In the previous chapter it was noted that the provision of services could be disaggregated into four components and that there was no need for the same sector, organization or level of governance to deliver them all. The objective of this chapter is to examine some of the factors that need to be considered in determining the appropriate level or form of governance to deliver particular functions. This is done by posing the following three questions. Who should pay? Who should produce? Who should regulate?
One of the reasons why governments have provided agricultural services in the past is because it has been considered desirable or necessary for the government to pay for them. The decision to decentralize offers an opportunity to question this decision - indeed, as we noted earlier, in some cases the reason for reform has been the unsustainability of the financial burden on government of paying for services. To understand the issues involved, we will first look at how the market mechanism determines this payment issue. This is followed by a review of how central governments finance the services they provide and how different forms of decentralization alter these arrangements. We will then consider how cost recovery procedures could be introduced and some of the problems that might be encountered.
3.1.2 Market-based pricing principles
Demand and supply
It will be useful to start this discussion with a review of the role of prices in a market economy. In a market economy the price that a rational consumer is willing to pay for another unit of a service is less than, or equal to, the benefits that they hope to obtain from their entitlement to that service. Typically, consumers are willing to demand a smaller quantity of a service as its price rises, in part because of income constraints and in part because other goods and services become more attractive substitutes to them.
From the producers point of view, a seller will be reluctant to continuously sell a service for a price that is less than the full replacement cost of the resources required to provide it because they will make a loss and would be better off using their resources elsewhere. As the price of a service rises, producers typically will be willing to offer more of it because they can make a more profitable use of their resources in this way. In a competitive economy with freedom of firms to enter or leave the industry, the balance of effective demand and supply will determine the price of the service. Market price can be thought of as a rationing device. If demand increases price will rise and existing firms will expand production. If prices rise sufficiently to generate economic profits new firms will be entering the industry. This increased supply will tend to depress prices and eliminate the excess profits and we can conceptualize an equilibrium price that is socially efficient where supply just balances demand.
Effective demand and need
It is important to appreciate that, in a market situation, the demand for a good or service is measured by the amount that consumers are willing to purchase, and have the ability to purchase, at a given price. The stipulation that the consumer must be able to pay for the good distinguishes effective demand from need. A farmer may need hybrid seeds to increase her yield but if she cannot afford them she has no effective demand for them.
When a buyer pays a seller for a good or service they are, in effect, exchanging property rights or entitlements. The nature of these entitlements differs according to the type of property right associated with the service being transacted. Private rights in a service means that the particular person concerned and no one else has the authority to decide how the resource should be used (Ricketts, 1987)
To be truly effective it is desirable for this right to be backed up with the force of law to establish a legal entitlement to the right. This is because if a private property right is ill-defined, some of the entitlement will be in the public domain and not under the control of the owner. This will reduce the price that a buyer would be willing to pay for the service or even make it impossible to sell. However, even full legal entitlement does not allow that person completely unconstrained use of the service - the property right may also impose limitations on how the service can be used.
3.1.3 Government fiscal arrangements
We noted in Section 1.1.2 that the government agencies set up to deliver agricultural support services were often monopolies and the services they supplied were frequently subsidized or free to the consumer. Very often there was little control over the agencies costs. We also noted that the increasing costs of subsidies contributed to the government deficits that led to the reforms including calls for the decentralization of government activities. This implies, inter alia, a clarification, or revision of the fiscal arrangements for services that are to be decentralized.
Fiscal arrangements with central government provision of services
When services are provided centrally through a ministry, such as the Ministry of Agriculture, typically they are financed by public funds. Public finance is managed by the principle of the unified consolidated current account (UCCA). All revenue from fiscal (taxes, levies, duties) and non-fiscal origins (fees, charges for service rendered, revenue from other sources) is deposited in a central account held at the Treasury (the government bank) and released from this single account to different government agencies. The balance of the UCCA thus signals the current cash flow position of the government. The Treasury regulates the release of funds in accordance with the availability of funds during the fiscal year.
In principle, the Treasury releases funds to a ministry in accordance with the budget vote of the national assembly that empowers the government to collect and spend financial resources. The officer heading each ministry or government agency of similar rank, (normally the Principal Secretary) is also its chief accounting officer (CAO). The Treasury issues to the CAO the authority to incur expenditure (AIE) in accordance to the budget vote, generally on a quarterly basis. The AIE specifies the amounts for each item of expenditure identified in the approved budget vote. Each CAO in turn issues subsidiary AIE (warrants) to designated officials of the agency. These warrants authorize them to incur expenditure up to the limit stated in the AIE. The warrant holding officer is authorized to issue payment vouchers up to the limit stated in the warrant for each authorized item. These vouchers instruct the Treasury to make payments to suppliers. AIE and warrants do not include authority to pay government salaries, which are directly paid by the Treasury.
The transfer of financial resources and assets from the central administration to other levels of government is called fiscal decentralization. Fiscal decentralization affects the revenue as well as the expenditure side of the budget, though the impact at local level is generally much stronger on the expenditure side. In particular, fiscal decentralization concerns the:
Authority to raise taxes locally, and/or to borrow on the capital market
Authority to budget independently the use of own and central government transferred financial resources
Design of the instruments to transfer central government financial resources to decentralized organizations
Procedures to make use of the resources transferred under those instruments
System of controls over the use of resources transferred by the central government.
Decentralization can have profound effects on central government control of the financial aspects of the economy and through this on the performance of the whole economy. It is also the case that in many countries the effectiveness of decentralization to local government is adversely affected by weaknesses in revenue raising and in transfers of resources from central government.
Transfer of functions vs. resources
Historically, there have been several cases in which the transfer of functions has not been accompanied by an adequate transfer of resources. In these cases the provision of services has deteriorated. There are several reasons that can account for this:
The central government did not really wish to transfer the functions
They did not have the resources to discharge them
More frequently, the opponents of decentralization manage to block its effectiveness, with a view to providing evidence for reversing the decision to decentralize, so that they could conserve positions of influence and power, which they would have lost if decentralization succeeded.
Deconcentration of service provision to peripheral locations does not involve fiscal decentralization. With deconcentration the officer in charge of the deconcentrated unit, for example the chief district agricultural officer (DAO), becomes an accounting officer in his own right. The DAO reports hierarchically to the Principal Secretary of Agriculture, and prepares the district work programme for approval by the central ministry. The DAO also prepares an annual budget, which after approval by the central ministry and the cabinet is included in the budget vote. However, the DAO receives the respective AIE directly from Treasury rather than from the Principal Secretary, and issues payment vouchers which are normally honoured by a district (or regional) sub-accountant who is the local representative of the Treasury.
Another change that may be introduced with deconcentration concerns the appropriations-in-aid (AIA). These essentially consist of two categories:
Non-fiscal revenue, such as charges for services rendered, sale of goods (for example vaccines), and occasional sale of equipment
Funds made available by external loans or grants.
Deconcentrated units of a central ministry may be allowed to retain this revenue, although this rarely occurs. When the non-fiscal revenue component of AIA is not deposited in the UCCA, the Treasury tends to release funds to deconcentrated units on a net basis, that is assuming that the cash flow requirement is equivalent to the expected expenditure, less the targeted revenue from AIA sources. Failure to collect the targeted revenue by the deconcentrated unit generates a cash shortfall at local level.
Deconcentration may improve administrative efficiency by drastically reducing the number of operations involved in authorizing, certifying, and making payments, and because deconcentrated offices, including those of the Treasury sub-accountants, are smaller and the number of transactions much less at district level. Such efficiency gains are actually achieved only if the local level is adequately staffed with trained accountants and funds are indeed released by Treasury in accordance with the vote. In many developing countries these conditions are not always met.
Devolution to local governments does imply fiscal decentralization and significant changes to procedures. The most far-reaching is that the UCCA principle is released with each devolved local government handling its own financial transactions independently, under the supervision of the Treasury. Local governments are given the authority to:
Raise fiscal revenue through local taxes
In some cases to borrow from the capital market
Retain their own collected fiscal and non-fiscal revenue
Retain the grants and loans part of their AIA, to which private donations may be added.
Local governments in rural areas may independently raise resources in four ways:
By imposing local taxes
By selling goods and services
Through voluntary (cash and labour) contributions in local projects by communities/grassroots organizations
Through donations from private sources.
The main sources of revenue for local governments include licences and other fees, stamp duties, fines, road tolls, and charges for services rendered such as water and power supply in urban areas. However, very few local governments in rural areas manage to raise significant amounts relative to the cost of discharging their responsibilities. The reasons for this include:
The very poor tax base
The reluctance of local politicians to take unpopular measures
The resistance of potential tax payers who have little confidence in the capacity of the local administration to put the money to good use
A widespread feeling that it is the rich central government that should take care of the need of the poor local administrations.
These complex factors take time to change, so that the rate of growth of local revenue tends to be generally modest.
Inter-governmental financial transfers
To complement the local revenue, the central government transfers funds from the Treasury account to the local government accounts: these are called Inter-Governmental Financial Transfers (IGFT). IGFTs often represent the largest share of the local government budget and the way they are designed determines the degree of devolution that the central authorities allow in practice. At the local level, revenue from all sources is spent in accordance with a local budget, approved by the local government assembly, and generally submitted to controls by the central government. The degree and content of such control is another determinant of the degree of devolution allowed.
There is a strong possibility that the ready availability of IGFTs generates transfer dependency. This is where local governments adopt an attitude of expecting central government to pay for everything. Local politicians then tend to see their role as lobbying central government for more resources rather than attempting to increase local self-reliance.
Delegation implies a variety of financial arrangements. One form of delegation consists of a contractual relationship between the central government and an independent agent (in the private, voluntary, or public sector). The agent is required to perform a specific public function, or to supply a set of specified goods and services, in exchange for an agreed sum of money. For example a private company or an NGO can be delegated to implement a rural development project or some of its components. Such contracts may be renewed on an annual basis or cover longer periods. Delegated agents must submit their annual budget to be approved by the delegating authorities (normally a line ministry and the Ministry of Finance). Treasury transfers funds to them on the basis of procedures agreed within the framework of the contractual arrangements.
Some forms of delegation consist of establishing autonomous agencies, for example the research institutes or an irrigation agency. These agencies have a separate budget and a separate administration, but their financial procedures are often those of a government department: the budget is approved in the same way, and the release of funds is through the Treasury. However these agents are normally authorized to retain their AIA.
Other forms of delegation may involve the creation of a public enterprise, for example a national water supply or a power company. These are commercial operations that keep commercial accounts. The companies sell goods and services and retain the income from such sales. In these cases the central government makes a payment to establish the equity capital of the company, guarantees the companys borrowing from the capital market, and often undertakes to refund part of the expenditure incurred in providing the services by way of subsidies approved under the budget vote.
Partnerships with CSOs
Partnerships with CSOs are not subject to the same financial control requirements since the central government is under no obligation to honour their financial commitments, except for disbursing the funds granted to the CSOs under contractual arrangements. Managers of CSOs are subject to the legal consequences of unsound financial practices like all managers of private enterprises. The transfer of funds to CSOs requires special instruments that technically are not IGFT, but may have similar features. In particular, it is preferable that:
The instrument clearly indicates the reason why the partnership is established, its objectives, and the strategy and the approach that the CSO commits to take to achieve the objectives
The instrument establishes the principle that resources will be made available over a reasonable period of time, subject to a good record with respect to agreed indicators of performance and of financial discipline. This still applies even if fund release is done on an annual basis to fit in with government budget management procedures
The financial commitment is in the form of a lump sum and disbursement is not linked to specific projects. This may affect the CSOs independence in allocating funds among different activities, and its ability to respond to the demand of its members. Moreover, it would enormously complicate bureaucratic procedures with respect to budget approval and fund disbursement.
Especially designed instruments for financial transfers to CSOs have been introduced in some countries. They are used on an ad hoc basis and donors projects often experiment with them. These instruments (conventions, contract-programmes) are in the nature of contracts which delegate certain functions such as support to grassroots organizations, group formation and training and may include handling of funds to subsidize self-help projects. Conventions provide a longer-term framework to the contracts, and an agreement in principle that the government will provide financial support to the contractor. Contract-programmes are more specific, and commit the release of government funds in a particular financial year. In these cases the CSOs have considerable autonomy in the way they perform the assignment. However, such autonomy can be diluted when the central administration requires that the annual budgets of the CSOs be subject to a detailed ex ante approval, or when central approval is required for individually presented community projects promoted through the CSO.
Deconcentration involves only a change in budgeting and release of funds procedures within a unitary administration. However, delegation, devolution and partnerships involve changes in the sources and the flow of fiscal and non-fiscal resources from one level of government to another and from government to non-government actors.
3.1.4 Cost recovery and pricing regimes
Full cost pricing
So far, in considering the fiscal arrangements associated with decentralization, we have implicitly assumed that the services to be decentralized are to be publicly financed. However, one of the reasons for, or consequences of, decentralization may be to remove or reduce the subsidy element from the services being provided. This involves the establishment of an appropriate pricing regime. But if the government has been a monopoly supplier there is no guide from other existing firms in establishing this price. Nor has there been the discipline of new firms entering the market if the price is sufficiently high to generate economic profits. How then should the price for the service be set? If the intention is that the service should be placed on a financially sustainable basis and consumers of the service should not be subsidized by, or subsidize, taxpayers then the full average cost pricing principle would seem appropriate. In principle, this would require the supplier to estimate the full replacement costs of providing the service and then to set a price per unit so that over an appropriate time period these costs were just met. In this way sufficient revenues would be generated to sustain the service even when the fixed resources and infrastructure needed replacing.
A few minutes reflection soon illustrates some of the practical problems of implementing this principle. How is the effective demand for the service to be estimated? How sensitive will quantity demanded be to price? How are costs allocated to a particular service, if an agency is providing several services? How are the costs of resources used to provide a service to be estimated? What is to prevent the costs, and hence price, being inflated by unnecessary expenditures? Although these issues appear complex, private sector companies deal with all them of in their day-to-day business. How well they deal with them determines whether they succeed or fail. Similarly, if sustainable unsubsidized pricing systems are to be established for agricultural services as a prelude to, or as part of, a process of decentralization, these self-same issues will need to be addressed.
In the absence of guidelines provided by market forces there are sound reasons for agencies to adopt an incremental, or step-by-step, approach to cost recovery measures. One reason is that it takes time for farmers to adjust to paying for services that were previously provided free or at low cost. Whilst they may need a service, or use it if it costs them nothing, this does not represent their effective demand when they have to pay for it. Experience suggests that farmers may initially boycott a service, or substantially reduce their use of it when charges are first made. It takes time to establish whether this decision adversely affects their business and to decide on the optimal level of usage if payment has to be made for the service.
Another reason is that it may take time for the agency to establish how much it actually costs to provide the service. As we have just seen, it may take time to establish the likely effective demand for the service and the most effective means of delivery. The agency previously supplying the service may have had an unnecessarily high level of expenditure that cannot be reduced immediately. In particular, it may take time to adjust employment levels. Moreover, it may be difficult to estimate replacement costs for items of fixed equipment and infrastructure.
3.1.5 Continued fiscal support for service provision
Central vs. local government priorities
In practice there are a variety of reasons why, historically, government agencies have not applied this full cost pricing principle. In Section 1.1.2 we stated that governments frequently subsidized the services provided by public agencies. In terms of the analysis in Section 2.1 it could be argued that this was done on merit goods or redistribution of income grounds. If provision of these services is decentralized these reasons for subsidization are likely to be re-examined. One of the reasons for this is that the changed fiscal arrangements are likely to make the costs of the subsidies more transparent. Another reason, particularly applicable with devolution to local government, is that priorities may be different to those at central government level. Those services deemed to be merit goods at central level may be viewed differently by some local governments. Equally, views on which groups in society should be provided with subsidized services may vary because political or social priorities differ at the local level compared with those of the national government. Central government will have to provide an appropriate IGFT to local governments or other agencies responsible for service provision if it wishes these subsidies to continue, but local government does not consider them to be a sufficiently high priority. Central government may even alter its own priorities, as the costs of subsidizing services become more transparent.
Local collective goods
In Section 2.1 we also saw that there are market failure grounds where it would be inappropriate, or difficult, to use full cost pricing. For example, we saw that with pure public goods the free rider problem makes it difficult to raise revenue and as a result they are normally paid for out of general taxation. Any attempt at decentralization will have to recognize this. However, it is conceivable that 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 decentralization 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. Examples include 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, subsidization, 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 organizing or monitoring the facility.
It could be the case that several other agricultural support services assume the characteristics of club goods at the local level. This is likely to be an empirical question that depends on local circumstances. 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.
Willingness to pay
Local government, like central government, can meet demands as well as needs in the provision of goods and services. Thus farmers may have a demand for the provision of local collective goods that the private sector is unable to supply and be willing to pay local government for this provision. This willingness to pay is an important criterion in assessing if an effective demand for a good or service exists. However, the willingness-to-pay criterion will fail to identify priority needs and local officials and politicians will have to make judgements on these themselves.
Subsidiarity and specialization
Given the ability to disaggregate the components of provision, central government has a choice of continuing to produce a service itself or passing this function to other levels of governance when considering the decentralization of service provision. The principles of subsidiarity and specialization have important implications for this choice.
Economies of size
As we saw in Section 2.3.1 one of the economic factors determining the optimal level of governance for production is the existence of economies of size in the production process. If there are substantial economies of size relative to the size of the market this will tend to reduce the extent to which decentralization of production is economically desirable. This is obviously an empirical issue that needs to be investigated in each situation. However, as an example, while it may be appropriate for a large, rich country to devolve most powers and functions associated with, say, agricultural research to the provincial government level, this might be entirely inappropriate for a small, poor country. Indeed, for some small countries the reverse might be the case and they should be seeking collaboration with neighbouring countries to solve common research problems.
Even within the production area, disaggregation into different functions is often feasible, which means that it may be possible to decentralize certain functions, such as day-to-day operations and maintenance while the capital infrastructure remains centralized. It is also important to question arguments relating to the importance of economies of size on three grounds. First, the constituency responsible for the existing arrangements may be exaggerating the importance of economies of size because they are enjoying the benefits of a monopoly situation with unnecessary costs and expenditures and the absence of competition. Second, modern technologies can often operate on a smaller scale than older technologies. Third, increased competition often squeezes out unnecessary expenditure and organizational slack and hastens the search for productivity enhancing technologies.
Another important reason for considering the decentralization of production decisions is the existence of significant variations in local demand. One of the claimed benefits for devolution is the ability of locally elected officials to identify and respond to local needs. These advantages also apply to developing partnerships with CSOs to meet local needs.
On the other hand there are often claims that local organizations have insufficient capacity to produce services effectively. Opponents of devolution may use this argument in attempts to avoid losing positions of influence and power. Instead they will favour deconcentration measures. However, in developing countries, local units of the central government are also plagued with very low capacity problems, so that a real competition for limited capacity-building resources is set in motion between those who wish to strengthen deconcentrated central government units and those who wish to strengthen local government administrations. A similar view is presented in a recent World Bank document (Box 3.1). Capacity building is indeed a prime component of decentralization programmes, irrespective of the form of decentralization that the government has opted to pursue.
Box 3.1 Problems of limited local administrative ability
Local administrative capacity is sometimes considered inadequate because bureaucratic requirements imposed by the centre are inappropriate for local decision-makers. In reality, if the appropriate requirements were assigned to each level of government according to the information required for them to perform their unique function, capacity may not be so problematic.
Source: Litvack et al. (1998).
Contracting out or outsourcing
Any organization responsible for providing services must have access to appropriate administrative skills regarding legal, personnel, procurement, and financial matters. However, it is conceivable that at least some of these skills can be obtained from external professional service providers. For example, private professionals can be recruited on a job-by-job basis to assist with project design, tendering, contracting, construction supervision, certification of contractors invoices for payments, etc. Similarly, relatively small organizations can contract out book-keeping and financial reporting services to private accountants/auditors, providing that agreed procedures are followed. But do these private service providers exist in rural areas? In some places the answer is certainly no, a typical case of market failure. However, private professionals may be operating in medium size towns not too far away, and could be interested in collaborations that do not require full time employment. It is likely that more will be forthcoming if the public organizations provide a market to expand their enterprises, rather than preventing their development by the crowding out effect resulting from recruiting their own capacity. Examples of contracting out in Cape Verde and Mali are given in Box 3.2.
Several factors have to be weighed up in considering this issue. For example, private professionals normally want to be paid higher rates than public servants. On the other hand, they are paid only for the work actually done, they take care of their own social security, and can be substituted if they do not perform well. Employment of private professionals may result in higher or lower administration costs, but it is their cost effectiveness that needs to be considered. Moreover, it is still necessary to control contracted out services and internal management must establish what is the minimum of administrative skill that an organization must retain to be able to do this.
Box 3.2 Contracting out professional services
In Cape Verde, the USAID assistance is channelled through ASDI, an American international NGO. ASDI has promoted a network of 140 groups of rural people in several islands, who are interested in undertaking projects of various kinds based on labour intensive technologies. Groups have a minimum of organization, including bookkeeping with a simple method required by ASDI. Every year, ASDI undertakes a survey to estimate the cost of the basic tasks required to implement the group projects. Many unit cost estimates are lower than the rates applied by the government for similar tasks. The list is distributed to all groups demanding finance for their projects. Local private technicians are recruited by ASDI to provide technical assistance for project preparation, to supervise implementation and to certify expenditure. Private accountants are recruited in Praia to audit regularly the groups accounts.
Each group sets its own wage for members performing project implementation tasks and the number of days paid to perform each task. As a result, the total cost of the project may be in practice higher or lower than estimated at the time finance is granted. The ASDI experience is that most groups manage to save on the original cost estimate. The surplus can be distributed to group members as a bonus, or deposited in a group fund to develop other activities. ASDI believes these deposits may be used in the future to guarantee loans supporting income generating activities.
Another example is the development of small civil works enterprises in the Office du Niger area in Mali. The maintenance of the irrigation system and, to a lesser extent, some of the large scale rehabilitation works provide a market for the tacherons, artisans capable of constructing simple cement structures and of handling earth moving jobs with manual or simple mechanical equipment. In 1996 as many as 20 such enterprises were active in the area. They are family business started by a single worker, but many of them have managed to get their younger boys trained as quantity surveyors or as accountants to help in the business and raise the standards of services they can provide.
Source: R. Pantanali (personal communication).
Technical skills are less necessary in decentralized organizations because technical know-how lends itself more easily to external recruitment practices than administration skills. Local governments and CSOs may contract out construction and maintenance and, sometimes, even the operation of service facilities in areas such as water supplies, irrigation and rural roads. Leaders of interest groups at community level can be trained to acquire the skills necessary to run their own sponsored projects, which are normally of a size and complexity within reach of the proponents. Indeed, one of the objectives of devolution is to tailor the technical characteristics of community and local government projects to what local people can implement. Once again, this raises the issue of rules and standards, since in many cases the higher standards imposed by central administrations are a cause of inadequate implementation and of incapacity to maintain facilities properly. Naturally, the less developed a rural area is, the more difficult the task of contracting private service suppliers would be. In very isolated areas local governments may have to rely entirely on their own resources or on the assistance of central government deconcentrated units. In the latter case, the tasks of the deconcentrated units must be modified to reflect the objectives of government policy.
Contracts and governance
One of the advantages of contracting out is that the discipline of private sector incentive systems - the ability to reward good performance and penalize poor performance - can be brought into play. There are, however, two major reasons why it may prove difficult to assign production of all services to the private sector. One relates to problems of regulation and supervision and is discussed in the next Section. The other relates to the difficulty of drawing up appropriate contracts with the private sector. We saw in Chapter 2 that risk and uncertainty may make the private sector unwilling or unable to provide a service, leading to an incomplete market. In theory it should be possible for the government to draw up a contract with a private sector organization to produce a service and to financially compensate them for expected risks and losses. In practice, a whole range of principal-agent problems arising from asymmetric information and opportunistic behaviour create severe difficulties in implementing this approach. The problems of obtaining sufficient accurate information to ensure that the rate of return to the organization is reasonable but not excessive, or of drawing up a contract that takes account of all possible eventualities, can be so complex that it is cheaper for the government to continue to provide the service itself.
Similar contractual problems may arise with private sector provision of merit goods and the provision of other goods and services where the government intervenes on grounds of income redistribution. For example, the contracts would have to specify in some detail which groups were allowed to obtain the good or service and how much they were entitled to.
3.3.1 Rules, regulations and institutions
A few minutes reflection will soon lead to the conclusion that all economic transactions have to be governed by some formal or informal set of rules and regulations that define the conditions under which trade and exchange can take place. To give just one example, it would be very difficult and risky to purchase goods and services if property rights defining their ownership were not clearly specified. Would you be willing to buy a house if someone else could come along and claim it belonged to them?
The New Institutional Economics
The New Institutional Economics is concerned with this whole area of institutions, or rules and conventions and how they shape, and are shaped by, economic transactions and behaviour. Box 3.3 contains some of definitions of institutions.
Box 3.3 Some definitions of institutions
Institutions are the rules of the game in a society or, more formally, are the humanly devised constraints that shape human interaction. In consequence they structure incentives in human exchange, whether political, social, or economic (p.1). (North, 1990).
a public system of rules that define the kinds of exchanges that can occur among individuals and that structure their incentives in exchange. Economic institutions include markets and property rights, systems of land and animal tenure, obligations of mutual insurance within lineage groups, and other systems of exchange that are determined by implicit contracts or social norms (Hoff et al, 1993).
the set of rules actually used (the working rules or rules-in-use) by a set of individuals to organise repetitive activities that produce outcomes affecting those individuals and potentially affecting others (p.19). Working rules are those actually used, monitored and enforced when individuals or groups make choices about their actions. To be effective, all the people affected by them must know of their existence, expect others to monitor behaviour with respect to these rules, and anticipate sanctions for non-conformance. Enforcement may be by the decision-makers themselves, by agents they employ, or by an external (regulatory) agency. Some of the functions of working rules are to determine who is eligible to make decisions in some arena, what actions are allowed or constrained, what procedures must be followed, and what costs and payoffs will be assigned to individuals as a result of their actions (p.19). (Ostrom, 1992).
It can be seen that institutions consist of both formal and informal rules. Examples of formal rules are laws, statutes, international agreements and contracts. Examples of informal rules are social conventions and codes of behaviour. The working rules, to which Ostrom refers (see Box 3.3) may, or may not, closely resemble formal laws as found in national legislation, administrative regulations and court decisions. Ostrom states that a system that is governed by rule of law is one in which formal laws and working rules are closely aligned and enforced. However, there are numerous instances where working rules may assign de facto rights and duties that are contrary to the de jure rights and duties of a formal legal system. For instance, the formal rules of irrigation agencies may forbid the acceptance of bribes from farmers for delivering water, but the practice may be so routine that the exact price for various types of services is well known to all farmers and most officials (Ostrom, 1992).
It is very important that the policy analyst is aware of this reality. However, discovering what working rules are in use may require patient and long-term observation - the rapid rural appraisal approach is unlikely to work here. Ostrom cautions that the difficulty of observing institutions can result in two errors. The first is to assume that rules-in-use are identical to formal laws. This leads to misplaced confidence in the effectiveness of changing behaviour merely by changing formal law. The second is to assume that no institutions exist except for those formally created through governmental actions. In this latter case, actions may be taken that actually destroy existing institutions that may have been functioning effectively.
Reducing transaction costs
The major role of economic institutions in a society is to reduce uncertainty by establishing a stable structure for human interactions and economic transactions. Appropriate institutions will reduce the information and transactions costs of economic activity. Regulations can also reduce the excesses of opportunistic behaviour that arise from the presence of asymmetric information. Equally, certain institutions can increase informational problems and increase the costs of economic transactions. The more that policy analysts understand of the roles, forms and functions of institutions the greater is the likelihood that they can give appropriate advice on how to improve the performance of economic transactions in the presence of missing or incomplete markets. In some cases this will result from suggesting institutions that can be adopted by existing organizations. In other cases it may require the encouragement of new organizational forms that can then craft appropriate institutions.
3.3.2 Decentralization and regulation
Law and order
Laws, rules and regulations have many of the characteristics of public goods. Indeed, the national legal framework can be thought of as the ultimate public good with the government creating the framework and developing a set of procedures for detecting deviations from these rules and regulations (a policing system) and for enforcing compliance with them (a judiciary system). Decentralization may bring about the need for new institutional arrangements to govern those aspects of provision that are decentralized. The principle of subsidiarity applies equally well to the regulatory function as to the other components of provision. The power to set and monitor rules and regulations can be retained by central government, or delegated or devolved to lower levels of governance including the possibility of self regulation by organizations or groups of organizations in the private and voluntary sectors. Central government may want to retain a regulatory role where it wishes to see a specific policy or standard of service implemented over the entire country or where it is playing a major role in financing a service.
The regulatory burden
A major problem that arises when provision is transferred to the private sector is that the contracts have to be supervised or monitored to ensure that they are effectively fulfilled. A similar need for monitoring would apply if the private sector were to take over the provision of goods and services where natural monopoly conditions apply or where competition is weak. Thus the regulatory burden on the government is likely to increase as the private sector assumes more responsibility. This monitoring and regulation can only be done effectively where there is strong governance. Paradoxically, where governance is weak it may be more appropriate for the government to continue to provide those goods or services where provision cannot be left to the market mechanism. The same consideration applies when central government delegates functions to other public agencies or when local government contracts out activities.
There are other problems associated with government regulation. There are considerable information costs associated with many types of regulation. Those being regulated often hold much of the information required, rather than the regulators, that is, there is asymmetric information. When this is severe, capture of the regulators can occur with those being regulated manipulating the information supplied, to suit their own ends. Again, this is likely to be more of a problem where governance is weak and is another factor that has to be carefully considered when deciding whether, and/or how, decentralization of services should occur.
There may also be significant indirect costs associated with regulation as it generates rent-seeking activities. Highly regulated markets also tend to become inflexible. There are two reasons for this (FAO, 1991a). First, where large rents result from regulation, there will inevitably be those with vested interests who wish to keep the regulations in their initial form, regardless of changing economic circumstances. They may only be willing to forgo these favours if there are other benefits that offset the rent reductions resulting from changes in the regulations. Second, when regulations are imposed by central government common rules are often set for the whole country regardless of local circumstances. Moreover, when economic conditions change there may be no cheap and easy method of modifying the regulations. In some countries the state will have great difficulty in enforcing the regulations except in particular areas or with certain (ethnic) groups, or on a random basis. The net result is an increase in the risk costs associated with transactions and a decrease in the credibility of the state.
In this case, a possible solution is to encourage a more decentralized regulatory system with greater autonomy and thus more flexibility to accommodate to local conditions. However, to be effective this requires the regulators at the local level to have sufficient powers and credibility to enforce the regulations.
If governments try to regulate too rigorously, the result may be a transfer of activity from the formal, regulated economy to the illegal, unregulated parallel market. This will occur when the costs of compliance outweigh the risks of operating illegally. If this occurs generally it may ultimately undermine the governments authority and credibility. Thus governments should have an incentive to restrict their regulation to areas where they can hope for a credible degree of enforcement.
One other issue that has to be considered is who is going to finance the regulatory functions after decentralization. While central government may wish to pass the responsibility of regulation to local government, the local authorities may find it difficult to finance the regulatory functions and some form of IGFT may have to be organized. On the other hand, we have argued that many services that initially appear to be public goods at the national level, assume the characteristics of club goods at the local level. In these circumstances it becomes easier to recover the costs of regulation from those being regulated, or to encourage them to partially or wholly self-regulate their activities.
The decision to decentralize the provision of agricultural services raises many issues and offers the opportunity to question many previously held views. It has been argued that decentralization need not apply to all components of provision and that different forms of decentralization can be used for the various functions. It is also apparent that the efficacy of decentralization is strongly influenced by the capacity of local government and the strength of local governance systems.
In terms of financing provision, deconcentration involves only a change in budgeting and release of funds procedures within a unitary administration. However, delegation, devolution and partnerships involve more fundamental changes. Given the weak revenue raising capacity of many local governments, IGFTs may be required if public financing of services is to be maintained. However, decentralization offers a fresh opportunity to explore the introduction of cost recovery measures.
The extent and means of decentralizing production is influenced by a variety of factors. The existence of large economies of size may limit the extent of decentralization, but significant variations in the nature of local demand may increase the advantages of decentralization. The limited capacity of local government to cope with the production of services may be a constraint but it is possible that this can be overcome by contracting out certain functions. However, this is unlikely to offer a complete solution.
Strength of governance is a major factor in considering the most effective way of regulating services. Indeed, there are circumstances where the regulatory burden and the abuse of regulatory mechanisms may make effective decentralization of certain services difficult.
 In the short run a
seller may be willing to sell at a price that covers the variable costs of
producing that unit and in the longer term to continue production whilst
variable and current fixed costs are covered. However, it is not rational to
continuously sell at a price that does not cover the full replacement cost of
both variable and fixed resources.|
 Economic profits being defined as revenue in excess of the full (opportunity) cost of production.
 This 'equilibrium' price is a socially efficient price in the sense that the benefit that the satisfaction-maximizing consumer derives from the last unit just equals the price paid for it (if benefits exceeded price the rational consumer would purchase more of the good). Equally, the price received by the profit-maximizing producer will just equal the extra costs to the producer (and to society if there are no externalities) of producing that last unit, otherwise increases or decreases in production are justified.
 Private rights are not the only form of property rights. Rights may be 'communal' where the right is held in common with another person or group of people, or 'collective' when the decision about the service is taken as a group (Ricketts, 1987).
 The expert reader may recognize that things are a bit more complex. However the simplification helps to make the basic concepts easier to understand to non-specialists.
 Strictly speaking, a socially efficient pricing system requires price to be equated with full marginal cost. The full average cost is a socially efficient price if marginal (social) cost is also equal to average cost and hence price. However, in situations such as natural monopolies this is not the case and setting price at marginal cost would result in financial losses.
 A profit maximizing company uses the concept of 'opportunity' cost, that is, what a resource could earn in its best alternative use. In this way a rate of return on capital can be built into cost calculations. An appropriate risk premium is similarly built into costs. Are these approaches appropriate for a government?
 See Section 2.1.
 These problems stem from the fact that information has value and is costly to acquire. It is often acquired most cheaply as a by-product of production and consumption activities in which individuals are engaged. Thus individuals who frequently engage in an activity, and/or are relatively wealthy are likely to have access to much more information than infrequent users or poorer individuals (for example, large traders buying and selling daily compared with small farmers only selling small quantities occasionally). This generates asymmetric information where one individual, or group of participants, in a transaction has better access to information and hence greater bargaining power than others have. This problem of asymmetric information is reinforced when it is appreciated that there is a large 'fixed cost' element associated with acquiring and analysing information. Thus the average cost of information will tend to be higher for the small transactor, other things being equal. When this information advantage is used with guile or deceit to gain a benefit at the expense of another party to a transaction it is termed opportunistic behaviour.
 Recall the difference between 'institutions' and 'organizations' discussed in Section 1.5.
 FAO (1991a).
 'Rent seeking' is the use of resources in an attempt to gain control over artificially created monopoly situations.