Posted February 1998

Food and Agriculture Organization of the United Nations

United Nations Capital Development Fund

International Fund for Agricultural Development

German Agency for Technical Cooperation

Swiss Agency for Development and Cooperation

World Bank

Rome
16-18 December 1997
Technical Consultation on Decentralization
Documentation

Decentralization, Local Capacity and Regional Rural Development: Experiences from GTZ-Supported Initiatives in Africa

Abstract

The paper presents experiences from programs in Africa, which are supported by German Development Cooperation (GTZ). The focus is on experiences at micro-level (communities) and meso-level (sub-region / district). Although a comprehensive approach towards regional rural development in the context of decentralization has not been elaborated yet, some practical experiences have been generated. Five examples of regional rural development measures in Africa are provided. Their aim is to develop workable models, in a specific region, adapted to local conditions, in line with national decentralization processes. Capacity-building measures for rural development are key areas of support. At micro-level, communities need to take the initiative and become the owners and managers of development projects. At meso-level, local government entities need to involve communities, develop regional "visions" and action plans, and effectively allocate resources. It is argued, that if locally compatible solutions are developed, then local government entities can turn into the driving force for development efforts in their region. A set of (8) criteria is suggested, against which the success of regional rural development in the context of decentralization can be assessed.

Sector investment programs (SIP) provide a particular challenge to decentralized rural development efforts. Although the SIP instrument is considered to have good potential to improve both the effectiveness of donor aid and the delivery of public sector services, there is a danger of counteracting or circumventing decentralization efforts. Sector investment programs induce a deconcentration of sectoral line ministries, but provide obstacles to a comprehensive decentralization. Cross-sectoral issues are not easily addressed. This could reduce the contributions to poverty alleviation. It is argued that linking of Sector Investment Programs with decentralization is the key to keep SIP programs manageable and achieve the desired impact of decentralization efforts.

1. Decentralization: new scope for regional rural development

Building-up the capacities of institutions at micro-level (community), meso-level (district / region / province) and macro-level (national) is crucial to seize the opportunities provided under decentralization and achieve sustainable development in rural areas. Capacity-building measures for rural development are key areas of support.

• At micro-level, communities need to take the initiative and become the owners and managers of development projects.
• At meso-level, local government entities and community organizations need to involve communities, develop regional "visions" and action plans, and effectively allocate resources. They also have the responsibility to monitor performance of ongoing projects.
• At macro-level, the political, institutional and fiscal arrangements need to provide flexibility for self-determined development in the regions. National level must also support and strengthen local governments.

Evolution of rural development approaches

Implementation was in the hands of either Government line agencies or new, autonomous implementation units, often parastatals (Regional Development Authorities). Neither community-based organizations, nor local government institutions played a role in the process. The parastatals lacked integration into existing institutional structures, depended on financial and administrative support from donors and dissolved in most cases shortly after the donor support ended. Those projects which were integrated into sectoral line agencies were not very successful and those operated by Regional Development Authorities were initially successful, but at the expense of longer-term sustainability.

Institutional arrangements, in particular the absence of local authorities with a clear regional mandate and accountability towards the communities, posed one of many difficulties faced during project implementation. The complexity of the cross-sectoral approach was another difficulty. Projects often turned out to be too complex to be manageable. The complexity of the approach further aggravated the problems with local administrative structures.

Consequently the IRD approaches came under heavy criticism in the mid 1980s. The World Bank commissioned an analysis of 184 projects (some of which were merely relabeled sector programs) and identified a number of shortcomings, which included:

• Difficult integration with existing (government) institutions
• Great complexity of multi-sector projects
• Adverse macroeconomic frame conditions
• Lack of conceptual basis for implementation
• Inflexible, rigid top-down planning methods
• Lack of compatible technical solutions for the regions concerned
• Inadequate knowledge of communities and failure to involve them

Bilateral donors, including Germany, took a less radical view and tried to reform their rural development approaches (GTZ, 1993). They did not see it justified abandoning a directly poverty-oriented, regional, supra-sectoral project approach to rural development. This was supported by further analyses, which demonstrated, that many of the crucial problems were not linked to the approach itself. For all deficiency areas, positive experience with problem-solving strategies could also be identified. This showed that difficulties could be overcome. Towards the end of the eighties, "Regional Rural Development" was further developed as a concept, which promotes the development of the particular region by giving emphasis to the spatial dimension of development and a regional framework of action. These regional rural development efforts received a major impetus when decentralization processes in various African countries gained momentum and the link to such decentralization efforts was established.

Decentralization in Africa

Zimbabwe

When Zimbabwe became independent in 1980, the country inherited a dual system of local government: one being a relatively autonomous and effective system of representative local authorities in "European" and urban council areas, and on the other hand, the weak and unrepresentative African councils. The former fell under the 'Ministry of Local Government and Town Planning' and the latter, under the Ministry of Internal Affairs in the then Tribal Trust Lands (Mtakati, Engel, 1997).

The new government aimed at removing the dual system and its first step was the transformation of African Councils into District Councils and bringing all local government under the jurisdiction of the Ministry of Local Government, Rural and Urban Development (MLGRUD). The Rural Councils (in former white large-scale commercial farming areas) and District Councils (in communal lands) - which had formerly been separated along economic and racial lines - were amalgamated into Rural District Councils (RDCs) through RDC Act No. 8 of 1988. The implementation of this act started on in July 1993 - following local elections. These elections resulted in the creation of 57 RDCs and subsequently a transfer of a wide range of powers, authorities, roles and functions from central government to the RDCs (main concept of decentralization) in the areas such as education, health, water supply and sanitation, housing etc. This meant that "Zimbabwe now had a representative local self-government structure that could respond more effectively to the country's development needs in general, and to local communities in particular" (World Bank, 1997). However, implementation of the RDC Act is still in a rudimentary and trial phase and therefore various RDCs adopt their own, "flexible" approach.

Each district consists of wards and each ward consists of villages. The number of villages and wards differ from district to district, but there are about 6 - 10 villages per ward and 15 - 40 wards per district. The actual numbers are mainly determined by the size of the area and population of a given district. The RDC consists of one elected councilor for each ward of the RDC area. Under specific circumstances, up to one one-quarter of the councilors can be appointed by the minister.

The official statements by the Central Government have always emphasized the commitment to the policy of self-reliance for local authorities, but major problems still persist. Most notably, funds are still centralized, means of generating revenue to attain self-sufficiency are limited, and ministerial control is overwhelming. During the first decade of independence, central government assumed many responsibilities of local authorities by its weighted presence in communal areas, aiming at:

• resettle approximately 165 000 families of poor, landless and displaced persons on farms acquired from white farmers;
• rapidly expand primary health care and education in all districts;
• re-direct agricultural state services towards the peasant farmer, e.g. extension, credit and marketing services;
• rapidly expand physical infrastructure particularly roads and water supplies, coupled to a policy of growth points and district service centers to strengthen the spatial structure of the communal areas.

South Africa

In South Africa, the current status of decentralization efforts and the scope of local government can only be understood against the background of the apartheid history of the country, as well as against the current tension in the country. The democratically legitimized government faces the challenge of integrating the population into a transformation process, based on the provisions of the Reconstruction and Development Program (RDP) which stipulates that the people be involved in the political, social and economic process of rebuilding South African society. Linked to this is the call to bridge the gap between the areas neglected under the apartheid regime - especially the "homelands" which were of little importance to the South-African economy for recruiting manpower - and the formerly privileged areas populated by whites.

Poverty and deprivation are widespread in rural areas, in particular in the former homelands. Unemployed rates are extremely high, in some communities reaching 60 and 100%. Some, mostly men and male youths from the age of 15 upwards, find work as unskilled laborers in the mines around Johannesburg. A high percentage of the male population do not live in their native communities, due to former apartheid structures which allocated the most infertile regions to the majority of the people, who was consistently deprived of education. Thus, a pool of unskilled workers with very limited employment prospects was provided for the South African economy. Therefore, the cash income flowing into communities is mainly income transferred by migratory workers and pension payments to individual inhabitants.

Decentralization is seen as a key policy tool in the reform process. The South African constitution foresees a highly decentralized, federal system with highly independent lower levels of Government. Legislation under the old regime was passed in 1982 to provide for elected councilors in Black Local Authority areas (GTZ, 1995). Although this was the first evidence of the South African Government acknowledging the permanency of blacks outside the homelands, these councils were vehemently opposed by ANC aligned forces. Councilors were intimidated into resigning, rent boycotts were organized, service charges on municipal services were boycotted, and a campaign of ungovernability was pursued.

The new Government introduced Transitional Local Councils (TLC's) that comprise in each town the previously separated white and black local authorities. The TLC's were to be constituted with 50% of its representatives coming from the so-called statutory side (existing councils) and 50% coming from the non-statutory side (Civics, ANC, SACP & PAC). The TLC's are functional since local authority elections in 1995. At regional level district councils were established (previously known as Regional Services Councils). The chairmen are appointed and local authorities elect the remainder of the Council. District Councils originally were formed to assist with funding for infrastructure development in previously neglected local authority areas.

There is some conflict and tension in the decentralization process between the governmental initiatives on the one hand and civil society's interventions on the other hand, e.g. in control over training of councilors and officials at local government level. This struggle for control, is, however, also part of the democratization process of the country. On the one side the newly elected ANC as the dominant political party is trying to establish as much control as possible, whilst the various NGO's and other actors from civil society are trying to ensure the establishment of sufficient bridgeheads to prevent state domination. Another important issue in local government decentralization is the intervention at rural level. A potential conflict arises between modernization and traditionalism, as the tribal chiefs are not as influential with the structure as they expected to be.

Inadequate community oriented planning is a key constraint among local government entities. This includes inadequate planning procedures and instruments as well as technical capacity for development planning. Participatory planning processes are often long winding, ineffective and do not lead to consensus building in the communities. There is no vertically and horizontally integrated planning system. Planning responsibilities of communities, regions, provinces are not known and local demands are not met by financial plans at higher level. Sectoral plans in a given region are not co-ordinated. Community involvement often has not consequences for the fund allocation or service provision. Economic promotion strategies in the communities hardly exist.

Mozambique

Since the end of the war, the Mozambican Government aims at an economic and infrastructural reconstruction and some poverty alleviation measures in the country. Officially, political and administrative decentralization is given high emphasis within this strategy. In 1994 efforts were made to create the legal and substantial prerequisites for an administrative reform that could tackle the mentioned challenges. It basically emerged from within the FRELIMO government, but found fairly wide initial support. It contained the following reform aspects (Weimer, Fandrych, 1997):

• the administrative division into 393 rural municipalities (povoações, corresponding to the former postos administrativos) and 151 urban municipalities (municípios, corresponding to the cities, towns and district capitals);
• directly and secretly elected municipal assemblies (assembleias municipais) and mayors (or administrators in the case of rural municipalities) as local representative bodies, and administrative councils (conselhos municipais) as executive organs;
• a clearly defined catalogue of duties and functions of the bodies of local government, including genuinely local tasks, such as public security, water supply as well as culture, land use, developmental measures and protection of the environment;
• a clearly defined distribution of competences and prerogatives between central state administration and municipality;
• autonomy with regard to budget and property, as well as fiscal, planning and organizational autonomy, but plans and budget have to be approved be central government institutions;
• budget subsidies (block grants, tax transfers) from the central state budget;
• the integration of traditional authorities ("autoridades tradicionais") in the local consultation and political decision-making process, particularly pertaining to questions of conflict management and land issues;
• the right of collective advocacy of interests of local governments, in the form of an association of municipalities;
• the "gradual" establishment of local governments according to infrastructure, socio-economic and administrative "minimum conditions" of the municipalities/districts;
• legal and financial care of the municipalities by the Ministries for State Administration and Planning and Finance.

Only towards the end of 1997 steps were undertaken to give more planning responsibilities to the district administrations in order to define their priorities to be considered in the sector plans and the investment plans. A key institution at provincial level is the Provincial Planning Unit, which resulted from a recent merging of the provincial directorates for Planning and Finances. They are to secure the link to district administrations and play a coordination role for the National Reconstruction Program.

Zambia

Since 1993 the government of Zambia has been making efforts to introduce a reform program for the civil service. Important elements of this program are a reduction in staff numbers in conjunction with an improvement in efficiency. This is to be achieved by:

• structural reform of the civil service,
• a "Human Resource and Performance" management system,
• strengthening of the local administration

The Council is the actual decision-making and resolving body in the district, which is made up of elected representatives of the population of the district (Ward Councilors). It monitors the relevant district administration, headed by a Council Secretary (CS) or Deputy Council Secretary (DCS), as central planning and administration authority. The councils also coordinate the sector departments.

The structure of the relevant district administration is composed of several sectoral and functionally classified sections, each headed by a Section Leader (Administration, Treasury, Works, Planning). The Council Secretary heads the district administration. The district administrations typically have about 150-200 staff. The level of staff qualification is low. Only the Council Secretary and the Section Leader have sector-specific qualifications. The remaining staff is auxiliary staff. Some councils have established District Planner posts. The planning tasks currently arising can only be carried out to a very limited extent by the staff available, since these have neither the necessary technical background nor sufficient practice-related experience in the field of development planning.

The implementation of the reforms necessitates continuous support by means of training, advisory services and strengthening of independent responsibility (ownership). This includes the necessary transfer of expertise in planning, coordination and management to local decision-makers.

Inadequate planning and coordination skills and lack of administrative competence at district level are core problems of the Zambian decentralization strategy. Example: a council in a remote province about 10 hours drive from Lusaka, recently initiated a "consideration fee" of 600.000 Kwacha (approx. 600USD) for businessmen from outside the province who come with investment proposals for small businesses.

Several sector investment programs have been initiated in Zambia. However, the link between these and the decentralization process have not been taken into consideration be policy makers. Local government has not been involved in SIP planning and sectoral plans generally lack the linkages to district plans. In some cases, parallel structures were created, and undermining of local government responsibilities occurred.

Ghana

Ghana is one of the countries in Africa pursuing decentralization with most determination. Prior to decentralization, Ghana underwent economic reform and structural adjustment programs. They started in 1983 and 1987 respectively and aimed at fighting inflation, restructuring national government and the productive apparatus and promoting agriculture. The decentralization of legislative and executive bodies initiated by the Decentralization Law (PNDC law 207) passed in 1988. It transfers public-sector tasks from national and regional level to districts and municipalities, aiming to ease the pressure on the central government, increase productive efficiency and increase the share of costs assumed by the beneficiaries. In the same year, 110 districts were created in Ghana's 10 regions and the members of the relevant district assemblies elected for a four-year term. The district administrations and, to the extent available, 22 sector departments, were put under the district assembly (GTZ, Fastenau, 1995).

According to the 1993 constitution, district assemblies are the second-highest legislative organs in Ghana. The legal basis is laid down in the 1993 Local Government Act and relevant implementing provisions. The district administrations and the subordinate sector departments are the district assembly's executive organs and are therefore also responsible for the technical implementation of the measures. The district assemblies are responsible for governing and developing the district. Two third of the members are directly elected; one third are nominated by the President on the basis of their expertise. Staffing of the district administrations varies from district to district, however, the multitude of new tasks poses problems. Rural districts usually have 3-5 qualified staff members in the district administration. In many districts the positions of planning and budget officer are vacant. The number of staff in sector departments in the districts varies greatly; in general, there is a shortage of field staff.

The constitution stipulates that at least 5% of national tax revenue (or Common Fund) be redistributed to the districts. Relevant implementing laws were passed in 1994, adequately specifying the framework conditions for decentralized development processes. However, actors in the districts lack the required know-how and expertise; this hinders decentralization processes being implemented consistently in most of the 110 districts.

The Ministry of Local Government and Rural Development (MLGRD) is responsible for the promotion of the decentralization process. MLGRD also assumes the responsibility for rural development. MLGRD's tasks focus on adapting relevant laws and implementing regulations, on training and upgrading personnel at district level and on strengthening capacities required to this end. The MLGRD has supervisory control of the district administrations, however has no authority over the district assemblies, provided they keep within the legal framework. MLGRD had a staff of about 100, of which one quarter are university graduates. The planning unit, with a staff of about a dozen experts (all university graduates), provides support and advisory services for the decentralization process. The Chief Director is responsible for coordinating the GTZ-supported Program for Rural Action and is in charge of project coordinators in the regions targeted by the Project.

Major shortcomings in the decentralization process are that villagers' thinking and behavior are still strongly influenced by an attitude that the Government is responsible for development, as was the case in the past. Local self-help capacities are not sufficiently mobilized and individual initiative at district level is not used to its potential. In addition, members of the assembly are not fully aware of their role as representatives of their constituency or of their decision-making powers as members of the legislative body. Officials of the district administration often still consider themselves to represent the central government and make decisions without the district assembly's mandate. Staff in sector departments continues to follow standards set by their ministries; measures are implemented without the approval of the district assembly. No comprehensive district development plans exist.

Towards a new regional rural development in the context of decentralization

• Local government institutions are in place with a "downstream accountability"
• Communities have better opportunities to participate in the design, implementation, and monitoring of projects, which are suited to their needs.

Strategy guidelines were developed which are critically dependent on decentralized arrangements:

Adaptation to the location: Every regional rural development program needs to be tailored to the unique problems of "its" region and "its" population. Such an adaptation of innovations and support services to the particular region, however, is severely limited under centralized arrangements, especially if they operate on the basis of national sector plans and implement "downwards" through sectoral line agencies. Adaptation to local conditions is best achieved when a project, which is identified/owned by the communities is linked with a regional development strategy (of the local government entity) and technical skills and services of the (non)governmental service structure.

Multi-sectoral commitment: Every regional rural development program is planned as a "demand driven" set of support measures, which will allow a wider range of development constraints to be overcome. Any "sector" of regional development can be used as a starting point within the community. The resulting complexity of support measures and subsequent management and steering problems can only be overcome if,

1. a local Government has co-ordination powers in the given region,
2. sectoral line agencies, as well as private sector agencies can provide technical and implementation support and
3. communities are the driving forces for the identification and implementation of projects.

Step-by-step progress: The outcome of complex development processes can not be anticipated with certainty. Errors are human, especially, when future human action has to be predicted. Planning participation-oriented development is, therefore, not a drawing-board procedure. Activity planning must be constantly adaptable as an ongoing process during program implementation. The pace of development is determined, amongst other things, by the practical and technical capabilities of people and by the speed with which consensus is achieved.

Field of action is the region: The impact area of an RRD program is the natural and economic environment of the rural population. It is often identical with the government administrative unit - the province, district, etc. In order to achieve maximum effectiveness while keeping inputs to a minimum the development activities have to be organized so that they are geared to the economic and social relationship network, which is generally tightly woven in the region. Within the framework of national policies, regional authorities must be given scope for independent activities in which the rural target groups actively participate.

Self-help orientation: Active participation by the local population is the basis for "regional rural development" These people must themselves identify, plan and implement RRD measures. Self-help strengthens the feeling of self-responsibility. RRD programs encourage self-help; they guide and support it. Successful self-help stimulates creativity and promotes self-awareness. It creates a conducive environment for private initiative and becomes the basis for an independent development of the region.

Institutional capacities need to be developed to fully utilize the potential. The main addressees for capacity building measures are local government institutions at (sub)regional level and the community organizations. For a decentralized system to function, the procedures and instruments for strategic planning at regional level need to be in place and participatory approaches have to be organized within the communities. Communities generally have a more specific, localized understanding of constraints and potentials. They need to get together, analyze their problems and potential, set priorities, plan and implement self-help projects and reap the benefits from these projects. Through decentralization, local government is put into a position to address those needs and aspirations.

Although a comprehensive approach towards regional rural development in the context of decentralization has not been elaborated yet, some practical experiences have been generated. The following chapter provides some examples of regional rural development measures, which are based on the guidelines given above. Their aim is to develop workable models, in a specific region, adapted to local conditions, in line with national decentralization process.

Examples of capacity building for rural development

Zimbabwe

At community level, a community-level planning and development approach and the processes of village development are supported and extended to other communities. The previously supported sectoral measures (agriculture, resource management, infrastructure, and income generating measures) were gradually integrated into this process.

At provincial and district level measures on the reorientation of institutions and on the building of relevant capacities were continued to an increased extent. The newly formed Rural District Councils (RDC) have been given particular priority. In this connection, the structure of the RDC is to be strengthened through transference and self-responsible administration of funding in the form of a financial contribution for the partial covering of costs in the partner country, training, etc., with advisory support being given by the seconded long-term experts. The provincial level will in future have to take on an increased advisory function in respect of the districts.

At national level it is a matter of creating improved framework conditions (an "enabling environment"), in order to support the lower levels in the performance of the changed function. At the same time the support of the Ministry of Local Government, Rural and Urban Development (MLGRUD) was to be given much attention within the framework of an inter-ministerial "Core Group" for rural development. In this way a better adaptation of institutional and financial framework conditions to support decentralized planning, decision-making and implementation was to be achieved.

The measures are closely connected and serve to give institutional and planning support to the decentralization process at provincial and district levels and the building of corresponding competence, particularly in the RDCs. At the same time, the organization of a dialogue process between the different levels and the various national decision-makers is important, in order that the ideas of decentralized development are gradually and independently realized.

At the beginning of the current phase the political responsibility of the project was to pass from the Ministry of Agriculture to the Ministry of Local Government, Rural and Urban Development (MLGRUD). MLGRUD had been assigned the overall control of rural development, and is the organization under which the Rural District Councils (RDC) fall. Since the end of 1994 this Ministry has also taken on the responsibility for the implementation of the decentralization policy, and particularly for the corresponding capacity building in the RDCs.

However the transfer of political responsibilities for the project did not materialize and the final form of institutional arrangements for interdisciplinary and decentralized rural development has not been decided. The current portfolio discussion of the ministries, the creation of new ministries of Rural Resources and Water Development and subsequent restructuring posed a major difficulty for the support of decentralization processes. Thus, support to national level was ineffective.

Support to rural communities to plan implement and evaluate projects on their own was quite successful. This required mobilizing communities and training in problem identification project planning, implementation and evaluation. In addition, staff of the participating institutions and representatives of the target groups were trained and supply of services was organized.

Major activities included:

• the qualification of Rural District Councils (RDCs) to coordinate measures on district development,
• agreement on planning and coordination functions as well as responsibilities of district and provincial levels,
• strengthening of planning competence at district and provincial level and grassroots-orientated planning processes,
• support to the drawing up of plans, strategies and draft policies at national level with a view to realizing a decentralized rural development, and
• effective planning, steering, coordination and administration of the program at provincial level

A promising development is the increasing level of co-operation between RDCs and private investors. In another district in Southern Zimbabwe, bordering South Africa, the RDC could secure a joint venture with big private company (Caltex Oil company). The venture will involve building a Z$50 million "ultra-city" to accommodate a bus terminus, a number of shops, restaurant, service/fuel station and a truck inn. This means creation of employment and higher tax revenue for the RDC. Decentralization support of this type can only work if restrictive controls and administrative procedures by central institutions do not circumvent them and if the Rural District Councils" can tap adequate implementation capacities and funds for the implementation of projects.

South Africa

The Project operates at different levels: At community level capacity building is geared towards traditional leaders, community groups and one or more NGOs. This level is reached indirectly through various actors at municipality level, esp. facilitators. Measures are geared towards formulation of needs, prioritizing, formulation of (project) requests, contributions to implementation and promoting the readiness to pay for services.

At municipality level the main actors are rural councils, Transitional Representative Councils (TRC) and the urban councils, Transitional Local Councils (TLC), the TRC subcommittees, the Administration, and District Development Forum (joint TRC/TLC plus civil society groups, but not line agencies). At this level it is anticipated that through capacity-building:

• participatory planning methods are applied,
• a "lead function" for development initiatives will be assumed,
• a "response attitude" towards community requests is initiated,
• an integrated development plan for municipality elaborated,
• effective service delivery (incl. administration) is ensured,
• funds are generated,
• monitoring of service delivery is carried out.

1. Municipalities draft integrated development plans and ensure effective service delivery
2. Municipalities in have a basic municipal financial system established (funds acquisition, revenue generation and financial management).

At provincial level the Office of the Premier (executing institution), civil society groups (Eastern Cape LG Association, ANC, COSATU) and sector departments are key players. At this level support is geared towards: the provision of clear directives to TRC/TLC in the planning process, coordination of various provincial-level plans, articulation of issues by civil society groups; clear administrative guidelines for municipalities; systematic involvement/selection of NGOs and consultants in planning and implementation. It is intended that the Office of Premier in Eastern Cape Province establishes practical examples (Prototype Projects) for effective use of public funds by TRCs/TLCs, involving sector departments, NGOs, CBOs, or consultants.

Indicators have been developed concerning the capabilities of municipal service providers, the government administration and non-governmental implementing organizations in participatory development planning. Indicators formulations include: a) the number of communities which have elaborated a hierarchy of needs using participatory planning, b) the number of (micro)-projects which have been successfully implemented by the Eastern Cape government in terms of sustainability and the extent to which the community can identify with them.

The Project provides a suitable basis to help eliminate apartheid structures and, with the assistance of governmental and non-governmental institutions, to help improve and institutionalize the participation of the non-white (rural) population in the development and transformation processes. The methodical approach is geared towards incorporating rural communities into development processes and the development dynamics of political reform in South Africa. To this end, the existing administrative and organizational structure of regional and local governmental- and non-governmental implementing institutions are to be provided with suitable methods and instruments required to implement the RDP. In addition, the existing social and physical municipal infrastructure is to be used to build local capacities for participation and communal planning.

The Project is essentially geared towards tapping local and regional potentials for involving the population in development planning at local level. Above all, it is aimed to generate and package political and social self-help capacities and initiate and network appropriate activities to mobilize these potentials.

Capacity building must ensure that:

• Implementing institutions in the Eastern Cape province are familiar with strategies and methods for self-help promotion, as well as needs-oriented and gender-specific planning.
• Communities apply participate planning methods for developing local-level structures under the guidance of facilitators.
• Selection procedures will be drawn up for assistance measures at local-authority and local-infrastructure level.
• Participatory planning methods will be incorporated as a regional development-planning instrument.
• The role which regional and local NGOs and CBOs play in the planning process, as facilitators between communities and responsible government authorities, must been recognized and institutionalized.

Mozambique

Districts are supported in public planning processes and the elaboration of an adequate planning methodology. Emphasis is given on district planning, because it is directly related to decentralization of state procedures and to possible introduction of political entities at district level at a later stage. It is further linked to elaboration of the public sector investment program (PTIP), co-ordination between sectors, and data banks. Other support includes the establishment of data banks in the provinces, assistance in the reorganization of the directorates, improvement of co-ordination with district sectors and administrations, formulation of provincial investment plans in the province. The project is working together with the Ministry of Planning and Finance, which recently made some important steps and introduced district planning as a public planning instrument. Guidelines for the planning methodology were published and all Provincial Planning Commissions were directed to assist district administrations to elaborate their plans. One important element of a district plan is the involvement of local communities or their representatives in the planning and decision process. A district plan is the only planning instrument in Mozambique, which foresees community participation; thus it contributes to the democratization process.

Some of the characteristics of district planning in Manica Province are:

• Action orientated, prioritisation of intervention areas and preparation of concrete project proposals, following a cross-sectoral planning approach.
• Planning is based on available data and studies and plan elaboration should be done within 4-6 month. The district administration is responsible.
• In the medium term, the costs of the plan elaboration should be fully supported by provincial and district institutions, therefore an adequate planning methodology should be developed which considers the costs of the planning process and the depth of analysis, data collection etc.
• Involvement of NGOs, and other active organizations in the planning and implementation process.
• A training program supports the introduction of district planning. Training includes: district planning methodology, moderation of public planning meeting with the communities, methods of problem and objective analyze, project formulation, and preparing of project proposals.

1. Public meeting on plan and the methodology/procedure of plan elaboration
2. Data Collection
3. Identification and collection of data
4. Identification of problems and potentials
5. Identification of the needs and priorities through public planning meeting
6. Syntheses of problems and potentials
7. Objectives strategy
8. Elaboration of development objectives and strategy
9. Decision-making
10. Public planning meeting to discuss proposed district plan
11. Approval of the plan through district and provincial government
12. Plan implementation

Zambia

Support activities aim at two levels. At the "micro-level" it is hoped that eventually communities in the districts analyze their potential and constraints, set priorities, plan and implement self-help projects and reap the benefits resulting from these projects. At the "meso-level" it is expected that Local Authorities, GO's, NGO's and private sector organizations deliver services to the communities in an effective, coordinated and demand-oriented way.

Capacity building measures for district administrations aim at:

• Effective data management, planning and co-ordination of all development activities of Local Government Organizations, Government Organization (e.g. Departments of Agriculture, Health, Education) NGOs and Private Sector Organizations.
• Empowering the communities in the respective districts to effectively participate in the process of social and economic decision making by initiating and co-ordinating a participatory community development approach involving all organizations in the respective districts.
• Delivering improved basic social services to the majority of the population with special emphasis on the rural population, on women and on poverty reduction.
• DDP has District Project Teams in all 5 districts. Members of these teams are the Director of Planning/District Planning Officer, the Director of Engineering/ Director of Works, 2 or 3 other officers of the District Administration, and one GTZ advisor. Each team has its own offices and facilities and is responsible for planning and implementing DDP activities in the respective districts.

Ultimately, the elaboration of a viable district development plan stands at the center of attention. The plan will also been an important instrument for coordinating developmental activities in the district. Provision of advisory services, improvement in the general physical capabilities of the district council administration, on-the-job training as well as carefully selected training measures are all designed to improve the planning and management competence of key district council staff (e.g. the district planner, director of works, treasurer) as well as key staff of the thrust sectors (e.g. agriculture, health, water, and roads). It is foreseen that in addition to the selected districts, participants from all other districts can also take part in certain selected training measures in order to ensure the spread of the concept throughout all districts in the Southern province.

Community level planning and management will be improved through a Participatory and Integrated Development Approach (PIDA). This is expected to initiate and sustain a developmental process at the community level based on a participatory approach. The elected councilors and traditional leaders play an important role in the process and will therefore benefit from specific training measures to be carried out by the project. Specific sectoral improvements are designed to improve the socio-economic base of the communities. The measures should enhance and strengthen the communities self-help capabilities and should be viewed as a form of joint venture between the sector departments and the communities. The gradual development of community action plans will eventually culminate in the elaboration of a district development plan. However, a developmental vision for the district needs to be elaborated in advance in order to guide the communities in the selection and implementation of developmental activities.

The process involves three phases, the initiating phase, the PRA phase, the follow-up phase (Schubert, 1994):

1. Initiating phase

• Application for the process
• Identification of organizing agency

• First meeting in the community
• Selection of development agencies, forming a PRA team and logistical arrangements
• PRA training
• Second meeting in the community
• Planning the village workshop
• Conducting the village workshop
• Report writing
• Evaluation of the PRA process

• Draft project proposal (based on the Community Action Plan and the village workshop report)
• Discussion within the community and with other parses involved on the project proposal
• Project application
• Channeling information for soliciting support and funding
• Initiating other development activities, identified through the PRA village workshop within the community
• Operational planning of the approved project
• Project implementation
• Monitoring and evaluation (M&E)

Ghana

PRA evolved in the mid 80's from a large-scale agricultural development program. Initially, the project concept embraced a two-pronged approach: regional planning support and co-ordination of German-supported projects in the Northern Region and promotion of regional rural development in 2 districts. Partner institution was a semi-autonomous regional development body under the Ministry of Finance. In 1992 a major re-orientation of the project concept and shift towards the newly created decentralized district structures was undertaken. This change in emphasis entailed the dissolution of the project field workers, steering committees and other traditional project organizational structures and the local government system became the focal point of the support provided by the project. The project executing organization changed to the Ministry of Local Government and Rural Development (MLGRD). The district assemblies in the chosen districts are responsible for implementing the Project.

Roles and responsibilities of the assembly, administration and departments were jointly reviewed and tasks from the Program for Rural Action's past assistance were re-assigned in line with the legal provisions of the Local Government law. Enthusiastic support for this process was given by the assembly persons (i.e. the legislators) and the district administration. On the other hand, the communities initially viewed the process quite skeptically, preferring the former approach where the project initiated and implemented many of the developmental measures (i.e. "receiver mentality"). This attitude changed quite quickly and a considerable number of communities were involved in the first community based planning cycle. This exercise was spearheaded by the district assembly persons. While the enthusiasm of the communities grew, that of the district administration and the sectoral line departments waned, mainly due to the increasing demands being placed upon them by the communities. Once it became apparent that the project would not in any way substitute and take over responsibility for tasks which were to be performed either by the communities or the district administration and sectoral line departments, a six month period of intransigence was observed. This status was overcome when all partners were finally convinced that a participatory approach could prove fruitful given the newly defined local government frame conditions.

In order to support the integration of the communities, the district assembly organizations and the administration in the planning and implementation process, training programs were designed and implemented jointly with the Program for Rural Action. Gradually a change in attitude took place, which was further strengthened once the benefits of the approach became apparent to all partners in the process. An iterative self-improving process was set in motion where the focal point was the local government system and the communities in the two selected districts. Growing recognition of the national importance of the program led to a further re-orientation of the program. Since the second implementation phase, the project support included the national level. Results and recommendations of the project's work in selected districts are fed back into the Ministry of Local Government and Rural Development where the project also provides advisory services. Thus a better vertical linkage between field experiences and national policy development could be supported

Contrary to the common belief that technical solutions and inadequate financial resources are key bottlenecks to rural development, the program's analysis shows that fundamental causes are rather low self-help capacity, insufficient collaboration between organs of local government, municipalities, national government and services from the public and private sectors. Support in the fields of strategic planning, community organization, economic development, administrative and financial management are some of the key elements of the assistance being provided. The experiences made are to be documented for wider circulation and replication in other districts. The results expected are that the:

• capacity of the selected districts for participatory medium and annual planning are strengthened;
• financial, organizational and managerial capacities of the institutions in the district are improved;
• decentralized institutions, including the district assemblies, are better able to promote community based local level development;
• local governments are enabled to identify issues at stake in economic development and to facilitate a conducive environment and to pursue qualified promotion for economic development;
• successfully tested support is replicated in other districts in Ghana by the Ministry of Local Government and Rural Development.

• Training and upgrading activities that familiarize members of the district assembly with their role and train them in problem analysis project evaluation, rational decision-making and supervision of executive bodies.
• Upgrading activities to familiarize staff from administrative bodies and sector departments with their role in decentralization and enable them to apply participatory methods of situation analysis and planning, target group differentiation and communication.
• Support and advisory services on drawing up participatory district development plans aim to deduce mid-term development perspectives to aid decision-making.

• to increase target group's self-help capacity, thereby lowering their passive dependence on external assistance;
• to actively involve village communities in "bottom-up" annual and mid-term development planning in the district, and
• improve the implementation and upkeep of self-initiated and district-assisted measures.

At national level MLGRD is being assisted to implement training and upgrading activities for several districts. Decentralization measures are to be implemented swiftly and at low cost by evolving standardized training programs from training modules conducted by the Program and other donors. An important measure in this context is support to the planned establishment of the Institute of Local Government Studies as the ministry's training institution.

PRA has followed a specific approach to the regional rural development (RRD) concept that takes into account the special circumstances of a country in the process of decentralization. It is applicable only when certain features of political, fiscal and institutional aspects of decentralization are applied by national decision makers and allow for a certain leverage of local decision makers. However, once such basic framework conditions are in place or at least point in the "right" direction, the PRA approach of "decentralization support" has excellent prospects for a substantial and sustainable development impact (Bartels/Engel 1997). This points to the need for a competent and thorough assessment of political, fiscal and institutional arrangements of the respective decentralization process which is to be supported. The framework for such an assessment still needs to be developed. It would guide in the formulation of country assistance strategies, appraisal missions and project design.

Broad-based rural development efforts which aim at responding to the needs expressed by communities can not narrowly pursue pre-conceived ideas of development (e.g. along sectoral lines). Therefore, PRA found it important that support measures do not impose strict sectoral limitations to the project proposals. (Micro)-projects are identified in communities not by sector departments. Communities may submit a whole range of project ideas, for (partial) support from district assemblies development funds. The entire "spectrum of demand" is considered for support. This should not preclude local government decision-makers to take into account meaningful selection criteria when prioritizing project proposals for funding. These should be derived from thorough district framework planning. A targeting (e.g. towards poverty alleviation) could still be applied. By not posing any sectoral limitations to project proposals, the self-responsibility of grassroots decision makers (communities) is taken acknowledged.

The provision of financial resources into a district fund allows for testing of decentralized planning and implementation procedures. However, provision of such local subsidies should only be provided on a temporary basis and based on reliable prospects for secured own funds (own revenue generation), or reliable and transparent fiscal transfers from central Government. Fiscal provisions under the respective decentralization act have to allow for sustained operations by the local Government entities. Once this is not assured, local subsidies would be meaningless.

The program does not exclude specific groups. In principle, every (group of) citizen and any community can forward project proposals. Public funds from district assemblies (and PRA support) are open to all actors in the district. This approach necessitates the establishment of mechanisms to prevent that resources/benefits are captured by local elites without benefiting broader development in the district. Such mechanisms might be:

1. Clear criteria for fund allocation based on a "vision" and supported by district framework planning.
2. Earmarking of funds.
3. Sound and practical monitoring instruments which feed into a single management information system for local decision-makers.

Sector investment programs and decentralization

The SIP instrument is considered to have good potential to improve both the effectiveness of donor aid and the delivery of public sector services through: (a) providing a cohesive framework for the government's public expenditure program; (b) enhancing the effectiveness of public sector management by eliminating fragmented project management units; (c) fostering local ownership and commitment by empowering the country to determine its development priorities within a decentralized framework; and (d) increasing the efficiency and sustainability of development aid through effective donor coordination and harmonization of their implementation arrangements. A SIP should allow donors to more effectively integrate their (multilateral and bilateral) assistance into national programs and thereby increasing the overall aid effectiveness.

However, the SIP instrument is still new and hard evidence on its performance is very limited. Since there is only one (agricultural) SIP at implementation stage, a conceptual and organizational enhancement of sector investment programs is still constrained by a lack of analyses of specific experiences, critical assessments, and practical guidelines for identification, design and implementation. The knowledge gained so far has not been synthesized and the concept is not fine-tuned to the complex environment. For such a conceptual refinement and an assessment of its impact, partner organizations and donors alike are lacking a clear basis for decision making. It is still unclear how the development of a coherent sectoral policy, which forms the basis for a program definition and serves as a reference point for funding decisions can be ensured. Factors, which determine a full responsibility by local stakeholders, are not established. The process of coordination among major donors in the sector and the need for and functions of a "lead donor" are not instituted. The particular role and tasks of technical assistance for capacity building under a SIP program have not been clarified. The integration of ongoing projects into a sector investment program is only based on crude guidelines. Quality control measures for sector investment programs have not been established. These issues need to be addressed to perfect the instrument, avoid failures and enhance sustainability.

The effects of SIP on decentralization efforts need to be thoroughly assessed. A concentration of resources in a given sector can possibly even undermine decentralization efforts. Sectoral planning for the SIP might ignore or counteract district/regional-planning efforts, which reach across sectors. Sectoral imbalances in the allocation of funds might occur. There are examples of parallel institutions being created specifically for the SIP, which then assume responsibilities of local councils or other local decision making bodies. However, if Local Government is included in SIP preparations, SIP may turn into a tool which supports decentralization. Unfortunately Local Governments have not been involved in SIP preparations yet.

Rural poverty targeting is not necessarily a priority for SIP planners; emphasis is usually given on sectoral growth. Further limitations to poverty alleviation might result out of the strictly sectoral approach of SIP. The definition of the sub-components, which is carried out with staff from the sector department has the tendency to go for a package approach which leaves little scope for community identification of measures. "Systematic Client Consultation, SCC", which is often propagated has not been made operational. However, the Rural Investment Fund, Zambia is an example of a cross-sector, poverty-oriented component within the framework of a SIP.

In ASIP Zambia, institutions were created at district level (District Agricultural Committees), which operate parallel to the local government structure. The pressure for quick results, and ineffective local government institutions are given as reasons for their creation. However, similar arguments have also been used for the creation of Regional Development Authorities in the early days of the IRDPs. The comparatively massive amount of funds going through a SIP compared to what is available to local government bodies (ratio of about 10/1) makes linking of SIP with local government structure difficult. SIP does provide a greater scope for institutional reform of public service providers (key ministries). Planning and implementation responsibilities are delegated to lower (sector department) levels and the institutional restructuring of the respective sector ministry is seen as a key to success. These measures may all be supportive of decentralization.

Linking SIPs with decentralization processes is an important, though not an easy task: while SIPs generally lack a regional focus/regional plan, have a mono-sectoral orientation and operate with (comparatively) large funding, Local Governments in turn have a cross-sectoral orientation, limited funds (exception Ghana, South Africa) and limited capacity. As a result, decentralized decision making about regional rural development by Local Governments and other non-sectoral decision-making bodies can be hampered.

Lessons learned and conclusions

Criterion 1. Local Government vision and regional development plan

Criterion 2. Community initiative and response by local government

Criterion 3. Rural poverty targeting

Criterion 4. Reform of service public provision

Criterion 5. Channeling of fiscal flows

Criterion 6. Financial management

Criterion 7. Regional flexibility

Criterion 8. Cross-sectoral approach

Considering these criteria for successful regional rural development, several fundamental changes in attitude and behavior are required (Schubert 1994):

• From top-down to bottom-up.
• From sectoral to integrated.
• From teaching and informing to listening and learning from communities.
• From supply orientation to demand orientation.
• From centralized to decentralized planning and decision-making.
• From agency ownership to community ownership and management.

Concerning the growing importance of Sector Investment Programs in the context of decentralized regional rural development, it can be preliminarily concluded, that:

1. Sector investment programs induce a deconcentration of sectoral line ministries, but provide obstacles to a comprehensive decentralization. Cross-sectoral issues are not easily addressed. This could reduce the SIP contributions to poverty alleviation.
2. SIP planners need to realize that linking of Sector Investment Programs with decentralization is the key to keep such programs manageable and achieve the desired impact of decentralization efforts.

References

A.Engel, M.Bartels: Program for Rural Action (PRA), Ghana; Summary Report of Project Mission, Accra, 15 - 22 April 1997.

A.Engel, N.Okidegbe: International Workshop on Sector Investment Program in Agriculture, Summary Report, Lusaka, 12-14 February. 1997.

IBRD: World Bank Experience with Rural Development, 1965-1986. Washington DC, Oct. 1987.

GTZ: Local Government & Decentralization in South Africa, Namibia & Botswana. Report by Rural Urban Interaction Consultants for the Mafeteng Development Project. Maseru, 1995.

GTZ: Regional Rural Development (RRD) Update. Elements of a Strategy for Implementing the RRD Concept in a Changed Operational Context. Eschborn. 1993.

GTZ, A.Fastenau: Program for Rural Action (PRA). Project Document. Accra. 1995. GTZ, R.Kressirer: Promotion of Community-Oriented Development Planning in the Eastern Cape Province, South Africa. Project Document. East London. 1995.

GTZ, J.Ising: Strenghtening Manica Provincial Planning Commission. Chimoio. 1997.

C.Mtakati, A.Engel: Decentralization and Rural District Councils (RDC) in Zimbabwe. Harare.1997.

A.Parker: Decentralization: The Way Forward For Rural Development. Washington. 1995.

B.Schubert, et. al.: Facilitating the introduction of a participatory and Integrated Development approach (PIDA) in Kilifi District, Kenya. Berlin. 1994.

B.Weimer, S.Fandrych: Administrative Reform in Mozambique: A Contribution to Peace and Democracy, Maputo. July 1997.

World Bank, Report on "Zimbabwe-Rural District Council Pilot Capital Project", Washington, USA. 1997.