The Kingdom of Swaziland is a small land-locked country situated in the south-eastern part of the African continent and almost entirely surrounded by the Republic of South Africa (RSA). The economy is strongly dependent on that of RSA. The country is sub tropical, with wide variation in annual rainfall between 760 and 1 140 mm. Most of the precipitation occurs in summer (October-March). Supplementary irrigation is needed either to grow vegetables and perennial tree crops and/or to reduce uncertainties of rainfall. The agricultural sector is dualistic in nature, with the modern plantation export subsector, covering 45% of all rural land and the subsistence or traditional sub-sector occupying 55% . Mean farm size varied in 1987 between 1.6 and 2.5 ha. Most homesteads grow subsistence crops, principally maize, and cotton as cash crops and about 65% own cattle.

Project design and objectives

Target group

An estimated 22 900 homesteads comprising 226 000 people in the Rural Development Areas (RDAs) and about 700 homesteads growing vegetables on the Vuvelane and Magwanyane private irrigation schemes outside the RDAs were to be taken as the target group. From this group, 8 000 homesteads or about 80 000 people were expected to benefit directly from the project.

Objectives and component

Objectives. The project aims at assisting the Government of Swaziland (GOS) in its efforts to achieve food self-sufficiency and to improve farm incomes and standards of living among the rural population, in particular for low income smallholders. The main focus of the project was to be on institutional development, in particular the strengthening of the smallholder marketing and credit support services. The project also included the rehabilitation of twelve irrigation schemes.

Components: (a) Irrigation: rehabilitation of 12 smallholder-group irrigation schemes for the production of rice and vegetables; (b) Marketing: creation and establishment of a National Agricultural Marketing Board (NAMB); provision of market facilities for fruits and vegetables, and maize; and establishment of a Market Intelligence Office under the Ministry of Agriculture and Cooperatives (MOAC) Marketing Advisory Unit; (c) Smallholder Credit: improvement in the lending policy of and provision of funds to augment and sustain Swazi Development and Savings Bank (SDSB) smallholder Agricultural Advisory Credit Scheme; (d) Smallholder Inputs Distribution and Maize Purchases through strengthening the Central Cooperative Unions (CCU): provision of working capital and/or credit facilities to improve CCU's smallholder farm inputs distribution and maize purchase activities; (e) MOAC Tractor Hire Pool: provision of tractors and other facilities and improvement in the organization and management of the operations and services of the unit; (f) Technical Assistance and Staff Training; (g) Special Trial and Studies; (h) Monitoring and Evaluation.

Expected effects and assumptions

Expected Effects. The project intervention would increase significantly yields from year 3 onwards of all winter and summer crops, in particular maize (40%), rice (50%) while vegetables vary between 20% and 50%. Based on the yield projections and on the phasing of farmers participation and land utilization, it is estimated that incremental output would be about 2 780 tons of assorted vegetables and 583 tons of paddy rice. At the point of full development, 491 vegetable producers will be direct beneficiaries of the project. Farmers will benefit from enhanced levels of real and money income. The estimated increase in farm surplus of farmers will vary between 40% and 89% . The provision of marketing and communications facilities will reduce price and market uncertainties, and should enhance the bargaining power of farmers vis-a-vis larger traders, retailers, and institutional buyers.

Assumptions. The expected benefits of the irrigation component depend on timely implementation of the rehabilitation of the physical structures and the effective organisation of the farmers. It was implicitly assumed that the project would receive adequate counterpart funding (borrower's contribution is 25% of project costs) for the execution of these works. The availability and quantity of surface water was not considered a constraint to the development of the irrigation schemes. In marketing, it was assumed that remunerative prices will be obtained by farmers through the rationalization of imports by NAMB. Equally in the case of maize, it was assumed that a ready market exist and that CCU's functions will ensure the provision of inputs for maize and its marketing.

Evaluation

An Interim evaluation mission comprised of four experts on credit and marketing, water management, socio-economics and monitoring and evaluation visited the project for a period of about three weeks. The Mission used information from three sources: (a) national and project authorities (MOAC, CCU, SDSB, NAMB), international organizations (UN agencies), AfDB, bilateral aid agencies and NGOs; (b) project documents; MTE, MTR, supervision mission, progress and monthly performance reports and (c) field observations and undertaken a Rapid Farm-Homestead Survey. The objective of the survey was to develop a preliminary impression of the socio-economic characteristics of project area rural population, their access to project services and benefits.

Implementation context

The SCMP was designed to complement the Rural Development Area Programme (RDAP) then under execution by MOAC. Its unexpected termination in 1985 had led to retrenchment of MOAC staff and construction workers who were needed for the irrigation rehabilitation work under SCMP. Re-employment of MOAC construction workers did not occur until 1989. The RDAP, was also expected to assist Land Use Planning (LUP) financed by USAID, in the design of irrigation rehabilitation. But due to the termination of LUP funding, the Technical Assistance (TA) was terminated prematurely, and the anticipated land development equipment was consequently unavailable. These factors, which were beyond the control of the project, coupled with the slow release of counterpart funds and the lack of an effective full time project management, all contributed to an implementation context leading to delays.

Project implementation gained momentum after the mid-term evaluation in 1988. IFAD intensified efforts to put the project on track and closer supervision by AfDB, the Cooperating Institution was helpful. The eventual appointment of a full-time project coordinator in 1989 was instrumental in improving implementation environment.

Project achievements

The rehabilitation of irrigation schemes had not followed engineering procedures resulting in poor design and low quality construction works. There were diffusion of institutional responsibilities, with many differing lines of authority, in the rehabilitation work. Three separate MOAC departments were responsible for different aspects of the work of planning/design and implementation of rehabilitation work; while assistance to farmers organization was provided by staff from other foreign aid missions. Furthermore, in trying to compress all the rehabilitation works in two years (before loan closing) and execute construction simultaneously in all 12 schemes without adequate supervision, quality has suffered.

The delay in implementation of irrigation rehabilitation has adversely affected the farm-level organization and production. Land use rates estimated by IE and production in these 12 irrigation schemes are very much below appraisal estimates. This has, in turn, affected negatively the full utilization of other facilities and services, including the wholesale market and packing sheds.

Marketing. The impact of the market structures established by NAMB has been limited since smallholders had made little use of these facilities. Farmers have preferred to sell to private traders in the field against cash and not through the wholesale market and commission agents. The establishment of a subsidized wholesale market has primarily benefitted imports of fruits and vegetables, while its merit for smallholder marketing has not been proved.

For maize production in rainfed areas, the objective was to increase yields through improved marketing of inputs and purchase of maize. For this purpose the CCU, has been strengthened through the provision of working capital and funds to build two maize transit stores. The grants provided have facilitated the operations of CCU.

Credit. Financing has been provided to SDSB to strengthen its credit services to smallholders. The project is aiming to disburse loans to 7 000 smallholders. In spite of efforts made regarding improvement of operations by the bank, the number of smallholders who received credit amounted to only 750 in 1990/91. High collateral requirements, high transaction costs on the part of borrowers, the limited staff capacity, and the use of EC funds to serve smallholders are part of the explanation. Repayments rates were about 80% at the beginning of the project but declined to 74% in 1989/90. Most of the loans disbursed were for maize production and only a small number was for the production of vegetables in the irrigation schemes. There has been limited demand for vegetable production credit due to the delay in the rehabilitation of the irrigation schemes and to the fact that farmers use their own funds to finance a production process with a rapid turnover.

Monitoring and evaluation. While the monitoring of the physical and financial progress of the individual project components by the respective institutions has in general been adequate, the activities of the Monitoring and Evaluation Unit (MEU) have proved to be unsatisfactory. The main responsibilities of MEU with regard to the project, were a socio-economic survey of the irrigation schemes and the compilation of quarterly progress reports until 1990, when this was taken over by the Project Coordinator. A delayed draft report of the socio-economic survey was produced in March 1991. The main weakness of the survey design was the lack of a clear concept that could have sharpened the focus of the survey and increased its usefulness.

Effects assessment and sustainability

In the absence of a beneficiary contact monitoring system, there is hardly any quantitative or qualitative information available on the effects of the project on the target group. From field observations, it appears that the impact so far has been limited due to delays in irrigation rehabilitation, low input use, labour constraints, inclement weather, pests and diseases.

Beneficiaries and incomes. To have an idea about the impact of the project on the household income, the Mission attempted as a proxy, to measure the changes (improvements or losses) on crops cultivated. In maize, for example, the survey carried out in June 1991 found that the average maize yield among the sampled farmers was about 3 120 kg/ha. While this result from a small sample should be interpreted with care, the yields are considerably above the 2 000 kg/ha or less normally expected from smallholders. Although the gross margins for maize appear to have reached a satisfactory level, the extent to which it has improved as a result of improved project services cannot be quantified. In case of vegetables, the SAR projected that the cropping intensity would reach or even exceed 200% in the twelve irrigation schemes 5 years after implementation of the SCMP. Field observations supplemented by the sketchy and incomplete data available on vegetable productions on these twelve schemes suggest that this is unlikely to be achieved without major reorganization and further investment in these schemes. Net income from rice after deducting charges for irrigation operation and maintenance, repairs and depreciation of farm machinery and imported costs of household labour would be negligible. Rice production will not be a profitable undertaken, except with heavy subsidies.

Nutrition. The National Nutritional Survey undertaken in 1983, identified stunting as the most chronic nutritional problem in the country. Stunting is found to be closely linked to poverty and subsistence farming. Though SCMP lacked any specific focus on nutrition, its focus is on the poorest and most vulnerable groups in the country. The SCMP aims at reaching these people and is locationally well directed in terms of the most nutritionally-deprived groups. Further support in the provision of training in income-generating skills, for women's groups in particular could indirectly assist household nutrition.

Specific effects on women. Despite the omission of women as an explicit target group the Mission observed that a substantial part of the benefits from the rehabilitation of the schemes appears to be accruing to female farmers and their families. In fact, an estimated 80% of the labour on the irrigation schemes is provided by women on farms operated by themselves and in which they control the decisions. This has emerged as an important secondary benefit of the project with considerable developmental potential for income- generating activities and improving nutritional status.

Environmental Effects. The physical environment in the twelve irrigation scheme areas have not improved, and might actually have deteriorated over the project life. The irrigation rehabilitation did not incorporate any water protection or erosion control works in the catchment area. The river diversion structures were inadequate to prevent also inflows and sedimentation of dams and conveyance systems. None of the silted reservoirs in four schemes have been cleaned. Cattle grazing and felling of trees were continuing in the reservoir catchment areas. Though, increased farm incomes could lead to a higher stock of cattle, overgrazing and soil erosion, there is no evidence that the project has, so far, contributed to this.

Credit. No assessment have so far been carried out on the impact of the credit component on the target group's income due to the absence of an adequate M&E system. It can, however, be assumed that the maize farmers have not benefitted for a number of reasons: (a) the delay in irrigation rehabilitation; (b) vegetable growers appear to prefer financing from their own resources, since the turnover is fast; and (c) the main vegetable growers are women who have no direct access to credit. Furthermore, potential poorer borrowers had no access to credit because they could not meet the collateral requirements, i.e. cattle, or that access costs to credit were considered by them too high.

Participation and group formation. Despite the existing potential, groups in the irrigation schemes did not organise savings clubs and savings and credit unions which could improve rural financial services and reduce transaction costs. The groups did not administer their own irrigation schemes, nor provided for adequate water management, maintenance of irrigation infrastructure and charging of appropriate fees for water use. There is scope for encouraging support from qualified and low cost NGOs. The promotion of small voluntary groups could later contribute to the rehabilitation of the primary cooperatives. The IE found that in none of the project implementation institutions (PCC NAMB, SDSB) are the beneficiaries adequately represented. Although the CCU serves primary cooperatives, it has not held a general assembly in recent years. Project participation and implementation would be strengthened if beneficiaries were adequately represented at the district and national levels.

Sustainability. Sustainability of the irrigation systems is in question. The design of the irrigation structures and the quality of construction are such that they will need further rehabilitation in the near future. Farmers groups have not been adequately formed and trained for sustaining the schemes. Production of rice and vegetables is at present heavily subsidized, without which it is not profitable to the scheme farmers who are mostly women with only limited labour and other resources. Further investments and major reorganization of the production systems are needed for sustainability. By IE the project had insufficiently focussed on the sustainability and cost effectiveness of the support services. The credit programme is not sustainable and requires reorientation focusing more on self-help and joint liability groups administering their own savings and credit programmes. The grants provided to CCU have improved the quality of its operations. But, its financial viability and sustainability would be enhanced by charging import levies on marketing agents.

Main issues and recommendations

To improve the design of irrigation component.

a) Design and/or modify the existing river diversion structures to improve diversion efficiency and eliminate sediment inflow into the conveyance canals.

b) Clean and desilt storage reservoirs, improve system layout, distribution network, and field drainage to enable intensive cultivation.

c) Harness small-scale hydro-electric power for domestic use and processing of farm produce.

d) Establish in the irrigated schemes community-managed woodlands for watershed protection, erosion control, and production of scarce firewood and lumber.

e) Introduce in the irrigation schemes a "garden-style" farming system, incorporating fruit trees and inter-cropping of a variety of vegetable crops.

To enhance Marketing Processes.

a) Strengthen the vegetable marketing extension service.

b) Establish the Fresh Wholesale Market at Nokwane as an autonomous commercial operation which should not be eligible for any further subsidies.

c) NAMB should change payment procedures so that farmers can be paid cash upon delivery.

d) NAMB should earmark adequate funds from levies collected from wholesale agencies as well as importers to the smallholder production and marketing development.

e) CCU accounting system should be reformed to reflect its contribution to the development activities associated with smallholders.

To improve Credit services to smallholders and enhance the viability of credit system.

a) Establish a viable rural financial system for smallholder and improve their credit worthiness through the promotion of joint-liability groups, saving clubs and savings and credit unions.

b) Strengthen the collaboration between extension and loan officers to provide integrated services to smallholders.

c) Reduce the transaction costs and improve loan recovery rates through a performance/incentive system to its loan officers and/or cooperation with credible NGOs.

d) Replace the commission fee received by SDSB for administering the loans on behalf of the project to a system in which the Bank shares the risk of lending, so as to improve loans administration.

e) Adopt procedures that would allow women access to savings mobilization, working capital and credit extension.

To increase the Effectiveness of Monitoring and Evaluation.

a) Introduce a system of beneficiary contact monitoring.

b) Assess the scope for a system of self-evaluation of the irrigation scheme farmers.

c) Review the reporting system.

Lessons learned

Project Management. Complex rural/irrigation development need a highly coordinated management system because of the interdependencies between components. The SCMP lost valuable time at the beginning because in the absence of a full-time project coordinator and without an active PCC, the project lost the sense of direction. If anything had been achieved, it was because of its perceived value to the agency involved and not necessarily to the project or its beneficiaries.

Irrigation rehabilitation. Irrigation development and rehabilitation without assessing a priori the conditions of the watershed and the availability of water resources, and without making provisions in the project for their protection and conservation cannot be expected to be sustainable. Watershed degradation, erosion, and sedimentation are the major factors which impinge on the performance and sustainability of the irrigation systems. Standard engineering procedures for review and quality control of scheme preparation, design, construction and acceptance of completed works should be built into project design and insisted upon before loan disbursement. In no case should loan disbursement be made without the project providing a proper description of the irrigation scheme, the works required, cost estimates, and the design document including technical, economic and financial feasibility studies.

Marketing. Marketing is a process which is very much linked with the import/export, subsidies and pricing policies at the macro-level for all tradables. To achieve project objectives, implementation of project components designed to serve marketing purposes need to be synchronized with the national policy. The construction of a wholesale and packing sheds and provision of subsidies on their services proved counterproductive to the smallholder marketing prospects. The strengthening of the market information service has proved an effective instrument from the beginning of the project since it contributed effectively to improved market transparency and strengthened the bargaining position of smallholders.

Credit. The assumptions made in the SAR on the demand for credit among smallholders and their repayment rate (98 - 518), were unrealistically optimistic. There was insufficient awareness in the SAR of the high transaction costs of smallholder credit. More emphasis should be given to the promotion of voluntary savings and credit groups which have been operating in Swaziland over the last 30 years. Rigorous monitoring and evaluation of these programmes would prove essential for their success.

Participation. A participatory development approach is essential to ensure the involvement of the beneficiaries and their training to plan, design, construct, operate and maintain the project investment. Efficient operation and sustainability of the investment can be enhanced by imparting these skills and a sense of ownership to the beneficiaries. The involvement of beneficiaries in project activities form the design stage to their future maintenance and sustainability is a very specialized and skillful exercise. It needs not only highly trained and motivated staff, but also an accumulation of experience in the technical as well as the social aspects of the project investment. It also requires continuity, monitoring, evaluation, and follow-up. Such attributes are not likely to be readily found in bureaucratic government departments. In Swaziland, like elsewhere in the world, a large community of NGOs have accumulated experiences and could provide the service.