PROJECT BACKGROUND
In 1985 and 1986 FAO undertook the important study "African Agriculture: the next 25 years". During the study soil conservation projects in Africa were examined to see why some have been more successful than others. The results were interesting so a more detailed study was undertaken. The findings of this detailed study are given in this publication.
The first plan was to select a few representative projects and study them in detail. However it soon became clear that projects vary so much that a small sample could not be representative. The plan was therefore changed to a less detailed study of a much larger sample. It also became apparent that there is a large amount of literature on assessments which has been carried out by other agencies, and although none of these were specifically addressed to soil conservation projects, their data and conclusions would still be very relevant to this project.
METHOD
The plan of work was to study all the available books, reports, assessment studies and project evaluations, and from this database to form some tentative conclusions. At the same time a fairly detailed evaluation was made of a sample of 40 projects which had a significant soil conservation component. A four-page assessment report was completed for each project, divided into sections relating to:
- planning and preparation before the project starts;
- implementation during the project; and
- assessment and continuity after completion of the project.
The projects covered all types of agency: multilateral, bilateral, and NGO, and geographically were spread worldwide. The simple six-point evaluation method used in the FAO Review of Field Programmes was also applied to all assessed projects.
CONCLUSIONS - BEFORE PROJECT IMPLEMENTATION
By far the most significant conclusion from the study is the importance of project design. Every major assessment reaches the same conclusion, for example the World Bank study of almost 1000 projects over a ten-year period found that of projects which had poor performance, 86 percent had design faults, that is the problems were designed into the projects. In one third of these cases poor design was the most important single factor leading to poor performance.
Design errors are mainly the result of incorrect assumptions made at the design stage.
The main donor errors are:
- over-optimism, including over-estimating the effect of new practices;
- over-estimating the rate of adoption of new practices;
- over-estimation of the ability of the host country to provide backup facilities;
- under-estimation of the time required to mobilize staff and materials for the project; and
- frequently a quite unreal estimate of the economic benefits.
Some of the main problems arising from the assumptions of the host governments are:
- over-estimating their capacity to provide counterpart staff and the funds for the recurrent costs arising from the project;- a tendency to under-estimate the problems of coordination among different ministries or departments;
- a tendency to over-estimate the strength of the national research base and its ability to contribute to the project.
Other design problems on the recipient side may include the absence of a long-term policy for the development of natural resources and the political stability necessary to ensure continuity of political support. There was sometimes a reluctance to make full use of low-level institutions (e.g. farmers' associations, cooperatives, village councils), which are usually very important for the successful implementation of agricultural projects. Finally, some technical problems are not adequately considered or resolved at the planning stage, such as the pricing policy for agricultural products, the effect on the project of variations in exchange rates, and the existing social background, e.g. land tenure.
A checklist of good project design requirements
1. A project must have clear objectives and definable measurable targets which must be limited and obtainable. It must be clear whether the project is primarily directed towards achieving national objectives or those of the individual farmer.
2. Project design must be flexible so that the project can grow and change and adapt during the course of the project. The old style of a detailed blue-print type of plan may be appropriate for engineering construction projects but is highly unsuitable for agricultural development.
3. Project design must include details of implementation. The duties and responsibilities of project staff and counterparts must be clearly defined, and the details of operating, down to who pays for the petrol for the project vehicles. An annual workplan is essential.
4. Changes in agricultural systems take place slowly, and projects must be designed with a realistic duration. In the sample of soil conservation projects, of the successful projects 75 percent had a duration of five years or more, and only 25 percent less than five years.
5. Many projects have failed because of the technology.
The technology should:
- be appropriate and tested locally;- offer short-term, on-site benefits, and large increments (50-100 percent);
- require affordable inputs, especially labour;
- not include foregone benefits, e.g. giving up land;
- not include any increased risk;
- be in tune with existing social factors, e.g. the separate roles of men and women in agriculture.
6. Where it is clear at the project design stage that there is a weakness in the capacity of local institutions, there should be heavy front-end loading of the training component. A three-year project which includes overseas post-graduate training means that the key counterparts are unavailable for most of the project. In soil conservation projects there is a trend away from overseas university training and towards more management training, widening of experience, and exchange visits with neighbouring countries.
CONCLUSIONS - DURING IMPLEMENTATION
Projects or programmes which set up a separate Project Management Unit (PMU) are in rapid decline, partly because these were frequently associated with multi-sector or multi-component projects, like the Integrated Rural Development Projects which are no longer in favour. The trend is towards projects which are based within existing line departments.
A high proportion of projects report problems arising from delay in staff appointments, both international experts and local counterpart staff. This occurs in many donor agencies from the largest to the smallest. Since appointing procedures are unlikely to change, the important thing is that the project design should be based on realistic assessments of the necessary lead time.
In project design the importance of flexibility was noted. This implies that there must be delegation of responsibility so that necessary changes and adjustments can be made on the spot by field staff without having to refer back to headquarters. When former project managers were asked why a project was successful, a common reply was, "Because head office left me alone to get on with it".
Most agricultural projects involve the introduction of new concepts or methods or materials but these have seldom been fully tested under farm conditions, and so a continuing backup from the national research service should be available.
Another common weakness is monitoring during the project. All the agencies which have conducted detailed evaluations of their operational procedures recommend that much more emphasis should be placed on regular monitoring.
CONCLUSIONS - AFTER THE PROJECT
Along with the need for better monitoring is a widely expressed need for more and better evaluation. It is only by the careful evaluation and study of past projects that the lessons and experience can be applied in the design of future projects. The increasing awareness of the low level of success of agricultural development projects has led to much greater interest in evaluation as a step towards improvement.
Current development jargon is so full of the buzz words sustainability, continuity, and replicability, that it is surprising to find how few of the projects designed more than five years ago gave any serious thought to the question of what would happen at the end of the project. Certain types of project have a specific objective to be achieved within a fixed time span (e.g. many FAO TCP projects) but in all cases where the objective is to introduce or encourage change, the question of the long-term future should be built into the project design.
SOME OBSERVED TRENDS
Projects on the whole are getting better. The FAO Evaluation Centre reports that in the period 1980-84, 50 percent of projects were rated good or satisfactory. In the period 1985-86 this had increased to 73 percent.
In the sample of soil conservation projects where, of those started in the 1970s, 25 percent were rated successful, but of the projects started in the eighties the proportion of successful projects has risen to 56 percent (Appendix 1).
Evaluations by the World Bank give a less encouraging result. They report that failure rates were 12 percent in 1979, 37 percent in 1984 and 40 percent in 1986, but not too much importance should be attached to these figures because the Bank assessment is based only on whether the project achieved its expected economic rate of return.
Projects are getting smaller and simpler. There is little interest today in Integrated Rural Development Projects which were thought to hold out so much promise in the seventies (Sections 6.1 to 6.5). There is also a noticeable decrease in multiple-donor and multiple-sector projects.
Projects are getting longer. The record clearly shows the need for a steady sustained pressure on agricultural problems rather than a short hammer blow. The three-year project is relatively uncommon now, apart from the narrow, specific-objective project mentioned earlier. The problem inhibiting the more widespread use of long projects results from the budgeting and accounting procedures of most donors. Where a country has an established long-term development programme, a donor can fund a time-slice of the programme.
The study of soil conservation projects does not give a clear picture of the usefulness of incentives, particularly subsidies and Food For Work. The most common view seems to be that all such incentives should be avoided if possible, relying only on the farmer's own perceived wish to carry out conservation practices. However many projects report that terracing programmes could not have been achieved without the incentive of FFW.
