Previous Page Table of Contents Next Page

Chapter 6: Agricultural research


In the first five chapters we have developed a framework and set out many of the issues that need to be considered in deciding whether to decentralize agricultural services. The purpose of the next six chapters is to apply these findings to specific agricultural support services. In this chapter we first examine the nature of agricultural research products. This is followed by a study of the factors that affect the supply and demand for these products. Finally we discuss special factors that need to be considered with regard to financing, production and regulation in deciding the appropriateness of different forms of decentralization. A similar format will be followed in each of the ensuing chapters.


6.1 The nature of research products

Types of research products

Three types of research products can be usefully distinguished[51]:

Saleable vs non-saleable products

In agriculture, it is useful to recognize two broad categories of research products from the point of view of their marketability: saleable and non-saleable products[52]. Saleable research products are those that embody characteristics which ensure their excludability, for example hybrid seeds cannot be produced without access to genetic material, which is the property of those who have developed the variety. Non-saleable research products are those which are not easily excludable (for example, biological pest control, silage-making technology, composite seeds, nematode-free cassava seedlings) and have most of the characteristics of public goods. Many have features more akin to those of processes rather than products.

If there is sufficient effective demand and the technical features of the products embody a substantial degree of monopoly, saleable research products can be sold at prices that remunerate research and development costs. Non-saleable products cannot be sold at prices which remunerate research costs, because they can be easily imitated by suppliers, produced or utilized by users who did not share the cost of research. For these products, the direct costs of production might be recovered, but, in practice, the cost of research cannot.

Research as a public good

We have seen in Chapter 2 that a product is a public good if it is non-excludable and non-rivalrous. According to this classification all products of basic research ought to be public goods which should be mostly provided by government. Historically, governments have indeed played a major role in basic research. In part this is because of their public good nature and the desirability to make them available to all producers of prototypes and of useable innovations. Another reason is because, for many types of research, a critical mass of resources are required leading to economies of size. Moreover, risks are involved in producing such products as a successful outcome cannot be guaranteed. As one moves towards the usable innovation end of the research product spectrum, the products become more excludable and rivalrous and increasingly assume the characteristics of private goods.

Basic research as a private good?

The sophistication of modern technology, the globalization of markets, and the concentration of large financial resources in private companies, however, have made matters more complex by closing the links between the three categories of research products. In the field of bio-genetics, for example, there are ways in which the excludability of products of basic research can be secured through the critical links between these products and the way in which they can be used in prototype development. This turns basic research products into private goods.

Control of new technology

Control of the features of a new technology is an important source of monopolistic power for the suppliers of technology innovations. In itself this is a positive feature of the industry, since it makes research profitable for private enterprise, and thus stimulates the development of technological innovations, which are the springboard of economic growth. The monopolistic power results either from difficult imitation, or the law. One example of difficult imitation would be a new implement where the effectiveness depends on the special material used. Copies by a competitor supplier would be an effective substitute only if competitors can obtain the same material and they possess the equipment required to forge it. Another example would be a chemical product that includes ingredients whose formulae cannot be readily discovered or that requires sophisticated and costly equipment for its production.

Patents and intellectual property rights

The law provides the other form of protecting commercial enterprises engaged in research and development from unfair competition, that is by competitors’ ability to produce the same product, or a very similar product, without incurring research and development costs. In the absence of such protection, there would be a ‘free rider’ problem and private enterprise would not risk the investment in technology development. The required protection consists in enhancing the monopolistic power naturally embodied in new products by way of patent rights, granted by governments, usually for a limited period of time and specifically for the characteristics of a single product. This permits the cost of research to be amortized, making research profitable, and thus attracting private investment. In a competitive environment, protection of intellectual property rights stimulates competition and thus promotes the development of new technologies.


To summarize, the main features which must be taken into account in assessing the nature of the products of research is the extent to which:

6.2 The demand for and supply of agricultural research products


From the demand side, many products of agricultural research that are likely to have an impact in developing countries are not saleable at prices which remunerate research costs. The main reasons for this are:

In the first three cases, there is a low degree of monopoly, and patents are meaningless or do not provide sufficient protection. This problem is made worse because protection of intellectual property rights is seldom enacted, except in a few large countries, or ineffectively enforced, which seriously reduces the scope for research products to acquire the features of private goods. Private enterprise would be reluctant to incur research costs that would not be recovered, let alone remunerated. These factors help to explain the limited scope for purely private involvement in agricultural research in developing countries, although this situation has often also been the result of deliberate government policies.

As a country’s economy develops, agricultural producers become more sophisticated and the share of cash transactions in their business increases. As a result, more inputs and more sophisticated tools tend to be used, more competition among input suppliers becomes possible with the enlargement of the market, marketing techniques become more articulated, and the scope for private involvement increases. Input and equipment quality and effectiveness acquire more importance for agricultural producers. This strengthens the capacity of suppliers to exploit the quality of their products, and to obtain better prices and/or more sales through marketing techniques. Marketing costs per unit of sales decline, and profit margins increase. At this point, the incentive system encourages private input and equipment suppliers to engage in research activities which would improve their position on the market.

Even in well developed agricultural economies, however, there is always an area of non-saleable research products and there is always need to fill gaps in basic research and potentially saleable products left unattended by private enterprise.


On the supply side, as mentioned earlier, the size of the financial resources required, the risk involved, and the presence of economies of scale are very important factors. Agricultural research costs money, often a great deal of money. It is an industry that often involves significant economies of scale. Research is, by definition, a risky business, although the risk probably diminishes as the size of the overall research operations increases. Economies of scale and risk factors have three important consequences:

The world market for biotechnology developments, for example, is in the hands of about five major private companies. However, even in the case of much simpler research fields, the cost of conception, design, on-station experimentation and field testing for prototype development, and the time required to place a useable innovation on the market, may far exceed the price at which, under the best circumstances, new products can be sold.

6.3 Appropriate forms of decentralization

6.3.1 The role of central government

These demand and supply factors and the nature of agricultural research will all have a marked influence on the appropriateness of decentralization, the forms it may take, and the levels of governance concerned with regulation, financing, and production.

The regulatory role

From the earlier discussion it is apparent that legislation governing intellectual property rights is essential to encourage the private sector to enter the agricultural research field or to allow public sector agencies to use cost recovery measures for its own innovations. The principle of jurisdictional spillover indicates that this is a central government concern, indeed regional and global legislation covers many of the issues involved in intellectual property rights. However, central government often delegates the implementation and monitoring of this legislation to a specialist agency or agencies.

Research has the potential to generate negative externalities and the principle of jurisdictional spillover indicates that it is a central government function to pass legislation to control potentially harmful research. Again, central government may establish a regulatory body or bodies to monitor research activities and guard against these possibilities. This may extend to banning or severely restricting certain types of basic and prototype research.

There is also a potential fraud problem associated with agricultural research products. Researchers may make misleading claims for their products that individual farmers or consumers have insufficient information to substantiate. Central government may need to legislate against such possibilities and may appoint specialist agencies to monitor this situation. This particular regulatory issue is discussed in more detail in Chapter 9.

Government’s granting of patents, certification of product quality, and import licensing of technology products can generate monopolistic positions. Therefore, utmost care must be exercised to avoid bringing about non-competitive environments. Protection of intellectual property rights is one thing, allowing a dominant position in a market is another thing. In a non-competitive environment with barriers to entry, the monopolistic power of new technologies allows the supplier to enjoy an economic rent. Furthermore, the supplier does not need to spend money to improve the quality of his products thus generally impeding further development[53].

Again, it is a central government role to establish a framework which encourages a competitive research environment. As we indicate below, there are a variety of ways in which this can be assisted by a process of decentralized reform.

The financing role

The earlier discussions have identified two areas where central government is likely to maintain a financing role. One is the financing of basic research that is unlikely to be financed by the private sector and which the principle of jurisdictional spillover suggests is most effectively financed at central government level. The other is with regard to non-saleable research technologies. These products will not be provided by the private sector, except in exceptional circumstances by philanthropic organizations. Many of these products generate positive externalities and/or assist food security and poverty alleviation. As we show below, the generation of non-saleable technologies may also play a role in reducing monopoly positions enjoyed by certain saleable technologies. If governments do not intervene, the spectrum of technologies available to agricultural producers would be sub-optimal and this would have two main negative consequences:

Thus government financing interventions may be justified on technical, economic efficiency, and poverty alleviation grounds. However, it is conceivable that in larger countries some of this funding could be by provincial or local government, rather than central government, where the benefits of new technologies would only benefit producers in specific locations.

The production role

In theory there is no reason why central government needs to play a role in the production of research products because appropriately specified contracts should enable this function to be carried out by a wide variety of organizations with central government providing the necessary financing and regulation. In practice, the very nature of research, where the outcome is not known with certainty, may make it very difficult to develop an appropriate contract. Moreover, where governance is weak, adequate supervision and monitoring of research contracts may prove difficult. There are therefore situations where it is more cost-effective for central government to conduct the research itself although this is often delegated to specialized institutions.

6.3.2 The reform process: the initial stages

One of the general objectives of decentralization includes responding better to peoples’ requirements and demands, improving efficiency in the provision of goods and services and improving the transparency and accountability of service providers. One of the aims of decentralization of agricultural research may be to bring researchers closer to the ultimate users’ of their products. This involves several aspects:

A first step in designing the reform policy is to assess the organizations currently entrusted with agricultural research and those to whom responsibilities may be decentralized. The issues raised in earlier chapters and the check lists in Chapter 5 are all relevant to this assessment. Following the issues raised in Section 2.4 it is of paramount important to understand:

We can anticipate that the results of the assessment of public research institutions will vary considerably given the great variety of situations that are extant in different countries.

Organizational culture

Breaking closed systems, when they exist, and developing a competitive framework, is an important element of the reform process. Many research institutes in developing countries are starved of resources because they are inefficient and, above all, ineffective. This stems from their power configuration and because they are closed systems, not for lack of rules, directives, or resources. They would have a greater claim on resources if the environment were right. If resources are made available regardless of performance, it is most unlikely that they will contribute to improving efficiency and effectiveness. A government’s commitment to reform is gauged by its capacity to change the environment of the institutions, which, in several cases, is not a painless operation. Unfortunately, the formulation of few agricultural research programmes addresses these issues. Instead, emphasis is put on the technical qualification of the staff, on the need for more and better-qualified technical staff, on inadequate equipment and resources and sometimes on organizational structures. It is seldom placed on organizational culture and, surprisingly, rarely on the relevance of research programmes. The result is invariably a list of programme objectives unrelated to the organizational reality, and a shopping list of items which donors are asked to finance.

The choice of the supervisory ministry

The choice of the supervisory ministry, its power configuration and that of the research institutes are very important in determining:

Even when the Ministry of Agriculture is the delegating authority, neither adequate planning nor accountability mechanisms, nor an adequate role of users in setting research agendas and evaluating results are guaranteed. In fact, experience suggests that, so far, they have been quite difficult to secure, and when they were, pressure from donors was a major factor of bringing them about.

6.3.3 Delegating and deconcentrating research activities

Central governments usually do not conduct research themselves but tend to delegate responsibility to agricultural research institutes in accordance with the principle of specialization. These institutes, particularly in small countries, are usually structured with several divisions, with the central management retaining close control over the divisions’ programme formulation, work plans and budgets. In other countries, especially in large countries and often in federal states, there may be several delegated institutes, responsible for regions, and/or specific crops. In nearly all cases the subsidiarity principle requires some form of deconcentration of the delegated agency and the establishment or strengthening of regional centres (Box 6.1).

Box 6.1 Deconcentrating research services in Mali

In Mali the Institut d’Économie Rurale has been deconcentrated to regional level, and central and regional research users’ commissions have been created. This has been done, with a view, on the one hand to strengthening the farmers’ participatory process and, on the other hand, to coordinating initiatives, avoiding duplication of efforts and fostering the flow of information.

Source: R. Pantanali (Personal communication).

Producers’ associations and cost recovery

The central government may devolve research responsibilities for specific crops directly to marketing boards or to producers’ associations, who would act independently of the public research institutes, set their own research agenda and implement their own programmes. In these cases, the transfer of responsibility is combined with the authority to levy a tax on the commodities produced to pay for the cost of research. This practice represents a return to an old system, which had worked rather well during colonial times in many countries, when the commodity research units were controlled directly either by strong commercial interests having a stake in the effectiveness of the research products or by commercial farmers. The authority to levy the tax may be granted by government or spontaneously levied by CSOs on the authority of their members. In the latter case the activity is exclusively private with the government’s role limited to authorization if this is required by the law. In effect, the association produces products that are saleable to their members. An example of such an arrangement is the Tea Research Foundation of South African Countries with headquarters in Malawi.

Participatory research

There is growing evidence of important results obtained by associating farmers in adaptive research activities. This includes decisions about the agenda and implementation of the experimentation work, with the result that research and extension activities are not just ‘more closely coordinated’, but become parts of a single tripartite field task. Box 6.2 provides some interesting examples of the new approach.

Box 6.2 Farmers’ participation in experimentation

The change in approach applies devolution principles, going beyond the stage of consulting farmers about their needs. It consists in moving from field verification of a few fine tuned innovations, developed in research stations, to participatory experimentation of many prototypes, by teaching farmers controlled experimentation methods. Examples of this include:

In Rwanda, farmers are experts in managing bean diversity. They use 500 cultivars, up to 20 on a single farm. The national research institute, ISAR, offered only limited diversity: from 200-300 research entries, only 2-5 reach an on-farm trial stage. Involving farmers in earlier evaluation (2-4 seasons initially, 5-7 after), had dramatic results: farmer selections out-performed controls by 38%; no gain from researcher selections was noticed; 21 cultivars were adopted by participating farmers after 2 years, an outcome equivalent to that of 25 years of ISAR work (Sperling et al., 1996).

In Latin America, a programme of developing locally elected committees (CIALs), composed of experimenting farmers who manage and conduct research on behalf of the community as a whole, records a success story in institutionalizing participatory research (Ashby et al., 1995). The strategy involves:

  • Participatory research is implemented through farmers committees

  • National research and extension agencies and NGOs help CIALs with decisions on research agendas

  • Farmers take the initiative in local testing and in creating demand

The following operational instruments and functions are adopted:

  • Committees are elected with four farmer members

  • CIAL conducts its own diagnosis and decides research topics

  • A host institution assists CIAL in designing the experiment

  • The CIAL Fund finances inputs, locally managed, on a revolving basis.

Results include:

  • After 4 years, 55 CIALs have been set up in Colombia, others in Bolivia, Ecuador, Honduras, Peru

  • 75% of involved farmer communities claim benefits: better seeds, improved practices, better information on technical recommendations

  • The cost of on farm research is about one third of the experiments managed by national research and extension institutes

  • Farmer agendas are different

  • Conservation and development of local genetic resources is achieved

  • Improved seed production

Research contracts

Delegation to private enterprise can take various forms, regulated by contracts for specific objectives and for a limited period of time. These contracts (outsourcing) specify the research work to be done and the remuneration to be paid for the work performed. Such contracts normally envisage agreements between private enterprises and national research institutes, but they can also involve universities, colleges and other educational institutes in the public or private sector having the competence to do the work. In these cases the research service providers are simply contractors.

Research partnerships

However there may be cases of private enterprise joining national research programmes by undertaking, and (partially or totally) financing, a sub-programme of interest to them and sharing the results with the host agency (or its deconcentrated unit), in a sort of joint-venture. Universities may undertake research on a cost-sharing basis with the specialized institutes. Contracting has the advantage of mobilizing highly specific competence, often not available in the institutes. It also limits the commitment of public funds and makes the contractor directly subject to performance.

Devolution to local government

Devolving responsibility to local government is really only a feasible option in large countries where a critical mass of human and physical resources is available at the provincial or district level. Even then some form of centralized organization and direction might be necessary to encourage the exchange of ideas and to avoid wasteful duplication of activities. Devolution does offer the possibility that research agendas would reflect local needs providing local elites do not ‘capture’ the agenda to meet their specific needs perhaps to the neglect of the wider farming population.

6.3.4 Encouraging private sector research

When research products naturally embody a high degree of monopoly, private sector involvement is likely to be forthcoming. In these circumstances the main role of the government is to ensure that a healthy competitive market develops. In addition, the government may wish to privatize all or part of its own assets engaged in agricultural research, especially those parts involved in ‘near market’ research developing useful innovations from prototype development. Market liberalization and encouragement of private enterprise research outfits (normally a part of commercial operations) are policies designed to achieve a healthy competitive market. In this context, ‘encouragement’ means:

Even when these conditions apply, however, there may be cases when private interests are not forthcoming. For example, when the impact of a series of technological breakthroughs in prototype development in one production line has been exhausted, and a renewed effort in basic research becomes essential for further growth. The experience of the Chilean fruit export agriculture (Box 6.3) teaches a lesson in this respect.

Box 6.3 Public and private roles in Chilean agricultural research

During the fruit boom of the 1970s, the Chilean government withdrew from research into export fruits. At the time, private companies readily took its place. However, when the boom ended, the kind of adaptive research that these private companies had been conducting became less productive, especially since most of the relatively easy gains had already been made. What was required at this stage was a renewed emphasis on basic research, but private companies proved unwilling to invest in this area, and the government capacity to do so had severely diminished.... Such cycle undermined the overall effectiveness of the Chilean research system.

Source: Carney, 1998, quoting a problem described by Jarvis (1994).

The case for and against privatization

To what extent then is it safe to withdraw government from research activities producing ‘saleable’ products? A powerful argument in favour of privatization is that activities are more efficiently carried out by private enterprise, on account of closer relationships with users, strict accountability to financiers, and a more professionalised personnel policy and environment.

Arguments against privatization are that it would not ensure that a competitive market will develop, because of economies of scale (in small countries multiple entry might not be a realistic perspective), and because multinational companies may easily circumvent competition by inter-company agreements and other means of enhancing monopolistic positions. Privatization, it is felt, will open the door to powerful multinational companies who would acquire a dominant position in the domestic market, and would tax farmers by selling their products at prices which reflect profit margins well beyond what is justifiable by the cost of research. The counter argument would be twofold:

A strong argument in favour of retaining a public function in the field of saleable research products is that the research agendas of large-scale private enterprises are not necessarily influenced by the requirements of the relatively small market of a particular country. It may be dominated by the interests of influential producers’ lobbies in their own political environment, and sometimes may not be sufficiently concerned with environmental conservation objectives relevant to a particular country.

Competition through non-saleable products

Multiple entry of suppliers in the private goods market of technology products is not the only way to enhance competition. There are cases in which alternative technologies may reduce the market for saleable technologies by substituting them with technologies which can be made available to a large number of farmers at a cost which is not related to the cost of research. In these cases, non-saleable technology products substitute saleable products. For example, techniques which improve farmyard manure production, silage and hay making (typical non-saleable technologies) can compete with chemical fertilizers and industrial animal feed, respectively, and thus reduce the need for purchased inputs with beneficial effects of the farmers’ cash flows and on the environment. Biological pest control techniques can compete effectively with chemical pesticides (Box 6.4).

Box 6.4 Non-saleable competition for saleable products

A particularly interesting example of a breakthrough in this line is integrated pest management (IPM), a technology which has been successfully promoted by FAO. This shows how monopolistic positions of large private sector suppliers of saleable technology products can be broken up, in practice, by providing competitive non-saleable technologies. IPM (a product that is difficult to sell at prices remunerating the cost of research) reduces the consumption of chemicals, the price of which can be easily controlled by the supplier because of the large economies of scale of the chemical industry and of the international patent rights they hold. The FAO IPM project in Indonesia represents an important example of government intervention in the field of agricultural research (and related extension activity) which has helped to considerably reduce the monopolistic role of chemical industries in crop protection, by substituting what is essentially a public good for a private good in the market.

Source: Pantanali (Personal communication).

Agricultural Research Funds

Another way of decentralizing and widening the number of providers of non-saleable products is to introduce and maintain a competition for the allocation of public funds for agricultural research. Competition can widen the channels of public financial resources allocated to research well beyond the national research institutes to include all professionally qualified competitors, domestic and foreign, in the public or private sector. The latter would include private foundations, several of which have recently begun to operate particularly in Latin America, and NGOs. The essence of the competition is that each competing organization presents its own research agenda, justifies it, and brings to bear its own specialization and competence. Competition among research outfits has the potential to improve the environment within which researchers work and offers more and better opportunities to them. This in turn may attract more talent to train and to devote their energies and imagination in the domestic research arena.

To this end, several agricultural research funds (ARF) have been established, often with substantial World Bank contributions, for example, in some African countries such as Kenya, Tanzania, Ghana, and more recently Cameroon. They offer opportunities to different agencies, in the private and public sector, to propose research programmes and to bid for the public funds available to finance approved programmes. The ARF are a decentralization tool that offers considerable advantages. Among others, the following appear of particular interest:

Managers and staff of national research institutes tend to oppose the ARF as they break a monopolistic regime. The experiences of the first attempts at introducing the ARFs show that bureaucracies have employed resistance and delaying tactics to boycott the smooth functioning of the funds[54]. Over time, and after the initial difficulties, however, the situation seems to have improved in most countries where the experiments have been undertaken. The success of the ARF depends on the proper design of its implementation mechanisms and, above all, on the political power of the committee or board members who are responsible for administering the funds. These committees have been designed to include a significant representation of ultimate users of the research products, among them poor farmers. In practice, the capacity of the representatives of poor farmers, primarily subsistence-oriented and women, to influence decision making in committees or board meetings has been marginal. This should not be a surprise since such representatives are hardly on an equal footing with other board members in discussions and lobbying. Thus establishing an ARF per se does not guarantee that the interests of the poor are automatically protected.

6.4 Conclusions

There is a range of reasons for government intervention in the provision of agricultural research services. This is particularly the case in low income, small countries where economies of size combined with low effective demand may make private sector provision unprofitable. In all countries the public sector is likely to continue to play a significant role in the provision of basic research. There are also likely to be types of non-saleable near-market technologies that justify some form of government provision. However, as the scientific base of a country expands it becomes increasingly possible to decentralize the production component through delegation to the private and voluntary sectors, providing the government is competent to devise and monitor the research planning and contractual arrangements. The private sector is also likely to assume an increasing role in the provision of research once the regulatory framework provided by the government creates excludability mechanisms such as intellectual property rights.

In larger countries it is more feasible to deconcentrate public service research provision. This has the advantage that it may make the research more relevant to local people’s needs but this will only occur if mechanisms are found for local farmers to participate in setting the research agenda. Moreover, deconcentration per se does nothing to foster competition between different potential providers of research services.

There are various innovations being experimented with in order to generate competition in the provision of agricultural research and the use of Agricultural Research Funds appears to offer considerable promise.

[51] This distinction follows the scheme elaborated by Evenson and developed by Pray and Echeverria (1990).
[52] The reader is reminded that 'non-saleability' of certain research products has nothing to do with lack of market development. Rather it is the effect of the intrinsic qualities of the products that are not those of private goods.
[53] Writing about one century ago, the great Austrian economist, Schumpeter, introduced the distinction between the monopolistic power that results from developing new technologies in competitive markets, which he called the 'God blessed monopolies' because they foster growth, from that which results from administrative measures which impede competition, and block development potential.
[54] Carney (1998), (Chapter 4) quotes the Kenya ARF as an example. She also mentions several design and administrative problems that caused similar ARFs in Tanzania and Ghana to perform initially below expectations. The approach has considerable potential, but experience is required to learn how to design the implementation mechanisms properly and to deal with vested interests that the approach is meant to overcome.

Previous Page Top of Page Next Page