China WGPAP CE

PROJECT EVALUATION

People's Republic of China

West Guangxi Poverty-Alleviation Project

International Fund for Completion Evaluation Agricultural Development Via Paolo di Dono, 44 00142 Rome, Italy Tel: +39 06 54592048 September 2010 Fax: +39 06 54593048 E-mail: [email protected] Web: www.ifad.org/evaluation

IFAD Office of Evaluation Bureau de l’évaluation du FIDA Oficina de Evaluación del FIDA

Document of the International Fund for Agricultural Development

People’s Republic of China

West Guangxi Poverty-Alleviation Project

Completion Evaluation

September 2010 Report No. 2200-CH

Photo on cover page: Soil improvement. The comparison of project with county achievements reveals that the relative contribution of the project was highest in terms of areas improved by irrigation schemes and soil management techniques . Source: Ernst Schaltegger

People’s Republic of china

West Guangxi Poverty-Alleviation Project Loan No. 552-CN

Completion Evaluation

Table of Contents

Currency Equivalent iii Weights and Measures iii Abbreviations and Acronyms iii Map of the Project Area v Map of the Project Area – Close Up vii Foreword ix Acknowledgements xi Executive Summary xiii Agreement at Completion Point xix

I. EVALUATION OBJECTIVES, METHODOLOGY AND PROCESSES 1 A. Methodology and Processes 1

II. COUNTRY AND SECTOR BACKGROUND 2 A. The Economy 2 B. Policies 4

III. PROJECT BACKGROUND 6 A. Location, Rationale and Partnerships 6 B. Design Features and Process 8

IV. IMPLEMENTATION RESULTS 9 A. Agricultural Development 10 B. Rural Financial Services 13 C. Social Development 14 D. Infrastructure 15 E. Project Management 16

V. PROJECT PERFORMANCE 17 A. Relevance 17 B. Effectiveness 19 C. Efficiency 22

VI. RURAL POVERTY IMPACT 25

VII. SUSTAINABILITY AND INNOVATION 32 A. Sustainability 32 B. Innovation, Replication and Scaling Up 34

VIII. PERFORMANCE OF PARTNERS 37

IX. CONCLUSIONS AND RECOMMENDATIONS 40 A. Conclusions 40 B. Recommendations 42

TABLES

1. Basic Project Data 7 2. Project and County Level Output Indicators, Agricultural Development 11 3. Project and County Level Output Indicators, Rural Financial Services 13 4. Project and County Level Output Indicators, Social Development 15 5. Project and County Level Output Indicators, Infrastructure 16 6. Project and County Level Output Indicators, Project Management 16 7. Trends in Infrastructure Coverage in Project Administrative Villages 21 8. Topics and Impact Domains in Schoolchildren Drawings 25 9. Results from the Ranking Exercise at the Evaluation Wrap-up Meeting 26 10. Gaps between Design and Implementation of VIGs 31 11. Evaluation Rating Summary 41

FIGURES

1. WGPAP – From Demonstration to Extension, Areas and Animals 12 2. WGPAP – From Demonstration to Extension, Incremental Output Values 12 3. Village Coverage per Project Activity 22

APPENDICES

1. Mission Itinerary 45 2. List of Persons Met 47 3. Bibliography 55 4. National and Regional Policies and Programmes 59 5. Financial Progress including Contingencies, in ‘000 CNY 61 6. Project Costs including Contingencies by Funding Source, in ‘000 CNY 63 7. Evaluation Framework 65 8. Definition of the Evaluation Criteria used by the Office of Evaluation 71

ANNEXES (*)

I. Client Perspective on Rural Finance II. Household Survey Data III. Social Development, Rural Infrastructure, Gender, June 2009

(*) Available upon request from IFAD’s Office of Evaluation [email protected]

Currency Equivalent

Currency Unit Yuan (CNY) US$ 1.00 8.26 (at Appraisal, December 2000) US$ 1.00 6.84 (at Completion, September 2008) CNY 1.00 US$ 0.147 (at Completion, September 2008)

Weights and Measures

1 mu = 0.66 ha 1 ha = 15 mu

Abbreviations and Acronyms

APR Asia and Pacific Division, Project Management Department, IFAD COSOP country strategic opportunities paper ERR economic rate of return ESSN Environmental Screening and Scoping Note FFT food-for-training FFW food-for-work FPA Farmers Professional Association GACFFPA Guangxi Administration Centre of Foreign Funded Projects for Agriculture GDP gross domestic product GZAR Guangxi Zhuang Autonomous Region IOE IFAD Office of Evaluation M&E Monitoring & Evaluation MDG Millennium Development Goal MIS management information system MTR mid-term review OECD Organisation of Economic Cooperation and Development PCR Project Completion Report PIA Participatory Impact Assessment PMO Project Management Office PRA participatory rural appraisal RCC Rural Credit Cooperative RCCU Rural Credit Cooperative Union RCF Rural Credit Foundation RIMS Result and Impact Management System UNOPS United Nations Office for Project Services VAM vulnerability assessment and mapping VDP village development plan VIGs village implementation groups WEILAI Water, Economy Investment, Learning and Assessment Indicator WF Women’s Federation WFP World Food Programme WGPAP West Guangxi Poverty-Alleviation Project

iii

vii

viii

Foreword

The overall objective of the West Guangxi Poverty Alleviation Project (WGPAP) was to achieve sustainable and equitable poverty reduction in the karst area of Guangxi Zhuang Autonomous Region (GZAR) in south-western China. To that end, the project aimed to increase productive capacity, both on- and off-farm, and to offer better access to economic and social resources, including financial services, education, health and social networks in ten of the poorest GZAR counties. Total project costs were estimated at US$107.3 million, including an IFAD loan of US$30.4 million, a World Food Programme grant of US$11.2 million, and contributions from the Government and beneficiaries of US$54.0 million and US$11.7 million, respectively.

The Office of Evaluation (IOE) of IFAD undertook a completion evaluation of the WGPAP in 2009. The evaluation noted that the project was relevant to the needs of the rural poor inasmuch as it addressed the major dimensions of rural poverty. Its design was consistent both with national policies and with the IFAD- Government country strategy, and it also integrated lessons from previous IFAD-financed interventions in China. Provision of rural credit helped to support agricultural investments and led to increased off-farm income-generating activities, including opportunities for migrant labour. The evaluation also discerned a significant impact in the domains of social capital and empowerment, food security, natural resources and the environment. Systematic, broad-based training activities and literacy courses, especially for women, greatly contributed to this success. As far as institutions and policies are concerned, remarkable progress was seen in the service delivery capacity of both the county/township authorities and the Rural Credit Cooperatives network.

Overall, the success of WGPAP can be attributed to strong synergy of three key ingredients, namely, quality at entry, implementation capacity and a conducive policy and institutional environment. Quality at entry was marked by the early unwavering determination, by and among the main partners, to reach agreement on key fundamentals of the project: solid partnerships forged along proven patterns, strong participation and ownership on the side of national partners during design, and participatory mechanisms involving large numbers of the target population. The project’s implementation capacity was driven by the experience and ability of the provincial project management office to provide decisive guidance, by the presence of committed and qualified human resources, and by the continuity of such human resources. The project environment was conducive as implementation was supported by a set of government policies with a clear rural poverty-reduction focus that emerged at WGPAP design, and the local government structure showed itself to be capable of integrating the project as a mainstream undertaking. The Government of China displayed a strong sense of ownership in the project, particularly at the local level. Finally, the stable political environment and booming economy during the life of the project constituted a robust bottom line.

Concerning areas for improvement, the evaluation noted that the strategy of reducing women’s workload was pursued in water and biogas access, but not as a cross-cutting theme in agriculture, thus underpinning the persistent labour-intensive nature of farming activities. In addition, the project achieved little impact in the field of women leadership development, one of the IFAD-Government country strategy priorities. Regarding environmental protection, the evaluation observed that, while the WGPAP was innovative in the agricultural development component in general terms, it failed to convey strong and consequent messages regarding fertiliser use. Given the scarcity of land in China, farmers try to maximize land productivity by extensive use of fertilizer, three times higher than the world average, thus adding to land pollution.

This completion evaluation report includes an Agreement at Completion Point summarizing the main findings of the evaluation. It sets out the recommendations that were discussed and agreed upon by IFAD and the Government of the People’s Republic of China, together with proposals as to how and by whom the proposals should be implemented.

Luciano Lavizzari Director, Office of Evaluation

Acknowledgements

This completion evaluation was led by Mark Keating, Evaluation Officer, and prepared with contributions by consultants Ernst Schaltegger, Claude Saint-Pierre, Rauno Zander and Giulia Mezzi.

Internal peer reviewers from the IFAD Office of Evaluation (Andrew Brubaker, Ashwani Muthoo and Jicheng Zhang) reviewed the approach paper and the draft report. Melba E. Alvarez, Evaluation Communication Assistant, provided administrative assistance.

The Office of Evaluation is grateful to IFAD’s Asia and the Pacific Division for their insightful inputs and comments at various stages throughout the evaluation process.

Appreciation is due to the Government of the People’s Republic of China for the valuable comments and support provided throughout the evaluation process, including the organization of the learning workshop held in January 2010. Appreciation is expressed also to the provincial Project Management Office for the excellent support provided to the evaluation mission, and to the county authorities for their outstanding hospitality.

______Director, IFAD Office of Evaluation (IOE): Luciano Lavizzari Lead Evaluator, IOE: Mark Keating Consultants’ Team Leader, IOE: Ernst Schaltegger

xii

People’s’ Republic of China

West Guangxi Poverty-Alleviation Project

Completion Evaluation

Executive Summary

I. Introduction

1. In line with the Evaluation Policy of the International Fund for Agricultural Development (IFAD), the Fund’s Office of Evaluation (IOE) has undertaken in 2009 a completion evaluation of the IFAD- financed West Guangxi Poverty-Alleviation Project (WGPAP) in the People's Republic of China. The main purpose of this evaluation was to assess the performance and impact of the project and generate lessons learnt that will guide the Government of China and IFAD in ongoing and future agriculture and rural development programmes in the country.

2. The basis for the completion evaluation was the self-assessment prepared by the Asia and Pacific Division of IFAD (APR), building upon own insight during project implementation and on the Project Completion Report (PCR). The Desk Review Note, prepared by IOE prior to the main mission, was another other important source of evidence. Mission activities concentrated, among others, on discussions with focus groups and individual households, and drawing contests organized in five primary schools, one in each county visited. These contests aimed at revealing impact domains in terms of poverty impact as perceived by schoolchildren. The mission also prepared 48 scorecards focusing on village implementation groups (VIGs), 20 detailed household interview records, related to the rural finance component, and conducted a ranking exercise of project strengths and weaknesses at the wrap-up meeting in Nanning, which will be referred to in this evaluation report. Thus, the mission attempted to triangulate perceptions of project performance and impact, by taking into account viewpoints of the project’s stakeholders, beyond the documentation prepared by the provincial and county Project Management Offices (PMOs) for the evaluation.

3. With a population of 1.31 billion inhabitants in 2007 and a land area of about 9.6 million square kilometres, China is the most populous country in the world and the third largest in terms of land size. During the past 25 years, China achieved remarkable economic growth of more than 9 per cent per year and managed to keep inflation in check over the last ten years. In 2005, the country ranked fourth in the world in terms of gross domestic product (GDP) as a result of the economic reforms introduced three decades ago. Because of this sustained rapid economic growth, China is ahead of schedule in achieving the Millennium Development Goals for national poverty reduction. More than 400 million people have moved out of poverty by increasing agricultural productivity and creating productive employment opportunities in the service and manufacturing sectors. The economy has become more urbanized than before, increasing the urban poor population up to 41.8 per cent in 2004. The distribution of economic benefits among economic and social groups has not been even, as inequalities still persist in three dimensions: i) rural-urban; ii) east-west (geographic); and iii) inland-coastal regions.

4. The national poverty reduction strategy had a marked focus on mountainous and other marginal areas, and on multi-sector poverty reduction programmes. Poverty alleviation efforts were focused on poorer administrative villages and on reaching out to all natural villages within one administrative village. Participatory approaches and capacity building at community level were increasingly seen as a major leverage factor in the development of marginal areas.

5. The WGPAP was approved by the IFAD Executive Board in December 2000 and declared effective in March 2002, with an IFAD loan portion of US$30 million and total project costs of US$106 million. The long term goal of the project was to achieve sustainable and equitable poverty reduction in the karst area of the Guangxi Zhuang Autonomous Region (GZAR). The overall objective

xiii

was to achieve a sustainable increase in productive capacity, both on- and off-farm, and to offer increased access to economic and social resources, including financial services, education, health and social networks in ten of the poorest counties of GZAR. WGPAP was one among five integrated rural development projects that have been launched during the period 1999-2005 in the framework of the 1999 country strategic opportunities paper (COSOP), but the only IFAD-funded project designed in Southwest China during that period. The project included five components, namely: i) Agricultural Development; ii) Financial Services; iii) Social Development; iv) Rural Infrastructure; and v) Project Management. In line with the 1999 COSOP, explicit partnerships were built into the design of WGPAP with the World Food Programme (WFP) and the Rural Credit Cooperative Union (RCCU) network in the project area. While a provincial PMO oversaw project implementation, the bulk of project activities was carried out by county and township PMOs.

II. Project Performance

6. WGPAP was relevant for the needs of the rural poor and addressed all important dimensions of poverty. Its design was consistent with both national policies and IFAD’s strategies, and integrated lessons from other IFAD projects in China. The whole design process was predominantly based on national expertise and can be considered as good practice. Effectiveness was highly satisfactory specifically measured by indicators in the dimensions of food security, income generation, health, education and enterprise start-ups. One of the key ingredients of effectiveness was the high coverage achieved across the ten project counties. Overall efficiency was satisfactory, but the project failed to conclusively prove that the economic rate or return (ERR) at completion was probably higher than the ERR assumed at appraisal. The study commissioned to this effect by the project, while commendable as such, lacked the necessary methodological rigour.

7. Poverty reduction impacts of WGPAP were in the satisfactory range for all five impact domains. Despite the existence of mainstream poverty-alleviation programs funded by the Government, impact in the project counties and townships could be attributed to the project with a fair degree of certainty, whereby the mission consciously triangulated impact perception from observation points that were additional to the PCR, which was admittedly the main source of evidence. Household income and assets progressed at a significant pace, inferring that the poverty reduction targets have been attained. In the domains of social capital and empowerment, food security and natural resources and environment, the evaluation mission also perceived significant impacts, derived from capacity building thanks to broad based literacy courses and agricultural training of which women were key beneficiaries. Food security was improved due to substantial productivity increases in cereals and sweeping crop and animal production systems diversification. The evaluation mission noticed, in relation to the impact domain of institutions and policies, remarkable progress in service delivery capacity of county and township authorities as well as of the Rural Credit Cooperative (RCC) network, but also had to recognize that generic institutional capacity at administrative village level remained weak.

8. Overall sustainability is rated as satisfactory. Two key factors of sustainability are the existence of mainstream government programmes in poverty alleviation - to which WGPAP was complementary - and the seamless integration of the PMO at county and township levels into local government structure. These factors outweigh areas of constrained sustainability, such as the still perceivable subsidy elements in the formation of interest and refinancing rates of the RCC network and the high level of public service delivery in remote areas. In terms of innovation and replication, the project went at length in capturing and documenting innovations, but it received insufficient innovation inputs from IFAD, despite the pronounced innovation drive in the 2005 COSOP.

9. A salient trait of the WGPAP evaluation is that all involved partners exhibit performance ratings in the satisfactory range. Government and beneficiary performance ranked best, not only for their ability to implement the project effectively and efficiently, but also for the determination of the PMOs at county and provincial level to establish and run a comprehensive monitoring and evaluation (M&E) system. WGPAP would have had the opportunity of transforming this M&E system into a genuine and real time management information system (MIS) if appropriate guidance and support from IFAD had been provided. Despite this shortfall, both IFAD and the United Nations Office for Project Services

xiv

(UNOPS) performance ratings are in the satisfactory range, marked by continuity and consistency in personnel and concepts that have marked a clear line of business. In fact, WGPAP was seen as a largely problem-free operation by all involved actors.

III. Findings and Recommendations

A. Findings

10. Summarizing the main evaluation findings, a storyline of relative success emerges. The table below underlines this with ratings that are all in the category of satisfactory achievements. Compared to the ratings contained in the Desk Review Note and to those made in the self-assessment of APR, only minor variations emerge. On the other hand, WGPAP ratings fare significantly better than the ones of the Annual Report on Results and Impact of IFAD Operations (ARRI) 2008.

Evaluation Rating Summary Ratings Evaluation Criteria Completion Evaluation A. Project performance Relevance 5 Effectiveness 6 Efficiency 5 Overall project performance 5 B. Rural poverty impact Household income and assets 6 Social capital and empowerment 5 Food security and agricultural productivity 5 Natural resources and the environment 5 Institutions and policies 5 Overall rural poverty impact 5 C. Overarching factors Sustainability 5 Innovation, replication and scaling up 5 D. Performance of partners IFAD 5 WFP 5 UNOPS 4 Government 5 E. Overall project achievement 5

11. Although the rating summary is a comprehensive and well tested analytical framework for the assessment of project success or failure, it falls short of convincingly explaining its underlying contributing factors. In a simplified manner, project success may be seen as a function of quality at entry, implementation capacity and conducive environment. Quality at entry of WGPAP was marked by: (i) an early and unwavering determination of and agreement on key fundamentals of the project, by and between the main partners, (ii) solid partnerships established along proven patterns, (iii) strong participation and ownership of the national partners during design, (iv) participatory mechanisms involving the target population on a broad basis, and (v) a component mix and cost sharing mode that reflected perceived needs as well as comparative advantages. WGPAP’s implementation capacity was driven by: (i) the experience and credibility of the provincial PMO to provide decisive guidance, (ii) the presence of committed and qualified human resources, sufficient in numbers and able to transform guidance into action, and (iii) the continuity of these human resources. The fact that poverty impact, sustainability and innovation were satisfactory stemmed from such implementation capacity, on one hand, and from a conducive environment, on the other hand. The project environment was conducive because: (i) project implementation was supported by a set of government policies that had a clear rural poverty reduction focus and that emerged when the WGPAP was designed, (ii) local government structure was capable, possibly by design and certainly in reality, to practically integrate the project as a mainstream undertaking, and (iii) a stable political environment and a booming economy during project.

B. Recommendations

12. The evaluation mission makes four recommendations as detailed below. The first refers to the design of future IFAD projects in China, the second to key aspects of institutional partnerships, in particular regarding rural financial services, while the third recommendation points to innovation and IFAD’s role of its promotion. The fourth recommendation deals with M&E and MIS. It is understood that the government of GZAR intends to propose a second project to IFAD that would be based on lessons learnt form the WGPAP.

13. Recommendation 1: Design of future IFAD-funded projects in China. The design of future IFAD-funded rural development projects in China should continue to be responsive to the multidimensional character of poverty and the growing expertise of national stakeholders in this kind of operations, in particular:

(a) Targeting of project townships should continue to be based on factual indicators. In projects focused on service provision, all villages within one township should be included in the project, and natural villages more clearly identified. Likewise, a transparent assessment of the special needs of the various ethnic minority groups is critical.

(b) The design process should replicate WGPAP’s good practice of strong involvement of local expertise at all levels, including from the anticipated target population.

(c) In settings where food-for-training and food-for-work is deemed essential for broad participation and coverage, a second-generation solution regarding food aid from domestic sources should be devised with the Chinese authorities, in order to replace WFP food aid.

14. Recommendation 2: Institutional partnerships. Project partnerships should extend over the boundaries of provincial and local government in the strict sense, e.g.:

(a) VIGs or analogous bodies at village level should not only play a role for planning and monitoring purposes, but should be strengthened in their core management capacity, with special regard to service delivery functions and equitable gender representation in positions of responsibility.

(b) Partnership with provincial RCC networks should be put on a completely new footing, by taking into account the ongoing reform and by agreeing on a set of information to be shared that is conducive to a real-time assessment of loan portfolios and banking performance indicators in general.

(c) Key areas of policy dialogue should be determined systematically and given due and documented attention over the project implementation period.

15. Recommendation 3: Role of innovation in IFAD-funded projects. Building on the orientations of the 2005 COSOP, new IFAD-funded projects should put significantly more emphasis on innovation and its promotion, in particular:

(a) Opportunities for innovations should be screened systematically from formulation stage onwards and be periodically reassessed.

(b) Particular attention should be devoted to the identification of local sources of innovation and their mobilization during project implementation.

(c) Innovations that exist or emerge in the global context should be made available to IFAD- funded projects more pro-actively, possibly also via IFAD grants.

xvi

(d) Priority areas should encompass innovations that respond to global challenges, such as food security, soil fertility, alternative energies and climate change.

16. Recommendation 4: From M&E to MIS. New IFAD-funded projects should take stock from the accumulated experience of the WGPAP and actively foster a quantum leap in the state-of-the-art of M&E and MIS, with the aim to make advanced M&E and MIS methods a mainstream feature in China. This would include the following:

(a) The generic requirements of an overarching, yet concise, M&E system for multidimensional poverty-alleviation projects in marginal rural areas should be defined, best in a joint effort between national expertise and IFAD resources.

(c) The above recommendations would give a particularly concrete meaning to the development of local evaluation capacity, an initiative that IFAD has decided to support in China.

xvii

xviii

People’s Republic of China

West Guangxi Poverty-Alleviation Project

Completion Evaluation

Agreement at Completion Point

A. Core Learning Partnership and Users of the Evaluation

1. In line with the Evaluation Policy of the International Fund for Agricultural Development (IFAD), the Fund’s Office of Evaluation (IOE) undertook in 2009 a Completion Evaluation of the IFAD-financed West Guangxi Poverty-Alleviation Project (WGPAP) in the People's Republic of China. The main objective of this evaluation was to assess the performance and impact of the project and generate lessons learnt that will guide the Government of China and IFAD in ongoing and future agriculture and rural development programmes in the country.

2. IFAD’s Evaluation Policy, while underscoring the need for independence, recognizes the importance of involving the main stakeholders throughout the evaluation process. This is fundamental in ensuring the engagement of stakeholders in a fruitful collaboration and to facilitate the discussion around the recommendations emerging from the evaluation and their adoption. To this end, a core learning partnership was formed, composed of the main users of the evaluation. Specifically, it included: (i) the Ministry of Finance of the People’s Republic of China; (ii) the Foreign Capital Project Management Centre, State Council Leading Group of Poverty Alleviation and Development; (iii) the Agricultural Department of the Guangxi Zhuang Autonomous Region (GZAR); (iv) the Guangxi Administration Centre of Foreign Funded Projects for Agriculture (GACFFPA) having functioned as Project Management Office (PMO); (v) the Rural Cooperative Credit Union (RCCU) of GZAR; (vi) the Women’s Federation (WF) of GZAR; (vii) the Asia and Pacific Division of the IFAD Programme Management Department (APR); (viii) the IFAD Country Presence Office in China; (ix) the World Food Programme (WFP), China Office; and (x) the United Nations Office for Project Services (UNOPS) that acted as the project’s cooperating institution until 2007.

3. On 15 January 2010, members of the core learning partnership, along with other stakeholders, convened in Nanning, the regional capital of GZAR, and held the final learning workshop on the evaluation. This Agreement at Completion Point summarizes the main findings and recommendations as contained in the evaluation report. It also benefits from the discussions around the main issues emerging from the evaluation debated at the learning workshop.

4. The ACP has been reached between IFAD (represented by the Programme Management Department) and the Government of China (represented by the Ministry of Finance) and reflects their understanding of the main findings from the Completion Evaluation, as well as their commitment to adopt and implement the recommendations contained in section C of this ACP within specific timeframes.

B. Main Evaluation Findings

5. The West Guangxi Poverty-Alleviation Project was relevant to the needs of the rural poor in that it addressed all important dimensions of poverty. Its design was consistent with both national policies and IFAD strategies, and integrated lessons emerging from other IFAD-funded projects in China. By selecting poverty counties through a focused targeting in accordance to the vulnerability assessment and mapping (VAM) method developed by WFP, dimensions that were highly relevant to the needs of the poor in the project area were included right from the outset at project start. The inclusion of a rural financial services component based on the Rural Credit Cooperatives (RCC) network substantially enhanced relevance. The whole design process can be considered as good

xix

practice, given the strong synergy of three key ingredients, namely quality at entry, implementation capacity and a conducive environment.

6. Effectiveness was highly satisfactory. That is, the overall project objective of achieving a sustainable increase in productive capacity, both on- and off-farm, and to offer increased access to economic and social resources, including financial services, education, health and social networks was successfully met. One of the key ingredients of effectiveness was the high coverage achieved across the ten project counties. Overall efficiency was satisfactory, but the project at completion did not conclusively prove that the economic rate of return (ERR) at completion was probably higher than the ERR assumed at appraisal, as the study commissioned to this effect by the project was not entirely rigorous. Such studies, if carried out with the necessary methodological rigour, are highly commendable with regards to evaluation capacity development, for which China shows strong interest.

7. Poverty reduction impacts of WGPAP were in the satisfactory range for all five impact domains. Despite the existence of mainstream poverty-alleviation programs funded by the government, impact in the project counties and townships could be attributed to the project with a good degree of certainty, as; 75 per cent of the project beneficiaries received support from the WGPAP only, inferring that the impacts observed can be attributed to the project. Household income and assets progressed at a significant pace, inferring that the poverty reduction targets were attained. In the domains of social capital and empowerment, food security and natural resources and environment, the evaluation mission also perceived significant impacts. Systematic and broad-based training activities and literacy courses, especially for women, greatly contributed to this success. The evaluation noticed, in relation to the impact domain of institutions and policies, remarkable progress in service delivery capacity of county and township authorities as well as of the RCC network, but also had to recognize that generic institutional capacity at administrative village level remained weak. The project strengthened the coordination capacity of village cadres, but did not focus on management capacity. For instance, the village implementation groups (VIGs) were to include a village accountant, whereas none of the village committees interviewed by the evaluation mission had one.

8. Two key factors of sustainability were, and still are, (i) the existence of mainstream government programmes in poverty alleviation - to which WGPAP was complementary - and (ii) the seamless integration of the project management organization at county and township levels into local government structure. These factors outweigh factors that may constrain sustainability, such as the still perceivable subsidy elements in the formation of interest and refinancing rates of the RCC network and the high level of public service delivery in remote areas. In terms of innovation and replication and scaling up, the project went at length in capturing and documenting innovations, but it received insufficient innovation inputs from IFAD, especially in terms of innovative agricultural technologies to address soil fertility, despite the pronounced innovation drive in the 2005 country strategic opportunities paper (COSOP).

9. The WGPAP storyline of success is the result of a strong synergy between three key ingredients, namely (i) quality at entry; (ii) implementation capacity; and (iii) a conducive environment in terms of support provided by the central and local governments, a stable political environment and a booming economy. The Completion Evaluation formulated the four main recommendations detailed below with the intention of fostering such synergy in future projects and programs between the Government of China and IFAD.

C. Recommendations Agreed Upon by all Partners

10. All the recommendations derived from the Completion Evaluation have been accepted by partners involved. The paragraphs below provide details on the nature and on the implementation arrangements, including assigned responsibilities and deadlines as applicable, for the four main recommendation and the 13 derived sub-recommendations.

11. Recommendation 1: Design of future IFAD-funded projects in China. The design of future IFAD funded rural development projects in China should continue to be responsive to the

multidimensional character of poverty and the growing expertise of national stakeholders in this kind of operations, in particular:

(a) Targeting of project townships should continue to be based on factual indicators. In projects focused on service provision, all villages within one township should be included in the project, and natural villages more clearly identified. This is recommended in order to enhance critical mass in service delivery and would be congruent with the “block development” approach, which the Government of China is currently implementing. Likewise, a transparent assessment of the special needs of the various ethnic minority groups is critical.

(b) The design process should replicate WGPAP’s good practice of strong involvement of local expertise at all levels, including from the anticipated target population. This should include the joint calculation of ERRs at appraisal, and a set of guidelines that enable projects to revisit the ERR calculation at completion based on own expertise.

(c) In settings where food-for-work (FFW) and food-for-training (FFT) is deemed essential for broad participation and coverage, a second-generation solution should be devised with the Chinese authorities, for instance by linking donor-funded projects with government programs for infrastructure and human capacity building. The government has increasingly stepped up investment in rural infrastructure, and in human capacity building (e.g. training for migrant workers). The IFAD country program is linked with government programs to build similar complementarities. In countries where food aid continues to be policy, the lessons from the IFAD-WPF partnership in the WGPAP should be learnt.

12. Recommendation 2: Institutional partnerships . Project partnerships should extend over the boundaries of provincial and local government in the strict sense, that is:

(c) Key areas of policy dialogue should be determined systematically and given due and documented attention over the project implementation period.

13. Recommendation 3: Role of innovation in IFAD-funded projects. Building on the orientations of the 2005 COSOP, new IFAD-funded projects should put significantly more emphasis on innovation and its promotion, in particular:

(a) Ambitions of the project partners regarding innovations should be screened systematically from formulation stage onwards and be periodically reassessed.

(b) Particular attention should be devoted to the identification of local sources of innovation and their mobilization during project implementation.

(c) Innovations that exist or emerge in the global context should be made available to IFAD- funded projects more pro-actively, possibly also via IFAD grants.

(d) Priority areas should encompass innovations that respond to global challenges, such as food security, soil fertility, alternative energies and climate change.

xxi

14. Recommendation 4: From M&E (monitoring and evaluation) to MIS (management information system). New IFAD-funded projects should take stock from the accumulated experience of the WGPAP and actively foster a quantum leap in the state-of-the-art of M&E and MIS, with the aim to make advanced M&E and MIS methods a mainstream feature in China. This would include the following:

(b) The basic architecture of a real-time MIS, based upon such M&E requirements, should be designed in a way that it can be used for a wide array of IFAD-funded projects, in China and possibly also elsewhere. For this to substantiate, the data capturing and storing capabilities of an M&E system must evolve into a real time management system allowing easy analysis across all relevant variables, and thus require a program basis going well beyond the accumulation of mere spread sheets. Again, a partnership between national expertise and IFAD would be advisable.

(c) The above recommendations would give a particularly concrete meaning to the development of local evaluation capacity, an initiative that IFAD has decided to support in China.

D. Implementation Responsibilities and Proposed Timeframe

15. IFAD and the Government of China, with the contribution and support of local partners as appropriate, will be responsible for implementing all of the above recommendations and sub- recommendations. Such recommendations will be taken into account in formulating the new results- based COSOP and new IFAD-funded operations in China.

xxii

People’s Republic of China

West Guangxi Poverty-Alleviation Project

Completion Evaluation

Main Report

I. EVALUATION OBJECTIVES, METHODOLOGY AND PROCESSES

1. In line with the Evaluation Policy of the International Fund for Agricultural Development (IFAD), the Fund’s Office of Evaluation (IOE) has undertaken in 2009 a Completion Evaluation of the IFAD- financed West Guangxi Poverty-Alleviation Project (WGPAP) in the People's Republic of China. The main objective of this evaluation was to assess the performance and impact of the project and generate lessons learnt that will guide the Government of China and IFAD in ongoing and future agriculture and rural development programmes in the country.

A. Methodology and Processes

2. Methodology. The basis for this Completion Evaluation was, in the first place, the self- assessment prepared by the Asia and Pacific Division of IFAD (APR). Building upon own insight during project implementation and on the Project Completion Report (PCR),1 which was prepared by the provincial Project Management Office (PMO) in Nanning with the assistance of an IFAD consultant, APR gave ratings along the same evaluation criteria as the ones contained in table 11, chapter IX. The PCR and the Desk Review Note,2 the latter prepared by IOE prior to the main mission, were other important sources of evidence for the completion evaluation.3 Convergences and divergences with the evaluation mission rating pattern will be highlighted in para 137. The extensive literature available on China in general and poverty and food security issues in particular provided further insight into the project context. The mission also availed of the multiple records kept by the provincial and county PMOs, based upon the project’s monitoring and evaluation (M&E) system, which provided valuable insight into the workings of the project in the ten target counties. The fact that most of the achievements were not only recorded for the project in its strict sense, but also with regard to the relevant universe of the ten project counties, made the evaluation particularly rewarding. The due assessment of the abundant documentation was only limited by the language barrier, which may have caused some loss of access to otherwise meaningful evidence. Hence, the abundant and partly detailed references on the general environment and the project itself have caused the evaluation mission to aim for a report that attempted to build a “formula” containing the ingredients for successful project implementation. (See chapter IX, Conclusions and Recommendations).

3. Mission activities concentrated, among others, on discussions with focus groups and individual households, and drawing contests organized in five primary schools, one in each county visited. These contests aimed at revealing impact domains in terms of poverty impact as perceived by schoolchildren. The mission also prepared 48 scorecards focusing on village implementation groups (VIGs), 20 detailed household interview records, related to the rural finance component, and conducted a ranking exercise of

1 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, Draft, Rome, 2009. 2 IFAD, Office of Evaluation, Desk Review Note, West Guangxi Poverty-Alleviation Project, People’s Republic of China, Completion Evaluation, Rome, 29 April 2009. 3 The general purpose of a Desk Review Note is to summarize the information on a project’s performance and impact as reported in available evaluative documents and literature produced by IFAD, such as supervision reports, project progress reports, project status reports, the mid-term review, and the PCR. Moreover, the WGPAP Desk Review Note specifically identified specific hypotheses and issues that deserved deeper enquiry during the main project evaluation mission.

project strengths and weaknesses at the wrap-up meeting in Nanning, which will be referred to in this evaluation report. Thus, the evaluation mission attempted to triangulate perceptions of project performance and impact, by taking into account the viewpoints of the project’s stakeholders, beyond the documentation prepared by the provincial and county PMOs for the evaluation.

4. The evaluation methodology related to rural financial services was based on a simple two- pronged approach. First, a “demand assessment” to capture project impact for the clients of rural and microfinance services through WGPAP was carried out, by thoroughly analyzing the changing income and asset status of the clients. The twenty household interviews conducted as well as the impact surveys made at project completion provided the required evidence. Second, the mission proceeded to a “supply side institutional evaluation” to gauge the impact of the project on the rural and microfinance service providers. The evaluation process was participatory at all levels. The household interviews checks were carried to allow heeding the voices of the people for whom this project was designed. At institutional level, Rural Credit Cooperative Union (RCCU) and Women’s Federation (WF) representatives were invited to join household level interviews, and PMO representatives accompanied the evaluation team.

5. Processes. A preparatory mission was carried out in March 2009, both in Beijing and in Nanning, including a short visit to one of the project counties. At that time, a meeting with the main IFAD counterpart institution in the People’s Republic of China, the Ministry of Finance (MOF), was not possible, but IOE held discussions in Rome with representatives of MOF during the Executive Board Session of April 2009, MOF indicated that it wished to hold a debriefing session in Beijing on 27 May 2009, which took place as planned. The main evaluation mission was fielded from 4 to 27 May 2009. At mission start-up, the evaluation mission held a briefing session with the PMO in Nanning, the provincial capital of Guangxi Zhuang Autonomous Region (GZAR). Subsequently, it went on for an eleven-day field trip, together with a team of the provincial PMO. The mission visited five out of the ten project counties, 15 townships, 22 administrative and 20 natural villages,4 interacting with the county PMOs, the participating government agencies, Rural Cooperative Credit (RCC) agencies and RCC county unions (RCCUs). Four non-project township and four non-project villages were included in the visits as a control group. The evaluation mission produced an aide-memoire, which was discussed at a wrap-up meeting in Nanning on 26 May 2009 that was also an opportunity for rating project strengths and weaknesses by the invited stakeholders (see paragraph 3). The main evaluation report was prepared between June and September 2009, during which time it underwent an IOE internal peer review and another review by IFAD’s Asia and the Pacific Division (APR) of the Project Management Department. A learning workshop with the core learning partnership and other key partners was held in Nanning on 15 January 2010.

6. A detailed list of places visited is provided in appendix 1 while appendix 2 shows the persons met during the mission. The evaluation mission team wishes to express its gratitude to the provincial PMO for the excellent support provided to the mission, and to the county authorities as well as to the households, for their outstanding hospitality. Their active participation was instrumental for having gained evaluation relevant insight on a wide array of issues.

II. COUNTRY AND SECTOR BACKGROUND

A. The Economy

7. With a population of 1.31 billion inhabitants in 2007 and a land area of about 9.6 million square kilometres, China is the most populous country in the world and the third largest in terms of land size, extending 5,500 km from north to south and about 5,000 km from east to west. However, only about 10 per cent of the total land area is arable land, concentrated in the plains and river valleys of the lowlands in the east. The climate is continental, with extremes of temperature and is subtropical in the south-east. 5 From an institutional perspective, the second administrative tier of the country beyond the

4 Natural villages or settlements are grouped around an administrative village. 5 Economist Intelligence Unit (EIU). China Country Profile, 2008.

national government is organized into provinces, autonomous regions, special administrative regions and four major municipalities.

8. During the past 25 years, China achieved remarkable economic growth of more than 9 per cent per year and managed to keep inflation in check over the last ten years. In 2005, the country ranked fourth in the world in terms of gross domestic product (GDP) as a result of the economic reforms introduced three decades ago. Because of this sustained rapid economic growth, China is ahead of schedule in achieving the Millennium Development Goals (MDGs) for national poverty reduction. More than 400 million people have moved out of poverty by increasing agricultural productivity and creating productive employment opportunities in the service and manufacturing sectors. The economy has become more urbanized than before, increasing the urban poor population up to 41.8 per cent in 2004. The distribution of economic benefits among economic and social groups has not been even, as inequalities still persist in three dimensions: (i) rural-urban; (ii) east-west (geographic); and (iii) inland-coastal regions. 6

9. Agriculture and rural development. Despite the urban migration and the widening of the gap between rural and urban incomes,7 agriculture is a predominant sector of the Chinese economy, with rice being the most important grain crop.8 Even though the agricultural sector accounted for only 11.7 per cent of the GDP in 2006, 325 million people still made a living from farming, forestry, animal husbandry and fisheries, accounting for 44 per cent of total nationwide employment. The rural reforms introduced by the Government in 1978 transferred the control of land from collectives back to individual families. Consequently, the effect of reintroducing individual economic incentives in the countryside, allowing farmers a degree of freedom in planting crops, was highly positive; as a consequence, real net income per head in the countryside rose fivefold between 1978 and 2001. Nevertheless, senior government officials continue to express concerns about the health of the rural economy for both economic and climatic reasons. The proportion of the sown area devoted to grain production is continuing to fall; between 1999 and 2006 it fell from 72 per cent to 67 per cent, as farmers diversified into other, more profitable crops.

10. In comparison with the primary sector, China’s development is marked by a heavy reliance on investment and exports for economic growth, at the expense of domestic demand, especially consumption. This was reflected in a large and growing external current account surplus that reached 7 per cent of GDP in 2005. Domination of industry is substantial, especially heavy industry, over the services sector (services make up 40 per cent of GDP in China, compared with an average 54 per cent in middle income countries and 70 per cent in high income countries). China’s primary energy consumption grew by 62 per cent during 2000-05 largely because of its capital-intensive, industry-led growth. China’s energy intensity also climbed, reversing a trend decline since the start of market reforms. By 2005, it was 43 per cent higher than in India and 73 per cent higher than in the U.S., based on PPP GDP. Excessive withdrawal of water from surface and underground resources caused acute water scarcity in the northern plains. These basic facts put both the achievements in and opportunities for agriculture and rural development in a perspective or relativity.9

6 Ibidem. Further information about regional disparities is mentioned in the most recent country strategic opportunities paper, dated December 2005. As a result of a strong concentration of growth and investments in the eastern urban economies, regional disparities have been increasing, notwithstanding government programmes for the development of the western provinces. Moreover, rural and urban income differences grew, as shown by an urban/rural ratio of 3:1. Looking at regional disparities, the ratio of average urban incomes in the east (CNY 8,448) to rural incomes in the west (CNY 1,640) is even larger, set at 5.1:1. These regional disparities explain why, in 1999, around 89 per cent of all the rural poor and 70 per cent of all urban poor households were located in the central and western regions. 7 The 59 per cent rise in net rural income per head in 2000-2006 failed to match the 87 per cent rise in urban income per head over the same period. 8 IFAD, Report and Recommendation of the President to the Executive Board on a proposed loan to the People’s Republic of China for the West Guangxi Poverty-Alleviation Project, 2000, Part I. 9 World Bank, Mid-term Evaluation of China’s 11 th Five Year Plan, World Bank Office China, 18 December 2008.

B. Policies

11. Poverty reduction policies in the 2000s . The national poverty reduction strategy, in its 2001- 2010 phase, retained a marked focus on mountainous and other marginal areas.10 In this reference, the efficiency of multi-sector poverty reduction programmes and the importance of improved targeting are acknowledged. Poverty-alleviation efforts are now focused on poorer administrative villages and on reaching out to all natural villages within one administrative village. Participatory approaches and capacity building at community level are increasingly seen as a major leverage factor in the development of marginal areas. A total of 148,131 administrative villages have joined the integrated village development programme in 2001 and have established their own village development plan.11

12. Gender in poverty and related policy framework. Both IFAD’s country strategic opportunities paper (COSOP) of 1999 and the above cited World Bank study noted that gender labour division tended to be rigid and women workloads were high, especially in mountainous areas, with many hours spent on water and fuel wood collection. Despite many years of government-led efforts, gender biased attitudes continued to sharply reduce opportunities for rural women.12 Migration to urban areas became a compounding factor in the mid-1990s when active male labour in marginal areas typically migrated out most of the year, leaving women and weaker community members behind in the villages.13 The supportive government agenda to reduce gender inequalities has however remained strong. National and provincial programmes for development of women are in place and the Women’s Federation is a mass organization increasingly playing an advocacy role for women’s economic and legal rights.14 Women’s literacy programmes have retained a key role in this regard in marginal areas.15

13. Rural financial services. Towards the end of the 1990s, China’s rural financial institutions were restructured with the dissolution of Rural Credit Foundations (RCFs) as financial institutions, the separation of the RCCs from the Agricultural Bank of China and the gradual shift towards reduced subsidisation and loan volumes of priority sector lending programmes. At the time of the WGPAP design, supervisory and financial control systems were still incipient, and effective internal audit routines not adequately developed. Similarly, asset quality remained weak in the absence of an encompassing approach to debt restructuring and asset management that would also include the RCC network. Risk management systems were underdeveloped, and despite some efforts, rural financial markets were still highly regulated with interest rates determined by the central bank and not acting as a settlement price between supply and demand for financial services.16

10 Ibidem. 11 (i) World Bank, Poor Rural Community Development Project, Project Appraisal Document, Washington DC, 2005; (ii) World Bank, From Poor Areas to Poor People: China’s Evolving Poverty Reduction Agenda. An Assessment of Poverty and Inequality in China, Poverty Reduction and Economic Management Department, East Asia and Pacific Region, Washington DC, March 2009. 12 World Bank, China: Country Gender Review. East Asia Environment and Social Development Unit, Washington DC, 2002. 13 World Bank, From Poor Areas to Poor People: China’s Evolving Poverty Reduction Agenda. An Assessment of Poverty and Inequality in China, Poverty Reduction and Economic Management Department, East Asia and Pacific Region, Washington DC, March 2009. 14 World Bank, China: Country Gender Review. East Asia Environment and Social Development Unit, Washington DC, 2002. 15 UNDP and China Development Research Foundation, China Human Development Report 2007/08, Access for all: Basic public services for 1.3 billion people, Beijing, 25 July 2008. 16 R. Zander, The Way Ahead: Financial Instruments and Proposals for Continuing Reform of Rural Financial Institutions, IFAD Thematic Study on Credit in China, Workshop – Financial Services Delivery under IFAD Project Framework, Beijing, 03 July 2000.

14. Access to health and education . China made marked progress overall in social development in the rural areas. Achievements related to MDGs indicators are good at national level as evidenced in the respective reports on progress towards the MDGs.17 However, limited access to basic health care and education remained an important dimension of poverty in China’s marginal areas.18 At the end of the 1990s, access to health services was constrained not only by the lack of facilities and qualified practitioners but also by increasingly market-based services, which the poor could not afford. Many parents could not bear the cost of school attendance until recent reforms, while dilapidated schools and teachers with low or no qualifications were factors discouraging school attendance in many areas. In both rural health and education, gender disparities were assessed to be high.19 Access to maternal healthcare was markedly low until the mid-2000s. Illiteracy rates among women were still 2.6 higher than among men.20

15. Health and education policies . The national policy framework for health and education went through important changes. Compulsory 6-year education had been launched in 1996, but low central government expenditures were a driving factor in the lower school enrolments in western regions.21 Tuition fees have been gradually waived starting from 2002 and fully waived in 2006. Only miscellaneous fees remained an issue for the poorest.22 In a concomitant manner, primary schools have been centralized at administrative village and township levels by closing teaching points. This, together with higher qualifications of teachers, increases quality of primary education. Broad numbers of schoolchildren from remote villages have become boarding students, and this is seen by many as a major challenge in the current access of the poor to basic education.23 The new rural cooperative medical scheme has instituted health insurance and is in the process of being fully established in remote rural areas. The central government has recently decided to complete village health post coverage by 2011.

16. Access to water, sanitation and roads. Many households in China’s marginal rural areas still lack access to a reliable source of water for domestic use. Access to domestic water supply was one of the core indicators of absolute poverty in the 1994-2000 National Poverty Reduction Programme.24 Resource-poor regions are not only arid or semi-arid regions but also the subtropical mountain, and especially karst areas.25 Rural sanitation coverage remains today below 45 per cent on average in many of the poorer provinces.26 The road network in the mountainous poor areas of China typically extended to administrative centres only in the 1990s with low quality roads mostly made by local villagers. The

17 Ministry of Foreign Affairs of the People’s Republic of China and UNDP, China’s Progress towards the Millennium Development Goals, Beijing, October 2005 and the 2008 Report. 18 UNDP and China Development Research Foundation, China Human Development Report 2005, Beijing, 15 October 2005. 19 UNDP and China Development Research Foundation, China Human Development Report 2007/08, Access for all: Basic public services for 1.3 billion people, Beijing, 25 July 2008. 20 Ibidem. 21 World Bank, China: Overcoming Rural Poverty. Joint report of the Leading Group for Poverty Reduction, UNDP and the World Bank, Washington DC and Beijing, 2001. 22 World Bank, From Poor Areas to Poor People: China’s Evolving Poverty Reduction Agenda. An Assessment of Poverty and Inequality in China, Poverty Reduction and Economic Management Department, East Asia and Pacific Region, Washington DC, March 2009. 23 (i) Department for International Development, China Briefing Paper: Basic Education, Beijing, May 2008; (ii) World Bank, From Poor Areas to Poor People: China’s Evolving Poverty Reduction Agenda. An Assessment of Poverty and Inequality in China, Poverty Reduction and Economic Management Department, East Asia and Pacific Region, Washington DC, March 2009. 24 IFAD, China, Country Strategic Opportunities Paper (COSOP), Report No. 906-CN, Rome, February 1999. 25 World Bank, China: Overcoming Rural Poverty. Joint report of the Leading Group for Poverty Reduction, UNDP and the World Bank, Washington DC and Beijing, 2001. 26 Ministry of Foreign Affairs of the People’s Republic of China and UNDP, China’s Progress towards the Millennium Development Goals, 2008 Report, Beijing, 2008.

focus in the rural road network has shifted from township-village roads to intra-village tracks around 2000. Large-scale household biogas programmes have been launched in the early 2000s with a dual purpose of improved quality of life and environmental restoration.27 They have a broad geographical scope and include marginal areas. Water-related investments have become increasingly centred on clean water in the latest years under the national safe water programme.

17. A summary overview on the evolving policy framework and on rural development programmes relevant for the project is shown in appendix 4. It appears that there was an acceleration of policy adjustments and a growing emergence of rural development programmes, both starting with the new millennium. Thus, the WGPAP evolved within this framework of increased attention to marginal rural areas.

III. PROJECT BACKGROUND

A. Location, Rationale and Partnerships

18. The WGPAP was approved by the IFAD Executive Board in December 2000 and declared effective in March 2002. The long term goal of the project was to achieve sustainable and equitable poverty reduction in the karst area of GZAR. The overall objective was to achieve a sustainable increase in productive capacity, both on- and off-farm, and to offer increased access to economic and social resources, including financial services, education, health and social networks. WGPAP is one among five integrated rural development projects that have been launched during the period 1999- 2005 in the framework of the 1999 COSOP, but the only IFAD-funded project designed in Southwest China during that period.

19. The project area is located in southern China, in the West of GZAR, on the eastern edge of the Yunnan-Guizhou Plateau, and part of the karst region of Southwest China. It consists of two disjoint areas. The four southern counties, stretching from the provincial capital city of Nanning to the Vietnamese border, are almost fully affected by karst topography. The other six counties belong to North Guangxi bordering Guizhou Province and are a mix of karst and red soil mountains. Ethnic minority people account for 85 per cent of total population in the project area. The Zhuang represent 96 per cent of the population in the southern counties, 2 per cent being from the Han majority and 2 per cent from smaller groups, mostly Yao and Miao.28 In the northern counties, the Han account for 29 of county population and, in addition to the Zhuang, smaller groups make up 11 per cent of the population.29 Both regions only became served by the expressway network in 2008. The project rationale mentioned in the Appraisal Report 30 was based on tackling, in a coherent and integrated fashion, four major constraints, i.e.: (i) low productivity of labour and land resulting in low income, (ii) low access to social services, principally formal education and health, (iii) the incomplete set of economic infrastructure facilities, and (iv) the insufficient poverty and gender awareness and focus of main support institutions. Low productivity in Western Guangxi was due to the fact that almost 90 per cent of arable land was rainfed, and about 50 per cent on slopes. The Report of the President to IFAD’s Executive Board regarding the WGPAP noted that rainfall was unreliable, with regular droughts and floods, and that farmers were not able to take advantage of improved technology for lack of funds and training. This resulted in a food deficit for most households for a period between two to four months per year.31 The presence of significant segments of ethnic minorities with limited access

27 World Bank, Eco-farming Project, Project Appraisal Document, Washington DC, 2008. 28 The Zhuang, from the Thai-Lao ethnolinguistic family, are the larger minority nationality in Guangxi which is a Zhuang Autonomous Region. The Miao and Yao belong to the Miao-Yao linguistic family (called Hmong- Mien in Southeast Asia). 29 Official poverty statistics and ethnic minority data provided by county PMOs to evaluation team. 30 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, Rome, December 2000, paragraphs 64- 65. 31 Report and Recommendation of the President to the Executive Board on a proposed loan to the People’s Republic of China for the West Guangxi Poverty-Alleviation Project, 2000, Project Brief, vi.

to social services and the low endowment of the project area with economic infrastructure were also strong drivers for the choice of Western Guangxi as an IFAD project location.

20. IFAD-World Food Programme (WFP) partnership . The partnership between IFAD and WFP had been initiated in 1996, and the project was incorporated into WFP’s China Country Programme 2001-2005. In social development, priorities in that programme were to “enable poor households to invest in human capital through education and training” with investments in both human capacity and social infrastructure. WFP funds were disbursed in the form of food aid earmarked for these priorities. The WFP partnership also implied since 1998 the determination of target areas through vulnerability assessment and mapping (VAM), the definition of poorer households as target of food aid, the preparation of project activities on the basis of participatory village development plans (VDPs) and a shared commitment to support women in projects.32

21. IFAD-RCC partnership. IFAD has been cooperating with the RCC network since the mid- 1990s. Initial policy dialogue with this largest and most prominent rural finance provider in China focused on operational independence in borrower selection and loan approvals. Consequently, loan sanctioning and appraisals in IFAD projects were left entirely with the financial institutions and outside of the control of the PMOs or other public agencies. The policy dialogue then shifted towards the adequacy of the interest rate regime and issues of transparency of RCC networks with respect to their general portfolio and the IFAD managed funds.

22. The project was implemented in 9,878 natural villages located in 676 administrative villages within ten of the region’s most underdeveloped counties.33 It reached a total number of 896,000 beneficiaries from 239,000 households, where women accounted for 60 per cent of the total number of beneficiaries. The target group was defined at the beginning of the project, under the guidance of WFP experts and using the VAM methodology. It classified families by setting three categories of households as follows: i) category A: better-off; ii) category B: poor; and iii) category C: very poor.34

Table 1. Basic Project Data Country China Project title West Guangxi Poverty-Alleviation Project Board approval date 06 December 2000 Loan signature 20 February 2001 Loan effectiveness 21 March 2002 Project completion date 31 March 2008 Loan closing date 30 September 2008 IFAD loan US$ 30.4 million Lending terms Highly concessional Project type F – IFAD initiated and co-financed Sub-sector classification Agricultural Development Contribution of the Government US$ 53.0 million Contribution of cofinancier (WFP) US$ 11.2 million Contribution of beneficiaries US$ 11.7 million Cooperating institution United Nations Office for Project Services (UNOPS) Source: Project Appraisal Report

32 WFP, Evaluation of WFP’s Assistance to China (1979–2005), Rome, 6 January 2006. 33 The ten counties are: Long’an, Debao, Napo, Tiandeng, Tianlin, Leye, Fengshan, Tian’e, Nandan and Huanjiang. All ten counties are classified as poverty counties by the GZAR Government and all are included in the list of some 280 counties identified as poverty counties by the Chinese Government. 34 The indicators for this classification were annual grain production and net income per capita.

23. The project included five components, namely: (i) Agricultural Development; (ii) Financial Services; (iii) Social Development; (iv) Rural Infrastructure; and (v) Project Management. 35 Eight main outputs were to be produced by the project, namely: (i) responsive farmer extension services and trained farmers; (ii) stronger marketing support systems; (iii) improved land and land use; (iv) poverty and gender-sensitive rural financial services provided by the RCCs; (v) upgraded social service facilities such as village health facilities and township clinics; (vi) imparted income generation activities skills; (vii) rehabilitated or constructed rural infrastructure such as biogas systems and rural roads; and (viii) established, operational, participatory and gender-sensitive VDPs.36

B. Design Features and Process

24. Being an important part of its background, the project design features and processes need to be briefly revisited. The key stages of the project design process were: (i) the tripartite memorandum of understanding signed between the Government of GZAR, IFAD and WFP on 29 November 1999, (ii) the Formulation Report dated April 2000, (iii) the minutes of the Technical Review Committee dated 9 May 2000, (iv) the joint IFAD-WFP Appraisal Report of December 2000 and (v) IFAD Board approval in December 2000.

25. Tripartite memorandum of understanding. The memorandum of understanding set forth the fundamentals of the project on less than two pages. These included: (i) the number and location of the ten project counties, (ii) the selection of townships based on poverty and food-insecurity through the VAM system,37 (iii) the allocation of the IFAD loan in priority to the rural financial services component through the RCC network, indicating a minimum spread of 4.5 per cent, (iv) the conscious allocation of the WFP lead food-for-work (FFW) and food-for-training (FFT) tools aimed at rapid poverty alleviation, (v) the setting of minimum counterpart funding, and (vi) the mention of key project approach principles, such as participatory rural appraisals (PRAs), the formation of VIGs and the set-up of a result-oriented M&E system based on a gender analytical baseline survey.38

26. Formulation Report. The Formulation Report was the work of a team of exclusively national consultants, with the exception of the team leader. It was comprehensive and congruent with the fundamentals set forth in the memorandum of understanding referred to in paragraph 25. The formulation of the project goal was derived from the 1999 COSOP, and it remained the same through the Appraisal Report. The number and nature of the five components did not change either, with the exception of a rural electrification sub-component under rural infrastructure, which was dropped at appraisal.

27. IFAD internal quality control. The Technical Review Committee minutes 39 assessed the Formulation Report as the strongest project design document so far prepared in China, and recommended only five issues to be looked into during appraisal, namely: (i) more clarification on participatory processes, (ii) refinement of the output indicators in the logframe while acknowledging that the overall logframe was considered to provide acceptable impact indicators at the level of the purpose, (iii) a clarification that the PRA exercises allowed a free choice of beneficiaries to participate in project activities, (iv) more explicit strengthening of the operational capacity and restructuring of

35 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, Rome, December 2000. 36 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, p. 2, draft, Rome, 2009. 37 WFP/IFAD China VAM Unit, Geographical Targeting for WFP/IFAD New Project in Guangxi, Beijing, 2000. 38 WFP-IFAD, People’s Republic of China, Guangxi Zhuang Poverty-Alleviation Project, Baseline Survey Report, Nanning Fangzhi Agriculture Investment Limited Company, Nanning, November 2001. 39 IFAD, China, West Guangxi Poverty-Alleviation Project, Technical Review Committee Minutes, Rome, 9 April 2000.

RCCs and RCCUs, and (v) a mention of the findings of the Environmental Screening and Scoping Note (ESSN).40

28. Appraisal Report. The response of the Appraisal Report to the five recommendations given by the Technical Review Committee (see paragraph 27) was mixed. With regards to participatory processes, this was complied with.41 The output indicators in the logframe were indeed revised, mostly by lowering the threshold values to some extent. A discrepancy however persisted between the logframe at appraisal and the logframe in the PCR concerning the number of loans. The former set these at 680,000 and 284,000 for seasonal and investment loans for income generating activities, respectively, while the latter retained a total target number of loans of 68,000, this probably having been the overdue correction of a typing error of one order of magnitude (see table 3). With regards to PRAs being used to facilitate a free choice of crops and mixed farming schemes by the beneficiaries, the mission did not find any explicit notion in the Appraisal Report. The RCC and RCCU strengthening aspect was taken care of to some extent while an explicit reference to the ESSN cannot be found in the Appraisal Report (see also paragraph 61). This being said, the Appraisal Report, totalling more than 400 pages in two volumes, was a document that went into painstaking details in terms of implementation schedules, cost tables and also of recommending M&E criteria, indicators and methods for the implementation period.

29. A swift exercise without disruptions. Spanning a period of slightly over 12 months between the initial memorandum of understanding between the parties and board approval, the project design process can be considered relatively short, and indeed appears seamless in hindsight. There were no major shifts in paradigms, and it is noteworthy that already the memorandum of understanding of November 1999 determined all major conceptual cornerstones of the project (see paragraph 25). Both the Formulation and Appraisal Report were prepared by Chinese consultants, with the exception of the respective team leaders and the Country Programme Manager in the case of the Appraisal Report.

IV. IMPLEMENTATION RESULTS

30. Project implementation in terms of financial progress including contingencies, as reported in the PCR, is shown in appendix 5. Financial target adjustments at mid-term review (MTR) were minimal, and the cumulative financial progress at project completion was close to the adjusted overall target. Appendix 6 indicates the project costs at appraisal including contingencies by funding source. The Government contributed over 50 per cent to project cost and bore the brunt in hardware related components and sub-components, such as irrigation, dry land improvement, tree development including biogas tanks, drinking water systems and rural roads. The Government’s share was also more than two times higher than IFAD’s contribution in the project management component. IFAD resources focused on crop and livestock development, the rural financial services and social development components. WFP intervened in all components and sub-components where FFW and FFT played a major role while the beneficiaries were substantially involved in crop and livestock development and rural financial services. This reveals a selective cost sharing pattern based on comparative advantages and installed capacities.

31. Project output indicators versus results achieved at respective county level. The fact that the county and township PMOs were an integral part of the respective local governments made it possible, across the five components, to compare the project output indicators with the county level achievements in the same subject areas. This is an uncommon feature of development projects at large, and the evaluation mission took the opportunity to make these comparisons in the sections below related to implementation results. This will allow placing the project implementation results in the wider context of efforts undertaken by the national and provincial governments in the ten counties served by the WGPAP.

40 IFAD, China, West Guangxi Poverty-Alleviation Project, Environmental Screening & Scoping Note, Rome, 14 April 2000. 41 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, Rome, December 2000, paragraphs 18, 125.

A. Agricultural Development

32. The implementation results of the agricultural development component (41.6 per cent of base cost) are captured in table 2. The respective output indicators are the ones found in the logical framework of the Appraisal Report. All targets of the agricultural development component have been met or exceeded. The comparison of project with county achievements reveals that the relative contribution of the project was highest in terms of areas improved by irrigation schemes and soil management techniques. On the other hand, the project’s contribution to plantation of trees on common land was less than a third compared to county performance.

33. Crop experiments. The crop experiments on farmers’ fields were the starting point of a promotional sequence, followed by on-farm demonstration and finally extension and adoption. All county PMOs have prepared summaries on crop experiments conducted under the project, and five of the counties also reported results on trials made with farm animals. More than 95 per cent of the crop experiments concerned annual crops such as maize, rice and soybean, and were focusing on varietal comparisons for the most part. The summaries include: experimental areas, years, yields obtained in comparison to local varieties, and mean yield differences. While the areas used for crop experiments were more or less equal among the ten project counties, the “technology leaps”, expressed by the average of productivity difference between the new and local technology for each trial series, vary widely among counties, in the range between 4 per cent and 38 per cent. These differences are substantial as they reflect experimental activities over six years. An analysis of the available summaries reveals that productivity differences are negatively correlated to their variability within a given county. In other words, a county reporting a high average productivity progress is likely to have obtained this with less variation between the individual trial results than counties with low overall productivity progress. One of the reasons may have been highly varying trial layouts and conditions between counties. Data reported for farm animals, mostly based on pig or ox feeding and other animal management improvements over local techniques, indicate average productivity progress between 10 and 34 per cent. Data is too scarce to calculate the correlation between productivity progress obtained and its variability. As noted above, the crop experiments were heavily concentrated on varietal assessments annual crops, mostly cereals, while the on-farm demonstrations, as described in paragraph 34, focused on agronomy and animal husbandry, or mixed farming systems.

Table 2. Project and County Level Output Indicators, Agricultural Development Per Cent of Project Indicator No Project Output Project Project Total County Project Target Contribution and Units Indicator Narrative Targets Achievements Achievements Achievement % 1.1.1. Crop experiments 2,250 2,374 106% 5,050 47% mu On-farm 1.1.2. demonstrations, on land 92,000 110,240 a/ 120% 192,860 57% mu of poor households 1.1.3. Farmer training days 550,000 615,262 112% 1,251,475 49% days held, gender data kept Farmer professional 1.2 associations (FPAs) for 5 5 100% 33 15% FPAs specialty crops 2.1 mu of irrigation schemes 199,000 205,305 103% 341,660 60% mu upgraded or built 2.1 Tanks built n.a. 20,186 n.a. 52,501 38% Numbers 2.1.1 Km of canals n.a. 554 n.a. 1,441 38% Km 2.2.2 mu improved by soil 91,000 91,200 100% 158,259 58% mu management techniques mu individual plots 2.3 planted with economic 254,000 256,265 101% 565,285 45% mu crops mu common land 2.4 planted with fuel and/or 35,000 34,957 100% 109,095 32% mu economic trees mu degraded forest 2.5 regenerated through 1,300,000 1,377,000 106% 3,360,015 41% mu closure a/ 38.7 per cent on land of category B households and 61.3 per cent on land of category C households Source: PCR and additional data from the provincial PMO on county-level achievements.

34. From on-farm demonstrations to extension and technology adoption. Again, the documentation made available by the PMO to the mission provides an evidence trail of productivity progress obtained both at the level of on-farm demonstrations and extension of the demonstrated technologies. Areas devoted to on-farm demonstrations were in total 450,000 mu over project life, i.e. four times the area indicated in table 2 above. The probable reason of this discrepancy is that several technologies may have been demonstrated over the years on the same plot, e.g. diversified crops on land improved by terracing. Figure 1 infers that by far the biggest areas were devoted to land conservation and oil seeds/trees technologies, both in terms of on-farm demonstration and extension. Figure 2, related to gross output value generated by on-farm demonstrations and extension of the demonstrated technologies, shows that the two technology categories of mixed farming and crop diversification have generated the biggest leaps in gross output value addition. The field visits confirm these trends, and the evaluation mission commends the project’s record-keeping practice that is indeed providing evidence-based insight that is uncommon in comparable projects. The provincial PMO is finalizing a comprehensive compendium on 46 agricultural technologies generated and promoted under the project, which will be referred to in chapter VII (sustainability and innovation).

Figure 1. WGPAP - From Demonstration Figure 2. WGPAP - From Demonstration to to Extension, Areas and Animals Extension, Incremental Output Values

800'000 250'000'000 700'000 200'000'000 600'000 150'000'000 CNY 500'000 100'000'000 mu 400'000 50'000'000 300'000 0 200'000 Demonstration, Extension, incremental output incremental output 100'000 value (CNY) value (CNY) 0 Origin of Gross Output Values Demonstration area Extension area (mu/heads) (mu/heads) Organic farming technologies (plant production; mu)

Areas/Heads Involved Organic farming technologies (animal production; Organic farming technologies (plant production; mu) heads) Mixed farming technologies (plants-animals) Organic farming technologies (animal production; heads) Energy crop technologies (oil seeds/ oil trees; mu) Energy crop technologies (oil seeds/ oil trees; mu) Land conservation technologies (mu) Land conservation technologies (mu) Feature agriculture (crop diversification; mu) Mixed farming technologies (area/heads not shown) Feature agriculture (crop diversification; mu)

Source: Provincial PMO

35. Farmer training. Broad-based farmer training was the third ingredient in the extension strategy and complementary to crop experiments and demonstrations. Training activities were carried out under two distinct methods, i.e. the so-called “farmhouse classroom” and the “work-for-study” approach. The first method built on central households endowed with additional technical training and guidance by the county and township PMOs, enabling them to train fellow farmers on the spot and in their own language. More than 60 per cent of the trainees were women. The second method was an interactive training approach where household members would work for state owned enterprises, plantations or farms for a period of time. The evaluation mission collected positive feedbacks from trainees in a successful enterprise growing organic tea in Leye county that served as a training ground for individual tea growers.42

36. Farmer Professional Associations (FPAs). As shown in table 2, the project promoted the set- up of five FPAs, out of a total of 33 in the ten project counties. The mission visited two of them, the Liu Hua Pheasant Association in Tien’e county and the Napo Pig Industry Association. The latter has grown from 42 (2002) to 1,300 members now, and offers comprehensive services to its members, such as artificial insemination, feed, medicines and marketing support. The latter built on the initiative of a returnee migrant worker who found out, in 2001, how to artificially incubate eggs from wild pheasants in Tien’e county, which is reputed to be place of origin of pheasants in China. The Liu Hua Pheasants Association counts 182 members, and 243 households reared pheasants in Tien’e at the end of 2008. Both associations ventured into areas with high demand and opportunity to add value, not unlike the remaining three associations formed under the project, specializing in pepper, star anise and oil seeds and trees. All five FPAs together have pooled more than 4,800 households around such agro- marketing value chains.

37. Irrigation and soil improvements. The sub-components related to land improvement and conservation was not only exhibiting the largest share in area covered by the project, but these were also absorbing 65 per cent of the overall cost of the agricultural development component. The data in table 2 infer that the project apparently supported irrigation schemes that required comparatively less investments in canals and tanks than the county mainstream schemes (60 per cent project contribution

42 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, appendix 9, draft, Rome, 2009.

to newly irrigated land area versus only 38 per cent in terms of tanks and km of canals built across all ten counties in total).

Irrigation tank and paddy fields Soil improvement (terracing and soil deepening) Source: Provincial PMO Source: Ernst Schaltegger

38. Plantations and forestry. Land use change via plantations of economic and fuel trees and forest closure was another output category supported by the project. As table 2 shows, the project attained or overshot the corresponding indicators, but contributed to overall county level outputs with less than 45 per cent. The achievements related to forest closures were however possible only in conjunction with the promotion of biogas tanks (see paragraph 48), which provided the required substitute to fuel wood.

B. Rural Financial Services

39. The rural financial services component (22 per cent of base cost) marked by far the highest IFAD contribution within its loan allocation to the project, i.e. 49 per cent at appraisal. In terms of financial progress, this component closed with over 100 per cent compliance with targets while table 3 below informs on the status of the two component relevant indicators. The reason why the number of loans is smaller than the project target is that the individual loan ceiling has been increased at mid- term review. Close to 35 per cent of the RCC borrowers were women.

Table 3. Project and County Level Output Indicators, Rural Financial Services Per cent of Project Indicator N o Project Output Project Project Total C ounty Project Target Contribution and Units Indicator Narrative Targets Achievements Achievements Achievement % Small loans given to 3.1 resource poor 68,000 59,227 87% 469,126 13% Numbers households for seasonal agriculture 3.2 All RCC staff trained All 782 n.a. 1,080 72% Numbers Source: PCR and additional data from the provincial PMO on county-level achievements.

40. According to the PCR, the extension of loans to poor and very poor farmers in the project area showed a marked increase. By end December 2007, the average amount of credit per capita made available in the project area was CNY 1,723, roughly four times higher than it was prior to the project. Data on the increase of deposit and associated financial services were unfortunately not collected.

41. Similarly, 4,119 women obtained micro loans from WFs for income generating activities. The mission noted that their loan utilization seemed to be less restricted to farming only, a noticeable

difference to the RCC loan portfolio. Spot checks made by the mission indicated that some WF loans were used by borrowing women for income generating activities at household level in a wider sense.

42. There was a significant extent of training to RCC and RCCU staff in different areas related to microfinance to individuals. Interviews indicated that these trainings were assessed as significant value added to the existing skill sets and immediately usable for work. RCC and RCCU staff were invited to attend farmer demonstrations and crop and livestock trainings as well. The regular attendance of many of the RCC and RCCU staff improved their own technical and management skills, which helped them for their credit appraisal of small project loans.

C. Social Development

43. Health . The county health bureaus trained village practitioners to the tune of 1.4 person per village on average. They launched, jointly with the WF, a complete set of maternal and child health prevention activities. Village health posts were set up in 35 per cent of the project villages. The township clinic program was completed at 90 per cent, with 10 clinics completed instead of 12. Township clinics were upgraded to standard quality facilities. According to the M&E records, two persons per village attended annual training for village practitioners, and two additional practitioners were trained in each township. Six-month refresher courses were organized for those with low qualifications. Close to 165,000 households in 98 per cent of project villages took part in a hygiene awareness campaign that widely disseminated illustrated brochures. Immunization campaigns were carried out through two trained practitioners in each of the 74 townships, and a prenatal check system was initiated. It covered 100 per cent of project villages with an average 88 births per village over six years. Training one female village doctor in each village was the only initial target with a low achievement rate, at 50 per cent.

44. Education . The county education bureaus and PMOs have gradually adjusted the facilities and equipment foreseen at design stage. A total of 1,400 classrooms or dormitory rooms were built or rehabilitated in 38 per cent of project villages. One third of the rooms were teaching outposts, compared to the appraisal target of 70 per cent minimum. Half of the counties built a total of 110 dormitory rooms. Tables, chairs, black boards and dormitory berths were provided to facilities funded by the project or other facilities lacking such equipment. Several new schools with no clean water access benefited from a new water cistern. The teachers attended training both in multi-grade teaching training and in standard Chinese as initially planned. The latter was a precondition to the qualification of all teachers that took place at the end of the training period. The 2,073 teachers trained are equivalent to half of primary school teachers in the project area.43 The schoolmasters of 100 per cent of villages in the project area undertook management training with qualified trainers.

45. Women’s Federation Programme . The project allowed a many-fold increase in each project township of what was an emerging set of WF activities. These activities supported the access of poorer households to primary education, skill development and small loans, giving priority to girls and women. According to the project M&E documentation, about 27,000 school children received grants during one to two years. Part of them were out of school. Around 39,000 adults attended evening courses in village schools with qualified trainers providing both literacy skills and farming skills, and received illustrated brochures. These courses were organized jointly with primary school teachers and agricultural technicians. Income-generation courses were held for more than 72,000 women, mostly in agriculture. A four-wheel drive car, a computer and a photocopier was planned for each county WF, an unprecedented improvement in their level of equipment. Eighty six per cent of the equipment value was completed. A meeting room was built in 60 townships instead of the 74 foreseen, often as a stand- alone training centre.

43 Based on five teachers per administrative village, the average figure in the five primary schools visited.

Table 4. Project and County Level Output Indicators, Social Development Per cent of Project Indicator No Project Output Project Project Project Total County Contribution and Units Indicator N arrative Targets Achievements Target Achievements % Achievement 1,320 classes 4.1.1 constructed/or 1,320 1,405 106% 2,957 48% Numbers upgraded Teachers trained in 4.1.2 management and 2,000 2,073 104% 4,942 42% Numbers teaching in Mandarin Chinese Pupil-years of drop- 4.1.3 outs' presence in 39,000 39,783 102% 51,605 77% Pupil-years primary schools Village health 4.1.4 facilities upgraded 225 238 106% 506 47% Numbers or built; practitioners trained Township clinics 4.1.5 upgraded; all 12 10 83% 72 14% Numbers practitioners trained Maternity wards 4.1.5 upgraded; all 34 30 88% 148 20% Numbers practitioners trained 4.2.1 Illiterates trained 30,000 38,692 129% 48,868 79% Numbers Semi-literates given 4.2.2 skills for income- 57,000 72,177 127% 126,156 57% Numbers generating activities Source: PCR and additional data from the provincial PMO on county-level achievements.

D. Infrastructure

46. Access roads . The project funded as foreseen an average 19 km of all-season roads in each county. Rural road construction improved or opened access to three villages per county on average.44 In some cases these were the first good quality roads linking townships to administrative village centres. Some counties, which had already a better road network, used the project to build rural tracks to natural villages, also all season but more narrow. These roads were among the first rural transportation works in the counties to be mostly completed with mechanized equipment.

47. Domestic water supply . Close to 16,000 water cisterns were constructed as foreseen for drinking water for human and livestock consumption. These are on top of the 20,000 cisterns primarily built for irrigation. The project has also built 137 gravity systems, serving around one natural village each. Most of the cisterns were for individual households but a small number of collective cisterns, typically for natural villages, have also been built.

48. Biogas . Although the promotion of biogas tanks was included in what the project called tree development (see paragraph 30 and table 2, positions 2.4 and 2.5), the targets and achievements regarding biogas tanks were reported separately in the PCR and shown on table 5. The project contributed to the implementation of the provincial biogas programme, immediately after this programme was launched in 2001. The fact that biogas promotion was a government priority underpins the relatively modest project contribution of 14 per cent across all ten project counties. Household biogas tanks were built in the form of a standard package used throughout China that includes improved pig sheds and sanitation. More than 22,600 household biogas tanks were accounted as project expenditure. The mid-term review mission noted that the project acted as a pro-active demonstration vehicle of biogas.

44 M&E system data in PCR annex. The PCR reports 76 villages.

Table 5. Project and County Level Output Indicators, Infrastructure Per cent of Project Output Project Indicator No Project Project Project Total County Indicator Contribution and Units Targets Achievements Target Achievements Narrative % Achievement Beneficiary 5.1 households with 24,000 32,513 135% 90,887 36% Numbers new domestic water supplies Km rural roads 5.2 upgraded to class 193 192 99% 1,256 15% Km 4 Household 5.3 biogas systems 22,000 22,649 103% 158,970 14% Numbers installed Source: PCR and additional data from the provincial PMO on county-level achievements.

E. Project Management

49. Targeting method . All administrative villages within a given project townships were selected as project villages except for eight larger township government seats. Natural villages were not ranked into poverty categories. On the other hand, the household poverty ranking system was fully used for targeting as well as monitoring purposes. It was operational for the delivery of school drop-out grants, WFP rations, targeting of RCC credits and the persons selected for training or demonstrations.

50. Village implementation groups and development plans. The VIGs established in each village remained active throughout the project in the coordination and monitoring of household activities and income levels. The county PMOs and technical bureaus generally decided the actual sets of activities on the basis of the preferences expressed in the VDPs prepared by the VIGs. These were five-year master plans and no provision for annual updates were made. Thus, the VDPs were not updated after the initial formulation exercise, in contrast to what was foreseen at appraisal.

Table 6. Project and County Level Output Indicators, Project Management Project Per cent of Project Indicator N o Output Project Project Project Total County Contribution and Units Indicator Targets Achievements Target Achievements % Narrative Achievement Overall village plans 6.1 drafted in 684 684 100% 1,008 68% Numbers each administrative village Source: PCR and additional data from the provincial PMO on county-level achievements.

51. Food aid . Food aid was delivered as foreseen in the form of FFW or FFT. Household labour was mostly used for individual investments in water cistern, and for community infrastructure including irrigation canals, forestation and minor works in road construction. As a result, FFW was largely an incentive for broad participation of households as much as FFT, and not a source of employment for poorer households as initially envisioned. Termination of WFP operations in 2005 resulted in a gap of CNY 10.4 million or 13 per cent in the WFP plan of 80,000 tons of food aid. By then, 97 per cent of the FFW expenditure had already been completed. The reduction affected training activities with the exception of the health awareness campaign. Seventy to 90 per cent of the appraisal targets have been completed according to the progress reports. Activities resumed outside the framework of the project, but achievement data is not available.

V. PROJECT PERFORMANCE

A. Relevance

52. Consistency with Government and IFAD strategies. The review of the ample evidence available on policies (paragraphs 11-17) leaves little doubt: The WGPAP design was coherent with government policies and strategies valid at the time of project design, in terms of area and population targeting, including ethnic minorities, making gender a visible concept and rural financial services a key dimension, and because of the participatory and multidimensional approach. The WGPAP was designed in the wake of the COSOP publication in 1999, which noted that the relevant IFAD and WFP experiences concurred with the policy and strategic lines of the Government.45 The 1999 China COSOP was remarkable in that it brought forward a joint IFAD-WFP strategic framework, which subsequently substantiated into the WGPAP design. Government policy changes after project inception in 2002 did not affect the relevance, although the phase-out of WFP as a provider of food aid did lead to the completion of the components that required food aid as key input by 2006. IFAD’s Regional Strategy Paper for Asia and the Pacific 46 was published when the WGPAP became effective. Thus, it was not a conceptual driving force for the WGPAP design, but coincided to some extent with key traits of the WGPAP, e.g. the inclusion of marginal areas, ethnic minorities and women as agents of social transformation. A further reaching ambition of the regional strategy was to shift “from problem projects to projects and programs with strategic potential to influence policies and to generate ripple effects”. The 2005 COSOP, this time without inclusion of WFP, stated that “the Fund’s niche is defined by emphasizing its catalytic role and developing innovative, strategic pilot programmes with governments and other donors for scaling up. The development of future IFAD poverty reduction programmes will be driven by two strategic thrusts: access and innovation”.47 Thus, there was a paradigm shift between the 1999 and 2005 COSOP, away from a rather general poverty-alleviation focus toward innovative, strategic pilot programmes. Both these strategic documents expressed considerable conceptual ambitions that deserve to be revisited in chapter VIII on performance of partners.

53. Multidimensional approach. The WGPAP was designed as a multisectoral, service-oriented project, comprising roads, domestic water supply and sanitation, and health and education around the main component, agricultural development. It therefore included dimensions that were highly relevant to the needs of the poor in the project area at project start. The baseline survey in 2001 confirmed the magnitude of poverty in all these terms, while the Water, Economy Investment, Learning and Assessment Indicator (WEILAI) Report in 2007 48 acknowledged that poverty was a multidimensional phenomenon that logically also required multidimensional tools for project planning, monitoring and evaluation, thus indirectly confirming the relevance of the WGPAP dimension and component mix chosen.

54. Poverty reduction in the karst . The Government’s rationale in the selection of the project area was three-fold: all counties were nationally designated poor counties, they were fully or partly composed of karst areas and they were ethnic minorities areas.49 The project was designed as a karst initiative, consistent with government policies, and came at the right time in this respect as the Guangxi Government had just decided to undertake a series of rural infrastructure projects in the karst region. There was however a gap between the goal statements and the conditions of the actual project

45 IFAD, China, Country Strategic Opportunities Paper (COSOP), Report No. 906-CN, Rome, February 1999, paragraph 32. 46 IFAD, Regional Strategy Paper Asia and the Pacific, Asia and the Pacific Division, Project Management Department, Rome, March 2002. 47 IFAD, People’s Republic of China, Country Strategic Opportunities Paper, Rome, 13 December 2005. 48 IFAD, Alisdair Cohen, WEILAI and Rural China: Development and application of a multidimensional tool for project planning, monitoring and evaluation, Technical Report prepared for the Asia and the Pacific Division, Rome, December 2007. 49 Guangxi project leading group director interview in the project’s TV series.

area in North Guangxi, which included a fair proportion of red soil mountains. On the other hand, it is true that ethnic minorities in the six northern project counties are predominantly concentrated on red soil mountains.

55. Natural villages in targeting strategy . The 2000-2010 Government Poverty Reduction Strategy offered opportunities to separately target natural villages. It appears that this has not been seized at design stage, except for schools and food aid. The mentioned programme has also created a full monitoring system of access to services at natural village level, which the WGPAP had not accessed.50 The project instead chose a targeting method whereby all administrative villages within each project township targeted as poor were included. This was highly consistent with the purpose of effective service delivery with broader impact. In education, the Appraisal Report noted that focusing on building teaching points in remote natural villages might conflict with government policies. This is precisely what happened when primary schools became centralized, but this policy discrepancy was not reviewed in detail.

56. Gender strategy . The project had a three-pronged strategy by: (a) mainstreaming attention to gender from all service providers, especially agricultural extension, (b) developing a full maternal and child healthcare system, and (c) retaining a specific subcomponent for women capacity building and girl education under the management of the WF. This was highly relevant to the needs of poor women. Agricultural work and decisions are largely carried out by women. Maternal healthcare was almost non existent in 2000, and the gender gap in literacy and primary education was substantial. Hence, the WGPAP strategy was also consistent with the government programme for women.

57. Participatory planning . Prior to the establishment of VDPs, PRA methods were used, a principle agreed upon between the provincial government and IFAD even before project formulation.51 The adoption of this participatory approach was expected to strengthen ownership by the key stakeholders. Chapter VI on rural poverty impact (specifically paragraphs 104-105) will shed light on whether this has happened.

58. Inclusion of rural financial services . At the time of appraisal, the RCCUs were adequately liquid and did not require external lines of credit for general purposes. However, they would have not been willing to expose their own and depositor–mobilized funds by financing a client group that was considered risky in the RCCU assessment at that point in time. Moreover, the term structure of the funding sources within the RCCU network was strongly lopsided towards short-term maturities. A line of credit that far exceeded the longest available period for term deposits within the RCCU system (five years) therefore strengthened the balance sheet and enabled RCCUs to direct more loan resources to investment finance, rather than just restricting lending to short term agricultural input finance. Hence, the inclusion of rural financial services into the component mix was relevant in the case of the WGPAP. It was not the size of the IFAD contribution, admittedly modest in comparison with other international financial institutions, but its systemic added value in terms of quality that made a difference.

59. Learning from other projects . The design of the WGPAP was clearly learning from previous IFAD-WFP projects in China, such as the Southwest Anhui Integrated Agricultural Development Project 52 and the Qinling Mountain Area Poverty-Alleviation Project 53 of which Completion or Interim Evaluation Reports are available. The project has also integrated experience from at least two relevant projects funded by the World Bank: the Southwest Poverty Reduction Project 54 and the Agricultural

50 Several supervision missions recommended that coordination in monitoring be enhanced. 51 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, 2000, appendix 1. 52 International Fund for Agricultural Development, People’s Republic of China, Southwest Anhui Integrated Agricultural Development Project, Completion Evaluation, Report No. 1785-CN, Rome, September 2006. 53 International Fund for Agricultural Development, People’s Republic of China, Qinling Mountain Area Poverty-Alleviation Project, Interim Evaluation, draft, Rome, 25 March 2009. 54 World Bank, Project Completion Report of the Southwest Poverty Reduction Project, Report no. 26132, Washington DC, 30 June 2003.

Support Services Project.55 The former was reviewed in detail during the preparation of the WGPAP and the latter was being implemented by the same provincial PMO. Both were multi-sectoral rural development projects with an explicit poverty reduction focus.

60. Logical framework . The logframe contained in the Appraisal Report 56 had the advantage of being relatively concise and endowed with a strong intrinsic logic, by making the underlying development hypothesis from outputs to the goal of the project clearly understandable. As is often the case in many project designs, the objectively verifiable indicators tend to be more precisely formulated at output than at purpose and goal level. In the case of the WGPAP logframe, this was avoided with only two exceptions (reduction of preventable disease incidence and number of enterprise start-ups) where no indicator magnitudes were set. At output level, some few indicators were blurred. The indicator related to number of classes built did not allow to identify the actual use of the investment (teaching points, classrooms and/or dormitories). Likewise, collective water cisterns, which provide a much lower level of service, were lumped together with individual ones, and the number of villages served by roads was not highlighted as an output, but was monitored outside of the logical framework. Despite these shortcomings, the logframe was a useful reference framework for the project’s M&E system. The column related to important assumptions, a good indicator of risk awareness, or aversion, contains only a few assumptions, and none of these was a killer assumption.

61. Environmental and social impact assessment . While the design of the WGPAP was explicitly conscious of environmental and social issues, neither the ESSN nor both the Appraisal Report and the previous Formulation Report, mentioned the Long Tan hydroelectric dam project located at the upper reaches of the Pearl River in Tien’e county. It was, at its design, the world's largest roller-compacted concrete dam, displacing a total of 80,000 people, of which 30,000 from Tien’e county. According to the Tien’e county government, many of the displaced households have benefited from the project, especially as members of FPAs that focused on value addition via animal production. However, a utility project of this dimension, with considerable impacts on people in the project area, should have been taken into account at project design.

62. Summing up. The overall relevance of the project would have deserved the maximum rate of six. Considering the above mentioned design shortcomings, a rating of five (satisfactory) appears to be justified.

B. Effectiveness

63. The effectiveness criterion focuses on the question to what extent the project objective has been achieved. The width and depth of such achievement evidently builds on the outputs generated by the project, which are described in chapter IV related to implementation results. Thus, the evaluation of effectiveness evolves around the essence of the project objective, which was, as mentioned in paragraph 18, to achieve a sustainable increase in productive capacity, both on- and off-farm, and to offer increased access to economic and social resources, including financial services, education, health and social networks. A detailed assessment of the degree of achieving this objective must rely on objectively verifiable indicators, as determined by the logframe. An analysis of these indicators follows in paragraphs 64-68.

64. Per capita grain availability. With over 49 kg of per capita grain availability increase by 2007 versus a target of 45 kg, the project was effective in delivering this food security relevant target. The PCR 57 further substantiates this achievement, by relating it to the incremental household level crop, livestock and fruit tree product outputs, of 95, 78 and 55 per cent, respectively. Food security indeed increased as more than 95 per cent of the surveyed households reported food sufficiency for more than 12 months. All these trends refer to 2001 baselines, and the mission concurs with the PCR that these

55 Completion Report not available. 56 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, 2000, appendix 3. 57 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, Draft, Rome, 2009, table 4.

are a logical consequence of agricultural productivity leaps achieved under the agricultural development component (see paragraphs 32-38).

65. Per capita income increase. In the logframe, the per capita income increase between 2000 and 2007was assumed to be CNY 300, expressed in 1999 values. This increase was effectively CNY 1,045 by the end of 2007, but it is not clear whether this difference is expressed in constant 1999 CNY. In any case, the effective per capita income growth overshot the target significantly. The PCR complements this assessment by indicating that not only household incomes grew over project life, but also household assets. Household incomes and assets are a poverty relevant impact domain and will be revisited in chapter VI.

66. Reduced incidence of preventable diseases. The logframe does not specify which preventable diseases were to be measured against this indicator, for which no magnitude was set at appraisal. The PCR chose to report on the trends, over project life, concerning the incidence of diarrhoea and of the diseases targeted by the MDG 6 (combat HIV/AIDS, Malaria and other diseases), exhibiting incidence decreases of 5 and 2 per cent, respectively. The mission complements this information with an aggregated indicator of recorded mortalities (maternal, infant and under five), which can also be the consequence of preventable diseases. The Project Participatory Assessment (PIA) 58 attached to the PCR presents data that allow to formulate such an aggregate indicator. Between 2000 and 2008, these aggregated mortalities show an average decrease of 45 per cent, with a coefficient of variation of 40 per cent only between the counties. The evaluation mission attributes this success to the synergies between village health posts and townships clinics serving as referents and providing maternity services, and the pervasive health training activities (see figure 3). Delivery in township clinics has widely been adopted, reducing maternal mortality and improving health of mother and child.59 There is also anecdotic evidence: in one Han project village, only one woman went to the clinic for delivery in 2001, and only one did not in 2008. The project contributed specifically through a prenatal check campaign financed by IFAD.60 Another key factor were the health awareness manuals, which were provided in very high numbers and are still visibly well kept to date by the beneficiaries. They include straightforward illustrated messages in hygiene, AIDS and other sexually transmitted diseases. Messages have been repeated in trainers’ training programmes and in literacy manuals. Immunization and health awareness campaigns have reportedly been adopted in the prevention programmes.

67. Reduction of school drop-outs, mostly girls. The PCR 61 provides evidence that the target of reducing primary school drop-out rates by 50 per cent, compared to the pre-project status, has been exceeded. The school-age child enrolment reached 99.5 per cent by the end of 2007, and girl and boy drop-out rates decreased by 95.8 and 89.3 per cent, respectively. The project has contributed to this achievement by rebuilding and equipping dilapidated schools, funding teacher training and providing grants to poorer households. These “school drop-out grants” have subsidised education costs of 11 per cent of project households whereby 60 per cent of beneficiaries were girls.

68. Enterprise start-ups, in particular women. The noteworthy feature of this indicator is that: (i) no magnitude was given at project design and (ii) it assumed that its compliance would be measured by the number of micro-credits extended by the RCC network and the WF’s micro-credit facility. The PCR 62 therefore mentions the 52,255 RCC loans that were used specifically for income- generating activities, with 34.7 per cent directed to women, plus the 4,119 micro-credits from the WF micro-credit facility for the same purpose. It could be argued that the extension of such credit facilities

58 Ibidem, Working Paper 1, Participatory Impact Assessment. 59 Rates of hospital delivery went up from 60 per cent in 2001 to more than 90 per cent in 2008 according to the PCR, or from 48 per cent in 2000 to 87 per cent according to county data in the PIA. 60 This is an interesting example of an activity that was not included in regular government programmes but achieved an attitude change towards delivery in hospital. 61 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report (PCR), Draft, Rome, 2009, table 4. 62 Ibidem.

was not automatically equivalent to successful enterprise start-ups but, on the other hand, the relatively low default rates observed would indeed suggest that tangible enterprise start-ups may have been promoted thanks to micro-credit. The project thus was effective in directing the attention of the RCC network to smaller farmers in the lower and lowest income brackets of their village economies. This is probably the greatest and most durable success of the financial services component of the WGPAP, namely the demonstration, to the RCCUs in Guanxi, that with specialized loan appraisal and credit analysis procedures, and through prudent pre-selection of loan applicants based on their training and skills levels, providing financial services to rural poor people was both profitable and strategic.

69. Rural infrastructure. There was no explicit indicator at purpose level referring to rural infrastructure, but “to offer increased access to economic and social resources” would not have been possible without an upgrading of rural infrastructure to which the project contributed, except for electricity (see table 7 below and table 5).

Table 7. Trends in Infrastructure Coverage in Project Administrative Villages Lack of Lack of road No Lack of Per cent of project administrative village with: electricity access a/ biogas domestic water 2000 10% 18% 46% 51% 2005 3% 9% 16% 31% 2008 0% 2% 3% 20% Average project contribution at county level 0% 15% 14% 36% (from table 5) a/ The road indicator does not differentiate between all season roads and other roads. Source: Calculated from county PMO data in the Participatory Impact Assessment.

70. Effectiveness and underlying reasons. The achievements mentioned above suggest that all relevant dimensions of the project objective were effectively met. There are a host of reasons why the effectiveness of the WGPAP was high, and these will be further explored in particular in chapter VII on performance of partners. A good measure of assessing one ingredient causing effectiveness is highlighting the geographic coverage achieved by the project across all relevant activities carried out. This is shown in figure 3 below. Project activities that covered more than 90 per cent of project administrative villages, and concurrently more than 70 per cent of associated natural villages were, in descending order, farmers’ technical training, care of expecting mothers and health training. A second set of underlying causes for effectiveness were the stable political context and the overall economic growth of China during project life. There were no overwhelming factors in theses domains that could have depressed effectiveness to a significant extent.

Figure 3. Village Coverage per Project Activity

Credit activities

Care of expectants

Dropout support

School buildings

Clinics and hospital

Biogas systems

Drinking water tanks

Drinking water systems

Irrigation tanks Project activity Canal construction

Hills closure

Ecological planting Per cent natural village coverage Per cent administrative village coverage Village roads

Land improvement

Livestock demonstrations

Crop demonstrations

Health training

Literacy training

Farmers' technical training

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Per cent coverage

Source: PCR, appendix 6.

71. Effectiveness rating. Given the extent of indicator achievement or over-achievement, a rating for effectiveness of six (highly satisfactory) is given.

C. Efficiency

72. A first glance. A generic assessment of efficiency can be made by comparing the magnitude of the achieved project purpose with the effective costs incurred. The two facts that: (i) the project objective was achieved in excess of the targets set at appraisal, but (ii) practically at the cost originally estimated, would infer that project efficiency was high. In addition, the following considerations have been taken into account in the following paragraphs 73-77.

73. Monitoring of unit costs. The provincial PMO prepared 14 component and sub-component related spreadsheets 63 with the aim to monitor the differences between targeted and effective quantities

63 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, draft, Rome, 2009, appendix 11.

and unit costs of all cost items where applicable. The percentage range of higher or lower than targeted unit costs at appraisal span from plus 32 per cent (county PMOs) and minus 76 per cent in the case of the irrigation sub-component. Unit cost overruns mainly occurred with vehicles and hospital equipment while computers were generally procured at lower than estimated costs at appraisal. The case of the irrigation sub-component, however, must have been caused by a typing error, because the negative unit cost difference is mainly due to an exaggerated unit cost at appraisal for engineering fees, for check dam rehabilitation only, at several orders of magnitude below the stipulated 7 per cent flat rate.64 Given the high variability of design input intensities, such flat engineering fees may have been justified in the case of irrigation canals and roads, but less for land improvements and drinking water and biogas tanks, the latter of which were standard designs. Despite this shortcoming, the WGPAP’s monitoring tool for unit costs and quantities is highly commendable, and it indicates that costs were under tight control (see paragraph 72). It underpins the project’s determination to monitor key determinants of performance. On the other hand, and as the case on engineering fees shows, it failed to proceed to a plausibility test.

74. Administrative cost share of the project and per beneficiary. One method of gauging project efficiency is calculating the share of administrative and total project cost per beneficiary. The effective cost of the project management component at project completion, which included ocean freight and local handling of WFP rations, was 10 per cent of total cost in accordance with appendix 4, and per beneficiary equivalent to CNY 98 or US$14.4 at the exchange rate at evaluation, or US$144 in terms of total investment per beneficiary. The interim evaluation of the Qinling Mountain Area Poverty- Alleviation Project (QMAPAP) 65 proceeded to an analogous calculation, including a comparison with the WGPAP, which precisely concurred with the figures above given. The corresponding magnitudes for the Qinling Mountain Area Poverty-Alleviation Project were 12 per cent of administrative cost share and US$82 investment per beneficiary. The completion evaluation of the Southwest Anhui Integrated Agricultural Development Project 66 did not make similar calculations and comparisons.

75. Economic rate of return at evaluation. At appraisal, the economic rate of return (ERR) was estimated at 20.8 per cent, compared to 23.8 per cent at project formulation.67 At completion, the WGPAP assigned a reassessment of the economic analysis 68 to a national consultant. Whilst the mission highly commends this initiative, especially in view of evaluation capacity development, the related working paper has several shortcomings. There is no bibliography, and indeed, it is not clear to what extent it refers to the method of calculating the ERR at appraisal, which would have been necessary to establish comparability. It is understood that the consultant proceeded to on-site surveys in the ten counties to gather farm model relevant information, but no details are given in the working paper. Some productivity increases assumed to be caused by the project are by and large congruent with the appraisal estimates,69 and others differ widely. Apparently for practical purposes, the economic analysis at completion assumes average input and product prices over project life from 2002 to 2008, but uses labour prices of 2008, which were a multiple of the ones used in the ERR calculation at appraisal. This last fact would have constrained the ERR at completion, but the working paper concludes that the ERR at completion is conspicuously close to the one calculated at appraisal, i.e. 20 per cent versus 20.8 per cent, respectively. Considering the above mentioned labour cost distortion in the calculation, the ERR at project completion may have been considerably higher than assumed at

64 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, 2000, paragraph 87. 65 International Fund for Agricultural Development, People’s Republic of China, Qinling Mountain Area Poverty-Alleviation Project, Interim Evaluation, draft, Rome, 25 March 2009, paragraph 103, table 4. 66 International Fund for Agricultural Development, People’s Republic of China, Southwest Anhui Integrated Agricultural Development Project, Completion Evaluation, Report No. 1785-CN, Rome, September 2006. 67 The Technical Review Committee suggested forecasting more conservative productivity increases, without formulating a recommendation in this regard, however, see paragraph 47. 68 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report (PCR), Draft, Rome, 2009, Working Paper 2, Economic Analysis. 69 IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, 2000, volume 2, working paper 7, appendix 1.

appraisal, but the lack of method clarity and traceability in this working paper infers a different conclusion. In order to take stock of the project’s initiative to recalculate the ERR at completion, a recommendation is formulated in paragraph 143.

76. Value for money. What is explored here is whether the achieved dimensions of the project purpose were of adequate intrinsic quality. Increased per capita grain availability was based on systematic crop experiments, on-farm demonstrations and ensuing training and extension efforts whereby much emphasis was put on new varieties, especially cereals, and improved crop and animal husbandry in mixed farming systems. Such investments have a long lasting effect, as much as irrigation schemes and soil improvements. As such, they can be considered as key ingredients for incremental household incomes (see also paragraphs 86-87). The significant reduction of both preventable diseases and school drop-outs constitute sound investments in terms of quality, building the basis for human and social capital and empowerment. The same can be said for the widespread and successful support to enterprise start-ups, in addition to the income generating effects that were generated. The mission can also confirm the quality of infrastructure built by the project. It is therefore fair to conclude that project money bought good, and in many cases, long lasting quality.

77. The time dimension. The WGPAP was implemented in accordance with the established time schedule. The only noteworthy delay occurred between the loan agreement signature in February 2001 and project effectiveness in March 2002. The main reason for this was the time lag in the signature of a subsidiary loan agreement regarding rural financial services. This was also the case with the Qinling Mountain Area Poverty-Alleviation Project, where a delay of 18 months was the result, while the Southwest Anhui Integrated Agricultural Development Project reached project effectiveness only three months after IFAD board approval.

78. Efficiency rating. In view of the multiple evidence of efficiency, but taking into account the deficiencies related to the ex-post verification of the project ERR, a rating of five (satisfactory) is indicated.

Project Performance in Summary

Relevance • The WGPAP design was consistent with both national policies and IFAD strategies as put down in the 1999 COSOP. • It was specifically directed to marginal areas and had an explicit gender and participatory focus while it included learning from other IFAD projects. • It included a rural financial services component, enhancing relevance to a considerable degree. • The design did however not take into account a big dam project that affected populations in the project area.

Effectiveness • The project objective, as summarized in paragraph 18, was achieved, especially in terms of per capita grain availability and income increases, reduction of preventable diseases and school drop-outs, and of enterprise start-ups. • Satisfactory effectiveness can be attributed to the support of rural infrastructure and the resulting broad geographic coverage of the project.

Efficiency • WGPAG has achieved or overshot the anticipated objectives at practically the same cost assumed at appraisal. • Unit costs of goods purchased and of infrastructure built by the project was under tight and explicit control. • However, the ex-post evaluation of ERRs was not satisfactory due to weak methodological rigour in the respective study.

VI. RURAL POVERTY IMPACT

79. Perception angles of impact. While the PCR remains the main source of project impact data, the mission relied on additional triangulation points especially in terms of poverty issues and perceptions of project achievements. The most uncommon glance at poverty-alleviation awareness was the drawing contest in five primary schools (fourth and fifth grade) carried out by the project prior to the evaluation mission. The task given to the students was to depict the impact domains as perceived by them, of which a brief analysis by the evaluation mission is given below in table 8. It appears that the winners clearly captured the multi-dimensional character of poverty and the tangible and intangible substance of its reduction. This coincides with the multi-dimensional poverty reduction approach of the project, which was corroborated by the WEILAI Report (see paragraph 53).

Table 8. Topics and Impact Domains in Schoolchildren Drawings a No. of Results and Impact Management System No. of Topic Citations Impact Domain Citations Home, water, biogas 20 Income and assets 24 Environment 14 Crosscutting 19 Cash crops 14 Human capital 17 Road and transportation 9 Agricultural productivity 15 Grain crops 8 Natural resou rces and environment 14 Village, community 7 Local governance 1 Animal husbandry 6 School 4 Leisure time 4 Women and girls 3 Comprehensive impact 1 Total 90 Total 90 a Drawn from the 15 winning drawings, 11 boys and four girls.

Source: Provincial PMO

80. The second triangulation point used by the evaluation mission were the 48 scorecards established during focus group discussion with VIGs. The revealed satisfaction levels did correspond, by and large, with the ones reported in the PCR.70 The focus group discussions also captured the concerns that remained an issue at project completion. Access to schools, as a consequence of their concentration in administrative villages, ranked highest with 63 per cent of responses. The ones regarding agricultural training and credit infer that access to these services remain high on the

70 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report (PCR), Draft, Rome, 2009, Working Paper 1, Participatory Impact Assessment.

population’s agenda with 61 and 54 per cent, respectively. The rankings of poverty reduction for the poorest (25 per cent), drinking water (25 per cent) and medical care (21 per cent) seem to suggest that these issues have been addressed to a substantial extent by the project.

81. The mission also undertook 20 extended household interviews around rural financial services with the intention to complement component relevant information in the PCR. All households were selected at random and on the spot by the mission. Their present situation can shed some light to what extent households - that were all classified as poor or very poor (categories B and C) at the time of contracting the first RCC loan funded by the project - have developed. Average cash incomes as high as US$702 equivalent for all household members including minors are considerable, although still slightly below the threshold value of US$2 per day, when compared to other rural poor areas in Asia. According to these 20 interviews, average household income composition was 73 per cent from agriculture and livestock, 34 per cent from livestock alone and, more specifically, 28 per cent from pig breeding and fattening. Cash transfers to households from work outside the village amounted to 14 per cent and from off-farm undertakings to eight per cent. These patterns will be compared with the PCR data on household income and assets in paragraphs 86-87.

82. The last viewpoint considered in terms of impact perception was the ranking exercise carried out at the evaluation wrap-up meeting in Nanning on 26 May 2009, with the participation of 40 respondents. The questions asked evolved around the strengths and weaknesses and successful components of the project as perceived by the main project stakeholders, the majority of whom were from the provincial and county PMOs, the associated technical bureaus, the RCCUs and the WF. The respondents were given ten points to be placed in a total of 13 boxes at their choice, whereby cumulation and/or abstention were explicitly allowed. The findings in terms of project strengths, weaknesses and component success perceived is shown in table 9.

Table 9. Results from the Ranking Exercise at the Evaluation Wrap-up Meeting Project strengths Points The project took place in the context of a comprehensive rural development strategy in the project area 39 The county governments considered the WGPAP as their own project 39 The project used participatory working methods 35 A very big number of villages was covered 30 Poor households received benefits 28 Other strengths, not specified 6 Project weaknesses The investment portion of the Government was too high 3 Resources were limited 2 Lack of scientific research 1 Several components were particularly successful Training in general 21 Agricultural development 18 Project management 15 Rural financial services 13 Social development 13 Infrastructure 13

83. Attribution of impact. According to the PCR,71 75 per cent of the project beneficiaries received support from the WGPAP only, inferring that the impacts observed can be attributed to the project with a fair degree of certainty. Several senior county government staff clearly stated that without the project, service improvements would have taken place in priority in better-off townships within the counties. The project provided a very comprehensive and balanced set of improvements

71 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, draft, Rome, 2009, Working Paper 1, paragraph 60.

based on local preferences. On the other hand, the WGPAP was embedded in a sweeping government effort of poverty alleviation practically across all components. It is understood that the data previously mentioned refer to project outputs and corresponding total county achievements, but these are pre- requisites for attaining the project objective and goal. The concurrent existence of government programmes may blur attribution of impact, but the detailed records of project output achievements compared to the respective county level achievements provide an uncommon insight into the (derived) impact contribution of a development project. These ranged from 14 per cent (biogas tanks) to 79 per cent (training of illiterates), with an average of 43 per cent of project contribution share to total corresponding achievements in the ten project counties. The targeted project population was 34 per cent of the rural population in the same counties. Hence, the project did proportionally more than just substitute local investments, without however causing significant discrepancies.

84. The general pattern of IFAD and WFP resource allocation (see appendix 6) fairly well reflects output achievement at county level. In the agricultural development component, the aggregate contribution of IFAD and WFP was less than 40 per cent of the aggregate project share of the Government and the beneficiaries, and the average contribution to aggregate county level achievements was in this range (44 per cent). In the social development component, where IFAD and WFP were the most prominent financiers, this same ratio reached 48 per cent. Twenty two per cent contribution was attained in infrastructure, despite the fact that only WFP provided funds for FFW. An exception is the rural financial services component where IFAD devoted 49 per cent of loan resources but had only a contribution of 13 per cent in terms of number of RCC loans extended during project life at aggregate county level. The RCCUs however acknowledged that IFAD funding was instrumental in integrating poorer households in their borrower portfolio (see also paragraph 58). IFAD funding made a significant pitch in the training of RCC staff (72 per cent), similar to the support of VDP preparation (78 per cent) under the project management component.

85. Non-project villages . The mission visited four non-project villages in order to avail of a control group, albeit not commensurate with the numbers of project villages visited. The PRC also referred to one non-project village only, for the same purpose. The general trend observed was that social and infrastructure development was visible also in non-project villages, however with less density and completeness compared to project villages. One non-project village, Fengdong in Leye county, showed dysfunctional symptoms as most of the inhabitants were absent as migrants, leaving only the aged and children behind in a dilapidated village with difficult road access. Absence of coordination in the community to build water cisterns, new houses and biogas tanks was also an issue, according to the household interviews conducted. The school building was rehabilitated thanks to corporate sponsors in Guangdong province, but classroom furniture was not included in the package. The one example cited in the PCR,72 the non-project Baji village, exhibits a steep increase in migration between 2001 and 2007, and an insignificant progress in per capita incomes. The insight on development patterns in non-project villages is insufficient to reach a statistically conclusive assessment on project driven impact attribution, but it is fair to say project villages seem to have benefited in a more comprehensive way from service delivery across the five project components than non-project villages. With these considerations on impact attribution, a detailed review of five key impact domains follows below.

86. Household income. The project goal aimed at a sustainable and equitable poverty reduction in the karst area of GZAR, measured by one indicator, i.e. a reduction of 75 per cent of the population below official poverty lines. At design, the target population was classified into four categories, i.e. better off, poor, very poor and poorest.73 During implementation, the project chose to work only with three categories, namely A (better off), B (poor) and C (very poor), setting net per capita income thresholds at above CNY 1,200, between CNY 800 and 1,200 and below CNY 800, respectively, jointly with annual per capita grain availability ranging from more than 250 kg for category A to less

72 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, draft, Rome, 2009, Working Paper 1, tables 5 and 6. 73 Ibidem, volume 1, paragraph 54.

than 200 kg for category C. By setting the poverty threshold bar lower, the pro-poor focus was enhanced, and targeting priority was given to categories B and C.

87. At completion and according to the PCR, the number of category C households diminished from 129,000 in 2002 to 41,700, equivalent to a 68 per cent, thereby boosting the numbers of category B households that evolved from 111,000 to 154,000. A movement of 51,000 households from category B into category A took concurrently place, increasing the latter from 28,700 in 2002 to 79,700 in 2008. This achievement is commensurate with the indicator initially set al project goal level, as well as with the rural poverty decline reported by the World Bank over this period.74 It is also confirmed, albeit in a small sample size, by the detailed household survey related to rural financial services (see paragraph 81).

88. Migration. The PCR does not provide detailed accounts on income composition by sources. Thus, the role of cash transfers from migrant household members cannot be assessed in precise terms. The overall population, household labour force and income dynamics are however documented in the PCR.75 The overall trends referring to project households in the ten counties are the following:

• Project population grew by six per cent between 2000 and 2008, but household numbers by 14 per cent, thus resulting in smaller households in terms of members at the end of the observation period. • Total labour force increased by 11 per cent while migrant labour grew by 33 per cent in the same period. • Concurrently, per capita income increased by 115 per cent.

89. Opportunities to work in urban areas have remained an overarching factor in the project area. Changes over the 2000-2008 period highlight how migration opportunities have increased for both men and women. Household members engaged in agriculture accounted for two-thirds of active members in 2000. This proportion decreased to slightly less than 50 per cent in 2008. Year-round migration accounted for around one third of the total in 2008.76 This does however not answer the question of which part of the observed income increase was due to cash transfers from migrant household members. paragraph 81 infers that this is relatively modest (14 per cent). This surprisingly low contribution of migrant cash transfers may be skewed by the fact that the 20 household interviews were conducted exclusively with RCC borrowers that invested the loan proceeds in income generating activities, mostly in (cash) crops and livestock. In households without credit access, the portion of migrant cash transfers may be higher. The baseline survey of 2001 indicates an average of migrant to total income portion of 18 per cent.77 The PIA concludes that the project has helped households adjust in a fast-changing period, thus to cope with a situation where migration remains a necessity. The mission tends to concur.

90. Household assets and liabilities. The PCR also highlights changes in household assets, which are an important poverty impact indicator as well:

• 98.3 per cent of households reported family property improvement, • 89.6 per cent of households reported family property improvement of 3 kinds or more than 3 kinds, • 67.4 per cent of households reported family property improvement of 5 kinds or more than 5 kinds.

74 World Bank, From Poor Areas to Poor People: China’s Evolving Poverty Reduction Agenda. An Assessment of Poverty and Inequality in China, Poverty Reduction and Economic Management Department, East Asia and Pacific Region, Washington DC, March 2009, Executive Summary. 75 Ibidem, table 4. 76 Assembled from data on migration and household size provided by county PMOs during Participatory Impact Assessment. 77 WFP-IFAD, WGPAP, Baseline Survey, Report, Nanning Fangzhi Agriculture Investment Limited Company, Nanning, November 2001.

91. The 20 household surveys conducted by the mission shed some additional light on the nature of household assets at the time of evaluation. Eight out of the 20 surveyed households have built new houses in the last few years, and the average present housing value indicated is CNY 65,000 (16 households). Compared to these assets, livestock and other small and medium enterprise productive assets are in average only CNY 6,700 and 9,000, covering 12 and 13 households, respectively. The most noteworthy fact is that agricultural machinery and implements amounts to only CNY 670 in average over 12 households. This is an indication that the nature of agricultural household operations remained very low-capital but high-labour intensive, a feature highlighted both by the Formulation and Appraisal Reports already. Average total assets per household are equivalent to CNY 93,400, liabilities to CNY 19,600, mostly RCC loans but much less informal loans and WF credits (one case), resulting in an average net worth of CNY 73,800 per surveyed household. In terms of “poverty graduation” over project life, three of the surveyed households remained in category C, 11 moved up from category C to B, one remained unchanged in category B, and five graduated from category B to A while all of these five have moved from category C to B some years earlier. This, however small, survey sample seems to underpin that the rural financial services component of the project indeed had its share in the observed poverty-alleviation impact related to household income and assets.

92. Household income and asset rating. The extent of household income and asset change warrants a rating of six (highly satisfactory).

93. Human and social capital and empowerment. The 4,800 households that are members in one of the five FPAs promoted by the project (paragraph 35) have gained better market access as well as better leverage on the sale of their products, which is an expression of empowerment. On a more general plane, the transcendental role of training in all components, partly thanks to the WFP FFT, had a strong impact on social capital, especially when considering its high coverage (see figure 3). Beyond training, the project triggered a series of changes in land use systems, credit access and health and education, which together were conducive to higher status of women in their family. The proportion of households in which both men and women make important financial decisions is on an increasing trend according to the PIA. Confidence is the empowerment factor prominently put forward in the MTR.78 High participation of women in all agricultural extension activities and access to credit contributed to confidence building. Literacy courses were one of the project activities with the strongest focus on women and poorer households. Overall illiteracy rate was 17 per cent in the baseline survey, and 8 per cent at project completion, according to the PCR. The 2005 gender assessment reviewed the project area and concluded that overall female workloads had remained stable. The strategy of reducing women’s workload has been pursued in water and biogas access, but not as a crosscutting theme in agriculture, thus underpinning the persistent labour-intensive nature of farming activities. In fact, the farming chores of women may have increased, due to smaller household sizes, higher cropping intensities and crop diversification, and mixed farming systems involving fodder collection and animal care. During the evaluation mission, women did not actively join focus group discussions that were not specifically women-centred in all villages except two. This is one of the weaknesses highlighted in the PIA.

94. Contrasting situations were observed in household satisfaction regarding access to health and education in their village. Medical care was perceived the most satisfactory service in the scoring card exercise. All indications point to a marked improvement in health status among community members including the poor. Village health posts improved the lives of poor people, because those living in remote villages can walk to that health post for basic care instead of going to the township clinic. In the absence of a health post, village practitioners used to provide care in their own homes. With basic healthcare and prevention, risks of developing important diseases and of falling back into absolute poverty are reduced. On the other hand, education raised the highest proportion of concerns in household scoring cards, not in terms of quality, but access.

95. Rating of human and social capital and empowerment. Considering the substantial progress achieved, but also the efforts that remain on the agenda, a rating of five (satisfactory) is given.

78 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Mid-term Review, Report No. 1731, Rome, December 2005.

96. Food security and agricultural productivity. It is highly likely that child nutrition status has improved. The main rationale for this is the combined positive change in water accessibility, grain availability (paragraph 66), hygiene behaviour and gender attitudes brought by the project. The 2001 baseline survey, carried out by a professional group using the VAM methodology, found that proportions of under height and underweight children were very high, at around 45 per cent and 30 per cent, respectively, with no significant gender differences. Unfortunately, comparative figures assembled in the PIA of 2008 may have a problem of plausibility. These indicate a practically constant rate of these two malnutrition indicators among boys since 2001 while that for girls decreased substantially, to 31 per cent and 12 per cent, for under height and underweight, respectively. One reason may be that the PIA did not follow the same methodological rigour of the baseline survey of 2001, a flaw that could have been avoided.

97. The food security improvements reported in the PCR in terms of per capita grain availability are all the more remarkable as they took place in a context of reduced basic grain area as a consequence of the growing crop diversification over project life. Food consumption patterns have been preserved according to spot checks. Grain is still mostly self-produced, and maize remains the staple crop in the karst villages visited for two meals a day out of three. Project interventions related to high yielding maize varieties were very timely. They were available at project start and, according to PMO interviews, among the first ones to take grain quality into account. The hybrids promoted in the 1990s were not appropriate for human consumption.

98. Paragraphs 32-34 infer that crop and animal productivity progress achieved was substantial in a period only covering six years. While the productivity leaps achieved in crop and animal experiments are referred to in paragraph 32, the ones reported for on-farm demonstrations are in the range of five to 20 per cent over project life, i.e. reasonably lower than the ones under experimental conditions and thus plausible. Reported productivity progress achieved under conditions of extension were also lower than with on-farm demonstrations, again an indication of plausibility. Productivity levels achieved were remarkable where high yielding varieties were involved, and still substantial in the cases where crop diversification and/or mixed farming systems were at stake. Given the high degree of intensity due to small plot sizes, the productivities reported are in the range of the feasible, and access to farm inputs is apparently not a problem. The overall impact on agricultural productivity was the result of a synergy between crop and animal experiments, on-farm demonstrations and subsequent extension efforts, broad-based farmer training, and also the promotion of FPAs.

99. Rating of food security and agricultural productivity impact. Weighing the still noticeable child nutritional deficiencies, as a proxy indicator for food security, against the substantial progress in agricultural productivity and household grain availability, a rating of five (satisfactory) for this impact domain is given.

100. Natural resources and the environment. Forest cover in the project area is reported in the PCR as having increased from 51 per cent to 61 per cent. Only part of this improvement can be attributed to the project. Hill closure was implemented in 43 per cent of project villages, an impressive proportion, but only 12 per cent of households received project support for biogas to reduce fuel wood consumption.79 The on-going Guangxi Karst Rehabilitation Programme, and the Sloped Land Conversion Programme with its annual environmental payments covering many households in the project area, have been the main drivers of rehabilitation. The increase in migration may also have reduced pressure on forest resources in the villages. In addition to biogas and forestation activities, agricultural productivity enhancement and crop diversification (increased availability of pruning waste from orchards) may have helped to reduce extractive uses of vegetation in the hills. To some extent, environmental vulnerability has decreased.

101. Rating. Given the sweeping and rapid changes observable in forest cover and protection, and the reported contribution of the project to aggregate forest closures (41 per cent in table 2), despite

79 Reduced fuel wood consumption is 1.8 ton per year according to the PCR, but only 1 ton per year according to the Eco-farming Project, a World Bank-funded biogas development project covering also GZAR, see paragraph 116.

modest investments (7 per cent of project cost), the rating of impact on natural resources and the environment is five (satisfactory).

102. Institutions and policies. Service-oriented project implementation, according to the PCR one of the main lessons learnt, may well have strengthened mainstream government orientation toward service delivery. “Thinking about households, implementing for households” became the project’s motto. The Guangxi Government was merging its agricultural support services into service centres when the project started. These range from one-stop government service centres, legal mediation, to rural libraries and marketing facilities such as grocery shops. Services are set up as networks linking county, township and administrative village level, as well as natural village level in one case. County government staff in several of the counties visited explained that the project contributed to the introduction of this new approach by building capacity among staff and village cadres and changing attitudes. The existence and consequent implementation of national policies related to poverty alleviation, referred to in paragraphs 11-17, have helped in achieving this impact. In line with improved service delivery of the local government, credit access has also been made more equitable, which was a consequence of improved service delivery capacity of the RCC network.

103. Institutional capacity of villages . Administrative villages in Guangxi remain institutions with remaining gaps in management capacities. The project has strengthened the coordination capacity of village cadres, but apparently has not focused on management capacity. Thus, the project objective has been achieved in a narrow sense for both VIG and VDP related indicators. One of the main factors behind this limited impact of the project’s participatory approach at village level is that its purpose and processes were loosely defined at project start. The absence of management training for VIG members is one of the weaknesses highlighted in the PCR. For instance, the VIGs were to include a village accountant, whereas none of the village committees interviewed had one.

104. Village implementation groups. The PCR describes VIGs as invaluable in the practical management of project activities in their village and in basic data collection. Evaluation mission focus group discussions confirmed that interaction among village committee members has been enhanced. Meetings took place once a month, a remarkable achievement. VIG members were each in charge of a block of natural villages, and held frequent information meetings in the natural villages. This “block management” had just become a new operation mode in GZAR when the project started, and a focus group discussion with the village committee in a non-project township revealed that a similar management system was in place. According to the PIA, village cadres attributed a higher sense of community generated through collective action as one of an important positive impacts of the project. Project requirements in terms of VIG membership had a limited effect only, and gaps between compliance in M&E indicators and reality is visible (table 10).

Table 10. Gaps between Design and Implementation of VIGs Appraisal Report M&E Reporting Evaluation Findings in Five Villages 5 to 7 members 5.4 persons per village Often fully composed of village cadres, or with some 2-3 households 40 per cent household successful farmers, or natural village heads selected as VIG representatives representatives representatives. But many other household representatives have attended natural village meetings. Headed by officially All VIG members Village party secretary or village head making primary by an elected village “elected” for project decision (positions are often merged) head purposes 46 per cent of women WF representative in each village. Another woman VIG head or deputy head sometimes is listed as member, but absent in all cases but 40 per cent of women was a women in 2 per cent one in focus group discussions. Roles mostly in consulting of villages with women and data recording. Sources: Appraisal Report, MTR, appendices 4 and 5, PCR annexes, discussions or interviews with five VIGs.

105. Village development plans. The VDPs have been a useful planning exercise at project start to take local needs into account, but less a tool to reinforce management capacity or prepare institutionalization of village-level planning. As a consequence, the PMOs have generally derived annual plans from the initial VDPs. The PIA observes that training has been supply-driven rather than built on local training needs. Ownership of VDPs by VIGs may therefore have been limited. The 31

VDPs reviewed by the evaluation mission were kept in township PMOs, not in the villages. The VIGs were able to report to the evaluation mission scales of activities and number of households in each category of poverty, but did not know total project investment in their village, and the VDP format did not specify the location of activities in the various natural villages.

106. Women leadership development . Negligible impact in this field, which was one of the 2005 COSOP priorities, was one of the main project weaknesses highlighted in the PCR. Women gradually acquired leadership positions in the project area through formal village elections and through employment as township technicians. The project has reinforced neither of these two channels. A very small number of cadres were promoted from township deputy head to county deputy head due to their successful implementation of the project. The women listed as VIG members in the M&E system (table 10) took a limited part in decision-making in the villages observed during the mid-term review and the evaluation. The 1999 COSOP strategy for women leadership development, which was skill development in business management, was in reality rather geared toward simple tasks in relation to animal raising or cash crops, sometimes in handicrafts or low-skill urban employment such as maids.

107. Rating of institutions and policies. Within the overall government policy framework and its implementation, the project was certainly instrumental for having enhanced the drive of the institutional stakeholders for more people-oriented service delivery. The VIG management capabilities and the role of VDPs as bottom-up planning and monitoring tools remained below expectations at appraisal. Consequently, the impact on institutions and policies is rated four (moderately satisfactory).

108. Overall impact rating. Taken all impact domain ratings together, an average of five (satisfactory) is emerging.

Impact in Summary • Progress in household impact and assets is rated as highly satisfactory. • Social capital end empowerment was significantly enhanced, not least by literacy courses and agricultural training activities that included women to a high degree. • Agricultural productivity increased notably, substantiated by a consistent string of evidence. Food security at household level also improved, but child malnutrition indicators reported are not entirely consistent with household food security data. • The project contributed significantly to forest closures and thus had an impact on the environment and natural resources. Biogas promotion appeared to have played a substantial role in this development. • WGPAP strengthened service delivery capacity by local government agencies. While the involvement of VIGs was considerable in relation to project planning and monitoring activities, mainstream management capacity of villages remained weak, and women in positions of responsibility were rather the exception.

VII. SUSTAINABILITY AND INNOVATION

A. Sustainability

109. Capacity building. The substantial efforts made in various sorts of training is one of the main factors of impact sustainability because enhanced capacity is generally a durable investment good. Diversified groups of people have benefited from training, with probably lasting impacts. Without the project, technical training would have mostly been organized for selected farmers. The proportion of project funds allocated to training has been substantial compared to the other international projects in China reviewed.

110. Durability of project benefits and resilience against shock and stress. Project benefits attained under the agricultural development component are likely to be sustainable, especially new crop varieties and animal breeds, as well as investments made in land use, such as irrigation schemes, land improvements, tree plantations and forest closures. Demand for staple and cash crops and farm animals are expected to be robust and lasting, barring dramatic events, for instance a massive outbreak of swine flue. In the face of climate change, Southern China is estimated to be better off, with slightly more humid years ahead, than other regions of China in the North and East where increasingly

frequent and severe droughts are anticipated.80 The situation is somewhat different in terms of cold events that are forecasted to occur with more severity in Southern China in the future.81 Unusual and severe snowfall and frost in Guangxi in early 2008 appear to underpin this threat. The fact that the project has promoted crop diversification and mixed farming, as well as investments in sustainable land use and infrastructure, represents a coping potential against climate-driven shock and stress, to some extent at least.

111. Project exit strategy and further Government commitment. Two factors have contributed to an apparently seamless transition to a normal pace of activities since project completion in March 2008. First, government mainstream routine and special programmes continue to be deployed in many dimensions where the project was active. Agricultural extension and livestock stations are open to the public, health posts and schools are staffed and equipped, road improvements works are still under way and, of course, rural financial services continue to be available through the extensive RCC network. The financial performance indicators of county RCCUs, where available, infer that their recent history was healthy and thus conducive to further development and expansion.82 The second factor of government commitment was the integration of the project in the county government structure. The county and township PMOs were and still are one and the same as local governments. The head of a county PMO was and is also vice-governor of the county. The county and township PMOs are still functional today, albeit with a comparatively lower level of activities compared to project life. In such a context of transcendent project integration, the exit strategy was sort of inbuilt from the start. The PCR reaches similar conclusions.

112. Sustainability of the RCC network. Appendix 7 of the PCR suggests an overall default rate of 4.78 per cent for RCCs. However, cumulative data are not helpful in assessing default rates, and the reference framework has to be clearly spelled out.83 According to the PCR, recovery rates in the RCC network have improved in the project and as a result of IFAD financed trainings to RCCU staff. Discussion on sustainability will be restricted here to the issue of interest rates and their appropriateness at different levels of the financing chain, and to the sustainability prospects of WGPAP supported financing processes and procedures in the partnering institutions beyond project closing. First, the need to adequately price loans in order to cover costs and risks of lending to lower income households has been stressed repeatedly and for more than a decade by IFAD. Progress was slow, even though in the meantime, the legal provisions have made it possible for RCCs to lend within a much larger top up to the PBC base rates than a decade ago. To date, loan pricing is possible with an interest rate increment of 130 per cent as related to the base rate of the PBC. Concretely, this means that a 12 months loan with a PBC base rate of 5.31 per cent could be granted for 12.21 per cent. In reality, loan pricing under the project did not even reach half of this threshold. The second concern refers to the rates of interest that RCCUs at county level were charged when taking out IFAD funds from the Finance Bureau at county level. These were set at a fixed 1.2 per cent for long-term loans. Comparing these rates with the RCCU´s own cost of funds for these term finance, i.e. 5.85 per cent for a 60 months term deposit, it becomes clear that the subsidy element that IFAD itself introduced into the operations was much higher in proportion than the subsidy element that the RCCUs charged on the sub-borrowers. Although it is understood that the subsidy of the IFAD loan was long term (i.e. 15

80 Ministry of Science and Technology of China (2001BA611B-01) and Ministry of Water Resource of China, Climate Changes of the Past 100 Years in China, in Climate Change Newsletter (2003/2004). 81 Song Ruiyan, Gao Xuejie, Shi Ying, Zhang Dongfeng, Zhang Xiwa, Simulation of Changes in Cold Events in Southern China under Global Warming, Advances in Climate Change Research, Vol. 4 No. 6, 6 November 2008. 82 The evaluation mission attempted to obtain a list of banking indicators from the ten county RCCUs, such as asset and liability structure, equity, operating cost and income, and trends in loan portfolio classification from the year 2000 to 2008, as well as derived indicators (return on assets and equity, asset utilization, etc.). The information received was not consistent enough for a thorough analysis, except for the conclusion that the RCCUs that reported data had a robust financial history over project life. 83 For details on best practices in monitoring and accounting for impaired loan portfolios, see: Von Pischke, Yaron and Zander, Why credit project repayment performance declines, Savings & Development, 2-1998.

years) and aimed at strengthening the institutional capacities of the RCCs to deliver microfinance, subsidies may often induce unwanted short term distortions.

113. Areas with uncertain sustainability. During project life and until today, the local governments have shown a remarkable determination of decentralized and demand based service delivery. Areas with possibly low sustainability levels will likely be the ones that already today show signs of a dilemma, generally between access and quality. With regard to teaching posts in natural villages, the decision in favour of quality has been taken and implemented, resulting in a concentration of better equipped schools in administrative villages, with the ensuing need to add dormitories. A similar move may occur with village health posts and maybe also with individual water supply schemes in remote natural villages. In relation to farmers’ training, county PMOs confirmed that this level of effort will most likely not be sustained. Households independently concurred with this view in the scorecard exercises: Concerns over continued access to agricultural training was ranked second after school access as a remaining issue in the villages.

114. Sustainability rating. Despite some concerns of constrained sustainability, especially in subject areas requiring high levels of service delivery, sustainability is rated five (satisfactory).

B. Innovation, Replication and Scaling Up

115. Overall assessment . A comparison with the IFAD-funded Qinling Mountain Area Poverty- Alleviation Project, which was designed and implemented during the same period, indicates that some innovative contents of WGPAP were taken from other projects, which is commendable. The WGPAP was successful in replicating existing experience from the VIG/VDP approach and from other multisectoral poverty-alleviation programmes. In microfinance, the project approach contained subject matter such as borrower training, loan application, review of client’s capacity and refresher training to ensure credit eligibility to provide access to microfinance for those who were considered previously not to be eligible. The innovative dimension was the partnership of both PMOs and relevant RCC offices that was practiced in the WGPAP. It was scaled up in succeeding IFAD funded projects and recent reports indicate that RCCs intend to continue serving the poorer populations, following the “test phase” implemented in the WGPAP. Similarly, the PCR points to the village credit funds, which were innovative in the country context. The assistance to setting-up FPAs can be considered as an innovation in China, supporting a very recent policy change in 2006. The Chinese government approved a law that allowed and fostered the establishment of self-managed farmer cooperatives/associations. Biogas promotion is another example where the WGPAP took up existing technologies, applied them successfully and also took care that this was widely publicized.84 The introduction of the biogas to poor farmer households was an innovative poverty reduction tool, besides its provision of cheap energy and good effects on soil fertility. Biogas triggered a number of other effects like improved village and household hygiene, reducing pressure on forest utilization and alleviating labour of women. Those immediate effects triggered economic opportunities like investments in pig farming, training opportunities for women, credit access, etc.

116. Innovations in agricultural development. The project did venture into the promotion of organic agriculture, reaching close to 100,000 mu of aggregate areas under extension. Organic agriculture in this context is however not defined as a scheme that is systematically audited by independent certification bodies, but rather an effort to raise awareness towards sustainable farming and windows of opportunity in increasingly sophisticated consumer markets, thus confirming trends observed in an IFAD thematic evaluation.85 A noticeable exception visited by the evaluation mission was an organic tea estate in Leye county that is integrated into the Organic Crop Improvement

84 (i) IFAD, Stories from the Field, China biogas project turns waste into energy, Rome, June 2008; (ii) IFAD documentary movie “gas, gas, gas” which was prepared in the WGPAP and aired on BBC, CNN, DW-TV, RAI 24 and other international channels. 85 IFAD, Organic Agriculture and Poverty Reduction in Asia, China and India Focus, Thematic Evaluation, Report No. 1664, Rome, July 2005.

Association network.86 These project initiatives, which were not foreseen explicitly at design, are commendable. Another innovation, this time mentioned in the Appraisal Report, did not materialize, i.e. the inclusion of farmers into multiplication schemes of high quality cereal seeds. With some reason, the PMO heads were of the opinion that the involvement of small farmers into a seed multiplication process with a high degree of sophistication and stringent traceability requirements would have been premature. Within the context of innovation, IFAD grants were implemented by CIFOR and CIAT in Guangxi and linked to the WGPAP. The CIFOR grant focused on Forests that benefit the poor: Linking income generation to influence among forest communities in Asia . The CIAT grant focused on integrated upland agriculture development using participatory methods. Despite these innovation relevant inputs, the fact remains that the PCR notes that more support during project implementation should have been forthcoming from IFAD.

117. Fertilization versus soil fertility. The mission noted that both recommended norms by the project for mineral fertilization and observed application rates were comparatively high. The essence of this observation is confirmed by the World Bank funded Eco-farming Project,87 stating that “scarce land in China motivates farmers to maximize land productivity. As a result, the use of fertilizers is very high and continues to increase, though varying amongst provinces…. China's use of fertilizers per hectare is almost three-fold higher than the global average and higher than in the majority of OECD countries”. As shown in extension leaflets and in the technology compendium prepared by the provincial PMO, the WGPAP also promoted the use of biogas slurry as fertilizer and included hints of integrated crop management in their extension messages. However, the crop models of the Eco- farming Project are much clearer in showing that the use of biogas slurry, compost or manure causes mineral fertilizer doses to be reduced and, consequently, external input costs as well. In fact, most of the mineral fertilization norms proposed by the Eco-farming Project are lower that the ones applied in the WGPAP. Hence, while the WGPAP was innovative in the agricultural development component in general terms, it failed to convey strong and consequent messages regarding fertilizer use. On the other hand, there is evidence that successful elements of the WGPAP have implicitly found their way into the design of the Eco-farming Project, such as on-farm and applied research, the strengthening of the agricultural extension set-up in connection with farmer and extension staff training, and biogas promotion. Moreover, the compendium on technologies promoted by the project, which will be published as a book, are an indication of the project’s determination to document innovation.

118. Both the WGPAP and the Eco-farming Project fell or fall however short of much needed solutions that address soil fertility in an innovative way and over the long term. Until 2020, China must produce additional 50 million tons of cereal equivalents, or ten per cent of the present output, in order to cope with the population increase, enhanced food security requirements and expected growing consumption levels.88 Because China has very little room for agricultural area expansion, increased food production can only come from further significant productivity increases. While the ones achieved under the WGPAP were respectable, continued growth is not possible without substantial accumulation of soil carbon stock that remains stable over very long periods. Related technologies have caught growing interest worldwide under the term “biochar”,89 over the past five years. The WGPAP would have deserved to be exposed to such innovative agricultural technologies, which did not occur sufficiently. This assessment is clearly expressed in the PCR, not only in connection with

86 Organic Crop Improvement Association. OCIA International is one of the world's oldest, largest and most trusted leaders in the organic certification industry. 87 World Bank, Project Appraisal Document on two Proposed Loans to the People’s Republic of China for an Eco-Farming Project, Report No. 39781-CN, Washington DC, 3 November 2008, annex 10, paragraph 13. 88 Anthea Webb, WFP Country Director, completion evaluation briefing, Beijing, March 2009. 89 Biochar is charred biomass from agricultural and forestry waste that is incorporated into soils. It significantly enhances soil fertility, remains stable over long periods and thus is an atmospheric carbon sink. See www.ibi.org ; www.biochar.info .

agricultural technologies, but also related to know-how in project management, monitoring and evaluation, gender and rural financial services.90

119. Gender and innovation . First, an innovative institutional partnership was set up between the project and WF that consisted in supporting a paradigm shift of the Guangxi WF, away from promoting social development mostly as a civil society role, to supporting women’s rights, gender attitude change and economic development. All these were pursued through a set of simple and multisectoral activities as reported by the PIA.91 The regional and county WF representative confirmed that this had been an important shift. Second, in four of the ten counties, the county PMO included at least one female staff with important actual responsibilities: one was deputy director, one was chief agronomist, and two were in charge of planning or M&E. These staff described how the project’s encouragement to hire female PMO staff has made a difference. These two points are best practice in the current recommendations of national experts to promote gender attitude change.

120. Replication and scaling up. The WGPAP not only replicated pre-existing innovations but also undertook efforts to disseminate these, or home-grown innovations, beyond its immediate borders for third party replication and scaling up. The project had and still has a strong presence in the media, and two publications in particular went at some lengths to demonstrate project approaches and results.92 The first publication is also available as a CD, while the second is a comprehensive reflection on the significance of international cooperation within the agricultural sector in GZAR, with a full chapter on the WGPAP. The author is presently Deputy Director of the Guangxi Department of Agriculture. Over project life, cadres of the provincial PMO were called about ten times for consultancies in favour of other projects with the objective to disseminate knowledge and experience, mostly on project management methods and tools.

121. Rating. The WGPAP undertook considerable and numerous initiatives around innovations; at the same time, there is evidence of missed opportunities of exposure to up-to-date technologies and methods for their replication and scaling up. Still, such innovations hold a high degree of replicability potential. Based on such considerations, a rating of five (satisfactory) is given.

Sustainability, Innovation and Replication in Summary Sustainability • The capacity building effort of WGPAP, and the results thereof, were solid elements of sustainability. • Investments in land improvements as well and in diversification of crop and animal production systems enhanced resilience against shock and stress, including against anticipated fallouts from climate change. • Thanks to the far-reaching integration of WGPAP into local government structures and procedures, exit strategies were sort of inbuilt into project design, thus an element enhancing sustainability. • Uncertain sustainability is perceived where subsidy elements prevail, such as in the RCC interest structure. It is also not certain that local governments can sustain the high level of service delivery at village level. Innovation and replication • Innovations in generic project design were taken from other IFAD projects while WGPAP proved to be capable of replicating these effectively and efficiently. • In the area of agricultural technologies, WGPAP was innovation conscious to a high degree. However, long-term soil fertility issues were not taken up with adequate rigor. More support from IFAD would have been instrumental to address this shortcoming. • WGPAP went at great length in disseminating and replicating innovations, including in project management methods .

90 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report, draft, Rome, 2009, Executive Summary, paragraphs 18-23. 91 Attention to domestic violence, for example, one of the 2005 COSOP priorities, was one of the themes included in the health awareness and literacy course manuals. 92 (i) GACFFPA, WGPAP, The Passion and Love to West Guangxi, Nanning, March 2007; (ii) Wei Jitian, Knowledge and Activities related to the Internationalisation of Agriculture, Guangxi People’s Press, Nanning, February 2005.

VIII. PERFORMANCE OF PARTNERS

122. IFAD. As mentioned in paragraphs 24-29, IFAD was in a position to conduct the project design swiftly and based on relevant experiences in other poverty reduction projects of IFAD and on a solid track record related to the involvement of WFP and the RCC network. This referral to existing routine was clearly an advantage over lengthy fact findings, evidenced by the early agreement on most of the project’s fundamentals and the subsequent adherence to these agreed principles. Despite the expedient nature of the design process, the project went at length in assuring a participatory approach. Another noteworthy feature of IFAD’s performance was the continuity of the respective country programme managers, only two over project life, with Mr Eric Martens starting from the tripartite memorandum of understanding in 1999 and with Mr Thomas Rath from the MTR onwards to date. The WGPAP was considered as a practically problem-free project,93 with the logical result that IFAD was driven to devote more attention to other projects. Here may precisely lie one of the causes of what the PCR observed in its Executive Summary: “Project management could be improved by funds or other provisions to readily access special technical knowledge or technical assistance when required. Once the design was completed and implementation began, the project experienced the need for technical knowledge and information that it did not have in order to interpret the project design and to translate it into concrete operations. This was the case in a number of areas ranging from approaches to project management, exploration of targeting methods, new agricultural production technologies and engineering techniques for the design and construction (of) infrastructure”. The mission tends to concur and has made a corresponding case already in paragraph 117. In particular, IFAD did not live up entirely to the ambitions of the 2005 COSOP that consciously shifted its priorities from a general poverty reduction focus to access and innovation. The case could also be made that IFAD should have sought to shift the project into higher gear at MTR, on the basis of the excellent performance record until that date. In hindsight, however, a more decisive provision of technological and methodological innovation inputs all over project life would have been an investment with higher returns.

123. IFAD performance rating. Considering the overwhelmingly positive performance assessment in general, but also taking into account the lack of targeted support in favour of innovations, IFAD performance is rated five (satisfactory).

124. WFP. The partnership with WFP in the WGPAP was part of a broader blueprint and used in other IFAD projects prior to the WGPAP. It was a mature venture and found its expression in the fact that the 1999 COSOP was explicitly a joint IFAD-WFP document. This allowed WFP to play its cards that had a high comparative advantage, i.e. the ability to make available food rations for FFW and FFT. These tools appear to have been highly adequate as an incentive for broad participation of households, as mentioned in paragraph 51. It is highly doubtful whether the target population’s participation would have been so broad-based without these tools in the case of infrastructure building, and also of training, which constituted a pivotal axis of project design. In conjunction with the government, the provincial and county PMOs in particular, WFP was able to implement the planning, logistics, monitoring and evaluation tasks related with the procurement, shipment and distribution of food aid, which went without major problems and were completed by the end of 2005. In terms of know- how, the VAM methodology jointly developed by IFAD and WFP was another asset that contributed to good performance. A systematic participation of WFP in the UNOPS supervision missions would however have provided synergies that went untapped, with the exception of the first supervision mission in November 2002 in which one representative of WFP was part of. With the withdrawal of WFP from direct food aid interventions in China by the end of 2005, there remains the question on how to include FFW and FFT in future IFAD projects (see recommendation 1, paragraph 143).

125. WFP performance rating. Considering the above, WFP performance is rated five (satisfactory).

126. UNOPS. As in all IFAD projects prior to 2007, the United Nations Office for Project Services (UNOPS) was the designated cooperating institution in charge of loan administration and supervision. While there is nothing on record pointing to problems in loan administration or compliance with loan

93 Verbal communication by Mr Sun Yinhog, IFAD Country Presence Officer, 5 May 2009.

agreement covenants, it is indicated to look into the observations regarding supervision missions contained in the PCR and reported verbally to the evaluation mission. UNOPS fielded five supervision missions, in 2002, 2003, 2004, 2006 and 2007. The MTR in 2005 was jointly composed of IFAD and UNOPS participants. The PCR 94 notes that continuity of supervision members from mission to mission was a problem, which was later addressed in the 2006 and 2007 supervision missions where the members were identical. The same PCR highlights somewhat conflicting messages in the mission recommendations from 2002 to 2004, and the lack of implementation support, in particular in terms of training for project management. Apart from these remarks, a cross-sectional analysis of the supervision reports reveals that the supervision approach was systematic. Each report referred to the recommendations made by the preceding mission and recorded compliance with such recommendations.95 Mission composition was always limited, except for the 2005 MTR and the 2007 supervision mission, and the areas of social development and gender were only covered during these last two missions. In general terms, the time allotted to supervision missions was always in the order of ten days, except for the MTR. This did not allow conducting field visits to the desired extent and was not commensurate with the complexity of the project. This may have been due to the contractual arrangements between IFAD and UNOPS, but was perceived by the provincial PMO as inadequate attention once project implementation started.96 Again, a project that was deemed to be problem-free may have caused this neglect (see paragraph 121), a missed opportunity because more decisive implementation support would have had particularly high returns in a well performing project. Since 2007-2008, IFAD opted out of the sub-contracting arrangement with UNOPS for loan administration and supervision functions, in all of its projects.

127. UNOPS performance rating. The performance of UNOPS is given a rating of four (moderately satisfactory).

128. Government. The existence of explicit government policies with regard to poverty alleviation in marginal areas, coupled with their systematic implementation (see paragraphs 11-17), constituted a solid reference framework that contributed to project performance, in particular regarding effectiveness. Moreover, implementation capacity was widely present for matters that concerned project operations in particular. The evaluation mission attributes this to a fortunate combination of: (i) a well functioning and specialized structure at provincial level and (ii) an apparently seamless integration of project management capacity at county and township level (see paragraph 110) into local government routine. The provincial PMO is in fact called Guangxi Administration Centre of Foreign Funded Projects for Agriculture (GACFFPA). It was established in 1996, subordinated to the Agricultural Department of GZAR, in order to manage foreign funds, information, advanced technology and managerial experience for the integrated development of agriculture and the rural economy.97 Thus, it was and is more than just a PMO but a centre of excellence in the areas stated above. This may help to explain why the PCR highlighted the relative absence of inputs related to advanced technologies and methods from IFAD. Although attached to the Department of Agriculture, the provincial PMO was in a position to streamline project management requirements with the ten county PMOs, which were closely integrated into the local government hierarchy (see paragraph 110) that encompassed all branches and not only agriculture. This denotes a substantial capacity in the coordination of major stakeholders and implementing agencies down to VIG level. This is an uncommon feature compared to other government set-ups that often replicate line agency compartmentalization at local government level. As a result, it can be concluded that the government was fully supportive of project goals. For other ongoing and especially new IFAD projects, the above mentioned ingredients for government performance may be helpful to be used as a yardstick.

94 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report (PCR), Draft, Rome, 2009, paragraph 94. 95 It is interesting to note that the provincial PMO maintained a detailed track record of all supervision, including MTR, recommendations and their respective compliance, and attached it as an appendix to the PCR. This can be considered good practice, which goes to the credit of the PMO. 96 IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report (PCR), Draft, Rome, 2009, paragraph 94. 97 GACFFPA, Illustrated Flyer, Nanning, not dated.

129. The provincial PMO was also instrumental in establishing and implementing a far-reaching M&E system that was congruent, in terms of activities and outputs, with the reporting system of the county governments in general. This was a major achievement that allowed, to a certain extent, gauging project outputs compared to overall county performances. Some dimensions however appear to have given less importance in the M&E system, such as the correlation between households and their geographic location, especially in natural villages. With no mention of ethnicity in the M&E system, project impact on the poverty-ethnicity linkages could not be analyzed.

130. Monitoring data were compiled, starting from administrative villages to township and then county PMOs, and finally consolidated by the provincial PMO. The WGPAP M&E system however fell short of becoming a real-time management information system (MIS), because data were compartmentalised horizontally and vertically. Although township PMOs used computer-based templates, the records were transmitted separately to the corresponding higher level for consolidation, often in hard copies. A comprehensive MIS would have consisted, with the prevailing information technologies at hand, in individually tagging the basic unit, i.e. the beneficiary household, and to integrate all data along the processing and consolidation chain. This would have allowed to assure complete traceability and also to run statistical and plausibility tests as no single entry would have been submerged in level-specific aggregations. With adequate access authorities assigned, such an MIS would have constituted not only a powerful management tool, but also the basis of variance and correlation analyses that could have shed light on questions that remained open, such as household income composition. Data compartmentalization was particularly striking with regard to the RCC network and individual portfolio performance indicators. A comprehensive MIS would not only have been helpful for loan monitoring, but also for the assessment of county RCCUs in terms of banking relevant performance indicators. Obviously, a modern MIS would require to build on baseline surveys designed along the lines of the MIS variables, and subsequent impact assessments would do the same. The otherwise remarkable PIA lacked this kind of methodological stringency. The transfer of MIS related know-how would have been particularly useful and a sound investment, given the strong determination of the project to build and maintain a good M&E system.

131. Despite the missed opportunity of establishing a cross-cutting and real-time MIS, which the evaluation mission tends to attribute to a lack of adequate implementation support by IFAD, Government performance is rated five (satisfactory).

132. Beneficiaries. In the context of the WGPAP, the term “beneficiary” is incomplete and maybe outdated. In fact, the so-called beneficiaries were rather clients or stakeholders. Their ample and active participation was not an expression of poverty or desperation, but reveals a profound capacity to deploy substantial own efforts and to assume ownership. The WGPAP unleashed this capacity to a high degree, without being an artificial island, thus underpinning the prevailing disposition of the people to overcome poverty. Compared to other rural development projects in other settings, where the openness for change of the target population is either not understood or not expressed by impact, WGPAP client performance makes a significant difference.

133. Client performance rating. Given the above, beneficiary performance is rated five (satisfactory).

Performance of Partners in Summary • IFAD coached the project design process in a seamless manner and had a remarkable staff continuity over project life. WGPAP would however have deserved more targeted support in agricultural technologies, and project management and monitoring methods. • WFP was a proven partner of IFAD and able to implement its part of WGPAP components expediently. • UNOPS fielded regular supervision missions but devoted only limited time for these, thus being unable to provide meaningful project implementation support. • The Government of China, at all echelons, contributed significantly to a basically problem-free project implementation. • The beneficiaries, finally, were the ones who proved to be open for change and improvement of their lives.

IX. CONCLUSIONS AND RECOMMENDATIONS

A. Conclusions

134. The performance assessment in paragraphs 52-62 strongly suggests that the project was relevant to the needs of the rural poor and that it addressed all important dimensions of poverty. Its design was consistent with both national policies and IFAD’s strategies, and integrated lessons from other IFAD projects in China. By selecting poverty counties and thanks to a stringent targeting according to the VAM method, the WGPAP set a solid relevance-driven foundation for success. The whole design process can be considered as good practice and warrants a recommendation. Effectiveness was highly satisfactory, specifically measured by indicators in the dimensions of food security, income generation, health, education and enterprise start-ups. One of the key ingredients of effectiveness was the high coverage achieved across the ten project counties. Overall efficiency was satisfactory, but the project failed to conclusively prove that the ERR at completion was probably higher than the ERR assumed at appraisal. The study commissioned to this effect by the project, while commendable as such, lacked the necessary methodological rigour.

135. Poverty reduction impacts of WGPAP were in the satisfactory range for all five impact domains. Despite the existence of mainstream poverty-alleviation programs funded by the Government, impact in the project counties and townships could be attributed to the project with a fair degree of certainty, and the mission consciously triangulated impact perception from observation points that were additional to the PCR, which was admittedly the main source of evidence. Household income and assets progressed at a significant pace, inferring that the poverty reduction targets have been attained. In the domains of social capital and empowerment, food security and natural resources and environment, the evaluation mission also perceived significant impacts. The evaluation mission noticed, in relation to the impact domain of institutions an policies, remarkable progress in service delivery capacity of county and township authorities as well as of the RCC network, but also had to recognize that generic institutional capacity at administrative village level remained weak. Thus, new forms of project partnerships would be recommendable.

136. Overall sustainability is rated as satisfactory. Two key factors of sustainability were, and still are, the existence of mainstream government programmes in poverty alleviation - to which WGPAP was complementary - and the seamless integration of the project management organization at county and township levels into local government structure. These factors outweigh areas of constrained sustainability, such as the still perceivable subsidy elements in the formation of interest and refinancing rates of the RCC network and the high level of public service delivery in remote areas. In terms innovation and replication, the project went at length in capturing and documenting innovations, but it received, in its own perception, insufficient innovation inputs from IFAD, despite the pronounced innovation drive in the 2005 COSOP. A conclusion may be that projects that are innovative by their own choice and determination must receive an above average attention related to innovations as they are capable of making good use of it. A recommendation in this regard is given in paragraph 145. 40

137. A salient trait of the WGPAP evaluation is that all involved partners exhibit performance ratings in the satisfactory range. Government and beneficiary performance ranked best, not only for their ability to implement the project effectively and efficiently, but also for the determination of the PMOs at county and provincial level to establish and run a comprehensive M&E system. WGPAP would have had the opportunity of transforming this M&E system into a genuine and real time MIS if appropriate guidance and support from IFAD had been provided. In order to keep such opportunities open for future IFAD operations in China, a respective recommendation is given in paragraph 146.

138. Summarizing the main evaluation findings, a storyline of relative success emerges. Table 11 underlines this with ratings that are all in the category of satisfactory achievements. Although the rating summary is a comprehensive and well tested analytical framework for the assessment of project success or failure, it falls short of convincingly explaining its underlying contributing factors. In a simplified manner, project success may be seen as a function of quality at entry, implementation capacity and conducive environment. These factors are relatively easy to identify, and the underlying good practices in project design and implementation should systematically be strengthened, by all project partners where applicable.

Table 11. Evaluation Rating Summary Evaluation Criteria Ratings Completion Evaluation A. Project performance Relevance 5 Effectiveness 6 Efficiency 5 Overall project performance 5.33 B. Rural poverty impact Household income and assets 6 Social capital and empowerment 5 Food security and agricultural productivity 5 Natural resources and the environment 5 Institutions and policies 4 Overall rural poverty impact 5 C. Overarching factors Sustainability 5 Innovation, replication and scaling up 5 D. Performance of partners IFAD 5 WFP 5 UNOPS 4 Government 5 E. Overall project achievement 5

139. Quality at entry . In the case of the WGPAP, this important ingredient was marked by: (i) an early and unwavering determination of and agreement on key fundamentals of the project, by and between the main partners, (ii) solid partnerships established along proven patterns, (iii) strong participation and ownership of the national partners during design, (iv) participatory mechanisms involving the target population on a broad basis, and (v) a component mix and cost sharing mode that reflected perceived needs as well as comparative advantages. By and large, quality at entry has been captured under project relevance.

140. Implementation capacity . WGPAP’s implementation capacity was driven by: (i) the experience and credibility of the provincial PMO to provide decisive guidance, (ii) the presence of committed and qualified human resources, sufficient in numbers and able to transform guidance into action, and (iii) the continuity of these human resources. This implementation capacity lead to effectiveness and efficiency and was also a function of performance of partners. The fact that poverty impact, sustainability and innovation were satisfactory stemmed from such implementation capacity, on one hand, and from a conducive environment, on the other hand.

141. Conducive environment. The project environment was conducive because: (i) project implementation was supported by a set of government policies that had a clear rural poverty reduction focus and that emerged when the WGPAP was designed, (ii) local government structure was capable, possibly by design and certainly in reality, to practically integrate the project as a mainstream undertaking, (iii) a stable political environment and a booming economy during project life.

B. Recommendations

142. The evaluation mission makes four recommendations as detailed below. The first refers to the design of future IFAD projects in China, the second to key aspects of institutional partnerships, in particular regarding rural financial services, while the third recommendation points to innovation and IFAD’s role of its promotion. The fourth recommendation deals with M&E and MIS. It is understood that the government of GZAR intends to propose a second project to IFAD that would be based on lessons learnt form the WGPAP.

143. Recommendation 1: Design of future IFAD-funded projects in China. Based on the conclusions illustrated in paragraph 132, the design of future IFAD funded rural development projects in China should continue to be responsive to the multidimensional character of poverty and the growing expertise of national stakeholders in this kind of operations, in particular:

(i) Targeting of project townships should continue to be based on factual indicators. In projects focused on service provision, all villages within one township should be included in the project, and natural villages more clearly identified. This is recommended in order to enhance critical mass in service delivery and would be congruent with the current “block development” approach, which the Government of China is currently implementing. Likewise, a transparent assessment of the special needs of the various ethnic minority groups is critical.

(ii) The design process should replicate WGPAP’s good practice of strong involvement of local expertise at all levels, including from the anticipated target population. This should include the joint calculation of ERRs at appraisal, and a set of guidelines that enable projects to revisit the ERR calculation at completion based on own expertise.

(iii) In settings where FFW and FFT is deemed essential for broad participation and coverage, a second-generation solution should be devised with the Chinese authorities, for instance by linking donor funded projects with government programs for infrastructure and human capacity building. The Government has increasingly stepped up investment in rural infrastructure, and in human capacity building (e.g. training for migrant workers). The IFAD country program is linked with government programs to build similar complementarities. In countries where food aid continues to be policy, the lessons from the IFAD-WPF partnership in the WGPAP should be learnt.

144. Recommendation 2: Institutional partnerships. Project partnerships should extend over the boundaries of provincial and local government in the strict sense, e.g.:

(i) VIGs or analogous bodies at village level should not only play a role for planning and monitoring purposes, but should be strengthened in their core management capacity, with special regard to service delivery functions and equitable gender representation in positions of responsibility.

(ii) Partnership with provincial RCC networks should be put on a completely new footing, by taking into account the ongoing reform and by agreeing on a set of information to be shared that is conducive to a real-time assessment of loan portfolios and banking performance indicators in general.

(iii) Key areas of policy dialogue should be determined systematically and given due and documented attention over the project implementation period.

145. Recommendation 3: Role of innovation in IFAD-funded projects. Building on the orientations of the 2005 COSOP, new IFAD-funded projects should put significantly more emphasis on innovation and its promotion, in particular:

(i) Ambitions of the project partners regarding innovations should be screened systematically from formulation stage onwards and be periodically reassessed.

(ii) Particular attention should be devoted to the identification of local sources of innovation and their mobilization during project implementation.

(iii) Innovations that exist or emerge in the global context should be made available to IFAD- funded projects more pro-actively, possibly also via IFAD grants.

(iv) Priority areas should encompass innovations that respond to global challenges, such as food security, soil fertility, alternative energies and climate change.

146. Recommendation 4: From M&E to MIS. New IFAD-funded projects should take stock from the accumulated experience of the WGPAP and actively foster a quantum leap in the state-of-the-art of M&E and MIS, with the aim to make advanced M&E and MIS methods a mainstream feature in China. This would include the following:

(i) The generic requirements of an overarching, yet concise, M&E system for multidimensional poverty-alleviation projects in marginal rural areas should be defined, best in a joint effort between national expertise and IFAD resources.

(ii) The basic architecture of a real-time MIS, based upon such M&E requirements, should be designed in a way that it can be used for a wide array of IFAD-funded projects, in China and possibly also elsewhere. For this to substantiate, the data capturing and storing capabilities of an M&E system must evolve into a real time management system allowing easy analyses across all relevant variables, and thus require a program basis going well beyond the accumulation of mere spread sheets. Again, a partnership between national expertise and IFAD would be advisable.

(iii) The above recommendations would give a particularly concrete meaning to the development of local evaluation capacity, an initiative that IFAD has decided to support in China.

APPENDIX 1

Mission Itinerary

Date Place (County) Activity 04 May 2009 Beijing Arrival of mission members. Briefing with IFAD Country Presence Office at WFP 05 May 2009 Beijing Resident Mission, internal mission organization. 06 May 2009 Beijing Nanning Document analysis, travel by air to Nanning. 07 May 2009 Nanning-Tian’e Briefing at the PMO, travel by road to Tian’e. Bala township, Long’e administrative village. Laopeng township, Madong administrative village and 08 May 2009 Tian’e school, Wufu administrative village, Shenli and Mogu natural villages. Wrap-up meeting with institutions involved in the project, 09 May 2009 Tian’e-Leye travel by road to Leye. Briefing by county authorities and PMO. Gantian township. Paping natural village. 10 May 2009 Leye Xinhua township, Nashe administrative village, school and 1 natural village, Guili administrative village. Visit of Lewei administrative village, Dashiwei Karst Tiankeng geopark, Lansai administrative village (2 natural villages). 11 May 2009 Leye Leye township, Fengdong administrative village (non- project) Visit of and dinner at organic tea farm, Long Yun Mountains. Travel by road to Tianlin Langping township: Tanghe village school, health clinic, 12 May 2009 Leye-Tianlin township agricultural extension and animal husbandry stations. Visit of administrative village Pingjiu and natural villages Napo and Xiaolan. 13 May 2009 Tianlin Liulong township, 1 natural village, township PMO. Meeting with county PMO, bureaus. Ganoderma mushroom experiments. Travel by road to Napo, lunch halt at Funing county (Yunnan Province) 14 May 2009 Tianlin-Napo Meeting with Napo county authorities, PMO and Women’s Federation. Baisheng township: Nameng administrative village and 1 natural village, school, township PMO. 15 May 2009 Napo Visit of Sangai and Nangmin administrative villages, Gugai natural village. Meeting with Napo Pig Producers Association. Travel by road to Long’an, halt in Debao county and interaction with Debao PMO. 16 May 2009 Napo-Long’an Briefing session with Long’an PMO and bureaus, RCCU and Women’s Federation. Doujie township: Longmin administrative village and 1 natural village, township PMO, WF training room. 17 May 2009 Long’an Visit of Aihua and Luanzheng administrative villages. Yanjiang town, Dongli administrative village (non-project): village committee meeting, school.

Date Place (County) Activity Nanxu township, clinic, township PMO, agricultural extension and animal husbandry stations. 18 May 2009 Long’an-Nanning Lianwu administrative village: school, health station. Arrival of Mr M. Keating in Nanning. Debriefing session with PMO. 19 May 2009 Nanning Dinner with Deputy Director of Guangxi Agriculture Department. 20 May 2009 Nanning Writing of aide memoire. 21 May 2009 Nanning Writing of aide memoire 22 May 2009 Nanning Clearance of aide memoire by IOE, IFAD 23 May 2009 Nanning Revision of aide memoire 24 May 2009 Nanning Translation of aide memoire and layout of evaluation report Dispatching of aide memoire to participants of wrap-up 25 May 2009 Nanning meeting. Wrap-up meeting. 26 May 2009 Nanning-Beijing Travel by air to Beijing. 27 May 2009 Beijing Debriefing session at Ministry of Finance. 28 May 2009 Beijing Departure of mission members.

APPENDIX 2

List of Persons Met

Name Place and Institution Position Beijing Deputy director, International Financial Li Rui Ministry of Finance, People’s Republic of China Institution Division III, International Development Foreign Capital Project Management Center, State Su Juan Council Leading Group of Poverty Alleviation and Project Officer Development Sun Yinhong IFAD, Sub-regional Office Head Anthea Webb WFP, China Office Country director Zhang Yan (Zoe) WFP, China Office Donor relations officer Nanning Deputy director Wei Jitian Guangxi Agricultural Department general Guangxi Administration Centre of Foreign Funded Liu Dian Zheng Director Projects for Agriculture (GACFFPA) Deputy director and Chen Wei Chao GACFFPA head, PMO, Eco- farming Project Song Yuejia GACFFPA Senior agronomist Li Jian Hua GACFFPA Financial officer Huang Dianke GACFFPA Cadre Qin Yu GACFFPA Staff Jiang Qiu GACFFPA Staff Luo Zhixiong GACFFPA Staff Liao Yannian GACFFPA Staff Jiang Shitao GACFFPA Staff Xiong Qiao GACFFPA Staff Vice manager of credit Li Mingdong RCCU of GZAR department Xie Zuo RCCU of GZAR Staff Liao Fengkai RCCU of GZAR Staff Laing Xiuqing WF of GZAR President Lin Xin Financial department of GZAR Staff Huang Shenguang Debao PMO Director Huang Dianke Debao PMO Vice director An Yongfa Fengshan PMO Director Ya Hanjing Fengshan PMO Vice director Luo Caijun Fengshan PMO Technical director Nong Youbian Tiandeng PMO Director You Bing Tiandeng PMO Technical director Long Wen Tiandeng PMO Technical director Li Chunlu Huangjiang PMO Vice director Liu Xunyi Huangjiang PMO Technical director

Name Place and Institution Position Mo Rongmao Huangjiang PMO Director Chen Yingdi Nandan PMO Vice director Feng Hongliu Nandan PMO Technical director Tian'e county Tan Haifan Government of Tian'e county Vice governor Guo Changyong County PMO Vice director Wei Haiyang County PMO Cadre Lu Min County WF Vice presidenet Lei Shilong RCC, Bala township Head Peng Mingjing Agricultural extension station, Bala township Head Liu Fufeng PMO, Bala township Cadre Sun Chuangyu Government, Bala township Secretary Sun Xiaowen Long'e village Party secretary Wei Rongxin Long'e village Vice head Liu Huaying Primary school, Madong village President Li Changqing Wufu village Party secretary Tian Xinyang Wufu village Head Tian Guangqiu Mo Guzai natural village Team leader Luo Zhengfa Mo Guzai natural village Team leader Luo Encai County PMO Cadre Lei Shiqun Credit department, Tian'e county RCCU Manager Wang Rongjiang Agricultural extension station, Bamu township Head Huang Chengjiang Government, Bamu township Vice head Lu Xiuzhong WF, Bamu township President Lu Jinglin Yongli village, Bamu township Party secretary Lu Xiuzhong WF, Bamu township President Wei Mingjian Senli village, Bamu township Party secretary Wei Lianri Nadong natural village Head Leye county Wang Cegong County PMO Vice director Zheng Zuohua County PMO Technical director Yang Changqi County agriculture department Vice director Zhou Taochuang County forestry department Cadre Huang Yuanbing County animal husbandry department Vice director Huang Ruixiao County animal husbandry department Cadre Ren Yuantong Nashe village, Xinhua township Head Yang Fayi Nashe primary school, Xinhua township President Peng Zhangyun Nashe primary school Student Zhu Minghua Nashe primary school Student Liu Chunping Nashe primary school Student Xie Qishun Yanzhan natural village, Nashe village Villager Zhu Youwen Yanzhan natural village, Nashe village Villager Liu Chunxian Yanzhan natural village, Nashe village Villager Zeng Xianhuan Ye Zhuduo natural village, Langshai village Villager Zhang Qun Ye Zhuduo natural village, Langshai village Villager Mr. Tang Ye Zhuduo natural village, Langshai village Villager Wang Mei nian County WF Head

Name Place and Institution Position Huang Xian fa County education department Head Huang Zhong qiang County health department Cadre Zhang Xing si County PMO Cadre Zhou Dabang Langshai village, Huaping township Director Ye Zhuduo natural village, Langshai administrative Wang Xian Villager village, Huaping township Ye Zhuduo natural village, Langshai administrative Yang Cheng'en Villager village, Huaping township Tang, Zhang, Yang Fengdong village (non-project), Tongle township Villagers Liu Qikang Daping natural village Head Yang Shuhua Daping natural village Team leader Zhao Yuanying, Liu Fengmian, Liu Daping natural village Villagers Qizhuang Luo Gui'e Daping natural village Villager Lu Guihong Fada natural village Head Gong Shizhen Fada natural village Villager Lu Mengfang Fada natural village Villager Lu Guifeng Fada natural village Villager Lu Ming Fada natural village Villager Wang Meinian County WF Vice director Chen Shichun County RCCU Vice director Tianlin county He Haikang County PMO Director Ming Fangchun Agricultural extension station, Langping township Head Gan Haican PMO, Langping township Head Bi Chenglong Pingliu village, Bagui township Vice head Pan Jiren Pingliu village, Bagui township Head Li Jianjun Guopu village, Nongwa township Secretary Fan Zhixiong Guopu village, Nongwa township Head Ming Feng'en, Qin Weisun, Huang Gonaderma planting center Cadre Yanfang Nong Xuemin, Qin County PMO Cadre Weisun Huang Jiaping Clinic, Tanghe village, Langping township Cadre Tang Sheng'ai, Yang Dagao school of Tanghe village, Langping Teachers Shengzhao township Yang Guowen Langping township Party secretary Yao Yuanyin Tanghe village, Langping township Secretary Yange natural village, Gongyang village, Liulong Lin Anyuan Secretary town Qin Zhiming, Yan PMO, Liulong township Cadre Yongjie Yange natural village, Gongyang village, Liulong Cen Shijun Vice head town

Name Place and Institution Position Huang Lishen, Yang Meilian and her husband, Huang Yange natural village, Gongyang village, Liulong Villagers Yumei, Lu Guanglian, town Huang Yulan, Li Yulan, Cen Shijun Huang Zhilin County government Vice governor Huang Zhu'en County Development Reform Committee Cadre Luo Wenfu County transportation department Cadre Huang Weidong County water conservancy department Cadre Xian Keqi County Poverty-Alleviation Committee Cadre Liang Minde County PMO Vice director Li Wehong County PMO Cadre Dagao natural village, Tanghe village, Langping Huang Jiahua Head township Wu Zhengdong, Yao Dagao natural village, Tanghe village, Langping Villagers Yuanpin township Yao Caihua Jiufeng natural village Villager Zhu Guanghui RCC, Langping township Chief Huangzongqi RCC, Langping township Cadre, county RCCU Nong Xueming RCC, Langping township Cadre, county PMO Cen Shihong, Luo Pingzai natural village, Xiaolan village, Liulong Villagers Minwu township Pingzai natural village, Xiaolan village, Liulong Nong Xuemin Cadre, county PMO township Pingzai natural village, Xiaolan village, Liulong Peng Yanhua Cadre, WF township Pingzai natural village, Xiaolan village, Liulong Yan Yongjie Cadre, township PMO township Pingzai natural village, Xiaolan village, Liulong Yangzhu Cadre, township RCC township Huang Zongqi, Shi Tianlin RCCU Cadre Quanhui Nong Xuemin Tianlin RCCU Cadre, county PMO Napo county Lin Rongchang County PMO Director Li Weijian County PMO Vice director Wu Anyu County PMO Technical director Huang Yubing County PMO Vice director Cui Shujin County PMO Technical director Su Chanbin County PMO Vice director Huang Weilan County water conservancy department Cadre Huang Jiajin County agriculture department Cadre Lin Qidong County financial department Cadre Dai Yukang County government Vice governor Pan Shaoxin County government office Vice director Liang Yindan County WF Vice president Mo Chengge County RCC Director

Name Place and Institution Position Li Peijia Government, Baisheng township Secretary Li Zhengfeng Government, Baisheng township Vice head Nong Zhongqian Government, Baisheng township Cadre Wei Xiaoli Government, Baisheng township President of WF Cen Guozheng RCC, Baisheng township Director Liangfeng VIG of Shanggai village, Bainan township Team leader Yang Heming VIG of Shanggai village, Bainan township Vice team leader Nong Guangqiang VIG of Shanggai village, Bainan township Member Nong Zhongmin VIG of Shanggai village, Bainan township Member Nong Yixun VIG of Shanggai village, Bainan township Member Nong Lianqing Shanggai village, Bainan township Villager Huang Jiajin County agriculture department Cadre Long Guofeng County agriculture department Cadre Xu Zhenhua PMO, Bainan township Cadre Lu Shiwang VIG of Nongming village, Bainan township Team leader Nong Changlong VIG of Nongming village, Bainan township Vice team leader Nong Rongsheng VIG of Nongming village, Bainan township Member Huang Yuansheng VIG of Nongming village, Bainan township Member Xu Jinyong VIG of Nongming village, Bainan township Member Lu Libao VIG of Nongming village, Bainan township Representative of WF Nong Meiqi Nongming village, Bainan township Villager Nong Xiulan Nongming village, Bainan township Villager Nong Lihua Nongming village, Bainan township Villager Pan Shaoxin County government office Vice director Long Guofeng County agriculture department Cadre Huang Zhuguang Pig producers' association Vice president Qinyong Pig producers' association Member Liang Dongping Pig producers' association Member Dai Yukang County government Vice governor Liang Yindan County WF Vice president Li Peijia Government, Bainan township Secretary Li Zhengfeng Government, Bainan township Vice head Nong Zhongqian Government, Bainan township Cadre Wei Xiaoli Government, Bainan township President of WF Cen Guozheng RCC, Bainan township Director Liang Feng VIG of Shanggai village, Bainan township Team leader Yang Heming VIG of Shanggai village, Bainan township Vice team leader Nong Guanagqiang VIG of Shanggai village, Bainan township Member Nong Zhongmin VIG of Shanggai village, Bainan township Member Nong Yixun VIG of Shanggai village, Bainan township Member Nong Lianqing Shanggai village, Bainan township Villager Nong Yulan Shanggai village, Bainan township Villager Huang Guobing Shanggai village, Bainan township Villager Liang Yindan County WF Vice president Nong Shangming County education department Cadre Mo Wenxin County PAC Cadre Zhou Zhengzhen County PMO Cadre

Name Place and Institution Position Nong Hanfei County PMO Cadre Liang Chaozhi County PMO Cadre Li Jianwei County PMO Vice director Zhou Zhengzhen County PMO Cadre Nong Zhongqi PMO, Baisheng township Cadre Xu Yongjun PMO, Baisheng township Cadre Su Zhongcheng Government, Baisheng township Secretary Huang Xinxue Government, Baisheng township Head Mofan Government, Baisheng township Vice head Li Peizhen Government, Baisheng township President of WF Nong Wenchang Nameng primary school President Du Junxiong Nameng primary school Teacher Su Guofeng VIG of Nameng village Team leader Meng Yuanqing VIG of Nameng village Vice team leader Wei Jianrong VIG of Nameng village Vice team leader Nong Xinzhong VIG of Nameng village Member Zhao Youqiang VIG of Nameng village Member Meng Meiquan VIG of Nameng village Representative of WF Luo Xiumei Nameng natural village, Nameng village Villager Deng Xiuhai Nameng natural village, Nameng village Villager Xiong Mumei Bajiao natural village, Nameng village Villager Nongli natural village, Nongming village, Bainan Nong Zhaowen Team leader township Ma Yongda Namo natural village, Nameng village Villager Su Yanping Namo natural village, Nameng village Villager Meng Lihua Namo natural village, Nameng village Villager Yao Xiuhe Namo natural village, Nameng village Villager Long'an county Lin Yi County PMO Director Zhang Zengshou County PMO Technical director Xu Tingyao County PMO Vice director Lei Biliang County PMO Cadre Nong Gaobo County forestry department Vice director Xiang Lu County transportation department Cadre Liu Kangyin Credit department, county RCCU Manager Zhou Yiliang RCC, Yangwan township Director Lei Biliang County PMO Cheng Zhangsheng Ai Hua village VIG team leader Nong Youpeng Yang wan village Vice head Deng Yifen PMO, Yang wan township Cadre Chen Litian Nanxu hospital President Xu Meirong Gynaecology department, Nanxu hospital Doctor Zhang Yanmei Electrocardiogram department, Nanxu hospital Doctor Zeng Yinyan Assay department, Nanxu hospital Doctor Xu Zhi Radiation department,Nanxu hospital Doctor Pan Liangxiang Pharmacy, Nanxu hospital Doctor Deng Yifen PMO, Nan Xu township Cadre

Name Place and Institution Position Bna Xiuliang PMO, Nan Xu townsip Cadre Yan Ming Animal husbandry station,Nan Xu township Head Chen Shaoliang County animal husbandry department Vice director Nong Youpeng Government, Nan Xu township Vice head Zai Riwei Government, Nan Xu township Head Lin Yi County financial department Director Huang Anli County financial department Vice director Zhou Yiliang RCC, Yangwan township Director Qin Xiaosong VIG of Luan zheng village Team leader Huang Liangying VIG of Luan zheng village Member Qin Zhenren VIG of Luan zheng village Member Zhou Yiliang RCC, Yangwan township Head Lin Fuying County WF Vice director Huang Xiaoying County WF Cadre Cheng Zhangsheng VIG of Ai Hua village Team leader Nong Youpeng Yang wan village Vice head Deng Yifen PMO, Yang Wan township Cadre Nong Lansong Ai Hua village Villager Nong Jinhua Ai Hua village Villager Zhou Ruicheng County RCC Cadre Yuan Qizhen County agricultural department Head Zhang Zengshou County PMO Cadre Hunag Lianhai VIG of Long Min village Team leader Liang Deyi VIG of Long Min village Member Huang Lianwen Long Min village Villager Huang Lianguo Long Min village Villager Liang Jinxin Long Min village Villager Huang Honghua Long Min village Villager Huang Meihua Long Min village Villager Huang Jingqi Long Min village Villager Huang Huihong Long Min village Villager Huang Meizhi Government of Dujie township Vice head Nong Liliang PMO, Dujie township Cadre Huang Zhaoqun PMO, Dujie township Cadre Zhang Zengshou PMO, Dujie township Cadre Xun Yaoting PMO, Dujie township Cadre Zhang Zengshou PMO, Dujie township Cadre Liang Guanshen Government, Yan jiang town Vice head Su Zhenhui VIG of Dongli village Team leader Pan Hong VIG of Dongli village Vice team leader Lu Fukang VIG of Dongli village Statistic staff Zhao Shijian Agricultural extension station of Yan jiang town Vice director Xu Yaoting PMO, Yan Jiang township Vice director Long Degao VIG of Lian wu village Team leader

APPENDIX 3

Bibliography

DFID, China Briefing Paper: Basic Education, Beijing, May 2008.

Economist Intelligence Unit (EIU). China Country Profile, 2008

GACFFPA, WGPAP, The Passion and Love to West Guangxi, Nanning, March 2007.

GACFFPA, Illustrated Flyer, Nanning, not dated.

IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Project Completion Report (PCR), Draft, Rome, 2009.

IFAD, Office of Evaluation, Desk Review Note, West Guangxi Poverty-Alleviation Project, People’s Republic of China, Completion Evaluation, Rome, 29 April 2009.

IFAD, Report and Recommendation of the President to the Executive Board on a proposed loan to the People’s Republic of China for the West Guangxi Poverty-Alleviation Project, 2000, Part I

IFAD, West Guangxi Poverty-Alleviation Project Appraisal Report, Rome, December 2000

IFAD, China, Country Strategic Opportunities Paper, Report No. 906-CN, Rome, February 1999.

IFAD, China, West Guangxi Poverty-Alleviation Project, TRC Minutes, Rome, 9 April 2000.

IFAD, China, West Guangxi Poverty-Alleviation Project, Environmental Screening & Scoping Note, Rome, 14 April 2000.

IFAD, Regional Strategy Paper Asia and the Pacific, Asia and the Pacific Division, Project Management Department, Rome, March 2002.

IFAD, People’s Republic of China, Country Strategic Opportunities Paper, Rome, 13 December 2005.

IFAD, People’s Republic of China, West Guangxi Poverty-Alleviation Project, Mid-term Review (MTR), Report No. 1731, Rome, December 2005.

IFAD, Alisdair Cohen, WEILAI and Rural China: Development and application of a multidimensional tool for project planning, monitoring and evaluation, Technical Report prepared for the Asia and the Pacific Division, Rome, December 2007.

IFAD, People’s Republic of China, Qinling Mountain Area Poverty-Alleviation Project, Interim Evaluation, draft, Rome, 25 March 2009, paragraph 103, table 4.

IFAD, People’s Republic of China, Southwest Anhui Integrated Agricultural Development Project, Completion Evaluation, Report No. 1785-CN, Rome, September 2006.

IFAD, Stories from the Field, China biogas project turns waste into energy, Rome, June 2008.

IFAD, Organic Agriculture and Poverty Reduction in Asia, China and India Focus, Thematic Evaluation, Report No. 1664, Rome, July 2005

Ministry of Foreign Affairs of the People’s Republic of China and UNDP, China’s Progress towards the Millennium Development Goals, 2005, Beijing, October 2005.

Ministry of Foreign Affairs of the People’s Republic of China and UNDP, China’s Progress towards the Millennium Development Goals, 2008 Report, Beijing, 2008.

Ministry of Science and Technology of China (2001BA611B-01) and Ministry of Water Resource of China, Climate Changes of the Past 100 Years in China, in Climate Change Newsletter (2003/2004).

R. Zander, The Way Ahead: Financial Instruments and Proposals for Continuing Reform of Rural Financial Institutions, IFAD Thematic Study on Credit in China, Workshop – Financial Services Delivery under IFAD Project Framework, Beijing, 03 July 2000.

Song Ruiyan, Gao Xuejie, Shi Ying, Zhang Dongfeng, Zhang Xiwa, Simulation of Changes in Cold Events in Southern China under Global Warming, Advances in Climate Change Research, Vol. 4 No. 6, 6 November 2008.

UNDP and China Development Research Foundation, China Human Development Report 2007/08, Access for all: Basic public services for 1.3 billion people, Beijing, 25 July 2008.

UNDP and China Development Research Foundation, China Human Development Report 2005, Beijing, 15 October 2005.

UNDP and China Development Research Foundation, China Human Development Report 2007/08, Access for all: Basic public services for 1.3 billion people, Beijing, 25 July 2008.

Von Pischke, Yaron and Zander, Why credit project repayment performance declines, Savings & Development, 2-1998.

Wei Jitian, Knowledge and Activities related to the Internationalisation of Agriculture, Guangxi People’s Press, Nanning, February 2005.

WFP/IFAD China VAM Unit, Geographical Targeting for WFP/IFAD New Project in Guangxi, Beijing, 2000.

WFP-IFAD, People’s Republic of China, Guangxi Zhuang Poverty-Alleviation Project, Baseline Survey Report, Nanning Fangzhi Agriculture Investment Limited Company, Nanning, November 2001.

WFP-IFAD, WGPAP, Baseline Survey, Report, Nanning Fangzhi Agriculture Investment Limited Company, Nanning, November 2001.

World Bank, China: Country Gender Review. East Asia Environment and Social Development Unit, Washington DC, 2002.

World Bank, Poor Rural Community Development Project, Project Appraisal Document, Washington DC, 2005.

World Bank, Mid-term Evaluation of China’s 11th Five Year Plan, World Bank Office china, 18 December 2008

World Bank, From Poor Areas to Poor People: China’s Evolving Poverty Reduction Agenda. An Assessment of Poverty and Inequality in China, Poverty Reduction and Economic Management Department, East Asia and Pacific Region, Washington DC, March 2009.

World Bank, China: Overcoming Rural Poverty. Joint report of the Leading Group for Poverty Reduction, UNDP and the World Bank, Washington DC and Beijing, 2001.

World Bank, Project Completion Report of the Southwest Poverty Reduction Project, Report no. 26132, Washington DC, 30 June 2003.

World Bank, Project Appraisal Document on two Proposed Loans to the People’s Republic Of China for an Eco-Farming Project, Report No. 39781-CN, Washington DC, 3 November 2008.

APPENDIX 4

National and Regional Policies and Programmes

Major Changes in the National Policy Framework Sector 1990s 2000-2003 2004-2005 2006-2010 Grain quotas re-established National rural Agricultural (1996-98) Recognition of development Agricultural tax and rural National ecological rural migrants strategy removed (2006) development rehabilitation programmes starts New socialist launched countryside policy Poverty 8-7 poverty reduction plan Poverty reduction strategy (2001-2010) reduction (1994-2000) West China Development Programme launched Rural health insurance system expanded Rural health (2006) Health insurance system Plan to compete initiated village health stations (2009-2011) Primary education Compulsory education Education reform initiated law revised Programme for the Gender development of women Programme for the development of women (2001-2010) (1995-2000) Sources: various background reports, in particular, Frank Pieke, YEDP (Yunnan Environment Development Project) – Selected National Policies – Summary Analysis, Final Report, Institute for Chinese Studies, University of Oxford, 12 th August 2004.

Rural Development Programs of Relevance to the Project Programmes 1990s 2000-2003 2004-2005 2006-2010 Great karst Guangxi mountains infrastructure Border counties Three karst counties West Guangxi five programmes counties Guangxi Karst rehabilitation (biogas and hill closure) ecological Sloping land rehabilitation Sloping land conversion conversion programmes maintenance Tuberculosis control (Department for Poor Rural International Southwest poverty Communities Development/World reduction Development Bank 2002) International- (World Bank 1995) Integrated Forestry Basic education in funded projects Karst environmental Development Western areas rehabilitation (Ausaid (World Bank 2005) (Department for 1998) a/ International

Development/World Bank 2003). Sources: Guangxi PMO for infrastructure programmes, county bureau interviews, international donor websites a/ Project implemented in other than WGPAP counties

APPENDIX 5

Financial Progress including Contingencies, in '000 CNY

Adjusted Cumulative Per cent of Per cent of Appraisal Targets Financial Components Appraisal Adjusted Targets after the Progress as per Targets Targets MTR 31 March 2008

Agricultural Development Crop & Livestock Development 74,949.50 83,017.16 82,771.85 110.44 99.70 Irrigation Development 148,076.91 148,076.93 148,271.67 100.13 100.13 Dry Land Improvement 84,468.24 84,468.24 83,255.31 98.56 98.56 Tree Development 60,051.88 60,051.88 59,895.37 99.74 99.74 Sub-total Agricultural Development 367,546.53 375,614.21 374,194.20 101.81 99.62 Rural Financial Services Revolving Fund 185,984.37 186,706.37 196,818.14 105.83 105.42 RCC Support 9,243.38 6,210.31 4,438.10 48.01 71.46 Sub-total Rural Financial Services 195,227.75 192,916.68 201,256.23 103.09 104.32 Social Development Health and Nutrition 28,386.11 28,543.05 27,471.32 96.78 96.25 Education 46,579.71 46,579.37 46,113.24 99.00 99.00 Women Income Generation 45,172.14 42,989.42 37,031.78 81.98 86.14 Sub-total Social Development 120,137.97 118,111.83 110,616.35 92.07 93.65 Infrastructure Drinking Water Supply 68,403.40 68,403.39 66,532.92 97.27 97.27 Rural Road Construction 31,919.66 31,919.66 31,766.32 99.52 99.52 Sub-total Infrastructure 100,323.07 100,323.05 98,299.24 97.98 97.98 Project Management Province PMO 34,127.80 23,662.42 25,248.23 73.98 106.70 County PMOs 46,990.72 46,346.58 55,393.15 117.88 119.52 Township PMOs 19,674.48 12,407.59 7,232.40 36.76 58.29 Sub-total Project Management 100,792.99 82,416.59 87,873.78 87.18 106.62 GRAND TOTAL 884,028.31 869,382.36 872,239.79 98.67 100.33 Source: PCR, appendix 4, summarized.

APPENDIX 6

Project Costs including Contingencies by Funding Source, in '000 CNY

Appraisal Report Targets Total Components Appraisal Report target IFAD WFP Government Beneficiaries

Agricultural Development Crop & Livestock Development 74,949.50 31,176.89 6,224.01 15,233.47 22,315.22 Irrigation Development 148,076.91 0.00 16,343.90 131,733.02 0.00 Dry Land Improvement 84,468.24 0.00 33,936.82 50,531.42 0.00 Tree Development 60,051.88 9,427.63 7,099.30 43,524.96 0.00 Sub -total Agricultural Development 367,546.53 40,604.52 63,604.03 241,022.87 22,315.22 Rural Financial Services Revolving Fund 185,984.37 113,375.50 0.00 0.00 72,608.87 RCC Support 9,243.38 8,170.38 0.00 1,073.00 0.00 Sub -total Rural Financial Services 195,227.75 121,545.88 0.00 1,073.00 72,608.87 Social Development Health and Nutrition 28,386.11 16,207.13 3,175.81 9,003.16 0.00 Education 46,579.71 19,112.73 10,298.92 17,167.72 0.00 Women Income Generation 45,172.14 25,948.95 7,811.45 11,412.10 0.00 Sub -total Social Development 120,137.97 61,268.81 21,286.18 37,582.98 0.00 Infrastructure Drinking Water Supply 68,403.40 0.00 7,647.75 60,755.64 0.00 Rural Road Construction 31,919.66 0.00 0.00 31,919.66 0.00 Sub -total Infrastructure 100,323.07 0.00 7,647.75 92,675.30 0.00 Project Management Province PMO 34,127.80 5,546.59 0.00 28,581.25 0.00 County PMOs 46,990.72 16,942.23 0.00 30,047.97 0.00 Township PMOs 19,674.48 6,854.54 0.00 12,819.94 0.00 Sub -total Project Management 100,792.99 29,343.36 0.00 71,449.17 0.00 GRAND TOTAL 884,028.31 252,762.57 92,537.96 443,803.32 94,924.09 Per cent by Funding Source 100.00 28.59 10.47 50.20 10.74 Source: PCR, appendix 4, summarized.

Evaluation Framework

CRITERIA EVALUATION QUESTIONS DATA SOURCES I. Project Performance A. Relevance • Are project objectives realistic and consistent with national agriculture and rural • President’s Report development strategies and policies, the COSOP and relevant IFAD policies, as well • Loan Agreement as the needs of the rural poor? • Formulation Report • How coherent was the project with respect to the policies, programmes and projects • Appraisal Report undertaken by the Government and other development partners? • Mid-term review • How was the internal coherence of the project in terms of synergies and • Supervision Report complementarity between objectives, components, activities and inputs? • Project completion report • Was the project design participatory in the sense that it took into consideration the • Interviews with country authorities inputs and needs of key stakeholders, including the Government, executing agencies, • Group discussion with beneficiaries co-financiers and the expected beneficiaries and their grassroots organizations? • Did the project benefit from available knowledge (for example, the experience of other similar projects in the area or in the country) during its design and implementation? • Did project objectives remain relevant over the period of time required for

65 implementation? In the event of significant changes in the project context or in IFAD

policies, has design been retrofitted? • Was the project design and implementation approach (including financial allocations, project management and execution, supervision and implementation support, and M&E arrangements) appropriate for achieving the project’s objectives? • What are the main factors that contributed to a positive or less positive assessment of relevance? B. Effectiveness • To what extent have project objectives been attained? • President’s Report • Did the project succeeded in providing sustainable rural infrastructure and services to • Appraisal report the target population? • Loan Agreement • Did access to these infrastructure and services generate an improvement in living • Supervision reports conditions of poor households? • Mid term review

• Did smallholders benefit from improvement in agricultural productivity? • Project completion report APPENDIX 7 • What factors in project design and implementation account for these results? • Individual interviews in the field and • Did changes in the overall context (e.g., policy framework, political situation, with country authorities institutional set-up, economic shocks, civil unrest, etc.) affect project results? • Group discussion with beneficiaries • Survey • Direct Observation

CRITERIA EVALUATION QUESTIONS DATA SOURCES C. Efficiency • What are the costs of investments to develop specific project outputs compared with • President’s Report national standards? Is the cost ratio of inputs to outputs (including cost per • Appraisal report beneficiary) comparable to local, national or regional benchmarks? • Supervision reports • How does the economic rate of return at evaluation compare with project design • Mid term review estimate (both at project level, by component and by investment)? • Project completion report • What were the administrative costs per beneficiary and how do they compare to other • Individual interviews in the field with IFAD- or other donors-funded operations in China? beneficiaries and with country • How much time did it take for the loan to be effective, and how does it compare with authorities (at central and local level) other loans in the same country and region? By how much was the original closing • Government data (i.e. for date extended, and what were the additional administrative costs that were incurred benchmarking) during the extension period? • Did the project deliver expected results in a timely manner? • What factors help account for project efficiency performance? II. Rural Poverty Impact A. Household income and • Did the composition and level of household incomes change (more income sources, • Supervision reports assets more diversification and higher income)? What changes are apparent in intra- • Mid term review household distribution of incomes and assets? • Group discussion with beneficiaries • 66 Did households access to water, livestock ownership and endowment of productive • Individual interviews in the field with

assets change? Did other household assets change (houses, bicycles, radios, television beneficiaries and with country sets, telephones, etc.)? authorities (at central and local level) • Did the rural poor benefited from improved access to community infrastructure? • Direct observation • Did poor households’ financial assets change (savings, debt or borrowing)? • Self Assessment • To what extent the rural poor benefited from higher income through better access to • Project completion report financial markets more easily? Did the rural poor benefited from better access to input and output markets (for example through extension services)? B. Human and social • Did rural people’s groups and grass-root institutions change? Are changes in the • Supervision reports capital and social cohesion, collective capacity and local self-help capacities of rural • Mid term review empowerment communities visible? • Group discussion with beneficiaries • To what extent did the project empower the rural poor vis-à-vis development actors • Individual interviews in the field with and local and national public authorities? Do they play more effective roles in beneficiaries and with country decision-making? authorities (at central and local level) • Did the rural poor gain access to better health and education facilities? • Direct observation • What were the project effects on household health through the repair and • Self Assessment rehabilitation of water infrastructure? • Project self-assessment report

CRITERIA EVALUATION QUESTIONS DATA SOURCES C. Food security and • Did cropping intensity change? Was there an improvement in land productivity (for • Supervision reports agricultural example through adoption of improved technologies) Did the returns to labour • Mid term review productivity change? • Group discussion with beneficiaries • Did children’s nutritional status change (e.g. stunting, wasting, underweight)? • Individual interviews in the field with • Did household food security change? beneficiaries and with country • To what extent did the rural poor improve their access to input and output markets authorities (at central and local level) (for example through credit, access to nurseries or extension) that could help them • Direct observation enhance their productivity and access to food? • Self Assessment • Project completion report D. Natural resources and • Did the status of the natural resources base change (land, water, forest, pasture, etc.)? • Supervision reports the environment • What were the effects of community investments on natural resources? • Mid term review • Did local communities’ access to natural resources change (in general and • Group discussion with beneficiaries specifically for the poor)? • Individual interviews in the field with • Has the degree of environmental vulnerability changed (e.g., exposure to land beneficiaries and with country degradation, soil erosion)? authorities (at central and local level) • Direct observation • Project completion report 67 E. Institutions and • What improvements were discernable in local governance, including government • Supervision reports

policies departments, elected bodies and officials? • Mid term review • Group discussion with beneficiaries • Interview with key informants of CACB and AREA • Individual interviews in the field with beneficiaries and with country authorities (at central and local level) • Project completion report III. Other Performance Criteria A. Sustainability • Was a specific exit strategy or approach prepared and agreed upon by key partners to • President’s Report ensure post-project sustainability? Was this effective? • Loan Agreement • What are the chances that benefits generated by the project will continue after project • Appraisal report closure, and what factors militate in favour of or against maintaining benefits? What • Supervision reports is the likely resilience of economic activities to shocks or progressive exposure to • Mid term review competition and reduction of subsidies?

CRITERIA EVALUATION QUESTIONS DATA SOURCES A. Sustainability • Is there a clear indication of government commitment after the loan closing date, for • Group discussion with beneficiaries (continued) example, in terms of provision of funds for selected activities, human resources • Individual interviews in the field with availability, continuity of pro-poor policies and participatory development beneficiaries and with country approaches, and institutional support? Did the IFAD project design anticipate that authorities (at central and local level) such support would be needed after loan closure? • Project completion report • Did project activities benefit from the engagement, participation and ownership of local communities, grassroots organizations, and the rural poor? • Are adopted approaches technically viable? Do project users have access to adequate training for maintenance and to spare parts and repairs? • Are the ecosystem and environmental resources likely to contribute to project benefits or is there a depletion process taking place? B. Innovation, replication • What are the innovation(s) promoted by the project? Are the innovations consistent • President’s Report and scaling up with the IFAD definition of this concept? • Appraisal report • How did the innovation originate (e.g., through the beneficiaries, Government, IFAD, • Supervision reports NGOs, research institution, etc.) and was it adapted in any particular way during • Mid term review project/programme design? • Group discussion with beneficiaries • Are the actions in question truly innovative or are they well-established elsewhere • Individual interviews in the field with 68 but new to the country or project area? beneficiaries and with country • Were successfully promoted innovations documented and shared? Were other authorities (at central and local level) specific activities (e.g., workshops, exchange visits, etc.) undertaken to disseminate • Direct observation the innovative experiences? • Self Assessment • Have these innovations been replicated and scaled up and, if so, by whom? If not, • Project completion report what are the realistic prospects that they can and will be replicated and scaled up by the Government, other donors and/or the private sector? IV. Performance of partners A. Performance of IFAD • Did IFAD mobilize adequate technical expertise during design? • Formulation report • Was the design process participatory (with national and local agencies, grassroots • Appraisal Report organizations) and did it promote ownership by the borrower? • Supervision reports • Were specific efforts made to incorporate the lessons and recommendations from • Mid term review previous independent evaluations in project design and implementation? • Individual interviews with country • Did IFAD adequately integrate comments made by its quality enhancement and authorities (at central and local level) quality assurance processes? • Project self-assessment report • Did IFAD (and the Government) take the initiative to suitably modify project design • Project Status Reports (if required) during implementation in response to any major changes in the context, • Interview with IFAD Country especially during the MTR? Programme Manager for China

CRITERIA EVALUATION QUESTIONS DATA SOURCES A. Performance of IFAD • How effective was IFAD in working with the UNOPS to carry out the mandated • Interview with IFAD Country Presence (continued) task? Has IFAD exercised its developmental and fiduciary responsibilities, including Officer compliance with loan and grant agreements? • Were prompt actions taken to ensure the timely implementation of recommendations stemming from the supervision and implementation support missions, including the MTR? • Did IFAD undertake the necessary follow-up to resolve any implementation bottlenecks? • What was the role and performance of IFAD’s country presence team? Did IFAD headquarters provide the necessary support to its country presence team, for example, in terms of resources, follow-up and guidance, adequate delegation of authority, and so on? • Has IFAD made proactive efforts to be engaged in policy dialogue activities at different levels in order to ensure, inter alia, the replication and scaling up of pro- poor innovations? • Has IFAD been active in creating an effective partnership and maintaining coordination among key partners to ensure the achievement of project objectives, including the replication and scaling up of pro-poor innovations?

69 • Has IFAD, together with the Government, contributed to planning an exit strategy?

B. Performance of the • Has the Government assumed ownership and responsibility for the project? • Supervision reports Government of China • Judging by its actions and policies, was the Government fully supportive of project • Mid-term review and its agencies goals? • Interview with IFAD Country • Has adequate staffing and project management been assured? Have appropriate levels Programme Manager for China of counterpart funding been provided on time? • Interview with IFAD Country Presence • Did project management discharge its functions adequately, and has the Government Officer provided policy guidance to project management staff when required? • Individual interviews with government • Has auditing been undertaken in a timely manner and have reports been submitted as authorities (at central and local level) required?

• Did the Government (and IFAD) take the initiative to suitably modify the project design (if required) during implementation in response to any major changes in the context? • Was prompt action taken to ensure the timely implementation of recommendations from supervision and implementation support missions, including the MTR? • Was an effective M&E system put in place and did it generate information on performance and impact which is useful for project managers when they are called upon to take critical decisions?

CRITERIA EVALUATION QUESTIONS DATA SOURCES B. Performance of the • Did the Government (and IFAD) contribute to planning an exit strategy and/or Government of China making arrangements for continued funding of certain activities? and its agencies • Have loan covenants and the spirit of the loan agreement been observed? (continued) • Have the flow of funds and procurement procedures been suitable for ensuring timely implementation? • Has the Government engaged in a policy dialogue with IFAD concerning the promotion of pro-poor innovations? C. Performance of the • Has the supervision and implementation support programme been properly managed • Supervision reports cooperating institution (frequency, composition, continuity)? Has the cooperating institution complied with • Mid-term review (UNOPS) loan covenants? • Interview with IFAD Country • Has the cooperating institution been effective in financial management? Programme Manager for China • Has the cooperating institution sought to monitor project impacts and IFAD concerns • Interview with IFAD Country Presence (e.g., targeting, participation, empowerment of the poor and gender aspects)? Officer • Have implementation problems been highlighted and appropriate remedies • Individual interviews with government suggested? authorities (at central and local level) • Has the cooperating institution promoted or encouraged self-assessment and learning processes? • Has the supervision process enhanced implementation and poverty impacts? 70 • Has the cooperating institution been responsive to requests and advice from IFAD when carrying out its supervision and project implementation responsibilities? D. Performance of WFP • How did WFP perform with respect to the delivering of grains? Was WFP supportive • Supervision reports of project rationale and objectives? To what extent WFP performance contributed to • Mid-term review project objectives and performance? • Interview with IFAD Country Programme Manager for China • Interview with IFAD Country Presence Officer • Individual interviews with government authorities (at central and local level)

APPENDIX 8

Definition of the Evaluation Criteria used by the Office of Evaluation

Criteria Definition a

Project performance Relevance The extent to which the objectives of a development intervention are consistent with beneficiaries’ requirements, country needs, institutional priorities and partner and donor policies. It also entails an assessment of project coherence in achieving its objectives. Effectiveness The extent to which the development intervention’s objectives were achieved, or are expected to be achieved, taking into account their relative importance. Efficiency A measure of how economically resources/inputs (funds, expertise, time, etc.) are converted into results.

Rural poverty impact Impact is defined as the changes that have occurred or are expected to occur in the lives of the rural poor (whether positive or negative, direct or indirect, intended or unintended) as a result of development interventions. • Household income and assets Household income provides a means of assessing the flow of economic benefits accruing to an individual or group, whereas assets relate to a stock of accumulated items of economic value. • Human and social capital and Human and social capital and empowerment include an assessment of the changes empowerment that have occurred in the empowerment of individuals, the quality of grassroots organizations and institutions, and the poor’s individual and collective capacity. • Food security and agricultural Changes in food security relate to availability, access to food and stability of access, productivity whereas changes in agricultural productivity are measured in terms of yields. • Natural resources and the The focus on natural resources and the environment involves assessing the extent to environment which a project contributes to changes in the protection, rehabilitation or depletion of natural resources and the environment. • Institutions and policies The criterion relating to institutions and policies is designed to assess changes in the quality and performance of institutions, policies and the regulatory framework that influence the lives of the poor.

Other performance criteria • Sustainability The likely continuation of net benefits from a development intervention beyond the phase of external funding support. It also includes an assessment of the likelihood that actual and anticipated results will be resilient to risks beyond the project’s life. • Promotion of pro-poor innovation, The extent to which IFAD development interventions have: (i) introduced innovative replication and scaling up approaches to rural poverty reduction; and (ii) the extent to which these interventions have been (or are likely to be) replicated and scaled up by government authorities, donor organizations, the private sector and others agencies.

Overall project achievement This provides an overarching assessment of the project, drawing upon the analysis made under the various evaluation criteria cited above.

Performance of partners • IFAD This criterion assesses the contribution of partners to project design, execution, • Government monitoring and reporting, supervision and implementation support, and evaluation. • Cooperating institution The performance of each partner will be assessed on an individual basis with a view • NGO/CBO to the partner’s expected role and responsibility in the project life cycle. a These definitions have been taken from the OECD/DAC Glossary of Key Terms in Evaluation and Results- Based Management and from the Methodological Framework for Project Evaluation as agreed upon with the Evaluation Committee in September 2003.

PROJECT EVALUATION

People's Republic of China

West Guangxi Poverty-Alleviation Project

International Fund for Completion Evaluation Agricultural Development Via Paolo di Dono, 44 00142 Rome, Italy Tel: +39 06 54592048 August 2010 Fax: +39 06 54593048 E-mail: [email protected] Web: www.ifad.org/evaluation

IFAD Office of Evaluation Bureau de l’évaluation du FIDA Oficina de Evaluación del FIDA