Customary system as ‘constraint’ or ‘enabler’ to peri-urban land development: Case of Kisumu city, Kenya

Scholarly research on land delivery processes and land tenure dynamics in the context of accelerated urbanization and massive rural-urban land use conversion in sub-Saharan African cities remains limited, especially in Kenya. Kisumu city, Kenya’s third largest urban agglomeration, is currently experiencing spontaneous growth and uncontrolled outward expansions of urban development from the urban centre. This is taking place in the peri-urban areas and rural hinterland, where urban-based actors, under the pressure and anticipation of urbanization, undertake land transactions, land tenure conversions, and land development. Making use of three case studies that represent contrasting phases of settlement formation (incipient, intermediate, and consolidated), this article attempts to provide knowledge, particularly on the relationship between the land delivery processes, land tenure dynamics, and the role of various actors involved in the context of land development. The results indicate that the current trend with regard to landholding in the case study areas was a gradual shift from community control to individualisation. There is a need to integrate both customary and public authority’s efforts towards orderly urban development. This study thus advocates for partnerships between the municipal authority and neo-customary/customary actors and, where appropriate, uses their creativity in solving problems to adapt municipal by-laws and procedures.


INTRODUCTION
A common view of sub-Saharan Africa's context of land tenure is that the system is characterised by dualism: a tenure system founded on customary values, servicing a livelihood system dependent on agriculture, the use of natural resources and social relations of patronage and reciprocity, and a tenure system regulated through a legal framework derived from the colonial era and servicing modern needs and globalising economy (Adams & Turner, 2005: 2). Despite the increasing importance of private ownership on land and the establishment of a well-operative property market in sub-Saharan Africa, customary land tenure represents the dominant system for landholding that is vibrantly active (Payne, 1997: 167). Moreover, many still view the current customary land tenure situation in sub-Saharan Africa as regressive. Various scholars, however, have demonstrated that customary tenure systems can be flexible and responsive to changing economic circumstances (Cotula, 2007: 103;Wehrmann, 2008: 78;Durand-Lasserve & Selod, 2009;Wiley, 2011: 4).
This article focuses on Kisumu City, Kenya's third largest city, to contribute to the contemporary discourse on urban land tenure policy and delivery systems in sub-Saharan Africa and their implication for urban land administration, planning and development in rapidly evolving African cities. Urbanization and urban growth in Kenya have been proceeding at a rapid pace over the past five decades since independence. According to the Kenya National Bureau of Statistics (2011: 1), the population of Kenya has considerably increased since it stood at 2.5 million in the first count in 1897, rising to 5.4 million by 1948. During the first post-independence census in 1969, the population was estimated at 10.9 million, and had increased to 38.6 million by the time of the 2009 census. The crude birth rates (CBR) and crude death rates (CDR) are the primary determinants of the increase in the population growth, since international migration to Kenya is negligible. The current growth rate increases the total population by roughly 1 million individuals every year, with an expectation that it will double within the next 23 years.
As per the 2009 population and housing census, Kisumu City's urban population stood at 504 322 persons, representing 52.1% of the total population (Kenya National Bureau of Statistics, 2011: 8). This growth puts Kisumu City among urban centres with the highest population growth rate in Kenya, bringing with it associated complexities in urban planning such as, for instance, considerable pressure on land demand for urban development, in general, and housing, in particular (Olima, 1997: 325). The problem is further compounded by the fact that government alienated land in the urban area that is either unsuitable or too small and scattered to be of any substantial and adequate use for urban development.
Logically, a major requirement for all urban development is land. Baross and Linden (1990: 13) and Kaitilla (1993: 253) emphasise that no land should be brought forward into use for urban development without proper advance integrated planning to take care of infrastructure provisions. At present, the form and structure of Kisumu City's urban fringe is far from the traditional development sequence of planning, servicing, building, and occupation.
Despite the abundance of land in Kisumu City's expanding urban area, its availability is critically constrained, due to protracted and complex land tenure arrangements (Olima & Obala, 1998: 113). In the peri-urban areas, successive local authorities have shown an apparent weakness in controlling or guiding urban land use development. Property owners can do as they wish with their lands, subject to family and communal obligations that may apply depending on the prevailing land tenure. Municipal officials have found it nearly impossible to achieve harmony, delight and efficiency in the planning process when ownership of large parcels of land adjacent to the city is held under customary tenure.
Against this background, this article seeks to explore how the customary system of landholding in peri-urban Kisumu adapts and adjusts to the needs of rapid urbanization and how it enhances or constraints urban land development.

Urban population growth and the conversion of rural land into urban land
As cities grow in population and extent, the task of managing them becomes ever more complex.
Various studies have shown that rapid urban growth throughout the developing world has seriously outstripped the capacity of the vast majority of cities to provide adequate basic services for their citizens (Farvacque & McAuslan, 1992;Mabogunje, 1992;Olima & Obala, 1998;Fekade, 2000;Lupala, 2002;Rakodi, 2002;Dowall, 2003;Sliuzas, 2004;Rakodi & Leduka, 2004;Kombe, 2005). Yet each year cities  Cohen (2006: 63) concurs that this "rush of people from the countryside to cities" frustrates local authorities' attempts to proactively improve basic infrastructure and deliver essential services, especially in the newly developed areas of the city. Firman (2004: 347) acknowledges that recent urban development in cities of developing countries has been characterised by rapid changes in the use of land in urban centres, as well as by the conversion of prime agricultural lands to residential areas and other urban land use on the peripheries.
Land is essential for all urban development; yet in the majority of cities, there have been no effective measures to control land development. One consequence of this is that the urban morphology in most of the developing cities and towns is characterised by a 'leapfrog' development pattern. Kaitilla (1993: 254) maintains that leap-frog planning does not proceed in an orderly manner, making planning objectives of convenience and efficiency obsolete. He also explains that, due to obvious difficulties in acquiring land, only those lands that are easily acquired are developed, leaving large tracks of undeveloped land in between. Nonetheless, there are obvious benefits to property owners in expanding urban areas, as urban development (for instance, utilities and infrastructure provision) encroaches upon such land. Property owners stand to benefit from improved land values through the transfer thereof to urban land uses.
In his study of the urban land management system in Kenya, Olima (1997: 319) points out that the expansion of town boundaries often resulted in land that was previously held on freehold tenure by individuals and (or) trust land (not registered in individual or group titles, but occupied by a given community and governed by customary law) being brought into the jurisdiction of local authorities. For that reason, referring to their study of secondary towns in Kenya, Olima and Obala (1998: 114) note that the provision of facilities and utilities in the urban fringe has gradually fallen short of need. They found that the inability to control or guide land use development added to the rising costs of land and made it virtually impossible for the local authority to find adequate and appropriate sites for low-income housing, schools, health services, and other infrastructural facilities and utilities.
A number of studies reviewed (Olima & Obala, 1998;Rakodi & Leduka, 2004;Kombe, 2005) point out that, due to limited capacities, local authorities often fail to proactively contain, at an early stage, the growth and consolidation of informal settlements. The entry point in most of the cases is reactive to existing situations. Intervention takes place when development has already taken shape. Thus, people in urban areas not only wallow in abject poverty, but also do not live in planned urban areas, which are "orderly, safe, convenient and healthy living environments" that is promised as the gains of land use planning (Falade, 2003: 6). This clearly undermines the efforts to achieve patterned development and increases the misery of unplanned settlements; a setback to meeting the United Nations Sustainable Development Goal 11; aimed at making cities and human settlements inclusive, safe, resilient, and sustainable, by 2030.

Defining land tenure and property rights
Simply stated, land tenure refers to the relationship among people with respect to land (Payne, 1997: 3). Property rights are a bundle of characteristics (exclusivity -'who'; inheritability -'how'; transferability -'where', and security -'when'), that define the use of land (Owusu & Agyei, 2007: 14). Figure 1 illustrates the characteristics of property rights as measured along the axes in four-dimensional space.
The higher the numerical value (on a scale from 0 to 1), the more perfect the property right. Thus, a measure of 1 implies complete certainty in the respective property right. Zero implies lack of certainty.
Briefly, the content of these characteristics includes: • Security/quality of title refers to the ability of the owner to withstand challenges by other individuals and maintain his/her property right. • Exclusivity refers to the ability of the holder of property rights to utilise and manage the resource in question without outside interference. Enforcement is an important aspect of exclusivity. • Inheritability refers to the time span of the property right. • Transferability refers to the ability to transfer the property right to someone else. The ability to subdivide (divisibility) the property into smaller parts for the purpose of transfer is an important aspect, in this instance.
According to Payne (1997: 8), any rights that a person holds in/on land may be considered property. He maintains that rights exist in parallel with ownership, for instance, use rights; access rights; control rights, and transfer rights. Since multiple property rights can be associated with land and the resources on/under it, this is often referred to as a 'bundle of rights'.

Customary tenure and property rights
Customary tenure, also referred to as traditional, indigenous or communal land tenure, is the mode of holding rights in land that exists through customs and traditions (Kalabamu, 2000: 305). Various scholars note the distinctive features of this system: • Customary land is not subject to personal ownership, although use rights are alienable (capable of being sold or transferred) within and between members of the community (Agbosu, 2000: 9); • There is an individual right of occupation and use of land -the 'right of avail' -that is uniformly applied to all and automatically shared by all people belonging to a particular community, tribe, or clan (Kalabamu, 2000: 306); • The rules governing the acquisition and transmission of property rights are usually explicitly and generally known, although not normally recorded in writing (Payne, 1997: 4), and • Authority over the allocation of rights may be vested in a group of elders, tribal chief, or king (Ogolla & Mugabe, 1996: 5).
Recent studies on land-related issues in sub-Saharan Africa reveal that lands held under customary tenure are coming under pressure from expanding rural populations or the growth of urban areas (Olima & Obala, 1998: 114;Kombe, 2005: 114;Mends, 2006: 12).

Private land tenure and property rights
Private land tenure is also referred to as individualised, modern, or statutory land tenure (which includes freehold and leasehold tenure) (Arnason, 1999: 18). Private land ownership and the registration of individual property rights is largely an imported concept within developing countries. Through a series of proclamations, decrees, and orders in councils, tracts of land were expropriated from Africans by European settlers and colonial administrators and, subsequently, territories were divided into native reserves and European land. While customary land tenure provisions remained operative in areas reserved for natives, private land tenure systems were imposed on expropriated land (Payne, 1997: 4). Settler populations were later issued with various forms of ownership titles by colonial governments. The remaining expropriated land was vested in governments of the colonising power as crown land for allocation to future settler populations (Payne, 1997: 5). Kalabamu (2000: 308) points out that, in private land tenure systems, land rights are defined by law and supported by documentary evidence; the title deed, unlike under customary law where active occupation or usage is the main evidence of ownership, entitlement, or existing interest. Thus, the private land tenure systems imported with them the concept of 'land ownership', which he notes, has two meanings: • "A bundle of rights to own, control, use, abuse, and dispose of land" under Roman law. This is closely related to freehold tenure, under which land ownership is exclusive and perpetual.
• "Ownership of certain interests (bundle of rights) in land which are defined, secure, guaranteed and, most important of all, can be transferred on the market at the will of the owner" under market capitalism. Payne (1997: 4) posits that private landownership permits the unrestricted exchange of land and property markets, in which the balance between supply and demand is achieved through the pricing mechanism. It is also intended to ensure the most efficient and flexible use of land and normally produces the easy transformation of rural land into urban use, subject to statutory controls and their enforcement. In many countries, however, a common issue of concern with this concept is its inability to ensure equitable access to land or property by lower income groups.

Neo-customary tenure system
In urban and peri-urban areas, the customary tenure system does not operate essentially as described above. Instead, a mixture of reinterpreted customary practices with other informal and formal practices are in operation, referred to as the neocustomary tenure system (Durand-Lasserve & Mattingly, 2003: 3). Cleaver's (2002: 11) concept of 'institutional bricolage' (a process whereby current and traditional social and cultural relationships are reshaped into new institutional arrangements -either consciously or unconsciously) provides a striking comparison to the neocustomary system. Based on their extensive studies in sub-Saharan African cities, Durand-Lasserve and Mattingly (2003: 11) make the following observations about the neo-customary tenure system: they treat land rights as market commodities; they are based on trust, which provides a degree of security of tenure that may not be given through other informal land delivery systems, and they involve central and local government institutions. Neo-customary actors play an important political and social role at settlement levels. They are often able to position themselves as necessary intermediaries between the communities and the administration (even where they are not officially recognised). The higher the numerical value (on a scale from 0 to 1), the more perfect the property right. Thus, a measure of 1 implies complete certainty in the respective property right. Zero implies lack of certainty. Payne (1997: 27) Figure 2 illustrates the urbanization process and resultant changes in land tenure and urban land use practices. Due to push-and-pull factors, migrants seek a foothold in the city to exploit opportunities for a meaningful and better life. This ignites the process of urbanization and its consequences for land access and rights. For instance, the acute shortage of land and high rents in the city, facilitated by an overheated urban land market, lack of employment opportunities, and weakening of sociocultural networks and ties. The deepening of this situation leads to a rush for peri-urban land and land speculation. The lack of, or inadequate institutional * According to Wehrmann (2008: 2); illegality implies violation of both land tenure and planning laws

STUDY AREA
Kisumu City, Kenya's third largest urban agglomeration, is situated on the Kavirondo gulf, an arm of Lake Victoria (Figure 3). Kisumu City's origin dates back to 1898, when it was a railway terminus. In 1901, Port Florence, later renamed Kisumu, was established after the completion of the Kenya-Uganda railway. In 1903, Kisumu was gazetted as a township occupying an area of roughly 4.025 km 2 . A grid structure largely determined the shape of the colonial town (Figure 4a). The development of several stations along Kisumu and its hinterland accelerated the development of the town between 1902 and 1914.
Between 1915 and 1930, Kisumu continued to grow within the grid structure and culture-specific sector frameworks, as defined in the first period. This phase was conspicuously characterised by a more concentrated development as opposed to the scattered and far-flung developments with enormous open spaces in between (Figure 4b). Additional development included a 1923 boundary amendment to include the aerodrome areas, which were later reverted to native reserve in 1930.
In 1930, the town boundary was reduced to make it more manageable (UN-Habitat, 2005: 17). Land was allocated along ethnic grounds. Hence, racial segregation divided the town and greatly influenced further planning and policy (Figure 4c). In 1941, Kisumu town was granted the status of a municipal board. The urban structure and development in this period were characterised by demarcation of new outlying areas. Colonial power relations became more pronounced, and race-specific neighbourhoods and segregation on environmental health basis developed (Figure 4d).
In 1960, Kisumu was elevated from a municipal board to a municipal council. Notably, Kisumu was witnessing an increased development of villages that were growing in its periphery, with particular focus on Nyalenda and Manyatta (Mwamburi & Adolwa, 2006: 6). UN-Habitat (2005: 17)  boundaries of Kisumu municipality were extended to include a total area of 53 km 2 . This process caused the division of the town into two distinct land tenure zones, namely land under leasehold in the old boundary, and land on freehold in the extensions. Until 1972, when the boundary of the municipality was extended, unplanned settlements fell under the jurisdiction of the Kisumu county council and the district administration of the rural Kisumu district. This meant that any planning that was applied in this peri-urban region was of rural type, and was, therefore, less demanding of developers and urban planning. In 2000, following the elevation of the town to city status, the municipal boundaries were further extended to include approximately 65% of Kisumu district. At present, the city covers an area of 417 km 2 (157 km 2 of water and 260 km 2 of land), with a population estimated at over 500 000 people.

RESEARCH METHODOLOGY
This study examines the nature and dynamics of customary landholding; identifies the existing land markets, and it assesses the impact of the customary tenure system on urban land development processes in three settlements in Kisumu City. Following a mixed research approach, quantitative and qualitative data are collected in parallel, analysed separately, and then merged (Creswell, 2014: 45). In this study, three case studies were selected that represent contrasting phases of settlement formation (incipient, intermediate, and consolidated), to provide insights into the nature and dynamics of the customary system of landholding. The interviews explored the nature and dynamics of customary landholding; identified the existing land markets and assessed the impact of the customary tenure system on urban land development processes. The reason for collecting both quantitative and qualitative data is to elaborate on specific findings from the cross-case analysis such as similarities and differences of customary landholding among the three cases (Creswell, 2014: 47).

Interviews
For the household survey, 90 customary and private property owners were interviewed on the nature and dynamics of the customary system of landholding in the three case study areas. Using property owners' details records from a spatial database of the Pamoja Trust, the number of property owners for the study areas totalled 5 382 (Nyamasaria = 113; Nyalenda 'B' = 4 178, and Manyatta 'A' = 1 091). The property owners were selected through systematic random sampling followed up by simple random sampling combined with purposive/judgemental sampling. This resulted in a sample size of 90 (30 in each study area) (Creswell, 2014: 47;Janesick & Defelice, 2015: 18). Based on their expertise and professional roles in land planning, another 18 interviewees were selected including 8 from government, 5 from the private sector, 2 from universities, and 3 village chiefs. This results in a total sample of 108.
Taking into account that the total population from the three study areas is 5 382, a sample size of 108 is very low, but it is representative of the customary and private property owners. The sample size is not valid and not within the recommended sample size of 357 for a population equal to or over 5 000 (Krejcie & Morgan, 1970: 608).

Rationale for case study selection
Case study selection was guided by two criteria. Cases were selected to represent contrasting phases of settlement formation (incipient, intermediate, and consolidated). In brief, this would provide insights into the nature and dynamics of the customary system of landholding. The cases selected were former rural villages that were incorporated into the municipal area, following the extension of their administrative boundaries. These settlements are often referred to as 'informal' by municipal officials, apparently because they have been developed without approval from the municipal authorities. Likewise, land development in these areas does not conform to town planning standards. In Nyamasaria, active conversion of rural agricultural land to urban uses is presently occurring. The settlement is undergoing rapid land subdivision and urbanising devoid of sufficient infrastructure and utilities.

Case 2: Nyalenda 'B' (INTERMEDIATE PHASE)
Nyalenda 'B' is a settlement in rapid transition. The original inhabitants of the area are selling land to new developers who have been putting up quality residential houses (UN-Habitat, 2005: 29). The area is thus characterised by a chaotic mix of new and older but expanding pockets of housing.
Most of the buildings have no approval by the municipal authorities.
Most of the land is held on freehold tenure. Although officially abolished and replaced by freehold tenure, it is common to find pockets of land where tenure relations are regulated by customary law. Families hand down land to successive generations. as is typical in a rural set-up

Case 3: Manyatta 'A' (CONSOLIDATED PHASE)
Manyatta 'A' is a fully consolidated and cosmopolitan settlement, exhibiting a sense of order. Most of the residents have land title deeds as a result of the previous government's slum improvement efforts, with the World Bank's assistance in the late 1970s.

Data collection
For the case study, the main method of data collection was cross-sectional, including interviews with key informants, and a structured questionnaire survey with the property owners.
To understand the nature and dynamics of customary landholding, identify the existing land markets, and assess the impact of the

Data analysis and interpretation of findings
Using the SPSS software program, content analysis (selective coding) was used to examine and group components mentioned in the interviews and questionnaire surveys on land delivery processes and land tenure in Kisumu City. Content analysis is a technique that relies on coding and categorising the data (Stemler, 2001: 137). Once the responses from the interview and survey results were analysed, the themes were coded and categorised based on frequency of occurrence, including the nature of customary land access and rights over time; typology of land markets; various land actors, including their specific roles and motives; characteristics of observed land markets; the nature and type of development carried out on land held under customary tenure, and the roles of traditional authorities in urban land development.
Data analysis took place in two stages, namely 'within-case analysis' and 'cross-case analysis'. In the 'within-case analysis', the findings in relation to the seven categorised themes, for each single case, were explained using the process of data reduction, data display, and drawing conclusions and verifications (Miles & Huberman, 1994: 75).
The 'cross-case analysis' used the seven categorised themes, in each case study, as a whole entity, by using line-by-line coding of each household's interview responses, followed by a comparative analysis of all three cases. In addition, the findings of each case study were comprehensively and individually presented, as each case study represented an independent information-rich experiment (Yin, 2014: 47).

The nature and dynamics of customary landholding
To establish the nature and dynamics of customary landholding in the study areas, the key informant interviews and questionnaire surveys find out how customary land access and rights have evolved over time under the pressure of market forces. The findings reveal that the customary tenure system in the case study areas is not static, but has adjusted and adapted itself over time through several stages: from the communal control stage, through to the family control stage, and lastly to the stage of individual titles and land dealings, as depicted in Table 3.
The case study findings indicate a progressive shift from communal to individual land ownership and related rights to land; hence corroborating Payne's (1997: 7) view that with urban land markets rapidly becoming more commercialized, land has a potential of being transformed from a resource with a use value to a commodity with a market value.

Existing land markets in peri-urban Kisumu
To identify the existing land markets in peri-urban Kisumu, questionnaire surveys and key informant interviews were used to find out how customary property rights are made accessible and acquired within the study areas.
In order to achieve this, the structure and operation of the peri-urban land market was examined. The study findings corroborate similar findings by other scholars (Kombe, 1995;Lupala, 2002;Olima & Obala, 1998;Rakodi & Leduka, 2004). Both formal and informal land markets operate in Kisumu's peri-urban area. Shortcomings of the state land management mostly in regard to land allocation have led to the development of informal land markets. Informal land market transactions in the case study areas are highly decentralised and enjoy local recognition and support as evidenced by the fact that most of the processes involved are undertaken and finalised by the local actors.

Customary land markets
It is clear from the data analysis that informal land transactions are by far the most widespread means of access to land in the case study areas. The expansion of Kisumu's urban area and growth of peri-urban areas over time has transformed what were initially customary-owned agricultural farms into peri-urban residential areas. According to the younger community members interviewed, "The little effective customary control over land sales has been diminishing. Nowadays, a buyer's origin is of little importance. Besides, selling land to outsiders brought development and availability of social services in the area." This different mindset reflects the gradual decline over time in the extent to which customary norms govern land transactions.
The study findings revealed that the sale of customary land was motivated by emergency situations (distress sales) such as payment for children's education; meeting unexpected medical expenses in the absence of alternative income sources, and family land disputes. Market forces rather than social considerations usually determined the selling price in transactions involving outsiders.

Neo-customary land markets
In the case study areas, neocustomary land markets usually operate within a monetary and commercial framework, which was foreign to the customary land norms. In principle, the land control board, which convenes in Kisumu City, must approve all proposed land market transactions within the former agricultural land in the peri-urban areas before they can proceed. The board's main purpose was to regulate land subdivision and sales of agricultural land; control fragmentation of agricultural land, and ensure that families were not left destitute because of land transfers. In practice however, this is not always adhered to.
More often, an intending seller or buyer directly declares to a known confidant (for instance, relative, friend, land agent) his/her intention to sell or buy land. In either case, the seller or buyer also declares the land size; location preferences, and his/her offer. Information about land availability is by way of mouth. Negotiations, site visits and preliminary price bargaining follow in that order. These findings corroborate those of Kombe (1995) in his study of informal land markets operation in Dar es Salaam.
Apart from the land reference numbers issued during the land adjudication process, the vast majority of customary property owners do not have land titles or administrative authorisation from the respective land control boards to transfer the land in question to the purchaser. In this case, a land sale agreement is issued; it is often typed to make it more "official" and signed by the buyer and seller, the witnesses of both parties, and authenticated by the local village elders and assistant chief. Once the buyer has acquired the land sale agreement, s/he may proceed to occupy and  Figure 6 shows the neo-customary land market transaction process. Table 4 presents the actors and roles played in the informal land market.

Impact of the customary tenure system on urban land development processes
To determine the impact of the customary tenure system on land development processes in the study areas, the study finds out how urban land development proceeds on land held under customary tenure. In order to achieve this, two aspects were examined: the nature and type of development carried out on customary tenured land, and the roles of traditional authorities in controlling and managing urban land development.
In relation to the nature and type of development carried out on customary tenured land, the study findings indicate that the techniques and procedures adopted by property owners often conflict, and are sometimes incompatible with the formal land development norms. The sequence of development follows what Baross and Linden (1990: 57) term the OBSP (Occupation -Building -Servicing -Planning) model. Under the existing planning legislation, unless an individual has a registered title to his/her land, s/he Technically speaking, any development on their land is considered illegal, due to lack of approval from the municipal council of Kisumu. Nevertheless, the type of buildings in these areas are not generally different from those on official land. Perhaps, this clearly reflects the confidence that the property owners have in the customary land tenure system and the similarity between planned areas and those areas under the customary land tenure systems.
Other transgressions were commonly found in the areas of low coverage and quality of infrastructure; habitation of hazardous land, and uneconomical household land sizes.

Customary tenure as "enabler" to peri-urban land development?
In relation to the role of traditional authorities in controlling and managing urban land use development, the study findings indicate that customary authorities still play a significant role in wider land management aspects such as encouraging community mobilisation and local participation in community development projects (for instance, building of stormwater drainage channels and ensuring that land developers do not block rights of way).
In Nyamasaria (incipient) and Nyalenda 'B' (intermediate) settlements, customary authorities were also involved informally in land transfers (mainly in terms of witnessing land sales), land access, and land dispute resolution.
The picture was quite different in Manyatta 'A' (consolidated) settlement, where local residents and public authorities distantly recognised the role of customary authorities. Community development committees, acting in partnership with the municipal council of Kisumu, had officially taken up the roles of customary authorities. These committees comprise various stakeholders, including representatives from neo-customary actors, who were responsible for policy decisions such as settlement planning and development.

CONCLUSIONS
The motivation for the study was twofold. First, to comprehend how the customary landholding system adapts and adjusts to the needs of rapid urbanization, the present study was guided by the expectation that, as urban development proceeds, the informal institutions that govern land transactions and use change over time, vary between residential areas, and sometimes break down (Rakodi & Leduka, 2004: 7;Ikejiofor, 2006: 13). Based on the study findings, one observes that, as settlements consolidate and the density of development increases (due to the pressures of the urban land markets and increased land demand), the ties between the customary authorities and the urban population weaken considerably with time and may eventually breakdown. Thus, the current trend with regard to landholding in the case study areas was a gradual shift from community control to individualisation.
Secondly, taken as a whole, the results of this study show that the role of traditional authorities in urban land management is especially strong in the incipient phase of settlement development. However, the consolidation of settlements, due to demographic pressures and overheated land markets, has curtailed the efforts of traditional authorities to intensify further urban land management. Given the strengths and weaknesses inherent in both public and customary land management approaches, the study concludes that there is a need to integrate both customary and public authority's efforts towards orderly urban development. The incipient phase of settlement development serves as a possible entry point to such initiatives before settlement densities build up and the ensuing consequences -the lack of apparent planning and sense of order -are often too ghastly to contemplate.

RECOMMENDATIONS
In view of the study findings and conclusions, the following measures regarding the management and development of urban land are necessary.
Past studies by several scholars have clearly shown that the capacity of the state to intervene and foster planned land use development has been declining year after another. The study thus advocates for partnerships between the municipal authority and neo-customary/customary actors and, where appropriate, uses their creativity in solving problems to adapt municipal by-laws and procedures.
Planning policy should recognise the danger posed by peri-urban areas, and the current practice of confining development plans to urban areas should be discouraged. A comprehensive policy statement for the development of these peri-urban areas is long overdue and should be effected soon to go hand-in-hand with the rapid urbanisation process in the country.