Interim evaluation

Introduction

Burkina Faso is a land-locked country in West Africa with a population of 13.6 million, of which 81 per cent live in rural areas. In 2006, GDP per capita was US$1 213 and the Human Development Index was 0.320, which was the second lowest in the world.

According¹ to Burkina Faso's poverty reduction strategy paper (2003), 46.3 per cent of the population live below the national poverty line of US$0.35 per day, and 94 per cent of poor people live in rural areas.

Community-based Rural Development Project (PNGT2) ² was originally formulated by the Government as a follow-up to the Programme National de Gestion des Terroirs (1994-1998), which was financed by the International Development Association (IDA). The project was identified and appraised by IDA between 1999 and 2000, and IDA was its main cofinancier. IDA also acted as IFAD's cooperating institution and was responsible for loan administration, project supervision and implementation support. For IFAD, the project financing type was therefore "type C".³

The Ministry of Agriculture, Water and Fisheries was the main executing agency.

At appraisal, the estimated total project cost was US$114.85 million. The IFAD loan of US$11.5 million on highly concessional terms was approved by the Executive Board in May 2000 and became effective in May 2002. It was provided on a pari passu basis⁴ along with the first segment (of US$66.7 million) of an IDA adaptable programme loan (APL).⁵ This APL was envisaged to provide financing for three consecutive five-year project phases under the umbrella of a unifying national programme called the National Programme for Decentralized Rural Development (NPDRD). The expected Government and beneficiary contributions to PNGT2 were respectively US$14.3 million and US$9.15 million. The Government of Denmark pledged a grant of US$4.2 million for the project's monitoring and evaluation (M&E) system.⁶ The project closed on 30 June 2007 after an extension of one year.

Evaluation objectives, methodology and process

The interim evaluation of PNGT2 was requested by the Executive Board as part of the annual work programme and budget of the Office of Evaluation (OE) for 2007. According to the IFAD Evaluation Policy, an interim evaluation is mandatory when the IFAD operational division concerned is envisaging a second phase of a project. The evaluation was conducted by OE from March to November 2007.

The first objective was to assess the performance and impact of the project against the project objectives presented to the Executive Board and in the loan agreement. The second was to generate lessons and recommendations for the next phases of the project and other similar operations in Burkina Faso. More specifically, as per OE's project evaluation methodology, the evaluation sought to: (i) assess project performance in terms of relevance, effectiveness and efficiency; (ii) assess the project's impact on rural poverty, the sustainability of benefits, and the project's contribution to innovation, replication and scaling up; and (iii) assess the performance of key project partners, including IFAD, IDA and the Government of Burkina Faso.

Each of the above-mentioned criteria has been rated on a six-point scale.⁷

The evaluation contained three distinct phases: (i) the preparatory phase, which entailed a desk review of documents, a preparatory mission held in March 2007, and a self-assessment by the project team; (ii) the field work phase, which consisted of the main mission fielded in June 2007, during which various stakeholders at all levels were interviewed and achievements were observed in 6 provinces out of the 26 covered by the project; and (iii) the report-writing phase, which entailed data analysis, report preparation and discussion of evaluation findings and recommendations during a stakeholder workshop held in Ouagadougou in November 2007. 

Project design

According to the report and recommendation of the President (RRP) for the project adopted in May 2000 by the Executive Board,⁸ the goal of the project was to reduce poverty and promote sustainable development in rural areas, breaking the spiral of rural poverty characterized by natural resource degradation, reduced production and lower quality of life. Specific objectives were to achieve: (i) improvements in the cost-effectiveness of publicly funded investments at the local level; (ii) increased management capacity of beneficiary groups and their institutions; (iii) greater absorptive capacity of rural areas; and (iv) better access for poor people to productive infrastructure and inputs, social facilities and means to preserve their environment. The project had five components, which are presented in the table below with their relative cost and component objectives.

Project components, relative cost and immediate objectives

Component	Percentage of Total Project Cost	Component Objective
Local capacity-building	7	Improve organizational, technical and managerial capacities of rural communities and their institutions to manage their development in a participatory way and to carry out village and intervillage level subprojects
Local Investment Fund	52	Finance village and intervillage-level subprojects using a local contract management (maîtrise d'ouvrage locale) approach, to improve rural peoples' access to social and economic infrastructure and services
Institutional capacity-building	25	Develop an enabling institutional environment at the provincial and national level for decentralized rural development
Land tenure security pilots	4	Promote sustainable and equitable improvements to the rural land tenure system to contribute to peace and social equity, and natural resource restoration and preservation
Project administration, coordination and M&E	12	Ensure good project coordination and administrative management

From 2004 to 2006, a UNDP project to support stakeholder consultation and local governance was integrated into the project.⁹ A Global Environment Facility (GEF) project to improve integrated management of lowland ecosystems was also incorporated into the project from 2005 and is still ongoing.¹⁰ Both projects provided funds and personnel for specific interventions under the overall coordination and administration of PNGT2. 

The project covered the whole country, but local capacity-building efforts and community investments were initially aimed at 2 000 villages in 26 provinces (out of the 45 provinces in Burkina Faso). In 2004, it was decided ¹¹ to extend project coverage to all villages in those 26 provinces and, ultimately, the project reached over 3 000 villages. The project financed also a small number of subprojects in seven additional provinces, in collaboration with development projects by other donors. The project relied strongly on contractual agreements with national and local service providers from both the public and the private sector, such as consultancy firms and public technical services, and also on the rural communities, to execute its main activities following a faire-faire¹² approach.

Implementation results

Local capacity-building

To pave the way for the Local Investment Fund (LIF) component and as part of the project's contribution to the decentralization process, the project relied on village land management commissions (CVGTs) to plan and manage rural development activities at the village level. The numerous consultancy firms under contract by the project for facilitating local development planning and CVGT establishment used a four-step approach. These were: community sensitization; participatory planning at the village level; formal establishment of the CVGTs; and training and coaching of CVGT members. As such, 2 986 local development plans (LDPs) were prepared in a participatory way, as compared with the 1 500 anticipated at appraisal.¹³ These LDPs were subsequently translated into annual investment plans (AIPs) with the help of the provincial project coordination teams.

In all, 2 981 CVGTs were established by the project, thereby significantly exceeding the appraisal target of 1 200 commissions.¹⁴ CVGT members received close to 260 000 person days of training, covering essential topics such as the functions of the CVGT, basic bookkeeping, subproject design, participatory monitoring, and facilitation and negotiation techniques. A minimum of coaching by local facilitators, on average three days per annum per village, focused on the facilitation of village meetings for annual planning and monitoring and, to a lesser extent, on the utilization and management of community investments.

To ensure sustainable operation and maintenance (O&M) of community infrastructure, the project supported the creation and basic training of O&M committees. Other local capacity-building interventions included: (i) literacy training, reaching 58 000 people at the beginner level and about 36 000 at the intermediate level; (ii) HIV/AIDS sensitization through the establishment of, and support to 231 departmental and village committees to combat AIDS; and (iii) environmental education for rural children, including the preparation of a teacher's guide and the training of over 4 500 primary school teachers in its use.

Local Investment Fund

The LIF was set up to cofinance social and productive investments in rural areas.¹⁵ The village-level facility of the LIF cofinanced subprojects included in the AIPs that were entirely managed by the village communities through the CVGT. The province-level facility was meant to cofinance more important infrastructure works to benefit large numbers of villages. Altogether, 9 622 contracts were signed between the project and CVGT, and funds were transferred to CVGT accounts opened in local financing institutions. Close to 18 250 subprojects with a total cost estimate of US$55.5 million were implemented under this component.

Over 3 000 villages benefited from project investments, which included, in order of importance in terms of financing: social infrastructure, investments for improving agricultural and livestock production, and natural resource-related activities (see table 2). The evaluation estimated that about 30 per cent of the activities planned for in the LDPs were effectively executed with project support. This percentage is satisfactory considering no limits were imposed on the volume of activities and investments that the villages could plan for.

Institutional capacity-building

This component aimed at developing a favourable institutional environment for decentralized rural development at the national and provincial level, by supporting both the human and the physical capacity of several key institutions.

In the first years of implementation, to ensure the effectiveness of the two components presented above (paragraphs 9 to 13), the project trained over 1 300 people, mostly associated with service providers under contract by the project (consultancy firms, trainers and local facilitators).¹⁶ These people were then directly involved in the implementation of activities supported by the project.

The project supported the 45 provincial technical consultation platforms (CCTPs), which were composed of public administration, technical services and development partners and responsible for identifying public investment priorities, and coordinating and harmonizing development approaches in the different sectors. Although the project's most substantial contribution probably was the financing of quarterly CCTP sessions, it also provided essential training to members on topics such as decentralization laws and processes, communication skills, and local development planning.

At the national level, the project actively took part in policy dialogue regarding decentralization in support of rural development. For instance, capitalizing on the project's community-driven development experience, project staff participated in technical review committees for the preparation of guiding documents for decentralization, the general code for local government, and the decree on the creation of village development councils. The project also commissioned several studies, developed guidelines for local development planning, supported the organization of national workshops and prepared training modules regarding decentralization.

Land tenure security pilots

Aiming at ensuring land tenure security for all users, this component supported: (i) the preparation of a practical guide for securing land tenure; (ii) the operation of a national forum for building knowledge and sharing experiences on approaches to land tenure security, namely the National Commission for Securing Land Tenure in the Rural Areas; (iii) the drafting of a national strategy and action plan for land tenure security; (iv) the development of an approach for securing land on which investments cofinanced by the project were built; and (v) the piloting of an approach for peaceful conflict management between farmers and herders through specialized village commissions.¹⁷

Project administration, coordination and M&E

Project staff was hired in adequate numbers and with the skills-mix foreseen in project design. The number of project staff was not reduced after two years of project implementation as originally planned, because the decision was taken during implementation to cover all villages in the 26 provinces of direct intervention (see paragraph 8). Training was provided to project staff, inter alia, on administrative and financial procedures, local development planning and M&E, but the emphasis was placed on learning-by-doing and experience-sharing because of the relatively limited budget for formal training.

Financial management of the project was found problem-free by supervision and audit missions. Several adjustments to project financing were made during project implementation. The ceiling of the special account had to be raised from CFAF 0.75 to CFAF 4.5 billion to allow a smoother flow of funds under the LIF component. In agreement with IDA and IFAD, government counterpart funding was reduced by half, given the difficult economic situation that the country was facing at the time. Finally, the ceiling for the amount of project cofinancing for subprojects was increased for smaller villages.¹⁸

The project set up an M&E system at three levels: (i) monitoring of outputs, with community participation through the CVGTs and supported by local facilitators; (ii) evaluation of outcomes, by means of annual surveys covering all 45 CCTPs and more than 500 villages (since 2006); and (iii) evaluation of rural development impact at the national level within the framework of the NPDRD. For the latter, a national socio-economic baseline survey was conducted in 2004, and follow-up studies were carried out in subsequent years. Several studies were prepared to allow for environmental impact assessment in the subsequent PNGT2 phases.

Project performance

Relevance

The evaluation found the project to be relevant. Although the project design pre-dated the major national strategies and policies regarding poverty reduction, rural development and decentralization, it was consistent with the main directions contained therein. The project design was also consistent with the main elements of the country strategic opportunities paper formulated in 1998. The project rightly aimed at meeting the needs of rural communities, initially in the most disadvantaged villages of 26 provinces, in terms of services and infrastructure through an innovative participatory and decentralized community- driven development (CDD) approach based on effective community empowerment. The IDA project appraisal document (PAD) – which was the guiding document for project implementation – was comprehensive and clear, although the logical framework had some limitations.¹⁹

The project was constructed as a national programme to support the ongoing decentralization process at the key levels (village, province and national) with a strong institutional capacity-building component. Local capacity-building relied mostly on a useful learning-by-doing approach and outsourcing to public and private expertise. Institutional arrangements were also found adequate, with adequately staffed, highly qualified national and provincial project teams. As mentioned in paragraph 19, the number of staff was not trimmed down after two years as foreseen at appraisal, so that sufficient human resources would remain in place to ensure support to the increased number of villages covered by the project. Administrative and financial procedures were found to be broadly adequate, with some relevant adjustments made during project implementation, such as those presented in paragraph 20.

The evaluation found, however, that project design and implementation gave little attention to the IFAD concern that all members of the target group – including the most vulnerable populations – must benefit from project interventions. The project lacked a targeting strategy and a monitoring system capable of capturing data in a disaggregated manner according to poverty and vulnerability criteria. Another limitation of the design was the shift in focus away from community-based natural resource management (NRM). This had originally been a major theme at the time of project formulation and still remains one of the most important issues for agricultural and rural development in Burkina Faso. It was also the evaluation team's opinion that the project missed out on strengthening intervillage coordination and experience exchange on subproject planning and implementation, in particular at the department level, an administrative level that has now become the rural commune, and the most important level of decentralized government.

Furthermore, the evaluation concluded that the project design had not allowed adequate resources for coaching the communities on participatory planning, design of sound subprojects and management of community resources. Owing to the fact that local planning was confined to village boundaries and to one-year execution periods, community NRM activities requiring intervillage consultation and coordination were usually excluded from the local development plans. Provisions to develop the capacity of small entrepreneurs and local service providers were also found to be insufficient, despite the fact that they were key players in subproject implementation and that the rural private sector was playing an increasing role in local development with the advancement of the rural decentralization process.

Effectiveness

Overall, the project was judged to be effective. At loan closing, on 31 December 2007, the project had disbursed 99.46 per cent of its total appraisal cost estimate, and 99.85 per cent of the IFAD loan.²⁰ With respect to the project's specific objectives (as mentioned in paragraph 6), effectiveness was very satisfactory in terms of providing social facilities and collective productive infrastructure to rural communities (see paragraph 30 below). The project was broadly successful with respect to building planning and subproject management capacity of beneficiary groups (see paragraph 27). There is no doubt that cost-effectiveness of publicly funded investments at the local level was increased (see paragraphs ¿37 and ¿38) and that the absorptive capacity of rural areas was improved. The project also obtained significant results regarding institutional capacity-building at the province and national level (see paragraphs 33 and 34). Effectiveness was sometimes found to be weaker, however, in terms of qualitative aspects such as the strengthening of local capacity to manage community resources and infrastructure.

Effectiveness was also found to be rather poor for the land tenure security pilots component.

A succinct overview of project effectiveness under each component is provided in the next paragraphs.

Local capacity-building

Most CVGTs have acquired the necessary skills for subproject planning and management, through the successful combination of coaching and learning-by-doing. The participatory LDPs prepared present a generally good assessment of village socio-economic conditions and development constraints, and propose a long-term village development plan with concrete activities for developing productive and social assets, well matched to the priority needs of the village inhabitants. Effectiveness was also considered high for HIV/AIDS sensitization and environmental education.

The evaluation found, however, that LDPs took too little into account local initiatives and opportunities, and gave inadequate attention to feasibility of planned activities. The subsequent translation of LDPs into AIPs often resulted in a "shopping-list" of investments eligible under the LIF, picked from a standardized menu. There was a clear preference for hardware investments with a maximum input from the project and minimum effort from the communities. As a result, most activities requiring a strong time investment by the population – such as community NRM activities – became a lower priority.

The standardized approach to set up CVGTs resulted in significant differences in capacity and dynamism among the latter, strongly conditioned by initial conditions at the village level. In many cases, CVGTs were driven by a small group of key members who did not regularly consult with the community on important decisions. Poor bookkeeping by CVGTs indicates less effective training and coaching on that aspect. In a similar way, capacity-building efforts to ensure O&M of community infrastructure were too basic and uniform, regardless of the type of infrastructure to be managed. As a result, the evaluation found that probably less than half the community investments have a functional and sustainable O&M arrangement in place.

Local investment fund

As the implementation results presented in paragraphs ¿11 and 13 show, the village-level facility of the LIF component was highly effective in quantitative terms. This was, in addition to adequate component design, mostly due to the existence of good quality local expertise (entrepreneurs, building site controllers, etc.) and adequate supervision of subproject works by CVGT subcommissions. The evaluation found the social and productive infrastructure built under the project to be of good quality. The evaluation tried to obtain a better understanding of a number of key elements in the local subproject management approach, two of which – cost-sharing and contract management by the village communities – are briefly discussed in box 1 below.

Some NRM investments and activities were less effective. For example, herd passageways, tree plantations and improved lowlands were often found in a poor state, mostly because of inadequate maintenance and management arrangements, or unresolved land tenure issues that usually went beyond village boundaries.

The province-level facility of the LIF was less effective due to long delays in project preparation and execution, because, inter alia, studies and works were subject to national tenders, coordination among provincial public agencies was difficult, and the local capacity to carry out major infrastructural works was lacking. As a result, the province-level facility was mostly used to finance village-level investments in villages where the annual cofinancing ceiling had been reached.

Summary of evaluation findings on community cost-sharing and contract management

Community Cost-sharing Depending on the type of subproject, the village community had to share a predetermined portion of the costs, labour force and primary building materials, to promote greater accountability by the CVGT to the community, and stronger ownership by the community. The financial contribution from the village had to be deposited in the CVGT bank account before the project would transfer its part of the financing for the subproject. The collection of financial contributions was often problematic due to coincidence of the collection period with low cash reserves at the household level, lack of trust in the CVGT or disinterest in the activities foreseen in the annual investment plan by certain groups within the community. Because cost-sharing by the community was a condition for obtaining funding from the LIF, local solutions to collection problems were found in most cases, ranging from solidarity within social groups to outright underhand recovery of contributions from contractors. Several cases were observed where the community provided its financial contribution as required by the project, but made an underhand deal with the local entrepreneur (the contractor) to recover part of that contribution from the latter, once the contract was closed. In many villages, contributions were unequal across social strata, sometimes with the adverse effect of promoting patronage and local mechanisms that differentiate rights of access to public infrastructure. For example, many cases were noted where a relatively well-off village inhabitant paid half or more of the village contribution to a project investment, such as a bore well or an input storage building. His family would then often have priority access to the facility and would quite frequently control its management. In many other cases, social groups that were unable to contribute as much as the rest of the village to the costs of infrastructure built under the project, or that had difficulties in paying the service fee to use the facilities, would not receive the same access to those facilities as the relatively better-off. In-kind contributions were the most difficult to collect, as these were not a precondition for project support but were part of the contracts with the local entrepreneurs. Shortages in many cases had to be compensated for by the contractors. Community contract management The tender for contracts to implement works under the AIPs was mostly managed by the CGVT in a transparent way and with close supervision by the project's provincial coordination teams. Irregular practices by contractors to win tenders were, however, quite frequent in certain parts of the country and documented by the project (bribery of CVGT members, below-cost tender offers, etc.). A subcommission of the CVGT was created to supervise works and a controller was hired for each building site. Despite the often insufficient community contribution (see point above), the sample of small and medium private entrepreneurs met by the evaluation team said that they had considerably benefited from the business opportunities offered by the project. They also praised the swiftness of payment at completion of the works.

Institutional capacity-building

At the national level, through policy dialogue, the project effectively brought forward certain ideas that were taken up in the decree on the creation of village development councils (VDCs), adopted in 2007. These ideas included the principle of transfer of mandate and funds from CVGTs to VDCs and the provision for village general assemblies to hold VDCs accountable to the village inhabitants.

At the provincial level, stakeholders met by the evaluation team agreed that the CCTP sessions sponsored by the project allowed for greatly improved consultation as well as increased accountability among public entities and development partners regarding public investments and development interventions. This consultation, however, has remained to a large extent project-driven and so far has not led – at least in the provinces visited by the evaluation team – to either adequate coordination of interventions or harmonization of approaches and procedures, which still depend very much on approaches and procedures dictated from above.

The training of service providers under contract by the project and directly involved in the implementation of project activities undoubtedly had a positive effect on the knowledge and skills of these providers, and therefore on the quality of services provided to the communities.

Land tenure security pilots

Thanks to its mostly financial support to the National Commission for Securing Land Tenure in the Rural Areas (CNSFR) and other agencies involved in land tenure, the project contributed to the discussions and validation of the important National Policy for Securing Land Tenure, which was adopted in October 2007.

The pilot work on approaches for securing community investments and peacefully resolving conflict over land between farmers and herders in six provinces also showed promising results. Land tenure was secured for plots on which community infrastructure was built under the project. In one province, a large proportion of conflicts over land between farmers and herders, mostly cases where herds had damaged crops, could be resolved peacefully by village communities without intervention by public administration.

This component, however, has achieved only part of what was originally intended. Poor effectiveness in the earlier project years due to design, personnel and external problems, led to the component's reformulation following the mid-term review (2004), with significant scaling-down of its ambitions and transfer of part of the component's assignments to the CNSFR. However, the CNSFR has not yet succeeded in involving all public and civil society players with valuable expertise in the field of land tenure. The evaluation also found the practical guide for securing land tenure very theoretical and of rather limited applicability, drawing insufficiently on recent experiences and best practices. The pilot work on securing community investments and resolving conflict between farmers and herders was little exploited and remained disconnected from the institutional and policy support at the national level.

Efficiency

The evaluation did not attempt a detailed economic and financial analysis of the overall project, as many project interventions were directed towards institutional strengthening and promotion of participation and capacity-building of beneficiaries for which the economic benefits are difficult to assess. Economic rates of return (ERRs) were estimated by a study under the project²¹ for certain types of community investments, based upon a limited sample, and indicate positive ERRs overall, ranging from 3 per cent for vaccination centres to 265 per cent for compost pits. An overall portfolio ERR for the LIF was estimated by IDA,²² based on the relative weight of the main subproject types implemented between 2002 and 2006, and points to an ERR of 74 per cent. These estimates appeared optimistic to the evaluation team, which had found O&M problems for many, as well as under-utilization for some community investments (as discussed in paragraphs ¿28 and 50).

Subprojects carried out under the LIF showed good efficiency, with average savings of between 5 and 10 per cent on the standard costs foreseen in the technical and financial reference used by the project. Project estimates indicate that costs were between 6 and 66 per cent lower than sector comparators, with the largest savings evident in social infrastructure. This satisfactory efficiency was largely due to beneficiary cost-sharing and supervision of works, and to the CDD approach, which stimulated the emergence of competitive local entrepreneurs with low fixed operating costs.

Regarding capacity-building and policy dialogue, expenditure was typically high for a project that combined both CDD and policy assistance. Although the project's contribution was remarkable, the quantity of villages reached by capacity-building efforts clearly prevailed over quality on certain aspects (mentioned earlier in paragraph ¿28). The results achieved by the land tenure security pilots component did not justify the resources mobilized.

The institutional arrangements were found efficient, considering that operating costs were maintained below 12 per cent of total expenditure despite the fact that staff were not cut back after two years of implementation as foreseen, and the project was extended by one year.

Overall, on account of the less than satisfactory quality of certain project-supported capacity-building and policy assistance interventions and sustainability issues related to community resources, the evaluation judged the project moderately efficient. 

Rural poverty impact

The project has made an impact in several domains. Impact was most significant on: (i) physical assets: 37 per cent of the country's rural villages were provided with some essential basic infrastructure, thereby increasing the rural population's access to potable water, basic health care, education and some productive physical assets; (ii) human assets: improvements were registered in literacy rate and school attendance, health conditions and awareness, and technical and local development management capacity in the villages covered; and (iii) institution and services: a definite contribution was made to the country's decentralization process, which is expected to raise the quality of public administration and service delivery in rural areas through more effective people's participation in decision-making and greater accountability on the part of the Government and service providers.

The project had a moderate impact on social capital and the empowerment of rural communities in that it strengthened village-level capacity for participatory planning and implementation of subprojects andempowered rural communities to a certain extent in their relationship with public and private service providers. Agricultural productivity was significantly improved on lands treated by soil and water conservation works or fertilized by compost pits, although these lands represent only 2-3 per cent of the cultivated area of involved villages.

Impact on common natural resources management was weak, despite the fact that access to natural resources, in particular land, is fundamental to the livelihoods of most rural households. Soil fertility was improved at the individual field level where soil and water conservation works were successfully carried out, but the areas concerned are too dispersed to have any positive effect at the watershed level. More importantly, few solutions were found to land tenure issues concerning productive land (fields, pastures and forests).

The evaluation found that economic and social benefits remained below expectations for the more vulnerable population groups, including women, youth, herders and immigrants, which were specifically mentioned as priority target groups in IFAD's RRP.

The evaluation also found that the project missed the opportunity to strengthen the capacity of rural communities to coordinate and defend their interests beyond the village level.

This would have been useful in the light of the subsequent creation of the rural communes, which as mentioned in paragraph ¿23, have acted as the lowest level of local government in rural areas since 2005. Finally, harmonization and collaboration between province-level institutions outside the collaborative framework of development projects did not appear to have improved significantly.

Considering the above, overall rural poverty impact was found moderately successful by the evaluation.

Sustainability

A series of positive factors argue in favour of sustainability, namely: (i) a satisfactory uptake of the CDD approach by involved communities, with improved skills at the local level for subproject planning and management; (ii) the adoption, in most cases, of cost-sharing principles (see box 1); (iii) the usefulness of the majority of community-level investments, and therefore the existence of a local incentive to maintain those investments in good condition; (iv) land tenure security for community infrastructure; and (v) the opportunities presented by the next project phases that could focus more on consolidating local O&M capacity in coordination with the rural communes.

However, certain factors may impede sustainability of benefits, such as: (i) the economic non-viability or suboptimal use of some of the economic infrastructure (cereal banks, stores, etc.); (ii) the insufficient contribution by users towards maintenance of social investments (water points, meeting halls, etc.); (iii) the generally poor management capacity and inadequate remuneration of local O&M committees; (iv) unresolved land tenure security issues for productive and NRM investments; and (v) the lack of public funding to sustain public technical service support for local communities and to finance CCTP sessions.

On balance, the evaluation found sustainability of project achievements moderately satisfactory.

Innovation, replication and scaling up

Overall, the first phase of PNGT2 was innovative. The CDD approach followed by the project on a very large scale was both courageous and innovative in Burkina Faso in the sense that the approach gave village communities full responsibility for all technical and financial aspects of subproject planning and management. The three-tier M&E system of the project and the pilots in land tenure security were also the first of their kind in Burkina Faso. Several innovative refinements in procedures and methodology were tried out by provincial coordination teams, for example, in the type of support provided to villages for the awarding of contracts.

 Innovations were, to a certain extent, successfully replicated within the project itself, and scaled up outside the project through policy dialogue. The combination of highly visible field interventions with a strong institutional capacity-building and policy dialogue component, allowed the project to scale up key features of its CDD and provincial consultation approaches, by making sure that these features were incorporated into the rural decentralization process at the national level.

However, the participatory planning approach did not, in itself, promote much innovation because it did not allow for the full exploitation of local opportunities and initiatives (see paragraph ¿27). Likewise, the mechanisms for community investment O&M lacked innovation and were too standardized and simplistic to cater for the highly diverse types of investments and socio-economic contexts (see paragraph ¿28). Innovative experiences in land tenure security were poorly documented and their replication outside the provinces covered by the pilots has not yet been realized. 

Partner performance

IFAD

For IFAD, the project was categorized as financing type C (see footnote 3). The Fund provided a significant financial contribution to the project, but IFAD-specific objectives were inadequately taken into account in project design, which was led by IDA. Relevant issues²³ were raised in the course of the technical and strategic internal quality assurance process during project design at IFAD. These issues were partly addressed in IFAD's RRP, but IFAD was not successful in ensuring that these concerns were adequately reflected in the final PAD.  

IFAD delegated loan administration, project supervision and implementation support to IDA, for which IDA performed below expectations (see paragraph 60). IDA did not give adequate attention to IFAD-specific concerns (see paragraph 24) nor did it promote IFAD experiences and expertise in the region during project implementation. IFAD's capacity to learn from the project has been reduced by the fact that six different country programme managers were involved between project design and completion. Finally, collaboration with other IFAD-financed projects in Burkina Faso was limited, despite the existence of promising opportunities.²⁴ The evaluation rated IFAD performance moderately unsatisfactory. 

Government

The evaluation judged government performance satisfactory. The Government of Burkina Faso showed strong interest in and commitment to the project throughout implementation, and its continuous support certainly contributed greatly to the project's achievements, especially in terms of institutional capacity-building. The Government has made optimal use of the project's resources and experiences to advance its rural decentralization agenda. However, the Government did not fully succeed in ensuring harmonization and coordination among rural development partners and projects in the country.

Most of the project's achievements can be attributed to the good quality of project management at the national and provincial level. As mentioned in paragraph 19, the numbers and skills-mix of the project staff at the various levels were adequate. The project team succeeded, to a certain extent, in fine-tuning the project approaches according to the specific context of each province, and worked at the provincial level in close consultation with technical partners. Project M&E, as presented in paragraph ¿20, was also found to be adequate and innovative, especially in terms of steering the project at the national level. The M&E system has, however, not yet been transferred to the appropriate government agency to become the national rural development M&E system, as originally intended.

IDA

Overall, the evaluation considered IDA's contribution as cofinancier moderately satisfactory. As the main cofinancier of the PNGT2, IDA was the main driver behind the project's innovative and courageous design, which was largely based on other World Bank CDD projects in the region. As such, the main strengths – but also some flaws – in the design of the project, as discussed under project relevance, are attributable to IDA.

While the main ministry involved in project execution has only praise for IDA's support to the project, the other project partners, including IFAD, UNDP and the Embassy of Denmark, found that IDA made little effort to develop a true partnership among donors and showed little interest in other donors' specific objectives.

As IFAD's cooperating institution, however, IDA's performance was found unsatisfactory. While it has to be taken into consideration that IDA administration of the IFAD loan, project supervision and implementation support were free of charge for IFAD, IDA did not adequately respect its commitments. Delays in non-objection statements were frequent, communication with IFAD on both fiduciary and technical aspects was poor and often untimely, and supervision reports, although regular, were of poor quality. The "fluid" way in which IDA organized its supervision and implementation support missions,²⁵ and the often late announcement of such missions to IFAD, made regular participation by IFAD difficult.

Summary of ratings

The table below summarizes the evaluation ratings for the project.

Performance of the Burkina Faso Community-based Rural Development Project

Evaluation Criteria	Ratings
Relevance	5
Effectiveness	5
Efficiency	4
Project Performance26	4.7
Impact	4
Sustainability	4
Innovation, Replication and Scaling Ip	5
Overall Project Achievement27	4
Performance of IFAD	3
Performance of Government	5
Performance of IDA	3

Conclusions and recommendations

Conclusions

Overall, the evaluation assessed the project as moderately successful. The first phase of PNGT2 supported relevant interventions at the village, provincial and national level that responded well to the needs of the majority of the rural people, while remaining aligned with government policies and strategies. This can be attributed to innovative and effective approaches at both the macro level (national programme for institutional capacity building) and the micro level (CDD), implemented by a high-quality project team and skilled service providers.

However, participation by vulnerable groups in local decision-making, project activities and benefits fell short of IFAD's expectations. The project lacked a clear and effective targeting strategy, and therefore its participatory local planning (CDD) approach did not effectively reach the most marginalized and disadvantaged groups, such as women and young people (see paragraph 49).

The sheer number of villages covered by the project meant that quantity often prevailed over quality, and a highly standardized planning and management approach was favoured over more flexible and adaptable mechanisms. This resulted in poor valorization of local development initiatives and opportunities, inadequate attention to the more complex NRM issues and fragile O&M arrangements for many community investments (as analysed in paragraphs ¿27 and ¿28).

Impact on shared NRM was unsatisfactory, because these activities were usually not a priority in local development plans for two reasons: (i) subprojects were confined to individual villages and one-year periods, which is often not appropriate for NRM initiatives (see paragraph ¿24); and (ii) rural communities gave preference to hardware investments (such as infrastructure) that required relatively less time and effort on their behalf (see paragraph 28). The idea of financing larger subprojects through a province-level facility was relevant, but planning and implementation procedures for this facility were found inadequate (as described in paragraph 32). The first phase of PNGT2 also demonstrated the limitations of an isolated and overly ambitious land tenure component (see paragraph ¿36).

The principles of cost-sharing in community investments and charging maintenance fees for services were intended to promote ownership and better management by beneficiaries. Cost-sharing, however, as analysed in more detail in box 1, induced sometimes irregular practices (such as underhand recovery of part of the community's contribution from the local entrepreneurs by those who had contributed). It also led to local mechanisms that differentiate rights of access for the rural population to public goods in favour of the less poor.

Government performance was satisfactory, but both IFAD and IDA performed below expectations. IFAD was unable to attract sufficient attention to its priorities of combating poverty vulnerability and ensuring adequate targeting of the rural poor, as recommended by internal reviews within IFAD, prior to project approval. This was mostly due to IFAD's weak involvement in project design, and supervision and implementation support (see paragraph 55 and 56).  

Recommendations

The evaluation makes five recommendations. The first recommendation relates to IFAD's further involvement in a subsequent PNGT2 phase. The other four recommendations relate to specific strategic issues to be considered by the Government and IDA for the next phase of the PNGT2.

Recommendation 1: Further IFAD involvement. The evaluation recommends IFAD's continued participation in the next phase of the PNGT2. In this regard, it is important for IFAD to engage in a dialogue with the Government and IDA to ensure that the recommendations emerging from this evaluation are taken on board by the main partners fully and in a timely manner. Additionally, in the next phase, IFAD should be more involved in project supervision and implementation support, and ensure that there are opportunities to promote learning across IFAD-supported projects, including PNGT2, in the country.

Recommendation 2: Inclusion of the most vulnerable population groups. Through their support to the next project phase, project partners should ensure that the poorest, most marginalized and most vulnerable among the active rural population fully participate in project interventions and fully share in project benefits. In particular, it would be useful to: (i) improve the project's understanding of the mechanisms of social and economic exclusion affecting the most vulnerable social groups; (ii) develop approaches for local planning and M&E that ensure full participation by vulnerable or marginalized groups and assign unequivocal priority to reducing vulnerability among such groups; and (iii) provide incentives and earmark resources for specific subprojects aimed at reducing vulnerability and exclusion of the rural poor.

Recommendation 3: Empowerment of rural communities in the recent decentralization context. Village and community-based planning approaches should place greater emphasis on endogenous potential and initiatives to develop local resources. The next project phase should support a changing role for public technical services, helping them devote greater attention to providing advice and assistance to local government and community initiatives. These service providers should help develop appropriate mechanisms to promote effective accountability on the part of VDCs and elected local officials in developing, financing, implementing and monitoring subprojects.

Recommendation 4: Sustainability of local investments. To ensure the sustainability of project investments, the next project phase should contribute to: (i) the development of mechanisms for cost-sharing in the construction and O&M costs of community infrastructure that are equitable and adapted to the specificities of different types of investments; (ii) the promotion of rural people's access to the means (inputs, technical advice, etc.) to maximize returns on investments; and (iii) the development of an enabling regulatory framework to finance O&M of public infrastructure in rural areas through the future communal fiscal system.

Recommendation 5: Natural resource management. The CDD approach should be adjusted to accommodate more easily community NRM subprojects that go beyond the geographic boundaries of one village or one rural commune, and take longer than one year to implement. The next project phase should help develop appropriate accompanying measures and financing instruments to support consensus-building processes among communities, villages and rural communes. These processes would aim at bringing users and decision-makers together to adjust the NRM rules to ensure sustainable exploitation or protection of shared natural resources. Land tenure security should be mainstreamed into all project components as a cross-cutting theme.

---

1/ Sources: World Bank: World Development Indicators 2007; United Nations Development Programme (UNDP): Human Development Report 2007.

2/ The project is known in Burkina Faso as Deuxième Programme National de Gestion des Terroirs – phase I. The project name mentioned on official IFAD documents in French is Projet national de gestion de terroirs ( phase II).

3/Project financing type C means that: (i) the project was initiated by another development partner; (ii) this development partner provides project cofinancing; and (iii) the partner acts as IFAD's cooperating institution.

4/According to the pari passu principle, for each disbursement on the IDA loan, an amount equal to one sixth of the IDA disbursement is disbursed from the IFAD loan.

5/An adaptable programme loan is a World Bank Group lending instrument that provides phased support for long-term development programmes by means of a series of loans. Progress in each phase of the programme is reviewed and evaluated on the basis of agreed indicators. The targets set for these indicators (triggers) have to be reached before a subsequent phase can be initiated and corresponding funding is made available.

6/ The Government of the Netherlands had originally envisaged a US$9.0 million grant for the project, but withdrew before project effectiveness following a change in development assistance policy in the Netherlands.

7/ Ratings are given on a scale from 1 to 6, with 6 = highly satisfactory, 5 = satisfactory, 4 = moderately satisfactory, 3 = moderately unsatisfactory, 2 = unsatisfactory and 1 = highly unsatisfactory.

8/ Project objectives were formulated differently in the project appraisal document (PAD) prepared by IDA. According to the PAD, the project development objective was to increase the productive capacity of the rural sector and improve the effectiveness of public investments by developing the institutional and organizational capacity necessary to enable local communities to plan, implement and manage their own development process. Specific objectives of the project in the PAD are to (i) build local capacity to plan and implement rural development and accelerate the pace of public transfers for decentralized rural development, and (ii) support the implementation of the country's decentralization framework. The component objectives in the PAD correspond to those presented in the RRP (EB 2000/69/R.17/Rev.1) presented to the Executive Board.

9/ The UNDP grant-financed Consultation and Local Governance Support Project had an estimated cost of US$2.4 million.

 

10/ The GEF grant-financed Sahel Integrated Lowland Ecosystems Management Project had an estimated cost of US$1.8 million.

11/ This decision was made by the project steering committee following a supervision mission by National Assembly members in 2004 that found project activities highly relevant and effective.

12/ Faire-faire means literally "make-do". Following this approach, the project sources important parts of its interventions out to service providers under contractual agreements. In PNGT2, for example, the creation of village-level institutions and participatory planning were outsourced to national consultancy firms.

13/ At appraisal, it was expected that at least 75 per cent of the 2 000 beneficiary villages would receive capacity-building support and have adopted a local development plan.

 

14/ At appraisal, it was expected that at least 60 per cent of the 2 000 beneficiary villages would have representative and participatory bodies.

15/ Individual income-generating activities were not eligible under the LIF.

16/ The project hired local facilitators to assist the CVGTs with planning and monitoring of project activities. Their number increased along the project implementation period in parallel with the number of villages covered. On average, each local facilitator supported five villages.

17/ In following a joint ministerial order on conflict management between farmers and herders dated 2000.

18/ Originally, the investment ceiling per village was calculated on the basis of a fixed annual amount per inhabitant of CFAF 5,000, regardless of the size of the village. As a result, the smaller villages could not receive sufficient funding for community infrastructure subprojects. After the mid-term review, the project adopted a new formula for calculating the cofinancing ceiling, which attributed to all villages a fixed yearly amount of CFAF 5 000 000, and – counting from the 1001th inhabitant upwards – an additional fixed amount per inhabitant of CFAF 3 000.

19 For example, there was a lack of clarity between objectives and outputs, and inadequate indicators for quality, efficiency and sustainability of project interventions and results.

20/ A financing gap of approximately US$14 million created by the withdrawal of the Netherlands, an overestimate at appraisal of IFAD cofinancing and a decrease in the Government's contribution after the first year of implementation were compensated for by IDA and higher than expected beneficiary contributions.

21/ Konate, S. April 2007. Economic and financial analysis of community investments under PNGT2, Final Report.

22/As part of the preparatory work for the formulation of the second phase of PNGT2.

23/ These included: (i) focusing IFAD support on one component only; (ii) clarifying mechanisms to ensure that marginalized population groups would not be excluded from local planning and, as a result, from project benefits; (iii) putting more emphasis on NRM; (iv) providing adequate support to O&M of infrastructure; and (v) clarifying IFAD's role in project supervision.

24/ Such opportunities included: strengthening the local private sector (Rural Microenterprise Support Project), financing income-generating activities and supporting true community-based NRM (Community Investment Programme for Agricultural Fertility [PICOFA] and Sustainable Rural Development Programme) or supporting rural financial organizations (PICOFA and others). IFAD project coordinators met by the team in Burkina Faso agreed that many opportunities for collaboration and mutual learning had not been capitalized on.

25/ This entailed IDA experts often visiting the project separately, more or less as they saw fit.

26/ The rating for project performance is, as per OE project evaluation methodology, calculated as the average of relevance, effectiveness and efficiency.

27/ The overall project achievement rating is, as per OE methodology, given by the evaluation team taking into consideration its assessment of project relevance, effectiveness, efficiency, rural poverty impact, sustainability, innovation, replication and up-scaling.