Updated December 1997
|Food and Agriculture Organization of the United Nations||United Nations Capital Development Fund||International Fund for Agricultural Development||German Agency for Technical Cooperation||Swiss Agency for Development and Cooperation||World Bank|
16-18 December 1997
Technical Consultation on Decentralization
by Lawrence D Smith
Professor of Agricultural Economics
Department of Economics
University of Glasgow, Scotland, UK
There are many reasons given for decentralisation but these usually relate to:
In addition, when we relate the concept of decentralisation to rural development (and more specifically to the role of agriculture in this process), the changing nature of agriculture itself and the increasing diversification of rural activities both contribute to the complexity of local issues and also the need to find new institutional solutions to overcoming the changing nature of rural development problems . However, despite these apparently desirable features of decentralisation, any over-enthusiasm needs to be tempered with caution as even a cursory examination of the three reasons listed above will show.
With regard to the first reason, many observers assume that giving increased powers to local government will compensate for the weaknesses of the central State. However, a weak state is unlikely to provide the basis for effective decentralisation. Several factors may explain why this is unlikely to be the case and we can look to each of the conventional aspects of decentralisation - political, administrative and fiscal - for partial explanations of this. First, taking the political aspect, national level politicians are unlikely to be willing to surrender any real power to local level politicians if they themselves do not feel secure in their positions. There is, admittedly, an offsetting reason observed by Crook and Manor (1994), namely, that governments may devolve powers to a local level in order to extend their range of patronage and grass roots support in the rural areas. Second, from an administrative viewpoint, strong central institutions create a framework within which it is easier for local government to operate. This particularly applies to the legal framework. A weak state with a poorly functioning legal system is unlikely to be conducive to effective local government. Moreover, if there is a shortage of administrative expertise at the centre, it would be surprising to find strong and effective local administrations. Third, from the fiscal viewpoint, a weak state is likely to be typified by equally weak revenue raising powers and it is most unlikely that local government will be any more effective in this respect. Indeed, if it was it is likely that the central government would find ways of clawing back these revenues for its own use. For these various reasons decentralisation is unlikely to solve the problem of a weak central state. Thus in practice decentralisation is likely to be most effective when a strong central State decides that it is desirable that certain powers should be devolved to an appropriate level of subsidiarity.
Referring to the second reason for decentralisation, namely an inability of the State to continue to finance a whole range of services, there is little evidence that local government is any more successful than national governments in generating fiscal revenues. Indeed, Crook and Manor (1994) suggest that, for social and political reasons, the lowest levels of government find it extremely difficult to raise taxes to pay for local services or to impose sanctions on those unwilling to pay. In practice fiscal 'decentralisation' provides central government with a convenient excuse for abandoning certain functions - it does not guarantee an improved delivery. There is also the problem that while central government may consider fiscal decentralisation a useful device to reduce their own budget deficits, it may in practice generate macroeconomic instability if local government, particularly at the state level, engages in massive borrowing to finance their activities.
The third reason, namely that democracy is best served through devolved functions and widespread participation, is supported by many observers who concentrate their efforts on empowering those groups in rural society that until now have been marginalised in the development process. Sometimes this appears to be taken to excessive lengths with the impression that, through self help, producers' associations or other group activities can solve all of the problems of disadvantaged. This approach tends to overlook the fact that group activities typically suffer from inherent weaknesses such as moral hazard problems of shirking, free riding, and the capture of benefits by an Žlite within the group. For this reason group activities are likely to be most effective when the group is small, self selected, and providing a good or service with collective good characteristics that is likely to be under-produced by the market. Even here, as Olsen (1965) astutely observed, group cohesion is likely to be increased if the delivery of the collective good is linked to the provision of a private good that it is difficult to provide on an individual basis.
These observations are not intended to deny the benefits of participation and group activities, but rather to keep them in perspective. If the market fails to deliver at all, or in an uncompetitive manner, then group activity may be a useful solution or a source of countervailing power to correct the threat of monopoly abuse. However, typically the comparative advantage of group activities does not lie in the provision of private goods and services and only in certain circumstances will they lead to the effective delivery of collective goods.
However, there appears to be little attempt to synthesise these two approaches in the sense of exploring the role of decentralised government in providing an enabling environment for private sector activities, and this is one of the major objectives of this paper. In addition it may be useful to examine the role of the private sector in financing, producing and regulating goods and services which have previously been provided entirely by local government or that were provided by central government prior to decentralisation. Both of these issues require a reorientation of the decentralisation debate and a transformation of attitudes. The extent and speed of this transformation in attitudes is perhaps reflected in the fact that until recently the emphasis of FAO's training in this area was on decentralised planning with the implicit assumption that a considerable array of activities was to be performed by local government. What might be more appropriate now and in the future would be for FAO, or other organisations, to consider training local government cadres on how to create an effective enabling environment for sustainable and widespread rural development.
While economic efficiency considerations are important and should not be overlooked, it is important to appreciate that many other factors will interact with these efficiency concerns to influence the political economy, and hence the configurations, of service delivery in a decentralised setting. For example, if the central, or higher-level, government places a high priority on equity there will be a greater tendency towards centralised control over the allocation of resources and even the selection of target groups for the receipt of services. At any level of government, if there are economies of scale in delivery so that a large group is expected to share the benefits, it will be easier for an Žlite to capture the benefits for their own use. In this case, higher cost delivery to smaller groups may be preferred on equity grounds. Moreover, although politicians at the highest level may be in favour of decentralisation, bureaucrats in some line ministries may prove extremely reluctant to devolve their powers and may only grudgingly agree to deconcentration. The particular technical characteristics of the support service may also influence its mode of delivery, as discussed in more detail later.
A further complication is introduced when it is appreciated that the provision of any good and service can be disaggregated into at least four components: financing or funding the good or service; its physical production; the regulation of provision; and the consumption of the good or service . There is no necessity for these four components to be provided by the same sector or organisation. The degree of subsidiarity of these various components may thus also differ according to various circumstances. For example, there might be some types of services that the State may decide should be regulated at the central level, but which might be financed by local government. At the same time, the local government may decide that the production of the service would be most effectively conducted by the private sector, perhaps under contract, with the ultimate consumers being target groups within the local rural communities. However, a different configuration might apply to every type of good or service.
Thus to a large extent, the appropriate level of subsidiarity is an empirical question which will be a function of the size of a country, the number of levels of sub-national government, the resource endowments of the country and the particular characteristics of different goods and services. We should therefore anticipate a continuum of decentralisation with widely differing roles being played by various actors in different circumstances. This is also likely to be a dynamic process with the degree and extent of devolution and the relative role of different actors varying over time.
In the context of intergovernmental decentralisation at least three major questions need to be resolved. The first, mentioned earlier, is the nature of any economies of scale associated with the service, and the extent to which these advantages would be lost, and costs raised, by varying degrees of decentralisation. The second relates to the nature of any externalities or spillover effects associated with the service. Pareto-efficient delivery implies that the service should be controlled and financed at that scale where there are no spillover effects. Thus, for example, a service should only be controlled and financed at the District level if there are no significant spillovers into surrounding districts.
The third question to be resolved is whether the case for, and nature of, intervention changes as one moves towards local levels of government. There seem to be two conflicting possibilities here. Following Oates (1972), one of the major factors claimed in favour of decentralisation is that as the size of the population covered by a local government decreases the homogeneity of the population increases and their common needs can be more clearly identified. With agricultural services this effect may be reinforced if local provision reduces the range of agro-ecological diversity. This allows local government to provide appropriate types and levels of goods and services. For example, empirical studies seem to suggest that local collective goods that are overlooked by central government are high on the list of priorities of local communities, e.g. local rural infrastructure such as the provision, upgrading and maintenance of rural roads, and the provision of rural water supplies and rural electrification.
On the other hand, as the size of community decreases, it also becomes easier to identify who, within the community, will gain most (and who might be disadvantaged) by the provision of certain types of 'collective' goods. For example, viewed from a national government level the provision, subsidisation, or supervision of cattle dips to control tick-borne diseases may appear justified because of the externalities to other cattle owners. However, within a local community it may be only the wealthy households that own cattle and they may also control local government. The remainder of the community may resent financing such activities from public funds, especially when these are raised from local taxation. Cattle dips thus assume more of the characteristics of a club good  when viewed from the local perspective. In these circumstances one might expect cattle owners to pay for their use although the local government may play some role in organising or monitoring the facility.
This raises a more general question of the extent to which different agricultural support services assume the characteristics of club goods at the local level. To the extent that this does occur it provides an economic rationale for the encouragement of the farmers' associations consisting of members with similar demands for agricultural support services, and the financing of these services through membership dues and user fees.
It may be useful to pursue this line of reasoning and consider some of the characteristics of one of the agricultural support services that has traditionally been provided as a public service typically at a national level, namely, agricultural extension services. The major reason for the public provision of agricultural extension services is the public good characteristics of general or 'pure' agricultural information. Given time for diffusion, much information that is not tied to the acquisition of specific physical technologies is both non-rivalrous and non-excludable (Umali and Schwartz, 1994). There are also externalities associated with various agricultural practices, as well as moral hazard problems arising from asymmetric information. The public provision of extension advice may help to overcome these problems. A further justification suggested by Umali and Schwartz relates to the 'infant industry' characteristics of new technologies. One can expect falling costs and increased competitiveness over time as new technologies become better understood, and hence there may be a justification for accelerating this process through the public provision of information. Public sector provision may also be justified on income distribution grounds if farmers, or groups of farmers, are considered to be too poor to purchase sufficient advice.
On the other hand, if information is embodied in a physical good, as is the case with many new agricultural technologies such as agricultural machinery, chemicals, hybrid seeds and livestock, veterinary supplies and pharmaceuticals, then it can be classified as a private good because of high rivalrousness and excludability. Private firms thus have incentives to provide these goods and information related to their effective use (Umali and Schwartz, 1994). Even 'pure' information becomes more excludable as it becomes more specialised. For example, farm business advice may relate to an individual farm, or group of farms, and the farmer(s) may decide not to divulge the information to others. Such information has the characteristic of a club or toll good which is suited to private provision. Nevertheless, problems of externalities and moral hazard may still occur which may justify some form of public intervention, probably in the form of regulation.
How do these considerations influence the debate on decentralisation? First, there do not appear to be significant economies of scale associated with the delivery of agricultural advice per se. There are occasions when advice is sought by a client which is outside the competence of a particular extension agent, but these problems can usually be overcome by suitable networking arrangements with specialists in different subjects. The problem of continuously updating the agent's knowledge base can also be overcome in similar ways. However, if charges are made for new information, there may be a case for cost-sharing with other agents which suggests the possibility of some economies of scale in the overall provision of advice.
Second, are the externalities associated with advisory work such as to inhibit decentralisation? For most types of advice they are probably not and most spillovers are likely to be of a local nature. However, one can envisage circumstances, such as the outbreak of a new disease with severe and potentially widespread negative externalities, where timely advice could have substantial positive benefits. A deconcentrated public advisory service might be in a better position than a decentralised service to rapidly disseminate such advice. However, even with a decentralised service this type of problem could be overcome if the central source of information, say the Ministry of Agriculture, maintained a network for disseminating information of this nature.
Third, do needs differ in various localities and will decentralisation lead to a closer identification of needs and the provision of appropriate advice? As discussed earlier, as the size of a local government area becomes smaller, it might be expected that it would cover fewer agro-ecological zones and farming types, and thus farming systems would become more homogeneous than at the national level. This allows closer identification of farmers' needs and increased specialisation by an extension service which should enhance the quality of the public good. But these advantages could also be achieved by deconcentration, that is. organising the ministry providing extension services on a regional or district basis, rather than delegating control over the service to the local level.
Local needs can be even more clearly defined through the encouragement of the formation of farmers' associations incorporating groups of farmers facing similar problems. Usually the cost-effectiveness of extension delivery is increased if it is delivered to a group rather than to individual farmers. Associations have the further advantage that they provide a vehicle through which charges can be made even for 'general' information. This reduces the fiscal burden of providing the service. In this system it is possible, in theory, to differentiate the fee charged so that farmers' associations comprised of the poorest farmers pay much less than associations comprised of commercial farmers.
Another potential advantage of deconcentration, and even more so of devolution of control to local government, is that it offers the potential for further improvements because incentive structures can be created which reward the extension service for responding effectively to local needs. One way of achieving this is to encourage farmers' associations to hire the services of an extension agent(s) from the public service to meet their particular needs. Tendler (1997) describes an interesting case of this in the Cear‡ State of North-East Brazil where a farmers' association has an informal performance contract with a particular public extension service agent. The agent is provided with a list of queries and problems to solve and agrees to return with solutions within two or three weeks. His regular visits are facilitated by the association paying for his petrol costs. The incentive effects of this particular package appear to approach those normally associated with private sector activities.
On the basis of these various considerations it would appear that deconcentration or devolution of agricultural extension services appears desirable. However, there are two potential drawbacks with decentralisation, namely the question of finance and training. Local governments are typically more fiscally constrained than national governments, and national extension services are under-funded in many countries. Devolving the fiscal responsibility for extension to the lowest levels of government may lead to scant public sector provision, unless the collection of user fees through farmers' associations is extremely effective. To overcome this problem, and also on income distribution, and perhaps on infant industry, grounds there may be a case for part-funding of extension services by central government. With regard to training, national extension services typically incorporate training facilities for potential and existing extension workers. There are likely to be economies of scale in providing these facilities and it is important to ensure the retention of some form of cost-effective training facilities for extension workers. if decentralisation is the chosen route for the extension service.
Some interesting examples of the former can be found in relation to agricultural extension. It has already been argued that there are areas of extension advice which are suited to private sector provision. Decentralisation, and the likely shortage of funds that may accompany it, is an opportunity to question whether the same array of services should still be publicly provided. For example, Tendler (1997) quotes examples of farmers' associations hiring their own private extension agents rather than relying on those of the public extension service. Some countries, such as Chile, have withdrawn funding for extension from commercial farmers entirely but finance the use of private extension services by smaller farmers (Umali and Schwartz, 1994). It is also possible, in theory, for local government to provide the finance for farmers, or farmers' associations, to hire the services of private sector extension agents as in Colombia (Umali and Schwartz, 1994). The fiscal burden of this on local government can be reduced through the use of appropriate user fees. There is also the possibility of NGOs providing agricultural support services such as extension and thus relieving local government, to some extent, of the need to provide this service.
With regard to the second issue, the provision of an enabling framework, at the national level the major factors that would need to be considered are: macroeconomic stability and opportunities for profitable investment; the credibility of government actions; an effective and low-cost legal code and accessible legal system; positive attitude to private sector activities from the bureaucracy; a minimum level of regulation consistent with orderly production and marketing; a 'level playing field' for private sector and public sector activities; encouragement of private sector associations (Chambers of Commerce, traders' associations, etc.); and public sector investments to enhance private sector performance such as infrastructure, market development, and training (Smith, 1992).
To what extent can these various actions be enhanced, or hampered, by local government activity? Each of the national level issues will be briefly examined to see whether they apply at the local government level.
What is desirable, from a private sector promotion perspective is a bureaucracy that considers it to be in its interest to foster the growth of the private sector as a means of increasing long-term wealth and hence tax revenue. This usually requires firm leadership from the highest political level, and the redeployment or dismissal of those parts of the bureaucracy that are no longer performing a useful role (Smith, 1992).
Using the case of the provision of agricultural support services, and specifically agricultural extension, the three major factors determining the desirability and degree of intergovernmental decentralisation are the extent of economies of scale, the nature of any spillover effects and the extent to which local needs differ from those of the nation in general. Whilst decentralisation is likely to reduce agro-ecological variability and hence favour decentralisation it also reveals the heterogeneity within farming communities. Partly as a result of this, it would appear that some 'collective' goods assume the characteristics of 'club' goods at the local level which makes their financing/ production/ regulation feasible by voluntary groups. It would be interesting to pursue this topic in more depth to ascertain whether this holds for a wide range of agricultural support services. It is also observed that the provision of agricultural extension can incorporate private sector activities and that incentive structures can be devised to enhance their effectiveness. Whilst, in principle, there is nothing to exclude this occurring at the national level, in practice this flexibility is more likely to be achieved with decentralisation and the provision of club goods.
The paper concludes by looking at several ways that local government could provide an enabling environment for the private sector. Many of these require a change of attitude from regarding the private sector as a source of short-term taxation and economic rent extraction, to an appreciation that a vibrant private sector can increase the productivity and wealth of the local economy and hence increase the long-term viability of local government. The question is whether local politicians and bureaucrats are sufficiently long sighted to respond to this?
2. This disaggregation is used by Ross (1988), but a similar breakdown, with different terminology, is used by Ostrom (1990).
3. Cornes and Sandler (1986) define this as a good whose benefits are excludable but partially nonrival. Efficiency conditions can be established justifying charging a toll for such a good, hence the term 'toll' good is sometimes used.
Bennett, R.J. (1990) 'Decentralisation, Intergovernmental Relations and Markets: Towards a Post-Welfare Agenda? Ch. 1 in Bennett, R.J. (ed.) "Decentralisation, Local Governments, and Markets: Towards a Post-Welfare Agenda". Oxford: Clarendon Press.
Cornes, R. and Sandler, T. (1986) "The Theory of Externalities, Public Goods and Club Goods" Cambridge: Cambridge University Press.
Crook, R. and Manor, J. (1994) Enhancing Participation and Institutional Performance: Democratic Decentralisation in South Asia and West Africa. A Report to ESCOR, ODA of Phase Two of a Two-Phase Research Project.
Jaffee, S and Srivastava, J. (1992) "Seed System Development: The Appropriate Roles of the Private and Public Sectors" World Bank Discussion Papers No. 167 Washington, DC: World Bank
Oates, W. (1972) Fiscal Federalism New York: Harcourt Brace Jovanovich.
Ostrom, E (1990) "Governing the Commons: The Evolution of Institutions for Collective Action" Cambridge: Cambridge University Press
Ross, R.L. (1988) "Government and the Private Sector: Who Should Do What?" New York: Crane Russak
Smith, L.D. (1992) "The Role of the Public and Private Sector in the Process of Agricultural Development in Egypt" Government of Egypt, Ministry of Agriculture and Land Reclamation, Foreign Agricultural Relations/ Food and Agricultural Organisation of the UN: National Workshop on Agricultural Policies in Egypt, Cairo.
Smith, L.D. and Thomson, A.M. (1991) "The Role of Public and Private Agents in the Food and Agricultural Sectors of Developing Countries" FAO Economic and Social Development Paper No. 105 Rome: FAO
Tendler, J. (1997) "Good Government in the Tropics" Baltimore: Johns Hopkins University Press
Umali, D.L. (1993) "Public and Private Sector Roles in Agricultural Research: Theory and Experience" World Bank Discussion Papers No. 176 Washington, DC: World Bank.
Umali, D.L., Feder, G. and deHaan, C. (1992) "The Balance between Public and Private Activities in the Delivery of Livestock Services" World Bank Discussion Papers No. 163 Washington, DC: World Bank
Umali, D.L. and Schwartz, L. (1994) "Public and Private Agricultural Extension: Beyond Traditional Frontiers" World Bank Discussion Papers No. 236 Washington, DC: World Bank