Completion Evaluation¹

Introduction

The completion evaluation mission is conceived as an independent yet consultative exercise and designed as a service to the Southwest Anhui Integrated Agricultural Project (SWAIADP) as well as to IFAD's ongoing and future operations in China. The evaluation also served the purpose of accountability of IFAD to its governing bodies.

The Mission visited China from 23 June to 15 July 2005.

Methodologies utilised included field-level questionnaires, focus group discussions, interviews with key informants and direct observation of some 25 project sites. The evaluation adopted the prescribed framework of the IFAD Office of Evaluation.

SWAIADP was the thirteenth of seventeen IFAD projects in China, all of which have concentrated on agricultural development or rural credit or both. Six of these, including SWAIADP, have been co-funded by the World Food Programme (WFP). Total IFAD investment in China has now reached USD 400 million. The main thrusts of IFAD's Strategic Framework in the country are equitable access to natural resources, finances and markets, and the strengthening of grassroots institutions.

China has experienced a spectacular expansion of its economy in the 25 years since the initiation of market liberalisation and land reforms, during which time the average annual growth in GDP was nearly 10%. This has been achieved against radical structural shifts: from central planning to a market economy, from agriculture to manufacturing and services, from a closed to a globally integrated system.

SWAIADP was a USD 56 million project, financed by an IFAD loan of USD 26 million, counterpart funding of USD 22 million, local contributions valued at USD 5 million, and WFP food-for-training equivalent to USD 3 million. The bulk of the IFAD loan (68%) was directed to the provision of micro- credit. The project was declared effective in December 1997 and closed on 30 June 2004 after an extension of one year. Supervision and loan administration were carried out by UNOPS.

The main goal of the project was to reduce the chronic food shortages among half-a-million people - 123 400 households - living in the 34 poorest townships of the five poorest counties in the mountainous area of southwest Anhui province. An estimated 80% of households were categorised as poor or very poor at appraisal.

The project was implemented through seven components:

• infrastructure (27% of base cost) – irrigation, roads, drinking water and flood control;
• food crops (25%), development of arable land;
• tree crops (26%), tea, mulberry, bamboo, chestnut;
• livestock and fisheries (9%),
• women's development (4%), training in literacy, health care and agricultural techniques;
• social support (4%) and
• institutional support (5%).

The provision of micro credit for poor households constituted the main instrument for implementation.

The Anhui Department of Agriculture was the chief implementing agency, under the control of the Ministry of Agriculture in Beijing.

Project Management Offices (PMOs) were established at the level of Province, Prefecture, County and Township, assisted by the relevant technical departments. The management and delivery of credit funds was to be handled by the existing Rural Credit Cooperatives (RCCs) at township level, the first instance in China of this innovative approach. A new institution, the Village Implementation Group (VIG), was created to promote and supervise village-level activities.

Implementation

The overall level of financial achievement at project closure stood at over 90%, with the outstanding 10% largely accountable to a faulty estimate of likely exchange rate fluctuations. In terms of physical achievements, around 16 000 ha of land was improved or put under new or rehabilitated cultivation, 320 km of roads were built or upgraded, 1 500 gravity tanks and wells were constructed for drinking water, 122 km of irrigation canals, 870 irrigation ponds and over 200 km of flood protection barriers were built, all of which matched or exceeded the physical targets planned at appraisal.

Credit management was handled jointly by PMOs and RCCs, with the PMOs normally bearing the credit risk and having the final say in loan approval. This arrangement was accepted by the People's Bank of China, which regulates cooperatives, and by the supervision missions. Nearly 100 000 households received credit under the project, or around 77% of all households in the project area. The largest proportion (46%) of the total loan amount was utilised for the cultivation of "economic trees", followed by short-term seasonal credit (25%). Overall recovery rates stand at around 85%.

Training programmes had a budget of nearly USD 2.5 million and included literacy and healthcare training specifically for women, technical training for farmers for the new crops, and loan management training for all households accessing loans. Some of the training materials and methods were found to be inappropriate for the illiterate and semi-literate. The WFP food-for-training acted as an effective incentive for attendance at the training sessions.

Relevance, Effectiveness and Efficiency

Project objectives were relevant to the causes of poverty in the project area, namely a shortage of cultivable land, low yields, lack of infrastructure, limited access to markets and poor levels of education, literacy and health, especially among women. SWAIADP was a ‘home-grown' project, first proposed by the Anhui Department of Agriculture and formulated largely by local consultants. However, the county PMOs argued with some cogency that the ‘menu' and extent of new crops might more usefully have been left to them.

Overall, SWAIADP was a highly effective and successful project in that it met or exceeded the planned targets. Supervision missions expressed reservations, however, regarding the targeting of the poorest and the effectiveness of the M&E system in capturing the impact of the project in this respect. The response of the PMOs was that in an area where virtually all households were poor, the example of more active and resourceful households would have the desired multiplicative effect. By project completion, 91% of targeted households had received loans, which appears to legitimise the PMO approach.

The prescribed monitoring and reporting systems were scrupulously followed but the Project Completion Report acknowledged some weaknesses: too much emphasis on quantitative data collection at the expense of data analysis, and difficulty in applying the results of M&E to the modification of village development plans, which accorded with the present Mission's verdict that there was a lack of qualitative impact monitoring and analysis.

The investment cost per beneficiary household was relatively low, at around USD 400, and the expected ERR, at 34.2% comfortably exceeds the appraisal figure of 22.6%. Management costs were far lower than comparable projects, a mere 2.8% of project expenditure. The major recurrent costs including salaries and transportation, as well as the provision and maintenance of buildings and offices, were met from existing government budgets. The commendable decision to maintain previous levels of government support in the project area meant that the operation of the SWAIADP constituted a genuine addition, rather than a partial replacement.

Some ‘collective' schemes were implemented, in which the planning and some of the inputs were provided by the implementing agencies. Plots are managed individually and the loans made to individual households, but the overall execution of the schemes was managed by officials. Instead of the mosaic of small plots with a variety of crops envisaged at appraisal, larger areas of monoculture have been established. The system was commended by UNOPS consultants and the Evaluation Mission was impressed by the efficient utilisation of scarce land resources.

Loan defaults of 15% indicate an area for concern. At appraisal, it was suggested that 80% recovery would be an acceptable rate initially but would be expected ultimately to rise to 98.5%, with the remaining 1.5% covered by the risk fund. The PMOs claim the recovery rate will gradually improve, but the only available sanction is the refusal of further loans.

Impact

The signs of increased prosperity were apparent everywhere: well-maintained metalled roads, renovated two-storey houses, well-stocked shops and markets. Rice, soybean, maize, mulberry, tea and chestnut plantations, wherever located, look uniformly healthy and productive. The total number of households reached by all activities represents 92% of households in the project area, comfortably exceeding the appraisal target. Nearly 5 000 ha of poor and unusable land has been improved by levelling, soil replacement and drainage, with areas of waste land transformed into highly productive orchards.

Per capita income and grain availability have been the main impact indicators utilised by the project. The average increases by county were 36.8% for income and 13.3% for grain availability between 1997 and 2003. From one mu² of tea, a household can generate an income of Y 2 000 p.a., and from one mu of mulberry Y 10 000 from four annual silkworm harvests. Chestnuts were a successful option for households with a shortage of labour capacity, since the trees need only minimal tending.

Determining the extent of food insecurity is hampered by the lack of recent official records, which were abandoned as food became freely available at markets. Income and grain availability figures deal in overall averages, which might increase due to large gains made by the better-off. Much emphasis has therefore been placed by supervision missions on reporting the ‘graduation' of households in terms of the three categories of poverty. According to PCR figures, the percentage of households in the ‘poorest' category has fallen from 15.1% in 1997 to 2.6% in 2003, while that of the better-off households has risen with increasing rapidity.

Overall grain availability rose from 135 kg per person per year to 281 kg during project life according to PCR figures, 280 kg being the approximate yearly requirement in China for an adult with an average workload. There is of course inequality in terms of food availability, but no household met by the Mission, even the poorest, was unable to afford daily meals. Further, household diets were richer in protein and fresh vegetables.

SWAIADP loans were the first experience of borrowing for the majority of households. Project officials stated that their greatest problem in the early days was the suspicion and hostility among all but the most progressive households to the notion of borrowing. In all villages visited by the Mission, project-funded loans were found to be familiar and sought after, and the disciplines of loan management and repayment had become part of everyday life.

Particular importance was given to women signing for loans for mulberry cultivation, silkworm raising, animal husbandry, primary tea processing and other IGAs. This may not have been an appropriate emphasis in the case of southwest Anhui. Questions on this subject provoked an emphatic response from men and women alike: it was immaterial who signed the loan. Many loans signed by husbands were in fact utilised by women.

Rural Credit Cooperatives have benefited from the project credit operation, mostly without risk to themselves. Their deposits have increased, their clientele base has been broadened and their capacities in terms of loan approval and recovery have been enhanced. They are now aware of the potential of small loans made to poor people, provided the costs are covered by the interest charged.

The question mark hanging over the RCCs is whether they now have the capacity to carry out credit delivery, loan approval and recovery and to bear the credit risk in line with the original design. The feeling among the PMOs is clearly that the first responsibility of non-governmental cooperative institutions such as the RCCs will be to their members, and their primary aim will be profitability. IFAD is now testing this in the ongoing Rural Finance Programme.

The Mission did not see a single trace of deforestation or soil erosion in all areas visited. An extraordinary degree of care is taken in the terracing and cultivation of the slopes. Soybeans are grown on terrace verges for maximum utilisation of the soil and the prevention of erosion. In fruit orchards, mulch is commonly used for moisture retention. All maize, mulberry and tea crops are grown at reasonable density to maximize benefits and minimize surface run-off water.

Protected ‘niches' have been created for farmers to develop their land by the ‘bombing' of hilltops and mountains with the seed of pinus species. Field-to-forest conversion is subsidised.

Rural women have been key participants in project activities. The PCR assigns 40% of loans to women, apparently a mix of signature and utilisation. Their income from mulberry and silkworms, from tea cultivation and primary processing, from chestnuts and improved livestock breeds average at more than Y 2 000 p.a. Health awareness and disease prevention have improved. In terms of workload, women told the Mission they are busier than before, and content that it should be so in view of their increased income.

Greater awareness of disease, hygiene and nutrition was reported by women participants in healthcare classes. The declining incidence of gynaecological diseases is indicated by township data and confirmed by doctors, and staff of the Women's Federation. The consumption of meat, eggs, fresh vegetables and edible oils has greatly increased according to household interviews.

Lessons learned from SWAIADP by government departments include: (i) the advantages of ‘loan money' over ‘grant money' through the disciplines of repayment and the emphasis on sustainability; (ii) the usefulness of PRA techniques in needs assessment; (iii) the importance of compiling gender disaggregated data. SWAIADP has also boosted agricultural extension in terms of the range of services, with ‘strategically important' crops such as rice, wheat and maize being joined by cash crops including tea, mulberry, bamboo, chestnuts, gingko, kiwi, pears, peaches, plums, melons and grapes.

SWAIADP has built-in advantages in terms of sustainability: (i) the operation of the revolving credit fund for a further 15-20 years; (ii) the Government commitment to continued support through Food-for-work, increased funds for farmer training, further investments in rural infrastructure, and the provision of market information; (iii) agro-industrial initiatives (notably in Huoshan county) including primary and secondary processing and marketing, for tea, silk and bamboo; (iv) the survival of experienced PMOs, albeit reduced in size; and (v) above all, an environment well-protected against soil erosion and deforestation.

Innovations of the project include: the successful utilisation of the RCCs, in a somewhat restricted role, in the management of microcredit; the WFP food-for-training (as opposed to the well-tested food-for-work); the successful twinning of household choice and village-level planning regarding economic trees; the creation of the project-specific VIGs, which are now regarded as permanent institutions in villages.

Performance of partners

IFAD can be commended for an intelligent project design, the boldness of the new approach to credit delivery and the strong personal support of the then Country Portfolio Manager, mentioned by staff at all levels. It can be criticised for the lack of an initial survey of RCC capacities, which might have given advance warning of the problems in that sphere. In addition, the counties commonly requested more autonomy concerning the stipulations as to areas and outputs of new or expanded cultivation and the lifting of the Y 3 500 ceiling for loans.

UNOPS fielded six supervision missions, the first two of which were longer in the field, visited all five counties and produced more detailed and objective reports. The PMOs recommended a greater technical expertise in the membership of Supervision Missions, but generally appreciated the collaboration. The 2001 mid-term review was critical regarding the targeting of the poorest, inappropriate training for illiterate farmers and inadequate support for women's IGAs.

The provincial level PMO was the nerve-centre of the project. The contacts between administrative levels are frequent and the managerial ethos is marked by enthusiasm and frankness. PMO leaders at county level have a strong sense of ownership of the project. The Mission visited 12 out of 34 townships, where informative reports were presented and even searching questions answered without confusion or hesitation. As for the VIGs, the Mission's initial doubts as to whether the necessary capacities existed at this level proved unfounded.

As co-financiers, the WFP took a close interest in the project and its supervision. The WFP office in Beijing provided IFAD with in-country representation. The management of the food aid was effectively organised and losses in transit kept to a minimal level.

Recommendations

Pilot schemes for the latest generation of microfinance approaches and/or the proposed Village Development Funds might be appropriate ways of building on IFAD's investments in Anhui province.

The experience of SWAIADP indicates that with time the ‘model farmer' approach to targeting was capable of reaching the poorest households. Where the necessary degree of homogeneity and social cohesion exists in villages, inclusion of all households is to be preferred. Microcredit lent to the poorest households should have more favourable conditions attached, such as longer terms and a greater number of instalments.

Saturation in the markets for raw silk and tea may pose a risk to the economic sustainability of project interventions. The diversification of income must be seen as an ongoing process sensitive to market needs and changes and as requiring further support, notably through appropriate training programmes for agricultural extension officers and ready access to market information for producers.

The training offered to women has not always been appropriate to their educational level and needs. Recommendations include: the replacement of traditional literacy training by simplified functional training; technical and credit training specifically for women; training materials with more illustration, a less densely presented text and a larger font; the training of trainers for more effective healthcare programmes; shorter and more frequent sessions; and smaller classes.

Data on the incidence of disease is not effectively utilised. Healthcare workers should submit the results of medical check-ups to the township PMOs who, with the assistance of the hospitals, can investigate the causes of high rates of disease in particular villages and tailor the healthcare training accordingly.

The aim of 100% loan recovery should be re-emphasised and ways of achieving it investigated and discussed at county and township level.

The baseline survey was over-complicated. The questionnaire ran to many pages and included superfluous data. Perhaps as a corollary of this, the sample of households (250 in all) was too small to be representative. For comparable IFAD projects in the future, a shorter questionnaire and a larger sample is recommended.

The approval of the AWPB required a cumbersome process, passing through three levels of aggregation and approval prior to the donor approval process involving IFAD, WFP and UNOPS. Simultaneous in-country approval of the AWPBs for ongoing projects should be mooted to ensure a more participatory consideration of the AWPB by all parties simultaneously, as well as the abbreviation of the process.

---

1/ The completion evaluation mission was composed of: Mr Roger Norman, team leader; Mr Pham Quang Hoa, agriculturalist; Mr Yusaf Mahmood, management and institutions specialist; Ms Han Zheng, gender and targeting specialist.  Mr Flemming Nichols, then OE Evaluation Officer, was overall responsible for this evaluation and joined the mission in its final week in the field.

2/There are 15 mu to a hectare.