4.1 Donor factors
4.2 Recipient factors
4.3 Technology
4.4 Research
4.5 Monitoring
Should projects be implemented through a special Project Management Unit (PMU) or through existing departments and agencies? There was a shift of opinion between the mid-seventies and mid-eighties, with a clear trend away from PMUs, and the project evaluation studies do suggest that more projects are successful when implemented through existing agencies. This is not to say that PMUs do not work; there are successful examples like the Swaziland Rural Development Areas Programme which started in selected areas but spread until it was a national programme and the PMU became institutionalized as a permanent government agency.
The arguments for a PMU may include:
- The need for the project to bring together a number of departments which are not in the habit of working together.- The project is to be implemented on the basis of watersheds, which do not coincide with administrative boundaries.
- Existing infrastructure is unable to cope with the logistics of procurement and distribution of project inputs.
- There is no local agency able and willing to handle the administration and management of the project.
Possible disadvantages of a PMU are:
- It may weaken line departments by attracting the better staff.- It may exaggerate jealousies between departments, or between departments and the PMU.
- It is less likely to lead to self-reliance.
- There are likely to be problems at the end of the project. Can the local staff employed by the project be absorbed into other employment? What will replace the functions of the PMU?
A number of agencies appear to have taken the decision that they will try to avoid using PMUs if possible (ODA, FAO), but larger programmes, such as those of the World Bank, are more likely to need a PMU because the size of the project is such that it could not be handled by existing agencies. An example is the World Food Programme in Ethiopia.
On the question of the factors affecting the donor side of project management, several issues arise. Quoting Harrison (1987) again: "Personalities are always important. The quality of leadership, from national down to village level, often makes the difference between success and failure. There are many cases where the presence of a charismatic leader, whether project manager or local chief or outstanding minister, has been an important ingredient of success. There are few successes with weak project leaders".
Project management usually feels that it is overburdened with administrative duties, either imposed from donor HQ, or the necessity to conform with local procedures. Several projects report difficulties because the most senior of the technical specialists was made the project leader without having appropriate experience or training in management. Aid agencies might profitably give some thought to ensuring that project managers are suitably trained in management.
A very common problem, reported by all agencies without exception, is time overrun, frustration, and reduced effectiveness caused by delays in getting staff in post. It is felt that there must be something fundamentally wrong with the personnel management of aid agencies for this problem to appear so frequently (more than 50 percent of the evaluated case studies), and it is suggested that this problem might be further studied elsewhere.
The importance of flexibility was stressed during the discussion of project design. This usually comes down to the ability to make changes as a result of field experience, and this implies that authority to make changes must be delegated to the project manager in the field. A common remark by staff of successful projects is "It worked well because they left me alone and let me get on with it".
Some other aspects of relations with local staff and the indigenous people also arise in the evaluation reviews. For example it is important to use the traditional decision-making processes. If local practice is autocratic decisions by the chief, then he must be part of a management committee. If all important decisions are taken by a party political group then they must be involved in project management. But ultimately the final decision to introduce new practices or ideas is made by individuals within the context of their society. There may be orders, suggestions or requests but, except in totally repressive regimes, these are only carried out by the action of individuals. It is not the project's job to change basic social structures; the project should accept and make use of the existing social culture.
This empathy with the local people and their social structure is explained by Harrison as follows: "Those who succeed in Africa also share a common attitude to Africa's people. They have respect for ordinary African peasants - not only theoretical respect for their human dignity, but practical respect for their views and their wishes, and respect for their accumulated wisdom and traditional practices. It is the converse of the Colonial, condescending attitude...".
A project needs inputs and support from the host country at three levels - from the government, from the executing agency, and from the farmers. Looking at these in turn; at government level one looks for both moral and physical support. A very positive verbal declaration of support can work wonders for the morale of all staff on the project, and so can symbolic demonstrations. For example in Kenya the personal participation by President Moi at Conservation Work Days has had a big effect on building political respectability into all conservation activities.
In terms of policy, what one looks for most from governments is consistency and coherence of all government activities. Support which waxes and wanes under different ministers or even different governments is very damaging, and so are inconsistencies between ministries or departments.
Given a political climate within which the project can prosper, the physical backing required from government is the provision of resources, particularly financial and manpower. There is no doubt that recipient governments, anxious for a proposed project to go ahead, are prone to overestimate their ability to provide the inputs. As a result of either over-optimism at project preparation, or subsequent deterioration of the national economy, the inability to provide local staff was observed the 35 percent of the projects evaluated. At the level of local agency or institution, there was a strong tendency for donors to have unrealistic expectations of the support which can be provided by the implementing agency, and there is also a tendency for these agencies themselves to overestimate what they can provide, particulary counterpart staff. Although these problems emerge at the stage of implementation, it is at project planning that they should be measured, and corrected or bypassed.
Irrespective of the quality of the management at higher levels, the effectiveness of a project is very dependent upon the staff at the lowest level of intervention, usually the extension agent or village worker. The poor performance of extension services in most developing countries is now well known, although many projects in the '70s floundered because it was not fully understood at that time. Attempts have been made to tackle this problem through massive interventions by large agencies, particularly World Bank. The initial acclaim for the T and V system was undoubtedly excessive and it is no longer seen as a universal panacea. This is important for soil conservation projects. The method is very effective in disseminating single items of advice such as what to plant or when or how, but attempts to include soil conservation topics into the system have so far been unsuccessful, at least in Zambia, Kenya and several other African countries. A common problem for extension services is that there are usually few suitable recruits available with a combination of a rural farming background and sufficient educational achievement. This can result in a service too dependent on using urban school-leavers not in tune with the farmers.
The desirability of contributions to the project from the farmers also has mixed reaction in evaluation reports. There seems to be general agreement that project activities are likely to be better continued and maintained when the farmers have invested their own time money or labour. But the best way to achieve this produces conflicting suggestions. The use of incentives and subsidies is outside the scope of this report but some opinions are briefly discussed in Chapter 6.
The generally low level of success in agricultural development projects, and soil conservation in particular, is certainly caused to some extent by attempts to use unsuitable technology. In some poor-performance projects this is identified as the only reason, or the main reason, for failure. In others it is one of the causes. It is seldom that the technology is fundamentally wrong, but rather that it was discovered, too late, to be unsuitable for the particular conditions. Cassen (1986) has an interesting observation: "The tendency for agricultural projects to extend unsound recommendations has both technical and economic origins. Agricultural research, especially in Africa, although in drier areas everywhere, has generally failed to generate the flow of relevant recommendations on which a resilient and dynamic agriculture depends. Much of the strength of Indian agriculture has been based on a seed-fertiliser-water package that does not have its counterpart for most other farming systems. Providing this counterpart for other areas, tailored to local circumstances, must be a prerequisite for better project performance, and represents one of the very highest medium-term priorities for aid programmes."
Combining the results of assessments covering several hundred agricultural projects, a list can be drawn up of what a technology should or should not do.
i. The technology must offer a quick payoff. For crops this should be in terms of one year or less. Forestry projects require longer but one recommendation was, "If you want to get trees planted, go for quick-growing species not slow-growing hardwoods". The enthusiastic take-up of growing eucalypts in Kenya is an example of applying this.ii. The technology must offer a high financial rate of return. A possible increase of ten or twenty percent will not stimulate rapid uptake. This requires a return of 50-100 percent or more.
iii. A new technology must involve low risk, preferably reducing the existing risk. This question is referred to again when discussing the testing of packages.
iv. The technology must not require foregone benefits, particularly food. Practices which take land out of production will only be acceptable if there are balancing benefits.
v. The technology must have low inputs; from the farmer, low cash requirement and low inputs of labour; for the government, the technology should avoid recurrent costs, it should have a low import content and maximum use of locally available materials. It should avoid maintenance which requires imported spares or high levels of skill.
vi. A technology may be sound in purely technical terms but still fail because it does not take account of special issues, for example the different roles of men and women in production, or the attitudes towards communal production systems.
vii. The technology should be easy to teach and demonstrate. For widespread adoption there must be the multiplier effect of farmers teaching other farmers. A project which requires large inputs of engineers, or of imported earth-moving machinery, will inevitably be restricted to small project areas. This argument applies also to local inputs. A technology which requires a saturation level of extension agents will be restricted to a short period of time, or a limited area.
viii. A technology which is the development or improvement of an existing practice will be accepted more readily than something which is completely new. Awareness of indigenous conservation farming practices has improved dramatically in recent years as a result of many studies.
The importance of testing technology
It has already been stressed that technology can only be successfully implemented if it is suitable for the particular conditions, and this requires that it must be tested before being built into a project. The World Bank 1986 review showed that the use of untested packages of technology led to poor results in 61 percent of projects, whereas tested packages had an 80 percent success rate. This is confirmed by several other reports, thus Cassen (1986): "Where projects have run into trouble, it is often because they have attempted to persuade cultivators to abandon existing practices in favour of new ones without sufficient testing of the new ones".
Harrison said much the same thing: "As a general rule, no innovation or practice should be offered that would leave farmers worse off in a bad year than they would be with their traditional techniques and varieties".
The same thought is expressed in one of the basic principles of the Indian Council for Agricultural Research that "Only practices based on good evidence may be tested on farmers' fields, which must be protected from speculative experimentation".
This consensus can be summarized by saying, in investment terminology, that there must be no downside risk, and that possible weaknesses must be cleared out of the way before the project starts. There is a clear case for much more use of pilot studies but this is not popular with agencies looking for tidy packages ready for implementation and quick disbursement. It is not necessary in this report to quote the many projects which attempted to use technologies imported from other countries or conditions. Two of the more obvious examples are the inappropriate attempts to use the classical North American system of graded channel terraces in Francophone West Africa, and the equally unsuccessful use of bench terracing on unstable steep slopes in the humid tropics.
Deficiencies in the recommended technology should not be discovered during the project. The proposed recommended practices should be critically reviewed during the project appraisal, and plans made to resolve any uncertainties by field trials, adaptive research, or a pilot project, as appropriate. A useful list of questions to be asked at appraisal is given by Douglas and Lai (1988, p 52), and may be summarized as follows:
- Is it technically possible?
- Is it practically feasible?
- Is it productive?
- Is it financially desirable?
- Is it stable?
- Is it sustainable?
- Is it universally applicable?
- Is it socially and economically acceptable?
Single steps or packages?
There is not a consensus view on the relative merits of simple technologies versus packages. One view prefers simple technologies applied one step at a time, consolidating each step before taking the next. Certainly simple technologies are easier to disseminate and spread. But another view is that the large benefit which the technology must offer is more likely to derive from a package where several factors interact favourably. This is reflected in the common system of recommending bunding or terracing combined with the introduction of improved varieties and improved fertilization. One difficulty is to separate the benefits accruing from the components within the package. There is probably no universal answer to this question because the local circumstances and conditions vary so much, but it strongly reinforces the earlier recommendation that all uncertainties should be resolved by local testing and pilot studies.
On the other hand, if the project is sufficiently flexible, technical packages and improved techniques can evolve during the project to suit specific local contexts. An example of this being applied successfully is the Oxfam project in Burkina Faso. The project started with the idea of encouraging forestry on communal lands using micro-catchments to increase moisture availability, but after listening to farmers' suggestions and adapting the programme it eventually focused on water harvesting for food crops on private lands, and is one of the most successful conservation projects in Africa.
In many developing countries the poor technology, or the absence of tested technology, is the result of weakness of the national agricultural research programme. The World Bank has found that only, "Perhaps 10 percent of developing countries already have adequate skills, good national research programmes and effective linkages with international research institutions".
Cassen (1986) suggests that: "The weakness results from an interaction of several factors, only some of which are directly addressable by donors:
i. low government commitment;ii. low status of research within government and lack of credibility with farmers;
iii. inadequate budgetary provision;
iv. rapid staff turnover, both of national staff and expatriates;
v. lack of information and analysis for setting consistent and appropriate research priorities;
vi. unsuitable institutional structures and poor coordination;
vii. the short donor project cycle which tends to lead to short planning horizons, inappropriate phasing, and discontinuities, when what is required is sustained and logical programme development".
Another commentary on agricultural research in developing countries gives a list of research problems:
a. farmer participation in design and trials is essential but almost invariably inadequate;b. researchers are under pressure by funding agencies to produce highly visible innovations in a short period of time. As a consequence the emphasis is on producing hardware at the expense of improvements to the software of research methods, institutional organization, management, and communication procedures;
c. the short-term (often donor-funded) project approach to research and development means that virtually all published reports contain only early assessments of the technology's potential. These tend to report mainly on-station trials, thereby avoiding problems when the technology is introduced to farmers, and giving an over-optimistic view of its potential;
d. pressure to produce hardware results in inadequate searches for relevant experience elsewhere, and so the wheel is reinvented;
e. there is a reluctance to publish reports of the failure of experiments. Publication of experimental failure can prevent the exploration of blind alleys from being repeated (ODI 1988).
The deficiencies of research are particularly noticeable in soil conservation projects. There are several cases in Africa where large programmes of terracing have been carried out with the help of FFW or donor subsidy, but there is no tested package of improved agronomic practices to take advantage of the terracing. The concept of 'conservation-based farming' is widely accepted, but putting it into practice is held up by the failure of agricultural research to produce agronomic innovations to go with programmes of physical conservation works.
The general weakness of national research in developing countries is now recognized. World Bank has several programmes; as early as 1982 USAID made a major commitment to "develop human resources and institutional capabilities"; the All Party Parliamentary Group on Overseas Development (1985) recommended that British aid should be increasingly directed towards supporting national agricultural research programmes.
There is a clear consensus that good monitoring throughout the project is essential, and also that it is frequently inadequate. It has been shown that blueprint projects which are finalized at preparation are less likely to be successful than flexible projects which can adjust to experience gained as the project develops. This implies that there must be a regular and reliable programme of measuring, recording and reporting. This in turn means that there must be close contact with the beneficiaries, and also defined indicators of performance.
Evaluation studies show that it is very common for insufficient attention to be given at project preparation to working out plans for monitoring, and the lack of a clearly laid out plan encourages project staff to give it a low priority. It is the impression that field staff think of all returns and reports as being unwelcome chores that interrupt the real work.
The present position is that it has become customary to pay lip service to the importance of monitoring, but there is room for making it happen more effectively.
The information to be gathered and reported will vary from project to project. Questions which every project should be regularly asking are: is progress satisfactory? If not what are the difficulties which need to be addressed? What new ideas are emerging? And, which ideas may suggest changes to the project? If new technology is being introduced, the farmer reaction and uptake needs to be closely monitored in case the technology needs modification or can be improved.
