The purpose of this chapter is to examine the particular features of agricultural extension services that need to be considered in deciding what forms of, and approaches to, decentralization may be appropriate.
Technical progress is the critical factor for fast growth of production in the medium and long term, and, under certain circumstances, technical innovations may contribute to improving the conditions of the rural poor. However, there is little point in generating new agricultural technologies if knowledge of these is not passed on effectively to producers. The purpose of the technology and advice services is to provide effective links between the research outfits and the ultimate beneficiaries of useable innovations. Loose, or no, accountability to their clients has been a major reason why many extension services have failed to provide such links, and so have contributed to the inability of the research system to provide effective responses to farmers problems.
Agricultural extension is one of those services whose economic character varies in different circumstances. It can be argued that general agricultural information, designed to improve existing cultural and production practices is a toll good in the short term (Umali and Schwartz, 1994). That is, it is non-rivalrous in the sense that one persons use of the information does not reduce its availability for others, but it is excludable, particularly if communications are not very effective. This makes it possible to charge people, either individuals or groups, for receiving technology advice. However, in the long term it is difficult to prevent people obtaining information and it thus tends to become a public good. The private sector will then not be interested in its dissemination unless it is part of a package that is excludable. On the other hand, it is possible to make specialized information that relates to a particular farm business excludable, it then assumes the characteristics of a private good and the private sector will have an incentive to provide it.
The stage of economic development is another factor that determines the economic character of extension services. For instance, there could be substantial positive externalities associated with extension advice in poor countries, where there is a need to expand food production and increase food security. This would then justify government intervention in provision, at least in financing extension advice. A review of the literature on this topic indicates that most authors maintain that governments have an important role to play, but this role varies according to country and sector specific circumstances, which evolve over time.
7.2.1 The effective demand for technology advice
Basic economic theory suggests that farmers will be willing to pay for technical advice if it promises to offer a high rate of return to the financial outlay required. This aspect is all too often disregarded by public extension services. Instead, a combination of ideology and vested professional interests has contributed to the widespread belief that smallholder farmers are unwilling to pay for technical advice. This is often ascribed to cash constraints and/or their non-commercial attitudes. When examined more closely, the reasons are more often likely to reflect the anticipated poor rates of return to the technologies given the resources available to farmers and the location-specific requirements of many innovations that limit their widespread use.
7.2.2 The supply of technology advice
In Chapter 6 we distinguished between saleable and non-saleable research products and argued that the private sector has an interest in developing saleable innovations because it can sell the products and recover the cost of research. This implies an interest in developing the technology transfer services required to develop their markets and this has indeed occurred with the liberalization of input and product markets. For these activities to expand successfully, new technologies capable of producing a product with the characteristics required by the market are necessary. At the same time, as argued above, the adoption of these technologies must be significantly cost-effective for farmers. This provides an incentive to farmers to pay for the technology and advice service received. There are many instances of such developments involving input supplies, and new crops (Box 7.1).
Box 7.1 Private sector provision of technology advice
Kenyas exports of horticultural products, after the liquidation of an initial disastrous experience in the public sector, has rapidly grown from scratch into several multi-million dollar businesses for a few Kenyan private companies, which are involved in adaptive research and technology transfer to contracted outgrowers. In these cases, research and extension costs are amortized against the export profit margins. In Kenya, the private seed industry also records a similar experience, working with outgrower-multipliers. In several large developing countries, integrated pig and poultry production has begun to develop. Private companies produce animal feed, chicken and piglets, slaughter and process the meat of animals produced by outgrower farmers. These companies provide the full range of technical advice and inputs to the outgrowers.
Source: Pantanali (Personal communication).
The supply of non-saleable innovation advice
On the other hand, we saw that private enterprise is not interested in developing non-saleable innovations because they cannot recover the cost of research. Does this mean that governments must provide the extension of non-saleable innovations, as they have a responsibility for research, or is there a function for the private sector as well? If farmers perceive that they will benefit from advice and they can only obtain it by paying for it then cost recovery is possible. Information is saleable to users, even if it is not entirely excludable and non-rivalrous. The very site-specific nature of on-farm management, for example, increases the saleability of information adapted to the local problem by the skill of technical service providers. Therefore, there are opportunities for cost recovery, and private and voluntary sector service providers can play a role in spreading non-saleable technologies, which may have been generated from government or non-governmental (including farmers themselves) sources. However, this role is of a different and more complex nature than in the case of saleable technology products.
7.2.3 Creating a market for technology advice
If an effective demand for extension advice and non-government sources of supply can be identified then the conditions exist for creating markets in extension advice, at least in some circumstances. Market-oriented users of technology products reward technical assistance services:
By paying a service fee to professional advisers
Through the price they pay for the use of the inputs they purchase
Through the price they are paid for the products they sell to specialized buyers.
In a competitive environment, unsatisfied users will turn away from ineffective service providers, and more efficient ones will fill the gap. This is why liberal policies advocate letting commercial interests provide technology transfer services whenever they are willing to do so at their own cost, risk and profit. In these cases, governments retain the responsibility for ensuring that a fair, competitive environment is established among competent private extension networks. This is likely to entail the provision of basic training for private sector specialists to avoid a free rider problem (employers would be unwilling to provide free training if employees could immediately leave and seek employment with competitor organizations). The government may also assume regulatory functions such as the certification of advisers competence by issuing professional licences, the inspection of their advisory activities and of the quality of the products they may sell.
Decentralization of extension services may be undertaken with a view to improving their relevance, responsiveness, and their cost-effectiveness.
There is a broad consensus that centralized extension systems tend to lack relevance. Many problems and solutions are location specific. This should give a clear advantage to the local provision of advice. What is the appropriate local level? Administrative boundaries rarely coincide with agro-ecological zones (or with socio-economic situations): there may still be a large diversity of situations within the territory of a local government. On the other hand, the capacity to adjust the advice to local conditions (or to group needs) may be negatively affected by decentralization. In particular, the linkages with agricultural research may become more difficult to maintain at local level than at national level, for example if there is no research facility in the region.
Responsiveness to farmers problems is another field where decentralization can reasonably be expected to provide benefits, if only because extension managers are physically closer to the clients. However, there are many other factors affecting responsiveness and decentralization does not guarantee equal treatment for all farmers.
Decentralization can help to moderate the financial non-sustainability of public extension services by:
Encouraging private advice providers
Letting input suppliers and crop buyers play their role
Reducing the scope and modifying the type of assistance given to commercial farmers
Promoting smallholders organizations and helping them to reach the stage when can pay private technicians for on-farm advice.
7.3.2 The role of central government
Core functions of a national extension service
The functions of a national government-run extension service may vary depending on the:
Development of the agricultural economy in a country
Competitiveness of the markets for agricultural input, equipment, and crops
Stage of development of the private and voluntary sectors
Degree of competition among local, regional, and international research outfits for public funds.
Nevertheless, some core functions may be considered of such priority in some countries that a case may be made for their provision by the government.
The regulatory role
Most regulations about technology products are formulated and enforced by the central government who have the relevant information and therefore the specific responsibility of diffusion through the public administration in the national territory. There is a potential problem with asymmetric information in the provision of technology advice as, almost by definition, the provider is better informed than the recipient. There is thus a regulatory role for the government in ensuring that farmers are not given misleading information. Another or additional way of tackling this problem is to finance the provision of correct information in critical areas.
The financing role
There are at least three areas where some government financing of services may required. The first relates to the provision of advice in areas where there are considerable potential externalities. For example the government may consider it desirable to ensure that farmers are aware of the public health implications of certain cultural practices. The second concerns the possible financing of advice to categories of farmers whom the government considers would be unable to obtain sufficient advice without financial assistance, for example, resource-poor farmers. The third possibility relates to training extension agents in situations where a potential free rider problem would make the private or voluntary sector, or even local governments, reluctant to train workers who might then move to other employers. In some circumstances it may be feasible to recover the cost of the training by imposing a tax on all firms, organizations and individuals licenced to provide extension advice.
The production role
As with the provision of research services, in theory there is no reason why the government needs to play a role in the production of extension activities. In principle, this applies even in situations where public sector financing is justified. For example, when urgent safety measures must be taken to avoid serious negative spillover effects, as in the case of an outbreak of epidemic diseases or of dangerous pest developments, the public administration can recruit specialists from universities and research institutes to appraise the situation and identify remedies. The public administration can then announce the mandatory remedial or preventive measures which should be taken, mobilize farmers through their professional associations and through the mass media, and carry out the necessary inspections to ensure that the measures are actually taken.
In practice, however, the balance of public and private production is likely to be an empirical matter and will depend, inter alia, on the quantities and capabilities of alternative providers.
7.3.3 Modes of decentralization
Experiments in the decentralization of technology advisory services have involved the combined application, in differing degrees, of deconcentration, delegation, devolution and privatization policies. In different countries, reforms have attempted the:
Deconcentration of national extension services/agencies
Devolution of responsibility to local governments.
Encouragement of farmers professional organizations to take over some responsibilities for technology advice to their members
Introduction of private extension services
The relative merits of each of these will be discussed below.
The subsidiarity principle and efficient management practices require that operational decisions are taken at the level where all relevant information is available to front-line managers. Optimum allocation of resources requires that operating units be deconcentrated to the level where significant economies of scale cannot be achieved by larger units and where the activities of the units have few or no spillover effects into other territories. Extension services, in particular, are craft organizations (see section 2.4.3) in which higher level supervisors of field staff are unable to observe and closely monitor the output of their subordinates. These organizations must be managed by creating a sense of purpose, allowing front-line operators to find their own way of performing tasks and letting professional norms and peers expectations play their role. In addition, clients need to be associated with the evaluation of the outcome of the agency, in such a way that the clients/beneficiaries effectively become part of the power configuration of the organization. This will become ultimately a powerful external coalition of the field workers.
These considerations call for the agencies to deconcentrate their structure. The main problems to be handled are:
Which specific functions/tasks should be centrally performed, and which ones should be deconcentrated?
What should be the territorial basis of deconcentration?
Box 7.2 The main tasks of an extension agency
Box 7.2 contains a simplified list of the main tasks of an extension agency. Decisions about which tasks should be deconcentrated would depend on:
The nature of the activity
The role played by the outside partners and the location of these partners activities
The need to reconcile the know-how, constraints, opportunities and experience currently and potentially available to clients, with the knowledge, present and potential, of the multiple sources of technological solutions to current problems and of other potentially interesting innovations which those sources may provide.
Administrative and financial management would have to be designed to reconcile controls and flexibility of action of the deconcentrated units.
Delegation to private or voluntary sector organizations
Delegation involves the transfer of resources and responsibility for specific functions from a central government authority to an agency in the private, voluntary or public sector. Under delegation, the task of the delegated agent and the resources required to implement the tasks are defined in advance. The delegated agent:
Has considerable autonomy of decision as to how to implement the tasks
Has its own procedures, but these must be approved by the delegating authority
Must have its programme of work and budget approved ex ante for every fiscal year
Is accountable to the delegating agency for the work performed
Must regularly report on the use of the financial resources received by the delegating agency
Is subject to regular audit of accounts.
The advantage of delegation is that inadequate service providers can be screened out through pre-qualification procedures and through contractual arrangements making continuation of obligations dependent on performance indicators been satisfactorily fulfilled. Normally, delegation is a more cost-effective way to perform tasks than departmental implementation, even if the contract costs may appear to be higher than departmental costs, which often include only part of the full cost of delivering a service. The Chilean experience described in Box 7.3 provides a good example of how such delegation (leading, in this case, to eventual privatization) can be put into practice.
Contracting the implementation of specific extension activities to NGOs, consulting firms, private individuals and farmers associations by the national extension service is a form of delegation that is increasingly used. An important new field, which offers good opportunities for contracting, is the use of mass media for spreading technology information to farmers. In many developing countries the technical capacity to design effective communication messages through radio and television networks is growing rapidly. Some private firms, in addition to national broadcasting agencies, are rapidly learning the trade and some are even specializing in communications with rural people.
Box 7.3 The Chilean experience with delegating and privatizing technology and advisory services
Chiles first attempt at privatization failed because of the false assumption that there was a market for technical assistance in the rural areas of Chile, and because of widespread fraud. In later years the Development Institute of the Ministry of Agriculture (INDAP) introduced subsidized private consulting services for small farmers, and drew up a plan to discontinue services to medium and large-scale farmers. Those farmers who remain eligible for service are assessed, and grouped according to their access to resources and productive potential. The poorest ones become the beneficiaries of a scheme which has primarily social emphasis (PTTB) and for which they make no financial contribution (15% is however planned for). Better-endowed farmers are catered to by a more market-oriented scheme (PTTI). This focuses on commercial developments and incorporates farmers contributions over the years. In both cases farmers receive services as a group, not as individuals.
Local groups, government authorities or INDAP staff identifiy demand for services. A cooperation agreement that lays out rights and responsibilities of both sides is signed between the local group and INDAP. After a process of bidding, a contract is signed between the service provider and INDAP or, more recently, the local group itself. Payments by INDAP are based on performance, with fines imposed for poor performance, and with the local group having the right to terminate the contract if performance falls short of expectation.
Private firms and NGOs seek certification by INDAP to allow them to bid for contracts to provide extension services. Those organizations which are rated to provide worst performance through the year are replaced as a matter of course. By October 1995, there were 139 technology transfer consultancies (TTC) registered by INDAP: 40% private companies, 36% NGOs, 17% farmers organizations, and 7% Universities, municipal governments, etc. At that time 75% of the expected contributions were paid by the beneficiaries.
Adapted from Carney, 1998.
Some countries have experimented with devolution of responsibilities for a variety of rural development activities to local governments. Some of these institutional reform experiments have received substantial support from international financial institutions, for example the World Bank, initially in Mexico and Brazil and later on in other Latin American and in several African countries. In Brazil, where the policy has been experimented with for a longer time, the experience has evolved towards forms of partnerships with CSOs. One common but important lesson in the present context is that few of the resources, which districts or municipalities were allowed to allocate freely to their own priorities, have actually been spent on providing extension services to farmers.
Potential problems with local government provision
Agricultural extension is rarely a priority for local governments, priority tending to be given to health, schools and public infrastructure. In Uganda, for example, a UNDP survey conducted in 1997 showed that only 3 out of the 32 districts surveyed had budgeted more than 3 percent of their expenditure on agriculture, most districts had budgeted to spend only 1 percent. Government response has been increasingly to tie financial transfers for local governments to specific purposes in the form of conditional grants, which runs counter to the objective of devolution.
In Colombia, the increased resources accompanying devolution to local governments were transferred on a conditional grant basis. This increased extension coverage by a factor of three. Costs also increased but not to the same extent, so that cost-efficiency may have improved (Garfield et al., 1997). However, to what extent larger coverage and more resources have improved relevance, responsiveness, and accountability to farmers is not clear. The law that created the UMATAs (municipal extension offices) also required the creation of municipal committees for rural development, with the function of co-ordinating and prioritizing rural development activities including extension. Rural community representatives are appointed to these committees, but there are no effective mechanisms for them to actually counterbalance the influence of the municipal government. By the end of 1998, less than half of the municipalities had created these committees and they seem rather powerless. The issue, of course, is not only one of farmer representation in local assemblies, but one of empowerment of farmers representatives and of cultural change of municipal extension officers as well. Local governments may not be interested, or may not have the human resources or the required skill and determination to bring these about.
Many other difficulties have been experienced by decentralization of extension responsibilities to local governments. The case of Uganda, for example, is a good illustration of both financial and staff management difficulties. Primary responsibility for allocating budget resources for extension rests with the district councils. But after five years the budget mechanism, flow of funds to district level and financial management procedures still need to be clarified (Government of Uganda, 1998). Available resources are all used for salaries, leaving practically nothing for staff development, mobility and fieldwork. Although the staff is now employed by the district councils they report technically to the Ministry of Agriculture, which has recently delegated its responsibility for extension to the National Agricultural Research Organization (NARO). NARO was not prepared to the new task and there is a clear risk that the organization sees extension staff primarily as an arm for transferring research results to the field. This could run counter to the intention of introducing farmer responsive services. On the other hand, the positive experience of the Uganda National FarmersAssociation (UNFA) (see Box 7.4) shows how much more effective it would have been to devolve the extension function to similar voluntary sector organizations.
In the Philippines, a recent survey (Malvicini et al., 1996) shows that devolved extension staff have lost career development perspectives; salary increases have been blocked to progressively align remuneration with that of other municipal employees who have mostly lower education levels; and there are no funds for staff training. Another issue, also mentioned in a number of countries, is the politicization of new recruitment. As a result of all this, the risk of staff quality degradation is high.
Perhaps it is in the field of staff policies that a strong central regulatory role is most crucial to successful extension services. This could:
Ensure suitable norms regarding staff recruitment and management
Ensure the supply of competent agents by supervising education, technical training and the approach to farmers problems
Facilitate linkages with research and the free flow of knowledge among regions and outside
Develop effective quality control and monitoring systems
Encourage organizational cultures which would spread:
- continuous farm diagnosis practices
- attention and appreciation of farmers constraints, survival strategies, experiences and points of view.
The role of farmers organizations
Encouraging farmers organizations in the voluntary sector to assume an important role in agricultural technology transfer is an integral part of decentralization policies. The need to develop strong grass-root level rural institutions at the receiving end of agricultural services has been recognized for a long time. An important feature of government decentralization policy vis-à-vis the CSOs is that the nature of support and promotion should preferably be devolved, not delegated. Devolution is characterized by the freedom of the receiver of government support to decide itself how best to use the authority and the resources obtained from the devolving agency. Resistance to devolution (often by people enjoying rent-seeking positions) is grounded on considerations such as:
The limited technical knowledge of the leaders of spontaneous farmers associations
The illiteracy of most members
The lack of training of their officers in extension methods
The absence of knowledge and vision about opportunities offered by technical progress.
The pertinence and relevance of these arguments is not always confirmed by empirical evidence. The most important features of success, such as leadership, entrepreneurship, and deep understanding and knowledge of local conditions, are common features among even the poorest farmers groups, whereas technical know how is a relatively secondary factor which can be acquired by the groups, either by training or in the market. Government may usefully provide general support to mobilize the energies and potentialities of farmers organizations, in areas such as:
Providing linkages with higher level organizations
Financing the training of leaders
Assisting in the acquisition of formal legal status if so requested by members or imposed by the law
Providing some initial physical capacity building, for example, simple office and communication equipment
Helping to overcome initial financial problems by offering some opportunity to perform small extension jobs on a contract basis.
More than that would inevitably create excess accountability of the associations to the devolving agency by comparison to the accountability due to the members of the association, and thus open the door to political and administrative interference. The pre-existence of formal or informal farmers organizations who get together to achieve a common objective identified by themselves is a pre-requisite of successful intervention. Artificially organized groups aimed at obtaining some temporary gain provided from outside their community must be avoided. Group formation has often been a self-defeating effort.
Limited support, mostly of training and management advice devoted to strengthening spontaneous development of farmers associations can produce very interesting results. The Uganda National Farmers Association (UNFA) experience (Box 7.4), for example, shows that CSOs can provide, on a largely sustainable basis, effective extension services in competition with the public sector agencies to a large number of farmers.
Box 7.4 Training and management advice to farmers associations in Uganda
One of the impressive examples of a farmer-driven service that offers a full alternative to the public sector is that established by the Uganda National Farmers Association (UNFA). UNFA, with assistance from the Danish Agency DANIDA, has established a demand driven, cost recovery extension service in a number of districts of Uganda. Members of the organization request training or advice on a particular topic and this is provided by UNFA employees who charge a fee to all beneficiaries. The success of this scheme has varied from district to district, but in some places demand has been very high and UNFA has been forced to expand its number of technical personnel. Many of the UNFA extension agents are former public extension officers, and are well connected with existing service networks, which is obviously an advantage. The apparent success of the scheme is such that UNFA has been requested to take over general extension provision in at least one of districts, after the Uganda government decentralization programme put extension in the hands of local governments.
Source: Carney, 1998.
Another interesting example of decentralization and of progressive introduction of private market features in technology advisory services comes from China (Yonggong, 1998). Since liberalization policies were introduced, as far back as 1978, many farmers associations have been spontaneously established in China, most of which have become strong, market-oriented enterprises. The Chinese government policy encourages these associations to enter contracts with the national extension service (Agrotechnical Extension Centres, which operates all over the country), and also directly with research institutes, universities, and individuals. They negotiate with any service provider of their choice regarding the technical assistance they wish to receive, and for which they are prepared to pay. Representatives of the sources of technical know how are invited to sit on the associations boards and share responsibility about decisions involving adoption of technology innovation by the members of the association. If unsatisfied with the service received, the associations are free to rescind the contract, and to address alternative service providers.
Private technology advisory services
Since the relevance, quality and timely delivery of technical assistance advice is crucial for farmers, their willingness to pay for effective services should not be underestimated. Indeed willingness to pay is a key to successful privatization policies, and a key proof of effective working of the accountability process. However, the introduction of private extension services is not a simple and easy matter. Governments have crowded out private services for a long time, but rushing drastic reforms may be counterproductive. In Chile, a first attempt at privatizing the extension services introduced in early 1980s failed (Box 7.3). The initial attempt was followed by a determined but gradual withdrawal of government from providing services. At first, wealthy farmers were encouraged to address themselves to private practitioners. Then commercially-oriented small farmers were encouraged to do the same while retaining public services for resource-poor farmers and introducing cost-recovery measures at the same time. These measures involved temporary subsidies to private extension enterprises. These were justified, from an economic point of view, on the classical infant industry argument.
The Chile experiment is probably the most advanced form of decentralization and privatization of extension services in developing countries to date. In Chile the central government has retained a public function, but has:
Mobilized groups of farmers to draw up specific demand-driven extension agendas
Financed private sector extension service providers, rather than directly provide such services
Decentralized the monitoring and evaluation function as well, through contracting to private professionals and requiring clients participation as part of the contractual arrangements.
In Latin America, most policies aimed at privatizing the agricultural technology information services have followed similar patterns:
Government withdrawal has been gradual, and guided by social welfare considerations
Public resources spent on extension were concentrated on activities addressed to resource-poor farmers
Private service providers were encouraged to service wealthier farmers on an increasingly financially self-sustainable basis.
Privatizing integrated crop production and processing companies
Integrated crop production and processing companies have traditionally provided input supply, credit and extension advice to growers under contractual arrangements. With privatization, because of the size of the operations, the sale of the public assets to private investors may result in splitting different parts of the integrated process. Alternatively, public assets may be sold in such a way that the size of each privatized unit may not achieve sufficient economies of scale to amortize the cost of providing services to farmers. In these cases, a service gap would result, and in the absence of strong growers associations, the opponents of reform would argue that government should fill the gap on the ground of offsetting the negative impact of market failures.
The privatization of cotton processing and marketing companies in West Africa, which are fairly large commercial enterprises in most countries, has met with difficulties and delays on account, among other things, of uncertain solutions to this type of problem. To prevent such situations arising, cotton growers associations have received considerable assistance in some francophone African countries, with a view to helping them to take over responsibilities, including extension advice from the integrated public sector companies under privatization.
In most developing countries, the reform of the extension services is prompted by the need to improve the relevance of its activities to location-specific conditions, the responsiveness to farmers production problems, and the need to reduce the burden on the government budget. Decentralization policies (deconcentration, delegation and devolution of services, promotion of farmers organizations and privatization) can help to achieve those objectives. Accountability to the ultimate clients/beneficiaries of technology advice is a key factor in bringing about such improvements, in making links between research and the diffusion of useable innovations effective and in providing research and extension officers with knowledge about valuable farmer-own experimenting and results.
Private enterprise has an interest in developing saleable research products and in diffusing information about them to develop the markets of their products. The development of non-saleable technologies is a government responsibility. Information about non-saleable technologies, however, can be provided on a cost-recovery basis and therefore there is a role that private and voluntary sector organizations can play in these products as well with respect to commercially-oriented farmers. In a fully developed agriculture with performing markets, the private and the voluntary sectors provide technical advisory services. At the same time, government retains an important but marginal function as supplier of last resort, serving poor marginal farmers. However, this function is often outsourced to private companies or to NGOs.
In developing countries, including poor countries, there is ample evidence of farmers willingness to pay for effective technology advice provided the technology offered enables the clients to make substantial financial gains. There is also evidence of emerging medium-size private domestic companies which develop their own extension networks to promote the production of crops which they want to market, such as exportable horticulture or seed, recovering the cost of extension from the purchase price of the crops. In large countries integrated livestock companies provide full technical advice and supervision to outgrowers of pigs and poultry.
Government has a role to play in providing technology advisory services to offset the impact of market failures, particularly asymmetric information. Many tasks in this field can be outsourced by financing private technical advice providers. Core functions are retained in the public sector. These functions may include providing advice on issues with significant negative externalities and servicing resource-poor farmers. Other possible core functions include the promotion and support of on-farm adaptive research programmes and of farmers groups and associations capable of addressing themselves to private advisers. This helps to offset the impact of asymmetric information in new technology products and in crop marketing. The scope for outsourcing these activities depends on specific country situations. Finally government retains the function of evaluating the outcome of extension programmes, and the performance of private contractors using public funds.
Experience with devolution of responsibilities and resources to local governments suggests that agricultural extension in not a high priority at local government level. Increased extension activities took place when resources were transferred by the central government on the basis of conditional grants. To what extent this resulted in more relevance and responsiveness is not known. In other cases, inadequate resources were transferred and the institutional reform made no difference in terms of performance. More information is required to draw conclusions about the impact of transferring public extension services to local government level. A fairly diversified picture would be likely to emerge from a larger number of more detailed case studies, suggesting that a doctrinaire approach, i.e. devolution is better than deconcentration, local is always good, private sector is by definition more efficient, etc. should be avoided.
In Latin America, there are several examples of extension privatization schemes that led to gradual withdrawal of the public technical advisory services. Concurrently they promoted the growth of private services and successfully introduced cost recovery for wealthy and commercially oriented farmers, while retaining subsidized services for marginal areas and poor farmers. Supporting farmers associations in developing extension networks has also been successfully experimented with in several African countries. In China, strong, commercially oriented, farmers associations contract technical advisory services with government extension and research institutes, pay for the service rendered, and are free to change contractor if they are not satisfied with the advice they receive. In other Asian countries, privatization experiences are still limited, largely on account of ideology, strong vested interest in the public services and the belief that new extension methods can achieve more relevance and responsiveness within the public sector.
 Participants interested
in the theoretical discussion about the economic nature of the extension service
are referred to the excellent summary presented in Carney (1998), Chapter 5 and