* DOCUMENT : 3

KENYA

Understanding what community federations bring to city upgrading strategies: the work of Pamoja Trust and Muungano in Nairobi and other areas in Kenya

I. INTRODUCTION

PAMOJA TRUST IS a non-governmental Kenyan organization that was set up in 2000 to help urban poor communities organize themselves to oppose demolition and forced evictions, and to develop their own plans to get adequate housing and basic services. When the Trust was established, the government was supporting or permitting many “slum” demolitions and evictions. But with a decrease in these demolitions, the Trust has focused more on supporting low-income communities to improve housing and basic services, working in partnership with Muungano wa Wanvijiji, the urban poor federation in Kenya. This reduction in demolitions is a process that began before the change of government in 2001, but was reinforced in part because of the new government and in part because of the opposition from low-income households and support NGOs.
Muungano, like many other urban poor federations in other nations,[1] has at its base saving schemes that provide a community-based structure for governance and housing initiatives. Muungano and Pamoja Trust seek to improve housing both through regularization and upgrading of the settlements in which urban poor groups currently live (including getting tenure of the land they occupy) and through new land tenure projects. Pamoja Trust is a member of Shack/Slum Dwellers International, the international network of slum dwellers federations and NGOs that work individually and collectively to improve conditions for slum and shack dwellers.[2]

II. BACKGROUND

ALTHOUGH NAIROBI IS Kenya’s capital, and a successful international city considered by many as the gateway to East Africa (with many international agencies located there), housing conditions for much of its population remain very poor. A slum inventory in 1995 found that over half the city’s population lived in informal or illegal settlements that were squeezed into one twentieth of Nairobi’s total area,[3] and Pamoja Trust’s updating of this inventory has found little improvement since then. A survey of Nairobi’s informal settlements in 1998 found very high levels of infant and child mortality.[4] In most informal settlements, there are high levels of overcrowding, very inadequate provision for basic infrastructure (piped water, provision for sanitation and drainage) and high levels of risk from accidental fires (Box 1).
Any attempt to improve conditions in Nairobi’s informal settlements is complicated by the potential conflict between landlords and tenants and by the conflicts between different ethnic groups that often have been exacerbated by the manipulations of powerful political interests. In most informal settlements, the inhabitants do not have tenure of the land they occupy. Even without official tenure, there are landlords (“structure owners”) and tenants, and these two groups have very different priorities within any programme to legalize land tenure. Structure owners want to acquire full legal tenure of the land on which their structures are built; tenants want recognition of their right to live there and the possibility of becoming land and house owners. Many structure owners are in effect large-scale (and often absentee) landlords, as they own large numbers of huts and make high incomes from renting these out.
From the time of Kenya’s independence in 1963 up to the late 1970s, official government policy was to demolish informal settlements, even though much of the urban population had no other means of getting housing. During the 1980s, this changed to a more permissive approach and there were few demolitions. During the 1990s, official policy alternated between ignoring the problem and demolishing settlements. In 1990, two large settlements (Muoroto and Kibagare) were razed to the ground, and an estimated 30,000 people were made homeless or displaced. In 1992, the return to multi-party politics and local elections in Nairobi reduced the demolitions but, until the government of President Moi was voted out of office in 2001, there were many demolitions, evictions and violent conflicts.
In Kibera, the largest informal settlement in Nairobi (with over half a million inhabitants), a rent strike precipitated violent clashes that led to many deaths, and some 30,000 people were left homeless when their huts were destroyed. There were also many deaths in violent clashes in another informal settlement, Kariobangi, in 2001, although this was linked more to the build up to the election than to landlord-–tenant conflicts.

A major focus for most urban NGOs during the 1990s was fighting these slum demolitions and the constant threat of eviction that faced so many low-income dwellers in Nairobi. For instance, the NGO Kituo Cha Sheria (literally meaning “place for all”) was a legal and human rights organization that undertook public interest litigation on behalf of low-income communities threatened with unlawful eviction. The focus was thus on using the courts and other legal means to stop or prevent evictions.
There was much discussion (and some disagreement) about what role Nairobi-based NGOs should take. At this time, with a repressive and generally anti-poor government, many community leaders and the NGOs that supported them felt that their work should centre on protests and on demands on the state to change their policies. From 1995 onwards, there were constant visits to Kenya by the Indian and South African members of Slum/Shack Dwellers Federation.[5] They suggested that the methodologies that had been developed by urban poor federations in other nations, based on strengthening and developing representative community organizations through savings schemes, should be tried in Kenya. They also argued that partnerships and working relationships had to be developed between urban poor organizations and the state – which went against the views of some NGOs and community leaders, who were reluctant to consider a partnership with government organizations. The success of the methods used by various urban poor federations (including those of India and South Africa) suggested the need to get urban poor groups doing things themselves (setting up savings groups, undertaking their own “slum” enumerations, developing plans for land regularization and housing). In part, this is because these can be done more effectively by urban poor groups than by the State. But, more importantly, these processes help build an internal community governance structure that has to be in place before a dialogue with the city government can be effective. Without internal capacity building among the poor, they end up being the “objects” or consumers of state solutions (or the lack of them). In addition, the urban poor groups also sought partnerships with government organizations where possible, because there are aspects such as land tenure legalization and trunk infrastructure provision that only the State can provide. This approach implied a considerable change in strategy for many NGOs.
During 1997 and 1998, there were many evictions and demolitions in Nairobi, and much corruption and oppression by the police and by powerful individuals. Individuals and community organizations within the informal settlements could do little without the permission of “leaders” and without making some payment to them – for instance, to be able to sell goods on the street or extend a shelter. The conflict between structure owners and tenants, and the threat posed to “leaders” and landlords by any accountable community organization, made any work with the urban poor difficult.
In addition, although the networks of NGOs and community organizations were successful in limiting the scale and scope of evictions, they were not developing models for significant improvements. Every time there was an eviction, there was much organization to prevent or oppose it but no further pro-poor activities. Meanwhile, each low- income community remained isolated. For any NGO working with the urban poor, it was not clear whether the focus should be on community organization or legal measures.

Muungano wa Wanvijiji, the Kenyan urban poor federation, began as a committee of community organizers that had been set up primarily to oppose evictions, with support from many Kenyan NGOs and from Catholic priests and human rights activists. It emerged out of dialogues and workshops organized by a Nairobi- based NGO, the Mazingira Institute, and, initially, it focused on a land rights campaign. Community organizers from Muungano took part in exchanges with the Indian and the South African urban poor federations, and these suggested the need to build representative community organizations through savings schemes and enumerations – although, again, there was disagreement among community organizers as to the usefulness of such an approach in Kenya.
Pamoja Trust was set up in 2000 as an organization to focus specifically on strengthening and supporting urban poor community organizations to access land, shelter and services, and to do so by building savings and loan schemes that the urban poor organized and managed themselves. This focus also implied a changed role for community leaders. At the time that it was formed, there were many local conflicts and difficulties. The State tried to close down Pamoja Trust and its offices were firebombed. Community organizers were arrested. And some community leaders did not want this change of direction, in part because it threatened their legitimacy and power base. With the support of MISEREOR (the German international NGO) and the Maryknoll priests who were based in Nairobi, and subsequently of OXFAM, Pamoja Trust began its work programme.

III. ENUMERATIONS

ENUMERATIONS REPRESENT THE first part of the process by which informal settlements become “regularized” with secure tenure, house construction to improve conditions, and infrastructure built or negotiated from local authorities. As this section describes, enumerations provide the means by which data are gathered to allow for local planning, but also the process by which consensus is built and the inclusion of all residents negotiated.
In December 2001, the president of Kenya (at that time Daniel Arap Moi) issued a directive that the residents of Korogocho “slum” (with a population of around 100,000) should be permanently settled on the land that they already occupied (which was owned by the government). This created considerable tension between the “house structure owners”, who were seeking formal title to the land, and the majority of residents, who were tenants and who paid rent to the structure owners. There was an active association of Korogocho structure owners (KORE), with 2,460 members, and they had lobbied the government to get this declaration. . Some structure owners owned many structures and derived large incomes from renting them out – and stood to make a lot of money, if they could get legal tenure of the land on which their structures were built.
Pamoja Trust and other groups in Nairobi sought a fairer process. The president had asked Nairobi’s provincial commissioner to oversee the process in Korogocho, so the challenge was to influence the commissioner. During a visit to Kenya from the Indian members of Shack Dwellers International, it proved possible to meet the commissioner, who then agreed to visit India to see the work undertaken by the Indian National Slum Dwellers Federation. He also agreed that the interests of tenants should be upheld in Korogocho, through a committee formed by elected representatives of both structure owners and tenants. Two structure owners and two tenants were to be elected from each of seven areas within the settlement, thus forming a committee of 28 persons. The provincial commissioner subsequently visited the Indian National Slum Dwellers Federation in April 2002 to see how the government and urban poor organizations worked together.
Jockin, the leader of the Indian National Slum Dwellers Federation suggested that there should be a full enumeration of Korogocho to produce detailed data and maps of the settlement. The Indian federation and other members of Shack Dwellers International had considerable experience with slum enumerations, and these followed a methodology that had been well established and tried in India and in other nations through SDI members. Pamoja Trust decided to try out the first enumeration in Huruma in 2000 (which was a much smaller settlement than Korogocho and where there was less conflict). Federation members from Zimbabwe, India and South Africa came to help with the Huruma enumeration.

It took a month to complete this first enumeration. A first draft of questions for the enumeration were prepared and discussed between Pamoja Trust and SDI, and then the enumeration began, undertaken by community residents. Numbers were given to each house and each household was interviewed. As work got underway, the difficulties became apparent, as did many inhabitants’ anxiety about the process. In some house structures, there were both structure owners and tenants; in others, only tenants, with the structure owner living elsewhere in the settlement; in others, only tenants, with the structure owner living outside Huruma. Some tenants would not talk (or told the enumerators that the landlord had said they should not be enumerated), while others gave false information (for instance, putting down their children as separate households or claiming that their kitchens were separate housing units in the hope that they would get two plots). The structure owners were keen to ensure that they were the people listed as having ownership of the land; in Kambi Moto, one of the settlements within Huruma, the enumeration showed that one person owned 50 structures.
But the community organizations had to work through these difficulties. Structure owners with more than one unit had to give these up if they wanted a regularized unit. A subsequent verification process, whereby the information was returned to the community, showed that the information in this first enumeration was not always accurate and that the number of resident households was much smaller than the initial numbers stated. Box 2 summarizes some of the findings from the Huruma enumeration.

Drawing on this experience, the Korogocho enumeration was planned. There was a recognition that this would prove more problematic, as the settlement was much larger and the inhabitants and community organizations more politicized. There was also a vicious informal political authority, whereby no inhabitant could build or repair their house or use land for growing vegetables without “permission” and a payment. Also, Korogocho had experienced many NGO interventions, which had created an expectation that NGOs would deliver for them. And there was a feeling that, since NGOs did not stay around, one should take as much as possible from any external intervention.
Designing the enumeration form required a long negotiation, especially with regard to how tenants would be enumerated. For instance, the issue of whether tenants should be enumerated as separate households or come under the landlord’s name was particularly difficult to resolve. This negotiation took four days. An information programme was also organized in Korogocho, with posters and fliers telling the inhabitants about the enumeration. The elected committee was formed, although many of those elected as representatives were village elders and structure owners who were part of the informal political control system.
Eventually, agreement was reached on the content of the questionnaire, but KORE (the structure owners association) still opposed this and, through its effective propaganda machinery, spread a rumour that this enumeration was part of a process through which Indians were coming to buy the land and that the director of Pamoja Trust was their land broker. Pamoja Trust was threatened, and KORE sought a court ruling to stop the enumeration. Although they failed to stop it, the enumeration had to begin under police guard. Serious violence was only avoided by good preparation – with a lot of mediation and with the Korogocho committee also mobilizing politicians to support it. At the time, there was a political alliance between the ruling party Kano (which represented the Kikuyu, the tribal group to which most structure owners belonged) and the LDP (in which the Luo were represented, a large proportion of tenants being Luo). This helped ensure high level political support for the enumeration.
On the first day of the enumeration, several truckloads of police and the entire provincial administration came to oversee the process. There was a lot of tension (and a threat to kill the director of Pamoja Trust, whose car had been identified), but in the end, there was very little violence. SDI members from India and Zimbabwe were also there to help with the enumeration. Tensions were reduced as the enumeration completed its first and then second days. The earlier experience with the Huruma enumeration had shown how to avoid some of the difficulties and bottlenecks. Three hundred community members were involved in (HELPING CONDUCT??) the enumeration, and a police escort was there to make sure that the enumeration forms got out of the settlement safely. Over a 10-day period, 18,500 forms were completed.

The structure owners association again tried to get a court order to stop the enumeration, but the provincial commissioner refused to accept it (although he too was threatened). In court, the association of structure owners sought not only to stop the enumeration but also to confirm themselves as the landowners. On the day of the court case, the settlement committee was able to mobilize 6,000 people who went to the court when the case was being heard. The case was not resolved and it has dragged on, with the date for the hearing constantly being set and then postponed. But one of the enumeration’s successes has been that it has provided the basis on which the residents have been enjoined as interested parties for the land.
As in Huruma, various difficulties emerged during the enumeration. Some enumerators asked households for payments, and many households gave inaccurate information, especially as structure owners sought to suppress any information on tenants. Many households pretended that there were two households, where actually only one household lived; some households were enumerated in places that subsequent verification procedures found were not occupied. Pamoja Trust followed the methods used by the Indian federation in having a strong verification process, with the information returned to households for checking. A verification process began in Huruma, with the information that had been collected and printed out being returned under the names of the people who had been enumerated, for review by individuals and community organizations. Initially, this verification process was strongly opposed by the local elected city council politicians. But the inhabitants insisted that they wanted to continue with this verification process, so the city council came as observers. Getting the city council involved proved very useful, as this began a working relationship between the council, the community and Pamoja Trust. As the verification process developed, people also came to realize the disadvantages of giving false information.

In Korogocho, the enumeration initiative had to be put on hold because of the court case – but during this lull, many savings schemes were set up, supported by the 28-person committee. Seventeen savings groups were formed and the communities became more organized. With nothing happening on the enumeration front, the association of structure owners weakened. The promises it had made to its members were not fulfilled, and the money it had taken from them never returned. By September 2002, it was possible to take the enumeration data back to the residents and start verification processes with the communities and the savings groups.
Enumerations should be seen as negotiations, as people set out their hopes (which may include false information that they feel will benefit them). With a strong verification process, people realize that it is not in their interests to cheat. In a recent exchange between community members from Huruma and Soweto–Kahawa, Huruma residents told Soweto–Kahawa members that it did urumanot pay to cheat and that you would be found out and be embarrassed. Compare this with the non-negotiating stance of any official data collection undertaken by state agencies, where, in the absence of clarifications and internal dialogue, some people get away with many houses or, if “caught”, are humiliated and punished. The community observing this process then assumes that this is the way to respond to external surveys.
After the experiences in Huruma and Korogocho, enumerations became easier, although they always take a lot of careful preparation. Now, enumerations take place to deal with specific problems. For instance, an enumeration was undertaken in a settlement called Deep Sea because its inhabitants were threatened with eviction from the private land they occupied. It was important for the residents to find out how many they were and to collect the information they needed to make a claim on the land – which they had occupied before the land had been allocated to a private landowner. The enumeration also demonstrated how many of the residents worked locally. In another settlement called Dagoretti, an enumeration was conducted in November 2000 in two settlements (COMMUNITIES?), and this was planned and executed by Muungano. Enumerations were also undertaken in Soweto Kahawa, Mitumba and Kiambiu. In Soweto, this was to identify beneficiaries of a Nairobi City Council land regularization process, while, in Mitumba, it was to support a community development programme. In Kiambiu, the enumeration was to counter an illegal land allocation process.
During 2003, an enumeration was also undertaken in Athi, one of the poorest settlements in Nairobi. The shelters are made of paper and waste materials, and are built on a site between two rivers which often floods – and there is a constant need to call in the Red Cross to help the inhabitants when this happens. It is close to a cement factory that generates a lot of air pollution. The enumeration showed that there were 500 households, more than the number of dwellings, because many people who had been evicted from the site came to register the fact that they had formerly been there. Thus the enumeration gave numbers for people as well as for shelters. Part of the enumeration had to be done at night because so many inhabitants were not there during the day. If enumerations are to collect information on everyone, especially the poorest and the most excluded, the enumerators need to visit more than once. For instance, in this settlement, there was one shack that always seemed to be empty, until a resident was found there during an enumeration visit made at night. It turned out that this was a woman with children, who was very isolated and with no ties to other residents, and who could easily have been left out of the enumeration. The community was not concerned that she be included in the enumeration, in part because she was from another tribe. In all settlements, there are groups and sub-groups and complex micro-politics that may act to exclude or hide some of the poorest households. Not all community processes are positive, and while surveys are being undertaken, mediation and negotiations are needed to ensure everyone is included – this is a vital role of the larger federation and the NGOs who assist such activities.
More than 30 enumerations have now been completed, and the information they collected has fed into a larger programme that is developing profiles for all Nariobi’s “slums”. One hundred and fifty-nine settlements have been identified, and profiles and histories are being developed for each of them. The experience with enumerations has shown the importance of having people from the community being enumerated (TRAINED?) to be the main enumerators – rather than developing a specialist team of external enumerators. How an enumeration is done and who does it is as important as the information it collects. It is a reminder to those who are planning slum upgrading of the critical choices that need to be made while designing timeframes. These procedures need to be undertaken at a pace that suits the process and not fitted into “slots” on the basis of how much time it takes to fill a questionnaire and process data. Multiple verification and consensus building is vital and, without these, no enumeration exercise will lead to real community involvement in solutions.
Enumerations are a means of organizing and federating communities. When communities undertake surveys and verifications, these acts in themselves produce federations, which, in turn, remain the basis of the sustainable involvement of communities in large- scale projects.

IV. THE DEVELOPMENT OF SAVINGS GROUPS

DAILY SAVINGS ACTIVITIES within the “slums” remain the core of Muungano, and their member groups are Pamoja Trust’s main partners. Daily savings is a mechanism, a technique or a simple system of mobilizing and organizing communities of the urban poor. It is based on the daily collection of money for savings. The people from the community who are charged with collecting the savings walk each day from door to door, collecting money for savings and loan repayments. They also collect and pass on information and, in so doing, they become better informed about what is happening. This information is used by the group to make important decisions about improving life, and thus, people are taking charge of their future. This ensures that people do not wait for the government or some outsider to come into the settlement, collect information and use it for those own purpose (and often end up using it to manipulate local residents).
Savings schemes formed by urban poor groups have long provided the “glue”, or the foundation, for the federations. It is now the organizing methodology most widely used by SDI groups worldwide – for example, in India, South Africa, Namibia, Zimbabwe and the Philippines.As a country, Kenya has had its own community savings practices, but these have had their limitations and weaknesses. Two examples can demonstrate this. One common savings practice is an agreement among a group of people to pool money, which one of them then receives. The group meets and pools money until everyone has been funded. But these schemes often suffer, as people move away before the circle is complete; and in many cases, the poorest are excluded. Another common practice involves a group organizing themselves, either through a political intervention or through the action of individuals to work together – for instance, to buy land. But at some point, the group breaks up, and often the funds disappear, benefiting a few people. In this case, no-one from the original group can question where the funds have gone or demand refunds, because there are no records. Such experiences made it very difficult initially for Pamoja Trust to promote the idea of daily savings, and communities found it difficult to accept this new technique. Another worry was that the whole process of mobilizing people to set up a savings scheme is too time consuming and cumbersome – the work required to set up a group includes developing a constitution for the group, getting it registered with the government, nominating collectors and treasurers, getting the daily collection walks organized, opening a bank account, filling in each member’s savings books and recording these on wall charts. However, exchange visits, which included federation members and staff from Pamoja Trust, showed the advantages of such savings for coping with crises, and this led to the rapid expansion and spread of these schemes. When one settlement burnt down, it was people from the savings schemes who immediately came forward with money and food. Savings groups also helped settlements develop their own governance structures. In the settlement of Deep Sea, the savings scheme members managed to get a chief arrested because he was making money by “selling” land to new people, thereby interfering with their negotiations for secure tenure of that land. Thus, savings schemes demonstrated that they were able to cope with crises and help organize groups to allow them to act collectively. In the same community (Deep Sea), the savings scheme has also allowed the residents to organize within an area that is dominated by middle- and upper-income groups, and to resist eviction (which these groups tried to legitimize as ridding the settlement of thieves).
There are now 100 savings groups in 60 settlements and around 10,000 savers. In some settlements, more than half the residents are members. The number of savings groups has grown rapidly, and the number of active groups nearly doubled during 2002. Many of the newest schemes are outside Nairobi – in Nakuru, Kisumu, Mombasa, Kitale, Meru, Thika and Kiambaa. Many of these schemes started with the hope of getting land, but other needs have since been identified. For instance, over 20 savings schemes have developed for water, and they continue to inspire other settlements to start savings. In July 2003, representatives of the savings schemes met with their parliamentary representative to familiarize him with what the schemes did and what kind of support they needed from him. At the end of the meeting, they realized that they needed more information if their member of parliament was to support them. They sought the help of Pamoja Trust to facilitate an enumeration, and this was subsequently undertaken; the verification process is currently underway. The report from this will soon be out, and the schemes will be able to present it to their member of parliament at their next meeting.
Membership of savings schemes is open to all residents. Usually, a scheme starts with a small number of members and then it expands. A big challenge is to keep the schemes as representative as possible. In some cases, savings schemes tend to be dominated either by the structure owners or by particular ethnic groups. There is also the issue of meeting administrative costs. Most savings schemes have insisted on charging new members up to 100 Kenyan shillings (around US$ 2) and this is too expensive for the poorest residents to afford.

By the end of 2002, 54 schemes had developed constitutions to govern themselves, and 43 had functioning bank accounts and regular savings collectors. Most new savings schemes were inspired by other savings schemes. Because many savings schemes develop with no formal contact with Pamoja Trust, it is difficult to estimate how many members they have.
To support the momentum of the process, Pamoja Trust undertook the following:
• They mobilized and facilitated savings from active savings schemes to conduct a savings revival campaign. This spread information about the federation and its structure, and about the need for all savings schemes to undertake daily savings schemes and to participate in exchanges.
• They accelerated exchange visits for peer mentoring, records checking, linkage and solidarity building.
• They facilitated the structuring and consolidation of the schemes to produce a federation. By mid-2004, they expect to have a Kenyan slum dwellers federation which would take the following form:
• Level 1: the savings schemes are the basic unit of the federation; when these are strong, the whole federation is strong;
• Level 2: between four and six neighbouring savings schemes link up to form a neighbourhood network.
• Level 3: each of the neighbourhood networks within a wider geographic area nominates a number of representatives to form a regional network.
• Level 4: The regional networks nominate a number of people to form the national federation.

Levels 2, 3 and 4 are to help coordinate the schemes’ activities and to provide a representative federation for all the savings schemes. The different levels also preserve the autonomy of the individual schemes within the umbrella framework of the federation.
In Kambi Moto, the number of savings groups has grown from 30 to over 120; savings groups also develop to become savings and loans groups. Enumerations encourage the spread and development of savings groups: twenty savings schemes have been developed by a number of rural squatters in Timau, and they, in turn, have stimulated savings groups in other nearby settlements.

Threats to land tenure or opportunities for regularization (and processes such as slum enumerations) tend to boost savings membership numbers. However, membership of Muungano is larger than that of its savings schemes, as many members support its land and shelter agenda but see no need to save.

Most savings schemes now have daily savings, which is convenient for the poorest members. It also provides for daily interaction between members and the management of the money collected.

To keep savings schemes equitable and just requires external pressure and regular audits. Currently, the system of weekly bank account and loan checks by members is well institutionalized in most cases. However, some savings schemes are still very top–down, and most savings schemes are reluctant to submit to external audits, although this reluctance is lessened if they are audited by another savings scheme. For instance, to verify an enumeration report, it proved possible to send in a team of enumerators from other savings schemes.
The issue of the right of urban poor groups to manage their own savings schemes versus the need for external auditing is one that is difficult to resolve. In other nations with urban poor savings groups and federations, where quite rightly there has been a stress on strengthening the autonomy of urban poor organizations, it has proved difficult to ensure that instruments are in place to make savings groups accountable to members and to the outside. This issue can be resolved where there are possibilities for external support for house construction, as it is possible to put pressure on savings groups to accept regular audits. This has been demonstrated in Kambi Moto, where construction is now underway. The savings group from Redeemed visited Kambi Moto and were told that they would have to have their accounts audited if they wanted to get support for construction. In this context, auditing is interpreted as meaning the checking of financial records. However, for the Trust it means taking stock of he total performance of the scheme and its outputs.

These issues, at their heart, are what makes savings so vital. It is not only about creating the financial capacity to repay housing loans but also about building unity and trust among people, building internal governance structures within communities, and focusing the internal capacity to hold one another accountable in a world where people, including the poor themselves, see each other as unworthy of trust. The savings process makes it possible for such a group to work together in the first place, to fight the corruption of others and to protect themselves individually against extortion. Money transactions produce many non-financial transitions that bring about change in the quality of relationships. They especially change the role that women play in informal settlements – both in their own eyes and in the eyes of their community – since they manage the savings.
The process of savings, especially when they are daily savings, brings saving volunteers in contact with each family on a daily basis. This, in turn, links and networks individuals and communities in a way that produces (builds) the capacity of communities to stay organized over the fairly extended periods of time it takes to negotiate with the State on the issue of security of land tenure and adequate housing.

V. EXCHANGES

AS IN ALL the urban federations that are represented within Slum/Shack Dwellers International, community-to-community exchanges are important for allowing savings groups and other community organizations to learn from each other. Most exchanges are local (i.e. between the inhabitants of two slum settlements in Nairobi), although inter-city and inter-national exchanges are also important. Exchanges are not only about sharing experiences but also about building solidarity between savings groups, drawing common lessons from their struggles, discussing their members’ migration histories. Through these connections, strong personal and community bonds are formed, which then sustain the federation. Communities that have visited each other also support each other during times of crisis, often with no intervention from the Trust.
During 2002, the Trust supported 32 local exchanges involving 41 communities. Many were to strengthen savings schemes. Many also involved helping with enumerations (as described above) and house modelling (as described below). During 2002, there were also 18 inter-regional exchanges – for instance, between the inhabitants of slum settlements in Nairobi and the towns of Athi River, Nakuru and Meru. Much of the growth in savings schemes in Nakuru and Meru was inspired by these exchange visits.
During 2002, Pamoja Trust and Muungano also took part in ten international exchanges. These included two exchanges to Uganda, where Shack Dwellers International had been invited by the Ugandan government to advise them on how to work with their urban poor communities, and where a memorandum of understanding was agreed between the Ugandan government, Slum/Shack Dwellers International and the UK charity Homeless International. Staff from Pamoja Trust and representatives from savings schemes from Huruma, Soweto Kahawa, Suswa, Muriambogo and KCC have visited Uganda.

Visits to Kenya by representatives from other urban poor federations, especially from India and South Africa, have also had particular importance in developing the work of Pamoja Trust and Muungano in Kenya – as described already. These visits challenge their members to do more, to try things in new ways – and the federation leaders and NGO staff who come on these international exchanges help show how this can be done. Box 3 describes an exchange visit to Zimbabwe and Namibia that took place between May 27 and June 5, 2002.

VI. HOUSE MODELLING AND STARTING HOUSE CONSTRUCTION

ONLY IN 2003 has house construction become possible in Nairobi, and construction is now underway in Huruma. What should be stressed, before describing house construction, is the process that had to precede it – the enumeration of 2,309 households in which both tenants and structure owners were included, and the measurement of the land and the agreement as to how to divide it equitably as new houses were built (a difficult process on which to reach agreement, given the number of structure owners with more than one structure and the fact that there are absentee structure owners who will not benefit, many of them politically powerful). Planning for infrastructure, development of the most appropriate design from both a need and a cost perspective, and agreement on financing and on loan conditions also had to be considered.
During 2002, house-modelling exercises were carried out in five of the Huruma settlements, and a permanent sample house built in Ghetto (which had the added advantage of boosting the number of savers). House modelling comes out of discussions within the community as to the most appropriate building design for the site – in this instance, the need was for relatively small plots so that everyone could be accommodated. Site layouts were developed that would accommodate all resident households.[7]
Financial support for house construction was available to savings groups who could collectively and individually contribute some funds, and who would agree to allow external audits of their finances. Out of five communities, only one was ready to start lending their own money and to open their books, and also to accept solutions for both landlords and tenants. Because they were able to do this, work began on building and on training the inhabitants to produce materials for construction (windows, doors) and also to do the lattice, structural beams and stairs.[MSOffice1] Work has begun on 39 houses and Nairobi City Council has agreed to release the land on which all the Huruma settlements are established. Existing shelters have to be demolished, so the inhabitants have to find alternative accommodation. There are loans to cover building costs, but this does not cover the unskilled labour input that the community organizes and that is undertaken on a rota, drawing on all the savings scheme members . Within this savings scheme, there are 84 tenants and 270 structure owners.

The construction of community toilets is also underway.Box 2 noted just how inadequate (or non-existent) provision for sanitation was. But it has proved difficult to get community contributions for toilet construction. The issue of whether the residents can provide labour for community toilets and other forms of infrastructure, or whether this is best done by using contractors, has yet to be resolved.

VII. OTHER WORK

PAMOJA TRUST IS currently up an urban poor fund called the Akiba Mashinani (grassroots savings) Trust, to provide extended help for the savings schemes and to help residents acquire land, build homes and develop livelihoods. The UK-based Ruben and Elizabeth Rausing Trust has made the first contribution to this fund. The Trust will be capitalized with finance from community savings and development agencies. The Kenya Community Development Foundation, a local funder, has agreed to double the capital. The Trust is now considering a number of applications from savings schemes.

This fund, created jointly by Pamoja Trust and Muungano, will support house construction. The fund lends to savings schemes, which then on-lend to members. Members have to pay a 10 per cent deposit, the savings scheme also contributes 10 per cent, and 80 per cent comes from the fund. The cost of a complete unit measuring around 55 square metres is around 200,000 Kenya shillings (ca. US$ 2,500). Now that house construction is underway in Huruma, there is a lot of demand from different communities for support for house construction, and this has also boosted the number of communities undergoing the planning process and the measures that precede it (enumeration and house modelling). For instance, in Dagorretti, there are more than 2,000 households planning for house construction, as enumeration and house modelling has been completed there.

It (THE FUND?) also recently began working with youth groups in many informal settlements, and these are developing into a youth federation. This will not focus on land and shelter, but mainly on the chance to play football and also on new possibilities for income generation, AIDS prevention and the development of a mentoring scheme. Savings schemes have started to help pay for football boots. Youth groups bring a lot of energy and innovation, and the scale and scope of the youth programme is likely to grow rapidly.

VIII. CONCLUSIONS

THE EVOLUTION OF Pamoja Trust and its evolving relationship with the federation reflect in many ways the emerging character of a model of urban organizing that SDI networks have begun to demonstrate across many countries in different parts of the world.

For many decades, organizations of the poor focused their energies on protesting the State’s lack of action, with a view to demanding that the State undertake a range of activities. Their relationship with the State further deteriorated when the State failed to respond in ways that suited poor groups. These were reactive responses in a world where evidence demonstrates that aspirations cannot be fulfilled by actions of the State alone. Instead, urban poor federations seek to build their capacities internally, in order to change the manner in which they negotiate with city or the State to produce solutions, both for themselves and for city and the State.
This federation model of action brings a critical light to bear on many development strategies adopted at present by both municipal authorities and governments at national level, and also on the method of developmental intervention adopted by bilateral and multilateral international agencies. While they constantly proclaim a commitment to community involvement and participation, the very tools that are used, and institutional forms that are set in place, seem more designed to respond to the needs of international agencies than to produce an engagement with the poor.

In a globalizing world where decentralization is being promoted, ultimately the relationship of the urban poor and the city is central to addressing issues of land, poverty and livelihoods. How will those marginalized but increasingly large sections of the city population who reside in “slums” develop voice and choice to fulfil their aspirations? How can NGOs who seek to assist the urban poor develop systems to create that voice and choice in ways that will survive the seemingly unending negotiations that have to be undertaken to address the issues of land, housing infrastructure and poverty in cities? The development of this experience in Kenya and Nairobi serves to highlight the capacity of SDI-affiliated federations to support and promote the creation of city and national federations, with support from each other. It also demonstrates that what took the earlier federations in India and South Africa one to two decades to achieve, can be achieved in shorter periods by more recent members such as Muungano. Box 4 provides a postscript to this paper – a reflection by one of the South Africans at the meeting in Nairobi in November 2003 for the launch of Akiba Mashinani.
[1] See Environment and Urbanization Vol 13 No 2 which had papers on the federations in India, Zimbabwe, South Africa, Cambodia and the Philippines; papers in this issue about Vietnam and Cambodia; also details of the work of Slum/Shack Dwellers International and its members on http://www.sdinet.org/
[2] See Patel, Sheela, Sundar Burra and Celine d’Cruz (2001), “Shack/Slum Dwellers International: foundations to treetops”, Environment and Urbanization Vol 13 No 1, pages 45-60; available at http://www.ingentaselect.com/09562478/v13n2/[3] Alder, Graham (1995), “Tackling poverty in Nairobi's informal settlements: developing an institutional strategy”, Environment and Urbanization Vol 7, No 2, October, pages 85–107; available at http://www.ingentaselect.com/09562478/v7n2/ [4] APHRC (2002), Population and Health Dynamics in Nairobi’s for instance Informal Settlements, African Population and Health Research Centre, Nairobi, 256 pages[5] This included members of the National Slum Dwellers Federation and Mahila Milan (cooperatives from by women slum and pavement dwellers) from India and their support NGO SPARC and members of the South African Homeless People’s Federation and the support NGO People’s Dialogue on Land and Shelter
[7] Two other house modelling exercises were carried out during 2002: the first in the KCC slum in Kariobangi to help support the development of the settlement after the Nairobi City Council announced its intention to regularize the settlement, the other at the World Urban Forum that was held in Nairobi in 2002 to sensitize both local and international groups to the community-based approach to upgrading.