1 Enumeration as a grassroot tool towards securing tenure in slums: An initial assessment of the Kisumu experience Working Paper, 2008 Marie Huchzermeyer School of Architecture and Planning, University of the Witwatersrand, Johannesburg Independent report of a peer evaluation mission to Kisumu 2-4 October 2007. Prepared in cooperation with Hakijamii (Economic and Social Rights Centre) Kibera Human Rights Foundation and Pamoja Trust and Slum Dwellers International 2 Abstract Global Land Tool Network (GLTN) recognises community-based slum enumeration as a potential grassroots mechanism tool which could assist in achieving its objectives relating to land management and tenure security. This paper analyses a community-based slum enumeration exercise in Kisumu in relation to these objectives. The enumeration was carried out in Kisumu as part of a city-wide slum upgrading initiative. The paper draws on a peer evaluation that included interviews with slum upgrading stakeholders as well as community-based focus group discussions, mainly with enumerators. The paper finds that coordination of the slum upgrading initiative, and beyond this of wider and often competing city initiatives, is imperative for a grassroots enumeration exercise to link up effectively with the planning authorities and for grassroots trust to be sustained for ongoing verification and updating of the enumeration data. Key findings towards securing tenure include the importance of various forms of mobilisation that accompany enumeration, and of the informal and formal knowledge generation that results from the enumeration process. Contents Abstract ............................................................................................................................... 2 1. Introduction ..................................................................................................................... 3 2. Enumeration as a global land tool towards securing tenure .......................................... 4 3. The Kisumu slum upgrading initiative and its challenges .............................................. 6 3.1 The context of Kisumu’s slums ...............................................................................6 3.2 The slum upgrading initiative ..................................................................................7 3.3 Governance, institutional capacity and coordination...............................................8 4. Enumeration in Kisumu’s slum upgrading initiative...................................................... 11 4.1 Conceptualisation of and reflection on the grassroots enumeration approach and questionnaire........................................................................................................11 4.2 Mobilisation through enumeration and the role of savings groups ........................12 4.3 Constraints and tensions in carrying out the enumeration ....................................14 4.4 Changes in landlord tenant relations.....................................................................16 4.5 Challenges of data verification ..............................................................................17 5. Conclusion .................................................................................................................... 18 References ........................................................................................................................ 20 3 1. Introduction Grassroots enumeration has been underway in Kisumu’s slums since 2005. The complex process was near completion in October 2007, when a peer evaluation was organised by UN-HABITAT, the Global Land Tool Network (GLTN) and other stakeholders. The purpose of the evaluation was to establish the relevance of grassroots enumeration as a global tool to secure land tenure. In particular, the evaluation was to identify lessons learnt from community-based slum enumeration in Kisumu, recommendations on grassroots engagement on the use, management and updating of the enumeration data and its integration into the official planning system. A further objective was to analyse stakeholders’ recommendations on securing of tenure, and grassroots involvement in this (GLTN, 2007:5). This paper draws on the peer evaluation process and report. It starts by looking more closely at the idea of global land tools and the origins of the particular enumeration approach that was envisaged for Kisumu. The paper then contextualises the enumeration exercise within the city of Kisumu, its slums, the slum upgrading initiative and the governance and coordination challenges that this initiative is confronted with. It then reviews the enumeration experience in Kisumu, covering aspects of mobilisation, tensions experienced in carrying out the enumeration, impacts such as changes in landlord-tenant relations and challenges of data verification. Whereas a parallel process of grassroots spatial surveying was undertaken in Kisumu’s slums, this paper reviews only the enumeration/data- collection exercise. In the conclusion, the paper highlights the importance of coordination and makes recommendations for a closer association between grassroots enumeration and securing of tenure. The team that participated in the evaluation mission to Kisumu included the Kenyan human rights organisation Hakijamii Trust1, a grassroots organisation Kibera Human Rights Foundation from the slum Kibera in Nairobi2 and an independent consultant (author of this paper). The Slum Dwellers International (SDI) participant was unable to attend. However, SDI organised an exchange group of Zimbabweans (two government officials, three Federation members and a staff member from the NGO Dialogue on Shelter) who joined the team in order transfer learning about grassroots enumeration to Epworth in Zimbabwe. The NGO Pamoja Trust3, which was facilitating the slum enumeration in Kisumu, participated in the evaluation team while also organising the focus group discussions. The peer evaluation involved interviews, mainly in groups, with representatives of the main stakeholders in Kisumu’s slum upgrading initiative and the enumeration exercise: the Kisumu Municipal Council, in particular is Planning Department and GIS (Geographic Information System) Team; the Physical Planning District Officers from all the districts in Nyanza Province in the Ministry of Lands; the Institution of Surveyors in Kenya (ISK); the Regional Centre for Mapping of Resources for Development (RCMRD); the NGO Pamoja Trust facilitating the slum enumeration exercise; the NGO Sustainable Aid in Africa (SANA) which is involved in sanitation and governance in several slums in Kisumu; and the NGO Network that operates in the region. Three extensive focus group discussions were conducted with grassroots participants of the enumeration exercise. Participants were selected by the NGO 1 Elijah Odhiambo (Coordinator, Hakijamii Trust, Nairobi). 2 Marcy Kadenyeka.. 3 Lawrence Apiyo (Senior Programmes Officer, Pamoja Trust, Nairobi). 4 Pamoja Trust and represented most of Kisumu’s slums. Participants were mainly enumerators and members of savings groups of the slum dweller’s federation Muungano, which were being established with Pamoja Trust’s support and in close association with the enumeration exercise across most of Kisumu’s slums. The composition of the focus groups generated useful insight into the practicalities and experiences of undertaking the enumeration on the ground, and useful grassroots recommendations were articulated. However, the focus group discussions did not elicit a representative view of the enumeration process among slum residents, those merely having provided their information without direct involvement in the enumeration activities. The focus group discussions required translation for many of the participants, and it is expected that some of the original detail and nuance was lost in this process. Given the sensitivities of some of the aspects discussed, the identities of both stakeholders and focus group participants are not provided in this paper. 2. Enumeration as a global land tool towards securing tenure Land and the recognition and administration of land rights remain a central concern in addressing the issue of slums (Lemmen, Augustins, Oosterom and van der Molen, 2007). Formal land administration systems have relied on registration of individual rights, often economically excluding ordinary people (Augustinus, 2005), and therefore contributing to the inadequate living conditions that are broadly captured under the term ‘slum’. The invisible slum characteristic of insecure tenure often underpins the more visible slum characteristics of inadequate shelter, lack of basic services, overcrowding and exposure to health risks and physical hazards (see UN- HABITAT, 2003:11). The need for new systems of land administration was confirmed at the 2006 World Urban Forum (Lemmen et al., 2007). This includes new systems for the management of land information. Lemmen et al. (2007:5) advocate for a Social Tenure Domain Model, through which ‘a spatial unit forms the basis for recording the people-land relationship’. The focus is on ‘social tenure relationships’ or recorded (but unregistered) rights, thus ‘personal and not real rights’ (ibid.). This model is understood as a means of including ordinary people such as slum dwellers in a recognised system that protects tenure rights, thus also forming a basis for the improvement of the lives of slum dwellers, as articulated in the Millennium Development Goal (MDG) 7 Target 11. The Global Land Tool Network (GLTN), an emerging international network that promotes ‘a holistic approach to land issues’, places a strong emphasis on the inclusion of grassroots, in addition to governments, non-governmental organisations, civil society and professionals, in developing appropriate land management tools (Langford, 2007:4). With the term ‘grassroots’, GLTN refers to those at the local level who are disempowered due to land deprivation and poverty, who would be users of land systems and/or who may be organised into social movements or civil society organisations (ibid.:8). They are considered therefore to have particular insights that should be harnessed into the development of land administration systems. GLTN’s work on ‘grassroots mechanisms’ is concerned with effective grassroots participation in the development of land administration systems. GLTN identifies the grassroots mechanism of enumeration as one global land tool in which grassroots participation should be evaluated in the context of an ‘ongoing land tool design and implementation process’ (Langford, 2007:32). Grassroots enumeration was developed in the first instance as a 5 [M]eans by which not only the data is gathered to allow for local planning, but also the process by which consensus is built and the inclusion of all residents is negotiated. Enumeration is a means to federate and organise communities and involve them in large scale slum upgrading projects (GLTN, 2006:4). This approach to enumeration that incorporates consensus-building and mobilisation was developed by the Slum Dwellers International (SDI) affiliated Indian NGO SPARC through its extensive work with pavement dwellers in Indian cities since the 1980s (see SPARC, 1985). The approach has since been promoted through SDI affiliates in several other countries, including the NGO Pamoja Trust in Kenya (see Weru, 2004). Within the SDI, enumeration is one component of a broader methodology which involves in the first instance mobilisation of (primarily) women through savings and credit groups in order to build membership-based federations of the urban poor. The SDI’s approach to addressing inadequate housing among Federation members is to promote modelling and self-construction of housing. This requires land, and where Federation members do not have access to land, as in the case of the Indian pavement dwellers, grassroots enumeration plays an important role in creating visibility to policy-makers, in improving organised demand-making and in forming the basis for partnerships with relevant government administrations or agencies (SPARC, 2002; Huchzermeyer, Karam, Stemela, Siliga and Frazenburg, 2006). Where relocations are imminent, enumeration is considered a vehicle for consensus-building within the community around the process of relocation (Patel, d’Cruz and Sundar, 2002). The various components of SDI’s methodology, be they mobilising through savings and credit groups, house modelling and self-construction of housing or enumeration and mapping, are replicated through international exchanges of grassroots groups, supported by SDI. The components are considered mutually reinforcing, thus an enumeration exercise also has the important function of disseminating information about, and mobilising for, savings groups. For instance, in the case of an enumeration exercise on access to water in Tanzanian informal settlements, a grassroots exchange with South African and Zimbabwean SDI affiliated federations promoted, in addition to the enumeration methodology, the idea of savings groups (Glockner, Mkanga and Ndezi, 2004). For the purposes of GLTN, grassroots enumeration as a land tool should capture tenure conditions, including various levels of insecurity of tenure (GLTN, 2006). However, this is not to the exclusion of the dimensions of enumeration developed by SDI, namely improving demand-making for basic services other than land, and mobilisation and building of self-reliance through savings groups. It is against this backdrop that in 2004 UN-HABITAT (though prior to the formation of GLTN), unable to identify any Kisumu-based NGOs with the capacity to carry out door-to-door data collection and mobilisation in Kisumu’s slums, requested the Nairobi-based SDI- affiliated NGO Pamoja Trust to facilitate a slum enumeration exercise in Kisumu. GLTN has a particular interest in the lessons learnt from Kisumu’s grassroots enumeration exercise and their relevance to the promotion of a continuum of land rights, through, for instance, a social tenure domain model. In addition to such lessons, the evaluation of the enumeration exercise in Kisumu exposed limits to what enumeration can achieve in a context of competing interests. 6 3. The Kisumu slum upgrading initiative and its challenges 3.1 The context of Kisumu’s slums Kisumu is Kenya’s third largest city. It hosts headquarters of Nyanza Province as well as the Kisumu District. In the 1999 census, Kisumu had a population of just under 350 000 people and the annual growth rate was estimated at 2.8%. With 53.4% of its population surviving below the food poverty line (compared to 8.4 for Nairobi, 30% for Nakuru and 38.6% for Mombasa), Kisumu is Kenya’s poorest city (City of Kisumu, 2004). Given its status as ‘leading commercial/trading, industrial and administrative centre [hosting various Lake Victoria regional bodies] in the Lake Victoria Basin’, it is also one of the fastest growing cities in Kenya (ibid.:10).Kisumu’s hinterland is agriculturally rich, supporting sugar and rice industries. Kisumu’s so-called belt of slums (see figure 1) surrounds the formally planned city centre in a semi-circle, though with a concentration in the topographically flat east, where development occurred largely without planning. Poor drainage in this area imposes high infrastructure costs. Kisumu’s slums resulted from the incorporation of formerly rural areas into the urban boundary. Luo villages had been subdivided through inheritance and developed into higher densities, and as agriculture became unviable tenancy catering for city-ward migrants resulted in increased densities. With incorporation into the urban area, land was largely adjudicated and registered as freehold (though with general boundaries, rather than accurate stakes on the ground, as explained by the representative of ISK). In this process, however, the municipal council did not acquire and set aside land for the development of infrastructure and public facilities. It is estimated that 60% of Kisumu’s residents live in areas that do not provide adequate access to basic necessities. (Onyango et al., 2005) Figure 1: The location of Kisumu’s slums (based on Onyango et al., 2005:x,13) A land market as well as a rental market emerged in Kisumu’s slums in response to the pressures of urbanisation. The Situation Analysis (Onyango et al., 2005), which was commissioned as a basis for Kisumu’s slum upgrading initiative, reports that 7 land prices are rising in the slum belt. It further points out that landlords do not enter into formal lease contracts with their tenants, as ‘renewals are based on willingness to pay rent’ (ibid.:8). Both rising land prices and informality of tenancy agreements result in tenure insecurities that are not immediately evident on the surface. However, the dominant position among stakeholders in the Kisumu slum upgrading initiative is that insecurity of tenure is not the primary concern in Kisumu’s slums. The Situation Analysis likewise observes that ‘issues of tenure’ in Kisumu’s slums ‘are not so critical’ (ibid.:8). In its dominant form of freehold titling, the land tenure situation in Kisumu’s slums is not representative of what is internationally referred to as ‘slums’. However, slums across Kenya’s cities are characterised by processes of commodification – of shelter (through a rental market) and of basic services (through vending). The land market in Kisumu’s slums, underpinned by freehold titles, is a further dimension of this Kenyan slum characteristic. Beyond the pockets of tenure insecurity that do exist at any given point in time on invaded land or through localised market threats to economically weak tenants, there is the larger process of market change which threatens tenure security. This emerging deepening of tenure insecurity (and residents’ fear thereof) is not unrelated to land values created through processes (or even promises) of slum upgrading (Onyango et al., 2005; Huchzermeyer, 2006). Given the hidden nature of the tenure issues, the slum enumeration in Kisumu only indirectly dealt with this aspect, therefore limiting to some extent its usefulness for evaluation as a global land tool or grassroots mechanism for tenure security. However, the process of enumeration itself led to changes in landlord-tenant relations that provide interesting insights on the possible relationship between land tenure and enumeration, particularly within a social domain tenure model. Class and tribal clashes in Kisumu since the allegedly rigged December 2007 national elections, resulting in massive demolition in Kisumu and internal displacement of people, have introduced new tenure insecurities. While the slum enumeration data will have been rendered inaccurate through the displacements, it might be useful in identification of internally displaced people. This paper acknowledges this new challenge. However, it analyses and draws lessons only from the pre-election situation in Kisumu. 3.2 The slum upgrading initiative UN-HABITAT has played an instrumental role in initiating a slum upgrading initiative in Kisumu. Firstly, since 2003, UN-HABITAT has a cooperation agreement with the Kenyan government on the Kenyan Slum Upgrading initiative (KENSUP). Since its inception, KENSUP identifies Nairobi and Kisumu as the initial two cities in which KENSUP will unfold. Secondly, within the Cities Without Slums (CWS) Sub-regional Programme for Eastern and Southern Africa of Cities Alliance (a multi-lateral initiative involving UN-HABITAT and the World Bank as well as country governments), Kisumu is one of nine selected pilot cities. Through KENSUP and Cities Without Slums, the Kisumu Municipal Council was encouraged to address the reality of its slums at a city-wide level. With support from UN-HABITAT/Cities Alliance, the detailed Situation Analysis (Onyango et al., 2005) was undertaken in 2005. This consultative and consensus-building process led to an Action Plan that identified priority areas and set out an institutional framework. The slum enumeration or ‘social mapping’ exercise began in March 2005 while the Situation Analysis was being prepared, and with the same intention of forming a basis for city-wide slum upgrading. Whereas the initiative of a community-based enumeration covering all Kisumu slums and generating data for use by the GIS 8 facility predated the Action Plan, it was not integrated into the plan. In practice, it has remained largely a parallel initiative – a coordination problem to which I return in detail below. The rationale behind this enumeration initiative was twofold. The data was to be made available to the municipal council for planning purposes, while also empowering the community and ensuring effective demand-making. In conjunction with the enumeration, Pamoja Trust was tasked with mobilising or organising communities into savings groups as well as settlement executive committees – SECs (UN-HABITAT, 2004). The model of SECs was developed in the context of the Kibera slum upgrading pilot programme of KENSUP and appears to be the model that KENSUP favours. By incorporating the establishment of savings groups in Kisumu’s slums into Pamoja Trust’s brief, UN-HABITAT was tailoring the enumeration work in Kisumu to SDI’s broader philosophy and methodology. Capacity to manage enumeration data and plan effectively was created in the Kisumu Municipal Council through the establishment of a dedicated GIS Unit. The GIS facility in the Kisumu Municipal Council appears to have been the most enduring of the institutional innovations that led out of the Action Plan. However, the failure of important coordination functions that were set out in the Action Plan and to which I turn next has hampered the effectiveness of the enumeration exercise. 3.3 Governance, institutional capacity and coordination Kisumu’s slum upgrading initiative depends on collaboration between the Physical Planning Department of the Ministry of Lands and the Planning Department of the Kisumu Municipal Council. According to the Action Plan, coordination between these departments and the non-governmental stakeholders in the slum upgrading, including NGOs such as Pamoja Trust in its facilitation of the slum enumeration, was to be handled by a slum upgrading Programme Secretariat within Kisumu Municipal Council, headed by the Director of Planning, with a ‘well developed technical support unit’ (Onyango et al., 2005:72). In addition, the Action Plan recommends the formation of a Multi-Stakeholder Support Group, with stakeholders being the municipality, KENSUP, the Government of Kenya, civil society organisations, neighbourhood associations, the ‘Programme Secretariat’ and the business community (ibid.). However, while these two coordinating structures were initially established (according to the Progress Report annexed to the Situation Analysis – Onyango et al., 2005:72), at the time of the evaluation in October 2007, neither existed. From the various stakeholder interviews, it was evident that the GIS unit, also established in the Kisumu Municipal Council in support of the slum upgrading initiative, had captured all the attention. The idea of a Programme Secretariat hat fallen by the wayside. As a result, by October 2007 the city-wide slum upgrading initiative was fragmented and delayed by lack of coordination. According to the representative from the NGO SANA, peer groups working in Kisumu’s slums were yearning for coordination. With the absence of a Programme Secretariat, NGOs were fearing that KENSUP funding for the slum upgrading initiative was captured by the wrong municipal departments, hampering efficiency of targeting and disbursement of these funds. A representative of the Institute of Surveyors in Kenya (ISK) highlighted that beyond coordination, there was also a need to build trust between the various non-governmental and governmental stakeholders in the slum upgrading initiative. 9 In the peer evaluation discussions, NGOs voiced doubts over the Municipality’s capacity to fulfil a coordination function. It was considered whether NGOs could instead fill the gap in coordination. In 2005, the Situation Analysis had already identified as a challenge the lack of coordination between the many NGOs active in Kisumu’s slums. According to the representative of the NGO Network, which operates in the Nyanza Province and Western Region, NGOs were still to some extent duplicating their work in the slums. An example mentioned was the duplication in setting up community organisations – NGOs such as SANA (Sustainable Aid in Africa) were in the process of establishing representative settlement committees at the time when Pamoja Trust was commissioned to do the same. The SANA representative recalled that at the early stages of the Cities Without Slums initiative the various stakeholders had agreed to coordinate their activities in Kisumu’s slums. However, this initial commitment had lapsed and had needed to be re-addressed by the various NGOs. At the time of the peer evaluation this was not resolved. It was unclear as to what extent Pamoja Trust, due to its presence in Kisumu’s slums through the extensive enumeration exercise and setting up of community-basd organisations through daily savings, was expected to play a coordinating role. The slum data generated by the enumeration was certainly of interest to other NGOs and needed to be shared. In terms of a coordination function, a limitation for Pamoja Trust was the fact that it is a national organisation headquartered in Nairobi and did not have a permanent office in Kisumu, unlike other NGOs that are regionally based (i.e. in Nyanza province). Pamoja Trust’s staff tasked with facilitating the enumeration exercise was not based fulltime in Kisumu for the duration of the enumeration. For similar reasons, Pamoja Trust had also not joined the NGO Network, which Pamoja Trust also felt was not optimally organised and focussed. However, Pamoja Trust indicated that it was planning a series of consultative meetings with stakeholders in Kisumu. Coordination was also lacking at the grassroots level. Representative Settlement Executive Committees (SECs) were to be established as part of the official slum upgrading initiative. However, the peer evaluation revealed that various grassroots committees or structures already existed in many of the slums incorporating a system of chiefs and elders that was not necessarily representative but enjoyed legitimacy at the grassroots. In addition, Pamoja Tust was tasked by UN-HABITAT to compliment the slum enumeration by mobilising slum dwellers into savings groups under the slum dwellers (Muungano) federation. At the time of the peer evaluation, it was evident that relevant structures for community involvement in the planning, management and implementation of slum upgrading interventions in Kisumu required review. A broader concern was raised by the representative of the NGO Network, namely that there were too many externally initiated and competing programmes within the municipality, which were insufficiently coordinated: City Without Slums (CWS); City Development Strategy (CSD); Sustainable Urban Mobility; and Millennium Cities. Kisumu’s City Development Strategy (City of Kisumu, 2004) is prefaced by the following vision, which makes no mention of improved conditions for its desperately poor and marginalised residents: [A] leading transportation, communication and commercial hub in the Great Lakes Region offering great tourism and agro-investment opportunities. From the brief (courtesy) meeting between the peer evaluation team and the Mayor of Kisumu, it became evident that the various donor-driven initiatives had diverging and partly conflicting objectives and would undoubtedly compete for resource 10 allocation and commitment. The Mayor voiced enthusiasm about the upcoming Millennium Cities event with Jeffrey Sachs, an initiative that seeks to augment donor funding and attract foreign investment with the aim to stimulate the urban economy and to end poverty4. The Mayor also enthusiastically mentioned plans to build an international airport, linked to the Millennium Cities initiative and its objective of attracting foreign investment. This was evidence of the conflicts in resource allocation that were facing the municipality. Further, lack of engagement by the municipality with residents who may be evicted for the new airport was highlighted by human rights organisations Hakijamii Trust and COHRE (Centre on Housing Rights and Evictions) in their Housing Rights update of August 2007 (see COHRE and Hakijamii Trust, 2007). The challenges of aligning, coordinating and harmonising the initiatives of multilateral and bilateral development institutions with country strategies were addressed internationally, with participation of the Kenyan government, in the Rome (2003) and Paris (2005) declarations. The focus of these declarations is on meeting the Millennium Development Goals (MDGs), with an emphasis on the need for sustained coordination, as well as country-level ownership of initiatives, mutual accountability and the achievement of results. What these declarations do not address are the inherent tensions between initiatives that seek to grow urban economies with implied spin-offs for poverty alleviation (such as Kisumu’s City Development Strategy and the Millennium Cities initiative) and those that seek directly to extend social infrastructure and service delivery with a direct budgetary burden (such as the Cities Without Slums initiative). The Situation Analysis for Cities Without Slums alludes only to one aspect of this tension, namely that the vision of regional economic development is directly associated with the burden of increased migratory pull to the city (Onyango, 2005:4). In a city context such as Kisumu, which is given the status of a regional centre for economic growth yet has substantial backlogs in social infrastructure and services, this tension presents a major political challenge, which translates into difficulties for sustained coordination. Reflecting this larger governance tension between the economic and social function of the city, the newly established GIS Unit in Kisumu Municipal Council, which according to the Situation Analysis has the purpose of supporting planning and slum upgrading in Kisumu, at the time of the peer evaluation was being targeted by the municipality as a tool for revenue collection (payment records being linked to spatial units on the GIS), thus supporting an economic rather than the social imperative. The GIS unit staff also mentioned the income generation potential for the municipality of selling GIS (and slum enumeration) data to researchers. These economic interests posed a challenge for Pamoja Trust, whose representative voiced fear that the enumeration data, once integrated into the GIS Unit’s data-base, would be used for purposes that were not in the direct interest of slum dwellers. This mirrored concerns voiced in focus group discussions with grassroots participants of the enumeration. It was feared that, because of the grassroots (and regional) opposition to the Kibaki government at the time, particularly in the wake of the upcoming December 2007 elections, the enumeration data would be used against those enumerated. In early 2008, the violent aftermath of the election that retained Kibaki in power gave tragic evidence to the deep-rooted and unmediated suspicion, sense of injustice and fear among ordinary people in this region. 4 In the contemporary development literature, there is no consensus on the relationship between increased donor funding and foreign investment on the one hand, and poverty alleviation on the other – see divergence in positions on poverty alleviation in Sachs, 2005 and Collier, 2007) 11 However, seen outside of the challenging socio-political context and the lack of coordination, the GIS facility represented a major advancement in Kenyan municipal capacity. A representative of the Regional Centre for Mapping of Resources for Development (RCMRD), which had been instrumental in setting up and training for the GIS facility, shared his view of the institutional context of the GIS facility. He emphasised that while setting up of the GIS facility had overshadowed the creation of a coordinating body, the importance of the GIS facility for the work of the Kisumu Municipal Council should not be underestimated. Indeed, this was the first municipal GIS facility in the country (Nairobi’s Municipality had outsourced this function to the company Geomap). The GIS Unit had enabled a fundamental change in the work of municipality’s Planning Department, replacing a space consuming paper-based system with outdated information. Setting up of the software and data-base had been a success, thanks to the support of the Mayor and Town Clerk. High resolution satellite images from 2003 were purchased through a partnering organisation in the US. RCMRD had trained municipal officials and had seconded some of its staff to Kisumu Municipal Council for 6 months to ensure the officials were coping with the new system. High turnover and inability to retain staff once trained on GIS were challenges the municipality was faced with in terms of maximising the impact of the new facility, in particular in relation to slum upgrading. Delays in handing over of the long-expected slum enumeration data had compounded the problem. The GIS facility was planning to initiate GIS access centres within the slums, to ensure that enumeration data was analysed and used by the communities. Practicalities of this were not yet resolved. It was not clear how this would impact on the relationship between communities and the municipality and whether resulting demand-making by communities would in fact lead to improved municipal services. In the focus group discussions it was acknowledged that the municipality had already constructed toilets in some settlements, e.g. in Kaloleni. However, in some settlements, the relationship with the municipal council had deteriorated due to awareness raised in the community about poor conditions and about the duties of the municipal council. In these settlements demand-making had increased through the enumeration, but had not yielded results. 4. Enumeration in Kisumu’s slum upgrading initiative Pamoja Trust was commissioned by UN-HABITAT late in 2004 to undertake the Kisumu slum enumeration. The mutually agreed budget also provided for community mobilisation into savings groups and the setting up of representative settlement executive committees for the purposes of slum upgrading (UN-HABITAT, 2004). Key components of the SDI approach to enumeration and mobilisation were to be implemented. The questionnaire used in the social and economic mapping of Nairobi’s slum was to be discussed widely for input from Kisumu’s stakeholders including slum dwellers, before being administered door-to-door. A major awareness- raising campaign needed to precede the data collection, whereas data-collection was to be followed by verification. Pre-dating the formation of the GLTN, the project proposal from UN-HABITAT (2004) made no direct reference to data-collection on security of tenure, requesting however that both tenants and structure owners be included in the process. 4.1 Conceptualisation of and reflection on the grassroots enumeration approach and questionnaire GLTN’s definition of grassroots participation in land tools is that ‘grassroots groups fully participate in all stages of the tool development process and not simply be 12 passive recipients of solutions developed for them’ (Langford, 2007:11, emphasis in the original). It first appeared that grassroots involvement in the conceptualisation or development of the enumeration tool in Kisumu was limited, as the methodology was transferred through SDI from India via Nairobi to Kisumu. However, Pamoja Trust indicated that while the methodology originated from India, each country domesticates it based on its situation, therefore differing from country to country and municipality to municipality. In the case of Kisumu, adjustments had been made after consultation with the NGO SANA, and other stakeholders endorsed the questionnaire, giving it local ownership. While the grassroots then did not adjust the approach or questionnaire or develop the tool upfront, the enumeration process was one that allowed responses to land tenure issues (primarily the tenancy relations) to emerge at the grassroots. The Kisumu experience suggested that it might be difficult for ordinary slum dwellers to give upfront suggestions on the conceptualisation or development of a complex data-collection approach without any prior experience of such work. However, focus group discussions with the enumerators generated important recommendations for improving the approach and tailoring it to the Kisumu context – the focus group participants enjoyed the opportunity to reflect and make suggestions. This suggested that rather than expecting upfront input from communities into the methodology, it would be relevant to ensure a feedback loop for continuous improvement/ refinement/ adjustment of the tool by the participating community. Several grassroots recommendations voiced in the focus groups related to the enumeration form: - Detail about the size of the plots needed to be handled differently, as it had made landlords suspicious of intentions to grab their land. - The detailed question about how people were using their money was considered sensitive and unlikely to yield honest responses. - Disclosure of identification numbers was also considered sensitive. Further, the useful suggestion was made that the form be completed in duplicate, with the enumerated household keeping one copy. This could then be used by the household to verify whether their data was correctly represented in the final set. With relevance to tenure security, it was suggested that absentee landlords also be enumerated. One focus group also considered whether an organisation should be represented on the enumeration form, so that the enumeration process be clearly identifiable with an appropriate organisation – currently, the enumeration form for Kisumu did not indicate any affiliation. Most respondents expressed the need for such affiliation, but felt that a Muungano association on the form would probably be rejected by the respondents. The focus group considered this issue sensitive and in need of further consideration. Enumeration teams had managed to resolve the tensions they felt were generated by the enumeration form. This had required lengthy explanation and the assistance of village elders or elected councillors who played a role in legitimising the process. Some felt that this generated additional work for the already time consuming exercise. 4.2 Mobilisation through enumeration and the role of savings groups As the enumeration approach in Kisumu was based on the broader philosophy of the SDI-approach it was difficult for the peer evaluation team to distinguish the impact of the community-based enumeration from the impact of the mobilisation and learning 13 (about process as well as settlement conditions) that was occurring at the same time through savings groups and exchanges. In some settlements (in Nyalenda, Manyatta A, B and Arab and Bandani), Pamoja Trust first established savings groups or Muungano structures, and then announced or introduced the idea of enumeration through Muungano. In other settlements, Pamoja Trust was able to rely on existing community organisations and authorities to introduce and launch the enumeration exercise. Thus one of the focus group respondents noted ‘enumeration gives birth to [Muungano] savings groups and vice versa’. This respondent also indicated how the enumeration exercise had directly encouraged mobilisation to resolve the issues that emerged through data-collection: Many savings schemes came out of the enumeration, because this process made us realise the need e.g. for toilets. The importance of mobilisation for self-reliance, as promoted through the SDI approach, was explained by a Muungano member directly involved in organising the slum enumeration. He noted the prior attitude of the community of expecting per diems for any meeting attended. This had changed to an attitude of voluntary participation, which was a basic requirement for a grassroots enumeration exercise to be carried out. The selection of the enumerators and their composition into teams was an aspect of mobilisation in itself. While the composition of the focus groups suggested that most enumerators were Muungano members, this relationship was clarified by Pamoja Trust using the example of Nyalenda: Muungano chose enumerators. However, the majority [of enumerators] were not Muungano members. Pamoja Trust gave Muungano members criteria for the selection of the 250 enumerators that were needed for Nyalenda. The enumerators had to enumerate across large parts of the settlement, therefore it could not be a matter of Muungano members only enumerating in the sub-areas in which their savings groups were localised. In settlements where there were no pre-existing savings groups, Pamoja Trust approached the Area Administrator, who requested the village elders to convene community meetings at which Pamoja Trust could announce and plan the enumeration. Team leaders for the enumeration would be identified at these meetings. Pamoja Trust explained that enumerators were selected through public meetings, through calls for volunteers or put forward by existing CBOs. In a small settlement, the enumeration team would consist of 50-100 enumerators. The team would be subdivided into smaller groups, each with a team leader, and each made responsible for a section of the settlement. As part of the enumeration procedure, the enumeration teams would meet at the beginning and end of each day, to bring back the forms and share experiences of the day. At these meetings, Pamoja Trust would also talk about savings and the need to get the community organised. Savings schemes were set up in all settlements, working in the first instance with enumerators. Pamoja Trust indicated that as a result of this process, savings schemes have grown from a mere five to 102. A city-wide slum dweller-led financing facility known as the Urban Poor Fund was also set up. An enumerator in a focus group discussion explained the importance of the relationship between Muungano and the enumeration process: Muungano members were trusted and known in their communities. They were able to ‘team up’ with village chiefs to lay to rest the suspicion among landlords that the enumeration 14 questionnaire would be used for land grabbing. Muungano gave support to the enumerators even if they were not Muungano members, and the allowance for enumerators was disbursed through Muungano. Muungano also played a role in the three day training of enumerators with a team from Nairobi. Enumerators from Bandani, Obunga and Manyatta mentioned that they had been trained through an exchange visit to Nairobi..Pamoja Trust explained that exchange visits provide a holistic learning exposure to all components of the SDI methodology. Pamoja Trust explained that enumeration was announced in the communities as a voluntary activity, therefore relying on mobilisation and conscientisation. Those selected or coming forward for enumeration were asked to indicate how many days they would be able to volunteer without a stipend (this differed from enumerator to enumerator), and for the remaining days, the stipend was KShS 200/day, the equivalent of an average daily wage among slum dwellers. Focus group participants did not perceive the stipend as a payment for a job carried out. One of the focus groups considered whether enumeration could be promoted as a daily practice similar to daily savings promoted by Muungano. This would be on a purely voluntary basis, independent of external funding, and could ensure regular updating of data. Mobilisation that resulted from the enumeration went beyond savings groups. One focus group respondent noted: In Bandani, a youth group formed, because the enumeration process had made the youth realise that they’re sidelined from development. In this settlement, it was mainly youth that were chosen as enumerators. As a result, these youths became mobilised and formed their own groups. Another respondent from Bandani shared that since the enumeration had showed the prevalence as well as hardships of widows, the Settlement Executive Committee had decided to mobilise funding to assist widows with grants to improve their income. A focus group participant from another settlement noted that [A]fter enumeration, landlords realised that they needed to improve the living standards of their tenants. Therefore it was decided to register with cooperative societies. As mentioned above, a further aspect of mobilisation that was included in Pamoja Trust’s brief was that of setting up representative Settlement Executive Committees. In Bandani, a Landowner’s Association was transformed into a Settlement Executive Committee including tenant representation. However, this process varied from settlement to settlement depending on existing structures. The Situation Analysis notes that migrants to Kisumu tend to join ethnic enclaves. In its analysis of opportunities and threats for the slum upgrading initiative, it highlights as one of the opportunities the ‘creation of multi-ethnic, multi-cultural communities … organising themselves around key socio-economic issues not ethnic politics’ (Onyango et al., 2005:23). Given the post-election class/tribal violence that erupted in Kisumu three months after the peer evaluation, it would be important to establish to what extent the non-tribal mobilisation linked to the enumeration, especially where it bridged landlord-tenant as well as tribal divides, had moderated what might have been even worse clashes. 4.3 Constraints and tensions in carrying out the enumeration Among the focus group participants, there was general consensus that although the Kisumu-wide slum enumeration exercise had already taken three years, enumeration 15 in each settlement required more time. Resource and time constraints had forced Pamoja Trust to insist on enumeration in each settlement over a very short period, and with long working days for the enumerators. Numbering of doors, filling in of forms and answering questions from respondents had to be completed in 3 weeks. Pamoja Trsut added that because of the communities’ mistrust that had been built up from previous externally driven experiences, more time was needed before data could be accessed/extracted. Focus group participants indicated that they interacted with sick, elderly or disabled individuals in the normal course of the enumeration, as questions about age and disease were included in the questionnaire. However, enumerators did not have the capacity to return and follow up on the plight of vulnerable residents (although Muungano set up some support groups for those affected by HIV/AIDS). During the enumeration, particularly the more vulnerable individuals had asked what benefits would arise from the enumeration. Enumerators admitted that this had resulted in ‘the creation of expectations’, meaning that enumerators had mentioned forthcoming benefits in order to gain willingness from these residents to be enumerated. However, resistance to being enumerated was not only from vulnerable groups – even landlords feared they would be losers in the process. Expectations that had been raised by enumerators’ attempts to advocate for the enumeration were a concern raised in the focus group discussions, as those enumerated were now holding enumerators to account: To get information from the landlords on the size of their plots, we had used the tactics of saying that this information might assist in allocation of funding. As a result, we are now facing problems in our communities, as people now have raised expectations and are asking us what happened to the promised improvements. The promises made by enumerators were not unjustified, given the city’s slum upgrading initiative. However, in many settlements improvements had not yet materialised and enumerated households were making demands or reminding enumerators of these promises. Enumerators were at a loss as to how to answer, and expressed this as a post-enumeration problem or pressure they were experiencing. Focus group questions about households’ resistance to being enumerated generated much discussion and consensus. With some exceptions, tenants had welcomed being enumerated, but landlords had resisted for various reasons. Some had feared that their land would be grabbed, which suggests limits to perceived tenure security even for holders of freehold titles. Others had feared demolition – possibly a larger fear that enumeration would be used to enforce development control without adjusting building standards to the existing building fabric of the slums (the Physical Planning Department of the Ministry of Land did mention alternative building standards for the Obunga pilot). In certain cases, landlords had feared outright eviction. In Kaloleni, the fear of eviction was exacerbated by the good location of the settlement, thus the assumption that there might be plans to build high rise buildings. Landlords had also raised their assumptions of gainful employment of the enumerators and had complained that this monetary benefit was not fairly distributed. In some instances, resistance to the enumeration resulted from perceptions that the enumerators were gainfully employed, rather than volunteering for a small stipend. Some households had felt discriminated against, as no enumerators had been drawn from their areas. These tensions had been resolved. 16 Households that had resisted enumeration, even rubbing enumeration numbers off their doors, had subsequently understood the purpose of enumeration and were now asking the former enumerators enumerate them. They had to be told that the time for enumeration had passed. A door to door data verification process would have enabled these households to have their facts included. I return below to the constraints that were experienced in verifying the enumeration data. 4.4 Changes in landlord tenant relations With important implications for tenure security in Kisumu’s slums, it was evident that the enumeration process impacted on the landlord-tenant relationship. Focus group discussions, while not representative, gave evidence that attempts by enumerators to gather information from resident landlords as well as their tenants brought to the fore tension in these relations and led to some adjustments. Landlords were initially suspicious of having their tenants disclose their rentals, or more generally about their tenants being enumerated, as they feared exposure of the poor conditions of their rental stock. Landlords in Manyatta Arab had perceived that the enumeration might entice tenants to take action against them, given that the low standards of living might be found not to match the rent charged. One focus group participant exclaimed that [I]in some cases the landlords are resisting their tenants being enumerated, they even use machetes to chase [enumerators] away. Landlords also had suspicions of being enumerated by tenant enumerators, and feared that tenants would write something on the enumeration forms to benefit themselves. Were landlords were illiterate (a tendency that was observed by enumerators), this fear was justified. Landlords generally felt that the questionnaire favoured tenants. However, most focus group participants who raised these concerns added that the tensions were resolved through briefings and explanations. In addition, accounts were given of landlord tenant relations having improved or been harmonised. One landlord participating in a focus group had felt the questionnaire gave tenants high expectations of being relieved of not having latrines and adequate housing. After understanding the questionnaire, this landlord came to a compromise with his tenants and made some improvements. One tenant in a focus group acknowledged that tenants experienced the enumeration as a ‘wake-up call’. They had not been sure how to engage their landlords in a fruitful process. Now we know how we can approach our landlords and how to approach them without causing tension. This change in relationship had also awakened landlords to business principles. While tenants now had the right to ask for services, landlords felt entitled to demand rent on time. As long as services were not provided, it had been hard to press for the rent. Therefore a general focus group comment from landlords was that the enumeration process had improved their business. After providing basic services, demand for their units had increased. The focus group discussions did not establish conclusively the extent of displacement of poorer tenants through this change in the market. One focus group participant noted that tenancies tended to be short term due to contract work, therefore rents could be increased between tenancies, avoiding direct eviction of poorer tenants. The Situation Analysis (Onyango, 2005:33) 17 mentions that the Bangani and Obunga slums house a significant number of short- term contract workers from the nearby industrial zone. However, the report does not link this to high tenant turn-over or short durations of tenancies. It is therefore not clear whether this focus group observation is representative of tenancies in Kisumu. From the focus group discussions, it was also not clear whether the accounts of improved landlord-tenant relations were mainly from landlords who were members of Muungano enumerator teams, or whether this raised awareness of both social responsibility and business opportunities had occurred throughout the settlements. 4.5 Challenges of data verification At the time of the peer evaluation, the enumeration process was in its final stages. Data for most of the slums was being handed over to the Kisumu Municipal Council’s Planning Department and GIS Team for use in its planning of slum upgrading interventions. Due to the resource constraints, one crucial aspect of the enumeration approach was compromised - enumeration data was not fully verified. In the published literature by Pamoja Trust and SDI (see Weru, 2004), multiple verification, referring to a repetition of the door-to-door process, is emphasised as an integral part of SDI’s enumeration methodology, improving accuracy of the data while also building consensus around development. For Kisumu, Pamoja Trust’s representative indicated that verification of the enumeration data had only been carried out in Obunga and Bandani, while further verification was being planned. This clearly raised questions about the data sets that were handed to the municipal council, and whether it was appropriate for the GIS team to integrate this data into the GIS if a second set of verified data was still to be provided. Pamoja Trust’s representative, while acknowledging that they did not have the resources for the verification process, suggested that the initial enumeration data would be 80-90% accurate without verification. The ISK representative indicated that for planning purposes a 20% error margin was acceptable. Where data verification was carried out in Kisumu, it was not through a door-to-door process as promoted by SDI, but through mass meetings with PowerPoint presentations of the household data and physical mapping. Pamoja Trust was considering this as a regular event, particularly where communities were organised and demanded information. It was envisaged that communities could be trained to undertake this form of verification themselves. With the establishment of Muungano savings groups in the slums, it was felt that capacity and willingness for processes such as verification could be built in these communities. Focus group participants from some settlements were not aware that any verification of the data was planned. All were concerned about data accuracy. Enumeration was done mainly during day-time when key household members were not at home, some households (a figure of 10% was mentioned by an enumerator) had refused to cooperate, some withheld information on sensitive questions such as expenditure patterns, and some landlords had refused having their tenants enumerated. The other concern was about data becoming out-dated with time as new structures were being constructed and others replaced. For Obunga, the first settlement to be enumerated, it was noted in particular that since enumeration, some tenants had been evicted, some had moved from one place to another within the settlement, some residents had died and others had given birth. Focus group participants suggested that enumeration be done quarterly, and alternatively, up-dating and verification should be done before the data is used for a particular project. 18 In Obunga, the first settlement to be enumerated, sense of ownership of the data was expressed by enumerators: The data bank that has been produced is beneficial to the community. Landlords are coming to ask questions about their ‘papers’ and the Ministry of Health has asked for data. Data verification appeared to go hand-in-hand with communities’ sense of ownership of that data. In Bandani, a report of the enumeration data had been brought to the community. However, the data was ‘only with the leaders’. It was felt that the community should own the data, and should have a document. It was suggested that the data could be displayed at the Chief’s office, the schools, churches or market places, thus also playing a role in data-updating. 5. Conclusion GLTN recognises community-based slum enumeration as a potential grassroots mechanism tool which could assist in achieving GLTN objectives of establishing a continuum of land rights, developing pro-poor land management and land tenure tools, and unblocking existing initiatives. This paper has analysed a community- based slum enumeration exercise in Kisumu in relation to these GLTN objectives. Community-based slum enumeration has been carried out in Kisumu as part of a city-wide slum upgrading initiative under UN-HABITAT and Cities Alliance’s Cities Without Slums (CWS) initiative and the Kenyan Slum Upgrading Programme (KENSUP). The slum upgrading initiative in Kisumu included a detailed Situation Analysis in 2005 with an Action Plan, and led to the establishment of a GIS facility in the Kisumu Municipal Council (KMC) as well as slum upgrading interventions in some of Kisumu’s slums. UN-HABITAT commissioned the Nairobi-based NGO Pamoja Trust to facilitate the enumeration in Kisumu’s slums. Through its affiliation with the international NGO Slum Dwellers International (SDI), which had developed a slum enumeration and mobilisation approach in India, Pamoja Trust had experience in implementing SDI’s enumeration approach in some of Nairobi’s slums. Pamoja Trust found its resources stretched in facilitating the enumeration and mobilisation in Kisumu. Therefore aspects of the enumeration such as verification of the data and ongoing collaboration with the Municipality were compromised and completion delayed. Institutionally, the slum upgrading initiative in Kisumu depended on collaboration between the Physical Planning Department of the Ministry of Lands and the Planning Department of the Kisumu Municipal Council (KMC), with its GIS facility. Coordination between these and the non-governmental stakeholders in the slum upgrading, including NGOs such as Pamoja Trust in its facilitation of the slum enumeration, was to be handled by a Slum Upgrading Secretariat within KMC. However, this had not been established, and the city-wide slum upgrading initiative was fragmented and delayed through this lack of coordination. On the ground, the slum enumeration approach, which incorporates community mobilisation, had to engage with local complexities. These included a diversity of existing community structures as well as a tenure form that is atypical of slums – a tenancy system in which landlords (largely living in the slums) hold freehold tenure and therefore have relative security of tenure in comparison to their tenants. The following conclusions result from the peer evaluation. 19 According to anecdotes from the focus group discussions, the enumeration exercise in Kisumu had an initial unintended impact on tenure security by changing landlord- tenant relations. While follow-up research is required to establish the extent of such change, in the cases that were reported, this change in relations occurred not only at the level of individual tenancies, which had become fairer though at the same time more business-oriented. It also occurred at the level of community structures through the various forms of mobilisation that accompanied the enumeration and which included landlords and tenants. This joint landlord-tenant mobilisation included the composition of enumerator teams, the setting up of savings groups, the formation of issue based structures to address insights gained through the enumeration process, and the improvement of representation on community-wide structures such as settlement executive committees. The level of community voluntarism required for an enumeration exercise to be carried through relied on conscientisation through mobilisation. Anecdotal evidence was that the Muungano/SDI savings group approach played a strong role in achieving the necessary attitudinal change among residents, both landlords and tenants. An important outcome of enumeration is knowledge generation, as a basis for informing ways of addressing vulnerabilities. Provided the enumeration questionnaire incorporates more direct questions about tenure vulnerabilities and fears, enumeration can contribute to addressing tenure security at various levels. On the one hand, the process of enumeration informally generates insights, particularly among the enumerators. These insights are about the process of enumeration (and in the focus group discussions led to suggestions for its refinement), as well as conditions within the slums. In some of Kisumu’s slums, this informal knowledge had led to issue-based mobilisation responding to, for instance, vulnerable widows. If more directly incorporating questions on tenure security, an enumeration could lead to similar community-based responses to tenure vulnerabilities long before official finalisation of the data collection. On the other hand, enumeration leads to formal knowledge generation, in the form of data-sets that can be spatialised through geographic information systems (GIS), analysed and used for planning purposes. In the Kisumu case, this process was slow. Given lack of coordination of the city-wide slum upgrading initiative, there was no evidence that tenure security would be addressed in the short term through application of this formal knowledge in a planning and land management process. The formal knowledge was also subject to competing interests within the municipality: planning for improved service delivery was clearly in the interest of the slum dwellers; and development control and revenue collection was clearly not in the interest of many current slum dwellers, landlords in particular, and their current tenants by implication. Mediation in the use of the formally generated knowledge is a major requirement in order for the enumeration initiative to sustain support from within the slum communities. The incomplete and partly continuous processes of data-verification and up-dating in Kisumu (especially in a context of high tenant mobility) required sustained support from communities. Resource constraints in the completion of data verification meant an increased reliance on voluntarism and cooperation from communities. What role for enumeration in GLTN, and particularly in developing the social tenure domain model (SDTM)? Given the area-based focus of this model, it would be relevant for an enumeration to be carried out on the basis of sub-areas of a slum, 20 with enumerators drawn from that area. With their informal knowledge of that sub- area generated through the enumeration, enumerators could fulfil ongoing functions in the governance of these areas, including management of the land. Savings groups and other forms of mobilisation would then also be encouraged on a sub-area basis, reinforcing the capacity of sub-area communities to address insecure tenure and other vulnerabilities within their areas. This recommendation requires further consideration for the various land ownership situations that prevail in the slums, engagement and discussion with SDI and its affiliates, where experience across diverse contexts has been generated and could enrich the development of a role for enumeration in the social tenure domain model. This discussion may also embrace technological advances, for instance verification and up-dating of data through the use of cell phones, which have experienced an unexpected spread among slum communities and are increasingly embraced in social programmes (id21, 2007). For the slum upgrading in Kisumu, it could be concluded that the enumeration exercise achieved a number of positive results. Among these is the improved overall awareness on the lack of service delivery to Kisumu’s slums, and of intra-city inequalities. This potentially paved the way for pro-poor resource distribution through an improved revenue collection mechanism. The enumeration initiative enabled community participation in the official physical plans for two of the slums. It also led to an understanding among stakeholders in Kisumu about the need for adequate information to develop detailed physical plans at neighbourhood as well as municipal levels. The establishment of the Urban Poor Fund is also an important achievement. The slum upgrading initiative in Kisumu was constrained by several factors. Only marginal resources were set aside for slum upgrading by the Kenyan government and the Kisumu Municipal Council, therefore only a few interventions were implemented on a piecemeal basis. International development partners did not adequately engage with the slum upgrading initiatives. In addition, lack of coordination at the city level posed a serious challenge for those engaged with improving the lives of Kisumu’s slum dwellers. For the purposes of the peer evaluation, this situation helped distil the critical importance of sustained coordination at several levels – between government departments, between externally driven initiatives, between NGOs and between all slum upgrading stakeholders. Coordination and mediation particularly of the competing social and economic interests of the city is essential for any grassroots tool, whether based on sub-areas or settlement-wide, to effectively link up with the authorities and yield tangible benefits on the ground. References Augustinus, C. 2005. Key issues for the future that support and prohibit a more pro- poor approach and why such an approach is needed. Schermerhorn Lecture, International Institute for Geo-Information Science ITC, The Netherlands, 2005. City of Kisumu, 2004. Kisumu City Development Strategy 2004-2009. Lake Victoria Region City Development Strategy for Improved Urban Environment and Poverty Reduction. Kisumu Municipal Council, Kisumu. COHRE and Hakijamii Trust, 2007. Kenyan Housing Rights Update, August. Centre on Housing Rights and Evictions, Geneva and Hakijamii Trust, Nairobi. INTERNET: www.cohre.org/view_page.php?page_id=165 21 Collier, P., 2007. The Bottom Billion: Why the Poorest Countries are Failing and what can be done about it. Oxford University Press, New York. Glockner, H., Mkanga, M. and Ndezi, T., 2004. Local empowerment through community mapping for water and sanitation in Dar es Salaam. Environment and Urbanization, 16(1), 185-198. GLTN, 2006. Brief Issues Paper in Preparation for the Workshop on Innovative Pro Poor Land Tools and LIS to Deliver Security of Tenure for the Majority in Kenya. Nairobi, 30 November. Global Land Tool Network (GLTN) and UN-HABTIAT, Nairobi. GLTN, 2007. Grassroots Mechanism Tool: Mission to Kisumu. Terms of Reference (ToR), 1 Septermber - 31 October 2007. Global Land Tool Network (GLTN), United Nations Human Settlements Programme, Nairobi. Huchzermeyer, M., 2006. Slum upgrading initiatives in Kenya within the basic services and wider housing market: a housing rights concern. Discussion Paper No.1/2006. Kenya Housing Rights Project, COHRE Africa Programme, Centre on Housing Rights and Evictions (COHRE), Geneva. INTERNET: www.cohre.org/view_page.php?page_id=165 Huchzermeyer, M., Karam, A., Stemela, I., Siliga, N. and Frazenburg, S., 2006. Policy, data and civil society: Reflections on South African challenges through an international review. In Huchzermeyer, M. and Karam, A. (eds.), Informal Settlements: A Perpetual Challenge? pp. 19-40. University of Cape Town Press, Cape Town. Id21, 2007. Mobile phones and development: The future in new hands? Id21 Insights, 69. INTERNET: http://www.id21.org/insights/insights69/index.html Langford, M., 2007. Grassroots Mechanisms: GLTN Concept Paper. Prepared in collaboration with Huairou Commission, Centre on Housing Rights and Evictions (COHRE), Slum Dwellers International (SDI) and Hakijamii Trust. Global Land Tool Network (GLTN) and UN-HABTIAT, Nairobi. Lemmen, C., Augustinus, C., van Oosterom, P and van der Molen, P., 2007. The social tenure domain model – design of a first draft model. Authors’ unpublished manuscript. Onyango, G., Wasonga, G., Asamba, I., Teyie, P., Abuya, J., Odera, B. and Ooko, E., 2005. Situation Analysis of Informal Settlements in Kisumu. Cities Without Slums Sub-regional Programme for Eastern and Southern Africa; Kenyan Slum Upgrading Programme. UN-HABITAT, Nairobi. Paris Declaration on Aid Effectiveness, Ownership, Harmonisation, Alignment, Results and Mutual Accountability. Paris, France, February 28-March 2, 2005. INTERNET: www.adb.org/media/articles/2005/7033_international_community_aid/paris_declarati on.pdf Patel, S., d’Cruz, C. and Sundar, B., 2002. Beyond evictions in a global city: People- managed resettlement in Mumbai. Environment and Urbanization, 14(1), 159-172. Rome Declaration on Harmonisation, Rome, Italy, February 25, 2003. INTERNET: www1.worldbank.org/harmonization/romehlf/Documents/RomeDeclaration.pdf 22 Sachs, J., 2005. The End of Poverty: Economic Possibilities for our Time. Penguin Books, London. SPARC, 1985. We the invisible: A census of pavement dwellers. Society for the Promotion of Area Resource Centres (SPARC), Mumbai. SPARC, 2002. Demolitions to dialogue: Mahila Milan – learning to talk to its city and municipality. Society for the Promotion of Area Resource Centres (SPARC), Mumbai. UN-HABITAT, 2003. The Challenge of Slums. Global Report on Human Settlements 2003. United Nations Human Settlements Programme. Earthscan London. UN-HABITAT, 2004. Project proposal between UN-HABITAT and Pamoja Trust for Social and Economic Mapping. Cities Without Slums Sub-Regional Programme for Eastern and Southern Africa. Kisumu Initiative (Preparatory Phase). United Nations Human Settlements Programme, Nairobi. Weru, J., 2004. Community federations and city upgrading: The work of Pamoja Trust and Muungano in Kenya. Environment and Urbanization, 16(1), 47-62.