CHAPTER 2. NATIONAL AGRICULTURAL RESEARCH SYSTEM (NARS)

2.1 Agriculture sector

The seven countries selected in sub-Saharan Africa were Cameroon, Ghana, Kenya, Madagascar, Malawi, Mali and Senegal. Their populations range from 9 million in Senegal to 29 million in Kenya. All countries have a high proportion of people living in rural areas. The growth of cities exceeds population growth rate in rural areas in all countries. Cities are growing especially fast in Ghana, Kenya and Malawi.

Gross National Product is over three times as high in Kenya (US$10.6 billion) as in Malawi and Senegal (US$2.6 billion). Kenya has, however, a much lower GNP growth rate than the other countries. Agriculture contributes heavily towards Gross Domestic Product (GDP). Mali with 46 percent is on top and Senegal at the low end with 19 percent. The average for the sample countries is 34 percent. Agriculture naturally represents the main source of employment for rural inhabitants with figures ranging from 86 percent in Malawi to 60 percent in Ghana and Senegal. An economic and social overview of the countries is provided in the tables in Annex 1.

Overseas development assistance (ODA) contributes heavily to the agriculture sector. ODA is falling, but not as dramatically as in the mid-1980s and early 1990s. The countries received on average US$78.6 million per year in the period 1996-1991 (OECD figures). Ghana, Senegal and Kenya received most with Mali, Malawi, Cameroon and Madagascar receiving less. Assistance to the agriculture sector represented roughly 16 percent of the total ODA to these countries. In francophone countries, France remained the major donor with the United Kingdom, Canada and the World Bank having a preference for the anglophone countries. The European Union and individual European countries also provide considerable financial assistance to the agriculture sector. Data for national support to the agriculture sector is unfortunately scant and not available for all countries. Based on OECD data national financial support show the following picture:

Table 3. Government expenditures in agriculture as a percentage of total expenditures in 1990

The Governments of Malawi and Madagascar spent the most on agriculture and the Government of Mali spent the least. The average government expenditure in 1990 for agriculture for sub-Saharan Africa was 6.3 percent.

All countries have been subjected to structural adjustment programmes. For the agriculture sector this has meant introduction of policies to liberalize provision and marketing of inputs and outputs, limit state intervention and promote private investments. Emphasis has also been on increasing the productivity of land and labour, export diversification and import substitution to improve balance of payments as well as improving natural resource management for sustainable development. The structure adjustment programme strategies have been geared at:

(i) creating a free market system as a means of improving producer incentives;

(ii) withdrawal of government services from activities better performed by others, whether small farms (e.g. seed production) or commercial companies (e.g. input supply); and

(iii) concentrating government financial and management resources on essential government services (research and extension) and the rehabilitation and maintenance of rural roads and irrigation networks.

Governments in the sample countries, as well as in sub-Saharan Africa overall, have realized that agricultural research is crucial for the achievement of important development goals. Subsequently agricultural research policies have been developed to match these development aspirations.

2.2 Research policies for the agriculture sector

There are major similarities in agricultural research policies among the selected countries. An outline of the policies is given in Annex 6. Key common traits are:

• the agricultural development policy is the basis for agricultural research policy;

• agricultural research should be responsive to the needs of end-users and stakeholders who should be involved in the decision-making process, in setting priorities and in monitoring and evaluation;

• agricultural research should be holistic, balanced in terms of subsector coverage and regional and agro-ecological zones;

• the network of facilities and its decision-making process should be decentralized towards regional and local levels;

• agricultural research should contribute to food security and alleviation of poverty and increase revenues, through diversification of production and more generally improve the living standards of rural people;

• agricultural research should protect the environment through a rational use of natural resources;

• funding of agricultural research should be consistent with its role in the national economic development, and foremost, it should be sustainable and timely disbursed.

The overall impression is that agricultural research should be more accountable and market oriented. These views might be welcomed by research managers and donor partners. The question is whether these policies will have an effect and whether governments will be able to make tough decisions on reform of management at national, regional and local levels. The policies entail that NARS need a reassessment in terms of their mandate, approach, size and composition, resources and other vital elements. The rest of this chapter and the next chapters deal with this issue.

2.3 Evolution of agricultural research systems

The evolution of agricultural research in the selected countries is intimately linked to the overall History of Agricultural Research in sub-Saharan Africa (FAO/SPAAR forthcoming). An historic overview of NARS in the seven countries is given in Annex 1.

2.3.1 Common features of agricultural systems during the colonial period before 1960

Agricultural research in the colonial period was completely dominated by the major colonial powers (France, the United Kingdom and to a lesser extent Germany) that ruled the countries in question. France in Madagascar, Mali and Senegal, Germany in Cameroon and the United Kingdom in Ghana, Kenya and Malawi initiated agricultural research in the late 19th and early 20th century.

The most prominent feature of this early period was the creation of botanical gardens. In Cameroon Germans established botanical gardens in three locations (Edea, Akonolinga and Victoria). The British created the Government Botanical Gardens in Aburi in Ghana in 1890. The Aburi Gardens were formally linked to the Kew Gardens in England, whose staff largely directed the work in Aburi. Research at that time focussed mainly on screening exotic material, such as oil palm, cocoa and rubber, for economic uses in the colony.

After the First World War colonial nations needed more exotic raw materials for their fledgling industries. A push was exerted towards creating more formalized research structures throughout the French and British colonies. Both France and the United Kingdom adopted a similar approach. Regional research entities were set up to cater for several local territories. Research facilities located and run at the federal level were deemed better and more cost-effective than having large national research institutions.

In the British colonies basic research was conducted in commodity or discipline-based regional centres linked to an extensive international network. Research facilities were more or less evenly distributed. Each territory being headquarters to at least one federal institution for which it was thought to have a comparative advantage. For the French colonies the approach was slightly different. As the headquarters of the Federation was established in Dakar, Senegal, Bambey served as the hub of the research network for Sudano-Sahelian West Africa under the responsibility of Federal Government. Senegal, therefore, came to have a lion share in running and administering agricultural research in West Africa. Mali had little infrastructure and a weaker network. Cameroon and Madagascar had for historical and geographical reasons been administered and organized also a bit differently.

In Ghana, Kenya and Malawi, a network of adaptive research stations administered locally were in place and connected to the regional centres prior to independence. This constituted the embryo of NARS. In the French colonies locally administered stations were non-existent or limited, except for Cameroon and Madagascar. However, soon after the Second World War, France created a network of individual export commodities research institutes, with their headquarters and direct management from Paris. Activities were spread all over the colonial territories. They constituted along with ORSTOM (Overseas Scientific and Technical Research Institute), for basic research the backbone of the French colonial research system for tropical agriculture. The dual system in the organization of research was a characteristic feature in the French colonies. There was no common administration of research activities at the territory level and the activities were already fragmented. Therefore, the embryonic NARS that existed in the British colonies did not exist in their French counterparts.

2.3.2 Early post-independence (1960-1975)

From 1960 to 1970 almost all sub-Saharan African countries, in particular those in the sample of the study, gained independence. The responsibility for agricultural research was transferred to each country. The evolution of the system was formed by political decisions made by the new national governments. It might be useful to reiterate the common definition of National Agricultural Research System (NARS) used throughout this study:

NARS are defined, in a given country, as encompassing all institutions public or private devoting full time or partially their activities to agricultural research and committed to a national research agenda. Generally, the following categories of such institutions are identified as follows:

(i) institutions whose mandate is to carry out research only, such as the NARI (National Agricultural Research Institute);

(ii) higher education institutions devoting their activities to teaching and research: they are the faculties of agriculture and related disciplines and the faculties of social sciences and economics of the universities;

(iii) technical departments of some ministries, development agencies that carry out some adaptive research programmes; and

(iv) NGOs and the private sector.

The international agricultural research centres (IARCs) of the CGIAR are not part of NARS, (as often found in literature). For obvious reason they are never committed to a national research agenda. Their focus is on a regional/international level. However, spill-over of their research is important for NARS as they represent partners to other foreign research institutions.

There were marked differences between the French- and English-speaking countries after independence. With the exception of Mali, the former, in the sample (Cameroon, Madagascar and Senegal) waited a decade and more before setting up their own national institutes. In the aftermath of independence bilateral agreements with France, whereby French tropical research institutes created just after World War 2, were given full responsibility for the management and execution of the research programmes. Co-financing between France and each country was almost on a par. However, apex policy formulation bodies came into being before the national institutions.

The English-speaking countries took immediate responsibility of the research structures in their territories. The colonial power under fledgling federal institutions tried to maintain some inter-country organizations, but they collapsed only a few years after independence. The most radical change was in Ghana where reorganization of the research infrastructure started and included the inheriting of locally and federally administered institutes.

2.4 Current national agricultural research systems (NARS)

The early 1970s was a turning point for the reorganization of NARS in sub-Saharan Africa, particularly for the French-speaking countries. A decade and a half after independence, they have had time to devise some research policy and reorganize their system. At present, NARS in each country are dominated by a national agricultural research institute/organization (NARI/NARO) that has quasi-monopoly of research activities and mobilizes 75-80 percent of all the resources, financial and human alike. In the following section dealing with the organizational structure and management, the analysis is limited to the NARI/NARO, backbone of NARS.

2.4.1 Organizational structure and management of NARS

In terms of organizational mode, all the countries except Malawi, have adopted the NARI mode with semi-autonomous management status. Malawi has kept the departmental mode inherited from the colonial era. Ghana has adopted a mixed organization combining the council for coordination and planning and the NARI for routine management and implementation under the direct supervision of a board for each individual NARI. For Cameroon, Senegal, Mali and to a lesser extent Madagascar, Kenya and Malawi, the dominant NARI has responsibility for research in all sectors.

Tables 4, 5 and 6 provide a summary of the policy guidance and coordination, organizational structure and management of the dominant NARI/NARO of NARS in each of the selected countries.

Table 4. Policy model chosen

Table 5. Organizational model

Table 6. Networks and agroecological coverage

NARS are all decentralized and have good coverage of all the agro-ecological zones with a network of national and regional research centres and stations and perennial experimental sites. The policy formulation and planning, priority setting and programme management processes differ from country to country and are summarized in the following sections.

2.4.2 Policy formulation, planning and priority setting processes

Policy formulation in Cameroon is vested with the Ministry of Scientific and Technological Research (MINREST). It supervises IRAD management and research activities and provides a legal and operational framework for its cooperation with universities and institutions abroad. Priority setting and long-term planning, have never been a pervasive process in Cameroon or in the individual institutions. However, the latest reorganization was made after a comprehensive strategic planning process. The capacity and the methodology are in place for future exercises provided the authorities decide to institutionalize the process.

In Ghana CSIR has the formal responsibility for advising government on science and technology policy. However, the mandate does not specifically include the formulation of research policy. Recognizing the need for such policy formulation, CSIR established the Planning and Analysis Group (PAG). The Technical Committee for Agriculture, Forestry and Fisheries (TCAFF), was established in addition, to advice on science and technology policy for the agriculture sector. However, an initial report prepared by the TCAFF does not concentrate on research policy as such, whilst the PAG has been unable to address the issue, due to limited resources. Consequently, CSIR does not at present, provide a lead in formulating agricultural research policies with clear linkages with national development objectives. The Agricultural Research and Development Advisory Committee within the MOFA, has terms of reference that include the formulation of research policy. It is, however, not certain if this is operational. None of the committees or units described above is responsible for the determination of research priorities, or has the authority to allocate resources to particular research proposals. The lack of a functioning mechanism to set these priorities, with the authority to allocate financial resources, is reflected by the dispersed nature of much of the research effort. The situation is clearly unsatisfactory and forms a major constraint to effective operation of the agricultural research system.

Kenya on the other hand, has a rather well established policy formulation process. The Amended Science and Technology Act (1979) provided for the establishment of the Agricultural Advisory Research Committee (ASARC). Its main function was to advise the minister responsible for agriculture, on scientific research and on a wide range of issues pertaining to agriculture, agricultural education and coordination and management of research. Thus, ensuring free flow of knowledge. This included the mechanism for establishing priorities for allocating research resources to research programmes within institutes. Unfortunately, although ASARC was constituted in 1980, it only operated in 1980 and 1981 and was disbanded thereafter. The National Council for Science and Technology, which was expected to be the apex organization for policy, has been starved of resources and moved around several ministries. As a result it has lost the pivotal role of overseeing the national research system. The Department of Research Development of the Ministry of Research and Technology, besides having its officers attend board meetings of the various research institutes, has no responsibility for directing national research. However, for the publicly funded components of NARS, the boards are responsible for policy-making and ensuring that they remain the apex organizations responsible for Kenya's agricultural research through development and application of science and technology.

Prior to the restructuring and reorganization of NARS, which took place in the early 1980s, there was not an appropriate and affective formal system for allocation of all research resources. Insufficient delegation of responsibility for design and the implementation of research programmes for key commodities, factors and disciplines exacerbated the situation. As indicated above the ASARC was responsible for policy guidance and determination of priorities and resource allocation. Unfortunately, it did not work and it was disbanded and the NCST could not fill the gap. The reorganized KARI has put into place a new mechanism for setting priority and allocating resources. Overall strategy and priorities have been worked out with the assistance of ISNAR and the donor community and the various stakeholders, this exercise yielded the National Agricultural Research Programme/Project (NARP I and II).

Regional research to adapt technology in different agro-ecological areas is carried out by the network of RRCs. Each research centre has a centre research advisory committee (CRAC) and a centre technical committee (CTC), which ensures that research proposals are consistent with the national/regional priority structure, respond to stakeholder needs and are technically sound. The membership of the regional CRAC comprise the RRC director as chairperson and representatives of farmer organizations, NGOs, agro-industries and exporters, extension workers and staff of the RRC. For the NRCs, the relevant assistant director chairs the CRACs; they have similar members as those of the RRC. Directors of relevant NRCs and RRCs are invited to participate. Proposals approved by these committees are submitted to the research coordinating committee (RCC) at KARI headquarters for consideration; KARI prepares a consolidated annual work programme including procurement plans.

Madagascar's Ministry of Scientific Research is in charge of the implementation of the National Research Policy, including the agricultural research policy. The Government recognizes that science and technology are essential elements for economic and social development; therefore, it is indispensable to define a national policy within the overall national development goals. Agricultural research master planning is commonly used in particular by FOFIFA. In the Ministry's structure, The Directorate of Planning and Coordination is responsible for the supervision of the agricultural research institutions such as FOFIFA.

Each research institution has its own bodies responsible for determination of priority and resource allocation. For FOFIFA they are:

(i) the Scientific and Orientation Council (CSO) that determine priority programmes for FOFIFA;

(ii) the Agricultural Research Funding Committee (COFIRA) determines the overall level of funding and the allocation for each research programme; and

(iii) at policy level, the board of management validates the decisions of the previous lower level organs.

The programme formulation and budgeting follows a bottom-up approach from the stations/centres through to regional centres and up to national level where the programme committee deliberate on the proposals from the lower levels. This is done under the chairpersonship of the director of research. The proposals reviewed are finalized and submitted successively to the CSO, the COFIRA and the board of management for final approval.

Malawi has an elaborate policy formulation process. The Agricultural Research Council (ARC) was established to set policy guidelines within which DAR operates; the functions of the ARC are to:

• ensure that the agricultural research strategy is consistent with the national development goals;

• outline an agricultural research policy that is compatible with and supportive of national goals;

• approve annual research programmes and projects for implementation by DAR;

• consider and recommend contract research proposals, including the level of funding;

• recommend the appropriate level of expenditure in agricultural research, with the target of eventually investing at least one percent of the annual value of the agricultural gross domestic product;

• prepare the agricultural research master plan, which allocates priorities according to national goals.

The ARC is made up of a cross-section of experienced persons in the public and private sector who have varying perspectives on the nation's economic and fiscal policies and agricultural development goals.

ARC is organized with a secretariat and subcommittees. For the ARC to conduct its functions efficiently, an ARC secretariat in DAR's headquarters was established. This secretariat is responsible for preparing the master plan in collaboration with all researchers, and organizing triennial reviews of the research system. It has two subcommittees: a technical subcommittee that examines and reviews research programmes and proposals including contract research and a financial subcommittee that examines the annual budget and scrutinizes the estimates of programme and project costs. The council at its general meetings considers recommendations of the subcommittees. The various NARS institutions are coordinated nationally by the Ministry of Research and Environmental Affairs (MOREA).

The task of determining priorities, programme formulation and resources allocation is assigned to the ARC, where the Agricultural Development Division, sets the overall research priorities. The ARC employs a set of guidelines for priority setting and draws heavily on the desired national goals embodied in the 1987-1996 Statement of Development Policies. The criteria used for priority setting include national goals and research and technical factors.

Within commodities, the task is assigned to the national research coordinators (NRCs) and commodity team leaders (CTLs) who ensure that priorities set within research commodities are in line with nationally set priorities, taking into account the human, physical and financial resources available to the commodity.

Prior to the 1992/1993 fiscal year the allocation of DAR funds under the revenue account was primarily done on the basis of research station needs. The allocation was not based on systematic planning of research programme priorities. After the reorganization of DAR the team approach to research programmes with commodity groups and team approach focus of planning and budgeting rather than the stations, provides a more rational and defensible basis for programming and budgeting. Under this system the budgeting process is coordinated and directed by the financial controller and the research economist. The CTLs are responsible for preparing work plans and preliminary budget requests for their respective commodity teams. These are assembled into work plans and budget proposals submitted to the respective NRCs who collate them into unified budget proposals for consideration by the Deputy Chief Agricultural Research Officer.

In Mali, the process evolved as in other countries in terms of policy formulation. Following the various reviews undergone by NARS and particularly during the strategic planning carried out by ISNAR, the mandate of the CNRA was changed in 1990. The new CNRA has the mandate to formulate agricultural research policy and strategy. It decides on research priorities, reviews and approves research programmes and budgets, reviews the research results and decides on their diffusion, monitors the use of resources and establishes linkages between research and external research organizations as well as within the country, with development agencies.

The membership of the NARC is as follows: nine voting members including the chairperson who is a renowned scientist. The sector directors of the concerned ministries and the director-general of IER participate in the meetings without voting rights. One mandatory meeting is held in November each year.

The NARC has three specific committees, namely,

(i) the Scientific Committee (SC);

(ii) the Financial Resources Committee (FRC); and

(iii) the Users Committee (UC). Furthermore, the NARC has a Permanent Executive Secretariat (PSE). It is composed of a research manager (the permanent secretary), a finance officer and two support staff. The secretariat is responsible for day-to-day implementation of policy decisions.

Planning, programming and resources allocation are made at three levels in Mali with the sharing of responsibilities:

• strategic planning: NARC is responsible for the preparation, regular review and updating of the National Strategic Plan for Agricultural Research. External programme evaluations, every three years, are part of the inputs for the strategic planning exercise. The strategic plan provides broad guidelines and only identifies research themes, their priorities for the long-term and the resources required;

• programme formulation: the scientists, in liaison with extension agents and the users, develop the identified priority themes of the strategic plan into fully-fledged research proposals. NARC reviews and approves proposals with their budgets assuming that the resources necessary for their implementation are committed for the duration of the research projects/programmes, unless the external review of the programme calls for adjustments;

• annual review, programming and budgeting: although activities and resources are agreed upon during project/programme approval, annual adjustments are needed in light of the previous year's results; a detailed annual budget also needs to be prepared.

For Senegal the research policy formulation body is the Interministerial Council for Scientific and Technological Research. The council meets every year under the chairpersonship of the Head of State. It deliberates on research priorities identified by standing sectoral committees. Moreover the quinquennial Economic and Social Development Plan that defines the overall development policy provides also for guidance on agricultural research policy and priorities. Since the cycle of sectoral adjustment programmes the country has gone through, the agriculture sector has issued several policy statements in which agricultural research policy and priorities are spelt out. However, these bodies are designed to provide a general framework only. Detailed policy is formulated by each institution under this framework under the supervision of its governing bodies. Furthermore, ISRA has had experience in strategic planning as its first strategic plan was prepared in 1979, the second in 1989 and the third in 1995, respectively for the periods of 1980-1985, 1990-1995 and 1998-2003. Over time the methodology has been fine-tuned with better involvement of all the stakeholders in particular the farmer organizations. The board and the STC played an important role in the process. The strategic orientations of ISRA are as follows:

• demand-driven oriented research underpinned by excellence and responsiveness to stakeholders needs;

• highly qualified, productive, innovative and motivated human resources;

• insertion of ISRA in NARS as its backbone in bettering linkages and developing partnerships based on comparative advantages.

Priorities are determined within the general framework of the government development policy and developed bottom-up in an iterative way. The programme formulation and resource allocation is carried out annually based on the action plan. The process involves all ISRA partners (technical services, NGOs, farmer organizations). The programme committees ensure at all levels (local, regional and national) that all these stakeholders participate in this dialogue. The scientific directorate serves as a go-between in the process, consolidates the proposals and makes the necessary adjustments before submission to the STC. This latter examines the pertinence of the proposed activities with regard to national priorities and the adequacy with the financial resources allocated to the research programmes. The STC reports and makes recommendations to the board that takes the final decisions on the programmes and levels of resources allocated. The board's report is submitted to the two ministries in charge of ISRA for final approval.

It can be concluded from the above-mentioned description that each country has some sort of overall apex body for coordination, setting policy at the national level for science and technology with generally sectoral committees. In the sample of countries it can be assessed that these bodies do not work properly either for lack of resources or capacity or both. The bodies attached specifically to NARS, either ARC or boards of the NARIs, work much better and regularly and discharge their functions in most of the cases efficiently and effectively. The priority setting is based on national development priority and the process of programming is bottom-up and involves the stakeholders at all levels. In the next section the programme management process is discussed.

2.4.3 Programme management process

The programme management responsibility in Cameroon is vested with the National Programme Committee that reports directly to its director-general. The operational aspect of the programme for implementation is under the responsibility of the directors of regional research centres in cooperation with the programme coordinators with delegation of authority for the stations and substations levels.

In Ghana, the process is under the management board of each institute. The management boards consider research proposals put forward by directors of institutes. Their role is thus, essentially, one of review, rather than planning of programmes. This is significant as the boards are the only forums in which users of research results, or their representatives from MOFA, have the opportunity to influence research programmes. Even research committees appointed by the management boards review ongoing programmes, rather than deciding on future programmes, though they may identify omissions, which may lead to an expansion of a particular programme. Furthermore, the management boards of individual institutions never consider relative priorities between programmes with responsibility vested with other agencies. The primary function of the management board is, in fact, the management of the physical means of carrying out research and of the substance of the research programmes.

In Kenya, national research programmes are coordinated and managed at the NRCs by programme leaders or programme coordinators based at the NRCs. The RRC directors, assisted by adaptive research coordinators, are responsible for adaptive research programmes of the regions, which are predominantly conducted in farmers' fields. Furthermore, authority and responsibility for implementing the research programmes have been progressively decentralized to the centres, including financial operations, once the annual work programme and budget have been approved.

For Madagascar, the execution of the research programme falls under the responsibility of the scientific director and the chiefs of department. For each programme the department chief appoints a national research programme coordinator who is responsible for ensuring the internal coherence of the programme and the adherence to a multi-disciplinary approach. The coordinators are responsible for annual programme and budget presentations and provide elements for FOFIFA's annual report.

In Malawi the task of programme management is the responsibility of research scientists under the supervision of the commodity team leaders (CTL) and the national research coordinators (NRC). The NRCs direct and integrate research on a commodity or project in a coordinated way to ensure that there is a balanced focus on the production problems faced by the farmers.

Mali, within a very decentralized system, the administrative management is decentralized and streamlined with direct responsibility for budget execution given to the programme heads and experiment station managers.

In Senegal, as for Mali the administrative management is decentralized and streamlined with direct responsibility for budget execution given to programme-heads and centre directors.

Overall, the programme management is rather similar from country to country, but quite decentralized within the research network set up with responsibility delegated where programme implementation takes place. The question mark is how this process is actually applied on the ground, effectively and efficiently. In the following section the resources available to NARS are analysed.

2.5 Resources of the national agricultural research systems

In this section the kinds of resources considered include: the human resources, infrastructure and equipment, and financial resources.

2.5.1 Human resources of NARS

Figures 2a and 2b portrait the situation of the human resources in the various NARS of the sample in terms of full time researchers equivalents as of 1990 for Mali and Senegal, 1991 for Malawi and Kenya, 1992 for Ghana, 1994 for Cameroon and 1997 for Madagascar. The following remarks can be made from these two figures:

• Kenya has the largest NARS in the sample and the smallest are Malawi, Senegal and Madagascar, Cameroon, Mali and Ghana have almost similar size;

• Madagascar has the highest percentage of national staff, followed by Ghana, Kenya and Cameroon. Senegal has the highest percentage of expatriates almost one third of total staff. This is an indication of the effort made by these countries to train and retain national agricultural research staff;

• in terms of qualification, on average 55.6 percent of the staff hold a post-graduate degree, with Senegal having the highest percentage followed by Ghana, Madagascar and Mali. In this sample it is no longer possible to discriminate countries in francophone and anglophone in terms of staff qualification, with the latter having a higher level of qualified staff;

• the ratio of professional staff to technical support staff is very low, less than two, with the exception of Malawi that has a ratio of four. This is an indication that highly qualified staff is not efficiently used in most of these countries.

Figure 2a. Human research capacity: levels of training

Figure 2b. Human resources categories of NARS in percentage

When analysing the evolution of human resources from 1981 to 1991, on average, the annual increase is 4 percent with only Senegal having a negative increase of -1.1 percent (Tabor, 1998). The highest percentage is for Madagascar, with 8.6 percent. However, a slow down can be observed when compared with the 1961-1985 period when the sub-Saharan Africa average was 6.8 percent.

2.5.2 Infrastructure and equipment of NARS

In the selected seven countries the equipment and infrastructure situation is described in the following paragraphs.

In Cameroon, the equipment and infrastructure constitute a liability. In 1994 both IRA and IRZV had a vast network of centres, stations and substations, in total 69. Their activities were often duplicated in the same agro-ecological zone. Total experimental land was more than 17 000 ha; offices, laboratories and equipment were deteriorating for need of proper maintenance not provided due to an acute financial crisis.

Ghana's NARS have the necessary land to perform research. However, buildings and housing are badly in need of rehabilitation. There was a serious lack of funding for maintenance and repair during the economic crisis of the 1970s and early 1980s. New buildings have been provided in some stations through donor funding. Elsewhere, planned improvements are often constrained by lack of funds. There is, therefore, a considerable need for upgrading and rehabilitation of many of the essential buildings and other structures of the agricultural research system. There is also a very considerable lack of equipment for undertaking research. Exceptions can be found, but these are nearly always due to specific provision from donors. The OPRI and two institutes outside the CSIR that contribute to agricultural research, the CRIG and the Forest Products Research Institute, receive substantial support under World Bank assisted projects.

Kenya, as the other countries, inherited a wide network of research facilities that has been streamlined over time. The public or semi-public components of NARS (headquarters, centres, outposts) have some of the best facilities among institutions in Kenya funded by the Government. The donors, mainly the World Bank, USAID, Germany (GTZ) and the Netherlands have also provided funding for purchase of equipment. In general, laboratories and equipment are often in very good condition despite little funding for maintenance from the GoK. This contrasts starkly with the poor state of physical structures and lack of maintenance of small pieces of equipment at public universities. The exception is the Jomo Kenyatta University of Agriculture and Technology, which has been generously endowed by the JICA's (Japan) assistance.

In Madagascar, the status of equipment and infrastructure is also variable depending on assistance received. NARS institutions inherited from the French a huge patrimony of infrastructure spread over the country including land beyond possible use of research institutions. Most buildings are old and built during the colonial era. They have not been maintained or rehabilitated. Within the streamlining process part of it has been rehabilitated with donor support and from government budget. However, the effort could not cover all needs for the priority programmes. As a result the infrastructure situation varies widely from one station to another depending on support received by donors. Limited budgetary resources of FOFIFA are used in priority for payment of salaries. The situation is the same with regard to equipment located in the central laboratories and stations. Much of FOFIFA's laboratory and agricultural equipment has become unsuitable or is redundant. The exceptions are those that have benefited from donor support.

For Malawi, the situation is similar to that of the other countries. Prior to the NARP project, infrastructure and equipment was in poor condition and insufficient. The project constructed and rehabilitated research buildings and required housing; provided much-needed scientific equipment, farm machinery, vehicles and office equipment/furniture. In terms of civil works: (i) the central buildings at Chitedze were rehabilitated and expanded including the construction of new administrative and library buildings, and soils and plant nutrition laboratories; (ii) Bvumbwe library, laboratories and selected other facilities were renovated; (iii) Lunyangwa's laboratory, library and office facilities were expanded; (iv) land was prepared and a laboratory, office and other buildings constructed at the new site of Mkondezi; and (v) 115 staff houses were built and 62 renovated to accommodate relocated staff in remote locations. It can be stated that DAR has adequate and appropriate physical facilities to carry out most of its planned experimental work. Its network of research stations, experimental stations and substations and trial sites has sufficient laboratory, office space and library facilities. However, there is still a need for specialized equipment and structures and more residential houses for staff at all levels.

In Mali, until recently, most of the infrastructure and equipment was concentrated around the capital Bamako. In the regions, before the merger of the two institutions in 1990, there was a lot of duplication between stations. Since then a streamlining of the network has been made and consequently equipment has been better distributed among the centres and stations. In 1992, there were 36 research experiment sites ranging from full-fledged stations with resident scientists to experiment sites, more than IER can manage adequately given resources available. The NARP project has, within the strategic plan streamlined the whole network and provided the required infrastructure and equipment for proper running of the priority research programmes.

Senegal, inherited a huge research infrastructure, which was maintained fairly well until the mid-1980s with the advent of economic difficulties and the era of structural adjustment. Moreover, during the implementation of the first master plan with the first World Bank and USAID funding, the network was expanded with new sites, the Kaolack Centre, the Tambacounda Centre and the Fanaye Station. The Djibelor Station was for example upgraded to a full-fledged centre, the Camberene Horticulture Centre (CDH). With the huge financial resources provided by these projects, infrastructure and equipment developed beyond the management capacity of the institute for long-term maintenance. Therefore, since its creation, ISRA as a medium-sized well-decentralized institution needing only to internalize a full regional concept within a multidisciplinary approach, developed into a huge bureaucracy in which the research scientists were marginalized. As a result its has been under perpetual restructuring and streamlining of its structure and is more or less heading to the previous organization set up in the late 1970s.

The infrastructure and equipment status as described above is an indication that none of the selected countries has assessed its needs through proper strategic planning in order to tailor the facilities inherited to the necessities of their priority programmes. Most of the facilities were for a regional scope and were beyond the needs of one country and hence the general poor status of maintenance. Some countries like Senegal have even added new facilities and increased the budgetary burden of recurrent costs particularly for maintenance. Therefore, all countries in the sample currently rely heavily on foreign assistance for maintenance of physical infrastructure and equipment. Downsizing of facilities, therefore becomes almost inescapable. The financial resources are analysed in the following section.

2.5.3 Financial resources of NARS

Financial resources are the most critical for NARS in developing countries and particularly the sub-Saharan Africa and in the selected seven countries. Complete data series are difficult to obtain and are missing for some countries. Data for this section came from ISNAR Indicator Series (Roseboom et al. 1993/1994)

The graphs in the figure below indicate the evolution of total funding of agricultural research in six of the selected countries.

Figure 3. Evolution of financial resources of NARS of selected countries

The graphs show three groups of countries, in the first group, represented by Kenya, there is a high funding level commensurate with the size of the research system that has grown steadily during the period (from 21 percent in 1983/1984 to 45 percent in 1992/1993). In this group, but at a lower level is Ghana, where funding also grew steadily. Cameroon and Malawi represent a second group, where since 1986/1987 there has been a notable decrease in funding. This decrease was very sharp for Cameroon, but it experienced a very high rise during the 1970s and early 1980s. The third group, composed of Madagascar and Mali, had a small increase or stable funding over the period.

For Senegal which is not represented for lack of reliable data during the period, considering the trend shown below for a more recent period, it is obvious that it belongs to the second group of countries for which the funding is decreasing.

Table 7. Evolution of financial resources of NARS in Senegal (current US$ million)

For a research institution the breakdown of budget in cost categories is as important as the total amount. The situation in the selected countries is as follows:

It is apparent from the figure below that there are two groups of countries, the first one with a high percentage of personnel cost between 69 and 77 percent (Cameroon, 69, Senegal 75.4 and Mali 76.7 percent). The second group is that of countries with a percentage of less than or near 50 percent for personnel costs (Madagascar, 43, Malawi, 44, Ghana, 48.7 and Kenya 49.9 percent).

Figure 4. Breakdown of cost categories of expenditure in percentage (1991, US$ million)

For Madagascar/FOFIFA, the cost of local personnel trended downwards over time due to a considerable reduction in support staff as well as low level of real salaries of researchers.

This situation is an indication of the heavy reliance of most NARS on donor funding for operating costs, indeed for actual research work, the research personnel, in the worst case, remain idle most of the time.

This over-reliance on donor funding is also clearly shown in Figure 5. The figure shows the source of funding of the major NARI/NARO of the selected countries. Again there are three groups of countries. The first group is funded at a level of more than 60 percent by donor assistance (Senegal, 65, Mali, 61 and Madagascar, 60.2 percent). The second group is donor funded at around 55 percent (Malawi, 54.7 and Cameroon, 56.3 percent). The last group funded at less than 50 percent (Kenya, 48.9; Ghana, 36 percent). This situation correlates well with the share of personnel and operating costs funded by national contributions except for Madagascar. Overall the resources from proceeds are very limited for all NARS, and do not amount to more than 3.5 percent.

Two other indicators are important regarding the funding of agricultural research, they are the expenditure per researcher and the research intensity, i.e. the national agricultural research expenditure as percentage of AgGDP.

Figure 5. Sources of funding (percent)

Evolution of the research intensity in some of the selected countries (Roseboom, Pardey, et al., 1993) is presented in Figure 6. There are three groups of countries: the first group with a high ratio well above one percent for all the period, comprising Senegal with an average for the period of 1.91, Malawi, 1.73 and Kenya, 1.54 percent. A second group comprising only Madagascar that started high until the break up with France in 1971-1975 and afterwards showed a constant decline of 4.1 percent per annum, down to 0.57 in 1991, however the average for the period is 0.89 percent. These two groups have an average for the period higher than the one for sub-Saharan Africa (SSA) of 0.78 percent. The last group is composed of Ghana that started fairly low and remains almost stable throughout the period with an average of 0.41 percent well below the continental average. This evolution actually reflects the importance of donor funding as with the national resources, only the situation is quite different as shown in the Table 8.

Table 8. Research intensities in selected countries

With national resources only, all the countries have research intensity well below the SSA average. In the particular case of Cameroon an historical analysis of the financial resources provided by national budget shows an annual increase of 9 percent from 1976/1977 to 1985/1986 that means in less than 10 years the budget increased nine times, this period of fast of economic growth was followed by the sharp decrease and the crisis of 1994/1995.

Figure 6. Evolution of research intensities in selected countries

A last aspect needing analysis within the review of financial resources of NARS is the expenditure per scientist, which can indicate if the scientific human resources are used efficiently and effectively. Table 9 shows the evolution of the expenditures per scientist for the selected countries. The data shows that there are three groups of countries. In the first group, comprising Ghana, Malawi and Senegal, expenditures per scientists are quite high and more or less stable, meaning that scientists have sufficient means for carrying out research. The second group, comprising Cameroon and Madagascar, for which the evolution is erratic with ups and downs but on average nearly acceptable for Cameroon and low for Madagascar. The last group comprising Kenya with a steady increase over the period, starting low to end and rising to an almost acceptable level.

Table 9. Evolution of expenditures per research scientist

(a) Current US dollars; (b) 1985 PPP dollars

However, this evolution during a rather short period does not tell the whole story, and for much of NARS these figures are meaningless as they include donors' contributions which are never available for all researchers with the system of enclave projects. Sufficient resources are often provided for scientists working on these projects while the others are almost jobless. This is the reason why FAO has introduced the notion of potential research capacity and actual research capacity. The former refers to the total qualified research staff and for the latter to the total qualified research staff with optimum means for carrying out research. Another factor that can bias these figures is the inclusion of expatriate research scientists. As a result, in 1990/1991, for NARS of the Sahelian countries as a group, the average for actual research capacity was 55 percent (for Mali, 41 and Senegal, 89 percent), while for Ghana it was 51 percent and Cameroon 60 percent.

Another feature is the skewed distribution of the financial resources between institutions in NARS. The academic institutions always have a token amount of funding for research. For the NAROs the commodity foundations have the lion share: for Ghana, in 1989 for CSIR institutes the recurrent expenditures net of salaries was US$3 900, while for the CRIG it was US$12 570; for Kenya, in 1991/1992 the operating cost for KARI was US$2 000, US$9 300 for the Tea Research Foundation and US$21 720 for the Coffee Research Foundation and; finally for Madagascar in 1997 the expenditure per scientist for FOFIFA was US$22 344 and US$140 050 for FIFAMANOR, which is six times more.

2.5.4 Linkages with the World Knowledge System

Over time all NARS have developed good linkages within and outside the system, however these are diverse from county to country, being mostly informal, but also formalized in some cases. The most widespread linkages are with the IARCs of CGIAR and regional/subregional organizations and with some bilateral donor organizations. Currently it cannot be said that the lack of linkages is a main weakness of NARS, the question is how these linkages can be optimized and transformed into a true partnership for strengthening their own institutions.

2.5.5 Conclusions

Diversified NARS exist in each country. All NARS are dominated by a NARI/NARO that accounts for at least 60 to 70 percent of all resources. NARI/NARO are always public organizations, funded by government. They have often inherited important research facilities from the colonial era. However, the infrastructure has often been too much for national needs. It has often been used with limited resources for maintenance. The result has been costly rehabilitation of infrastructure funded by donors or complete abandonment of infrastructure as part of a downsizing process. For all, human resource capacity has increased fairly well in terms of quantity as well as quality, but the inadequacy of the research environment has caused a high degree of instability amongst most NARS. Financial resources are the Achilles' heel for NARS. Funding has not increased at the same rate as human resources. Subsequently, expenditures per scientist have fallen to an inadequate level for most NARS. Full time employment is in many cases less than 50 percent.

Overall, NARS have had a positive impact, despite some weaknesses, on the production systems of these countries. They have also benefited over time from huge foreign assistance which impacts on the institutional development and sustainability analysed in Chapter 3.