the central area than in the surrounding middle- and upper-income residential areas subdivided under formal systems of tenure (Asiama, 1989; Acquaye and Associates, 1989).
The increasing problems caused by the undocumented nature of land rights and transactions are underscored by the increasing commercialization of indigenous tenure as urban pressures intensify. Evidence has accumulated of traditional payments approximating more closely to "prices" and of landholders selling land without the consent of their kin. This seems to be occurring both within pre-colonial settlements and on the outskirts of cities throughout the continent where peripheral land is held under "communal" tenure arrangements (Amis, 1990). In the area of communal tenure north of Lusaka, there was no evidence of payments being made to chiefs until the late 1970s, when reports began to emerge of rental being paid for land reclaimed from farms by the chief or headmen and of payments to the chief for consenting to the issue of statutory leases for individual plots (Van den Berg, 1984). Kironde and Rugaiganisa (1995) assert that there has been a market in land under customary tenure on the periphery of Dar es Salaam for decades. Wellings (1988) quotes similar evidence of chiefs in pert-urban areas in Lesotho terminating land rights, ostensibly to prevent overcrowding but in practice to sell plots. The subdivision and sale of land held under "communal" tenure have provided opportunities for the entry of brokers. In Burkina Faso, the traditional authorities sold land for urban development between 1960 and 1984, despite 1960 legislation that did not recognize their tenure rights (Bagre et al., 1995). In Bamako, Mali, since the late 1960s, traditional leaders have made the initial allocations of land in subdivisions, but, when these develop further, the leaders lose control over land and the remaining free land tends to be appropriated by middlemen, who act as intermediaries for rental housing developed by outside investors (van Westen, 1990). In Abidjan, customary owners of peripheral land have sold it to intermediaries for illegal subdivision since the 1950s, retaining some of the land for themselves (Dubresson and Yapi-Diahou, 1988; see also chap. 8 in this volume). Similar processes are described in Kinshasa and Douala (Caner and Girard, 1988) and on land belonging to the Lebou villagers of Cap-Vert (Navarro, 1988).
Although the traditional payment was seen as an expression of thanks, it increasingly approximates a price, albeit one in which the economics of the transaction are moderated by socio-political considerations. In Kumasi in the past, because the payment was not officially the cost of a lease, there was no incentive to lease a small plot, and most were about 1,000 m²; Ashanti chiefs made land available almost free to their subjects, and, on the basis of slightly higher payments, to other Ashantis, but charged considerably more to non-Ashantis. By the 1970s, although they offered discounts to their own subjects, they did not significantly tax non-Ashantis. Asabere (1981a) recognizes that the socio-political objectives of chiefs were satisfied by the granting of subsidized land to indigenes, but criticizes the practice, on the basis of very limited evidence, for encouraging inefficient utilization of land by Ashantis while repelling proposed investment by non-Ashantis from Kumasi to other places. Korboe (1994), on the other hand, considers the high densities produced by family land and housing to represent efficient space utilization, and Tipple and Owusu (1994) suggest that, as payments come more closely to approximate prices, the potential for leasing smaller plots is increasing. In Accra, however, the overall constraints on availability of stool land have led to shortages of subsidized plots, queues, and litigation (Asabere, 1981b; Asiama, 1995). In Conakry, Durand-Lasserve concludes that there are several prices depending on the tenure status of land sold by indigenous landholders, in particular related to the prospects of regularization. As a result, although prices are obscured by a "facade of customary procedures" (Durand-Lasserve, 1994, p. 59), he asserts that the market is in practice quite efficient.
Access to land through customary tenure procedures did in the past enable low-income urban residents to obtain access to land for house construction without resorting to illegal occupation. As a result, squatting has been less extensive than in other parts of the world (see below), and the major problem faced by urban residents has been inadequate provision of infrastructure rather than access to land. However, the increasingly commercialized process by which land held under indigenous tenure is disposed of, the markets that are developing in such land, and the processes of house production with which they are associated mean that the traditional responsibilities of community leaders are giving way to profit-making, opportunities are being provided for brokers with purely commercial aims, and the process of illegal subdivision and house production on this land is increasingly resembling that on land held under forms of individual tenure. Before examining the characteristics of the latter, the attempts of post-independence governments to intervene in land supply will be examined.
Urban land "reforms" and their consequences
In the years immediately after independence in many countries, the expansion of civil services and diplomatic corps, in some cases success in attracting foreign investment, and later the expansion of the indigenous bourgeoisie gave rise to increased demand for housing at the middle- and upper-income end of the market. In the ax-colonies that had been administered by expatriate civil servants, such as Zambia, the majority lived in subsidized rental housing. Undeveloped housing finance institutions, a construction sector with limited capacity, and the expectation of indigenous employees that such provision would continue led to continued reliance on employers to provide housing and only belated development of any volume of private sector formal housing. In Zambia, for example, it was not until demand from the new indigenous élite increased in the late 1960s (following independence in 1964) that prices of land started to rise again (Van den Berg, 1984). In the settler colonies, such as South Africa and Zimbabwe, in contrast, the European (and Asian) population had invested in property both for their own occupation and as an investment. A development sector and, in some cases, housing finance institutions had evolved to meet their needs, while indigenous successors in formal sector wage employment perceived the advantages of home ownership and did not expect employers to meet their needs.
Price trends in the formal residential and commercial sectors have reflected general economic conditions, although they vary geographically within cities and between sub-markets. In Nairobi, for example, Ondiege (1989) and Macoloo (1994) agree that real land prices increased between 1979 and 1982, fell during the years of drought and economic crisis between 1982 and 1986, and subsequently rose rapidly as economic growth picked up with the help of a boom in coffee prices. Within cities, status, location, and especially infrastructure give rise to price variations, as Obudho describes for Nairobi (chap. 9), while price levels and trends may also vary between tenures. Rising rents and prices and evidence of speculation in land, as well as the increasing difficulties governments were experiencing in fulfilling their independence promises to improve the quality of life for the African majority, led in the 1970s to a number of attempts at reform. These were related in one way or another to land in public ownership.
Part of the state land inherited at independence was used for public purposes. As the supply of appropriately located state land was exhausted, many countries attempted to strengthen their capacity to expropriate land in communal or individual ownership, their success depending largely on the level of compensation payable. State land was also available for subdivision for private sector house construction, potentially for all income groups. Government held the responsibility, under a variety of pieces of legislation and administrative agencies, for subdivision, survey, title registration, allocation, development and building control, and taxation. Generally, the legal basis for the sale of land to private developers or upper-income households was different from that for the provision of public housing or serviced plots to lower-income households. Invariably, however, the procedures were complex and time consuming and made more so by a lack of human, technical, and financial resources. In Gabon, for example, registration of title takes 6-24 months, while in Cameroon it takes 15-18 months in easy cases, and 2-7 years is not uncommon (Mabogunje, 1993). Cities attempted to maintain high planning standards and control over development, in part because of the political clout of high-income residents seeking to ensure their access to high-quality residential environments. However, bottlenecks in the process of survey and registration resulted in a shortfall of supply, so such land started to command a premium. Access was gradually restricted to the wealthy and well connected.
Sometimes, large stocks of publicly owned land and rapid subdivision resulted in a relatively well-planned process of urban development. In Côte d'Ivoire, for example, 11,000 building plots were made available between 1974 and 1981. This continued an earlier pattern by which, since the 1930s, the state had allocated plots on which allottees built courtyard houses; 85 per cent of Abidjan's households were accommodated in such houses as tenants. In 1984, 88 per cent of the city's area had been legally subdivided, although many plots were undeveloped (Dubresson and Yapi-Diahou, 1988, and chap. 8 in this volume). Still, by 1990, only a quarter of Abidjan households were living in unauthorized housing, 79 per cent were tenants, and the median rent was only 13 per cent of the median household income of renters (UNCHS/World Bank, 1993). In Conakry between 1963 and 1985, one-third of the extensions to the city's area were in public subdivision schemes (Durand-Lasserve, 1994). In the 1960s in Egypt, the construction sector and land development were nationalized and extensive sales of serviced land to public concessionary companies whose profits were limited by law, together with large-scale public sector house construction for rent, accommodated a large proportion of middle- and lower-income households (El Kadi, 1990; see also chap. 4 on Cairo in this volume). Elsewhere, large amounts of public sector land were used for sites and services schemes. In Malawi, for example, the allocation of state land to low-income residents kept pace with demand until the mid-1970s (Pennant, 1990).
However, sustaining such large-scale subdivision of state land was problematic in the face of diminishing supplies of land in public ownership and changing economic circumstances that either made land and property development more profitable or reduced the revenue available to finance infrastructure installation and land administration. Even in Zimbabwe, where government capacity is greater than in many other countries, slow land delivery for low-income housing has resulted in failure to fulfil targets, let alone needs (Rakodi with Withers, 1993; Musandu-Nyamayaro, 1994). Attempts by international agencies to finance sites and services schemes in the 1970s and 1980s were rarely successful in supplying sufficiently large quantities of affordable plots to reduce their attractiveness to households outside the intended target groups. Nor did they succeed in developing in-country capacity to formulate and implement appropriate land, planning, and construction sector policies (Rakodi, 1991). The supply of formal sector plots by either public or private subdividers is, therefore, generally restricted. As a result, the allottees or purchasers are either those with social and political influence, bureaucratic connections, and access to financial resources, or those who can trade a promise of political support for land allocation.
About half the countries in sub-Saharan Africa have nationalized land and converted freehold into leasehold tenure, mostly in the 1970s or at independence, although Burkina Faso and Côte d'Ivoire had done so in 1960 and Algeria in 1962 (Bagre et al., 1995; Ajavon et al., 1995; Boumedine, 1995). They included a few that believed they were carrying on the principles underlying traditional African communal tenure practices, colonies in which customary tenure had never been recognized, some that hoped to increase efficiency in the allocation of land for public and other uses, some that pursued the reforms for ideological reasons, and some that saw them as part of a wider programme of collectivization of production (Mabogunje, 1993; see also Simon, 1992).
In Tanzania, for example, land was nationalized in 1967 and rental buildings (mostly owned by Asians) in 1972. Since 1974 land has been leased through a process of administrative allocation, and property cannot, in theory, be transferred or mortgaged without government consent. In practice, the fees payable have been much less than the market price of the land and costs of infrastructure would have been, with the result that service provision has lagged behind subdivision and many plots have remained undeveloped (Kombe, 1994). In Zambia similarly, freehold was converted to leasehold, land declared to have no value, and transactions in property made subject to government approval at prices reflecting only the value of unexhausted improvements. The elimination of land prices, it was suggested, encouraged extravagant use, while delays in issuing leases created the need and opportunities for "oiling the wheels" in the relevant agencies.
The 1978 Land Use Decree in Nigeria aimed at introducing a uniform system of land administration, making land available to government for its needs, increasing equity, and curbing speculation. In practice, the nationalization of undeveloped land above a 0.5 ha ceiling in urban areas gave rise to resistance; the establishment of Land Use and Allocation Committees at state level to issue certificates of occupancy has provided scope for inefficiency, delay, and corruption; inconsistencies in the legislation have facilitated evasion; administrative requirements have favoured civil servants and businessmen with wealth and connections; inadequate mapping has inhibited implementation and enforcement; and land has become increasingly concentrated in the hands of the privileged. The reform has improved government access to land but has not replaced customary allocation systems or achieved its other aims (Okpala, 1982; Dickerman, 1988; Okolocha, 1993; see also Abiodun on land problems in Lagos, chap. 6 in this volume).
In Ethiopia, it was estimated that in the early 1970s 5 per cent of Addis Ababa's population owned 95 per cent of the privately owned land. The revolutionary government placed a ceiling of 0.5 ha on urban land, limited each family to one house, and prohibited the sale of this property; the remaining land and property were nationalized (O'Connor, 1983; Went et al., 1990; Mabogunje, 1993). Rents fell by up to a half for nationalized dwellings but maintenance responsibilities were shifted onto the public sector. Houses with low rents were managed by neighbourhood associations (Keble), which retained 15 per cent of the revenue. However, the low levels of the rents resulted in inadequate maintenance. Private house construction was discouraged by the prohibition of speculative house construction and private renting, while sales were deterred by limits imposed by government on the prices charged. Subdivision of state land was increasingly insufficient to keep pace with demand, and illegal occupation, as well as transfers for a price, proliferated. The decree was eventually revoked in 1990 and leasehold tenure re-established, although sale of leaseholds had not commenced by 1995, resulting in an explosion of peripheral illegal settlement (Assess, 1994; Gallup et al., 1995). In Uganda, the mailo land granted to the Buganda at the beginning of the twentieth century was nationalized, but the land information system was poor and much of the development that took place was illegal.
Because of declarations that land had no value or of administrative price setting that failed to keep pace with market prices, land subdivided and sold by the private sector was commonly subsidized, at least on the surface (Asiama, 1989). This, together with the shortfall of supply, increased its attractiveness to developers and the incentives for corruption (Simon, 1992). In practice, the official fees often did not equal the payment made for land, which should be taken to include bribes at the going rate, whether officially or illegally subdivided plots are being considered. The dividing line between these land markets is becoming increasingly blurred: the systems coexist and are closely related in both economic and political terms, as illustrated by Durand-Lasserve's (1994) account of how land prices are arrived at in Conakry. While price trends are related to wider economic and political conditions, in charging bribes for procedures to register an illegal plot, officials expect an amount equal to a proportion of the resale price of the plot. The "price" for an illegally subdivided plot is thus comprised of its market price, together with registration fees and bribes, and generally totals 70-80 per cent of the price at which registered and legally subdivided plots in public subdivisions resell. Durand-Lasserve quotes similar findings from other francophone West African countries.
Despite the ostensible "reforms" to land policy, the basic colonial rules and procedures have not been changed. Instead, states have used them to benefit their own clients, despite the lip-service paid to widening access to land and home ownership by means of public programmes and land nationalization. In some cases, for example Algeria and Burkina Faso, re-privatization of land has occurred (Boumedine, 1995; Bagre et al., 1995). Two other forms of informal land and housing supply remain to be considered: squatting and illegal subdivision of individually or state-owned land.
Illegal land supply
In colonial economies based on migrant wage labour, as has been noted, the government tried to limit the urban population to that needed to fill unskilled and semi-skilled wage jobs in the formal employment sector. No housing was provided for urban residents not in formal wage employment and in practice the supply of housing for wage-earners in the formal sector lagged behind demand, even before independence. Residents providing informal sector services or reluctant or unable to return to their villages were unable to gain access to public or employer rental housing. Where areas of indigenous tenure abutted the urban area, such residents were able to rent or even buy a plot or a house from a holder of customary tenure rights. Where undeveloped land was in public or private ownership under European tenure, this option did not exist. In some instances, European landowners allowed employees or retired workers to settle on their land, or permitted huts to be built on payment of a monthly rental. Around Lusaka, these owners soon lost control over the land, especially after independence, when it became politically unacceptable for them to charge rent. Legally, the existing and new residents on such land, on land with absentee European owners, and on state land became squatters. Further settlement occurred, sometimes incrementally, occasionally (for example in South Africa) by invasion, but more often by means of an adapted indigenous tenure allocation system.
Thus, in Lusaka in the 1960s and 1970s, the political party organization in a squatter area issued permission to occupy land owned by the state or absentee European owners, and even in some areas subdivided plots. Failure to understand the nature of private individual tenure meant that such house builders regarded the process of land settlement as an exercise of their familiar rights to usufruct tenure of unutilized land, whereas urban administrators regarded it as illegal and periodically attempted demolition. Interestingly, in the upgrading programmes of the 1970s, regularization was based on an occupancy licence that gave existing house owners use rights for 30 years. In the pressured situation of Nairobi, councillors, ward representatives, and members of the national legislature intervene in the semitraditional allocation system in squatter areas, to advocate community interests, represent formal sector interests, or mediate between the community and the bureaucracy. Alternatively, they may use their position to obtain access to land, often by initiating land companies or cooperatives, which have bought out many of the original owners (Lee-Smith, 1990; Yahya, 1990). Achieving access to land for the urban poor is thus a source of varying levels of cash income and political support.
As cities have grown, pressure on land has increased, and prices have risen, the scope for squatting has been reduced. In cities where indigenous tenure was predominant it was never widespread; elsewhere the proportion of existing land settlement that is squatting has been reduced by regularization and upgrading, and the proportion of new land occupation that is squatting has declined. Whereas in the late 1970s 12-25 per cent of the population of Egyptian cities, for example, lived in self-built houses on publicly owned land, squatting has become less important than illegal subdivision in the supply of new land. Squatting is now confined to the poorest, who occupy cemeteries, roofs, etc. (Soliman, 1988; El Kadi, 1988; see also chap. 4 in this volume).
Illegal subdivision has overtaken all other means of residential land supply in volume in many African cities. On private land, it is the subdivision and sometimes the sale and the construction that are unauthorized rather than the rights of the landholder. Illegal subdivision of state land, often by public officials, also occurs. In both cases the prospects for regularization vary over time and between cities - from almost certain to unlikely.
In Cairo between 1970 and 1981, 84 per cent of all housing units were produced informally, over half by adding additional floors to existing buildings and the remainder by illegal subdivision of private or government-owned desert land (UNDIESA, 1990; Metwally et al., 1995). As described by Yousry and Aboul Atta in chapter 4 in this volume, the interventionist housing policies of the 1960s were abandoned in favour of an extreme form of economic liberalism in 1973, leading to an explosion of investment in property, fuelled by remittances from migrants in the Middle East. Local development companies moved from the production of middle-income rental housing to the production of high-income housing and commercial development within the built-up area. Peasant owners of urban fringe land responded to the massive increase in demand by selling or subdividing their land. Such sales were not illegal, although the change of use from agricultural to urban use and the construction were (Steinberg, 1990). The opportunities available encouraged the evolution of subdivision and development companies able to develop and market on behalf of owners with insufficient capital. Other owners financed the process by building incrementally using rental income, or by securing key money from a prospective tenant with remittances to invest. Typically, walk-up blocks were constructed, especially along the main roads. Some developers were fined, but demolitions rarely occurred and eventually infrastructure was installed and several suburbs have been regularized. Many of the developer-entrepreneurs have created roles for themselves as local notables, using the indebtedness to them of land purchasers to gain political support and access to political position, and then using this position to obtain infrastructure connections. Analysts suggest that, despite the loss of agricultural land, which the state deplores, the satisfaction of popular demands by illegal subdivision and construction and the means of social control implicit in patronage relations between developers and residents have ensured state toleration of the activity (El Kadi, 1988, 1990; Pickvance, 1988). The massive surplus of capital available as a result of oil price increases and Middle Eastern demand for labour fuelled both extensive redevelopment, favouring large-scale capital and resulting in considerable oversupply of high-cost apartments, and profitable urban extension, which provides opportunities for smaller-scale capital, peasants, and international migrants. Although the process has slowed down more recently, companies offering a range of relevant services have continued to flourish (El Kadi, 1990). Although the government has now adopted a regularization policy, the approach is non-participatory, gives no scope for retaining land for infrastructure provision, and has given rise to fears that areas will become overdeveloped, leading to environmental deterioration (Metwally et al., 1995).
In Tunis, public habous (waqf) and French-owned land was nationalized on independence in 1956 and much of the agricultural land transferred to individuals in 1970. Because of land shortage within the urban boundary and the inability of the majority to afford the ready-built public housing being produced, one-third of the new housing in the later 1970s occurred in illegal subdivisions undertaken by both owners and agricultural tenants on state land. The whole process is less elaborate and sophisticated than in Cairo: information on land for sale circulates informally, no brokers are involved, and purchasers undertake building themselves. However, notaries work with subdividers to produce statements that land is vacant, as a basis for registering their ownership and to give purchasers the appearance of legality. The lack of infrastructure makes the land affordable to middle- and low-income groups, while subdividers, as in Cairo, use the indebtedness of purchasers buying land by instalments to achieve political office, which is in turn used to strengthen their position by obtaining infrastructure (Chabbi, 1988; Stambouli, 1990). Since the 1970s, many of these settlements have been regularized and upgraded (Chabbi, 1995).
Although the process of illegal subdivision was delayed in Fès, Morocco, by the availability of cheap rented housing in the inner city, some peripheral development is now occurring. As prices have increased and areas have been regularized and upgraded with World Bank assistance during the 1980s, the poor have gradually been excluded and driven back into the medina, further increasing its physical deterioration (Escallier, 1994).
Much of Nairobi's recent peripheral expansion has been on illegally subdivided privately and state-owned land, resulting in a dearth of land in public ownership for legal subdivision. The sale of undeveloped plots in sites and services schemes by officials also occurs. A Bill is currently before parliament to regularize illegal allocations of state land (Maine and Macoloo, 1995). In places, officials even took it upon themselves to allocate state-owned land where they were not authorized to do so, for example in Korogocho in Nairobi (Lee-Smith, 1990). The unofficial control exercised by state employees over land allocation in Kinshasa is described by Piermay in chapter 7 in this volume.
This extended account of the processes by which land is made available for residential development has drawn attention to a number of features that characterize cities continent wide, although to varying extents depending on the impact of colonization on indigenous tenure and social organization, subsequent economic conditions, and state policies. The informalization of land subdivision and development in the face of rapid urban growth and limited state capacity and the commercialization of formerly non-commercial processes are both widespread, although there has been little official adaptation of land policy and legislation to either. Inextricably linked to the supply of undeveloped land are the processes of house production and transactions in property within the built-up area. Both have been referred to in passing, but will now be discussed specifically.
The production of housing
Because the precondition for house construction is access to land, most attention has been devoted to this issue. Land is, however, only part of the residential package - the provision of infrastructure is closely associated with subdivision of land and has a major influence on land prices. In this respect, publicly initiated provision has lagged far behind needs, not least because of the failure of government to recognize and register various forms of land occupancy as a basis for property taxation (Durand-Lasserve and Tribillon, 1995). This section, is, however, concerned with the production of the shelter itself.
Despite urban housing policies since independence that have stressed the responsibility of governments and local authorities to meet the housing needs of the poor, state production of housing (even in partnership with individual households, as, for example, in sites and services schemes) has only ever met a small proportion of the need. In Nairobi, for example, between 1976 and 1987 only one-tenth of the increased numbers of residents were accommodated in publicly initiated housing projects, half in sites and services schemes (Ondiege, 1989). In Cairo, public sector housing also accounted for only about 10 per cent of total house production between 1960 and 1983 (Soliman, 1988; El Kadi, 1988). As Yousry and Aboul Atta note in chapter 4 of this volume, many of the public housing units constructed in new settlements outside the city were unaffordable despite subsidies, and they remain unoccupied (see also Metwally et al., 1995). In both cases, the proportion of total house production initiated by the public sector has declined over time and the dwellings or serviced plots made available have not been affordable by the lowest income groups. At the other end of the continuum has been Zimbabwe, where there is little squatting and illegal subdivision, and almost all low-cost dwellings have been developed in serviced plot areas in recent years (Rakodi with Withers, 1993).
Whether in the serviced plot schemes that have typified more recent public housing programmes, in areas of illegal subdivision, or in areas with tenure based on indigenous systems of land allocation, the majority of houses in both the formal and informal sectors have been built on an individual rather than mass produced or speculative basis. Some medium- or large-scale formal sector industrialized production has occurred, typically of public sector or employer housing or of private sector housing for rent or sale. In most middle- and upper-income housing areas, houses and walk-up apartments are commissioned by developers or their owners for occupation or rental from small and medium master builders. In the middle- and lower-income areas where subdivision and construction are unauthorized, housing may be commissioned from small formal sector and artisanal builders. Wherever there is de facto security of tenure, construction takes place by means of labour-only contracts and is generally incremental. The predominance of informal construction and the need for gradual construction are necessitated partly by the lack of credit for house construction, because of both the lack of title and the underdeveloped nature of the housing finance sector. These together account for the low ratio of total value of mortgages in 1989 to total formal and informal investment in housing, which was, for example, 25 per cent in Rabat, 20 per cent in Tunis, 12 per cent in Nairobi, 8 per cent in Abidjan, and 3 per cent in Dar es Salaam (UNCHS/World Bank, 1993).
Although sometimes the use of artisanal builders is supplemented by household labour, self-help (autoconstruction) has long been replaced by self-managed construction (autopromotion) (Coquery, 1990). Quite apart from gaining access to land, accumulating savings, the assembly of building materials, and the selection of a builder take urban knowledge and are, today, more rarely embarked upon by recent migrants than in the 1950s and 1960s when land, traditional construction skills, mud for bricks, and even thatch for roofs were relatively easily available. Today, only the makeshift dwellings of some squatter areas, temporary shelters used during the early stages of house construction, and backyard shacks are built by self-help and even these, in places, may be assembled from prefabricated components.
Although access to land is the single most important component in housing production, lack of capacity in the housing finance, construction, and building materials production sectors is in many African cities a constraint on house production. However, mechanisms to assist individual households to save enough to buy land or a house or to start construction have developed to alleviate the first constraint, including instalment purchases and saving via rotating credit associations. In Enugu, Nigeria, for example, recent data show that 78 per cent of the housing stock has been owner built, that three-quarters of the buildings contain tenants (and 23 per cent are wholly let), but that only 35 per cent of builders had had access to formal institutional credit. Almost all these had combined a bank loan with informal sources, including personal savings, rotating credit associations, social clubs, and inherited money (Osondu and Middleton, 1994). There seems little point increasing the capacity of housing finance institutions until a wider range and greater quantity of housing is eligible for formal sector credit. Lack of capacity in the formal construction and building materials production sectors in many African countries is caused primarily by lack of foreign exchange and skills. However, when building regulations are relaxed, as they are in most informally produced housing, it is clear that the small-scale sector has considerable capacity and is often more adaptable in times of economic crisis than are formal sector firms (Coquery, 1990). The liberalization of import controls on components for high-cost buildings, the lack of policy attention to the development of indigenous construction and building materials construction sectors, as well as the retention of inappropriate building regulations can hinder house production, as illustrated by the continued dependence of many countries on imported building materials even for housing (for example, 37 per cent by value in Nairobi, 45 per cent in Dakar, 35 per cent in Abidjan, and 25 per cent in several other cities; UNCHS/World Bank, 1993).
This section has focused on the production of housing. For the majority of urban residents in most African cities, house ownership is aspired to. This includes both for own occupation and for the generation of rental income. However, such an aspiration is by no means universal, either between or within cities. In parts of Africa, urban residents maintain strong ties with their home communities, often leaving their families behind to farm and directing their savings into remittances for both consumption purposes and investment in land, cattle, or housing. Although less commonly than in colonial times, many urban residents intend to retire to their home areas and prefer to rent in town, for example in Ghana (Korboe, 1994) or Kenya (Andreasen, 1987). An intention to retire elsewhere does not, of course, deter every resident from becoming an urban house owner, because the appreciating value of a house is seen as both a good investment and something to pass on to his or her children (Dickerman, 1988). Single people, newly formed households, and very poor households, as well as those living apart from spouses and children, may not aspire to ownership either because they prefer the limited responsibilities of rental housing or because they have insufficient resources to embark on house construction or purchase. However, for many poor households, especially those headed by women, access to home ownership forms an important part of their survival strategies if it can be achieved, because it potentially provides a secure source of shelter and income from renting. Rental housing markets will be further discussed in the next section.
House purchase or construction depend, however, not just on aspirations but also on opportunity. Even in areas of customary tenure, not all achieve owner-occupation. Peil's study of West African cities found that owner-occupiers were typically men aged 50 or more, especially those who were self-employed, and that ownership was often the result of a long process of planning, organizing, and saving (Peil with Sada, 1984). Barnes (1979), in an earlier study of Mushin in Lagos, found that, despite the attractions of ownership in terms of economic returns and social status, fewer than one in five of migrants succeeded in acquiring urban property. In areas of public land, access depends on political and bureaucratic connections, as well as luck (when sites and services schemes, for example, are massively oversubscribed) and economic status (often the ability to pay bribes in addition to proof of a regular income). The proportion of urban residents who achieve home ownership varies widely, from around two-thirds in Tunis, Johannesburg, Ibadan, and Dakar to under half in Algiers, Rabat, and Harare, one-third in Lilongwe and Cairo, little more than a quarter in Dar es Salaam, Nairobi, and Accra, and only one-fifth in Abidjan (UNCHS/World Bank, 1993).