4 Tackling Urban Poverty: Principles and Practice in Project and Programme Design in Kenya – Market Economy and Urban Change

Chapter 4

Tackling Urban Poverty:
Principles and Practice in Project
and Programme Design in Kenya

Carole Rakodi


Various lessons were learned from the sites-and-services and upgrading projects funded, primarily, by the World Bank and the United States Agency for International Development (USAID) during the 1970s. The response of these agencies during the 1980s, especially in the context of the shift in economic policy towards structural adjustment policies based on neo-liberal economic theories, was to emphasize the development of wider systems for the supply of infrastructure, housing finance and land in order to enable housing and land markets to work more efficiently and infrastructure provision to be placed on a financially sustainable basis.

By the late 1980s, it had become clear that, on the one hand, and especially in Africa, poor urban people had been one of the groups hardest hit by the impacts of stabilization and structural adjustment programmes, and, on the other, that the efforts to develop shelter support systems (especially housing finance) had not benefited the poorest. At the same time, recognition of the limitations of focusing almost entirely on economic policy and rolling back the state in favour of the market gave rise to increased emphasis, once again, on systems and institutions of governance, including democratization, renewed attempts to develop an efficient, effective and capable public sector, and decentralization. In addition, the increased availability of evidence that showed continued widespread and, in some cases, deeper poverty, especially from World Bank-funded Living Standards Measurement Surveys and poverty assessments, resulted in renewed attention to directly address the needs of the poor.

Spreading from the United Nations International Children's Fund (UNICEF) to many of the bilateral agencies, aid programmes were redirected in line with a renewed commitment to poverty reduction. Some of these agencies, including some that have provided little support to urban projects in the past, are reviewing the geographical distribution of their assistance and developing programmes to improve the management of urban development. Because of the concentration of poor urban households in informal settlements, poverty reduction activities in cities tend to concentrate mainly on measures to improve the living conditions of poor residents on an area basis.1 Such projects bear a close resemblance to the upgrading projects of the 1970s. It is timely, therefore, to revisit those projects in order to identify lessons relevant to contemporary attempts to reduce urban poverty.

The aim of this chapter is to briefly review some of the lessons of the upgrading and sites-and-services projects of the 1970s and recent research on the causes and characteristics of urban poverty in order to identify a number of principles that might form a basis for the design of a new generation of urban poverty reduction projects and programmes. Experience, to date, indicates that the principles and approaches that offer most promise raise a number of problematic issues, many of which pose difficulties for external funders because of the organizational characteristics of the latter. Some of these are reviewed in the final section of the paper.


Lessons from an earlier generation of low-income shelter

Many of the upgrading and sites-and-services schemes of the 1970s and 1980s depended upon external support. Indeed, the policies underlying them were often promoted by external agencies, instead of being home-grown. Evaluations were conducted by the agencies themselves, notably the World Bank's Learning by Doing (World Bank, 1983) (see also Keare and Parris, 1982). In addition, many independent evaluations have identified and explained the outcomes and impacts of these programmes (for example, van der Linden, 1986; Pugh, 1990; Rakodi, 1991). Here a brief summary of some of the findings provides lessons for new interventions.

First, few of the upgrading or sites-and-services projects were replicated on a large scale or made it possible for large numbers of urban residents (especially poor households) to obtain access to secure tenure, adequate services and improved housing.2 The reasons for this were political, organizational and conceptual:

  • Potential shortcomings, derived from weak local ownership of externally driven policy, were exacerbated by internal political factors. This led to disillusionment among politicians when upgrading of informal settlements failed to provide the dramatic ‘solution’ for which they had hoped and for which they could claim the credit. In addition, it was politically more expedient to retain the promise of future regularization as a bargaining counter in patron–client relationships with urban residents than to support widespread regularization. Moreover, political opportunities made it possible to provide access to publicly owned land by clients other than the intended occupants of serviced plot schemes.
  • The ‘project’ organization of most design, funding and implementation, often with special organizational and staffing arrangements, suited donor requirements for easy tracking of the use and impact of funds and was, in some cases, appropriate. More often, however, it generated resentment among poorly resourced departments responsible for overlapping aspects of land administration, service delivery or housing supply, and failed to build among the latter commitment to the new approaches and capacity to take on responsibility for operation, maintenance and replication.
  • The ‘project’ approach was also based on an area-based conceptualization of problems of insecure tenure, inadequate supply of land for low-income housing and deficient infrastructure. Purchase, subdivision and servicing of tracts of land were expected to provide opportunities for low-income households to access home ownership. Regularization and infrastructure installation in informal settlements was expected to stimulate continued improvement of services and investment in existing housing. In practice, beneficiaries of serviced plot schemes were not the poorest. Many were middle-income households above the eligibility ceiling, both for reasons of affordability and because of a failure to appreciate that market failures in the supply of land and housing for middle-income households would inevitably lead to downward raiding of publicly provided and often subsidized solutions aimed at the poor. In some cases, upgrading did benefit large numbers of low-income house owners; but where land supply was constrained, upgraded areas became attractive to non-poor households and investors, and rents increased, forcing out tenants. Finally, failure to tackle constraints on city-wide infrastructure supply and to strengthen the capacity of agencies responsible for ongoing operation and maintenance led to rapid deterioration of infrastructure services, compounded by a vicious circle of poor delivery and failure to recover costs.

Second, inappropriate and unsuccessful project components and failure to maintain or replicate improvements can be attributed to the approaches to design and implementation that were adopted. Project design processes which did not involve those responsible for implementation, which did not allow for intended beneficiaries to play a role in identifying needs and appropriate solutions, and which were designed and undertaken by central ministries or consultants (or, worse still, foreign consultants) produced inappropriate schemes. In addition, projects which involved the setting-up of new organizations, included capital costs but paid little or no attention to operation and maintenance, and were based on unrealistically high unit costs did not deliver sustainable and replicable improvements. Finally, external evaluations of projects often did not identify impacts perceived as critical by the poor, generated resentment among local staff and failed to maximize learning from experience.

Understanding urban poverty

A second source of inputs for the design of interventions to reduce urban poverty is the recent research that has built on an extensive reconceptualization of rural poverty to increase understanding of its nature. Much research on poverty has been based on the definition of a poverty line (PL), and this is still the basis for poverty monitoring using large-scale sample surveys. Expenditure-based PL monitoring does capture some important aspects of deprivation and provides a useful, if costly, tool to monitor the incidence of poverty. To fulfil this potential, it needs to incorporate the methodological refinements that have been developed during recent years, to be carried out periodically to give an indication of trends and to be disaggregated not only to the urban level but within urban areas. However, even with methodological refinements, sole reliance on PL measurements causes a number of problems (see, for example, Rakodi, 1995; Hanmer et al, 1997; Booth et al, 1998):

  • It is defined by outsiders and neglects people's own definition of poverty, which incorporates lack of access to assets and services, as well as ill health and insecurity, and recognizes different degrees of vulnerability related, for example, to household composition and its implications for the availability of labour (Rakodi, 1995; Wratten, 1995; Moser, 1996).
  • It does not provide a sufficient basis for understanding the causes of impoverishment (or improved well-being) for individuals, households and communities; the causes of persistent poverty and deprivation (and its reduction or deepening); the intergenerational transmission of poverty; or the immediate triggers of impoverishment.
  • It implies that the poor are passively waiting to become beneficiaries of external interventions when they, in practice, adopt positive strategies for coping with impoverishment and securing improved well-being (Rakodi, 1995; Moser, 1998).

Policy should be informed by an understanding of the ways in which households cope, adapt and manage in deteriorating economic situations, in circumstances of personal adversity and in response to opportunities to improve their well-being so that it supports rather than damages the efforts of the poor to help themselves (Amis and Rakodi, 1995; Satterthwaite, 1997).

This implies that an understanding of household livelihood strategies is necessary, although not sufficient to identify the most appropriate and effective ways of reducing poverty. Conceptualizing the resources available to households as a stock of assets, which can be stored, accumulated, exchanged or depleted and put to work to generate a flow of income or other benefits, is currently considered to be the most promising approach (Moser, 1998; Rakodi, 1999a; Rakodi with Lloyd-Jones, 2002). Social units need, it is suggested, to be able to call on stocks of all types of capital (natural, produced/physical, financial, human, social and political) (Rakodi, 2002). The capital available varies between households and over time, related to the life cycles of individual households, broader economic conditions and the urban policy regimes. Households’ strategies to manage their capital assets vary: the choices open to them may be very constrained (trapping them in vicious circles of permanent poverty or making them vulnerable to impoverishment) or relatively open (enabling trade-offs to be made between the use of, and investment in, different forms of capital to allow livelihood diversification or realization of improved well-being).

The effects of macro-economic fiscal and monetary policies are transmitted to the household level by the conduits of markets and infrastructure. Economic growth is generally associated with a reduction in poverty, transmitted in urban areas primarily through formal and informal labour markets (Figueirendo et al, 1995). Its labour market effects will, therefore, depend upon the sectoral composition of growth, its labour intensity, and the structure and ownership of enterprises. Macro-economic policies may be designed to promote expansion and to take account of their impact – for example, reducing the speed of liberalization to allow manufacturing and other sectors time to adjust (van der Hoeven, 1995). Some analysts would argue, in addition, for improved employment protection and the institution or reinstatement of minimum wages (for example, Guhan, 1994; ILO, 1995; Shaheed, 1995). However, economic growth tends to favour those above or just under the PL and not the poorest, whose limited human capital assets prevent them from taking advantage of income-earning opportunities.

Moreover, for economic growth to occur and for it to result in improved well-being, it has to be accompanied by investment in infrastructure and human capital. This, in turn, depends on, firstly, policies with respect to the reduction of fiscal deficits, which can be slowed to avoid harmful cuts in ‘soft’ sectors – for example, social services or maintenance (Stewart, 1995) – and, secondly, on the ability of public-sector agencies to capture a share of growth in revenue, which is then employed both to encourage further growth and to improve living conditions for the poor. Although it is possible that the ‘trickle-down’ effects of economic growth are the most effective means of reducing poverty, it is difficult to assess their impact because of the complexity of the effects, which are both direct and indirect and also change over time, and the heterogeneity of poor groups (Killick, 1995). Fiscal and public expenditure meso-policies may reinforce or compensate for the possible poverty-increasing effects of macro-policies through their impact on disposable incomes, their effect on the prices of goods and services consumed by the poor by means of indirect taxes and subsidies, and their influence on the availability (and price) of publicly provided goods, especially health, education and water (Stewart, 1995).

At the urban level, poverty incidence and characteristics result from the interaction between macro- and meso-processes and policies and the particular circumstances of city economies, residential areas and households. Urban agencies may be able to do little to influence the broader economic conditions that determine the level of demand for urban produced goods and labour other than to engage in macro-economic policy debates in order to ensure that the effects of economic policies on urban economies and residents, especially the poor, are recognized and that policies are designed to minimize their poverty-increasing effects and to maximize their poverty-reducing effects. In order to have an influence on economic policy-making, as well as for their own policy formulation, therefore, they need to develop an understanding of the impacts of economic trends on the incidence and characteristics of poverty in urban areas; the livelihood strategies of different groups of poor households in local economic, political and social contexts; and the way in which policies influenced or determined at the city level facilitate or constrain households’ attempts to make the best use of the assets available to them.

The scope for local policy is constrained both by the effects of national policies and by the allocation of responsibilities for social and infrastructure expenditure (see Dillinger, 1993; Smith, 1997). The potential for action depends upon the extent of local autonomy, which is limited in most countries despite recent decentralization trends, so that – although there may be scope for local decision-making on expenditure priorities and pricing/subsidy policies, the design of social policies and safety nets, and measures to support economic activities – the impact of local policies is likely to be outweighed by the effects of national policies. Nevertheless, devolution of decision-making to the local level potentially allows the initiation and design of policy packages appropriate to local conditions, the adaptation of national policies to local circumstances, and the adoption of a coordinated multi-sectoral approach to poverty reduction (see Vanderschueren et al, 1996).

This approach needs to be tailored to local circumstances. The assets upon which households draw in assembling their livelihoods can be used as a starting point for identifying potential interventions.3 Their access to these resources and ability to use them to prevent further impoverishment and to maintain or improve well-being is, in part, a product of household characteristics and strategies, but is also determined by the operation of wider systems of provision (for example, of piped water or education) or markets (for example, for labour, land and housing). The priorities of poor households will differ according to their needs and resources, the impacts upon them of economic trends, and the constraints on their ability to access services and take advantage of employment opportunities. It is not the purpose of this chapter to identify appropriate components for an urban poverty reduction programme. Decisions on priorities and the design of individual measures will be city specific in the light of local needs, allocation of responsibilities and availability of resources. However, an understanding of the livelihood strategies of urban households implies that interventions may be needed to increase their access to a range of assets and resources, including:

  • land with secure tenure;
  • physical capital, including housing and productive equipment for income-generating activities;
  • financial services;
  • labour market opportunities by
    • promoting the growth of good-quality employment and creating an enabling environment for the informal sector; promoting the growth of good-quality employment and creating an enabling environment for the informal sector; promoting the growth of good-quality employment and creating an enabling environment for the informal sector;
    • enabling people to take advantage of labour market opportunities;
  • physical infrastructure and social services, which are critical to the development of human capital resources (good health, knowledge and skills), employment generation (such as water and electricity) and enabling people to take advantage of urban opportunities (for example, public transport);
  • safety nets that are accessible to the chronic poor and to the vulnerable in times of stress or crisis, including provisions to ensure food security, crisis (consumption) credit, basic pensions and income subsidies, care for vulnerable groups such as the chronically sick and disabled, and support for children not attending school;
  • social and political capital through the development of adequate systems of representation, provision of access to legal entitlements and safeguards, improving security and providing protection against crime and violence, and supporting grassroots organization.

Responsibility for some of these will normally rest with central government, but there may be scope for local delivery agents to adapt policies to local circumstances. The responsibility for others may rest with central government, specific public agencies and local government, with precise arrangements varying between countries and cities. The allocation of responsibility for direct and indirect provider roles between central and local government, and between public sector and non-governmental entities (non-governmental organizations, or NGOs, and private enterprise), has been reconsidered in the course of economic and institutional reforms (Batley and Larbi, 2004). The organizational context is, therefore, often rather fluid. However, it is assumed that both the traditional responsibilities of local government and the new opportunities offered by democratic decentralization provide scope for local initiatives to tackle poverty and deprivation. The process of project design that will be considered in the remainder of this chapter is based on the assumption that local awareness, motivation and capacity exist.


The experience and increased understanding briefly discussed in the previous section suggest a number of principles upon which the design of city-level interventions to reduce urban poverty, including those supported by external agencies, should be based. These will be used to suggest an approach to project preparation, illustrated by recent Kenyan experience. The experience cited suggests that implementation of approaches based on these principles is feasible, although far from unproblematic (see ‘From principles to practice: Unresolved issues’).

Principles for the design of urban poverty reduction projects
and programmes

Principles include the following:

  • Political legitimacy and participation: both primary stakeholders (people, households and their organizations) and local secondary stakeholders (local authorities, local branches of national agencies, NGOs and private-sector organizations) should be involved in planning, not merely as consultees, but with an active role in decision-making. This moves beyond a conceptualization of participation as the involvement of intended beneficiaries in public-sector decision-making to a communicative approach to planning and collaborative or negotiated decision-making (Choguill, 1996; Healey, 1997; Rakodi, 1999b). The political context and system needs to give such an approach to planning legitimacy, and to support it by providing the means to resolve conflicts and to raise and allocate resources.
  • Needs and priorities of the poor: projects and programmes need to be based upon an adequate understanding of the situation and priorities of the poor, recognizing that these may vary between households, groups and communities over time. In addition, appreciation that the poor may not all be geographically concentrated in informal settlements and that not all of the residents of informal settlements are poor must inform project design.
  • Local ownership of proposals: this is vital to sustainability and replication. Priority must be given to negotiation and agreement between local actors over the technical superiority of project and programme design.
  • Partnership: despite moves towards devolution in many countries, local government often lacks policy formulation, and administrative and financial capacity. In addition, some of the activities needed to achieve developmental goals fall outside its traditional responsibilities and spheres of competence. To an increasing extent, NGOs, GROs (grassroots organizations) and the private sector have been trying to meet previously unsatisfied needs and to compensate for public-sector deficiencies. For some tasks – such as financial services for poor households; support to micro-enterprises; the supply of large quantities of low-cost housing; area-based improvements in informal settlements; and the operation of public transport and solid waste collection – one or other of these generally demonstrates greater capacity than government. However, they cannot ensure systematic coverage, redistribute resources or enforce regulations. Crude advocacy of ‘rolling back the state’, marketization and privatization are giving way to a more nuanced debate about alternative ways of dividing responsibilities for service delivery, appropriate roles for the public sector and ways of developing public-sector capacity to fulfil its agreed roles (Batley, 1997; Grindle, 1997). New governance arrangements are still evolving and the relationships between public-sector agencies and civil society organizations of various types are often characterized by mutual suspicion, conflict and resistance to change. Nevertheless, it is increasingly widely accepted that a solely public sector-driven approach to policy formulation and implementation is no longer either convincing or workable. Thus, in the context of urban poverty reduction, there is a need to involve a range of local actors in planning, implementation and service delivery, defining roles and responsibilities to make the most of their relative strengths, developing new working relationships between them and further developing the capacity of each to perform suitable tasks (Finsterbuch and Van Wicklin, 1989; Reilly, 1995; Abbott, 1996).
  • Institutionalization of planning and project design processes, service delivery and other interventions: these are also vital to sustainability and replication, implying a preference for working with existing and ongoing, rather than new, project-specific or limited-life organizations, and matching activities to their capacities, rather than to an ideal of maximum effectiveness. These capacities are a product of an organization's external context, as well as its internal resources, organizational arrangements and procedures. They are not fixed and can be developed, but in realistic incremental stages compatible with both external and internal constraints (Hilderbrand and Grindle, 1997; Throstle et al, 1997).
  • Participatory monitoring and evaluation: success should be assessed primarily in terms of local stakeholders’ perceptions of the extent to which interventions have achieved agreed objectives, and not merely in terms of externally defined criteria, implying collaborative design and implementation of outcome and impact evaluation (Narayan, 1993; Goyder et al, 1998).

Approaches to programme/project design

In order to design and implement policies to embody the above principles, alternatives to traditional approaches to planning, policy formulation and project preparation are needed. There is increasing evidence from around the world that, even if the process is initiated by public-sector agencies, it is necessary and possible to involve other stakeholders from the outset, and on a more equal basis than implied by traditional consultative modes of participation. While there has been some evaluation of collaborative approaches to planning, partnership arrangements and inter-organizational networks in Northern countries, accounts of their equivalents in the South have mostly been assembled by international agencies, which proceed to advocacy and replication on the basis of rather slender evidence on feasibility and effectiveness. Systematic evaluation of experience with approaches to project and programme planning and implementation, which attempt to comply with one or more of the above principles, is rare. Nevertheless, anecdotal evidence suggests not only that the number of attempts to realize them is growing, but also that some of the attempts are sufficiently successful to be identified as examples of good practice (UNCHS and CITYNET, 1997; Rakodi, 1999b). Here, an approach based on these principles will be outlined. This was developed during 1996 and 1997 in the course of designing the first project of a proposed urban poverty reduction programme in Kenya for the UK Department for International Development (DFID), building on some preliminary work supported by the United Nations Centre for Human Settlements (UNCHS, now UN-Habitat) urban poverty programme.

As implied above, there is a role for central government in the design of meso-policies (fiscal policies and sectoral policies related to education, health and physical infrastructure), the development of an overall policy framework, the design of regulatory systems, and the financing and provision of certain technical expertise; but the primary decision-making responsibilities for cities should be local.

A collaborative approach to poverty reduction at the city level needs to bring together local stakeholders within negotiated partnership arrangements, designed to make the most of their strengths, further develop their capacity to perform relevant tasks, and reach agreement on the most appropriate allocation of responsibility for project implementation and service delivery. The process will be influenced by the source of the initiative (external or internal, political or administrative), existing arrangements for planning, consultation and decision-making, and previous efforts at identifying needs and preparing policies, plans and projects at city and community levels.

Box 4.1 Partnership approaches to meeting the needs of
the urban poor in Kenya

The process used in preparing a proposed urban poverty project in three secondary cities and towns in Kenya started in each city or town with the following:

1   Preparation of a poverty profile by consultants, the results of which were presentedto a poverty consultation.

2   Organization of a poverty consultation by the municipal council but involving a wide range of local stakeholders, including local authority departments, councillors, local offices of central government agencies, non-governmental organizations (NGOs) working within the town, religious organizations, and community-based organizations (CBOs) where they existed. The consultation:

•   agreed an understanding of poverty in the town;

•   agreed a provisional set of possible responses to the problems identified;

•   agreed the composition of a steering committee representative of all stakeholders to coordinate the project planning process.

3   Establishment of a series of working groups, including steering committee members and others, to collect information on existing programmes and review possible solutions to the main problems identified. Since virtually none of those involved had any previous experience of project planning, the working groups were assisted by consultants tasked to support the project planning process rather than to produce proposals, as well as to carry out an institutional appraisal of the local authority and selected NGOs.

4   Initial training of staff of the local authority and NGOs in methods of participatory urban appraisal (PUA), and the carrying out of preliminary PUAs in a selection of typical informal settlements in order to improve awareness of the needs and priorities of poor residents and to inform the planning process, rather than to produce precise settlement level proposals, which were to be produced in a more collaborative and considered way with residents during project implementation.

5   Evaluation of the existing community-based healthcare system established with United Nations International Children's Fund (UNICEF) support during the 1980s, and a workshop to which results of the study were presented, in order to draw up proposals for further development of the system.

6   A logframe planning workshop, involving members of the steering committee and a range of other stakeholders, to develop the draft proposals suggested by the steering committee into a draft project proposal, including selection of settlements for the programme's launch.

7   Appraisal mission by UK Department for International Development (DFID) staff to review the draft proposals, leading to revisions and eventual approval of a project in Mombasa in 1999.

The process involved a number of departures for local actors: several local authority departments had no previous history of working together, let alone public-sector agencies with NGOs, or NGOs with each other; and few had had any experience of project planning since there had been little donor support to the towns, and previous projects had been planned by the ministry or consultants. Many wanted DFID staff or consultants to prescribe solutions. Neglected local authorities were desperately short of funds and senior staff; all public-sector employees were poorly paid and many were poorly motivated. In most places, communities were not systematically organized, and public-sector/community relationships varied in their intensity and levels of hostility. Virtually none of the staff of public agencies or NGOs were familiar with participatory methods, and none or few of the NGOs were using demand-driven methods in informal settlements.

The approach adopted, in the interest of local ownership, sustainability and potential replicability, resulted in a steep learning curve for a large number of local actors. Consolidation of new working practices and relationships, and of changed attitudes towards the urban poor and informal settlements, is threatened by resource scarcities, staff transfers and electoral change. The development of local institutional capacity to respond to urban poverty is central to the project. Inevitably, this commitment to developing local project planning skills and to the process of community organization and planning that will be needed during implementation takes time and funds before practical improvements to the lives of poor residents can be attained. In addition, the process and the proposed approach caused practical difficulties for the potential funding agency and there were considerable delays in securing (partial) project approval.

In the Kenyan context, local authorities lacked policy formulation, and administrative and financial capacity. Neither they, nor central government had explicitly considered the need for poverty reduction strategies at the urban level; and relations between local authorities and other stakeholders were weak and mutually suspicious. Earlier experiences were confined to large World Bank-funded upgrading projects implemented during the early 1980s – which had been designed by central government, experienced major implementation problems and had mixed and often adverse outcomes – and small-scale UNICEF-supported programmes during the late 1980s and early 1990s – which had avoided some of the worst features of the World Bank-funded projects but had had a limited and patchy impact. During 1996 and 1997, the initiative to commence project preparation (see Box 4.1) came from both Mombasa Municipal Council and DFID.

The more general application of the approach is now discussed, drawing on the Kenyan experience to illustrate issues and problems that would need to be addressed. The process of project preparation involves:

  • Preparing a poverty profile of the city or town, by consultants if necessary.
  • A local stakeholder consultation, bringing together a wide range of actors, including local government officials and councillors, relevant departments of central government, NGOs, religious organizations, community-based organizations (CBOs) and other agencies. Making use of the preliminary investigation of poverty reported in the poverty profile, such a consultation provides an opportunity to reach consensus on the nature and dimensions of poverty in the city or town concerned; to undertake a preliminary identification of the highest priority needs, the possible solutions and the actions required to achieve the proposed solutions; to identify responsible organizations; and to achieve agreement on a mechanism for organizing and coordinating a further process of programme development. Reaching agreement on the composition of a task force or steering committee to undertake further planning at a consultation with wide local representation is intended to increase transparency and accountability, although this can only be maintained if the steering committee periodically reports back to, and consults with, the wider group of stakeholders.
  • Establishing a local steering mechanism to ‘own’ the process of project planning and the proposal that results, and to request and coordinate external inputs. The appropriateness of including the mayor or political representatives in such a committee will depend upon the local political situation. In the absence of local experience of such an approach, this committee and the planning process are likely to need facilitation and support (Wight, 1997). The support might include both technical expertise absent in local agencies and mediation to assist and arbitrate as new ‘partnership’ relationships between public-sector organizations, NGOs and CBOs are developed. Local government may need support in developing institutional capacity; in changing emphasis from traditional controlling/providing roles to enabling/facilitating roles; and in handling tensions and conflicts both within its own ranks and between it and its political constituents. For example, appreciation may be developed that, if collaboration with other local stakeholders has the potential for increasing the extent and quality of service delivery, the legitimacy of the local authority itself may be enhanced, giving greater scope for improving its performance of other functions. Non-traditional consultants who adopt a facilitating and participatory mode of operation are needed if appropriate and locally owned proposals are to be developed, and if the skills to identify problems, analyse performance and draw up proposals are to be developed among local stakeholders (Wight, 1997).
  • Providing for poor residents to voice their experience of poverty, identify and prioritize their needs (through PUAs), devise solutions (jointly with secondary stakeholders through community action planning, or CAP), and participate in implementation, operation and maintenance, and monitoring and evaluation. There is a potential intermediary role for NGOs in facilitating this process, especially if local government is associated with unsuccessful past interventions and has to undergo a learning process on appropriate methods, if political control at the city level is not seen as being pro-poor, or if community organization is poorly developed or unrepresentative.
  • A forum for agreement to be reached on project proposals. In the Kenyan case, logframe workshops were held, at which local steering committee members and other stakeholders participated in agreeing project goal, purpose, outputs and activities, building on the deliberations at the initial consultation, preliminary PUAs and draft proposals prepared by working groups of the steering committee and NGOs. Such a workshop also provides an opportunity to consider any necessary changes to the composition and terms of reference of the steering committee once implementation commences, and to secure wide stakeholder backing for the central management mechanism. The proposals need to incorporate arrangements for implementation, including management and coordination and disbursement of funds; selection of settlements in which project activities will focus; arrangements for planning and implementation at settlement level; proposals for city-wide interventions to develop capacity, provide back-up to community-level measures and work with non-settlement specific groups, such as street children; and a menu of likely activities at settlement level derived from needs identified in the preliminary needs assessments. Whether it will be more appropriate to include political decision-makers in this forum or to institute a parallel or subsequent process will depend upon local circumstances.

This account implies that stakeholders are willing to collaborate, consensus can be reached and external consent will be forthcoming. However, none of these outcomes is straightforward and none can be assumed. In the Kenyan context, the marginalization of local government, the withdrawal of financial support for its activities by the central government, and the resulting failure in traditional municipal service delivery was, by 1996, leading councils to have somewhat more open minds than before regarding the need for working with other agencies, especially the private sector, but also NGOs. Moreover, gradual recognition that informal settlements and their residents could no longer be ignored or resettled had led to a change in attitude among many local politicians and officials, who were ready to adopt more constructive approaches to solving the problems of these areas.

Despite these shifts in attitude, resistance to change, inconsistency of attitudes and lack of capacity to operate differently hindered the development of new approaches. For example, senior municipal officials were reluctant to share decision-making responsibility with other stakeholders and to abandon municipal direct provider roles, despite their recognition that local authorities were incapable of fulfilling their responsibilities. Public commitment by local authorities to facilitate the informal-sector economic activities of the urban poor was accompanied by continued enforcement of by-laws under which informal sector entrepreneurs were cleared off city-centre streets. Staff of both local authorities and NGOs, while recognizing the need for residents to identify priority needs and preferred solutions, persisted in advancing supply-driven interventions. Thus, engineers and planners tended to propose roads and house improvements, despite evidence that neither were high priority to residents, while NGOs submitted requests for funding delivery of their traditional programmes (such as family planning and skills training) without evidence of demand. Staff members of both local authorities and NGOs, although expressing a desire to work with residents to identify their needs and priorities, lacked knowledge of participatory approaches and CAP methods.

Power struggles were manifest in the:

  • desire of local authorities to function as the funding channel despite the manifest lack of trust by all other stakeholders and many of their own staff in their ability to do so in an efficient and transparent manner;
  • reluctance of agencies with related responsibilities to work together, especially in the health and water sectors in Mombasa;
  • search for a model for community organization that would not threaten, antagonize or be de-railed by ‘the administration’ (the chiefs, or lowest level of national government, political organization and administration);
  • tendency of chief officers (appointed by central government) to keep elected councillors at arm's length from the project decision-making process, in the context of a sensitive political situation in which the large urban areas were actually or potentially under opposition control and symbolic mayors on the UK local government model were flexing their muscles and trying to adopt a stronger leadership role (Rakodi et al, 2000).

Agreement of policies and proposals may be achieved relatively easily because all of the actors involved recognize that the significant decisions are taken during the course of implementation, when finance has already been secured. Conflicts of interest and struggles for power may, therefore, remain concealed until resource allocation is under way and it may be difficult to sustain collaborative decision-making at city, project or community level throughout the implementation process. An effective project planning process attempts to reveal, mediate and resolve such conflicts early on.


At the time of writing, it was too early to evaluate the process of project implementation or its outcomes and impact in Mombasa. Ultimately, such an evaluation will (or will not) validate the principles set out above and the process of project design adopted. In the interim, however, it is useful to review some of the issues raised by the above account. Those discussed in this concluding section include both general issues, of concern to all stakeholders, and issues that have particular implications for external agencies and donors.

Politics, decision-making and the poor

Implementing interventions that are designed to reduce urban poverty implies political commitment at both city and national levels. Such political commitment has not been notable in many countries and cities in the past, and doubt has been cast on whether it is realistic to expect it. The motives of, and influences on, politicians, whether elected or not, depend primarily upon the extent to which their positions and legitimacy rely on support by the poor, compared with other urban groups. That reliance varies over time, with political circumstances, and may also vary between politicians with different party allegiances or motives for seeking political office. Where the votes of the poor count, they may be able to influence decision-making and policy agendas in their favour; but this may also be potentially divisive, rendering poor groups vulnerable to co-option. At present, we know relatively little about the political circumstances under which policy decisions and resource allocation in urban areas are pro-poor (although see Satterthwaite, 2000).

Political commitment to poverty reduction also implies willingness, on the part of politicians and officials, to listen to poor people's own analysis of their situation, their identification of priorities and their preferred solutions in order to ensure that interventions are appropriate. This implies attitudinal change, particularly with respect to gender, and a changed relationship between residents, officials and politicians. It gives professional expertise a different status and function and requires new skills on all sides. It implies a need for different channels and mechanisms for decision-making at community and city level from both traditional elected mayors or councils and hierarchical, top-down and isolationist local authority technocrats.

The need for political commitment thus raises questions about the necessity and characteristics of systems of representative and participatory democracy and about the relationship between them. The responsibilities of elected representatives at city level include promoting the economic health of the city, raising and distributing revenue between objectives and areas within the city, and responding to the demands of different constituency groups. The local accountability provided by elections of ward councillors seems to exceed that of a party list system, although the latter may make city-level decision-making easier. Furthermore, councillors can only be locally responsive and accountable if they are not overruled by an executive mayor. However, cities are large and differentiated, and even wards are mixed, so officials and councillors may find it hard to respond to the different needs and priorities of particular communities, while residents may perceive even a responsive local authority as distant and unapproachable. In addition, elections occur infrequently. In practice, even in democratic systems for urban local government, not only do parties rarely articulate coherent policy platforms, but also favouritism seems to prevail even in democratic politicial systems (Rakodi, 2004). Although representative democracy may function poorly, recent evidence indicates that residents still maintain a residual faith that, however ineffectual local government is, it holds out more hope of local responsiveness and accountability than either NGOs or the private sector (Rakodi, 1999c; Benjamin, 2000). However, participatory democracy is also likely to be needed at the local (settlement and community) level to provide an organizational framework for CAP and its implementation.

Uneven stakeholder bargaining power and negotiating

The relationships between national and local government, poor residents in cities and other stakeholders such as NGOs are characterized by uneven distribution of power. Not only is their bargaining power unequal, but their capacity to negotiate is determined by status, skills, expertise and experience, which vary. Complex power relationships are compounded by the presence of a donor and donor-recruited consultants. Control over decision-making should be exercised by a local mechanism formed by, and representing, relevant stakeholders, which takes responsibility for seeking external support and negotiating with potential donors, rather than, as so often, being donor driven. An incipient mechanism of this type has emerged in the Nairobi Informal Settlements Coordinating Committee since 1996.

It has been argued (Anzorena et al, 1998) that the most effective initiatives to reduce urban poverty will come from local groups/CBOs, working with local NGOs, with some support from external agencies. However, such a bottom-up approach depends upon the renegotiation of power relationships between CBOs and local government, difficult for donor agencies who generally work with public agencies and may be unwilling to challenge local political relationships. Where political changes have already occurred, existing procedures for consultation may provide a framework of prioritized needs and preferred solutions to which donors can respond. More commonly, however, and especially outside Latin America, this has not occurred and the capacity to undertake it prior to project approval may be lacking.

Methods have been developed to increase the negotiating capacity of the disadvantaged and to empower them to articulate their views and to influence decision-making, especially the techniques of participatory appraisal (PA) and community action planning. However, the needs and solutions identified through PA and CAP are likely to be influenced by who is facilitating the process, what questions are being addressed and what organizations are known to provide. Increased use of PA in rural areas, Warner (1997) concludes, has resulted in more aware and better-informed ‘outsiders’; but there is little evidence that it has improved the basis upon which the poor understand and interact with those outsiders. PA is, therefore, necessary but not sufficient, and projects may, in addition, need to include measures to develop the capacity of those less able to engage in negotiation (poor groups, but also some organizations and their staff), so that stakeholders can reach agreement to mutual advantage and satisfaction. In addition, non-adversarial techniques to resolving conflict and building consensus are required.

Community organization

Residential areas are often highly differentiated. They may be rich in associational life, including informal social networks and groups such as rotating savings and credit associations. They may also contain a variety of membership organizations, but there may be no overall organization that represent all interests to which agencies external to ‘the community’ can relate. Typically, residents organize on an area basis in order to solve problems of security, land and infrastructure installation, or in response to external threats. Once the problems are solved, the organization may become dormant or fail to actively involve residents. Even if resident involvement in planning and decision-making (and, perhaps, in service delivery) becomes institutionalized, a CBO is unlikely to be sustained unless leaders maintain their legitimacy and the benefits exceed the transaction costs for individuals and households. Demands on residents’ time, in particular, may be unrealistic. For example, suitable rewards will be necessary in most cultures and situations for community health workers to devote time to healthcare on an ongoing basis; and poor residents generally need to be paid for labour time devoted to implementing physical improvements.

Even if realistic approaches are devised, and support from CBO federations or NGOs is available, the pattern of resident involvement, and the viability and activity of CBOs over time, are likely to vary. In Sri Lanka, for example, the systematic and comprehensive process of CAP designed with the assistance of external consultants was adopted only in part and was adapted to reduce its demands on the time of both residents and facilitators from the National Housing Development Authority. In addition, following a peak in Community Development Council activity during the early 1990s, most either collapsed or became dormant, mainly because of the removal of political support, but also because immediate issues had been addressed and/or funds were not available to tackle outstanding problems (UNCHS, 1993; Hosaka, 1997; Fernando et al, 1999).

Donor support, project design and capacity-strengthening

The lack of structures, processes and capacity for identifying needs, developing solutions and preparing project proposals, especially in the absence of an appropriate political system and consultation arrangements, have been noted above as obstacles to implementing the principles for collaborative development and operation of projects advocated in this chapter. Identification of needs, CAP and formulation of proposals may be undertaken either in advance of the preparation of a project and its submission for funding, or as part of project preparation or implementation. In the case of the former, there is, however, a potential problem of disappointed expectations if funding is not forthcoming, from which donors are insulated, but which may cause political difficulties for the elected local government and community leaders, as well as practical difficulties for those from public agencies and NGOs who are working with communities. The latter is an alternative; yet funding agencies have difficulty in appraising and approving a project if its content is not known in advance. In addition, if capacity to undertake collaborative project design and implementation is weak, a relatively large proportion of project funding may need to be spent on capacity-strengthening activities, particularly in the early stages, rather than on delivery of practical measurable benefits to the poor. Various characteristics of collaborative approaches to project planning, therefore, cause difficulties for external agencies, which need to specify outputs, schedules and targets at project appraisal and to keep ‘administrative’ costs below strict and sometimes unrealistic ceilings. Capacity-building may be essential in order to achieve sustainability and to replicate projects; but measurable benefits count for more in project appraisal.

A staged project design may be used in which funds are committed for the whole package, but released in stages, initially for capacity-building and project design, and subsequently for further activities. Even at this stage, however, the project will not be a ‘blueprint’. City-wide sectoral programmes may be appropriate and can be designed separately or during the first phase of a project. However, it is unrealistic to expect poor groups to come up with fully formed CAPs at the outset, given their likely disillusion with the ability of government to keep its promises, transient populations and high proportions of tenants, especially where community organization is weak and fragmented. External agencies, therefore, need procedures that are sufficiently flexible to deal with evolving rather than blueprint project and programme design.

A further problem for the authorities, including donors, may arise from the trade-off between the technical quality of plan-making and inclusiveness. The technical quality of plans and programmes prepared in a collaborative way may be less than the authorities desire, especially in planning involving communities (and many NGOs). It is likely to arise from varying, and often limited, understanding and knowledge of project planning and differing perceptions of what is required among the participants, including local public-sector actors. Different perceptions of what is required may also cause difficulties between professionals and other stakeholders.

Multisectoral versus sectoral organization and delivery

Service delivery has traditionally been (and may most efficiently be) organized on a sectoral basis. The deficiencies of sectoral service delivery have stemmed, inter alia, from coordination failures and supply-driven planning that is unresponsive to user views and feedback. In order to improve on service delivery performance, better coordination, multi-sectoral priority setting and resource allocation, and greater responsiveness are needed. This is nothing new – these problems affected public-sector delivery systems long before the current reform attempts or donor-promoted focus on ‘poverty reduction’. Better-performing public-sector agencies and local authorities have both liaised with relevant external organizations and been responsive to user views and organizations. However, the recent emphasis on separating responsibility for direct from indirect provider roles to improve the efficiency of service delivery, and to make use of delivery agents other than the public sector, has given rise to an increased need for the development of effective inter- organizational networks. The jury is still out on whether current reforms will result in services that are available and affordable to the poor or more efficient and effective service delivery, and whether full or partial privatization, sub-contracting, commercialization or the development of arm's-length agencies have advantages over traditional central or local government departments (Batley and Larbi, 2004). However, it is clear that inter-organizational networks and partnership arrangements have costs as well as benefits, and that commercialization and privatization often adversely affect access to services by the poor.

In addition to the reform agenda in service delivery and renewed emphasis on decentralization, therefore, increasing attention has been paid to addressing the needs of the poor. As noted above, poor people see poverty, deprivation and well-being in multidimensional terms. Whereas a supply-driven project typically includes one or two interventions determined by the remit of an outside agency or NGO, a demand-driven approach requires flexibility that is related to project components and to which agencies should respond, as the needs of households and communities vary between areas and over time. Given multiple needs and limited resources, analytical techniques and decision-making processes are required to identify priorities and select the most appropriate interventions. Moreover, if a range of complementary interventions is to be implemented, clear links between sectoral service-delivery organizations are needed. If a project is only able to offer solutions to one or two problems following a general needs assessment, residents will become frustrated. Yet, not only do government ministries, local authority departments and NGOs tend to operate within their own sectors, but so do external agencies. Just as in governments, sectoral organizational structures within external agencies may hinder multi-sectoral responses to development problems.

Institutionalization of collaborative project planning and
implementation capacity

Project planning, decision-making, coordination of implementation, and the design and use of monitoring and evaluation should be carried out by a local coordinating mechanism rather than by a costly and unsustainable project-specific implementation unit established for donor convenience. Given the lack of experience of collaborative approaches in most cities, and the general deficiencies in policy-making capacity, coordination of implementation and accountability, the coordinating mechanism and planning and implementation processes are likely to need support. This may or may not include technical expertise. However, it is generally likely to require neutral outsiders to assist and mediate in the process of management and in the development of new working relationships between public-sector organizations, NGOs and CBOs. In the longer term, and not only for poverty reduction, strategic planning and policy-making capacity at city level needs to be developed for sustainability. Within a local authority, this is likely to be attached to the office of the chief executive or mayor, although it should involve not only sectoral departments, but also other relevant agencies.

The process of establishing and institutionalizing new mechanisms for coordination and for ensuring accountability (especially if government is distrusted by other stakeholders) is likely to be slow, as is the process of changing working practices and relationships. Progress in organizational and institutional change is incremental rather than dramatic. It is often affected by changes in the external context and unforeseen events (Throstle et al, 1997; Wight, 1997). These characteristics pose problems for donors, whose requirements are for short disbursement periods, rapid progress and clear achievements.

Funding arrangements for implementing city and community
poverty-reduction programmes

Urban services may be funded directly, through cost recovery, or indirectly, from general revenue at the national or local level. In either case, resources have to be distributed between achieving economic growth goals, meeting basic needs and increasing equity. Both user charges and allocations of general revenue should embody redistributive elements, channelling resources to poorer residents. Even in the short term, locally generated resources support most service delivery and project implementation, and in the longer term it is essential that they do. However, external funding is also likely to be needed for varying periods and on varying scales. Some donors are obliged to channel funds through central government, which guarantees loan repayment and audits expenditure. This reduces local autonomy and may cause delays.

It may be possible for projects to be prepared and implemented with external funds using established financial channels and reporting arrangements (for example, local authorities and their treasuries, semi-autonomous public-sector agencies and NGOs). However, this may be inappropriate if other stakeholders are suspicious of the fund holder, if the fund holder fails to coordinate with other bodies responsible for implementation of particular components, or if the sectoral composition of project components cannot be determined in advance. Funding arrangements must be transparent and CBOs, NGOs or public-sector agencies must be able to access funds as planning and implementation progress. At the community level, small-scale funding is required (loans or grants, as appropriate) for which decision-making on allocation is transparent, accountability is clear and enforceable and procedures are non-bureaucratic. One possible avenue for funding CBOs and local NGOs is a community development fund, several arrangements for which are possible (Anzorena et al, 1998). One model exists in the development trusts of the UK, which are supported by both government and non-government finance. In developing countries, there are recent examples of funds operated by central government (the Urban Community Development Office in Thailand) and at city level (the Sustainable Ibadan Trust Fund). A proposal for such an arrangement in Kenya, where levels of trust in local authority neutrality and accountability are low, gained substantial local backing. Experience with national-level donor-funded social development funds shows that they can provide a workable mechanism for NGOs and CBOs to access funds, but that there is a tendency for the funds to be disproportionately allocated to the best organized communities (Vivian, 1995). However, such funds are not a substitute for direct funding of public-sector agencies and the rules of some donor agencies prohibit their use.

Room for manoeuvre

A final issue is the extent to which structural constraints on reducing poverty or on changing the operation of the agencies involved can be tackled during a city-level project or programme, or have to be accepted. These constraints include national macro- and meso-policies, the legal framework for urban administration, the local government finance system, and local government powers, responsibilities and staffing. Lack of central government ability or will to tackle problems in these areas will constrain the effectiveness of interventions that are intended to reduce poverty. Donor strategies lie on a continuum between focusing on promising situations where commitment to shared goals is high and expected outcomes are dramatic; working within existing constraints because of a desire to address significant problems, even though only incremental progress can be expected; and deciding not to fund, despite the need, because it is considered that achievement of the desired results is precluded by the severity of the constraints. Unless the decision is not to fund, donors have typically dealt with constraints by imposing conditions. However, if conditions are imposed that are unwanted, they cannot be enforced and will be ignored. Ultimately, a balance has to be struck between negotiating conditions that are desirable, acceptable and likely to be fulfilled, and accepting that interventions must be realistic within the constraints.


The extent of poverty in urban areas varies with economic conditions between countries and cities over time. The current concern for the urban poor arises from their growing numbers in countries undergoing structural adjustment or in economic crisis, and from persistent problems even in countries and cities with good records of economic growth. Donor and recipient concern also extends to the potentially destabilizing effects for politics, society and economic reform policies of a disaffected urban population, and the important role of low-income people in democratization. However, even prosperous urban areas are typified by inequality, and poverty is not a new problem. Contemporary efforts to reduce it can learn from earlier experiences.

Remedial poverty alleviation measures may benefit poor people, but are unlikely to tackle the underlying causes of poverty and deprivation. Lasting solutions require appropriately designed macro- and meso-policies, changes in political and economic relationships, and new ways of working with poor people. The lessons of upgrading projects that were implemented, generally with external support, during the 1970s and 1980s, have been drawn upon in this chapter to derive a set of principles for project and programme design. In addition, recent research on urban poverty was discussed in order to identify likely interventions. This research uses a conceptual framework based on household livelihood strategies in which households manage portfolios of assets to survive, cope with impoverishment or increase their well-being in order to complement a consumption-based view of poverty in which welfare is linked primarily to labour market position. How the principles might be implemented is discussed, with reference to recent Kenyan experience. Although at the time of writing it was too early to evaluate implementation, outcomes and impact, the project planning experience demonstrates that, even in difficult and unpromising circumstances, the principles can potentially be implemented. However, progress was slow and uneven, a number of issues were unresolved, and problems occurred in the relationships between stakeholders. Central government distrust, on the one hand, and insistence that municipalities rely on locally generated revenue, on the other, resulted in difficult central–local relations and gross under-resourcing of local government in Kenya.4 Despite their lack of capacity, Kenyan local authorities were ambivalent towards potential partners, especially NGOs. Furthermore, there were a number of difficulties arising from mismatches between the mode of operation of the external agency and local needs.

For poverty reduction to occur, not only will such issues need to be resolved, but the economic and political causes of poverty, which constrain the scope for effective interventions at the city level, will also have to be addressed and further research carried out to assess the relative success of alternative components of poverty reduction strategies.


1   There are merits and demerits in the assumptions underlying such an approach and its practice, but insufficient space to discuss them here.

2   There were exceptions to this generalization, notably the Kampung Improvement Programme in Indonesia.

3   Reference is made here to households, although it is recognized that the concept of a household as a co-resident unit has problems given the mobility of household members and continued urban–rural ties, and inequalities of status, power, access to resources and scope for independent action within households. All of the suggested programme components need to be selected and designed in the light of, and to redress, gender, age, religious and ethnic inequalities.

4   Subsequently, agreement has been reached after long delays on the reintroduction of an element of central–local fiscal transfer in urban–local government finance, and implementation has commenced.


Abbott, J (1996) Sharing the City: Community Participation in Urban Management, London, Earthscan

Amis, P and Rakodi, C (eds) (1995) ‘Urban poverty: Concepts, characteristics and policies’, Habitat International, vol 19, no 4, pp 403–527

Anzorena, J, Bolnick, J, Boonyabancha, S, Cabannes, Y, Hardoy, A, Hasan, A, Levy, C, Mitlin, D, Murphy, D, Patel, S, Saborido, M, Satterthwaite, D and Stein, A (1998) ‘Reducing urban poverty: Some lessons from experience’, Environment and Urbanisation, vol 10, no 1, pp167–186

Batley, R (1997) A Research Framework for Analysing Capacity to Undertake the ‘New Roles’ of Government, Role of Government in Adjusting Economies Paper 23, Birmingham, University of Birmingham, School of Public Policy, Development Administration Groups

Batley, R and Larbi, G (2004) The Changing Role of Government: The Reform of Public Services in Developing Countries, Basingstoke, Palgrave Macmillan

Benjamin, S (2000) ‘Urban governance, economic settings and poverty in Bangalore’, Environment and Urbanisation, vol 11, no 2, pp35–56

Booth, D, Holland, J, Hentschel, J, Lanjouw, P and Herbert, A (1998) Participation and Combined Methods in African Poverty Assessment: Renewing the Agenda, London, Department for International Development, Social Development Department

Choguill, M B G (1996) ‘A ladder of community participation for underdeveloped countries’, Habitat International, vol 20, no 3, pp431–444

Dillinger, W (1993) Decentralization and its Implications for Urban Service Delivery, Washington, DC, World Bank, UMP 16

Fernando, A, Russell, S and Wilson, A with Vidler, E (1999) Urban Governance, Partnership and Poverty in Colombo, Birmingham, University of Birmingham, School of Public Policy, International Development Department

Figueirendo, J B, Lachaud, J-P and Rodgers, G (1995) ‘Poverty and labour markets in developing countries’, in Figueirendo, J B and Shaheed, Z (eds) Reducing Poverty Through Labour Market Policies, Geneva, International Institute for Labour Studies, pp11–26

Finsterbuch, K and Van Wicklin, W A (1989) ‘Beneficiary participation in development projects: Empirical tests of popular theories’, Economic Development and Cultural Change, vol 37, no 3, pp573–593

Goyder, H, Davies, R and Williamson, W (1998) Participatory Impact Assessment, London, ACTIONAID for Department for International Development

Grindle, M S (ed) (1997) Getting Good Government: Capacity Building in the Public Sectors of Developing Countries, Boston, Harvard Institute for International Development

Guhan, S (1994) ‘Social security options for developing countries’, International Labour Review, vol 133, no 1, pp35–54

Hanmer, L, Pyatt, G and White, H (1997) Poverty in Sub-Saharan Africa: What can We Learn from the World Bank's Poverty Assessments? The Hague, Institute for Social Studies

Healey, P (1997) Collaborative Planning: Shaping Places in Fragmented Societies, Basingstoke, Macmillan

Hilderbrand, M E and Grindle, M S (1997) ‘Building sustainable capacity in the public sector: what can be done?’ in Grindle, M S (ed) Getting Good Government: Capacity Building in the Public Sectors of Developing Countries, Boston, Harvard Institute for International Development, pp31–62

Hosaka, M (1997) ‘Involving communities: the CAP experience in Sri Lanka and its lessons’, in CITYNET (ed) Effective Participatory Urban Management, Yokohama, CITYNET, pp101–105

ILO (International Labour Office) (1995) ‘The framework of ILO action against poverty’ in Rodgers, G (ed) The Poverty Agenda and the ILO: Issues for Research and Action, Geneva, International Institute for Labour Studies, pp1–78

Keare, D H and Parris, S (1982) Evaluation of Shelter Programmes for the Urban Poor, Washington, DC, World Bank, Staff WP 547

Killick, T (1995) ‘Structural adjustment and poverty alleviation: An interpretative survey’, Development and Change, vol 26, no 2, pp305–331

Marr, A (1999) The Poor and their Money: What Have We Learned? London, Overseas Development Institute, Poverty Briefing 4

Moser, C O N (1996) Confronting Crisis: A Comparative Study of Household Responses to Poverty and Vulnerability in Four Poor Urban Communities, Environmentally Sustainable Development Studies and Monographs, Series No 8, Washington, DC, World Bank

Moser, C (1998) ‘The asset vulnerability framework: Reassessing urban poverty reduction strategies’, World Development, vol 26, no 1, pp1–19

Narayan, D (1993) Participatory Evaluation: Tools for Managing Change in Water and Sanitation, Technical Paper No 207, Washington, DC, World Bank

Pugh, C (1990) Housing and Urbanisation: A Study of India, New Delhi, Sage Publications

Rakodi, C (1991) ‘Developing institutional capacity to meet the housing needs of the urban poor: Experience in Kenya, Tanzania and Zambia’, Cities, vol 8, no 3, pp228–243

Rakodi, C (1995) ‘Poverty lines or household strategies? A review of conceptual issues in the study of urban poverty’, Habitat International, vol 19, no 4, pp407–426

Rakodi, C (1999a) ‘A capital assets framework for analysing household livelihood strategies – implications for policy’, Development Policy Review, vol 17, no 3, pp315–342

Rakodi, C (1999b) ‘Collaborative planning and decision making’, Paper to One World Action Seminar Influence and Access: Local Democracy and Basic Service Provision, London 25–26 May

Rakodi, C (1999c) ‘Water and health care provision: The views of users’, in Batley, R (ed) (1999) Findings in Health, Water, Agriculture and Business Development Sectors, Role of Government in Adjusting Economies Working Paper 40, pp75—96, Birmingham, University of Birmingham, School of Public Policy

Rakodi, C (2002) ‘A livelihoods approach – conceptal issues and definitions’, in Rakodi, C with Lloyd-Jones (eds) Urban Livelihoods: A People-centred Approach to Reducing Poverty, London, Earthscan, pp3–23

Rakodi, C (2004) ‘Representation and responsiveness – urban politics and the poor in ten cities in the South’, Community Development Journal, vol 39, no 3, pp252–265

Rakodi, C, Gatabaki-Kamau, R and Devas, N (2000) ‘Poverty and political conflict in Mombasa’, Environment and Urbanisation, vol 11, no 2, pp153–170

Rakodi, C with Lloyd-Jones, T (eds) (2002) Urban Livelihoods: A People-centred Approach to Reducing Poverty, London, Earthscan

Reilly, C (ed) (1995) New Paths to Democratic Development in Latin America: The rise of NGO–Municipal Collaboration, Boulder, Colorado, Lynne Rienner

Satterthwaite, D (1997) ‘Urban poverty: Reconsidering its scale and nature’, IDS Bulletin, vol 28, no 2, pp7–23

Satterthwaite, D (ed) (2000) ‘Poverty reduction and urban governance’, Environment and Urbanisation, vol 11, no 2, pp3—202

Shaheed, Z (1995) ‘Minimum wages and poverty’, in Figueirendo, J B and Shaheed, Z (eds) Reducing Poverty through Labour Market Policies, Geneva, International Institute for Labour Studies, pp111–132

Smith, B C (1997) ‘The decentralization of health care in developing countries: organizational options’, Public Administration and Development, vol 17, no 4, pp399–412

Stewart, F (1995) Adjustment and Poverty: Options and Choices, London, Routledge

Throstle, J A, Sommerfeld, J U and Simon, JL (1997) ‘Strengthening human resource capacity in developing countries Who are the actors? What are their actions?’ in Grindle, M S (ed) Getting Good Government: Capacity Building in the Public Sectors of Developing Countries, Boston, Harvard Institute for International Development, pp63–93

UNCHS (UN Centre for Human Settlements) (1993) The Urban Poor as Agents of Development: Community Action Planning in Sri Lanka, Nairobi, UNCHS

UNCHS and CITYNET (1997) Partnership for Local Action: A Sourcebook on Participatory Approaches to Shelter and Human Settlements Improvement for Local Government Officials, Nairobi, UNCHS and Yokohama, CITYNET

van der Hoeven, R (1995) ‘Structural adjustment, poverty and macro-economic policy’, in Rodgers, G and van der Hoeven, R (eds) The Poverty Agenda: Trends and Policy Issues, Geneva, International Labour Office, pp177–205

van der Linden, J (1986) The Sites and Services Approach Reviewed, Aldershot, Gower

Vanderschueren, F, Wegelin, E and Wekwete, K (1996) Policy Programme Options for Urban Poverty Reduction: A Framework for Action at the Municipal Level, UNDP/UNCHS(Habitat)/World Bank Urban Management Programme, Policy Paper 20, Washington, DC, World Bank

Vivian, J (1995) ‘How safe are “social safety nets”? Adjustment and social sector restructuring in developing countries’, European Journal of Development Research, vol 7, no 1, pp1–25

Warner, M (1997) ‘“Consensus” participation: An example for protected areas planning’, Public Administration and Development, vol 17, no 4, p413–432

Wight, A R (1997) ‘Participation, ownership, and sustainable development’, in Grindle, M S (ed) Getting Good Government: Capacity Building in the Public Sectors of Developing Countries, Boston, Harvard Institute for International Development, pp369–412

World Bank (1983) Learning by Doing: World Bank Lending for Urban Development 1972–82, Washington, DC, World Bank

Wratten, E (1995) ‘Conceptualising urban poverty’, Environment and Urbanization, vol 7, no 1, pp11–36