BANQUE AFRICAINE DE DEVELOPPEMENT - oecd.org · AFRICAN DEVELOPMENT FUND JULY 1999 Original :...

50
CONFIDENTIAL AFRICAN DEVELOPMENT BANK AFRICAN DEVELOPMENT FUND ADB/BD/WP/2000/14 ADF/BD/WP/2000/18 15 March 2000 Prepared by : OPEV Original : French Translated by : E. Alexander (10/03/2000) Probable Date of Presentation to the Committee Operations and Development Effectiveness TO BE DETERMINED FOR CONSIDERATION MEMORANDUM TO : THE BOARDS OF DIRECTORS FROM : Omar KABBAJ President SUBJECT : BANK’S EXPERIENCE IN POVERTY REDUCTION PROJECTS * Please find attached hereto, the above-mentioned document. Attach. : * Questions on this document should be referred to : Mr. G.M.B. KARIISA, Director, OPEV, Extension 4052. SCCD : N.G.

Transcript of BANQUE AFRICAINE DE DEVELOPPEMENT - oecd.org · AFRICAN DEVELOPMENT FUND JULY 1999 Original :...

CONFIDENTIAL AFRICAN DEVELOPMENT BANK AFRICAN DEVELOPMENT FUND ADB/BD/WP/2000/14 ADF/BD/WP/2000/18

15 March 2000 Prepared by : OPEV Original : French Translated by : E. Alexander (10/03/2000)

Probable Date of Presentation to the Committee

Operations and Development Effectiveness

TO BE DETERMINED

FOR CONSIDERATION

MEMORANDUM TO : THE BOARDS OF DIRECTORS FROM : Omar KABBAJ

President SUBJECT : BANK’S EXPERIENCE IN POVERTY REDUCTION PROJECTS *

Please find attached hereto, the above-mentioned document.

Attach. :

* Questions on this document should be referred to : Mr. G.M.B. KARIISA, Director, OPEV, Extension 4052.

SCCD : N.G.

GAF6281
Rectangle
GAF6281
Rectangle

SCCD : N.G.

AFRICAN DEVELOPMENT BANK AFRICAN DEVELOPMENT FUND

JULY 1999 Original : French Distribution : Limited

BANK’S EXPERIENCE IN POVERTY REDUCTION PROJECTS

THIS REPORT HAS BEEN PRODUCED

FOR THE EXCLUSIVE USE OF THE BANK GROUP

OPERATIONS EVALUATION DEPARTMENT

Consultant : KANE HAMIDOU BABA

GAF6281
Rectangle

TABLE OF CONTENTS

Page ACRONYMS AND ABBREVIATIONS

ANALYTICAL SUMMARY

1. INTRODUCTION

1.1 Bank’s Poverty Reduction Approach in the Rural Sector 1.2 Review of Bank Experience

2. PROJECT BACKGROUND

2.1 The Economic and Social Context 2.2 Summary Project Description

3. PERFORMANCE OF PROJECTS REVIEWED

3.1 Implementation Performance 3.2 Operational Performance 3.3 Institutional Performance 3.4 Economic and Performance 3.5 Social Performance 3.6 Environmental Performance

4. SUSTAINABILITY

4.1 Institutional Framework 4.2 Human Resource Capacity Strengthening 4.3 Sustainability of Social Results

5. BANK’S CONTRIBUTION

5.1 Project Design and Preparation 5.2 Bank Operating Modalities

6. TRENDS OF ONGOING PROJECTS 6.1 Strategic Options 6.2 At the Operational Level 7. KEY FACTORS FOR PROJECT SUCCESS 8. CONCLUSION, LESSONS AND RECOMMENDATIONS 8.1 Conclusion 8.2 Lesson 8.3 Recommendations ANNEXES

I. List of Projects Reviewed II. Project Retrospective Matrix III. Performance Matrix IV. List of Documents Consulted

(i-vii) 1 1 1 2 2 3 4 4 5 7 9 10 13

15 15 16 16

17 18 18

19 19 20

21

22 22 23 24

List of Abbreviations and Acronyms ADECAF : Agence de développement de la zone caféière (Coffee Zone Development Agency) AIDB : Ethiopian Development Bank AIC : Association d’intérêt collectif (Cooperative Society) CRDA : Commissariat régional au développement agricole

(Regional Agricultural Development Agency) IDRC : International Development Research Centre DDA : Division du développement agricole (Agricultural Development Division) RD : Rural Development FAO : Food and Agricultural Organization IFAD : International Fund for Agricultural Development FMG : Malagasy Franc GIR : Groupements d’intérêts ruraux (Rural Cooperative) OCV : Office des cultures vivrières (Food Crop Agency) OMVS : Office de mise en valeur de Souassi (Souassi Development Agency) OTD : Office des terres domaniales (Public Lands Agency) PDAR : Programme de développement agricole et rural

(Agricultural and Rural Development Programme) PDR : Programme de développement rural (Rural Development Programme) PIDER : Programme intégré de développement rural (Integrated Rural Development Programme) RMC : Regional Member Countries PPI : Périmètres publics irrigués (Public Irrigation Schemes) PRP : Poverty Reduction Project CAR : Central African Republic Tunisia – AC : Agro-Complex Project, Tunisia Tunisia – DIM : Mahdia Rural Development, Tunisia ERR : Economic Rate of Return UA : Unit of Accounts PIU : Project Implementation Unit

SUMMARY

The rural sector occupies a significant position in the Bank’s portfolio. Review of the Bank’s poverty reduction experience in rural development projects was conducted using available documents, comprising appraisal and completion reports, and country strategy documents on about ten completed projects. The review also tapped from interviews with key-players (Bank experts involved in rural development operations and poverty reduction projects1). Lastly, it focused on recent projects with a view to establishing a comparative base and delineating current trends. Although the review treated several aspects of traditional evaluation (project implementation modalities, institutional issues, social effectiveness, etc.), these criteria were reconsidered from the angle of the study itself, i.e. impact on poverty. Unfortunately, the absence of much vital information made the task an onerous one. Whereas the objectives were spelt out at appraisal, with figures to back them up in some cases, there were no objectively verifiable indicators to show the difference between the “no-project” and project situation. Risk analysis was more of a formality, and follow-up/evaluation instruments were hardly operational. There was no logical framework on previous rural development projects. I. Project Context 1.1 The projects here reviewed were initiated and implemented in the eighties – a period characterized by the introduction of structural adjustment and related policies, and the social dimensions of adjustment. Without doubt, the decade marked a turning point in the awareness of the scope of poverty. 1.2 Of the projects reviewed, a wide range of activities were undertaken in the rural sector, even though the agricultural sub-sector took the lion share. The projects reflect the Bank’s sectoral priorities as well as borrowing countries’ major orientations. All projects reviewed were locally implemented. Often, project implementation was considered as a means through which the project zone could catch up with other regions of the country (e.g. Tunisia, CAR and Malawi). 1.3 Practically all the projects considered aimed at increasing agricultural and livestock production. The construction of physical infrastructure was a major component for half the projects sampled. The health and institution building components were often taken into consideration through the strengthening of existing structures or establishment of new ones (health centres). Environmental protection appeared to be incidental except in the case of the Bush Fire Control and Reforestation Project in Senegal where it was the project’s raison d’etre. 1.4 Project beneficiaries are often mentioned in vague terms. Nevertheless, certain beneficiary categories can be identified, even if little is said about the fate of the poor or of women who have obvious problems with regard to access to land. 1 The said documentation mostly comprised project appraisal and completion reports, country strategy papers, country portfolio performance and the Bank’s sectoral policy papers. Mid-term supervision and review reports were generally used in establishing the recent trends of ongoing projects. In all, about fifteen interviews were conducted with Bank experts involved in rural development and poverty reduction projects.

ii

II. Project Performance 2.1 Although projects were programmed to cover 4 to 5 years, the average life of operations considered ranged from 9 to 10 years, indicating poor implementation performance. Practically all projects underwent significant component modifications compared with what the appraisal outlined. Several reasons account for this, especially: (i) project organizational difficulties, exacerbated by institutional changes brought about by instability in the countries themselves (Ethiopia, Guinea-Bissau, CAR); (ii) inadequate project appraisal planning, often leading to over-estimation of the borrower’s capacity particularly in connection with the setting up of physical infrastructure (the two projects in Tunisia, that in Malawi and CAR); (iii) inflationary trends in certain countries (Egypt, Ethiopia, Malawi) which worsens the poverty situation; (iv) climatic constraints (Mahdia Rural Development Project in Tunisia, the Irrigation Development Project in Madagascar, the M’Baiki/Berberati Project in CAR); and (v) limited community participation. Apart from the Reforestation Project in Senegal, the Irrigation Project in Madagascar and the Integrated Rural Development Project in Rwanda, all other projects considered lacked consensus-building mechanisms. 2.2 Institutional performance differed from one country to another. However, whatever the degree of complexity in project set-up, they all came under the supervision of the Ministry of Agriculture or Rural Development. Is there a correlation between the management or supervisory system of these diverse categories and project institutional performance? We may hazard a “yes” to that question. 2.3 Although it may seem too early to draw lessons on the institutional performance of projects considered, three comments may be made: (i) the first project category (Madagascar, Senegal, Egypt) stand out, with the proven existence of human resources, a high degree of project management autonomy vis-à-vis the supervisory body, and use of adequate sub-contracting; (ii) secondly, the project category with unsatisfactory institutional performance (Tunisia-DIM, Tunisia Agro-Complex, CAR, Rwanda) seem to have been pulled down by the “structural impact” of public agencies that lacked the capacity to adapt in the face of specific project requirements; and (iii) in the project category with poor institutional performance (Guinea-Bissau, Ethiopia, Malawi), the projects were conducted as if the potential functions were more important than the obvious ones, from the sociological perspective. Under those conditions, they certainly lost their initial objective and it is easily understandable why the projects took so long to implement. 2.4 The economic performance table shows that the economic rates of return projected at appraisal were generally fairly optimistic. For Guinea-Bissau and Ethiopia, the lack of reliable data bases with which to calculate the ERR on completion reflected the failure somewhere along the line of project follow-up systems. 2.5 It is difficult to evaluate the social performance of projects under review. There is some confusion in the completion reports between economic and social performance, or the latter is generalized and absorbed as part of the broad project impact. Usually, the social objectives were formulated in vague terms. Specific indicators were given only for production development and physical infrastructure. Variables upon which the indicators were based were not adequately broken down – all of which makes any evaluation of a poverty reduction project difficult. Such major variables as the socio-economic categories, gender, income and living standard were not clearly highlighted. Yet, during project follow-up and completion, they could have provided the basis for measuring the social effectiveness of projects in the light of relevant indicators, of which food security, employment, health, training, access to credit, rural exodus, etc.

iii

2.6 Food security for the target population is shared by practically all rural development projects. A review thereof indicates good performance for Madagascar, Egypt and Rwanda, and to some extent the Agro-Complex Project in Tunisia. The performance is poor for all other projects save Malawi (borderline). Where success was recorded (e.g. Madagascar and Malawi), emphasis was laid on close technical supervision, rural credit for procurement of factors of production, including seed quality and dynamic vocational organization of farmers, resulting in considerable food crop output (rice, sorghum, sweet potato). The ensuing surplus production contributed substantially to food self-sufficiency. 2.7 Indicators on access to facilities (potable water, housing, transport and household equipment) are, in our view, relevant in reviewing the “standard of living” variable, particularly allowing for observation of the behaviour of the income variable with regard to housing and household equipment. It is indispensable to take these indicators into account when evaluating the impact of poverty reduction projects. However, given the few available documents on projects, there is need for caution in interpreting the performance obtained. That notwithstanding, rural development project accommodating measures (road and access road improvement) formed a key component of nearly all projects reviewed. Performance varied from one project to another. 2.8 Lastly, the social performance of rural development and poverty reduction projects should also be analyzed from the angle of local development and impact on rural exodus. Since some of the projects were presented during appraisal as necessary for geographic balance (e.g. Tunisia, Malawi, Ethiopia, Guinea-Bissau), it is normal to wonder to what extent they contributed to social mobilization and to halting rural exodus. Completion reports often skipped both indicators. In connection with local development, performance recorded in Senegal, Madagascar and Egypt may be considered satisfactory, reflecting the people’s support of project objectives. 2.9 In nearly all projects reviewed, environmental performance was only reconsidered on project completion, partly indicating the minor attention given to ecological issues during the eighties (project appraisal period), and the growing trend towards taking those issues into account, given their close link with sustainable development. If the environmental performance were judged on the basis of the negative impacts, attenuation measures and positive impacts, then the rating would be very satisfactory for the North-East Senegal Project, satisfactory for the El-Beheira Project (Egypt), the Tsirihibina Project (Madagascar), the Agro-Complex Project (Tunisia) and the Byumba Project (Rwanda). With the exception of the Tunisia-DIM Project whose rating would be difficult owing to lack of information, the environmental performance of other projects is commensurate with their overall impact – nearly imperceptible and zero in their impact area particularly with regard to production, direct or indirect activities. In other words, based on the very modest results obtained, there is hardly any difference between the pre- and post-project situation.

III . SUSTAINABILITY 3.1 Poor performing projects stand little chance of showing sustainable achievements, nor are good results recorded here and there necessarily a guarantee for sustainability. The factors affecting the sustainability of projects considered are analyzed at three levels: the institutional framework, capacity building and sustainability of social results.

iv

3.2 Sustainable projects are usually those that receive a firm commitment from RMCs. For the El-Beheira Project, the Government of Egypt demonstrated its political will in support of the operation by increasing its contribution beyond the ceiling set at appraisal. The Egypt Project enjoyed considerable administrative and financial autonomy as did those of Senegal and Madagascar. Since it enhances community participation, Senegal’s decentralization edict is equally a guarantee for project sustainability. On the other hand, political instability and frequent change of staff in the wake of administrative reorganization is a permanent challenge to the sustainability of certain projects, irrespective of whether they were implemented within a regional development programme framework (Ethiopia, CAR, Rwanda, Malawi). 3.3 The sustainability of social results of projects considered is generally hypothetical and hard to evaluate. The underlying reasons are the same, not least of which the absence of social indicators during project start-up and lack of impact assessment. No stock is taken during the identification, preparation and appraisal phases. Information gathered is too general, borrowed, indeed anecdotal. “No-project” situations are hardly differentiated from projections. If described, social achievements are generally viewed from the physical infrastructure quantification perspective, whereas the sustainability of social results is basically evaluated on a qualitative scale. For instance, in connection with health, health centre buildings and rehabilitation works are very easily highlighted, but little is said about the provision of services. IV. BANK’S CONTRIBUTION 4.1 In view of the Bank’s limited participation in identification and preparation of rural development projects, it was not possible to list major project shortcomings in relation to quality. Indeed, little attention was given to institutional changes, to the detriment of the life of most projects. Appraisal missions, generally unevenly composed, did not focus on essentials, namely the definition of development objectives and social concerns, and cross-cutting issues. As a result, there were no focused social studies and precise beneficiary targeting. In like manner, land access issues were often overlooked. 4.2 It is also symptomatic that least performing projects are often those that received little Bank supervision. Missions to such projects focused more on the physical performance. Financial matters regarding the project expenditure circuit were never considered during such assignments. Supervision reports rarely, if ever, treated issues relating to the life of the beneficiaries, their active or passive support to the project and their real concerns. Furthermore, supervision mission feedback was often slow and forwarded so late to executing agencies that they rarely contributed to solving problems and correcting project slippage. V. TREND OF ONGOING PROJECTS 5.1 Projects here reviewed were completed between 1994 and 1996. Their reference framework was set a decade earlier. New rural development projects draw inspiration from ADF VII and VIII operational priorities as well as the Bank’s new vision. Whereas past rural development projects were designed to contribute to improving the living standard of rural dwellers, the new thrusts establish a more direct link between poverty and rural development. Therefore, emphasis is on support to agriculture and rural development as “factors essential to production growth and improvement of the living standard of the poor”. Poverty reduction projects now fall within that framework.

v

5.2 At the operational level, more attention was given to gender issues and project environmental impact. Moreover, ongoing rural development and poverty reduction projects have moved away from the nearly undifferentiated notion of beneficiaries as conceived in past projects. 5.3 These positive developments, equally reflected in the quest for parameters on which basis results can be measured as well as the preparation of project design and implementation matrix (logical framework), count among the innovations in contrast to the projects reviewed. However, there are still certain lapses, especially concerning the formulation of performance indicators and management of the logical framework matrix as a development tool.

VI. SOME KEY FACTORS TO PROJECT SUCCESS 6.1 Environment Assessment: the facts show that beneficiaries played a marginal role in project implementation and practically no role at all at the initial phase. Generally, the principle of target-group participation in rural development projects remained very abstract. The operational impact was so limited that it could neither be embodied in any methodological approach nor through objectively verifiable indicators. Additional in-depth studies would have enabled the Bank to avoid certain errors (e.g. the financing of an emigrant family resettlement component under the Tombali Rice Project when the whereabout of the families was not even known!). 6.2 Beneficiary Support: since project success and sustainability are closely linked to beneficiary support, beneficiaries should be considered as participants in the full sense of the term. Of course, without the people’s involvement, community participation becomes yet another chore. Successful projects (e.g. Senegal, Madagascar, Egypt) are those that were able to mobilize the people, especially women and children. 6.3 Institutional Issues: the performance of certain projects underscores the importance of institutional issues. Successful projects focused on limited objectives, in which case activities were implemented within a decentralized framework under the supervision of a technical directorate. Apart from the use of sub-contracting, the factors contributing to success also included the proven existence of local human resources and low sensitivity to the political environment. On the contrary, projects implemented within the framework of existing public agencies had more disadvantages than advantages. 6.4 Food Security: with a view to poverty reduction and for all rural development projects, this appears to be a decisive indicator in judging project success or failure. Where it was slightly concealed or overshadowed by other components, the results were hardly conclusive even when much had been achieved. 6.5 Basic Social Services (Health/Education): in projects in which these components succeeded, the quality of services played an enormous role. Even in projects whose key objective of increasing agricultural production was not attained, access to primary health care, potable water, women and youth training all make the failure of a rural development project very relative. 6.6 Rural Credit: the weakness of this component worsened project poor performance. Direct credit management by a given project or its administration by banks or state structures constituted a failure factor. However, although limited, proximity management especially through farmers’ cooperatives, produced far better results.

vi

VII. CONCLUSION 7.1 Multiple-component rural development projects of the eighties were difficult to implement, certainly due to institutional inadequacies and weak human resource capacities in borrowing countries. However, that fact does not contradict the relevance of the strategy and objectives of rural development projects. From an etymological perspective, the aim of such projects was not to combat poverty but to improve the living standard of rural dwellers. Thus, their stated ambition was to contribute to reducing poverty. It is true that for the projects reviewed and whose appraisal reports go back to the early eighties, even though the word “poverty” is not expressly stated, the designers clearly had the concept in mind. 7.2 However, the perception of a rural development project as necessarily integrated makes them multisectoral, i.e. comprising several components. Given the rural development strategy, the projects covered a wide range of activities, and resulted in the thinning out or dilution of the declared objectives. Hence, several projects concentrated on the physical achievements (building construction, rural roads, etc.), abandoning or losing sight of their core objective of poverty reduction and agricultural production. VIII. Lessons The key lessons are summarized below: i) Bank-funded rural development projects were designed around small-scale farmers, especially those who already owned land; hence even in the rural area, the projects could not attack the roots of poverty, namely landless men, women and youths (see Section 2.2). ii) The traditional implementation period of four to five years for rural development projects proved unrealistic, as a result of which account could not be taken of major issues such as the beneficiary characteristics, land tenure system, organization and availability of labour or women’s role in the agricultural production systems ( par. 3.1.1 and 3.2.1). iii) Although a longer project preparation period is no guarantee for success, it helps to reduce error margins where beneficiary involvement and support are sought and obtained (inset 3.3; par. 3.3.1; 3.3.4; 3.3.5; 3.3.6 and Section 4.1). iv) It is important to involve development beneficiaries in project/programme design, choice of priorities and implementation, thus motivating and making them accountable for the performance (inset 3.2 ). v) The institutional set up granting project implementation units management autonomy from national administrative structures resulted in greater transparency and efficiency (par. 3.3.2; 3.3.3 and 3.3.7). vi) In view of the land tenure systems and the social division of labour which restricts certain domestic or supporting activities to women (harvest management, backyard livestock farming, small-scale productive projects, etc.), the allocation of agricultural credit almost exclusively to agricultural production can only benefit women indirectly, randomly and marginally (see par. 3.2.3). vii) The people’s interest in a project’s environmental component is closely linked to the obvious short-term benefits that they can gain from it. Sustainability clearly seems to be secondary (par. 4.3.2).

vii

viii) From this review, it is clear that the impact and sustainability of the social results of a poverty reduction project depend largely on the quality of services and not physical achievements (par. 4.2 and 4.3). ix) Foreign technical assistance was virtually imperceptible, with focus concentrated on physical achievements and the short-term. The results may not be sustainable. Management procedures were not internalized. Successful projects were those that already had proven local human resources (Egypt, Madagascar, Senegal) (par. 4.2.1 and 4.2.3). x) Follow-up/evaluation of rural development projects was rated poor. Evaluation techniques do not permit measurement of changes relating to the impact of rural development project activities on poverty reduction, far less give reasons for such changes (par. 5.2.1; 5.2.4; 5.2.5; 5.2.6). xi) The evaluation of a project’s impact on poverty reduction is only useful if it is backed by quantitative and qualitative analyses, using measurable indicators (par. 3.5.1; 3.5.6).

1. INTRODUCTION 1.1 Bank’s Poverty Reduction Approach in the Rural Sector

1.1.1 The rural sector in Africa is in the throes of a crisis as reflected in the precariousness of the people’s living standard. There is a wide consensus on the symptoms and consequences of the crisis: agricultural production and output no longer meet the needs of a fast-growing population, and the number of victims of poverty, famine and disease is increasing. The Bank’s approach is mirrored in its lending thrusts. The agricultural sector policy paper stresses that “the Bank finances integrated rural development projects coordinating investments in physical infrastructure and social services”. Generally, according to the World Bank’s African poverty profile, poverty is more widespread and severe in the rural area. To quote the UNDP, “worldwide, more than a billion people live below the poverty line, a quarter of them in Africa. Among the latter, 75% are rural dwellers”.2 1.1.2 In the face of these major challenges, it is understandable why the rural sector takes a large proportion of the Bank’s portfolio (40%), and rural development projects are used as “vectors for reducing rural poverty”. The Bank considers rural development as one of the major areas of intervention, through which the poor are guaranteed food security and access to basic social services. Therefore, rural development projects cannot but form part of the Institution’s sectoral priorities for reducing rural poverty. Rural development projects necessarily have to cover a wide range of activities, and a review of project objectives highlights the fact that improvement of the people’s living standard is at the forefront. A sample of projects reviewed (i.e. multiple-component rural projects appraised in the eighties) falls within this financing category. 1.2 Review of Bank’s Experience

1.2.1 The evaluation of the Bank’s poverty reduction experience in rural development projects mostly aims at documenting the design and implementation processes of Bank projects with focus on their impact and sustainability, using about ten completed projects in nine African countries and five sub-regions. Going by the performance criteria of projects considered, the review evaluates the impact on poverty reduction. Projects reviewed were initiated between 1984 (earliest) and 1998 (the most recent), and completed between 1992 and 1996. Furthermore, the projects were implemented in various countries and under diverse implementation structures. The sociologically diverse nature of these projects weakens the comparative base. 1.2.2 The evaluation of the Bank’s poverty reduction experience in rural development projects was conducted using available documentation. It also tapped from interviews with key players (Bank experts involved in rural development operations and poverty reduction projects)3. These discussions provided useful lessons on poverty reduction and various phases of project implementation for both old and new operations. They confirm, within the context of rural development projects that the Bank financed in the eighties, that although the 2 In World Development Report, UNDP, 1997. 3 The said documentation comprised project appraisal and completion reports, country strategy papers, country portfolio performance and the Bank’s sectoral policy papers. Mid-term supervision and review reports were mostly used in establishing the recent trends of ongoing projects. In all, about fifteen interviews were conducted with Bank experts involved in rural development and poverty reduction projects.

2

interpretation of objectives was basically correct, the Bank fought “around” poverty instead of “against” poverty. Happily, the new thrusts appear to address the issue differently. The reflection was extended to similar projects implemented by other development agencies. 1.2.3 Although the review comprised several aspects of traditional evaluation, that is to say the results obtained vis-à-vis objectives set at appraisal (project implementation modalities, economic rate of return, social effectiveness, etc.), these criteria were reconsidered from the very angle of the study: impact on poverty. Unfortunately, much needed information was not available. Although the objectives were clearly stated at appraisal and sometimes with figures to back them up, there were no objectively verifiable indicators to establish the difference between a “no-project” situation and projections. Risk analysis appeared more as a formality, while the follow-up/evaluation tools were hardly operational. There was no logical framework for these rural development projects, worsened by the Bank’s limited involvement in the project preparatory phase. 1.2.4 The evaluation of the Bank’s experience will remain incomplete insofar as the project beneficiaries who are also actors in their own right are not heard. For instance, it is a known fact that the participation of target groups, the extent to which they are sensitized and their capacity for involvement are decisive factors in any evaluation concerning project sustainability. Among the projects reviewed, some are in the second phase while others ran out of steam. It was important to support the conclusions, lessons and recommendations in the review with “in-situ” field surveys on project supervisory structures and the beneficiaries themselves. When interpreting the results, these factors should be borne in mind. 2. Project Background

Generally, the projects reviewed were implemented under an unfavourable context, itself aggravated by national specificities. At the macro-economic level, difficulties mostly concerned the fall in international cash-crop prices, inflation and heavy currency depreciation in certain countries. Further in this review, it will be seen that where projects were sensitive to these three exogenous factors, performance was far from satisfactory (e.g. Ethiopia, CAR, Malawi). However, exogenous factors also led to project success and/or failure, linked to the choice and number of components, the type of institutional framework, the extent of beneficiary involvement and a country’s political stability. 2.1 The Economic and Social Context 2.1.1 Projects in this review (see Annex 1) were initiated and implemented in the eighties – a period marked by the introduction of structural adjustment policies and their corollary, social dimensions of adjustment. The decade in question clearly signalled a turning point in poverty awareness. Describing those trends, the Knox Report noted that in recent years, population growth and traditional extensive farming methods had brought much pressure to bear on the arable land. Food crop production per capita had nose-dived, food importation had increased. The report went on: “Despite increasing aid, about 100 million people are malnourished and Africa records the world’s lowest calory intake per capita”.4 4 David Knox, Report on the Quality of ADB Loans, April 1994.

3

2.2.2 The crises facing Africa’s rural sector show basic interdependent symptoms that limit the implementation performance of all Bank rural development projects. Various weaknesses have been highlighted, and concern the project cycle, the institutional capacity of management structures, non-involvement of key beneficiaries and difficulties to delineate relevant parameters on inter-sectoral – therefore necessarily cross-cutting – issues: poverty reduction, gender and environment. During appraisal, the project documents merely brushed over these matters. 2.2.3 The complex relationships between a project and a country’s political and social context are often the background that colour and determine the conditions for the success and/or failure of a rural development project. Thus, at the socio-economic level, projects that helped raise the living standard of target groups and improve their nutritional and health conditions are those with adequate follow-up and located in countries with relative stability (Egypt, Senegal, Madagascar). On the other hand, projects marked by frequent management changes or low social mobilization reported marginal social performance. The Sidamo-Gamo-Gafa Project (Ethiopia) is a borderline case since it was implemented within a specific national context typified by war and extensive administrative and political changes involving the rectification of old regional boundaries. These institutional changes which overtook the project led to the fragmentation of its activities and the dissolution of the Monitoring Unit. Any social impact that the project may have had could only be induced and, at best, marginal. 2.2 Summary Project Description 2.2.1 The ten rural development projects here reviewed are located in nine countries of the Bank’s five sub-regions. Excluding Tunisia where two projects were selected, one project on which a completion report had been prepared per country was used for the purpose of this review. Of the projects in question, a wide range of activities were undertaken in the rural sector, even though the agricultural sub-sector took the lion share. The projects reflect the Bank’s sectoral priorities as well as the borrowing countries’ major orientations. 2.2.2 Projects considered are characterized by the fact that they were implemented locally and often, such implementation was considered a means for the project zone to catch up with other regions of the country (e.g. Tunisia, CAR, Malawi). In principle, having focused on reducing regional imbalance, the projects aimed at improving the living condition of beneficiaries, thus contributing to strengthening the “poverty reduction” dimension. 2.2.3 Lastly, certain projects were implemented within the framework of ongoing regional development programmes (PDAR III in Ethiopia, Agro-Complex in Tunisia, Irrigation Development in Madagascar and El-Beheira in Egypt). Other projects were implemented under the direction of ad-hoc structures and supervised by several authorities. Institutional ambiguities in connection with project implementation will be discussed further in this review. 2.2.4 Practically all the projects had the set objective of increasing agricultural and livestock production. Construction of physical infrastructure was a major component for more than half of the projects sampled. The health and institution building components were often taken into consideration either with regard to strengthening existing facilities or establishing new ones (health centres). Environmental protection appeared to be incidental except in the case of the Bush Fire Control and Reforestation Project in Senegal where it was the operation’s raison d’être.

4

2.2.5 Project beneficiaries are often mentioned in vague terms. An analysis shows frequent repetition of “peasant families” without any attempt to clarify the target-groups. However, using the small farm size, certain beneficiary characteristics can be identified even if such indicators are not very revealing of the fate of the poorest or of women whose access to land remains a problem. Apart from the supervision and research structures, other beneficiary categories such as modern farmers (Madagascar), immigrants (Guinea-Bissau) and the unemployed (Tunisia) are explicitly mentioned. 3. PERFORMANCE OF PROJECTS REVIEWED

The absence of a project logical framework is a handicap when reviewing performance projections/gaps. However, the retrospective analysis of the project logical framework highlights significant trends, using performance categories as the yardstick. Hence, although varying from one project to another, poverty reduction performance is generally low. Performance review, especially social, on the basis of retrospective indicators, allow verification of achievements and the limits of probable impacts. Accommodating measures and attempts to create enabling conditions for poverty reduction mobilize more time and resources than direct support to beneficiaries. Notwithstanding the fact that the ten projects reviewed pursue the same development objectives (improvement of the living conditions of rural dwellers) often through production, farming and social mobilization, the performance is mixed. However, it is true that intrinsic results were obtained with regard to rural infrastructure. 3.1 Implementation Performance 3.1.1 On average, sixteen months elapsed between the loan signature date and entry into force, due, among other things, to borrowing countries’ difficulty in meeting necessary conditions, institutional changes connected with the political decentralization of the eighties and ignorance of Bank procedures. 3.1.2 These factors also affected the disbursement timeframes; lastly, they contributed much to extending the implementation phase, thus widening the gap between projections and achievements. Whereas the project implementation phase was generally planned to cover 4 to 5 years, on completion, the average life of projects considered ranged from 9 to 10 years. Major modifications were made to the components of practically all projects compared with what the appraisal set out. Although several reasons account for that, it is often due to:

i) project organizational difficulties, worsened by institutional changes arising from the instability of countries themselves (Ethiopia, Guinea-Bissau, CAR); ii) poor project appraisal planning often leading to over-estimation of the borrowers’ capacities, especially the capacity to set up physical infrastructure (two projects in Tunisia, one in Malawi and one in CAR); iii) inflationary trends in certain countries (Egypt, Ethiopia, Malawi), thus aggravating poverty; iv) climatic constraints (Mahdia Rural Development Project in Tunisia, Irrigation Development in Madagascar, M’Baiki/Berberati in CAR); and

5

v) the limits of community participation. Excluding the Reforestation Project in Senegal, the Irrigation Project in Madagascar and the Integrated Rural Development Project in Rwanda, nearly all projects considered lacked dialogue mechanisms. The quitch-grass (jujube) eradication component in Tunisia and the one on emigrant resettlement in Guinea-Bissau are a caricature of that phenomenon.

3.1.3 Implementation schedule overrun led to repeated modifications to the goods and services procurement lists. The relatively satisfactory project performance points to a net predominance of physical infrastructure. The civil engineering works components (buildings, tarred or improved pavements, wells) are well represented. Land development through agricultural intensification was more pronounced in North African countries (Egypt, Tunisia-AC) than in sub-saharan Africa, even though the impact on poverty varied from project to project. Other positive performance was obtained in connection with improved grazing land and soil conservation (Tunisia-DIM, Eastern Senegal, Rwanda-BYUMBA). Lastly, the quality of community participation through sensitization, extension and implementation of women’s or youth programmes or involvement of village committees, mixed in Egypt, seemed satisfactory for projects implemented in Senegal, Rwanda and Madagascar. 3.1.4 Despite the extended implementation phase for all projects, those in Guinea-Bissau and Ethiopia seem the least satisfactory, even from the physical implementation perspective. In Guinea-Bissau, the construction of two large dams, housing for technical assistance workers and resettlement of immigrant families from the war of liberation which were all components of the Tombali Rice Project were not implemented. In Ethiopia, the project even lost its main objective, i.e. direct support to production. Other major integrated rural development project components such as the production and supply of improved seeds, agricultural credit to farmers and livestock development performed unsatisfactorily on the whole (Ethiopia, Rwanda, Tunisia Agro-Complex). 3.2 Operational Performance 3.2.1 All the projects were implemented behind schedule. Some of the reasons are given above, but there are more fundamental ones: the Bank seems to have had little participation in the initial project design. The technician’s vision also prevailed, limiting the beneficiaries to a passive “target-group” role. In projects where the FAO and IFAD were involved, it was as if the Bank delegated the determining of production objectives to those Institutions, while it concentrated on “fields of expertise” and related infrastructure, generally physical. Mostly, except in the case of the El-Beheira Rural Project (Egypt), multiple-component integrated projects recorded unsatisfactory operational performance (e.g. Guinea-Bissau, Ethiopia). In contrast, projects with fewer objectives produced tangible results (Madagascar, Senegal), although the decisive impact on poverty cannot be confirmed. Project operational performance may be classified into four categories: very satisfactory, satisfactory, unsatisfactory and very unsatisfactory projects. 3.2.2 The only operation that may be classified as very satisfactory is the Tsiribihina Irrigation Project (Madagascar). No major modifications were made during the implementation phase and the delay recorded with regard to the civil engineering works are attributable to the damage caused by the 1991 cyclone and social disturbances between 1991 and 1993. The key objectives, of which the cultivation of rice twice yearly, installation of irrigation facilities planned for and settlement of modern and traditional farmers, were

6

attained. The project went as far as building an access road linking the pilot centre to the village where traditional farmers live even though the project had no such component. 3.2.3 The El-Beheira (Egypt) and the North-East Senegal projects come under the second category. The Egypt project promoted land development through intensification of agricultural infrastructure. It is true that performance could have been better on account of the high investment; nonetheless, output increased considerably (10% to 30%) compared with traditional agriculture. Above all, the project trained 10 000 rural women from 110 villages. Furthermore, it developed small-sale productive projects initially not planned, thus providing employment to young school leavers. 3.2.4 The Bush Fire Control Project in North-East Senegal is a success story in terms of social mobilization around environmental protection activities. Rural awareness was so much that the number of bush fire control committees increased from 200 at appraisal to 600. The availability of labour and employment prospects that the project offered were a powerful incentive. The project had a bush fire control component and another on community reforestation through planting of the gum tree (Acacia Senegal). The latter component was less successful owing to social inertia and, especially, the late discovery by farmers of the profitability of arabic gum.5 3.2.5 The two projects implemented in Tunisia, the one in Rwanda (Byumba) and another in CAR (M’Baiki Berbeti) come under the third category of fairly satisfactory projects. These projects share in common the existence of several components, generally unclearly identified or inadequately prepared at appraisal, leading in several cases to oversizing and significant changes to certain components during the implementation phase. Inset 3.2: Limited Beneficiary Involvement The case of the DIM Project in Tunisia is a good illustration: the project contained a quitch-grass eradication component because such operations were carried out in neighbouring regions. However, the component did not meet the expectations of the project’s target groups. Hence, of the 13 850 ha set aside, operations were carried out on only 220 ha. Furthermore, components on orchard plantations, stable or surface well construction encountered serious difficulties due to land and credit management problems, the poor perception by farmers of the real needs of such investments and likely benefits. Indeed, the absence of environmental study which would have permitted identification of constraints and issues specific to the project area explain the incorrect projections and lack of enthusiasm on the part of the target population for certain project components.

Particularly in the implementation phase, the DIM Project in Tunisia and the M’Baiki-Berberati Project

in CAR encountered the same problem of the people’s limited involvement. Although the components identified were relevant at appraisal, particularly vis-à-vis the project objective to integrate the development of export crops into larger programmes whose aim went beyond increasing the income of rural dwellers to incorporate general improvement of their living standard, it is doubtful how such an objective could have been attained without the full participation of the farmers themselves. Excessive centralization under the project adversely affected crop production; generally, farmers were skeptical of the need to use fertilizers which they found to be too expensive. 3.2.6 The poor performance of other projects (Guinea-Bissau, Ethiopia, Malawi) is attributable to an initial design laden with complex components, against the backdrop of the low managerial capacity, and inadequate cooperation among various project operators during the implementation phase. The very long implementation period (average 11 years) is symptomatic of the difficulties that these projects faced. A case in point is the Sidamo-Gamo- 5 Actually, the farmers hardly mastered harvest techniques because, traditionally, that activity was carried out by the Moors during transhumance.

7

Gofa Project in Ethiopia whose aim was to promote food self-sufficiency and increase employment possibilities in the rural sector. On hindsight, the development of farmers’ cooperatives that were meant to supervise credit to farmers appeared to work against the attainment of that objective. Credit operations directly funded by the project during the first three years were suspended following the dissolution of cooperatives (the only means through which credit reimbursement was made). The management of multiple-component projects became difficult as they were adversely affected by omissions that paralyzed the adequate implementation of certain components,6 and especially the nearly insurmountable coordination problems due to insufficient information dissemination, itself caused by the compartmentalization of various departments and conflict of authority. In addition, there were inflationary tensions. Under these conditions, it is difficult to see how these projects could effectively contribute to improving the living standard of the target population as envisaged at appraisal. 3.3 Institutional Performance 3.3.1 Project institutional framework differs from project to project. However, irrespective of the degree of complexity of project set up, all come under the supervision of the Ministry of Agriculture or Rural Development. That notwithstanding, projects surveyed may be put in three categories: (i) those supervised by ad hoc implementation units set up specifically to manage them (Madagascar, Rwanda, Senegal); (ii) those that are closely attached to the supervisory authority (Ethiopia, Guinea-Bissau, Egypt, Malawi); and (iii) those managed by existing public agencies (CAR and the two projects in Tunisia). Is there a link between the management or supervisory system of these categories and the institutional performance of projects? We could answer in the affirmative. The institutional performance of projects seems very satisfactory for Madagascar, Senegal and Egypt, fairly satisfactory for Rwanda, Tunisia-DIM, Tunisia Agro-Complex, and poor for Ethiopia, Guinea-Bissau and Malawi. 3.3.2 There are similarities between the legal framework of the Madagascar and Senegal projects. Both were implemented by autonomous units within the central directorates of the supervisory authority. In the case of Madagascar, the agency in-charge was the directorate of rural engineering, and in that of Senegal, the directorate of water and forest resources. This arrangement gave a more technical than policy and strategic perspective to project management. As for the Egypt Project, although highly dependent on the supervisory authority through the project steering committee headed by the Minister’s adviser, the autonomy of the management unit and the smooth coordination between the two levels contributed significantly to its satisfactory performance. 3.3.3 The Madagascar and Senegal projects are similar in yet another way: both concentrated on very few objectives, even if the aims were not the same. Although the Egypt project had several components and for that reason was more complex, it made ample use of sub-contracting. For the Egypt and Senegal projects, the national staff were competent and there was no need to use external technical assistance 7. Training was therefore limited. In sum, the decentralization of responsibility during the implementation phase of the three

6 For instance, the ‘‘health component’’ under which health centres and maternity wings were built, made no provisions for appropriate medical equipment. After completion, the Ministry of Health refused the handing over until the project found funds to procure necessary equipment. 7 The North-East Senegal Project was one of the rare Bank projects in Africa south of the Sahara solely designed and implemented by nationals. The success value is higher still since the project was openly welcome by the rural dwellers.

8

projects discussed was primordial and determinant of their success. In a way, the attachment of the El-Beheira Project with its supervisory authority was more a reflection of political will, further demonstrated through the increase in the level of Government contribution above the estimate set during the first project appraisal. 3.3.4 Then come the project category with unsatisfactory institutional performance, namely the two Tunisia projects, and those in CAR and Rwanda. The first three were implemented by financially and administratively autonomous legal and corporate bodies already in place. The evaluation highlights the fact that this umbilical link had more disadvantages than advantages. For its part, the Byumba Project in Rwanda was affected mostly by several changes in the internal structure and environmental pressure. 3.3.5 Institutional changes within the management agency had a direct impact on projects. A case in point is the Tunisia-DIM Project. Initially managed by the Souassi Development Agency (Office de mise en valeur de Souassi, OMVS), that Body subsequently came under the control of the Regional Agricultural Development Commissariat (Commissariat régional au développement agricole, CRDA). The poor performance of the Unit set up following the reorganization was manifest with regard to sensitizing project beneficiaries and supervising agricultural credit. Yet, these very components were essential for attaining the income improvement objective. Although the management audit function was adequately fulfilled, the follow-up/evaluation was limited to a physical and financial inventory without any environmental assessment or impact analysis. The project took little initiative on rural extension and farmers’ taking over project activities. Indeed, since it was not so programmed, the CRDA was ill-adapted to project requirements. Inset 3.3: Institutional Constraints of the M’Baiki-Berberati Project (CAR) Structural inertia also afflicted the M’Baiki-Bererati Project in CAR, implemented by the Coffee Zone Development Agency (agence de développement de la zone caféière, ADECAF). As its name implies, the initial role of the Agency was to develop coffee cultivation and it was organized for that purpose. Following its restructuring five years after entry into force of the projet, ADECAF’s brief was extended to include improvement of food crops, promotion of cooperatives and enhancement of road and social infrastructure. The reorganization whose explicit aim was devolution, if not decentralization, came up against a slow change of mentality. The grassroot staff responsible for extension were unable to rid themselves of their “crop line” approach to embrace a more open and diversified one based on participatory development, through which food crops would have been given their rightful place instead of being considered as support crops. Such a capacity to adapt was particularly necessary at a time when coffee prices were at their lowest, and the CAISTAB, a major link in the coffee sector, did not play its role due to lack of liquidity, causing farmers’ disaffection for coffee farming. 3.3.6 Projects with poor institutional performance share a common denominator: a profusion of administrative hierarchies which results in excessive decentralization, a very high degree of sensitivity to all policy and administrative changes and, paradoxically, individual unaccountability. In the case of the Byumba Project (Rwanda), the absence of an effective extension system was mostly due to the poor coordination of the six technical components that made up the project’s internal structure and unstable supervision. Lack of extension topics given as reason in the completion report is more of an epiphenomenon. 3.3.7 Although it may appear premature to draw lessons on the institutional performance of projects reviewed, the following comments can be made:

9

- the first project category (Madagascar, Senegal, Egypt) stand out with the existence of proven human resources, a high degree of management autonomy of projects vis-à-vis the supervisory authority and use of sub-contracting.

- the project category with unsatisfactory institutional performance (Tunisia-DIM,

Tunisia Agro-Complex, CAR, Rwanda) seem to have been brought low by the “structural pull” of public agencies which lacked the capacity to adapt to specific project demands. The Rwanda Project should be treated separately: it is highly probable that the project’s distributive capacities whipped up bitter competition and rivalry over management of various components 8.

- under the category of projects with inadequate institutional performance (Guinea-

Bissau, Ethiopia and Malawi), matters were conducted as if the potential functions were more important than the obvious ones, from the sociological perspective. Under those circumstances, although they lost their initial objective, it is easily understandable why the projects took so long to implement. If that assumption were right, the three projects would be among the least performing at the operational, economic and social level.

3.4 Economic Performance

Table : Presentation of the Economic Rate of Return of Projects Reviewed

(Appraisal/Completion)

COUNTRY/PROJECTS

ERR AT APPRAISAL

ERR AT COMPLETION

Tunisia-DIM 12% 04%Madagascar/Tsiribihina 10.55% 17.55%Guinea-Bissau/Tombali-rice 14.45% * Not calculatedEthiopia/Sidamo-Gafo-Goya 10.30% * Not calculatedMalawi/Blantyre-Shire 27% 20.6%Egypt/El-Beheira 18.1% 26%CAR/M’Baiki-Berberati 23% 0.22%Rwanda/Byumba 24% 6.1%Tunisia/Ghezala-Mateur 15% 10%Senegal/Bush Fire Control 17.05% 28.52%

* Rates given at appraisal; the ERR at completion are not available. 3.4.1 The above table shows that the economic rates of return calculated at appraisal were generally fairly optimistic. For Guinea-Bissau and Ethiopia, the impossibility of calculating the ERR at project completion due to lack of a reliable data base would mean that the projects failed because of inadequate follow-up. The absence of reliable data causes some confusion between the “no-project” and “with project” situation. Although the cost of certain projects (e.g. Tunisia-DIM and Byumba-Rwanda), the benefits are not so obvious, necessitating several assumptions regarding the irrigated area, crops cultivated and output. With the

8 The possibilites for training abroad which the project offers as well as a salary 30% to 40% above that of the civil service certainly contributed to the frequent change of national staff. Within the country’s context, both would be seen as incentives.

10

exception of a few projects (Egypt, Madagascar), the review highlights the relatively poor quality of project at portfolio entry: non-existence of the logical framework, no risk analysis vis-à-vis the project objectives, no analysis of probabilities of attaining project objectives, no relevant environmental impact assessment, inadequate analysis of institutional capacities, etc. 3.4.2 The optimism of the economic rates of return are perceptible in the calculation of assumptions. Whereas acceptable data on investment cost, renewal cost and operating expenses could be obtained in certain cases, production data was often more or less incorrect. It was also noted that assumptions for calculating the ERR rarely took into account the degree of a project’s sensitivity to its environment (e.g. El-Beheira/Egypt), nor were the price projections sound. The gap between the economic rate of return at appraisal and completion was particularly high for projects with a cash crop component. Indeed, production-based income calculations used world prices as reference, at the time on the high side. 3.4.3 The error margins are so wide that the extension of project duration becomes a cost-increase factor, likely to reduce return. Certain countries went through fairly high inflationary tension (Egypt, Malawi). Such investment options as the intensification of agricultural mechanization or fertilizer importation had a negative impact on the ERR at project completion. 3.4.4 Non-availability of data may be considered as an obvious sign of a project’s poor economic performance as in the case of Guinea-Bissau, Ethiopia and Malawi. Concerning Malawi, figures given should be considerably reduced 9. This fact which reflects the poor operational performance of the projects mentioned renders the ERR calculation, where it is made at all, approximative and random. Under these conditions, the risk exists of substituting projections for achievements. 3.4.5 In any case, the projects that strived to attain their objectives, especially in connection with agricultural and livestock production, had a direct positive impact on farmers’ income, thus confirming the importance of incorporating such a variable into poverty reduction projects. Consequently, the emphasis on other components of rural development projects (e.g. physical achievements) should be reassessed. The economic performance of the Tsiribihina Project (Madagascar) shows to what extent the income variable is significant. 3.5 Social Performance 3.5.1 It is difficult to evaluate the social performance of projects. There is some confusion in the completion reports between the economic and social performance when the latter is not included in the general project impact. Some projects skip the social performance while others explicitly state how difficult it is to quantify the social impact of projects. Actually, the problem resides in the methodological approach adopted at appraisal. Obviously, the social objectives were only stated in vague terms. Specific indicators were only provided in connection with production development and physical infrastructure. Enough details were not given on the variables underlying those indicators. Yet, such details are essential in evaluating the impact of any poverty reduction project. Furthermore, the appraisal was not forthcoming about the means to be used in verifying the stated social objectives, resulting in less attention and focus on project qualitative performance (quality being another key dimension in poverty reduction). Other major variables such as the socio-economic 9 Indeed, the completion report acknowledges that “consistent data on actual expenditure could neither be obtained from the Malawi authorities nor the project management … ”.

11

categories, gender, income and living standard were not explicitly stated, even though they could have permitted measurement, during project follow-up and completion, of the social effectiveness in the light of certain relevant indicators, of which food security, employment, health, training, access to credit, rural exodus, etc. Nonetheless, project evaluation has highlighted other equally important trends. 3.5.2 The food security indicators are identical in practically all rural development projects. A review shows that good performance was recorded for Madagascar, Egypt, Rwanda, Senegal and, to some extent, Tunisia’s Agro-Complex. Except for the Malawi Project whose performance is poor, the results were unsatisfactory for all the others. Where the performance was satisfactory (e.g. Madagascar and Egypt), focus was on high-yield food crops (rice, sorghum, sweet potato). For its part, the Senegal Project aimed at improving fodder for livestock. Therefore, the food security indicator should be considered from the dairy and meat production angle. It is on the same basis that Tunisia’s Agro-Complex Project is rated satisfactory. The project participated considerably in increasing basic food products, taking into account its function of regulating the egg and fruit market even beyond the project zone during certain periods of the year. For those projects that performed unsatisfactorily, reasons have already been given. In CAR’s case, for instance, coffee cultivation may have performed satisfactorily without the same being necessarily the case for food crops. To sum up, the success of a cash-crop based rural development project is relative. 3.5.3 Mention should also be made of the employment indicator. The Tunisia, Madagascar, Rwanda and Senegal projects performed best. However, except for Madagascar, the jobs were mostly temporary, therefore precarious. Impact on women and youths was only significant in Tunisia’s Agro-Complex Project and, to some extent, the Madagascar Project, even though the target beneficiaries could not be classified under the very poor category.10 3.5.4 Health-wise, of the ten projects considered, seven had a health component. The completion reports give no clue regarding the quality of services or care of the indigent. It would have been of interest if the following three indicators were considered: access to basic health facilities, access to primary care and access to curative care. The Tunisia-DIM Project, the El-Beheira Project (Egypt) and the Byumba Project (Rwanda) appear to have performed best in that regard. As for the Tunisian Project, the health centre provided services to the surrounding population. The Rwanda Project also performed creditably. The two health centres provide curative care to between 25% and 40% of the target groups, while the vaccination rate for children aged 0 to 2 increased from 30% before the project to 62% after the project. In other countries (Ethiopia, Malawi, CAR) the quality of services did not match the physical infrastructure set up. Earlier on, we demonstrated to what extent the Malawi case makes a caricature of this phenomenon! 3.5.5 Seven of the projects sampled contained a training component. The Tunisia-DIM, El-Beheira (Egypt) and Byumba (Rwanda) projects performed satisfactorily in that regard.11 Here again, the Tunisia Project stood out, given the focus on improving working conditions 10 In the Tunisia Agro-Complex project, 40% of agricultural jobs, although casual, went to women, the same as 80% of temporay and permanent poultry farming activities. The project’s success generated other activities, offering the possibility to modern farmers’ wives to set up such income generating activities as shelling, livestock farming and rain-fed agriculture. 11 In this case too, with a view to poverty reduction, such an indicator that highly contributes to capacity enhancement should be evaluated not on the basis of the project’s explicit objectives but in relation to the beneficiaries, therefore necessarily to external training.

12

by setting up a boarding facility especially for pupils from far towns, especially girls. Indeed the project enrolled more pupils than estimated at appraisal, guaranteeing a high teacher/pupil ratio at the same time. The El-Beheira Project focused on providing vocational training to rural women. The campaigns emphasized child care, food processing and conservation, literacy and family planning. Throughout the project, nearly 10 000 rural women were trained in about 110 villages. In Rwanda, the situation differed somewhat. The staff did not receive solid training before participating in the project. Continuous training was considered a key project objective and dispensed to all staff. Notwithstanding regular changes, the training component certainly contributed to enhancing the managerial capacity of project management. Farmers also benefitted through demonstrations in model farmers’ fields. Other categories such as cooperatives and livestock farmers who were to receive livestock loans were also trained. For other projects, the impact was either induced through extension action (Senegal, Madagascar) or marginal (Ethiopia, Guinea-Bissau). 3.5.6 In considering the “living standard” variable, we view as relevant those indicators on access to facilities (potable water, housing, transport and household equipment). Especially for housing and household equipment, they highlight the behaviour of the income variable, and necessarily have to be taken into account when evaluating the impact of poverty reduction projects. However, given the weak documentary base of projects reviewed, there is need for caution in interpreting performance in that regard. That notwithstanding, accommodating measures on rural development projects (road and access road improvement) constitute a top component in practically all projects considered. Performance varies from project to project: very satisfactory for the Tunisia-DIM Project where projections were exceeded, and the cost of untarred access roads was not only lower vis-à-vis the initial estimates but also in comparison with rural roads developed in the same region. The Madagascar Project attained all its objectives: the access road rehabilitation and infrastructure construction works were fully implemented despite the damage and delay caused by the passage of Cyclone Cynthia. 3.5.7 If there is a component of the Tombali Rice Project (Guinea-Bissau) that could claim some success, it would well be in connection with the construction of access roads in the Tombali Region. Although provision was made to build 4 km of access road, the project ended up with 55 km of road at various dam sites, in addition to supporting the maintenance of other access roads. The positive impact of these access roads is confirmed with the increase in communication and trade among the villages within the Region. However, should UA 4.45 million have been spent to obtain that result? 12 Regarding road infrastructure, the Blantyre Shire Project (Malawi) attained almost all its objectives (88%), although on the alter of an overdrawn implementation schedule and cost overrun (165%). Therefore, we are right to wonder whether the socio-economic benefits justify the investment made. Nothing could be less certain. Given the fact that such achievements basically make up the accommodating measures on which the Bank seems to base “its expertise”, even when successfully implemented, the scope of such measures largely depends on the effectiveness of other components. We saw to what extent certain projects have under-performed! Inset 3.5: Rural Credit In projects with an agricultural credit component, it is important to review the performance of the access to credit indicator in its organizational, beneficiary, gender and sustainability aspects. With the exception of Madagascar, the agricultural credit component in the six project samples ended up as failures. Although the reasons varied, the projects tended to grant loans directly. In Madagascar’s case, the project put at the disposal of modern farmers developed land and working capital for the first crop year. In view of the incentive nature of the 12 This is the actual amount of expenditure on ADF financing only. The total project cost stands at UA 11.47 million!

13

operation, the farmer would reimburse the loan with 0% interest over a ten-year period. Hence, this is not the traditional credit operation but a subvention. Paradoxically, the Ethiopia Project recorded the most reasonable performance. Within the framework of direct measures in support of production, the project had provided for an agricultural credit fund of UA 2.37 million. These funds were not directly managed by the project but on-lent to a development bank (AIDB). A subsidiary loan agreement stated that the bank could on-lend the credit fund to cooperatives at a rate not above 5%; it would then be up to the cooperative to grant loans to its members and recover the funds. Although the recovery rate is lower (69%) than the acceptable level for the viability of credit operations, the performance is better than those for the Tunisia-DIM (24%), Rwanda (44%) and CAR (50%) projects, not to mention Malawi where no figures are available and where loans were dispensed by extension workers due, it was claimed, to the absence of credit officers. Unlike the Ethiopia Project, agricultural credit in these four countries was directly administered by the projects. Although information available on these credit operations does not allow for precise identification of the type of beneficiaries, for a certainty, the allocation of funds to nearly exclusive agricultural production operations can only have an indirect, perhaps marginal, benefit to rural women. 3.5.8 Lastly, the social performance of rural development projects should also be analyzed from the local development and impact on rural exodus perspectives. Insofar as certain projects were presented at appraisal as geographic balance measures (e.g. Tunisia, Malawi, Guinea-Bissau), it is only right to ask to what extent they contributed to social mobilization and halted rural exodus. With regard to local development, the performance of projects in Senegal, Madagascar and Egypt may be qualified as satisfactory, reflecting the people’s support to project objectives. In Senegal, audio-visuals targeting all communities, including women, had a positive impact not only on bush fire control but also on health education and agro-sylvo-pastoral activities. It will be recalled that the project dynamism permitted the establishment of 600 village committees instead of 200 estimated at appraisal. Pasture has become so abundant in recent years that the region now provides refuge for transhumant livestock from neighbouring regions. Added to that, arabic gum plantations have raised much expectation. The least that can be said is that the project contributed to decelerating rural exodus. By obtaining improved yield through maize intercropping, farmers obtained additional income and reasons to persist. In Egypt, the development of small-scale productive projects (poultry, rabbit, bee and fish) and village industrialization (rice miller, cereal mills, fodder processing machines), although not mentioned in the initial project design became an essential component. These small-scale projects met the high demand from women and young school-leavers who, in addition, received special training. In Madagascar, the dynamics generated by the coexistence of two kinds of farmers (modern and traditional) promoted the sharing of technical knowledge and improved the general outlook for the zone. Output per hectare, nearly identical for the two types of farmers, reflects the quality of extension provided. Thanks to the project, the Monambolo zone which suffered chronic shortages of rice in the past began to receive steady supplies. Concerning other projects, although their performance varies from country to country, none seems to have had any tangible impact on local development and rural exodus. The two projects that would have fulfilled these conditions (Byumba and Tunisia-DIM) produced good results rather late (the Rwanda Project), but with no sustainability guarantee. Lastly, notwithstanding its having achieved significant socio-economic infrastructure, the Tunisia-DIM Project was unable to mobilize beneficiaries in support of the works. Therefore, the issue of appropriation remains. 3.6 Environmental Performance 3.6.1 In nearly all projects reviewed, environmental performance was only reconsidered on project completion, partly indicating the low consideration given to ecological issues during the eighties (project appraisal period), and the trend towards taking those issues into account,

14

given their close link with sustainable development13. If the environmental performance were judged on the basis of the negative impacts, attenuation measures and positive impacts, then the rating would be very satisfactory for the North-East Senegal Project, satisfactory for the El-Beheira Project (Egypt), the Tsirihibina Project (Madagascar), the Agro-Complex Project (Tunisia) and the Byumba Project (Rwanda). With the exception of the Tunisia-DIM Project that would be difficult to rate due to lack of information, the environmental performance of other projects is commensurate with their general impact – nearly imperceptible and zero in their impact area particularly with regard to production, direct or indirect activities. In other words, based on the very modest results recorded, there is hardly any difference between the pre- and post-project situation. 3.6.2 Factors with a negative effect on the environment were mostly noted in projects implemented in North Africa, attributable to agricultural intensification following the use of pesticide and fertilizers with harmful consequences. Furthermore, intensive livestock farming components (cattle and poultry farming) certainly have an impact on the environment, attributable to the storage of manure and droppings. At another level, other exogenous factors such as drought and cyclones hampered or delayed the results of some projects (Tunisia, CAR, Madagascar). Lastly, in the case of the Rwanda Project, on account of the population density, any project implemented has to strike a precarious balance between humans and their environment.

Inset 3.6: North-East Senegal Project – The Impact of an Environmental Project on Poverty The North-East Senegal Bush Fire Control Project is striking in terms of the environmental components. The long drought which ravaged Senegal in the seventies made spontaneous bush fires particularly devastating. The ecological, economic and social disruptions mostly resulted in the loss of biodiversity, rural exodus and impoverishment of the people. In the face of that situation, the implementation of a bush fire control project and the initiation of a reforestation plan with the active participation of rural communities had a positive impact on the region. Considerable results, full of prospects, have been obtained: the sylvo-pastoral landscape has been gradually restored. The herd is multiplying and the pasture can bear an increase in the carrying capacity of ecosystems. Fires rarely burn village plantations during the dry season (prohibition of grazing therein protects them from straying animals and other herbivores). The gum-tree (Acacia Senegal) stock, much of which perished in past bush fires, is increasing through combined spontaneous and directed action. In the next five years, gum production is expected to rise to past levels, generating substantial income for the people. At another level, given the fact that nine out of Senegal’s ten regions are short of wood fuel, the importance of diversifying reforestation species that take account of the people’s nascent needs is obvious. Lastly, the improvement of the vegetation cover in the region is a long-term factor for improving the weather and rainfall regime. The project’s major risk is that it could be a victim of its success, drawing into the region a large convergence of livestock not only from surrounding areas but also from neighbouring countries. 3.6.3 The attenuation measures taken had a positive impact, especially in the El-Beheira Project (Egypt) and the Agro-Complex Project in Tunisia. Drainage improvement through sub-grading, gypsum spreading and transformation of straw residue into compost with a view to producing organic fertilizer are considered to be ecologically positive. Furthermore, the Agro-Complex Project (Tunisia) recuperated lands previously set aside for cereals and fodder crops to implement the 907 ha of irrigated crops. No salinity or soil deterioration was noted following the use of dam water. Even the soil structure and texture improved.

13 Eight of the ten project appraisal reports make no mention of the environmental impacts of those projects. The Tunisia-DIM Project only mentions them to underscore the positive impact on the environment. In sum, only the Madagascar and Senegal projects adequately considered the environmental issue. In the latter case, that issue was the very raison d’etre of the project.

15

3.6.4 Measures taken to create and consolidate communal reforestation in Rwanda aimed at putting at the people’s disposal firewood and working wood through rational exploitation; other measures to repair the damage caused by the cyclone passage in Madagascar also considerably improved the people’s nutrition and health conditions. 4. SUSTAINABILITY

Several factors contribute to project sustainability. Where those are not confounded with the performance already described, the factors become interdependent and tie in together. Obviously, therefore, poor performing projects stand very little chance of sustainability. Having said that, the good performance recorded here and there is not necessarily guarantee for sustainability. Factors that affect the viability of projects here reviewed are analyzed at three levels: the institutional framework, capacity strengthening and sustainability of social results. 4.1 The Institutional Framework 4.1.1 A project’s link to the economic, administrative and political environment appears to be determinant. In that regard, the El-Beheira (Egypt), Agro-Complex (Tunisia), Tsiribihina (Madagascar) and North-East Senegal projects met the sustainability criteria, all of them having been implemented within the framework of national plans and regional development programmes. The programme-approach underlying the project design highly contributed to experience capitalization and institutional consistency. 4.1.2 Sustainable projects are those that are solidly backed by RMCs. Demonstrating its political will to support the El-Beheira Project, the Egyptian Government increased its contribution beyond the appraisal estimates. The project had considerably administrative and financial autonomy as did the Senegal and Madagascar projects. Since Senegal’s decentralization edict enhances community participation, it too will guarantee the sustainability of project achievements. 4.1.3 The quality of the physical and social infrastructure set up under certain projects encourages the managers to run a maintenance programme. In Tunisia, the guarantee that the Agro-Complex facilities will be maintained is given yearly through the systematic inclusion of a “maintenance” chapter in the budget. The project will also bear the recurrent expenditure on infrastructure. 4.1.4 Political instability and administrative reorganization that usually comes in the wake of frequent staff changes are ever-present challenges to the sustainability of certain projects even when such projects were implemented within the context of regional development programmes (Ethiopia, CAR, Malawi). Furthermore, the price trend of some agricultural products depends on the international situation (e.g. coffee, cotton, groundnut). In the case of CAR, Ethiopia and Malawi, project success depended to some extent on crop performance. Moreover, although the Byumba Project was implemented within the framework of the Rwandan Government’s second five-year plan (1976-1981), it was significantly modified from year to year owing to inadequate financial resources. In the face of the country’s environmental and demographic pressures and notwithstanding the positive results, the experience would hopefully have fulfilled the function of a “pilot project”.

16

4.2 Human Resource Capacity Strengthening 4.2.1 Some projects (e.g. the two Tunisia projects, the El-Beheira Project [Egypt] and North-East Senegal Project) already had intrinsic endogenous capacities for the preparation, implementation and follow-up of operations. Two other projects (Madagascar and Rwanda) used local expertise or structures to train staff. The performance of other projects (Guinea-Bissau, Ethiopia, Malawi, CAR), including the level of physical achievement, was most inadequate and show no promise of sustainability. The first category stands out through the active role played by national managers in the design and preparation process, without using foreign technical assistance during the implementation phase. Internal training was rather weak, with most of the attention focusing on the beneficiaries. 4.2.2 In Madagascar and Rwanda, project staff participated in several refresher courses and technical assistance played a significant role by offering on-the-job and field training. Project implementation units adequately played their role especially in the case of the Madagascar Project where management staff of the unit had extensive experience on Bank-funded projects (Rice I and Rice II). The main challenge facing the Madagascar Project is to maintain the same performance level and guarantee the sustainable management of social infrastructure (dispensary, community pharmacy, school). Indeed, the structure set up in the form of a users’ association, is still recent and inexperienced to satisfactorily manage all activities. The Byumba Project made extensive use of local labour to build roads, access roads and small-scale rural water supply works. However, the road network maintenance operation which should have been the responsibility of the communes is not well organized and cost recovery is still a problem, especially with regard to water supply. 4.2.3 Lastly, the projects in the third category (highly inadequate capacity) have several major weaknesses, especially: (i) lack of skilled staff associated with low technical assistance (Guinea-Bissau, CAR, Ethiopia); (ii) inadequate or unworkable training component (Guinea-Bissau, Malawi); (iii) the high uncertainty surrounding project infrastructure maintenance. In Guinea-Bissau, it became obvious early in the day that the borrower was unable to allocate enough credit for road maintenance beyond the project period. In fact, all project activities ceased on project completion. In Ethiopia, although the project was implemented within the framework of the PDAR III Programme, inadequate budgetary credit, shortage of foreign exchange for procuring equipment and materials, added to much reluctance on the part of farmers to participate, used as they were to receiving seeds free since the start of the project, offer little guarantee for project sustainability. 4.3 Sustainability of Social Results 4.3.1 The sustainability of the social results of projects considered is generally hypothetical and hard to evaluate. The same reasons are responsible for this situation, not the least of which the absence of social indicators during project start-up and lack of impact assessment. No stock is taken during the identification, preparation and appraisal phases. Information gathered is too general, borrowed, indeed anecdotal. “No-project” situations are hardly differentiated from projections. If described, social achievements are generally viewed from the physical infrastructure quantification perspective, whereas the sustainability of social results is basically evaluated on a qualitative scale. For instance, in connection with health, health centre buildings and rehabilitation works are very easily highlighted, but little is said about the provision of services. Generally, aspects related to healthcare were neglected. Having said that, a review of the projects highlights two trends: a project category with

17

relatively sustainable results, although qualified (Tunisia-DIM, Madagascar, Senegal, Egypt); a second project category with many constraints or worse, likely fatal to the sustainability of social results (Rwanda, CAR, Malawi, Ethiopia, Guinea-Bissau, Tunisia Agro-Complex). 4.3.2 The first project category is characterized by greater sensitivity to the social dimensions of the set objectives, and fall within the range of fairly satisfactory performance, backed by a promising environment. The Tunisia-DIM Project set up major social infrastructure even as it maintained a high level of access to social services (e.g. transport problems facing local pupils, reduction of health evacuation and availability of potable water within the reach of households). However, there is also a dark side: the project was weak in terms of beneficiary participation and environmental organization. It was hoped that the second phase of the project would consolidate the achievements by improving the participation of the target population14. In the Madagascar Project, the social results are closely tied with the project production. Their sustainability will also depend on the project’s sustained growth. Since the project’s remarkable success is also attributable to the fact that it is located in a zone that suffers chronic rice shortage, the elasticity of its potential market is better measured, the same as the sustainability of the project’s direct and indirect impact. Concerning the North-East Senegal Project, the social impacts already highlighted by the beneficiaries themselves (employment opportunities, construction of houses, constitution of a cattle rearing core, contracting of marriages, etc.), tied up to the regeneration of the ecosystems, were obtained within the context of an ecologically sustainable approach. This satisfactory performance will very likely be strengthened by the expected income from gum-tree plantations. 4.3.3 In the second project category, a few positive results were recorded although certain factors that call for caution persist. Even though, in the case of the Rwanda Project, the training, health and rural water supply components accounted for only 9.5% of actual expenditure of the Byumba Project, the impact on the living standard of rural dwellers was remarkable: giving 22 000 families in the zone access to potable water and improved healthcare management. However, in the light of the uncertainties already described, the sustainability cannot be guaranteed. For the other three countries (Guinea-Bissau, Ethiopia, Malawi), a review of their sustainability would be superfluous against the background of non-existing tangible social results. Commenting on the sustainability of achievements of the Malawi Project, the completion reports states: “Staff recruitment, provision of means of transport, other equipment and household furniture are among the project’s most important achievements”. It is as if the social results should be judged by the yardstick of its benefits to its key officers. The least that can be said is that that is a far cry from the objective set at appraisal. 5. BANK’S CONTRIBUTION

The Bank neither participated in the identification nor preparation of the projects considered. It contributed mostly to the appraisal and follow-up phases. Although its interpretation of the objectives of rural development projects is basically relevant, the project was designed by the authorities or other partners, especially the FAO. Discussions with Bank experts reveal that to be one of the weaknesses of rural development projects financed by the Institution. In its operating modalities, the Bank gave greater attention to physical 14 Discussions with Bank staff tend to confirm that the second phase of the project points in that direction and that sensitization of the people produced other results, for instance by helping the farmers to overcome their reticence to participate in the eradication of quitch-grass during the first phase.

18

infrastructure, more striking it is true, than on the project quality. It gave little focus to institutional changes even though such changes affected the project life. Given the complexity of rural development projects (necessarily multisectoral), the Bank’s performance may be rated as unsatisfactory, moreso in view of the poverty reduction objective which, besides, cannot be measured with the quantitative criteria yardstick only. 5.1 Project Design and Preparation 5.1.1 Of the ten project reviewed, only one (Tsiribihina, Madagascar) received an identification and preparation mission of only one week. Therefore, project appraisal missions were somewhat circumscribed by plans drawn up by other donors. In any case, the fact remains that where additional studies were indispensable for assessing project feasibility, the Bank did not judge it necessary to undertake them. In like manner, having perhaps only to approve the objectives defined, it was unable to detect when certain projects had gone off course. 5.1.2 The Tombali Rice Project in Guinea-Bissau amply illustrates these shortcomings. The Bank undertook to finance the project, the key component of which involved the resettlement of families that had emigrated to neighbouring countries during the liberation war. In that regard, provision was made for the development of 1800 hectares to contribute to the development of rice farming and the country’s food self-sufficiency. No serious environmental study was carried out to analyze the problem of beneficiary support for the project. The project started without any clear idea on where those the project was designed for lived. 5.1.3 The Tunisia-DIM Project appraised by a Bank team in March 1985 is another example. Several components were either ill-identified or inadequately prepared, resulting in over-sizing and often erroneous projections. Hence, proposals for the eradication of quitch-grass did not meet the farmers’ expectations, hence the 220 ha on which operations were carried out, compared with the estimated 13 850 ha. Due to land tenure and credit management difficulties, agricultural plantations were far below the objectives set at appraisal. Indeed, the project feasibility study most likely recopied the typical operations undertaken in regions bordering the project area, without first carrying out an environmental study that would have identified and helped to better overcome the constraints. The Bank focused more on loan conditions than on the project quality. 5.2 Bank’s Operating Modalities 5.2.1 Although the social objectives of rural development projects are clearly stated, missions comprised agronomists, agro-economists and financial analysts only. Specialists on social and institutional issues were not consulted throughout project preparation and implementation phases. As a result, there were no comprehensive social studies precisely targeting beneficiaries. It was as if the perception of poverty as a purely rural phenomenon was enough to justify all rural development projects. It is true that at the time, for projects under review, the very concept of poverty was very rarely used, substituted with references to improving the living standard of rural dwellers. 5.2.2 This uniform perception of rural poverty resulted in a vague and child-like vision of rural society. Issues related to access to land, gender, generation conflict and social reclassification were overlooked. In like manner, the dominating or adapted ideology of the

19

ruling class at the time moulded certain mentalities so much that resistance to reconversion to the market economy which underpinned the Bank’s rural development projects were under-estimated or not perceived. In Ethiopia, the farmers’ refusal to buy fertilizer put at the project’s disposal was due to the fact that they were used to receiving agricultural input free. The Bouenza Project in Congo encountered similar difficulties. 14

5.2.3 It is equally symptomatic to note that the least performing projects are those that received little Bank supervision. The Tombali Rice Project in Guinea-Bissau received only eight missions over its eleven-year life span, the same as the Sidamo-Gamo-Gofa Project in Ethiopia. The missions themselves comprised no more than one or two persons, and were of short duration. For dragging its feet, the Bank worsened project slippages. 5.2.4 Although the Malawi, CAR and Rwanda projects received more supervision missions, they faced other problems as well. The Blantyre-Shire-Highland Project in Malawi, complex enough in its conception, required supervision in several fields. It was simultaneously implemented in two different regions. The quality and efficiency of the missions suffered because the mission composition often changed and follow-up was limited to visits lasting one day. Such missions were hardly of any use to the project. Furthermore, whereas one of the loan conditions was that the accounts of the two zones were to be kept separate, the Bank insisted that the reimbursement request should be combined. During the reimbursement, the Bank did not indicate the respective share of either zone. The missions should have been multi-disciplinary in the light of the diversity of project components. 5.2.5 In all, the missions focused more on the physical achievements of projects. The unending review of the list of procurement of goods and services and of the financial statements was the key preoccupation. The internal follow-up mechanisms were rarely operational without the Bank calling for independent auditing or mid-term review. Apart from communication difficulties between the Bank, project managers and the supervisory authorities, fundamental policy issues such as cost recovery on services provided and service quality, measurement of impact of actions taken and their sustainability were not mentioned either during discussions with key project staff or with government authorities. 5.2.6 For projects implemented in Guinea-Bissau, Ethiopia, CAR and Malawi, apart from the reasons evoked, it was as if in the face of project slippage, discouragement gradually took over and led to a careless follow-up. A review of the periodicity of missions, close during the initial years and gradually spaced out, tends to confirm this analysis. The completion reports stress that for the projects reviewed, the borrowers considered the Bank’s reactions late, leading to loss of time and re-issuance of tender notices. 6. TREND OF ONGOING PROJECTS 6.1 Strategic Options 6.1.1 Projects here reviewed were completed between 1994 and 1996. Their reference framework was set a decade earlier. New rural development projects draw inspiration from ADF VII and VIII operational priorities as well as the Bank’s new vision. Whereas past rural development projects were designed to contribute to improving the living standard of rural 14 It is likely that the collective representation inherited from African socialist regimes of the seventies explain, at least partly, the failures recorded in the rural development projects in Ethiopia, Congo (Bouenza) and Guinea-Bissau.

20

dwellers, the new thrusts establish a more direct link between poverty and rural development. Therefore, emphasis is laid on support to agriculture and rural development as “factors essential to production growth and improvement of the living standard of the poor”. Poverty reduction projects initiated by the Bank in recent years now fall within that purview. The take-off of the AMINA programme also reflects the ADF’s willingness to promote microfinance intermediation in the rural area. 6.1.2 Furthermore, the new vision states that the Bank will “adopt a participatory approach throughout the project cycle to ensure that study conception meets the needs of the target beneficiaries”, clearly laying hands on the major weaknesses that stand out among the projects reviewed. Discussions with Bank experts as well as review of recent documentation comprising appraisal and supervision reports, mid-term and portfolio performance reviews show that reasonable results could be obtained where the participatory approach is patently followed. Without being exhaustive, an evaluation of some ongoing projects point to the impact of such performance on the socio-economically underprivileged, women, credit and sustainability. 6.2 The Operational Level 6.2.1 In such projects as the Selingue (Mali), Centre (Cote d’Ivoire), Food Security (Chad) or Poverty Reduction (Burundi and Mauritania), the socio-economic categories were well delineated. That marked a major step in the preparation of an indicator matrix for follow-up/evaluation. Beyond effort at identifying target groups, these projects develop a participatory approach by mobilizing socio-professional associations and non-governmental organizations. The Central Cote d’Ivoire Project, for instance, uses the trade council to train and grant credit to small-scale traders and farmers. Chad’s Food Security Project works in the field in close collaboration with village associations mostly comprising women. In both projects, each of which has a credit component, reimbursement is about 100%. That is far from the inadequate performance recorded under the direct credit extension system or by the major banks in projects reviewed. One lesson should be drawn: the issue of recovery is more closely linked with the behaviour of the person to whom the credit is extended and the organizational system of the structure granting the credit. Other examples sociologically founded tend to confirm that in the African context, joint credit has proved its mettle. 6.2.2 Considerations on gender issues and project impact on women received greater attention especially for environmental projects. Ongoing projects such as those in Burkina Faso (Bazega Land Management) and Benin (Fire Wood Plantation in Southern Benin) highlight the issues related to access to land and the positive impact of involving women in income generating activities. Studies conducted by the GTZ showed, using the Sahel rural market experiment (Burkina Faso, Mali, Niger, Senegal) that making certain socio-professional groups and village committees accountable for managing natural resources contributes to improving income from forest products (wood, charcoal, shear butter, etc.) and enhances natural resources protection. Confirming other researches carried out by WHO and UNICEF, the studies also revealed that the involvement of women in developing the Gonze natural forest in Burkina Faso (thus enabling them to increase their income) had a net positive impact on children’s education 15. Ongoing rural development and poverty reduction projects have moved from the nearly undifferentiated beneficiary notion as perceived in past projects.

15 GTZ/PED-SAHEL: Women’s Role in the Gonze Natural Forest Management, Burkina Faso, 1997.

21

6.2.3 These positive thrusts which can also be seen in supervision summaries, project indicators and the project design and implementation matrix (logical framework) are a contrast to the projects here reviewed. However, obvious inadequacies still exist. Needless to mention the Bank’s persisting weaknesses in project preparation. Whereas the project objectives outline the socio-economic target categories, the supervision reports and the mid-term reviews turn a blind eye on them when evaluating project impact, mostly likely due to difficulty in setting forth relevant follow-up/evaluation indicators. Furthermore, the quality of social services and their sustainability are still not sufficiently analyzed. Therefore, the Bank’s new follow-up guidelines could become a pure formality. Indeed, the problem lies not in adopting new methods but in abandoning old habits. 7. KEY FACTORS FOR PROJECT SUCCESS

An evaluation of the Bank’s experience in poverty reduction and rural development projects reveals unequal performance depending on project, even as it highlights certain factors for success and failure: 7.1 Rural Development Strategy: the Bank’s rural development strategy and sectoral objectives remain fundamentally relevant. Adherence to the strategy led some countries into tying their project with a more global sectoral development plan and a higher level of financial commitment above what was expected of the RMC at appraisal. 7.2 Project Preparation: one of the reasons for poor performance is the Bank’s limited participation in project preparation. Bank staff interviewed agree that that factor “considerably affected the quality of project appraisal reports”. The Bank did not participate in the identification of practically all the rural development projects reviewed (9/10). Is it by coincidence that the only project that received an identification mission (Tsiribihina) is one of the rare operations that attained and even exceeded the objectives set at appraisal? In any case Bank staff are of the view that “if the ADB does not prepare its projects, it is condemned to following and financing other donors’ ideas. Although ongoing poverty reduction projects in their complexity and multiple components are more like the traditional rural development projects, the Bank’s mission has evolved, although less so in its involvement of staff in project preparation. Staff are unanimous in maintaining that “appraisal missions only take two to three weeks, and they often comprise two Bank experts – clearly not enough”. 7.3 Participative Approach: facts demonstrate that beneficiaries played a marginal role in project implementation and practically no role during the initial phase. Generally, the principle of target group participation in rural development projects remained very abstract. The operational impact was so limited that it was neither embodied in a consistent methodical approach nor through objectively verifiable indicators. Additional comprehensive studies would have enabled the Bank to avoid such errors as financing the emigrant family resettlement component in the Tombali Rice Project (Guinea-Bissau) when no one knew the whereabout of the families in question! Such a study would also have, prior to any commitments, shown reasons for the lack of interest on the part of Tunisian farmers in the eradication of quitch-grass (Tunisia-DIM Project, Phase I). These weaknesses also explain the unsuccessful attempt to introduce sun-flower cultivation into Rwanda. Other comparative studies undertaken by WHO and UNICEF in developing countries tend to confirm that the impact of poverty reduction strategies is greater where emphasis is laid on improving women’s income especially in terms of education, health and child nutrition. See: Women’s Health, Improve the World, Position Paper, WHO, Geneva, 1995.

22

7.4 Beneficiary Support: since success and sustainability of project achievements is closely linked to beneficiary support, beneficiaries should be considered as full-fledged actors. However, it is a matter of course that without the people’s involvement, community participation simply becomes another burden. Successful projects (e.g. Senegal, Madagascar, Egypt) were able to mobilize the people, especially women and youths. In most cases, major issues like the characteristics of beneficiaries, the land tenure systems, labour organization and availability, and women’s role in the agricultural production system did not receive the attention that they merited. Yet, those issues come up again and again. Fortunately, although to a limited degree, the Bank’s new orientations seek to increase beneficiary involvement especially through poverty reduction projects. However, poverty reduction projects remain tainted by the “original sin” of too short a timeframe for project preparation. 7.5 Institutional Issues: the performance of certain projects underscores the importance of institutional issues. Successful projects reviewed either concentrated on limited objectives or activities were implemented within a devolved or decentralized framework, under the supervision of a technical directorate. Apart from the use of sub-contracting, other factors for success include the proven existence of local human resources and reduced sensitivity to the political environment. In contrast, projects implemented within the framework of existing public agencies had more disadvantages than advantages. 7.6 Food Security: with a view to poverty reduction and for all rural development project, this appears to be a decisive indicator in judging project success or failure. Where it was slightly concealed or overshadowed by other components, the results were hardly conclusive even when much had been achieved. 7.7 Rural Credit: the weakness of this component worsened project poor performance. Direct credit management by a given project or its administration by banks or state structures constituted a failure factor. However, although limited, proximity management, especially through farmers’ cooperatives produced far better results. 7.8 Basic Social Services (Health/Education): in projects in which these components succeeded, the quality of services played an enormous role. Even in projects whose key objective of increasing agricultural production was not attained, access to primary healthcare, potable water, women and youth training make the failure of a rural development project very relative. 7.9 Bank Supervision: where enough (average once yearly), supervision missions were factors for success. In contrast, projects that received fewer missions were less successful. However, the number of missions made little difference to multiple-component projects. Mission composition was a limiting factor. 8. CONCLUSION, LESSONS AND RECOMMENDATIONS 8.1 Conclusion 8.1.1 Multiple-component rural development projects of the eighties were difficult to implement, certainly due to institutional weaknesses and the inadequate human resource capacity in borrowing countries. Africa is in no way an exception in that regard. The World Bank encountered similar difficulties in its rural development projects in Haiti, Mexico (PIDER Programme) and Brazil (POLONORESTE Programme). However, comparative

23

studies conducted by the World Bank on rural development projects financed in Africa, Latin America and Asia, involving a wider range of projects tended to confirm, especially with regard to Africa, that the rural development experience had been a failure 16. 8.1.2 However, that fact does not contradict the relevance of the strategy and objectives of rural development projects. In the etymological sense of the term, the aim of such projects was not to combat poverty but to improve the living standard of rural dwellers. Thus, their stated ambition was to contribute to reducing poverty. It is true that for the projects reviewed and whose appraisal reports go back to the early eighties, even though the word “poverty” is not expressly stated, the designers clearly had the concept in mind. 8.1.3 However, the perception of a rural development project as necessarily integrated makes them multisectoral, i.e. comprising several components. Given the rural development strategy, the projects covered a wide range of activities leading to thinning out or dilution of the declared objectives (see the project retrospective matrix, Annex 2). Hence, several projects concentrated on the physical achievements (building construction, rural roads, etc.), abandoning or losing sight of their core objective of poverty reduction and agricultural production. For some years now, the adoption of the logical framework as an essential and compulsory element in project appraisal has improved project consistency. Unfortunately, the definition of indicators remains a problem.17 All international development agencies continue to face challenges in connection with appraisal methodology. 8.2 Lessons 8.2.1 Bank-funded rural development projects were designed as a strategy based on small-scale farmers, especially those who already owned land; hence even in the rural area, the projects could not attack the roots of poverty (par. 2.2.5). 8.2.2 The traditional implementation period for rural development projects proved unrealistic, as a result of which account could not be taken of major issues such as the beneficiary characteristics, land tenure system, organization and availability of labour or women’s role in the agricultural production systems (par. 3.1.1 and 3.2.1). 8.2.3 It is important to involve development beneficiaries in project/programme design, choice of priorities and implementation, thus motivating and making them accountable for the performance (inset 3.2 ). 8.2.4 Although a longer project preparation period is no guarantee for success, it helps to reduce error margins where beneficiary involvement and support are sought and obtained, both areas in which the Bank is weak (inset 3.3). 8.2.5 The institutional set up, granting project implementation units management autonomy from national administrative structures resulted in greater transparency and efficiency (par. 3.3.2).

16 World Bank: “ The World Bank and Rural Development”, 1965-1985, Operations Evaluation Department, Washington 1988. 17 For instance, alcohol abuse as a reason for poverty in Lesotho can be deduced using active participatory research methods, but could not have been determined through formal quantitative surveys.

24

8.2.6 Separation of various operational structures within the same development institution limits communication and sharing of experiences, and is therefore inimical to efficiency (par. 3.2.6). 8.2.7 In view of the land tenure systems and the social division of labour which restricts certain domestic or supporting activities to women (harvest management, backyard livestock farming, small-scale productive projects, etc.), the allocation of agricultural credit almost exclusively to agricultural production can only benefit women indirectly, randomly and marginally (par. 3.2.3). 8.2.8 The people’s interest in a project’s environmental component is closely linked to the obvious short-term benefits that they can gain from it. Sustainability clearly seems to be secondary (par. 4.3.2). 8.2.9 The impact and sustainability of the social results of a poverty reduction project depend largely on the quality of services (par. 4.2 and 4.3). 8.2.10 Foreign technical assistance was practically imperceptible, and focus was on physical achievements and the short-term. The results may not be sustainable. Management procedures were not internalized. Successful projects were those that already had proven local human resources (Egypt, Madagascar, Senegal) (par. 4.2.1 and 4.2.3). 8.1.11 Follow-up/evaluation of rural development projects was rated poor. Evaluation techniques do not permit measurement of changes relating to the impact of rural development project activities on poverty reduction, far less give reasons for such changes (par. 5.2.1). 8.1.12 The evaluation of a project’s impact on poverty reduction is only useful if it is backed by quantitative and qualitative analyses, based on measurable indicators (par. 3.5.1; 3.5.6). 8.3 Recommendations

Rural development projects such as those that the Bank has initiated in recent years (called “Poverty Reduction Projects”) are complex and similar in several ways. Whereas rural development projects suffered from insufficient social and institutional studies, poverty reduction projects could stand the risk of not receiving solid support from rural dwellers. Since the rural sector accounts for a significant portion of the Bank’s portfolio (40%), and given the fact that poverty is more widespread and severe in the rural area and the Bank has made poverty reduction a major challenge to overcome, the Institution should take serious initiatives commensurate with its ambitions. Hence the following recommendations:

Recommendations for the Bank

8.3.1 In the short-term, the Bank should put in place poverty reduction policies and strategies and establish a quality group bringing together experts, especially planners (macro-economists), sociologists/socio-economists, etc. especially for rural development projects and complex poverty reduction projects. 8.3.2 In the medium and long term, however, the solution resides in having a blueprint aimed at providing the Bank with a trans-sectoral poverty reduction programme, commensurate with the nature of the poverty phenomenon. The principal focus would

25

include: support to operational departments in preparing rural development and poverty reduction projects in particular, data collection and processing, dialogue on policies and programmes with RMCs and impact assessment on poverty reduction projects. Perhaps the scope of the AMINA programme should be extended to make micro-finance a means and not an end in poverty control. 8.3.3 The option taken notwithstanding, the Bank should participate more in project design, identification and preparation. The evaluation quality mostly depends on these phases. If it is unable to do so, it should give up on multisectoral projects, except it accepts to invest without having real control over uncertainties. 8.3.4 Comprehensive environmental studies and social analyses are a prerequisite to the appraisal of rural development and poverty reduction projects. Even when studies are carried out by other institutions, the Bank should critically review the terms of reference much more than it used to in the past. Poverty indicators, gender, institutional and land issues should be clearly stated. It seems that the Bank is in the process of setting up a fund to facilitate project preparation. The fund should be made operational as a matter of urgency. 8.3.5 The participatory approach should be strengthened at all stages of the project cycle. Beneficiaries should participate in identifying their problems and prioritizing them. At the intermediate level, the Bank should go beyond merely associating a few NGOs or civil society representatives in poverty reduction projects. Although there are no “ready-made” solutions in that regard, it could draw inspiration from the eco-development approach that the UNCDF has experimented in seven African countries. 8.3.6 For multiple-component projects, supervision missions should be more regular and comprise multi-disciplinary teams. The missions should ensure that focus is maintained on the real project objectives by emphasizing the economic and social dimensions and not only the technical and financial aspects as was too often the case in projects reviewed. 8.3.7 Mid-term reviews should be systematized and considered as an evaluation in the true sense of the word, to permit performance analysis vis-à-vis the objectives and results, and if necessary empower the decision to continue, reorientate or stop the project. 8.3.8 The Bank should define rural development project typology depending on their impact on poverty. It should especially strive to determine the social effectiveness rate (SER) much more than the economic return rate (ERR) when it comes to calculating the social impact of rural development and poverty reduction projects, as well as measure the opportunity costs/benefits. 8.3.9 The performance rating system is highly inadequate since the social reality is far more subtle. A seven-criteria (from 1 to 7) rating would be more relevant. Even so, each criteria should be weighted by being assigned a coefficient ranging from 1 to 3.

Recommendation to Rmcs

8.3.10 The project implementation units should guard against being divided by functional ambiguities regarding supervising and contracting. Instead, they should concentrate on coordination, information, control and follow-up/evaluation, delegating contracting to more qualified parties.

26

8.3.11 The repeated review of the list of goods and services as well as the weakness in internal project follow-up/evaluation should encourage RMCs to request the Bank to systematically organize training for project managers on contract award procedures and follow-up/evaluation indicators prior to project start-up. 8.3.12 Countries lacking enough competent human resources should limit the number of rural development and poverty reduction project components, instead of looking for a hypothetical solution through technical assistance. 8.3.13 Credit recovery is better guaranteed when the beneficiaries are associated with the organization of the structure that grants the credit and when they feel a collective responsibility. In the African context, mutual credit has proven itself. 8.3.14 With a view to taking greater cognizance of gender issues, rural credit should not only be allocated to direct agricultural production activities which, it is commonly known, tends to marginalize women who rarely own land.

Recommendations to Rmcs and the Bank

8.3.15 The Bank should agree with RMCs to limit project extensions to a maximum of two years, and to systematically prepare project completion reports on time to avoid loss of information. If necessary, a second phase could be programmed. 8.3.16 In addition to mail that remain impersonal and often lost in bureaucratic maze, the Bank and member countries would do well to develop personalized and direct contact. To a large extent, the success of rural development projects will depend on the Bank’s and RMCs’ ability to establish close dialogue with all actors. That requires: (i) the use of corresponding local experts to interface with the Bank; the size of the structure, basically light, will depend on the volume of the Bank’s portfolio in the country considered; (ii) the review of rural development and poverty reduction projects prior to start-up by organizing planning-by-objective workshops, bringing together project management staff and members of the steering or supervising committees; and (iii) the systematic delivery of mid-term and project completion reports to RMCs (par. 6.1.2 ) 8.3.17 Lastly, a dynamic system to disseminate useful information on the best practical models for designing rural development and poverty reduction projects should be set up. Such “success stories” as the Tsiribihina/Madagascar and North-East Senegal Bush Fire projects, for instance, should be known better not only by Bank staff but also by the RMCs.

ANNEXES

BANK EXPERIENCE IN POVERTY REDUCTION AND RURAL DEVELOPMENT PROJECTS

Annex I :List of Projects Reviewed

1. Guinea-Bissau : Tombali Rice Development Project 2. Central African Republic : Rural Development of the M’Baiki and Berberati

Coffee Regions

3. Ethiopia : Sidamo - Gamo Gofa Project (PDAR III) 4. Egypt : El-Beheira Rural Development Project 5. Madagascar : Tsiribihina-Manambolo Irrigation Project, Phase I 6. Malawi : Blantyre/Shire Highlands and Mamwera –

Mangochi Rural Development Project 7. Rwanda : Byumba Rural Development Project 8. Senegal : North-East Senegal Bush Fire and Reforestation

Project 9. Tunisia : Mahdia Rural Development Project, Phase 1

10. Tunisia : Ghezala Mateur Agro-Complex Project

BANK EXPERIENCE IN POVERTY REDUCTION AND RURAL DEVELOPMENT PROJECTS

ANNEX II RETROSPECTIVE MATRIX OF PROJECT LOGICAL FRAMEWORK

COUNTRY

Hierarchy of Objectives Objectively Verifiable Indicators Estimates Output

Assumptions/Risks

EGYPT/ EL-BEHEIRA Rural Development

Sectoral Objective: Improve the living standard of rural dwellers by implementing an integrated rural development policy in the El-Beheira Governorate. Specific Project Objectives : 1. Assist farmers in the El-Beheira Governorate to develop their agricultural and livestock production to better meet their food needs. 2. Improve the health conditions in the project zone as well as access roads and agricultural input facilities.

10 agricultural districts comprising 135 villages for an estimated population of 3 million (nearly 500 000 farming families) are the beneficiaries. 1.1 Land development (26 880 ha) 1.2 Cereal net income (in local

currency): 2793 to 5821(Egyptian Pounds)

2.1 Construction of 22 health centres 2.2 Construction of 8 warehouses and fertilizer stores, and 42 veterinary centres 2.3 Improvement of project roads

The project covered ten districts but there is no information on the beneficiaries.

1.1 Implemented and exceeded (125% to 150%) 1.2 3100 to 6512 Egyptian Pounds 2.1 22 Health centres built and equipped 2.2 100% reached 2.3 95% implemented

Matching of project objectives with the farmers’ true capacity to undertake intensive agriculture Availability of irrigation water Considerable budgetary allocation by the Government of Egypt.

ETHIOPIA/PDAR III. Sidamo-Gamo-Gofa Region Agricultural Development

Sectoral Objectives: Attain food self-sufficiency for the regions mentioned. Specific Project Objectives: 1. Strengthen agricultural production; 2. Enhance employment possibilities in

the project area; 3. Develop farmers’ cooperatives

100 000 farming families representing 600 000 people are the beneficiaries. 1.1 Production of 13 000 t of maize; 880

t of coffee; 2076 t of barley; 1285 t of beans.

2.1 Not stated 3.1 Not stated

7776 farming families as beneficiaries. 1.1 Information not available 2.1 Information not available 3.1 Dissolution of cooperatives after three years of project operation.

Country’s political stability and reconciliation of the administrative organization with project requirements. Farmers’ supervision and their support of project objectives Maintain and strengthen farmers’ cooperatives – the only means through which credit gets to the farmers.

GUINEA-BISSAU/Tombali Rice Development

Sectoral Objective: Improve the economic and social well-bring of the Tombali Region by increasing rice production. Specific Project Objectives: 1. Drainage 2. Resettlement of emigrants 3. Provide agricultural support services

Improvement of the living standard through increase in rice production of 13 400 t/annum. 1. 3400 ha of land rehabilitated 2. 720 emigrant families resettled 3. 30 extension officers trained and

equipped.

300 t of rice per annum, i.e. less than 5% 1. 546 ha were rehabilitated, i.e. 16%

of the objective 2. Components abandoned 3. 17 staff recruited and trained

Serious studies precedent to project start-up and assurance of the actual commitment of the Government of Guinea-Bissau Effective management by project staff. Precise knowledge of the country of residence of emigrants and their commitment to come back to the country.

MADAGASCAR/ Tsiribihina Irrigation Development

Sectoral Objective: Improve the living standard of the region’s dwellers through double rice cropping. Specific Project Objectives: 1. Development of 820 ha and related

works 2. Settlement of modern and traditional

farmers.

Attain average rice yield of 3.5 t/ha for both kinds of farmers (modern and traditional). 1.1 Development of 240 ha for single

cropping and 580 ha for double cropping

1.2 Development of an irrigation and drainage network, and access road of 39 km.

2.1 Settlement of 500 traditional and 10 modern farmers.

An agricultural intensification rate of 1.75 was obtained and rice yield per hectare stands at 4.5 t. 1.1 820 ha were double cropped 1.2 All works planned were

implemented and 41 km instead of 39 km of access roads were built

2.1 692 traditional farmers were set up instead of the estimated 500, as well as 10 modern farmers.

Frequent cyclones in Madagascar could cause damage and delay project implementation.

MALAWI/ Blantyre/Shire Highlands and Mamwera-Mangochi Rural Development

Sectoral Objective: Strengthen agricultural and livestock production by small-scale farmers with a view to reducing malnutrition in areas concerned through improvement in family income and supply of food items. Specific Project Objectives: 1. Increase in additional food crop

production, livestock products and fish

2. Increase in farmers’ income through

tobacco and cotton production 3. Poverty reduction by providing

infrastructure and social services.

The project should have an impact on the 57 780 farmers’ families in the two project areas. 1.1 Average cereal output: 1308kg/ha; 1.2 Procurement of 1000 finishing steers 1.3 26% fall in fish loss. 2.1 467kg/ha for tobacco and 773kg/ha for cotton. 3.1 Not stated

Information not available (the extension services did not supervise farmers in any project zone as planned at appraisal) 1.1 1132kg/ha of average cereal

output ; 1.2 94 finishing steers procured 1.3 2% fall in fish loss 2.1 458kg/ha for tobacco and

695kg/ha for cotton. 3. Construction of 3 maternities and 2 dispensaries; sinking of 72 wells in the Blantyre zone, 120 in Mangochi and 60 in Namwera.

Crop production depends on weather conditions Very complex project requiring smooth coordination among several partners

CAR/M’Baiki-Berberati Rural Development

Sectoral Objective: Promote integrated development in M’Baiki and Berberati Region with a view to improving the people’s living standard. Project Specific Objectives: 1. Increase coffee and foodcrop

production 2. Improve road and social

infrastructure

Impact of activities on a population of nearly 284 715. 1.1 Supply of input (590 t of fertilizer)

and coffee crop production equipment (8694 ha of old plantations to be regenerated)

1.2 Average coffee output: 1400kg/ha 1.3 Groundnut, maize, rice and cassava

production 2.1 Construction of 1700 km of access

roads 2.2 Construction of 17 health centres 2.3 40 wells sunk and 70 water sources

developed

Information not available (actually very little impact) 1.1 6159 ha affected 1.2 Average coffee yield: 600 kg/ha; 1.3 Implementation rate: groundnut

(30%), maize (35%), rice (41%) and cassava (43%)

2.1 Access roads built: 474 km 2.2 14 centres built 2.3 40 wells sunk and 19 sources

developed

Good world cocoa prices. Capacity of the executing agency to implement the programme (for which it was not prepared).

RWANDA/ BYUMBA Rural Development

Sectoral Objective: Improve the living standard of families in the project area. Special Project Objectives: 1. Agricultural and livestock production

development 2. Improvement of support infrastructure and establishment of extension and supervision structures for farmers’ organizations 3. Implementation of a preventive

medical programme.

19370 farming families benefitted from project operations and would have their income increase between 13 and 67% 1.1. Erosion control: replantation of live

hedges where there was none on 46 000 ha of arable land in the area

1.2. Reforestation programme on 5075 ha1.3. Selected seed production in 7

community centres and 66 agricultural sector centres

1.4. Construction of a sunflower oil extraction unit with a capacity of 400 t yearly

2.1 Construction of 7 warehouses with storage wings 2.2 Construction of 14 slaughter wings 2.3 Implementation of a road

programme: renovation of 350 km, opening of 100 km of community roads and building of 102 km of roads of regional interest.

3 Construction of 2 health centres.

Information not available (the production component that would have helped to increase income failed) 1.1 36 000 ha were protected 1.2 5085 ha reforested; 1.3 Inconclusive experience in terms

of quantity and quality; 66 centres were abandoned.

1.4 Components abandoned. 2.1 3 warehouses built 2.2 6 slaughtering wings built 2.3 250 km built (out of the 350 km

and 102 km). 3. The 2 health centres built and equipped.

Effectiveness of the extension system Farmers’ financial capacity to buy improved seeds and fertilizer.

SENEGAL/North-East Senegal Bush Fire Control and Reforestation

Sectoral Objective: Making the people accountable for managing the natural resources in their land through the development of forestry activities. Specific Project Objectives: 1. Protect the region’s sylvo-pastoral

resources against damage from bush fire

2. Set up Senegal Acacia plantations with a view to increasing the people’s income

n.s. 1.1 Establishment of 200 village

committees to undertake bush fire control and reforestation activities

1.2 Construction of 2500 km of fire wall 2.1 Establishment of community plantations (2500 ha) and on force account (500 ha)

Audiovisual training and sensitization sessions not planned at appraisal offered to 16 232 persons 1.1 600 committees set up 1.2 1610 km built 2.1 1215 ha of community plantations established and 669 on force account

TUNISIE/Mahdia Rural Development

Sectoral Objective: Reduce poverty and ensure regional balance by improving the nutrition, health and education of people in the project area. Specific Project Objectives: 1. Increase in agricultural and livestock

production

Increase in the income of 6300 farmer project beneficiaries. 1.1 Clearing of quitch-grass/jujube (13

850 ha), almond and pistachio cultivation (3910 ha)

1.2 Livestock development:

development of 5650 ha of grazing grounds, establishment of a ram selection centre and 2 veterinary centres, 186 stables and sale of 490 young heifer

Sure income improvement. 1.1 Quitch-grass clearing: 220 ha out

of the 1817 ha of plantation planned (all crops combined)

1.2 5315 ha of grazing grounds planned for the two centres were not set up, 2 stables were built out of the 186 units expected, and 186 young heifer were sold

Inadequate ground water table owing to prolonged drought Effective participation of the people Stability of the project institutional framework

2. Establishment of economic and

social infrastructure.

1.3 Irrigation development: SIP

renovation to irrigate 870 ha 2.1 Building of farm roads: 150 km 2.2 Construction of a health centre 2.3 Construction of a secondary school 2.4 Granting of credit to farmers: TD 2.7

million.

1.3 610 ha developed through the SIP 2.1 203 km built and tarred 2.2 Implemented and very operational 2.3 Implemented and very operational 2.4 Reimbursement rate: 24%.

TUNISIA/ GHEZALA MATEUR Agro-Complex Development

Sectoral Objective: Increase the productivity of various agricultural and livestock products with a view to reducing the region’s and country’s food deficit. Specific Project Objectives: 1. Intensify the production of various

agricultural and livestock products 2. Extension to the rural area of more

efficient product production and processing technologies.

Make 11 462 persons sharers in the project benefits. 1.1 Establish 400 ha of tree crop

plantations under drip irrigation, cultivate 250 ha of vineyard

1.2 Set up 500 ha of fodder plantation under sprinkling irrigation and improve pasture on 500 ha

1.3 Procurement of 1000 young heifers and 3000 sheep

2.1 Construction of hen houses for 200 000 layers and capacity for 54 million eggs yearly

2.2 Construction of a fruit processing station with capacity for 8000 t per annum.

Positive impact on poultry farming and dairy production. 1.1 137 ha developed, including 38 ha

of vineyard 1.2 78% implemented 1.3 1000 young heifer and 1500 sheep

procured; 2.1 Implemented. 2.2 Capacity reduced to 1500 t per annum.

Encouraging prices paid to farmers. Pedo-climatic conditions that could lead to soil deterioration for orchard cultivation. Individual capacity to absorb advanced technology.

Annex III Page 1 of 2

PERFORMANCE RATING

1 Conventionally, the Bank’s performance rating ranges from 1 to 4, corresponding to the following criteria: highly unsatisfactory, unsatisfactory, satisfactory, highly satisfactory. This tight rating poses a problem. A situation may be non-satisfactory without being classified as unsatisfactory. Furthermore, between the unsatisfactory and satisfactory categories, there is a certain dichotomy, preventing an intermediate (median) rating. The rating scale (1 to 4) tends to compress certain often qualified realities. It would be better to have a seven-criteria rating: extremely poor, poor, average, fairly good, good, very good, excellent. The rating scale would then range from 1 to 7. The rating for each criterion could be further refined by assigning coefficients ranging from 1 to 3. 2 However, to conform with usual Bank practice, the rating summarized below conforms with the existing system. The performance criteria considered here for each project concerns: the operational, institutional, economic and financial, social, environmental, Bank performance and sustainability. No INDICATORS COMMENTS 1 Operational performance The projects were implemented behind schedule. However,

in calculating the average, some projects that recorded tangible results (Madagascar and Senegal) were adversely affected by others (Guinea-Bissau).

2 Institutional performance Excluding El-Beheira (Egypt), projects implemented by the autonomous units and that concentrated on a limited range of objectives performed better. The decentralization of responsibilities was a key success factor. In contrast, frequent project staff changes and project sensitivity to political and institutional changes explain the poor performance recorded.

3 Economic performance The economic rates of return calculated at appraisal were generally too optimistic and erroneous. The calculation assumptions were neither homogeneous nor reliable. The absence of basic data generates some confusion between the “with” and “without” project situation.

4 Social performance The social objectives were only formulated in vague terms. Variables underlying the social indicators were not broken down. Most projects comprise a health component, but based on the results of the completion report, an opinion cannot be formed on the quality of services provided or care for the indigent.

Annex III Page 2 of 2

PERFORMANCE RATING

5 Environmental performance Except the North-Eastern Senegal Project, the raison d’etre

of which was to combat bush fire, the environmental performance of nearly all projects reviewed was only assessed on project completion, the reason being that ecological issues received little attention in the eighties.

6 Bank’s performance The Bank’s marginal involvement in project identification and preparation as well as limits it set on its operating modalities during project implementation weakened general project performance. Out of the ten projects reviewed, only two were considered as satisfactory.

7 Sustainability Project sustainability was rated according to their relation to national and regional development programmes, appropriation of experiences accumulated during the project implementation phase and actual beneficiary involvement.

PERFORMANCE RATING COUNTRY

O. P.

Inst. P

EFP

Soc. P.

Env. P.

Sus

Bank P.

TOTAL

Ave../Project

Guinea-Bissau 1 1 1 1 1 1 1 07 1 Tunisia DIM 2 1 2 3 2 3 1 14 2.0 Madagascar 4 4 4 3 3 3 3 24 3.4 Ethiopia 1 1 1 1 1 1 1 07 1 Egypt 3 4 3 3 3 3 2 21 3.0 CAR 1 1 1 1 1 1 2 08 1.1 Malawi 1 1 1 1 1 1 2 08 1.1 Rwanda 2 1 2 3 3 2 2 15 2.1 Senegal 3 4 3 3 4 3 3 23 3.2 Tunisia AC 2 2 3 2 2 2 1 14 2.0 O.P. = Operational performance Int P. = Institutional performance EFP = Economic and financial performance Soc. P. = Social performance Env. P. = Environmental performance Sus. = Sustainability Bank P. = Bank performance

Annex IV Page 1 of 2

List of Documents Used Completion Report : North-East Senegal Bush Fire Control and Reforestation Project

(Republic of Senegal). January 1995 Completion Report : Byumba Rural Development Project (Republic of Rwanda). February 1995 Completion Report : Tombali Rice Project (Republic of Guinea-Bissau). July 1998 Completion Report : Sidamo - Gamo Gofa (PDAR III) (Republic of Ethiopia). July 1996 Completion Report : Tsiribihina Irrigation Development Project

- Phase I Manambolo (Republic of Madagascar). Sept. 1997 Completion Report : M’Baiki and Berberati Coffee Region Rural Development

Project (Central African Republic). Dec. 1994 Completion Report : Mahdia Rural Development (Republic of Tunisia). July 1995 Completion Report : El-Beheira Rural Development (Republic of Egypt). Nov. 1997 Completion Report : Blantyre/Shire Highlands and Mamwera – Mangochi Rural Development Project (Republic of Malawi). April 1996 Completion Report : Ghezala Mateur Agro-Complex (Republic of Tunisia). Mars 1995 Performance Audit Report : Tsiribihina-Manambolo Irrigation Development, Phase I (Republic of Madagascar). January 1998 Performance Audit Report : Mahdia Rural Development (Republic of Tunisia). January 1997 Interim Project Performance Young Farmers’ Training Audit Report Project, Burkina Faso OPEV Dept. – African Development Fund Sept. 1998 Appraisal Report Performance Evaluation Review Retrospective 1994-1995. Sept. 1997 Appraisal Report Evaluation Performance Review (1992-93 operations). March 1995 Appraisal Report World Bank and Rural Development

– Retrospective of 1965-86 Operations. April 1988 African Development Bank: appraisal reports and country strategy papers for the ten

countries sampled. In IK Notes n° :5 “ Burkina Faso. Literacy for the Little Ones in Nowgene ”, World Bank,

Washington, February 1999

Annex IV Page 2 of 2

F.A.O : Report on Agricultural Development in Africa,

Rome 1998. ADB : Agricultural Sector Policy, Abidjan, 1990 ADB : ADF VII and VIII Lending Policy. ADB : Poverty Reduction Strategy and Action Programme,

Abidjan, Nov. 1992. ADB : Guidelines for Implementing the Poverty Reduction

Programme, May 1994. World Bank : Evaluating Assistance, Reasons for its Success, and

Failures, Washington, Nov. 1998 World Bank : Combatting Poverty in Africa, Washington, Oct. 1996.