Previous Page Table of Contents Next Page

Chapter 2: Factors influencing the decentralization decision


The objective of this chapter is to examine some of the main theoretical and practical issues relating to the decentralization decision. The various economic reasons why governments may wish to be involved in the provision of agricultural services will be the first issue to be explored. The nature of ‘provision’ will then be examined. This will be followed by a discussion of three of the main principles determining the form and extent of decentralization, namely subsidiarity, specialization and jurisdictional spillover. The practical issues examined mainly relate to organizational factors affecting the outcome of institutional reform.


2.1 Traditional roles of government in service provision

‘Market failure’

The institutional reforms carried out in recent decades to the provision of agricultural services have focused attention on justifications for government involvement in this area. One of the main reasons for government intervention has been the perceived presence of ‘market failure’ in various forms. This term is used to describe situations where the unfettered use of the market system or the private sector to allocate goods or services would not lead to a economically efficient[19] outcome where the price of the good adequately reflects the (opportunity) costs to society of producing it.

Excludability and rivalrousness

One source of market failure is low excludability[20] and low rivalrousness[21]. These features characterise pure public goods such as the law, the rules and regulations established by public agencies, and the services provided by the national defence system, the police, the judiciary system, and agricultural inspection agencies. These are typically provided by the government and paid for out of taxation as they potentially benefit all members of the community and ‘free riding’ makes it difficult to charge users directly for these services.

Classification of goods

However, for many agricultural services the degree of excludability or rivalrousness is often determined by the precise nature of the service and the conditions under which it is delivered. Thus similar services, such as extension advice, may be delivered by the private sector in some situations but can only be provided efficiently by the public sector in others. Some services, termed toll or club goods[22], are non-rivalrous (at least up to the point where capacity constraints may influence the marginal cost of further provision) but they are excludable. The private sector or civil society organizations supply many toll goods and charge users for them. However, if there is a monopoly provider there may be a case for some form of government intervention in their provision to prevent prices being charged that are higher than those justified on economically efficient grounds. The term ‘common property goods’ refers to rivalrous goods where control is in the hands of the state, a group of individuals or CSOs that can control access to the good. This often applies to grazing land where customary law defines who can share the grazing rights in land. As we will see in Chapter 11 it also often applies to water supplies where it is important to ensure adequate access to all parties. At the other extreme end of this spectrum are pure private goods that are both excludable and fully rivalrous. Fertilizer has both these characteristics and is an example of the typical goods that can be provided efficiently by the private sector and exchanged in the market[23].

Externalities or spillovers

The presence of externalities or spillovers[24] is another example of ‘market failure’ that may also justify government intervention. Where the externalities are associated with what are typically considered to be private goods, governments often intervene in the provision of the good through regulatory or fiscal instruments to bring production and consumption more in line with a economically efficient outcome. Direct controls on production, taxes and subsidies and the creation of property rights and markets in externalities[25] are the main instruments used for this purpose. However, where governance is weak and government regulations or fiscal instruments would be ignored, a government may feel that it has no option but to physically produce the product itself.


‘Market failure’ also occurs in the presence of monopolies where there is a strong likelihood that prices will be held above the costs of production and economic profits generated. Some monopoly situations actually arise from government actions such as the exclusive right granted by a government to a public or private sector agent to purchase agricultural products from farmers. Other monopolistic tendencies result from control of an exclusive source of supply (for example, a water well or the foundation material of hybrid seeds), from market fragmentation, economies of scale, poor infrastructure and lack of credit. Governments can reduce these tendencies by adopting policies that encourage competition and free trade and ensure that artificial barriers to entry are removed for domestic and foreign suppliers. However, in the case of natural monopolies[26], where economic efficiency requires price to be set below the average cost of production but unregulated monopolies are tempted to set a price above the full costs of production, governments are likely to assume responsibility for producing the good or service themselves. The alternative is for government to regulate the price and/or output of private sector provision.

Another source of ‘market failure’ is missing or incomplete markets stemming from imperfect information, risk, and poor transport and communications infrastructure. All of these particularly affect agricultural service provision to remote areas and may make it prohibitively expensive for the private sector to provide some types of goods and services.

Merit goods

So far we have considered the various types of market failure that may justify some form of government intervention to achieve economic efficiency. But another possible justification for intervention in provision relates to ‘merit goods’. These are goods or services that governments mandate individuals to consume, either to protect themselves from others or themselves where it is judged that people, left to their own devices, may not act in their own best interests[27]. This paternalistic reason for intervention, when a government intervenes because it claims to know what is in the best interest of individuals better than they themselves do, should not be confused with government intervention to correct externalities.

Redistribution of income

Even if an economy is economically efficient it does not mean that the allocation of goods it generates is consistent with the government’s objectives. For example, the government may wish to intervene by providing safety nets to ensure that people who are temporarily or permanently disadvantaged have access to minimum levels of certain goods and services. Equally, if a government’s main objective is to remain in power it may intervene in the provision of goods and services to increase the welfare of its potential supporters. If re-election depends on the votes of urban consumers rather than those who live in rural areas then one may expect to see increased provision of (subsidized) goods and services in urban areas and the relative neglect, or even increased taxation, of goods and services for rural areas. However, if a government is more concerned with reducing inequalities in income or in poverty alleviation then its interventions in the provision of goods and services may be entirely different, particularly if poverty is mainly a rural phenomenon.

2.2 Disaggregating ‘provision’

In the past governments have often reacted to market failure problems by undertaking the provision of the service themselves[28]. However, as will be stressed throughout this Sourcebook, one of the bases of the institutional reforms carried out in recent decades is the appreciation that the provision of any good or service can be disaggregated into four components:

Components can be provided by different agents

A crucial factor influencing the decentralization decision, and a point that will be repeated throughout this Sourcebook, is that there is no necessity for these four components to be provided by the same sector or organization or even level of government. For example, there might be some types of services that the state may decide should be regulated at the central level, but which local government might finance. 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. Equally, with merit goods that are also private goods, there is no reason why government should be concerned with their production if it can achieve its objectives through financing and regulating their provision. However, a different configuration might apply to every type of good or service.

2.3 Subsidiarity, jurisdictional spillover and specialization

2.3.1 Definitions

In Section 2.1 we stated that there are three main economic reasons - market failure, merit goods and redistribution of income - that are used to justify government intervention in the provision of goods and service. In similar fashion, there are three main general principles - subsidiarity, jurisdictional spillover and specialization,- that provide guidance on the level of government most appropriate for the provision of different components of various goods and services. These concepts are defined in Box 2.1.


There are various factors that influence the degree of subsidiarity that is likely to occur. One is the presence of economies of size in the provision of a service relative to the size of a country. If there are substantial economies of size then decentralization and the provision of the service in smaller units would lead to increased average costs and this is likely to reduce the extent of subsidiarity. Another factor is central government’s desire to achieve certain policy objectives, for example equity. In this case there will be a greater tendency towards centralized control over the allocation of resources and even the selection of target groups for the receipt of services. Any decentralization is likely to be through deconcentration or delegation that allows central government to maintain control, although conditional grants may be used with devolved local government to achieve certain objectives. On the other hand, the degree of subsidiarity is likely to be increased where it is recognized that services need to be tailored to local requirements.

Box 2.1 Subsidiarity, specialization and jurisdictional spillover


The principle whereby administrative responsibilities are assigned to the lowest level of governance capable of carrying out these responsibilities competently.

Jurisdictional spillover

The principle whereby a service should be controlled and financed at the jurisdictional level where there are no spillovers or externalities (see Section 2.1) on neighbouring jurisdictions.


The principle whereby limits are imposed on the domain within which an organization has authority to assume responsibility for autonomous decision-making. By ‘domain’ is meant the agendas for which an organization has responsibility to take independent decisions and the means to implement them.

The relevance of the principle of subsidiarity to the ability to disaggregate provision into its various components should be immediately apparent. It can help to determine the decision as to which functions are best provided by central government, by different levels of local government, by CSOs or the private sector. Other examples may also help to illustrate the application of the subsidiarity principle. In countries where overall water supply is the responsibility of a national organization, subsidiarity may indicate that autonomous operational responsibility for supplying specific townships or locations be entrusted to local offices. These would be better placed to serve the location, take routine operational decisions, undertake maintenance, and quickly intervene in case of breakdowns. The subsidiarity principle may require that a large irrigation scheme be subdivided into sections which can be managed in a unitary way, with operation and maintenance being entrusted to autonomous units, separate from the agency which is responsible for the parts of the scheme which serve all the sections.

Jurisdictional spillover

The impact of externalities can vary enormously in terms of the geographical area or number of people affected. This is an important consideration in decentralization because economically-efficient delivery implies that the service should be controlled and financed at that scale where there are no spillover effects. This reduces the extent of ‘free riding’. Thus, for example, a service should only be controlled and financed at the district level if there are no significant spillovers into surrounding districts. For example an access track to a single village, or a micro-irrigation project whose beneficiaries are members of a single community, or a health dispensary that serves only one village, is a community affair. A feeder road linking several villages or an irrigation scheme involving an area controlled by several villages or a health centre serving several villages, is a municipal or sub-district affair because the impact of these initiatives spills over the territorial boundary of a single community.


Specialization refers to the domain of an organization. Each level of governance is defined by its domain, that is by the agendas for which an organization has responsibility to take independent decisions and means to implement them. The domain may refer to:

Defining agendas according to the principle of specialization may be difficult. Local governments are not specialized organizations; they deal with a wide range of subjects. CSOs are highly specialized at grassroots level, but their higher level organizations are often involved in several lines of activities of a social and productive nature.

However, some domains seem to be easily defined according to the principle of specialization. For example, primary schools are often the responsibility of municipalities, middle schools of the provinces (districts), higher education of the regions, and universities of the central government. This division of responsibilities is often associated with a degree of deconcentration of a ministerial administration. However, primary school teachers are often on the payroll of municipalities, and, under devolution, local governments can construct new primary schools if they wish to use the resources allocated to them in this way. Would this authority modify the relationships between the primary school administrations and the Ministry of Education? Similar relationships apply to other services, such as health and water supplies.

There are similar problems relating to domains when responsibilities are devolved to local governments from line ministries. For example, to what extent does the transfer of responsibilities from line ministries responsible for various aspects of rural development, such as promoting education, agricultural production, health services, transport infrastructure, water supplies and communications also involve a transfer of authority to set independently the rules and standards under which services are provided? If central government rules and standards must be respected, what procedures should central government apply in order to enforce respect of these standards? To what extent would those procedures interfere with the detailed allocation of resources by local government to the extent that their freedom of initiative is jeopardized?

When clearly identifiable general interests are involved, such as utilization of scarce national resources (water for example), or prevention of epidemic and contagious diseases (with respect to animal and human health), there is little doubt that it is the central government’s responsibility to set the rules and devolved organizations must comply. Devolution of authority to perform a function would include the obligation to respect higher authority ruling in those matters and this establishes a boundary of the domain. But what about technical and cost specifications of services provided where questions of general interests are not evident or remote?

Equal treatment

If the central administration has equity as one of its major objectives, a further guideline for the provision of services financed by them is provided by the principle of ‘equal treatment to equal levels’. Some examples may explain the implications of this principle:

2.3.2 Forms of decentralization revisited

We can now briefly consider how the principles outlined above influence different forms of decentralization.


Deconcentration assigns specific functions and tasks performed by the staff of the headquarters of central administrations to staff posted in peripheral locations within the national territory. Staff, equipment, vehicles, and budgetary resources are transferred to units such as regional and district offices, and the managers of these units are given authority for autonomous decision making regarding the operations, which were previously taken at headquarters, or needed clearance from headquarters. In accordance with the principle of subsidiarity, deconcentration follows the good management practice that operational decision making authority ought to be entrusted to the level of management that directly commands all information needed to take that decision. Deconcentration is expected to increase the efficiency of the administration. The principle of specialization applies to a limited extent with reference to the territory that a deconcentrated unit is responsible for. The staff of deconcentrated units is central government staff, acting under the supervision of the central administration. Front-line managers have no authority to take independent initiatives as to what should be done: planning, programming, budgeting and release of resources is firmly under central control.


Delegation transfers responsibility for implementing specific tasks and delivering certain services from the central administration to independent organizations operating outside the public administration. Examples of delegation include: a national water supply company that may be entrusted with the responsibility to plan, construct, and operate water supply schemes (over a certain size) across the country; a water basin development authority; an agricultural research institute; a strategic grain reserve; a project management unit. Delegation involves both the principle of subsidiarity and that of specialization. Delegated agencies may be in the public or private sector. Delegation involves changes in the fiscal and political dimensions, but does not affect the central government function of setting goals, specifying the strategy to adopt and the specific objectives to be achieved or approving the detailed budget of the organization receiving delegated authority. However, delegated agents have considerable scope for autonomous implementation, particularly with regard to the way delegated tasks are implemented, including, in several cases, personnel and procurement policies. They also carry responsibility for the results. The difference between delegation and employing contractors is the degree of autonomy in interpreting the task. For example, a contractor must build a bridge where he is told by the Public Works Department (PWD), and according to the specifications of the PWD. On the other hand, a national water supply company decides independently where to build the facilities and their technical specifications.


Devolution is a more advanced form of decentralization. Devolution turns local governments from being agents of the central administration into autonomous suppliers of goods and services. It involves the transfer of responsibilities, authority, assets, and financial resources to lower levels of government, such as provincial or district governments in unitary states, in accordance to both the principle of subsidiarity and of specialization[30]. Devolution implies changes in the political and fiscal dimensions of government. Local governments to which authority and resources are devolved acquire the power of autonomous initiative and decision making with respect to setting their own rules, goals and objectives. They also acquire the power of elaborating and implementing their own policies and strategies, and of allocating resources to different activities within the domain assigned to them. In addition, they often are given authority to raise financial resources, through taxes, and in some cases, to borrow on the capital markets.

CSOs and partnerships

Part of the rationale for establishing a separate mechanism to decentralize support to local interest groups and communities to a lower level than local government, and outside the public administration, also rests on the principles outlined above. We have seen that the subsidiarity principle calls for mobilizing institutions to be as near as possible to the ultimate beneficiaries of services. Local governments’ capacity to meet the needs of small groups is limited and often overburdened with responsibilities related to services that must be handled at a higher level than the group/community level, because of jurisdictional spillovers.


Experience throughout the world has shown that, with a few notable exceptions, governments are not efficient producers of private and toll goods. This is because a government can rarely provide the same incentive structure as that available in the private sector to reward success and punish failure[31].

Thus with well-functioning competitive markets for goods, services and resources, the private sector and CSOs can be expected to produce and distribute private and toll goods where ‘merit’ or ‘redistribution of income’ considerations are absent. The public sector would then not incur the costs of production. Market forces would determine the price and quantity of these goods and services and potential consumers’ ability to buy them. If the government had been previously heavily involved in the provision of subsidized private goods and services the impact of government expenditure on the economy would be reduced. Moreover, the distorting effects of taxation on prices and private incomes would be less, with further beneficial effects on the efficiency of resource allocation mechanisms. As indicated earlier, even in the case of merit goods or those where redistribution of income considerations are important, it is conceivable that the private sector can produce the good or service, but the government may finance all or part of the cost and determine its allocation (consumption).

2.4 Organizational and political factors

2.4.1 The theory of power

How organizations function

The effectiveness of decentralization reforms is never automatic. A critical issue is the way that organizations react to reforms, or the threat of them. The way organizations in general, and complex organizations such as those of government in particular, set their own specific objectives, define functions and tasks of different units and interpret and implement their tasks is not related to the general goals and objectives assigned to them in any simple and straightforward manner[32].

In this section we discuss some of the major factors that determine the behaviour of the organizations, in particular the power relationships between different groups within and around organizations, that may influence the outcome of attempts at decentralization. These factors are specific to the situation of individual countries, of different organizations, regions, and localities within a country. They influence the policy formulation process and, to an even greater extent, the degree to which policy decisions are implemented in practice.

The theory of power

The theory of power in and around organizations attempts a general interpretation of the role of internal and external forces in determining the way organizations set their goals, establish their structure, develop their management dynamics, and respond in practice to the influence of the interplay of those forces. The theory recognizes internal and external coalitions of ‘people with influence’.

Internal coalitions

Within an organization, people with influence include the chief executive officer, the front line operators, the line managers, the planners/controllers/evaluators, and the support staff. In addition, another important factor of internal influence is (corporate) ideology that is also identified as the culture of the organization (see Section 2.4.3). People of influence in the different categories may get together to form an internal coalition. The internal coalition can run an organization by:

Thus the theory distinguishes the internal coalition as bureaucratic, personalized, ideological, professional and politicized, depending on the way its internal relationships are organized and on the management style.

External coalitions

People of influence around, or outside, an organization include:

These people can also get together to form an external coalition to influence the behaviour of an organization, a feature which is common to important economic enterprises as well as governments. External coalitions can be:

Power configurations

Power is the capacity to affect the behaviour, the output, and eventually the outcome of organizations. Decentralization redistributes shares of power within a society. A combination of the features of the internal and external coalitions determines the power configuration of the organization. The interaction of coalitions is illustrated in Figure 2.1 and different types of power configurations are summarized in Table 2.1.

Figure 2.1 The power configuration of organizations

The first case, which presumes a dominated external coalition and a rigidly bureaucratic internal coalition power configuration, is nearest to the assumption of a single actor and single goal situation associated with the elementary theory of the firm of neo-classical economics and the initial stages of the theory of management by objectives. Only in this case does the organization become the instrument of the dominant group/individual of the external coalition. Most economic analyses of government organizations, and many reform recipes of international advisers, implicitly assume that this is the configuration that prevails in the environment, sometimes despite evidence to the contrary.

Table 2.1 Factors affecting the power configuration of an organization

External Coalition

Internal Coalition

Power Configuration






Closed system












The political arena

The picture presented in Table 2.1 does not refer to a static situation. Indeed, conditions in the external and internal coalitions change over time. A change from a dominated to a passive external coalition, and a bureaucratic internal coalition, generates a closed system, which tends to have no goals except the organization’s own survival. A divided external coalition may generate a politicized internal coalition, resulting in an internal political arena, which risks leading to the organization being unable to work for growth and even for survival.

Different power configurations can be associated with different forms of government, as in Table 2.2. In practice, real situations will present combinations of the features presented above and different dynamics of the interplay among the various external and internal forces.

Table 2.2 Power configurations associated with different forms of government

Power Configuration

Form of Government


Colonies, overwhelming majority governments in democracies

Closed System

Communism, most large government agencies irrespective of the form of government


Dictatorships, dictatorial party leaders in democracies


Cultural revolutions

Political Arena

Revolutions, anarchy, pluralistic governments (most contemporary government agencies have features of this configuration)


The theory of power in and around organizations provides a powerful tool to assist in understanding the working of agencies, in the private, voluntary or public sector and in predicting their behaviour and performance. The analysis confirms the need to proceed with the elaboration of decentralization policies on an empirical basis. The main lesson we learn from this brief review of the principles of the theory of management is that decentralization of the government administration is likely to be easier when the external coalition is dominated and the internal coalition bureaucratic, that is when the power configuration is the ‘instrument’. In this case, however, the government machinery is also most effective, and its self-preservation forces strongest. Good reasons to justify decentralization may be hard to find, and difficult to sell to decision-makers. A dominated external coalition is necessary to design decentralization on rational bases. A divided external coalition may facilitate decentralization, but its design may not be the best, since it will reflect the different interests within the external coalition, rather than the general interests of efficiency, transparency and effectiveness of public services. A personalized internal coalition can successfully oppose decentralization if confronted with passive or divided external coalitions.

The problems are similar in any organization of a certain size and at any level of public governance. Local governments and autonomous agents of the central government are confronted with the same set of potential power configurations. A multiplicity of centres of governance will present many different situations, some conducive to better performance, others leading in the opposite direction. Under the circumstances one can easily understand why empirical evidence about the premises of decentralization is often not conclusive.

2.4.2 Politicians and bureaucrats

The role of politicians

Governments are organizations run by people with political power. By nature, the fundamental survival and development strategy of people with political power is to maintain and, if possible, to increase, their power. Political power requires the solid support of individuals and groups of people in the society. Support can be obtained in many different ways, ranging from the use of violence and monopolistic propaganda practices, to building consensus on the basis of free public debates at all levels of governance. Without solid support in the political arena, political stability, and the power of the people who run governments, are at risk. The supporters of people with power constitute a constituency, to which people with power are accountable for their action in the exercise of power. The concepts of constituency and accountability are closely linked. Accountability is the process whereby the people who represent a constituency are made to respond to the constituency for their action on its behalf.

Societies are complex and diversified. Within societies, people may differ according to economic interests, religion, culture, ethnic characters and allegiances, and many other features. Some of these features may identify specific groups as relatively homogeneous. However, other sources of differentiation overlap within such groups as well. Some groups are well organized, particularly, but not exclusively, common economic interest groups, some religious groups, and some ethnic groups. The tightness of the organization does not necessarily depend on whether the organization is formal or otherwise. The leaders of these groups often have great influence over the members of the groups and can determine the group members’ political allegiance. They therefore possess political power, independently of whether they are professional politicians themselves or not.

To organize their constituency, political leaders try to obtain support from a variety of individuals and groups of people often with different connotations within the society. They are politically accountable to their constituency, in the sense that they must satisfy the expectations of the constituency to retain its support. This is generally true, regardless of the form of government.

Democratic governance

Different forms of government provide different mechanisms to regulate the relationships between politicians in power, their constituency, and other groups of people who do not feel represented by those who currently run the government. The constituency of a dictatorial government may be the army, the police, and the party managers who control the mass organizations to the exclusion of other citizens. A democratic form of government is characterized by:

The relationship between the people in power and their constituency is generally looser the higher the level of government where the political power is exercised. At national level, consensus is achieved over matters of principles, general policy and strategy lines. In order to widen the constituency, the public political debate is often conducted in terms that are sufficiently ambiguous to accommodate different interpretations of groups with different interests and other characteristics. As a result, the national political constituency often appears to be a conglomerate of different interests, cultures, ethnic groups, etc. kept together by some common denominator, the implications of which are not always clear. Concurrently, however, specific and very concrete agreements are negotiated between influential leaders of important groups of people and the politicians wishing to obtain their support to secure the control of government. Such agreements are very seldom in the public domain and go well beyond the allocation of official positions in government such as those that result from agreements of different political parties about ministerial and sub-ministerial positions in a coalition government.

At lower levels of government, the variety of interests of different ‘homogeneous’ groups of people may or may not get organized into specific political constituencies. In a democratic local government, the political debate is generally more concrete and better focused about local issues and problems. However, the organization of political groupings is also influenced by factors other than local interests. National politics do play a role, at times a more important role than local interests. Local pressure groups may have stronger ties at local than national levels. Even when local societies are relatively homogeneous, for example from the point of view of income and wealth differentiation, conflicts of interest, such as those related to gender issues, do not easily emerge in clearly identifiable political constituencies.

Policy formulation

Policy formulation is a process that involves decisions by politicians who may or may not fully share the conclusions of the political debate at Cabinet level. Individual Cabinet members have different constituencies and represent different interests. Even at Cabinet level, governments are not monolithic organizations. They are the result of an equilibrium of different forces, some of which may be advocates of change, others of conservatism. Ideology and personal opinions may be important factors explaining such differences but the relationships of each member of Cabinet with his/her own constituency are relevant also. Decentralization changes the power of (and within) different ministries and government agencies. Some politicians may win and others lose in the process. Those who lose will try to cut their losses and oppose decentralization in various disguised ways.

The role of bureaucrats

Politicians are accountable to their supporters for the results of the activities at the level of government for which they are responsible. However, the operational responsibility for running public organizations rests with the bureaucracies. Bureaucrats operate under the instructions and supervision of politicians, and are accountable to them. Bureaucrats have a relatively stable career, guaranteed by the law, and may keep their position of responsibility for a long time, whereas politicians may change their office often. In practice, bureaucrats have power, often more power than politicians do, to influence the behaviour of public organizations. They also have power to act in such a way as to maintain and increase the power derived from their office. Bureaucrats as well as politicians must reckon with their constituency, which includes politicians in government, groups and individuals of their staff, and external people with influence.

Conflicting and ambiguous goals and objectives

Public organizations are normally confronted with conflicting goals and many tasks. Because of the complexities involved, many decentralization policy formulations and the related legislative instruments concerned with shifts of responsibilities between and within different layers of government are imprecise. Vague or ambiguous definition of goals and tasks opens the door to multiple interpretations by managers and staff, with the result that the effect of decentralization may differ according to the different interpretations. External pressure groups try to influence decision-makers to interpret goals and tasks in such a way as to undertake activities to their advantage. For example, some local wealthy traders will try to influence local governments to build the roads and markets they need. Similarly, influential farmers will apply pressure to build an irrigation project in their area, or on the land they control, irrespective of its comparative advantage. Other influential people will want to improve the water supply to their urban residential area and to the village they come from, and to higher standards that apply in other places, etc.

2.4.3 Organizational culture

The behaviour of organizations, and the results of organizations’ activities, depends to a large extent on the way tasks are performed. However, policy directives per se do not effectively influence this. For example, even when farmers’ participation is emphasized as an essential objective of government agricultural extension policy, the tasks assigned to the extension service can be implemented in different ways. Different relationships between agents and farmers can be established, more or less attention can be paid to poor farmers’ or women’s problems, stronger or looser links with research and with the CSOs can be established. The way that organizations implement their tasks is determined by the culture of the organization (Box 2.2), policy directives notwithstanding.

Box 2.2 The culture of organizations

“Every organization has a culture, that is a persistent patterned way of thinking about the central tasks of, and the human relationship within, an organization.”

Organizations may have several cultures, possibly in conflict with one another. Cultures are very evident in craft organizations. Two examples may illustrate this. First, the Public Works Department (PWD) is made up of engineers who want to build roads and bridges. They perceive this as the core task of the organization. They also have a distinct appreciation of their professional values, and tend to consider other functions in the organization as subsidiary and of lower priority. When this is the case, ‘engineering’ is the dominant culture of the PWD. Second, an NGO or a section of the Extension Department of the Ministry of Agriculture has adopted participatory rapid appraisal (PRA) as their professional approach. Field staff firmly believe that understanding farming systems and farmers’ survival strategies and assisting them to respond to their needs to improve livelihood is the core task of their organization, all other functions being less important. This establishes a uniform dominant culture, combining features of professionalism and ideology.

Some organization cultures may have negative effects. For example, when government officials interpret the core task of their office essentially as the ‘defence’ of government prerogatives, they apply laws and regulations rigidly without regard to real circumstances, and feel they are to serve the state rather than the public. This culture may have catastrophic consequences on private initiatives, and even on the administration of justice.

An agency’s culture is largely the product of the predisposition of its members, the technology they use, interest groups’ pressure, and the immediate situational imperatives they have to cope with. When these factors cause different interpretations of the agency’s core tasks for different groups of people (units), the organization will have multiple cultures (Wilson, 1988). Multiple cultures are evident, for example, by comparing attitudes in personnel departments and in operating departments of most organizations. Another example may be the different approach of extension field workers interested in spreading pre-fabricated technologies available ‘on the shelf’, and that of other workers keen on continuous participatory diagnosis of farmers’ problems.

Craft organizations

Managers play a fundamental role in establishing the culture of an organization. This function is particularly important in most organizations responsible for rural development, which are craft organizations (Wilson, 1988), like the corps of engineers, the forest service and the extension services. In these organizations, staff output cannot be meaningfully observed, since staff operate in the field on their own while performing their duties, but outcome (impact) can be measured. These organizations must be managed by creating a sense of purpose, allowing front-line operators to find their way of performing tasks, and letting professional norms and peers’ expectations play their role. Managers of decentralized organizations may establish different organizational cultures, some conducive to government decentralization objectives, some pulling in the opposite direction, with the result that the policies are implemented in various ways and results may be quite different from expectations.

The incentive system

The success of institutional reform also depends on establishing a system of incentives (Box 2.3) that helps to overcome some of the principal-agent problems that can bedevil large organizations and bureaucracies and brings about, in the most efficient manner, the production, allocation and consumption of goods and services in line with people’s preferences. One of the advantages claimed for the effectiveness of private sector organizations is that the incentives and penalties generated by the market system, in the form of profits and the possibility of bankruptcy for owners and bonuses and the threat of dismissal for employees, imposes a requirement for efficiency that is missing in many public sector organizations. Other types of rewards and penalties, however, are also important for the proper functioning of the economic and political system of a country. A special role in the system of incentives (or dis-incentives) is that of ‘bureaucratic procedures’[33], which may suffocate private initiative and discourage private enterprise. This is a matter of direct concern to decentralization policy designers.

Decentralization of the public administration is expected to improve the system of incentives, which confronts suppliers of goods and services where provision is not dictated entirely by market forces. The closer the administration of the service is to the clients, the greater the likelihood that decisions about which services should be provided, how much, where, and to whom, will be more responsive to the demands of users. It is expected that the quality of service will improve, and that the efficiency of the provider will increase as a result of closer accountability to clients.

Box 2.3 The system of incentives

The system of incentives

is the set of rewards and penalties (financial, social, political or administrative), which governs the demand and the production and distribution of goods and services

Rewards: prices, profit, power, status, promotion, winning the elections

Penalties: costs, bankruptcy, shame, demotion, marginalization, transfer to posts with no career prospects, losing the elections

2.5 Conclusions

There are three major reasons why governments may become involved in the provision of agricultural services and these relate to market failure, merit goods and the redistribution of income. However, it is important to distinguish between provision and production. Production is only one of the components of provision, the others being financing, regulation and consumption. A crucial factor influencing the decentralization decision is that there is no necessity for these four components to be provided by the same sector or organization or even level of government.

This is clearly extremely important in applying the principle of subsidiarity to deciding which functions are best performed by the different levels and forms of governance. The extent of any jurisdictional spillover is also of importance in determining the appropriate level of governance for each function. Applying the principle of specialization to define the agendas of various levels of governance can be difficult and this is indicative of the many problems to be faced in any decentralization reform.

Decentralization requires organizations, and the relationships between them, to be modified. To do this successfully requires an understanding of organizational culture and power configurations and the motivations and roles of politicians and bureaucrats in government. The success of institutional reform also depends on establishing an effective incentive system to help overcome some of the principal-agent problems that tend to adversely affect public sector organizations.

[19] In economics textbooks this would be referred to as a 'Pareto efficient' market where it is impossible to make one member of society better off without making someone else worse off
[20] Low excludability means that it may be difficult to exclude people from 'free riding' and enjoying the benefits of goods and services even if they have not paid towards their provision. Producers would find it difficult to recoup the full costs of their provision and, from a economic efficiency viewpoint, would thus tend to under-produce such goods.
[21] Low rivalrousness means that one person's consumption of the good does not reduces its availability to others. As the cost to society of additional consumers enjoying the benefits of pure public goods is zero, economic efficiency requires their price to be set at zero. As a result it would not be profitable for the private sector to attempt to sell these goods.
[22] The term 'toll good' is used because efficiency conditions can be established justifying charging a toll for such a good. The alternative term 'club good' is used because the nature of most clubs is that members can share the benefits, but non-members can be excluded. An example of an agricultural service that can be classified as a toll good is the use of cattle dip facilities. The use of the dip by somebody does not prevent others using it (providing there is spare capacity for additional users) but those who are not prepared to pay the price can be excluded from using it even though this is not economically efficient.
[23] Even so, there are exceptions. See the later sections on natural monopolies and merit goods.
[24] An externality or spillover exists whenever the production or consumption decisions of one individual unintentionally impact on the production or consumption decisions of others in some way other than through the market. There are many examples of negative externalities in agricultural production and marketing such as the careless use of pesticides and the failure to control certain types of pests and diseases. There are also many examples of positive externalities, one being the development of new crop and husbandry practices or new plant varieties if these can be copied or used by other farmers without paying for them.
[25] The recent discussions on tradeable emission targets are an example of this approach.
[26] A natural monopoly results from situations where average costs of production continuously fall as output expands. This means that one organization can supply a service more cheaply than two or more organizations. It also means that the cost of extra production, which determines the economically efficient price, is below average cost.
[27] This is the definition used by Stiglitz (1997) Some authors use a broader definition that embraces goods and services that are included under 'redistribution of income' goods in this Sourcebook.
[28] There is a lot that individuals or CSOs can do themselves to overcome or mitigate the effects of market failure. The New Institutional Economics (NIE) examines how institutions affect the working of imperfect markets, and seeks to provide guidance as to modifications of the institutions which may help to overcome the impediments to well functioning markets.
[29] Ross (1988) uses this disaggregation. Ostrom (1990) uses a similar breakdown but with different terminology.
[30] State governments in federal states have much independence of their own, sanctioned by constitutional law.
[31] The hierarchical structure of governments and bureaucracies creates situations of asymmetric information. Typically, bureaucracies operate by one party (the principal) delegating actions to another party (the agent). Usually, the agent who performs the task has the information advantage. The principal then has to provide incentives to the agent to encourage him/her to work in the principal's interest. If the government's terms of employment mean that it is difficult to give adequate incentives (e.g. low salaries, poor promotion prospects, difficult to fire workers for poor performance) then the performance of the whole government and bureaucracy is likely to suffer. The whole class of issues arising from asymmetric information is termed the principal-agent or agency problem.
[32] The key arguments of this section are derived from Mintzberg (1983). Mintzberg's theory represents a persuasive and relatively simple explanation of the forces that are at play in determining the current behaviour as well as the history and development of organizations. The author's approach is particularly useful in appreciating the complex reality of public governance institutions. Mintzberg's contribution focuses on the 'political' dimension of corporate decision making. By this is meant the role played by people with power and influence within and outside an organization, in shaping the organization's behaviour, the way it is managed, how decisions are taken, its evolution over time, and the cultures which are established within the organization.
[33] This is often referred to as 'red tape'.

Previous Page Top of Page Next Page