Mid-term evaluation

Introduction

Background. The Agricultural Resources management Project (ARMP) was designed in late 1995, and was the fourth IFAD supported intervention in the Hashemite Kingdom of Jordan. The previous interventions had assisted in promoting the government's strategies for rainfed agricultural development and the extension of credit to smallholders and women. A Mid-term Evaluation (MTE) of the Income Diversification Project (the third intervention) took place in 1998. The purpose of the present MTE, which is taking place in the fourth year of project implementation, is firstly to assess the achievements of the project so far and the impacts on the target groups, secondly to examine the need for adjustment and/or reorientation of the project design and implementation strategy to address identified constraints, and thirdly to suggest lessons from this experience which might be of relevance for similar projects. The outcome of the MTE will be a set of recommendations/lessons emerging from the implementation experience of the project so far, which can be used to modify the implementation methodology for the remainder of the project life, in order to address the identified constraints. These will be presented in the Agreement at Completion Point (ACP), which will remain as a reference point for future co-operation among the partners. The MTE mission was fielded in September/ October 2000.

The implementation environment

The macro-economic situation. Jordan has a small open economy with a Gross Domestic Product (GDP) at current prices estimated at JOD 5.3 billion (USD 7.5 billion). The country's population is currently 4.9 million with an annual growth rate of around 3% (1999 estimates) and per capita income is JOD 1 080 or USD 1 523. The economy is characterized by a dominant service sector (finance, government and other services, transport and communications, tourism and trade) which accounts for 69% of GDP. Jordan has limited natural resources and is highly dependent on imports, particularly oil and food stuffs. The economic prospects for 2000 were rated as encouraging, and the real growth rate is expected to outstrip the predicted population growth rate: annual inflation should be maintained at the present low rate of below 2%. The economy is also expected to benefit from increased remittances from Jordanians working abroad, and higher tourism receipts. Overall, with low inflation, a stable currency and a high level of foreign reserves (sufficient to cover about nine months of imports), Jordan has a very stable macro-economic environment. The country's major challenges are to resolve the unemployment and poverty problems and continue with the current economic reforms to achieve and maintain a good growth rate, curtail the trade deficit and further reduce foreign debt, particularly commercial borrowings.

Poverty and social initiatives. The government of Jordan (GOJ) has been allocating resources to poverty alleviation efforts for many years, but these were mostly in the form of social welfare programmes. Only recently has the government tried to target its efforts more specifically towards self-help activities and has also recognized the importance of employment creation, especially for the large youth sector. A 1997 study conducted by the Ministry of Planning (MOP) found that 21% of the sample population were under the absolute poverty line and 11% were under the abject poverty line, and that the rural poor were generally the landless, the small farmers, herdsmen and rural women. Following these findings GOJ prepared a Social Productivity Programme (SPP), the first three-year phase of which was launched in 1998.

The agricultural sector. Jordan's natural resources for agricultural production are very limited. Only about 5% of the land mass is considered arable, and Jordan is among the world's most water-deficit countries. The challenge for the Government is, therefore, to promote sustainable use of natural resources. The contribution of agriculture to Jordan's GDP, including forestry and fisheries, declined from 6% in 1995 to 4.5% in 1999 due to drought during the period 1997-1999. However, because of strong upstream and downstream linkages, an estimated 28% of the GDP is considered as agriculture-dependent. The sector provides livelihoods for about 20% of the population and employs about 7% of the labour force. Livestock is estimated to contribute about one half of the total agricultural GDP. Following sector reforms, agriculture in Jordan is now virtually free of all controls and restrictions and all direct subsidies have been removed. Credit to agriculture at low interest rates is now the single most important conduit for GOJ subsidies to agriculture.

Agricultural finance. The Agricultural Credit Corporation (ACC) is the sole institutional source of formal credit to both individual farmers and members of village cooperative societies. ACC is a government-owned corporation and not a bank: its current activities are confined to providing credit for agricultural production and development and investing in agricultural companies. ACC's operations are sustainable and its financial interest is protected because of a system of salary deductions from guarantors. Jordan's financial institutions, including ACC, are at present not geared to servicing small scale producers who, apart from the problem of easy access, often find it difficult to meet credit terms and conditions, particularly relating to collateral requirements . This void is being filled, on a limited scale, by some local and international Non Governmental Organizations (NGOs) who have been successful in operating micro-credit programmes for the benefit of both urban and rural poor .

Project objectives and design

Rationale and objectives. The rationale for ARMP is based on supporting rural households in the two Governorates of Karak and Tefila to intensify their agricultural practices by enhancing natural resource management in the designated project areas. This is to be achieved by arresting soil degradation and restoring soil fertility through the sustainable use of land and water resources. In the mid '90s the project area, lying in the southern highlands, had received proportionally less investment than other areas in the north which had more and easily accessible natural resources. As a result, rural households in Karak in particular, were reportedly the poorest in Jordan. The majority of farms are small and utilize low input techniques to minimize the risks inherent in an area of very uncertain rainfall. Whilst it was concluded that agriculture would remain the mainstay of the local economy, the project design also recognized that complimentary activities should be developed to allow for income diversification and the fuller involvement of all members of the family. The overall objective of the project, within the GOJ strategies for the development of rainfed farming, is to improve the income stability of vulnerable, resource poor farmers by safeguarding and improving the productive potential of their natural resources and enhancing their returns to labour.

Components. To meet the objectives the project was designed with five components, viz: (i) resource management (56% of total costs); (ii) agricultural development (18%); (iii) institutional strengthening (10%); (iv) women's development (10%); and (v) co-ordination and management (6%).

Target groups. At the time of design the number of rural households living below the poverty line in the project area was estimated at about 7 000. The basic target group definition adopted by the project was farmers with land holdings of less than 50 du (5 ha) in size: for this group the project provides soil and water conservation works (SWC) on a highly subsidized basis to protect a maximum of 25 du (2.5 ha). Above the 25 du maximum, farmers were expected to take loans for SWC from the credit line provided to ACC. Farmers selected for SWC were also automatically included for agricultural development activities, but this was not exclusive and some orchard rehabilitation works were extended to other groups. All women-headed households applying for SWC activities were to be considered eligible.

Implementation approach. Responsibility for project implementation was vested in the Ministry of Agriculture (MOA), with the Directorate of Projects (PD) assuming the lead role, but the actual implementation tasks were to be undertaken by the Directorates of Agriculture based in the two Governorates. Provision of credit was the responsibility of ACC, which was to have a separate subsidiary loan agreement for the project. The design called the selection of areas for project interventions based on both technical and poverty criteria. A participatory community approach was to be adopted, working in co-ordination and partnership with organizations already in the field. In order to introduce such an approach, in 1996 Care International received an Extended Cooperation Programme (ECP) grant towards the costs of undertaking community based water conservation schemes in four villages in the project area. The intention was to work with the government agencies to develop an acceptable methodology, which would then be adopted more widely throughout the project area. These schemes were completed in 1998.

To further define the socio-economic conditions in the project areas and to help with targeting and later, evaluation, the design called for the Project Management Unit (PMU) to carry out a baseline survey soon after the project start-up.

Project costs. The total project cost was estimated at USD 18.6 million, of which the IFAD loan is about USD 12.8 million. The project has a seven-year implementation period, closing on 30 June 2003. At the end of June 2000 overall project disbursement was 51%, the bulk of which was for civil works for soil conservation.

Implementation progress

Resource management

The Resource Management component consists of on- and off-farm SWCs, technical assistance and training and the provision of vehicles and equipment. The on-farm activities cover the construction of stone walls and terraces, contour furrowing and the development of cisterns for on-farm irrigation, whilst off-farm activities include wadi bank and flood protection measures, including check dams and gabion structures and the rehabilitation and improvement of springs and their associated water distribution networks. Findings are as follows:

• The project has made good progress in the on-farm soil and water conservation programme, and has provided conservation structures on some 36 000 dunum, or 44% of the appraisal target. Some 2 950 beneficiary farmers on 1 650 farms have participated in the programme. About 1 170 cisterns have also been completed. Overall, high standards of construction have been enforced. There has been a continuing high demand by farmers for participation in the SWC programme as a whole and for cisterns. However, under-estimation of unit costs at appraisal has resulted in expenditure at the time of the MTE of almost 79% of the allocated base costs for the SWC programme and 64% for cisterns.
• The implementation approach used by the project has been based on individual or family farms rather than on a whole village or watershed basis, as foreseen at appraisal, and also uses simpler land use planning methodology. Discussions with beneficiaries and PMU staff in the field and in the stakeholder workshop endorsed the present project approach and highlighted the difficulties which the project would have faced to introduce the more participatory village-based approach described in the design, which under-estimated the complexity of the land holding patterns. However, the minimum farm size criterion used by the project (4 du) for inclusion in the SWC programme is of concern, as this excludes some 20% of farms from participating.
• Following recommendations of the Water Harvesting Division in the Ministry of Irrigation and Water Resources (MIWR), larger water storage structures known as mini-dams and hafirs (off-wadi storage ponds) have been built instead of the smaller check dams described in the design. Eight mini-dams and four hafirs had been completed by 30 June 2000: in practice, these 12 structures have the potential to impound the same volume of water as the total number of check dams that could have been built, at a significantly lower cost per unit of water. This is partly because of under-costing at appraisal.
• ARMP has made good progress with spring rehabilitation, with 43% of the appraisal target for springs and 141% of the appraisal target for canals already achieved. Most spring rehabilitation works appear to be well constructed, but the approach adopted is only partially participatory. The MTE concluded that limited staff numbers and shortage of transportation was affecting this programme. Under present project arrangements, beneficiaries are not expected to make any contributions towards spring rehabilitation or canal lining works in terms of labour or cash. However, they were found to make substantial investments after the rehabilitation in on-farm distribution systems, including drip irrigation equipment, pipes and small storage ponds. Although the appraisal calls for the formation of formal spring user groups, the project has only worked with existing traditional groups.

Agricultural development

The Agricultural Development component consists of the establishment and/or rehabilitation of orchards and vineyards, support for the extension service, extension packages, demonstrations for field crop intensification, and training of staff and farmers. The MTE found that:

• Overall, about half of the total appraisal target for new orchard plantings has been achieved with almost 600 beneficiary farmers. At this stage, about a quarter of the area developed under the SWC programme has been planted to fruit trees, primarily olives. Essentially all of the new orchards are on undeveloped land without previous cropping. The project has also subsidized farm fencing, which was not included at appraisal.
• There has been limited uptake of fruit tree rehabilitation (about 17% of the appraisal target), partly due to farmer reluctance to deep prune bearing trees. The MTE concluded that the appraisal target overestimated farmer demand for this activity. A limitation of follow-up to the first year after pruning, is also probably restricting uptake.
• There has been little effective progress and no discernible impact from the intensification of field crop production activities, which involve demonstrations of new varieties and cereal/vetch rotations.

Rural financial services

A distinct Rural Financial Services (RFS) component was not included in the design of ARMP, rather various RFS activities were dispersed over three of the project components. However, all the activities were to be undertaken by ACC, funded through a subsidiary loan agreement. The anomaly in the design was remedied somewhat in the Loan Agreement, which described the project in terms of "parts" rather than components (see description of the project in Schedule 1 of the loan agreement). The RFS activities in total then became one "part", and are so grouped for allocation and withdrawal of IFAD loan proceeds. For the purposes of the MTE, RFS activities are grouped and treated the same way as in the loan agreement.

The RFS activities comprise a credit line to ACC for on-lending to individual farmers and women or their groups, and assistance for ACC to continue its restructuring and reform programme (which was initiated under the ongoing IFAD-assisted Income Diversification Project). The overall resource allocation for RFS amounts to USD 2.33 million, 80% of which is for incremental credit. About one-third of this is intended for women's Income Generating Activities (IGAs).

Implementation progress is as follows:

• ACC has developed a sound system for the delivery of credit under ARMP with due attention to financial viability and sustainability of its operations. ACC's loan policies and procedures for its regular credit programme also apply to project lending. However, The MTE found that there were variations in fixing loan amounts and maturity periods.
• Credit Disbursements. To date, just over JOD 1 million (USD 1.43 million) of credit has been disbursed, about 76% of the appraisal target. This achievement is due to much greater demand than anticipated for IGA loans from rural women, who use the credit for food processing, goat and sheep breeding and dairy/milk processing. Together these activities account for 86% of the total number of loans and 84% of the total amount disbursed. By comparison, credit disbursements for land consolidation and supplementary SWCs are far below the appraisal estimate and account for just 11% of the amount disbursed. This seems to have arisen because of insufficient knowledge at the design stage of the present land tenure arrangements and the social pressures leading to land fragmentation; credit alone is not an appropriate mechanism to address this problem.
• Loan Collateral and Maturity Periods. The ACC policy on collateral requires monthly salary deductions from the salary of the guarantor. Although this has ensured 100% loan repayment, it has some adverse implications. Firstly, the poorer women who are not able to comply with salary deductions do not have access to ACC credit, whilst women borrowers who can offer sufficient salary deductions can obtain larger loans than strictly needed for their IGAs, and can then divert excess funds to household and consumption purposes. Secondly, grace periods, even if justified, are not allowed and since loan repayment periods are not linked to the net cashflow of the IGA financed, the debt burden and interest payments are often much more than necessary.
• Beneficiary Constraints. Credit beneficiaries highlighted the two main constraints to loans as firstly, the lag between loan applications and disbursements (up to as long as two to three months, due to the two-level vetting system), and secondly the lack of support to market processed products.
• ACC Institutional Strengthening. ACC has been and remains reluctant to borrow funds for institutional strengthening, as these funds have to be repaid by ACC using the same mechanism devised for the credit funds. Only a small allocation of funds earmarked for institutional strengthening has therefore been used, and the use of the substantial balance seems unlikely. Progress with the restructuring of ACC and possible conversion into a rural finance bank has been minimal.

Women's development programme

The main development resource for women in ARMP is credit. It was envisaged that women's small enterprises would generate additional household income and provide opportunities for remunerative activities during slack periods in the agricultural calendar. Under the Women's Programme training and technical assistance for the preparation of feasibility studies is provided, whilst under the Institutional Strengthening component is included the identification of NGOs to support implementation of the programme. Progress has been as follows:

• There has been no identification of government organisations or NGOs to assist with the implementation of women's activities, as envisaged in project design, although there is both scope and need for the organisation of outreach to poor women and for production and marketing to take place on a collective basis.
• A participatory assessment indicated that many of the problems faced by women with respect to enterprise development, production and marketing, might be minimized or resolved with a group approach.
• Of the women who have taken loans from ACC, only 37% have actually started enterprises. The principal reason given was the inability to sell the produce from the IGAs. The proportion of loans actually utilised for IGAs was found to depend to a large extent on the type of enterprise, previous experience, commitment to the enterprise and household consumption and investment requirements.
• Whilst a WID unit has been established and the five female extension agents are technically qualified, they are not properly trained for the purpose of implementing the programme.

Institutional strengthening and co-ordination and management

The Institutional Strengthening component consists of support for the Field Directorates of Agriculture (FDA) and to the ACC Branch Offices at Karak and Tafila whilst the Co-ordination and Management component is mostly to support the establishment of a PMU. Progress has been as follows:

• The MTE found that the organisation and management aspects of the project were well developed and that adequate numbers of qualified staff had been recruited by MOA in order to implement the project. The organisation is essentially the same as proposed at appraisal, except that the position of Community Development Officer has been amalgamated with the position of Extension Officer in the PMU. However, the organisational arrangements do not consider the post-project sustainability of the Field Units that have been created and at project closure it appears to have been assumed that the Directorates of Agriculture would continue project activities. The project has obtained all the vehicles and equipment specified at appraisal
• There is little integration or handover of completed development activities to other Units for routine follow up operations. This may eventually affect such aspects as the final area planted, and hence the flow of benefits.
• Although many staff identified a need for more training, the project has actually undertaken an extensive programme of staff training, and staff field trips have also been organised to Egypt and Tunis.
• The design of ARMP included technical assistance (TA) in specific areas to support staff resources, enhance staff capabilities and to develop final recommendations for design parameters. Project costs included TA in inter alia water harvesting, software development and research. The MTE could only find TA had been used for water harvesting (from the International Centre for Agricultural Research in Dry Areas, ICARDA). In addition the project has made no use of the funds included for NGO support, and there has been very little use of the budget for research and studies. These funds were intended as an integral part of the project design, and were, for example, to identify activities to widen the scope of IGAs for women, or for contracting research on improved technology.
• Project Expenditure. PMU records indicate that at the end of June 2000 about 51% of the loan funds had been expended. Counterpart funds have been adequate and are mostly provided in a timely fashion. However, expenditure by component and sub-component activities show significant variations from appraisal estimates, either because of unit cost differences or because activities have not been adopted as predicted. Experience at the mid-term has led the Project Management to suggest the adjustment of physical targets and to propose a re-allocation of funds, shifting towards the activities which are more in demand by the beneficiaries.
• Monitoring and Evaluation (M&E). The M&E unit has developed its own software based on Microsoft Access: no assistance was received in the establishment of this system or the preparation of reporting formats. The MTE was impressed with the breadth and depth of the system that has been developed. The project reports are used as the basis for management decisions, hence form a valuable management information tool. Despite being included and costed in the design, no baseline survey has ever been carried out by the project. As project objectives are described in terms of improving the livelihoods of the beneficiaries, it is unclear exactly how the achievements of the project were to be judged, other than in purely physical or financial terms. Being aware of this problem, a socio-economic survey has recently been undertaken by a team from MOA, but while this provides more information on who has benefited and their preferences, it does not go far into actually assessing the impacts. The project management is aware of this problem and is actively seeking to include impact indicators in its regular reporting. However, the MOA socio-economic study proved an important source of information for the MTE.

The experience with care

An Extended Co-operation Programme (NGO/ECP Grant No. 068) was provided by IFAD to CARE in October 1996, with the objective of demonstrating a participatory approach to rehabilitating and managing spring water resources and developing methodologies that could be replicated in other springs and community activities. The ECP took place in four pilot villages in the project area. The CARE methodology was to work through communities that had local voluntary societies, with community contributions for construction and maintenance organised and managed by the societies under CARE supervision. A counterpart contribution of 12% for the construction works was expected from the community. A water committee was formed in each village, and regular payments were expected to be made by the beneficiaries for operation and maintenance of the rehabilitated system. The selection of springs, recruitment of local counterparts and mobilisation and training of the communities was carried out by CARE. The linkages to ARMP were in the design of the structures by ARMP staff, the participation of CARE in a training course for ARMP staff in PRA techniques, and direct involvement of two PMU staff in this programme.

Whilst the spring rehabilitation exercise in the four villages was successful, the CARE approach in its totality could not be used by a government agency (because it involved depositing money in a local development account). Visits by the MTE mission did not reveal any difference in the springs rehabilitated by CARE and those rehabilitated by ARMP. In addition, MOA already had an approach to undertaking spring rehabilitation and it would have required a policy decision for alternative methods to be adopted by project staff, who in any case were not in favour of this approach, as they considered that it would involve a substantial amount of additional work without significantly increased benefits. The MTE concluded that the CARE methodology had not clearly demonstrated its value added for ARMP.

Impacts and sustainability

Environmental Impacts. The recent IFAD Country Strategy and Opportunities Paper (COSOP) emphasises the critical situation in Jordan with respect to water availability, pointing out that water resources are presently being exploited at 150% of sustainable yield levels. For water conservation, the project should easily exceed its appraisal target, resulting not only in more water for livestock, but also increased recharge of groundwater. In addition the achievements for soil conservation measures are impressive - for stone walls so far about 70% of appraisal targets have been met. The impacts of soil and water conservation will not become immediately apparent as it take a number of years for the stone walls, in particular, to become effective in reducing soil erosion and water runoff. Nevertheless, as technical interventions, these conservation structures are virtually permanent (with minimum maintenance).

Specifically, in terms of environmental impacts the project is likely to result in reduced soil erosion and increased soil fertility (measured by crop productivity) and increased groundwater recharge. The combined effect of these improvements will result in the increased robustness of the farming systems, and greater water availability for both irrigation and municipal/domestic uses. No negative environmental effects were identified, although there is some concern at the possible concentration of livestock around the mini-dams and the extent of forage which will be available in the rangelands if animal numbers increase significantly.

Production Impacts. Based on mission estimates and farmer interviews, future yields of newly planted olive trees are expected to average about 437 kg of olives per dunum in a normal year at full production, provided supplementary irrigation is available from cisterns or other water sources. This yield would be expected to provide an incremental return of JOD 250 per dunum . Little impact has been observed as yet from the fruit tree rehabilitation programme, since trees that have been deep pruned will not produce a yield until the fourth year, but future yields are expected to increase by over 50%. This increase would produce an incremental return of about JOD 44 per dunum. The overall impact of the cistern construction is difficult to determine at this stage, although cisterns have a positive impact on the ability of the farmer to plant trees. Cisterns may also be used for domestic water supply. A major impact is the savings on the purchase of water. The MTE could find no impacts from the cereal demonstration programme.

Credit Impacts. Although impacts of the credit provided so far are limited, the MTE concluded that goat and sheep breeding loans, and to a lesser extent loans for dairy/milk processing, are in general having a positive impact in terms of return to family labour and addition to household income. The impact is marginal in the case of loans for food processing mainly due to marketing problems. There are clear indications of significant improvements in family nutrition with goat/sheep breeding and dairy/milk processing enterprises. The project lending so far has helped generate rural employment, developed entrepreneurial skills among the rural women and enhanced their economic clout and recognition. It was estimated that for the 220 women borrowers who have actually undertaken enterprise development with loans from ACC, the average increase in incomes has been about JD 150 per activity, which represents an increase in annual family income of about 10% to 12%.

Institutional Strengthening. MOA staff reported that the training received had been very beneficial in assisting with their work activities, but requested for training subjects to be extended. The MTE concluded that the training provided is contributing to enhancing the capacity of MOA generally. By comparison, the institutional strengthening provided to ACC has had almost no effect, simply because the arrangements under which it is to be provided are not considered appropriate, by ACC.

Sustainability. On balance the MTE concluded that the project interventions would be largely sustainable after project closure. The structures completed by the project require little maintenance and are most probably sustainable in the longer term. This also means that as soil and water conservation is improved, these improvements will also be sustainable. There are some maintenance requirements for the dams and hafirs, but this responsibility has been accepted by MIWR. As benefits increase from the planting or rehabilitation of tree crops, then annual inputs from the farmers should also be more likely to maintain yield levels. Repayments of the credit funds are almost 100% because of the system of guarantors, which means funds will continue to be re-cycled. Present IGAs give relatively low returns; the intervention would become more robust if a greater choice of more profitable enterprises were introduced - such an approach would also increase sustainability.

The staff in the PMU are young and have received useful training and work experience. At the end of the project life the PMU will no longer be needed and MOA needs to address the question of the re-location of the staff, or the continuation of project activities with alternative funding. The new Field Units created in the Directorates could be retained as separate units or absorbed into other FDA activities, on the basis that there will be additional agricultural activity in the Directorates, and continued support post-project will be needed to ensure impacts are optimised.

Issues

Re-allocation of funds

The changes proposed by the project management affect the resource management and agricultural development components, in particular. Under these headings some new activities have also been proposed. The changes proposed are justified by the PMU on the basis of the implementation capacity in the project, the expressed farmer demand and the results from the recent socio-economic assessment. Essentially the changes maximise the component activities and rationalise the various cost anomalies.

For Resource Management the PMU has proposed, for on-farm SWC: (i) a reduction in the total area of on-farm SWC measures from 82 500 to 75 500 du, plus a greater emphasis on fruit tree planting (see Agricultural Development); (ii) a change in the relative quantity of the different structures, in particular an increase in stone walls, in response to farmer demand; and (iii) an increase in the volume of water stored in cisterns, and the rehabilitation of old cisterns. For off-farm SWC the PMU has proposed: (i) to revise the number of dams to take account of the larger structures and to allow for the hafirs which are being built; (ii) to increase the length of gabions; and (iii) a new farm access road activity.

Under the Agricultural Development component the proposals are for: (i) an increase in the area planted so that a greater proportion of lands protected by the SWC activities is planted; (ii) the recognition of the existing activity of farm fencing; (iii) an increase in the length of canals rehabilitated as part of the spring rehabilitation; and (iv) a reduction in the area of orchard rehabilitation to reflect the project experience to date. In calculating the cost implications the PMU has also anticipated savings in other project activities. On balance, therefore, the net incremental cost amounts to JOD 1 064 000 (USD 1.52 million). The MTE estimated that about USD 1 million would be required from loan funds.

The MTE endorses the proposals submitted by the PMU. They are based on a full understanding of project realities, farmers' demands and the constraints faced. In the case of ARMP, the analysis also serves to highlight shortcomings in the design, and it is as well that these are addressed at this point. The revised targets will not change the nature of the project itself (and hence the objectives), but will simply alter and re-balance the targets on a demand-driven basis. The new activities proposed reflect and address the realities of project implementation. If there is a weakness in this request it is that the project management is unable to actually assess the likely changes impacts which will follow the re-allocation, because the M&E does not contain adequate impact indicators. This shortcoming cannot be addressed by the occasional socio-economic survey. The project management is aware of this, and is taking measures to address this in the second half of the project.

Issue 2: participation in the project

Although the project design describes an overall participatory community-based approach for the implementation of ARMP, in practice there is no systematic approach to community participation and in most cases ARMP and ACC staff are dealing with individuals. Community participation entails costs for both the implementing agency and the beneficiary community. As such, the benefits and value added of a community participation approach have to be clearly demonstrated for each aspect of a programme or activity in which community participation is envisaged. The CARE approach failed in this respect. The Ministry's strength lies in its technical abilities, and its policies are to improve production and productivity through the adoption of technical innovation, hence unless the additional benefits from participatory approaches are clear, there is little incentive for adopting a different approach to implementation.

However, the lack of community participation has led to the beneficiaries viewing the project interventions as passive recipients in some of the project activities such as spring rehabilitation, mini-dams and Hafirs. For MOA overall, a modified approach to include a greater degree of community participation in the second half of the project would provide the opportunity to realize the increased benefits from this approach and would expose staff to a learning process for the benefit of future projects. In its recommendations the MTE proposes that such an approach is included for the second half of the project and describes the mechanisms necessary to do this.

The participation of an NGO for social intermediation in some of the collective aspects of the project, such as spring rehabilitation, would be desirable. However, no suitable partners were identified in the field visits and the only real scope for participation of NGOs under ARMP seems to be with NGOs operating marketing programmes for women. For other components, it is concluded that ARMP itself needs to develop an approach to community participation that would add value to its work and enhance the sustainability of its investments. This will require a strategic commitment from the Ministry and the use of the funds available for technical assistance to develop the appropriate skills.

Issue 3: the approach to credit

Although mitigating poverty is a high priority for GOJ, ACC's loan terms and conditions are stringent and not appropriate to ARMP's poverty alleviation objectives. Almost certainly, project credit beneficiaries are not from the poorest groups. There is an urgent need to ensure the coverage of assetless rural women who belong to the most underprivileged groups and who need small loans to finance IGAs. If IFAD's development assistance strategy in Jordan is to continue support for GOJ's efforts to improve the social and economic well-being of the rural poor, especially women, then either an alternative window needs to be opened through ACC, with appropriate support to mitigate the increased risks, or another credit channel needs to be tried or developed. In its recommendations the MTE proposes a compromise approach through ACC. Its emphasis is on introducing a simple mechanism, which would be applicable for small loans, with minimum collateral requirements. The mechanism also allows for the protection of ACC's interests, through reduction of costs and risks of lending to the poor.

Recommendations

General recommendations

Participation in MOA Development Projects. It is strongly recommended that MOA utilise the remaining three years of project life to develop its own approach to participatory methods for use in its projects. This would involve a number integrated activities, as follows:

• A clear policy statement on participation needs to be made, and operational guidelines and instructions prepared for the implementation of the policy.
• In the case of ARMP, the post of Community Development Officer in the PMU, which was scrapped at the beginning of the project, needs to be reinstated, and four more field officers appointed in the Field Units.
• The participatory approach needs to be integrated into the work plans and work habits of all members of the ARMP team or it will be unlikely to succeed, hence an orientation, training and visits programme needs to be devised for ARMP and MOA headquarters staff in community participation approaches to agricultural development.
• ARMP should make use of its technical assistance budget to employ short-term TA for community participation.
• Once these steps have been undertaken, then the community participation approach (fully described in the main report) should be introduced in the remaining period of project life.
• There is need for careful monitoring of the value added and impact of adopting a community participation approach. The existing monitoring and data collection system should therefore be upgraded to include impact indicators for community participation.

Re-allocation of Funds. The project funds should be re-allocated along the lines put forward by the project management. However, before a re-allocation is submitted the project should examine the full consequences of the re-allocation on all project interventions, submitting justifications based on estimated relative impact assessments for the quantitative revisions. In addition the increased number of beneficiaries and the flow of benefits should be projected. ARMP should consider asking IFAD to assist with the preparation of this document, and should indicate the increased contributions expected from GOJ and the beneficiaries.

Resource management

For On-Farm Soil and Water Conservation it is recommended that:

• The development of land use plans and on-farm soil and water conservation works should continue to be on an on farm basis rather than village or watershed basis.
  
• The minimum size criterion of four dunums for participation in the on-farm soil conservation programme should be replaced with a programme to select farmers based on poverty criteria.
  
• There should be an expanded demonstration and field testing of other water harvesting methods (such as micro-catchments) by NCARTT/Extension system, with the project role including the identification of farmers and assisting in the organisation of field days and training.

For Off-Farm Soil and Water Conservation it is recommended that:

• Detailed analyses of the number of potential beneficiaries and livestock should be conducted to assist in the selection of sites for mini-dams and hafirs, the estimation of capacity required and to determine benefits from this activity.
  
• In principle, the size of the dams should be limited to maximum of about 15 000 m3 to allow a larger number of smaller structures to be constructed.

For Spring Rehabilitation it is recommended that:

• Additional resources, including a field engineer and vehicle, should be provided in each field unit to be responsible for spring rehabilitation and canal lining activities . This would allow for greater concentration of effort on spring rehabilitation, plus greater beneficiary involvement in all the activities.
• The size restriction of two dunums should be lifted to allow staff to work with groups of farmers in spring areas, both for technical reasons (common spraying against pests such as scale insects and olive stem borer) and to increase benefits from spring rehabilitation works.

Agricultural development

For Fruit Tree Planting and Rehabilitation it is recommended that:

• Project extension staff follow-up the maintenance of newly planted fruit trees until the trees start bearing, since optimum yields, and hence the best return on the investment, can only be obtained through correct and timely pruning, fertiliser application and the use of plant protection materials.
• Similarly, follow-up in the years after rehabilitation is also required, since full benefits cannot be achieved without continued fertiliser applications and pest control.
• There is scope for carrying out demonstrations and field tests of tree crops such as pistachios, particularly in southern parts of Tafila. This activity could be carried out by the NCARTT station at Shoubek under an agreement with the project, using project funds.

Intensification of Field Crop Production. Cereal crop demonstrations should be discontinued directly through the project. However, there may be a longer-term impact of the demonstration activity in support of GOJ's policy to encourage a shift in cropping patterns. Accordingly, it is recommended that NCARTT, which is the MOA agency directly responsible for technology development and transfer, should undertake this activity under an agreement with the project using the funds allocated for this activity.

Rural financial services

For rural financial services the following recommendations are made:

• Eligibility Criteria. To ensure that those from poor households benefit from project support, the PMU should ensure that a uniform family income ceiling is applied throughout the project area to identify credit beneficiaries.
• Loans for Supplementary SWCs. In order to ensure compliance with the project Loan Agreement, ACC should not allow loans for the construction of farmhouses as part of SWC activities.
• Broadening Credit Outreach. To provide women from the most disadvantaged households with access to credit, the project management and ACC should initiate action to develop alternative mechanisms that broaden the credit outreach. The objective should be to demonstrate the effectiveness and sustainability of a micro-credit development model. This would seek to benefit from the strengths of NGOs in understanding the problems of the rural poor as well as compensating ACC for the higher transaction costs and credit risks, and which could, based on experience gained, be replicated in other future projects. A suggested approach is described in detail in the main report.
• Determination of Loan Amounts and Repayment Periods. Loan amounts should be strictly limited to the outlay required for the IGA financed, as the burden of debt service for a higher loan than necessary results in a reduction of the financial impact in terms of increased income for the borrower. In addition, instead of the present standard period of eight years, loan repayment periods should be adjusted to the net cashflow of the activity financed and include, where justified, appropriate grace periods. The implementation of these measures would permit ACC to service a larger number of borrowers with the available loan funds, and this process would be further facilitated due to faster recycling of loan repayments through the Revolving Fund.
• Loan Processing Delays. In order to reduce delays in approving credit applications, the PMU and ACC should, in joint consultation, introduce a system under which all loan applications, after a preliminary scrutiny and certification of eligibility, are directly sent to the ACC branch for appraisal and approval.
• Revolving Fund. ACC should maintain the Revolving Fund in an account to which interest accrues at a positive real rate. This is also of benefit to ACC, as the subsidiary loan agreement provides for ACC's repayments to be in JOD, at the exchange rate prevailing at the time the funds are repaid.
• ACC Restructuring. The study on ACC restructuring and transformation into a rural finance bank should be expedited. This would, among other things, facilitate introduction of savings-linked micro-credit programmes, which have significant advantages for poor rural clients (e.g. consumption smoothing and saving for loan repayment or unforeseen eventuality).
• Increasing Funds Allocation for the Women's Programme. Since the demand for loans for land consolidation is very low consideration should be given to the re-allocation of these credit funds to meet increased demand under the women's credit programme. In addition, further use of substantial funds allocated for institutional strengthening is unlikely, and these might also be used in the credit programme.

Women's development

For the Women's Development Programme recommendations are as follows:

• Improved Technical Support. The appraisal included a range of TA intended to fill the knowledge gaps and refine the project design. This TA should be immediately initiated, covering:
• a feasibility study to identify suitable income generating activities;
  
• NGO support under contract to assist in pre-credit activities for women, including group formation and training;
  
• provision of funds to support action-oriented market research to assist women market their products
• Production and Marketing centres. The project needs to address the twin problems faced in establishing and maintaining production standards, and in marketing. The approach recommended is to establish, on a trial basis, district-based production and marketing centres, using funds which would otherwise be unutilised from the women's programme. These units should be established in districts where an NGO or CBO is willing to undertake the responsibility for implementation with the project supporting initial set-up costs. These centres should also be used for training. Technical assistance should be employed for the development of proper brand names, the standardising product specifications, proper packaging and sales. These centres can be developed to act as consolidation and marketing outlets for village based production units where women form groups to undertake production activities jointly. Group based activities should be encouraged, possibly by the provision of storage and processing equipment.
• Use of the ACC database. ACC has developed an excellent computerised database of its loan applicants and keeps good socio-economic data on its borrowers. However, this information is not used for programmatic decision-making or for feedback to ACC or ARMP management. This data should be analysed and used for regular reporting and evaluation of the loan applicants, and used by ACC for improving its loan products and portfolio.
• Staff Training. There is a need to provide institutional and technical support to Women in Development (WID) staff through additional training in enterprise development, group formation, communication skills and monitoring and evaluation skills.

Project management, co-ordination and institutional strengthening

The recommendations for these components are:

• Technical Assistance and Studies. Both the scope of the project and the capacity of staff have been restricted by: (i) the reluctance to make use of the technical assistance provided for in the project design; (ii) the non-engagement of outside agencies in areas where the Ministry does not have competence; and (iii) the lack of research and studies where necessary to develop the technology required for the project. This should be corrected in the second half of the project life. The TORs of the Permanent Steering Committee (PSC) should be extended to pay particular regard to these shortcomings.
• MIWR as an Implementing Agency. The MIWR should be recognised as one of the implementing agencies for the project, especially concerned with the optimisation of water resource use. The role of MIWR should be defined and MIWR should be represented on both the Regional Agricultural Coordination Committee (RACC) and the PSC.
• Post-project activities. MOA needs to address the question of the continuation of project activities post-project, and the role of the additional staff in the SWC Field Units. This should also include the expansion of the routine work programmes of existing Field Units to ensure that benefits and impacts of the development activities undertaken are fully realized.
• Clarification of Coordination Roles. The Coordination and Steering Committees both have important, but very distinct roles to play in project implementation. It would be useful if both committees prepared terms of reference for their areas of authority, in order to avoid wasteful and unnecessary frictions arising between them. In general decisions concerning only ARMP should be taken at the Regional level, whilst decisions with implications wider than ARMP should be taken by the PSC.

Lessons

Institutional Strengthening. The institutional strengthening for ACC was poorly conceived and is unlikely to be used. The reason for this lies in the terms and conditions applicable to the repayment of funds from the credit institution to the borrower government; for institutional strengthening these are the same as those for the credit line. For institutional strengthening this is unreasonable and places in jeopardy the attainment of the objective of ensuring a strong presence of the credit agency in the project area, for closer and regular interface with the beneficiaries. In this instance ACC should be released from this penalty, but in a more general sense grant funding should be preferred for institutional strengthening.

IFAD is currently supporting four projects implemented through PD in MOA. Although institutional strengthening is funded for ACC and to establish the new Field Units in the four Directorates of Agriculture under ARMP, no support is provided for PD, even though, under the management system in MOA, the Directorate has active roles in all the projects, and has central units to review progress. This task is made more difficult because all the IFAD-supported projects in Jordan use different implementation and monitoring mechanisms. This is a missed opportunity. If the monitoring and MIS systems for these projects had been unified, then a more strategic overview of the development impacts of the portfolio could be established, and monitored by PD. This would also have been of help to the M&E Directorate of MOA, which has to report on the progress of the whole portfolio (plus the other development projects).

The lesson is that in articulating IFAD's approach in the country it is worthwhile to try and identify agencies such as PD, which have a central role to play in the development of the portfolio, and to consider strengthening for such agencies to improve their monitoring role for the whole portfolio.

Credit Institutions and The Poor. Specific and well-defined interventions to broaden the credit outreach to cover the assetless poor in the target group should have been included in the project design. This would mean sufficient attention to capacity building and participation to motivate and promote self-help and self-reliance through group action and training, and assistance to cover transaction costs and credit risks of the credit network to encourage lending to the poorest in the target group. Standard ACC terms and conditions for loans are not appropriate to underprivileged rural women. It is unrealistic to expect a credit agency such as ACC to choose to develop a specific window for the poor, unless sufficient incentives and risk funds are both included in the project costs, and there is at least the outline of an agreement as to how the arrangements will work. Even then, the commitment of the credit agency may not be assured.

Commitment to Community Participation. Community participation entails costs for both the implementing agency and the beneficiary community. As such, the benefits and value added of a community participation approach have to be clearly demonstrated for each aspect of a programme or activity in which community participation is envisaged. The ARMP preparatory documents made a broad commitment to a community participation approach but they did not specifically outline the value added of this approach or outline how it would be implemented. Specific procedures have to be laid down for community participation, if it is expected that the approach will be followed in a systematic manner. Furthermore a community participation approach requires policy commitment at the highest level, staff training, beneficiary mobilisation, demonstration and proper monitoring and evaluation.

Impact Measurements facilitate Project Management. There has to be an internal demand from the project for regular data on outputs and impact of programme activities, in order for it to be valued. The main need for such data is to demonstrate progress towards objectives, which are difficult to measure, such as improved livelihoods, and to facilitate effective project management. In addition, the project objectives should be expressed in a concise and concrete manner so that project staff can be attuned from the beginning to the requirements of monitoring and assessing their achievements. In ARMP monitored data has given a sufficiently clear picture for the project management to be able to recognise the need to re-direct project activities. However, this re-direction is geared mostly to the project's implementation capacity, and does not benefit from a good knowledge of the impacts achieved, which would provide justification for such re-direction. It is the combination of physical, financial and impact data into a management information system that allows the project management to take informed decisions and guide implementation towards the agreed targets and objectives. The lesson is that impact data needs to be an integral part of the management process, from the beginning of the project.

Reaching Intended Beneficiaries. Farmers in ARMP essentially self-select themselves for participation in the various programmes and there are some aspects of the ARMP selection criteria and methodology which actually exclude the poor. Land holdings are split to take advantage of project SWC activities, and ACC collateral requirements and loan periods result in loans not going to the poorest. The lesson is that the poor and vulnerable households have to be specifically targeted if they are the intended beneficiaries of a programme. The approach to targeting the poor, female headed households and households with a greater reliance on agriculture has to be carefully determined and included in the operating policy and procedures of the project. ARMP is in a strong position to reconcile the twin objectives of poverty alleviation and agriculture growth, but only if the selection criteria are enforced.

The Sequence of Project Processing. The approach described in the project design is clearly participatory. It relies on partnership between the implementing agencies and the beneficiary communities. To support this, provision was made to employ assistance in terms of skills, extra capacity and training to fill the gaps in MOA capabilities. The design does not make it clear that this approach represented a major change for MOA. It would have been useful if one of the objectives had been to introduce such an approach. However, the fact is that a participatory approach has not been adopted. A possible contributing factor is that the implementation version of the appraisal was actually prepared after the loan had been signed. Whilst the formulation should properly conceptualise different and possible solutions to address the project goals, the appraisal has to be rigorously practical, fully describing the implementation methodology. The lesson for IFAD is simply that there can be a high price to pay for shortcuts in project processing.