1.

The Role of Local and Regional Government

This chapter:

introduces the theory of fiscal federalism and wide-world trend to decentralization specifies possibilities of decentralization of basic functions of public finance

describes the different types of institutions which make up the pattern of subnational government in different countries classifies the varying functions conferred upon, or exercised by, local and regional authorities

describes the main objectives and questions of intergovernmental financial system

Local self-government denotes the right and the ability of local authorities, within the limits of the law, to regulate and manage a substantial share of public affairs under their own responsibility and in the interests of the local population.... Public responsibilities shall generally be exercised, in preference, by those authorities which are closest to the citizen. Allocation of responsibility to another authority should weigh up the extent and nature of the task and requirements of efficiency and economy. (European Charter of Local Self-Government: Article 3 and 4) 1.1 A General Move towards Greater Decentralization The traditional approach to regional and local government economics is that of fiscal federalism. This is the result of the development of local and regional government economics in federal countries, particularly in the USA, Canada, but also in Austria, and Germany etc. The fiscal federalism literature uses efficiency criteria for assigning individual public sector activities and revenue sources to the various levels of government within the federal system. It refers to market failure, tax incidence, public choice, and other theories to determine the level of government most appropriate for delivery of specific public sector outputs. There is a presumption in favour of decentralisation because it facilitates the matching of public sector outputs and local preferences, so promoting allocative efficiency. Although developed predominantly for the federal system of American government (see more Oates 1972, Musgrave and Musgrave, 1984), economic literature emphasizes the general applicability of the prescription of fiscal federalism to non-federal countries. Those prescriptions can be used to analyze the design and operation of multi-level government in any country. However, this text is not solely restricted to fiscal federalism. Regional and local government has become much broader during 1980s and 1990s. The explosive growth of the public sector in the industrial countries in the post war period, which was associated with the expansion of central government's role in income maintenance, income redistribution, and stabilisation, has recently led to strong reactions. At the political level, the 1980s and 1990s have seen a swing towards more conservative attitudes, and especially the suspicion towards powerful central governments. The view that greater reliance should be placed on the market has been accompanied by the parallel view that less power should remain in the hands of the central government. Some influential economists have questioned the effectiveness of government action in stabilizing the economy and improving the distribution of income, thus reducing poverty and unemployment. This challenge has reduced the legitimacy of central government's action and created a presumption in favour of reducing the size of the public sector while giving more power to both the market and local jurisdictions. Many countries are considering devolution of some functions to local jurisdictions. In terms of resource allocation various arguments have been advanced to support the view that privatisation and decentralisation would lead to greater efficiency and a leaner public sector. Developments in specific countries, such as Canada, China, and some of the new states of the former Soviet Union, have forced a reassessment of multilevel finance. I n China they were driven by the need to re-establish some control over national public revenue. In the states that emerged from the break-up of the Soviet Union there was a need to create new fiscal arrangements and to give significant responsibilities to subnational governments from scratch. This proved difficult especially in Russia, with its regions of widely diverse cultural, ethnic, and economic composition. The interest of fiscal federalism in these countries was the logical outcome of discussion about what political organisation these countries should have after the breakup of centralized policy making. Other countries, such as Ethiopia, have been driven towards decentralization (or reorganisation) by ethnic diversity and by the belief that decentralisation would help hold the country together. During the 1980s another group of countries, including Argentina, Brazil, India and Nigeria, experienced macroeconomic problems that required major adjustment in their fiscal account, through revenue increases or expenditure cuts. In the last twenty years, European countries have also implemented reforms in favour of a greater decentralisation for local governments: increasing their responsibilities, apriori controls removed, tax autonomy consolidated, decreasing in the level of earmarked financial transfers in favour of general transfers, creation of new subnational government tiers. A number of these reforms concerned the regional level. This general move towards a consolidation of regions was undertaken with several goals in mind, such as the improvement of country planning or better living standards for local inhabitants. We can say that decentralization is an increasingly important phenomenon in many countries around the world, both large and small (Tanzi, 1995). Whatever the causes, the debate on decentralization raises serious questions about its potential impact, merits, and dangers. Whether the decentralization is appropriate often depends on many country-specific factors. However, many general issues are relevant to every country. 1.1.1 Defining Decentralization Before discussing the relationship between decentralization and economic efficiency it is useful to distinguish three forms of decentralization: 1) Economic decentralization is concerned with thelocationofeconomicdecisions, these being decentralized by definition within perfectly competitive markets (that is consumer sovereignty). Economic decentralization exists when subnational governments have the power, given to them by the constitution or by particular laws, to raise (some) taxes and carry out spending activities within clearly established legal criteria. Examples of this decentralization include the fiscal federations in Argentina, Australia, Brazil, Canada, India, Germany, Nigeria, Switzerland, and the United States (Tanzi, 1995). 2) Political decentralization refers to devolution of political decision-making to local and regional governments. Subnational governments have powers to levy their own taxes and user-charges to finance provision of a self-determined and level of public sector outputs. (This is the form of decentralization to which European Charter of Local Self-Government refers).

3) Administrative decentralization refers to the criterion of regional offices of central government departments with or without decisionmaking powers independent of sanction by the centre. The former is administrative discretion (that is the ability to make operational decision within set policy parameters). The latter is simply administrative decentralization. According to Tanzi (1995, p. 297) administrative decentralization exists when most taxes are raised centrally, but funds are allocated to decentralized entities that carry out their spending activities as agents of the central government and according to the guidelines or controls imposed by the central government 1. An example of administrative decentralization is
provided by Italy, where in 1992 local entities raised about 8 percent of the total net revenue of the general government but spent about 37 percent of total net expenditure.

However, these three forms of decentralization are not necessarily mutually exclusive. The main economic justification for decentralization rests largely on allocative or efficiency grounds. There can also be a political argument for decentralization if a country's population is not homogeneous and if ethnic, racial, cultural, linguistic, or other relevant characteristics are regionally distributed (as they are in Russia and Ethiopia). Decentralization may be needed to induce various regions to remain part of a federation. According to this argument, decentralization would be more desirable in, say, Russia than in Japan. By the same token the goal of national unity often has pushed non-democratic governments towards the forced elimination of regional differences. In democratic societies the economic and political arguments for decentralization tend to converge, since it is argued that decentralization strengthens democracy. Most people are more inclined to engage in local political activities because local policies have a more direct impact on their daily lives. 1.2 The Economic Rationale for Decentralized Government The conventional argument for subnational government is that it secures the public interest in facilitating representative democracy, a crucial component of the democratic state in promoting pluralism, participation, and public choice. Pluralism refers to the capacity of the system of government to accommodate alternative political views so as to avoid a tendency towards centralized authority. Participation refers to the role of subnational government in providing opportunity for people to take an active part in government. Public choice refers to the role of subnational governments in providing services accordance with local and regional needs and preferences, rather than according to uniformed national standards. 1.2.1 Possibility of Decentralization of Public Finance Function The economic theory emphasizes the four main economic roles of subnational governments: allocative, distributive, regulatory and stabilization roles. I t is usually argued that stabilization and income distribution are properly the concern of the central (national) government while resource allocation is primarily the concern of subnational governments. Although it is also usually argued that the regulatory function is best undertaken by central government, the division of that function between central and subnational necessarily reflects the division of service responsibilities between those tiers of government. The market failure in achieving allocative efficiency is usual justification for government intervention in the economy. Market failure at local and regional levels may relate to natural monopolies (such as water and sewerage services), to public goods (for example local street lightning, local environmental health), to merit goods (such as housing of adequate standard) or to externalities (for example relating to unacceptable or beneficial use in neighborhood). According to fiscal federalism and public choice literature, the subnational governments should not undertake redistribution of income (via taxis and subsidies) because of the inefficiencies it would create. The traditional argument for centralizing the redistribution policy is based on the mobility hypothesis (Thiebout, 1956), which applies particularly at the local government level, where there is a relatively high mobility of taxpayers between neighboring jurisdictions - especially within urban regions. The question one has to ask is whether a local government can implement an aggressive redistribution policy without jeopardizing its own existence in the long run (Oates, 1972). Subnational governments adopting programmes with substantial redistribution objectives would have to tax high-income groups in order to pay subsidies to low-income groups. Low-income groups would have an incentive to move in. As a result, tax rates on the higher-income groups would have to rise in order to compensate for the diminishing per capita tax base. In turn, this would increase the incentive for high-income groups to move out. So, it has long been argued by Oates (1972), Musgrave and Musgrave (1984) and others that redistributive policies are best undertaken by central government, since it reduces the incentives for migration of income groups from one local government to another. There are, however, a number of counter-arguments that may apply in practice.

1. 2. 3.

First, owing to several economic and non-economic reasons (particular local customs, established interpersonal relations, geopolitical idiosyncrasies, ethnic and religious factors, language barriers, social security heteronomy, labour market organizations, and so on), inter jurisdictional mobility is not very high, at least in the long run, thus allowing decentralized redistribution policies to be effective. Second, if central and subnational governments have heterogeneous preferences for local public goods, it may be optimal, on Paretoefficiency grounds, to decentralize redistribution policies (Dafflon, 1992). Third, it is often argued that costly information about the potential recipients of these policies can be obtained more effectively by the lower (or lowest) government tier.

Despite these particular counter-arguments, there is still a strong presumption for centralizing redistribution policies. Economic literature also argued that responsibility for stabilization policy cannot be left to local or regional government but must be conducted in a central fashion. Subnational governments will be ineffective in dealing with unemployment or inflation, because markets are interrelated so leakages will result. Such will clearly be the case within the national government where subnational governments share in an open market and resources and capital can flow freely. However, it also becomes increasingly the case across nations, thus calling for international coordination of macro policies. Among economic policy makers, the consensus for centralizing the stabilization function is general because of free-riding problems that exist with decentralized counter cyclical fiscal policy intended to enhance macroeconomic performance. In most cases local and regional governments will be acting as agents for central governments in carrying out the regulatory function. Regulatory functions may relate to town and country planning, policing central legislation relating to trading standards, the local or regional environment (for example pollution) etc. The Economics of Allocative Efficiency

Allocation theory applied to the public sector says that public services should be provided and their costs shared in line with preferences of the residents of the relevant benefit region. Moreover, given the fact that a political process is needed to secure preference revelation, it follows that particular services should be voted on and paid for by the residents of this region. This theory is based on the realization that not all public goods have similar spatial characteristics. Some, such as defence, benefit the entire country. Others, such as regional transportation systems or forestry services, benefit regions. Still others, such as street lighting or cleaning, benefit only municipalities, or particular districts. Furthermore, different areas have different preferences for public goods. Thus the supply of public goods must be adjusted to the different requirements of different groups. A centralised government might ignore these spatial characteristics and this diversity of preferences, or might not be well informed about them and thus might supply a uniform package to all citizens. When "the jurisdiction that determines the level of provision of each public good includes precisely the set of individuals who consume the good" there is "perfect correspondence" in the provision of public goods (Oates 1972, p. 34). The Figure 1.1 illustrates the situation when centralization is costly if it leads the government to provide a bundle of public goods different from the preferences of the citizens of particular regions, provinces or local governments. If these preferences vary geographically, a uniform package chosen by a national government is likely to force some localities to consume more or less than they would prefer to consume. Figure 1.1 Oates' Decentralization Theorem

Source: Bailey, S.,J.(1999) Local Government Economics,p. 20 Assume constant costs in producing the output such that the cost function is a horizontal straight line (so that average cost, AC,equalsmarginalcost,MC). Assume also that society is either composed of only two individuals (or, alternatively, of two groups with identical preferences within the group but different preferences between them) represented by demand schedules D1 and D2. Assume further that national government would decide to produce output Q2. However, at price P,group 1 demands Q1 while group 2 demands Q3, respectively less than and greater than Q2. Group 1 is therefore forced to consume more of the output than it wishes, the cost of the additional enforced consumption (Q1ACQ2) being greater than willingness to pay (Q1ABQ2). Likewise, group 2 is deprived of extra consumption for which it would be willing to pay Q2EDQ3, compared with a cost of only Q2CDQ3. Hence group 1 incurs a deadweight loss of welfare measured by the area ABC while group 2 experiences a loss of CDE. The creation of two local governments would therefore lead to a welfare gain through improved allocative efficiency compared with choices at national level. Decentralized public choices are Pareto efficient because of the removal of the deadweight loss of consumers' surplus. This result is referred to as Oates' decentralization theorem (Oates 1972). The actual size of allocative inefficiency or welfare loss depends on:

1) 2) 3)

The size of the disparities between central and local choices: the greater the heterogeneity of preferences, the greater the welfare losses, because the distance between the two demand patterns increases, and so do sizes of areas ABC and CDE. The slopes of the two demand schedules: the more inelastic (and hence steeper) they are, the larger are areas ABC and CDE (D1 and D2 each pivoting on A and D respectively). Hence, deadweight loss varies inversely with the price elasticity of demand. The deadweight loss of consumer surplus will vary from service reflecting the heterogeneity of preferences and price elasticises of demand. The extent of any economies of scale in production of the output: assuming ceteris paribus, an inverse relation between the rate of output and unit costs increases the optimum size of subnational government acting as a trade-off against loss of consumer surplus.

In addition to Oates's theoretical argument, which is based largely on the spatial characteristics of public goods, decentralization can be defended on the basis of other, more practical considerations. The other argument of allocative efficiency is linked to the Tiebout hypothesis discussed above that agents are mobile and can thus "vote with their feet" to choose the jurisdiction offering the combination of local public goods and tax system closer to their preferences (Tiebout, 1956). According to this, decentralized system can become a surrogate for competition, bringing to the public sector some of the allocative benefits that a competitive market brings to the private sector. The third criterion assigns the allocation function to the government tier that can provide a particular level of public goods at the lowest unit cost. Another potentially important advantage of decentralization is that it allows experimentation in the provision of the output. When the provision of a public service (e.g. education) is the responsibility of local jurisdictions and when these jurisdictions are free to provide the service in any way they see appropriate, some jurisdictions will discover better ways of providing the service, and other jurisdictions will emulate the successful ones. The more jurisdictions there are, the more simultaneous experiments will take place. When the service is imposed by a national

monopoly, which adopts a uniform approach to providing the service, there will be little or no experimentation, and thus dated methods may continue to be used even when there are better alternatives. This outcome is often noted by the supporters of decentralization, who point to the outmoded curriculums in countries with centralized educational systems. Still another argument in favour of decentralization emphasizes that individuals who are responsible for the results of their actions, and who thus have ownership rights over the outcome, are likely to have stronger incentives to perform better. Therefore, when local officials are directly responsible for providing a public service, and are praised for success and blamed for failure, they will have a greater interest in succeeding. In such cases the community may develop a sense of pride in successful service delivery. Additionally, when the cost of providing a service is borne by the local jurisdiction, the service is more likely to be provided costefficiently - to the point where marginal benefits equal marginal costs. This view, that accountability brings responsibility, motivates much of the support for the decentralization of various functions. Finally, at a time when large public sectors are considered wasteful and inefficient, some literature has argued that decentralization is desirable because it is likely to be associated with a smaller public sector and a more efficient economy (Ehdaie 1994). All the criteria converge on the idea that there ought to be a close correspondence between those who determine the level of provision of public goods, those who pay for this production to take place, and those who actually benefit from it. In fact, when there is an explicit, and simultaneous, link between the benefits of public goods and the taxes or user charges levied to finance them, fiscal accountability is enhanced and allocative efficiency can be attained more easily. 1.2.2 Challenges to Decentralization These arguments help explain why decentralization has become so popular in recent years. Some writers, however, have advanced counterarguments that challenge some of the above conclusions or, at least, outline conditions in which decentralization could be a less attractive policy. The point of this discussion is to identify situations in which decentralization might not lead to the expected results unless important changes are made in the existing conditions. While the theoretical case for decentralization is relatively straightforward, the practical case may be less so (Prud'homme 1994; Bird 1994; Oates 1994). As Oates (1994, p.1) puts it, "fiscal decentralization has much to offer, but it is a complicated enterprise". Insufficient Information As discussed earlier, part of the case for decentralization rests on the spatial characteristics of public goods - some benefit only certain areas of a country. As a result the central government may tend to under produce or overproduce them because it does not have the necessary information on local preferences, or it may not have the right incentives to act on the available information. The argument related to lack of information has been challenged on the grounds that central governments can and do assign government officials to local offices and that these officials may be capable of determining local preferences. The central governments of unitary countries often have representatives (for example, the prefects in France and Italy or the intendentes in Chile) who closely follow local developments and assess local needs. These agents are often highly trained and might even have an incentive to exaggerate the local demand for some public services in order to increase their own power or importance. Therefore, the main question is whether the information these individuals send back to the centre is any more or less correct or biased than that available to local policymakers. Whether local governments are more or less likely than the central government to respond to local preferences depends, of course, on the strength of various incentives and on how political decisions are made. A national government interested in local votes may have a strong interest in meeting local needs. A local government that is not democratic may have little interest in meeting these needs. It should not be automatically assumed that subnational governments are made up of democratically elected officials who necessarily have the public (though local) interest in mind. Where they do, decentralization has a greater chance of succeeding. The basic presumption behind the arguments made by proponents of decentralization is that local democracies are in place and do work. When they do not, the case for decentralization becomes weaker. Corruption Some authors (Prud'homme, 1994 and Oates, 1994) also mention corruption. Oates does not conclude whether corruption is likely to be greater at the local or central level. Prud'homme believes that corruption is a greater problem at the local level and mentions France and Italy to support his view. This issue cannot be settled by empirical evidence, so we must rely on impressions. According to Tanzi (1995) corruption may be more common at the local level than at the national level, especially in developing countries. The reason is that corruption is often stimulated by contiguity, that is, by the fact that officials and citizens live and work close to one another in local communities. They have often known each other all their lives and may even come from the same families. Contiguity brings personalism to relationships, and personalism is the enemy of arm's length relationships. When this occurs the public interest often takes a back seat, and decisions are made that favor particular individuals or groups. It should be emphasized that governance issues are problems at all levels of government in many countries; in some countries the local bureaucracy is certainly more honest than the national bureaucracy. The Quality of Local Bureaucracies Another factor that may reduce the benefits of decentralization is the quality of local bureaucracies relative to national bureaucracies. Central government bureaucracies are likely to attract more qualified people because they offer better careers, more possibilities of promotion and better salaries. This conclusion is strengthened by the argument that talented individuals tend to choose fields that offer better opportunities for advancement over the longer run. To the extent that national bureaucracies offer better opportunities to able individuals than do local bureaucracies, they may attract more qualified and more able individuals. But where qualified individuals are abundant, as is often the case in industrial countries, subnational governments may have staff as qualified as do national governments. On the other hand, where educational standards are low and there is a smaller pool of potentially efficient employees, the argument above carries more weight. This scarcity of local talent may impede decentralization efforts in, say, Ethiopia and other African countries.

Within countries there are often wide differences across regions in the quality of the personnel of local administrations. In Argentina there is a huge difference between Buenos Aires and some of the other provinces, and in Colombia between Bogota or Medellin and some other provinces. These differences are partly explained by differences in available resources, but cultural factors also play a role. Composition of Local Jurisdictions Another practical issue is that in most countries the composition of local jurisdictions is based on past political, rather than economic, considerations (apart from exceptional circumstances in which multilevel arrangements can be created from scratch and can thus be influenced by knowledge about the spatial characteristics of important public goods). Thus the sizes of states, provinces, regions, and metropolitan areas are fixed and largely the result of historical accidents. These are the subnational governments to which decentralization allocates fiscal responsibilities. The chance that the spatial characteristics of the public goods or services whose responsibility is assigned to the subnational jurisdictions will match the areas covered by these jurisdictions - and so achieves the "perfect correspondence" described by Oates - seems slim indeed. The smaller the degree of correspondence is, the smaller the potential economic advantages of decentralization are. Technological Change and Increased Mobility Two other aspects are important in today's world. First, the characteristics of public goods and services are subject to rapid change. Technological and economic developments ensure that new needs for public sector intervention arise continuously. For example, in the past there was a need to protect a city's population from outside attacks and to provide it with information about the time of day. Walls were built around cities and clocks were placed on bell towers to satisfy these needs. The public goods provided by these public services are no longer needed. On the other hand, the need to protect citizens from crime and pollution has become more important. Second, changing technology, combined with greater mobility on the part of citizens, implies that the spatial characteristics of public goods are also likely to change. For example, when mobility is limited, many of the benefits associated with public education are internalized by the jurisdiction that provides the service. But when mobility is high, extra jurisdictional externalities become important. The jurisdiction that finances education may not reap its benefits if those who are educated in its public schools move to another jurisdiction. Similar considerations apply to spending for health and many other services. On the other hand, this spillover problem can be partly solved through a reciprocity rule, especially if services can be standardized across regions. In such cases the existence of the spillover does not reduce the advantage of providing the service locally. But standardization eliminates one of the basic reasons for decentralization. These two aspects imply that, to be optimal, decentralization arrangements should be flexible over time. Either the geographical areas covered by local jurisdictions -and thus the number of these jurisdictions - should change over time, or the characteristics of public goods should be continuously reexamined in order to reallocate some of them across the existing jurisdictions. There is no simple mechanism that allows this process to take place. Once a federal structure is determined, local politicians and officials fiercely oppose major changes to borders and tasks. Consequently, fiscal federalism is at times characterized by a mismatch between the spatial characteristics of public goods and the responsible jurisdictions. 1.2.3 Definition of Local and Regional Government Types of Regional Authority Financing the regional and local governments means the system of decentralized administration - the range of public agencies responsible for undertaking public services and developmental programmes at the subnational level. For convenience, all types of decentralized government agency - state governments in federations, provincial administrations, local authorities, regional-urban development corporations, etc. - will be described as regional authorities. Such types of agency include:

a) a) a) b) c) d)

State/provincial governments under a federal constitution; Regional/provincial/local administrations comprising various branches of central government departments grouped under the leadership of the regional administrator (titles varying from governor or prefect to commissioner or collector); Local and regional governments/authorities directed by representative assemblies or councils; Regional development authorities - public corporations with a responsibility for the comprehensive development of a particular area; Single-purpose authorities: bodies with corporate status but responsible for a single function only; and Regional branches of central government departments responsible only to the latter.

Regional and Local Governments The general system of local and regional governments in any particular country we will call in this text as subnational government. This term will be used for describing any size or administrative level of subnational area from federal state - a province, a district, a town, or a village. Local governments (the term being used interchangeably with "local authorities", "councils" and "municipalities") are sometimes simply referred to in economic literature as "subnational authorities". Local governments can be thought of as democratically-elected bodies whose jurisdiction is of a local (rather than regional or national) scale, have powers to levy local taxes by which to exercise genuine discretion over services provision. However they vary substantially in (both geographic and demographic) size, often share subordination to the centre with other non-local public sector organisations, sometimes including non-elected bodies such as agencies mentioned above. Hence, there is no subset of characteristics by which to define 'pure' local government and distinguish it from regional, state or provincial government or from other subnational public bodies. The defining characteristics of local and regional government are its position on a continuum of more or less local criteria, rather then there being clearly distinct categorical differences between these various subnational bodies. Difficulties in deriving a clear definition of local and regional government make comparative analysis questionable because it is not possible to compare like with like. It may be even be difficult to construct a precise definition of local and regional government in a single country at a particular point of time.

Structurally, state governments in a federal system and local governments have much in common. They are bodies corporate with their own legal identity, their own property, budgets and accounts. They employ their own staff (though these may be seconded from central government service or from a central pool of regional authority personnel). They are normally governed internally by a representative assembly, elected as a whole or at least from its major part. Some part of their revenue is derived from direct taxation of their population. The differences between regional governments and local governments are in status, scale and function. Regional and state governments derive their existence and their authority from the constitution, while local governments are creatures of central governments legislation (or a state government in a federation). Regional governments customarily govern far larger territories in area and population and normally exercise a wider range of functions. This international practice only partially reflects the distinction between federal government at the centre and the subnational governments in regions, states or provinces. However, over 80 percent of world's nations are unitary states, the national parliament being the single source of constitutional power and legislative authority. In many European unitary states the constitutional status of local government is formalized and protected in a written constitution which guarantees the principle of local self-government. But there are exemptions - for example the UK has no such written constitution and local government is therefore constitutionally subordinate to central government. Regional Administrations Regional administrations are essentially deconcentrated arms of central (or state/regional) government. Their basic characteristic is some degree of integration of technical departments under the leadership of an administrator. Their status varies - a centrally appointed politician, a civil servant, a locally-elected governor. Their authority varies too - from formal control over all or most public services, including the direction of technical staff, to a purely informal leadership derived from prestige, from a responsibility for general order, or from control of particular resources. In some countries - notably with French and Ottoman traditions of administration - a regional administration has a dual character, operating partly as a deconcentrated arm of government and partly as a local authority, with staff exercising a dual loyalty. Regional Development Authorities Regional development authorities are more specialized institutions, outside the normal machinery of regional and local government. They are public corporations, usually established by statute, and managed by a board of directors appointed by the constituting authority normally central government but occasionally state or local government. They are similar in structure to other public corporations running public utilities, nationalized industries, agricultural marketing, etc. but are distinct in being responsible for a delimited area and its comprehensive development. This responsibility may be for all developmental or service functions within the area, or alternatively with those functions closely related to a particular type of development - for example the provision of irrigation (involving water engineering, drainage, survey and plot demarcation, agricultural extension) or the construction of a new housing area (involving housing, roads, water, sewerage and community facilities). The establishment of such authorities is usually associated with a development project attempting a transformation of a particular area of its people. Such authorities maybe created to integrate the various skills required for a project, to overcome territorial fragmentation of a project area between regional administrations or local authorities, or because the project is thought to demand a dynamic commercial approach to its management. 1.2.4 Local and Regional Government Functions To understand the variety of funding arrangements for subnational governments, we have to appreciate the diversity of the functions conferred upon them. The function can be classified in five groups: The first set of functions, most traditionally associated with local and regional government, is the provision of social and environmental services. The environmental services: local roads maintenance, cleansing, drainage, street lighting, refuse disposal, sewerage, parks and recreation, are habitually undertaken by local authorities although urban development authorities sometimes carry out the initial installation. Responsibility for trunk roads, flood control, and irrigation varies widely, from state governments to provincial governments (with much direct central funding). What varies widely is local government involvement in:

a) b) c) d) e) f)

public utilities, chiefly water and sewerage (sometimes provided by a national corporation, or typically by a metropolitan corporation or a municipally controlled enterprise) and electricity, (usually a national utility responsibility, but sometimes generated or distributed by local government); social services (local governments often provide primary schools and clinics, less frequently secondary education and hospitals); public protection (fire services are frequently municipal but police forces less frequently so outside Europe and North America); trunk roads (which can be a national, provincial or local responsibility); provision of rental or purchase housing or serviced sites (sometimes a local government activity, sometimes that of a special purpose authority); regulation of land use and development (usually local governments, but on occasion a provincial or metropolitan authority function).

The second group of functions may be regarded as regulatory - making or enforcing rules. Whilst defense and armed forces are invariably the central government responsibility, the maintenance of general law and order is frequently delegated to state government in a federation, to the heads of regional administrations (provincial governors, district collectors, etc), or even, as in France, to elected mayors. Most common to local authorities is power to regulate specific activities - land use, building standards, entertainments, the liquor trade, etc. A third group may be regarded as development functions. Subnational governments may engage directly in forms of economic activity such as the operation of factories, plantations, forests, or trading companies. The more widespread concern, however, is to promote and support private enterprise through the provision of infrastructure - industrial sites, employee housing, irrigation, warehousing, road access, etc. - or through extension services to the public in fields such as agriculture, animal husbandry, fisheries, or small industry. Some continental differences remain here. In Europe, direct concern with industry and agriculture has been largely a central government preserve, although regional authorities are heavily involved in the provision of infrastructure. In much of Africa and the Middle East, developmental responsibilities are devolved upon regional

administration; local authorities have little share in them. In parts of South and South-East Asia, however, rural local government is intended to play a largely developmental role with a prime emphasis on agricultural production. An obligation is frequently laid upon low-level governments to assist in communal mobilization - to promote literacy and adult education or to stimulate self-help schemas - but without major financial implications if these authorities do not have mainstream functional responsibilities. The representative function of subnational government - of expressing local opinion on matters outside its field of executive responsibility - is associated largely with state government and local governments. It is important in determining the weight and influence of local and regional governments, but has little direct financial significance. The final role of regional government lies in the coordination and planning of regional investment and land use. In Third World countries this is usually a role of deconcentrated regional administrations or special development authorities; it is rarely a local government function. Planning and co-ordination is not in itself a major spending function, but the role may attract responsibility for the execution of major internationally or centrally financed development programmes. The allocation of these major functional responsibilities between arms of decentralized administration is crucial to understanding the intergovernmental financial system of any particular country. They determine the scale of demands upon the financial resources with which regional authorities are endowed. 1.3 Objectives of Intergovernmental Financial System The determination of the spending responsibilities of subnational governments precedes the question of how resources will be generated to pay for the spending . The financing of the spending is often almost an afterthought. Yet for decentralization to be successful, it must include the decentralization of both spending and revenue, and these decisions must be made at the same time. It is therefore critical to understand the actual and potential impact of intergovernmental relationship on such key questions as the ability of central government to behave in a fiscally responsible way, and the efficient and accountable delivery of public services. In particular, four questions must be answered with respect to intergovernmental finance in any country:

1) 2)

Who does what? - the question of expenditure assignment. Who levies what taxes? - the question of revenue assignment.

1) 2)

How is any imbalance between the revenues and expenditures of subnational governments that results form answers to the first two questions to be resolved? -The question of vertical imbalance. To what extent should fiscal institutions attempt to adjust the different needs and capacities for different governmental units at the same level of government? - the question of horizontal imbalance or equalization. These questions of fiscal relationship and financial system must be answered "locally" in each particular country. Nonetheless, there are certain aspects of these complex questions that should be considered as "universal" principles. This means that the financial system should provide a rational distribution of power between levels of government over rising revenues and expenditure of public resources. Financial system should ensure that the devolution over financial resources is consistent with the general delegation of responsibilities. The system should ensure accountability to the public for the use of resources; those who determine levels of expenditure should have to face the taxing consequences for the community. The system should provide an adequate share of public resources as a whole for the functions - service and developmental - operated by local and regional government. The system should, so far as possible, distribute public expenditure equitability between regions. The cumulative effects of historic concentrations of development in particular regions and of different resource endowments cannot be corrected overnight or, indeed, completely. But the financial system can and should be sensitive to them, seeking improvements in distribution wherever possible, and giving priority to equalization of certain basic services such as education, at least. Finally, taxing and charging by subnational governments should conform with an equitable distribution of the total burden of public expenditure on the community. What matters primarily to the citizens is his total tax burden; wherever the taxes he pays go to central, regional or local government is of secondary importance for him. The text concentrates more on subnational government revenue. Expenditures is of concern only in its nature and scale affect the needs of subnational governments for revenue, the way in which it is allotted to them, and its use determined or influenced by higher levels of government. The next chapter will classify these resources and the different approaches to the matching of resources and functions.

Conclusions

The traditional approach to regional and local government economics is that of fiscal federalism. The fiscal federalism literature uses efficiency criteria in assigning individual public sector activities and revenue sources to the various levels of government within the federal system. It refers to market failure, tax incidence, public choice, and other theories to determine the level of government most appropriate for delivery of specific public sector outputs. During the 1980s and 1990s the decentralization became an increasingly important phenomenon in many countries around the world, both large and small. There were several reasons: a decrease of reliance credit of central government to be able to stabilize economies efficiently, too large and inflexible public sector to allocate resources effectively. Further on, a political and economic development in some countries led to reassessment of multilevel finance system. Decentralization may be needed to induce various regions to remain part of a federation if ethnic, racial, cultural, linguistic, or other relevant characteristics are regionally distributed. European countries have implemented reforms in favour of a greater decentralization: a consolidation of regions or creation of new subnational government tiers.

We can recognise three forms of decentralization: economic, political and administrative decentralization. They are not necessarily mutually exclusive. In democratic societies the economic and political arguments for decentralization tend to converge, since it is argued that decentralization strengthens democracy. Decentralization promotes pluralism, participation and public choice. The economic theory emphasizes the four main economic roles of government: allocative, distributive, regulatory and stabilization roles. According to fiscal federalism and public choice literature, the subnational governments should not undertake redistribution of income (via taxis and subsidies). The traditional argument for centralizing the redistribution policy is based on the mobility hypothesis (Thiebout, 1956). Also the responsibility for stabilization policy must be conducted with the central government. It is also usually argued that the regulatory function is best undertaken by the central government, but in most cases local and regional governments should act as agents for central governments in carrying out this function. Within the context of fiscal federalism and allocation theory the most appropriate function for subnational governments is the allocative function. It is based on the fact that the various goods and services have different spatial characteristics and benefit incidence. The spatially limited nature of benefit incidence thus leads to a fiscal structure composed of multiple service units, each covering a different sized region within which supply of a particular service is determined and financed. This results in allocative efficiency and is illustrated by the Oates' decentralization theorem. Another argument in favour of decentralized system is that it can bring competition into the public sector, it can provide public goods at the lowest unit costs, allows experimentation in the provision of the output, the officials at lower tiers of governments have stronger incentives to perform better and the service is more likely to be provided cost-efficiently -which means that accountability brings responsibility. There are some counterarguments that may reduce the benefits of decentralization. They depend on many country-specific factors and on existing conditions. This is the case of insufficient information when at the subnational level the officials should not necessarily have local interest in mind, the case of corruption at the local level, the quality oflocal bureaucracies relative to national bureaucracies, the composition oflocal jurisdiction based on past historical, political, rather than economic considerations and spatial characteristics and rapid change of public goods and services characteristics, technological changes and greater mobility ofcitizens implies the changes in demand for public services, creates the problems with externalities and spillover effects. Among all types of decentralized administration which are responsible for undertaking public services at subnational level we include: state or provincial governments in federations, regional/provincial/local administrations comprising various branches of central government departments, local and regional governments/authorities directed by representative assemblies or councils, regional development authorities, single-purpose authorities, regional branches of central government departments. Local governments can be thought of as democratically-elected bodies whose jurisdiction is of a local (rather than regional or national) scale. They have powers to levy local taxes by which to exercise genuine discretion over services provision. Local and regional governments are bodies corporate with their own legal identity, property, budgets and accounts. The differences between regional governments and local governments are in status, scale and function. Regional and state governments derive their existence and their authority from the constitution, while local governments are creatures of central governments legislation (or a state government in a federation). Regional governments customarily govern far larger territories in area and population and normally exercise a wider range of functions. Regional administrations are essentially deconcentrated arms of central (or state/regional) government with some degree of integration of technical departments under the leadership of an administrator. Regional development authorities are more specialized institutions. They are public corporations, usually established by statute, and managed by a board of directors appointed by the constituting authority. The establishment of such authorities is usually associated with a development project attempting a transformation of a particular area of its people.

The functions conferred upon subnational governments can be classified in five groups: the provision of environmental and social services, regulatory functions, development functions, the representative function, the coordination and planning of regional investment and land use. For understanding financing the subnational governments it is crucial to answer to four questions: the question of expenditure assignment (what spending responsibilities of subnational governments are); the question of revenue assignment (how resources will be generated to pay for the spending - who levies the taxes); the question of vertical imbalance (how any imbalance between the revenues and expenditures is resolved) and the question of horizontal imbalance, o r equalization (how fiscal institutions adjust for the differences in needs and capacities of different governmental units at the same level of government).

2. Revenue Generation - Introduction

This chapter: introduces the alternative methods of funding subnational governments

describes the different types of transfers from central government as grants, tax sharing, loans from central government budget and equity specifies in general the problem of tax assignment by different level of government, and the possible forms of tax revenues raised by subnational government discusses the grounds for charging as another source of revenue classifies other sources - borrowing and generation of revenues from the operation of commercial or productive enterprises of subnational governments discusses the different approaches to the financial relationships between central and subnational governments

Local authorities shall be entitled, within national economic policy, to adequate financial resources of their own, of which they may dispose freely within the framework of their powers. Local authorities' financial resources shall be commensurate with the responsibilities provided for by the constitution and the law...The financial systems on which resources available to local authorities are based shall be of a sufficiently diversified and buoyant nature to enable them to keep pace as far as practically possible with the real evolution of the cost of carrying out their tasks. (European Charter of Local Self-Government: Article 9)

2.1 Financing the Activities of Local and Regional Governments In this chapter we introduce alternative methods of funding subnational governments. At this introductory stage it is useful to classify funding sources as a guide to a detailed discussion in the remaining chapters. This chapter also looks back to the previous chapter by describing briefly alternative approaches to the relationship between functions and resources. Local and regional governments in most countries have to meet the costs of constructing and maintaining services that are rising because of inflation and the fact that more powers to provide goods and services are being shifted on them; and expanding because of an increasing population (mainly in the third world regional authorities). A major task is to raise adequate revenue for both capital investment and for recurrent expenditure the employment of staff, operation and maintenance and the servicing of debt. Such revenue will only be adequate if it grows with the demands made on it, i.e. in line with rising costs. According to Bird (2000, p.13) two basic principles of assigning revenues to subnational governments may be suggested. First, "own" source revenues should ideally be sufficient to enable at least the richest subnational governments to finance all locally provided services primarily benefiting local residents from their own resources. Second, to the extent possible, subnational revenues should be collected only from the local residents, preferably in relation to the perceived benefits they receive from local services. This approach is in turn based on principles:

1) 2) 3)

more attention should be paid to matching expenditure and revenue needs at different levels of government; in addition, more effort should be made to ensure that all governments bear significant responsibility at the margin for financing the expenditures for which they are politically responsible; and finally, subnational taxes should not distort the allocation of resources.

2.1.1 Central Government Allocations Allocations from the central government budget - appropriations from revenue accruing in the first instance to national government - are a significant and often predominant source of funding for subnational governments. They are frequently described as "transfers" and take several forms. Grants Allocations to devolved authorities, such as state governments in a federation or local governments, more frequently take the form of grants or subsidies (interchangeable terms). Although these are appropriations from the central budget, they involve a transfer of cash to a self-accounting recipient; (exceptionally a transfer may take another material form - services of seconded staff, or supplies from centrally purchased stores - but the cash value would nevertheless be reflected in the books of both granting and receiving authorities). Grants and subsidies take many forms - multipurpose, single-purpose, unit cost, percentage, equalization, deficiency, and so on. Tax Sharing The other type of allocation is the tax sharing. The sharing may be of (at least) two kinds. Different levels of government may tax the same base or one level may collect the tax from a given base and share the revenue with other levels. Examples of the first kind are the taxing of personal income in the United States and the taxing of sales in Argentina. In the United States personal income is taxed by both the federal government and by most states. Counties and municipalities surcharge on states' income taxes, which is sometimes called "piggy-backing". An example can be Scandinavian countries, where there is a surcharge on personal income tax. In Argentina, sales are taxed with a value-added tax at the national level and with a cascading turnover tax at the provincial level. When two government levels tax the same tax base, each retains its independence of action even though an increase by one level in its dependence on that base may limit the scope for the other level to tax the same base. At the subnational level the limits on effective tax rates on a given tax base are generally imposed by tax competition and by the potential mobility of the tax base. Examples of the second kind of tax sharing are quite common. They exist in Argentina, Brazil, Colombia, Pakistan, Russia, and in European countries, including the Czech Republic. Whether such shares should be regarded as a central allocation or direct regional revenue is open to question. In a unitary state, the sharing would normally be a purely discretionary act on the part of central government - virtually an alternative form of grant. In a federation, however, tax assignments may be built into the Constitution; central collection and transfer is an administrative device which is not (or should not be) abrogated by the "rights" of the State to their stipulated share of the proceeds. In practice, even this constitutional distinction may be blurred. For example the Indian Constitution guarantees the rights of the states to a share of income tax, but does not stipulate the size of the share or the manner of its territorial distribution. Two further refinements must be mentioned at this stage. The amount assigned to regional authorities may be the total yield of a tax or a fixed proportion of it. The share received by an individual authority may relate to the amount of tax derived from its region; alternatively, the assigned proceeds may be pooled and distributed to regions according to some formula unrelated (or at least, not wholly related) to their geographical derivation. Loans Loans constitute a further type of central allocation to subnational governments, particularly common to the funding of development corporations, but also provided to regional governments and local authorities for capital investment and, occasionally, for short term debt relief. They are distinguished from other forms of central allocation in that they specifically require repayment. They may, nevertheless, involve an indirect element of subsidy if they are interest-free or interest is charged below market rates. Equity The final form is central government investment of equity capital in a regional or local government. The recipient authority would normally be expected to operate in a quasi commercial manner and to invest the capital in a "self-liquidating" operation, i.e. in an activity which would earn revenue to cover its current costs and possibly yield a surplus to finance expansion or pay dividends. Central government will not expect to recoup

the capital sum invested, though it might hope to receive a return through dividends. Such funding is most typical of regional development authorities. It might be provided also to a semi-commercial undertaking run by another form of regional authority - a municipal water enterprise, for example. 2.1.2 Taxation It has generally been argued that local governments should finance their spending through "benefit pricing" or benefit taxation. For example, Musgrave and Musgrave (1989, p. 517) write that "the choice of tax instruments to be used by 'local' jurisdictions ... should conform to the rule that each jurisdiction pays for its own benefits." This is necessary because "benefit taxation - requiring as it does a balance of tax burdens and benefit gains -neutralizes the impact of fiscal operations on location choice" (p. 518). This is, of course, necessary for the result to be Pareto-optimal. However, Musgrave and Musgrave also recognize, however, that "the assumption of universal benefit taxation ... is unrealistic" (p. 518). The first approach, already discussed, is the sharing of taxes levied and collected by the central government. Secondly, subnational governments may impose a surcharge on a tax levied and collected by central government. Taxpayers within their jurisdiction normally pay this excess charge with their basic tax to the central government, which then pays over the proceeds of the surcharge to the regional authority. Swedish local authorities, for example, impose a surcharge on national income tax; in parts of the United States, local authorities place a surcharge on the state level sales tax; some Indian panchayats (rural local authorities) benefit from surcharges on land tax. The surcharge may be levied as an additional percentage of the tax base (e.g. of taxable income, in the case of income tax) or, alternatively, as a percentage addition to the actual tax paid to the central (or regional) government. The third source of tax revenue is the levies which subnational governments collect and retain themselves. (There are examples of subnational governments sharing the proceeds of their own collections with central government, but these are rare.). There is an almost infinite variety of such taxes. A major variable, however, is the legislative base. A tax collected by subnational government may be levied under central government legislation which may, in turn, be mandatory or permissive, i.e. it may require subnational governments to collect it or merely give them discretion to do so. The central legislation may also stipulate the tariff or leave this to the discretion of regional authorities within or without limits. Alternatively, taxes may be imposed by regional authorities under their own legislative power. The question of tax assignment by level of government has been discussed by many authors. The assignment of tax revenue to multilevel governments can follow several options. The first option is to assign all tax bases to the local authorities and ask them to transfer some of the revenue to the central government to allow it to meet its spending responsibilities. The amount transferred upward could be determined by rule, formula, or negotiation. This option is often unattractive and inefficient for a number of reasons. It is inconsistent with a national policy that aims to redistribute income through the tax system. It is inconsistent with a policy that calls on the public sector to stabilize the economy, using the tax system to achieve this objective. It may result in excessive fragmentation of the tax system, and it may provide the wrong incentives to the subnational jurisdictions if they know that part of the taxes they collect will be shared with the national government. There is also evidence from some countries (for example China and Mexico) that this policy leads to inefficient tax administration. The second option is for the central government to collect all taxes and transfer some of the revenue to the subnational governments. The transfer of funds to the subnational governments can be done by sharing total tax revenue or by sharing specific taxes. As argued earlier, the first approach is superior because it gives local governments a morestablerevenuesourceandgivesthenationalgovernmentmorefreedominpursuingits tax policy options. Still, there are problems with this option. Breaking the connection between decisions to collect tax revenue and decisions to spend that revenue destroys the concept of the tax price for public spending (that is, the idea that spending decisions carry a specific cost expressed through the taxes paid). Local officials and taxpayers may not connect the benefits they derive from public spending with the taxes they pay. Therefore local officials may not exercise the required restraint on spending, and taxpayers will be less willing to pay taxes. The third, most common option is to assign subnational governments some taxing power and, if necessary, to complement the revenue raised locally with grants from the central government. The taxing power can be provided to the subnational governments by assigning them exclusive use of some tax bases, allowing them to share some bases with the central government, or allowing local governments to surcharge on some national taxes. The assignment of tax bases to subnational governments must take into account several considerations. 1) The first is the importance of the objectives (other than raising revenue) being pursued through taxation. The more important these other objectives are, the less advantageous it is to leave these tax bases to subnational governments. For example, if the government assigns considerable weight to income redistribution (through progressive taxation) or stabilization (through built-in stabilizers), certain tax bases, such as the progressive income tax and the corporate income tax, should be left to the national government. The second consideration is the mobility of the tax base. If a tax base can easily escape taxation at the local level by moving to another jurisdiction, that base is not a good candidate for local taxation. Thus, the more mobile the tax base, the more desirable that it remains at the national level. The third consideration involves economies of scale. Depending on informational requirements (for example, the need for a national taxpayer identification number), technical requirements (the use of large computers), or other factors, economies of scale in tax administration for a given tax argue for leaving that tax to the national government. This consideration implies that the value-added tax and the global income tax should be nationally collected taxes. Regional governments and local authorities habitually derive part of their revenue from taxation, although in greatly varying proportions. Regional development authorities and deconcentrated regional administration might collect central government taxes as agents or receive tax shares but are rarely taxing authorities in their own right since they are not directly representative bodies; the traditional link between taxation and representation survives even under largely autocratic regimes. 2.1.3 Charging

Taxes normally have to be paid by members of the public as a legal obligation regardless of the extent to which they personally benefit from the services they finance. Charges, by contrast, are paid directly by those who consume a service and are normally appropriated to meeting all its costs or at least part of them. Examples may include rents for local authority houses, irrigation charges, and swimming pool entrance fees. The grounds for charging - the justification for charging the cost of a service directly to the consumer rather than the taxpayer at large - will be discussed in Chapter 4. The border line between charges and taxes is not always clear. A charge may fall on consumers only, but may have something of the character of a tax where variations in its scale are not strictly related to the volume of consumption - water rates based on property values rather than metering, for example - or where revenue exceeds service costs and is appropriated to other purposes. When making decisions whether a service should be financed by the means of charges or taxes, several aspects have to be concerned: What distinguishes a service financed by consumers from one funded by taxpayers? The first consideration is fairness. Broadly speaking, those who benefit should pay. If a service or, to be specific, the unit of a service benefits one person exclusively, that person should pay for it through a consumer charge. Examples might be a domestic electricity supply or a telephone extension. If a service benefits everyone collectively and indiscriminately, like defence or disease control, the cost should be borne by taxation. Many services fall between these two categories (known respectively as private and public goods). A service may benefit one person particularly but nevertheless have some impacts on others; such impacts are known as externalities and may be positive or negative. Domestic water supplies benefit the individual household and their cost can by measured; but they reduce communicable disease and therefore have wider benefit. Where the benefits are both private and public, a consumer charge subsidized from taxes may be justified if it enables or encourages a wider number of people to use the service. The mix of charge and subsidy should depend upon the balance of private and public benefit. Where the impacts are negative - for example, the congestion or pollution caused by city centre parking - one may wish to make the charge punitive, i.e. above the level of the service cost, - to discourage consumption. There is a further aspect of fairness. Some services may be largely private goods with little externality, but nevertheless regarded as "merit goods", i.e. basic human needs. Subsidisation or even total financing from taxation may be regarded as right if it enables consumption by those who are too poor to meet a full consumer charge. This is often applied to primary education, medical care and even housing, its extent depending on what an economy can afford and what contemporary values regard as a minimum standard of living. Critics argue, however, that such subsidisation should be directed to general income support so that the poor can exercise choice over the services they use. A second major consideration is economic efficiency. Where individuals are free to choose how much of a service they consume, charging enables the price mechanism to play a crucial role when allocating resources through:

rationing demand: on the basis that those who value the item or service most will be prepared to pay most

providing the incentive to avoid waste providing signals to the supplier concerning the scale of production providing the resources to the supplier to increase supply.

Without a price, demand and supply are unlikely to be brought into equilibrium and thus the allocation of resources will not be economically efficient. Water supplies and medicines are examples of costly goods for which charging is particularly supported on efficiency grounds. The problem, however, is that the market mechanism does not act perfectly. In many cases government is a monopoly supplier, and may be tempted to charge more than the necessary cost of a service, either to reallocate resources or because of inefficient provision. A third factor is administrative convenience. Charges are often a relatively easier form of revenue to collect than taxes, because they can in most cases be enforced through cutting off a service. Charging is a revenue source common to all forms of local and regional governments; it may well be the predominant source of current income for regional development authorities. 2.1.4 Borrowing Loans have been mentioned as a channel of central government finance. Regional authorities in different countries borrow also from a variety of other sources, including international and foreign agencies, centrally operated credit funds, commercial banks and other financial institutions (insurance societies, for example), private investors, or internal reserve funds. These will be discussed in Chapter 6. Borrowing serves different purposes:

1) 2) 3)

funding short-term cash flow deficits; financing investment which is expected to earn income; or paying for long-term capital development.

Regional development authorities are frequently capitalized by long-term loans and many regional governments borrow extensively for capital development. The extent of local authority borrowing varies radically between countries, both in legality and scope. Regional administrations normally lack the corporate status necessary for borrowing money in their own right. 2.1.5 Enterprise

Subnational governments may generate revenue from the operation of commercial or productive enterprises as a source of net income. The base, of course, is charging - but charging with the intention of profit to finance other purposes, not just to operate the enterprise itself. Such activities often represent incursions into what are otherwise private sector fields such as retailing, entertainment, or catering. Their significance in their contribution to general subnational government revenue is only occasionally large. All forms of regional authority may engage in profitable enterprise; there is often a greater bonus upon regional development authorities to do so. 2.1.6 Resources and Responsibilities

The previous chapter outlined a range of functions which may be discharged by different types of subnational governments and authorities in different countries. In this chapter we outlined a wide range of sources which subnational governments can use to finance their activities. Clearly, the critical elements of all the possibilities are:

1) 2)

the relationship between the allocation of functional responsibilities and the assignment of revenue sources - the match of tasks and resources, and the degrees regional discretion attached to the management of each. This is the nub of the central-regional financial relationships enshrined in the constitution, the statutes or the fiscal and budgetary policies of each nation.

These financial relationships vary enormously in detail, but are characterized by three different approaches. The first approach is based upon the assignment of income sources. In practice, subnational governments are given certain revenue sources (mainly taxes) to exploit, or a certain proportion of central tax revenues. The scale of their expenditure, and therefore the scope of the responsibilities they undertake, is effectively determined by the yield of these revenues. Some modifications may be made. Grants may be paid to standardize the potential revenue from the assigned resources - to compensate areas with lower fiscal capacity. The sharing of central taxes may be weighted in favour of areas with above-average expenditure needs. Basically, however, financial relationships revolve around the assignment of income rather than the financing of set levels of expenditure. Local and regional governments do what they can afford to do. The second approach might be defined as the expenditure approach. Grants, loans, and other allocations are based upon expenditure estimates - upon the cost of exercising a particular function at a particular level. This costs estimate may be the actual cost or some standard projection of what needs to be spent, and the central allocations may meet all or only part of the cost. But the financial relationships focus upon financing, by one means or another. Finally, there is a comprehensive or deficit approach. Revenue sources are assigned to subnational governments. Responsibilities are allocated too, with assumptions about their expenditure implications. Grants are paid to bridge the gap between these revenue sources and expenditure needs. These grants may be calculated according to actual revenue and actual expenditure, or upon estimates of revenue potential and expenditure need, using standard criteria. Either way, the financial relationships attempt an integrated matching of resources and functional responsibilities. These approaches are not mutually exclusive. A single authority might receive loans for self-liquidating projects, unit cost or percentage grants for a specific service, and tax shares attached to non-defined expenditure need (beyond, perhaps, a matching contribution to a percentage grant). The different approaches might apply to the financing of a local and regional government as a whole, or to different functional fields. The relationship between responsibility and resources can be seen in two ways: resources create functions, or functions search for resources. Two opposing principles are at stake. The first assigns resources to subnational governments and then permits them to undertake those functions appropriate to their means. The second principle assigns functions to subnational governments and then seeks to allocate the resources to match them. Practice tends to fall between these opposites. Some functions may be mandatory, implying a right to allocation of enough resources for at least some basic level of performance; other tasks may be purely optional and dependent upon what an authority can and wishes to afford. Functions can be assigned with varying degrees of prescription over their performance and thereby their resource requirements. This brings us to the second point. Delegation of responsibility and resources to regional government is not necessarily an absolute; it can be a matter of degree. Delegated discretion to spend money can vary from an unlimited right to spend any amount on any purpose, to achoiceof whether to buy blue or red paint. The right to raise revenue can vary from a constitutional discretion to impose any tax not specifically reserved for another level of government on the one hand, to collection of a tax according to predetermined assessments and tariffs on the other. Finally, the exercise of discretion is firmly constrained by the political, economic, and geographical environment within which subnational government operates. Individual types of revenue generation and their participation on composition of subnational government financial resources may be clearly seen in the following Figure 2.1. Figure 2.1 Composition of subnational government financial resources (As a percentage of total financial resources, 1991)

For non-tax revenue: 1984 for Switzerland, 1997 for France, Ireland, Luxembourg, the Netherlands and Spain, 1998 for Iceland, Norway, Portugal and the United Kingdom. 2. Non-tax revenues include: operating surpluses of public enterprises controlled by subnational governments; property income; fee, sales and fines; contributions to government employee pension funds and capital revenues. Source: Joumard, I., Kongsrud, P., M. (2003) Fiscal Relations across Government Levels. OECD Economic Studies No 36, 2003/1, p.183

1.

Conclusions
This chapter has classified the financial resources which may be made available to local and regional government. A wide range of options include: loans, equities, grants, tax shared, taxes delegated, locally imposed taxes, fees, charges, trading profits. Allocations from the central government budget are a significant and often predominant source of funding for subnational governments. They are frequently described as "transfers" and take several forms: grants, tax sharing, loans and equity. Allocations to state/provincial governments in a federation or local/regional governments, more frequently take the form of grants or subsidies (interchangeable terms). They take many forms - multi-purpose, single-purpose, unit cost, percentage, equalization, deficiency, and so on. The tax sharing may take (at least) two forms. In the first form different levels of government may tax the same base (e.g. personal income tax in United States); in the second form one level of government may collect the tax from a given base and share the revenue with other levels. This tax sharing arrangement is quite common in many countries - Argentina, Russia, Pakistan, and European countries, including the Czech Republic. The amount assigned to regional authorities may be the total yield of a tax or a fixed proportion of it. The share received by an individual authority may relate to the amount of tax derived from its region; alternatively, the assigned proceeds may be pooled and distributed to regions according to some formula unrelated (or at least, not wholly related) to their geographical derivation. Loans are provided to subnational governments for capital investment and, occasionally, for short term debt relief. They specifically require repayment and may involve an indirect element of subsidy if they are interest-free or if interest is charged below market rates. Another form of central allocation is central government investment of equity capital in a regional or local government. Such funding is most typical of regional development authorities. It might be provided also to a semi-commercial undertaking run by another form of regional authority - a municipal water enterprise, for example. The sources of revenues from taxation of subnational levels of government may take form of shared taxes levied and collected by the central government, or subnational governments may impose a surcharge on a tax levied and collected by central government; or subnational governments impose and collect taxes themselves. The assignment of tax revenue to multilevel government may be made by assigning all tax bases to the local authorities and asking them to transfer some of the revenue to the central government to allow it to meet its spending responsibilities; by collecting all taxes by central government and transferring some of revenue to the subnational governments; by assigning subnational governments some taxing power and, if necessary, to complement the revenue raised locally with grants from the central government. The assignment of tax bases to subnational governments must take into account the importance of the objectives (other than raising revenue) pursued through taxation, the mobility of the tax base and economies of scale. Charges are paid directly by those who consume a service and are normally appropriated to meeting all or part of its cost. In decision making whether the service should be financed through taxation or through charging it is necessary to consider several aspects: fairness, externalities of goods and services, regarding good as a "merit good", economic efficiency, and administrative convenience.

Subnational governments borrow also from a variety of other sources including international and foreign agencies, centrally operated credit funds, commercial banks and other financial institutions (insurance societies, for example), private investors, or internal reserve funds. The extent of local authority borrowing varies radically between countries, both in legality and scope. Subnational governments may generate revenue from the operation of commercial or productive enterprises as a source of net income. Their significance in their contribution to general subnational government revenue is only occasionally large. Since there is a significant diversity of financial resources to be used by subnational governments, it is essential to focus on the relationship between the allocation of functional responsibilities and on the assignment of revenue sources; and the degrees of regional discretion attached to the management of each. The financial relationships between central and subnational governments are characterized by three different approaches: an approach based upon the assignment of income sources, an approach defined as the expenditure approach, and a comprehensive or a deficit approach.

3.1 General Considerations of Local and Regional Taxation

Regional revenue sources have been briefly classified in the previous chapter. Following chapters examine them in greater detail, comparing ways in which particular revenues are exploited by local and regional government in different countries. This chapter deals with taxes exploited by subnational governments. It will also look primarily at alternative forms of local and regional taxes, their advantages and disadvantages. Taxes will be classified in three groups according to the object of taxation:

2) 3) 4)

taxes on property taxes on income taxes on goods and services

As discussed in Chapter 2, "regional and local taxation" can be interpreted in four ways: 1) taxes which subnational governments impose by their own legislation; taxes levied under national legislation but with tariffs determined by regional or local authorities (either freely or within statutory limits); taxes which subnational governments assess and/or collect; taxes which are levied and administered by central government but whose proceeds are given to, shared with, or surcharged by subnational governments.

Discussion in this chapter is concerned with all these categories. A tax source is of interest if local and regional authorities derive revenue from it, whether or not they levy or administer it, or determine its tariff. But the discussion is also concerned with the suitability of taxes for levying or administration by subnational governments, since they so often wish -or are expected - to play a part in raising the money they spend. To appraise the potential of taxes as subnational government's revenues, criteria are needed. These will be discussed in the first part of this chapter together with certain general issues of tax delegation and administration. Individual types of local and regional tax will be examined according to general criteria in the second part of the chapter. Criteria will be discussed under five headings. Four - adequacy and elasticity, equity, administrative feasibility, and political acceptability - apply to taxes whether they are centrally or regionally levied. The fifth concerns the suitability of a tax for imposition and administration by local and regional rather than central government. 3.1.1 Adequacy and elasticity The first and most obvious requirement for revenue sources is that they should be adequate to meet the costs. They have to yield substantial income - substantial in relation to the cost of the services which they are intended to finance, wholly or in part . It is not uncommon for subnational governments to be allowed a wide range of taxes on the one hand, but on the other hand the yield from them is not very important for subnational governments and represents a minimum percentage of their expenditure budget. This has many disadvantages: collection costs may well be high, administrative effort is diffused, equitable incidence is difficult to maintain. It is usually desirable to concentrate effort on taxes which can raise a substantial proportion of the cost of the services to which they are devoted. These costs are rising for a variety of reasons, as mentioned in Chapter 1. Inflation increases costs, often disproportionately if regional services are labour-intensive and public sector wage rises are running ahead of the general rate of inflation. Population growth -particularly significant in Third World countries in general and the larger cities in particular -increases the demand on services, employment creation, and infrastructure. The growth of the larger cities is often associated with diseconomies of scale - higher leakage as water has to be pumped over longer distances, greater traffic pressure on the existing city centre roads as well as an expanding suburban network, more sophisticated fire-fighting equipment for high-rise buildings. Rising standards of living associated with national economic growth lead to demands for higher standards of services. There is no point at which expectations become absolutely fulfilled (once universal primary education is provided, there is pressure for more secondary schooling; if every child has primary and secondary education assured, nursery schooling is demanded). Much of the improvement in the expansion of services is scheduled in national development plans. If costs are rising, revenue must rise also. It is highly desirable at least from the point of view of government-that taxes should exhibit elasticity, i.e. that their capacity to yield additional revenue should respond to the same pressures as give rise to increasing demands upon government expenditure, that the tax base grows automatically, e.g. when prices rise, when the population in an area expands, or when personal incomes increase. Elasticity in this context has two dimensions. The first is the growth of the potential tax base. The second is the ease with which the tax exploits that growth. The different responses of ad valorem sales and property taxes to inflation illustrate the point. (In times of rising prices the tax base for sales tax - the gross value of taxable sales - increases. If the tax is a percentage of the price of goods or services, its yields will grow automatically in pace with the tax base). Inflation usually increases the values of property or market rent levels; this growth in the potential property tax base is only exploited, however, if tariffs are raised or properties re-valued by the tax administration. In the one case elasticity is automatic; in the other it is dependent upon sensitive discretionary decisions or heavy demands upon administrative capacity.

Elasticity is an important quality in a tax source. It is also easily measurable by comparing yields over a number of years with changes in price indices, population, or the gross domestic product. An alternative measurement of elasticity would be to compare, over a period of time, the tax base per capita in real terms (i.e. discounting inflation) with changes in real per capita income. It is a better test of elasticity than actual tax yields since it reflects an automatic growth in tax potential independent of discretionary decisions to change tariffs. 3.1.2 Equity A second major criterion is equity - the principle that the burden of maintaining public expenditure should be borne by sections of the community in proportion to their wealth and their ability to contribute. This is a concept of social justice. As a concept, it implies that taxation is an instrument of redistribution, the rich contributing more than the value of public services to themselves, and the poor less. In practice, this too can be diluted if the rich enjoy disproportionate benefits from public expenditure. Equity in terms of regional and local taxation has three dimensions. The first is vertical - the relationships of tax incidence to differences in income levels. By widely accepted standards, a tax is good if it is "progressive", i.e. if the percentage of a person's income paid in tax increases with the level of that income. It is tolerable if it is "proportionate", i.e. if the percentage of income paid in tax is constant at all income levels. A tax is bad if its incidence is "regressive", i.e. if the percentage of income paid in tax declines with rising income levels. The second dimension of equity is horizontal - the relationship of tax incidence to the source of income. A salaried man should not pay more tax than one with an identical income derived from business or agriculture; a farmer who grows an export crop should not pay more than one with an identical income from domestic food crops. The third dimension is geographical. The incidence of taxation should be fair as between the populations of different geographical areas. People should not be taxed more heavily simply because they live in one area rather than another (something particularly likely to happen on the borders of a town). Equity has to be seen, however, in relation to both revenue and expenditure. It may not necessarily be unfair to tax people more highly if they live in an area which has an exceptional level of government service. Indeed, the whole argument for giving regional and local governments discretion to vary levels of taxation is to enable them to use different equations of tax burdens and service standards. Similarly, it would not be unfair to tax a particular income source more heavily if its generation was aided by an exceptional level of public investment. The equitable incidence of a tax is affected by its coverage - who pays and what type of income, wealth, business, etc. is subject to taxation - and by its rate structure. I t is also dependent upon the assessment methods and the degree of accuracy with which individual liabilities are calculated. Any inaccuracy in assessment causes injustice because people will be paying more or less tax than they should. But how much such injustice hurts depends upon the tax rates. Difficulties in assessing tax liabilities extend also to the award of exemption. Many taxes (and, to a lesser extent, charges) include some exemption from liability for groups such as the old, the young, the poor, or the physically handicapped. Reductions in liability may also be given according to the number of dependents supported by the taxpayer. The extent of exemption depends considerably upon the ability to identify those eligible. Industrialized countries may be able to rely substantially on documentary evidence of age, means, family circumstances, etc. In many countries, such records may be largely absent or unreliable. The decision then rests largely upon local knowledge of claimants and their circumstances; local tribunals are often used to arbitrate claims - a process prone to partiality. The equity of taxes is not always compatible with the other criteria governing their effectiveness. Administrative feasibility and the ease or difficulty of evasion also imposes severe limits on horizontal equity. What matters most is the distribution of the total tax burden upon society. The apparently inequitable implications of one tax may be compensated by the incidence of another. It is particularly important in this respect to see subnational taxes, not in isolation but as part of the national fiscal scene. 3.1.3 Administrative Feasibility The third major consideration in the assignment of revenue is administrative capacity. Revenue sources vary in the amount of skill, integrity, and determination required in their administration. They also vary in the amount of time and money involved in assessing and collecting them, compared with the yield. In many developing countries the majority of the population is self-employed smallholders, petty traders, or working for "small informal" sector businesses without clearly assessable income. Assessment and collection of any direct income or property tax involves personal visits. The administrative costs of such an operation are very high, though the average amount that can be collected per capita may be low. On the other hand, very substantial revenue may be collected through a duty on petroleum, for example, at negligible administrative cost. In such economies there is a heavy bias of administrative convenience towards dependence upon indirect revenues which can be levied on formal sector commercial transactions through large manufacturers, importers, or distributors. This is not always consistent, however, with considerations of equitable incidence or fiscal decentralization. 3.1.4 Political Acceptability No tax is popular. Some taxes, however, are more unpopular than others. Political will is needed to impose taxes, to determine a rate structure, to decide who should pay and how they are to be assessed, to collect taxes physically, and to enforce sanctions against defaulters. This, in turn, depends upon two factors: the sensitivity and visibility of the tax and the occurrence of discretionary decisions. Sometimes political sensitivity focuses on particular questions of social values (e.g. should you tax land, or charge for water) or of sectional interests (of landowners, civil servants, traders etc.). More generally, taxes are usually less sensitive politically if they are indirect (concealed) and do not involve too many overt political decisions such as a parliamentary/council decision to raise the rate of tax. Political sensitivity is a constraint upon the potential yields of a tax. It may be good for accountability, however. The need for a discretionary decision to increase a highly visible tax may force public bodies to think more closely about the justification for particular expenditures or the elimination of waste.

To make taxes more acceptable often it is made through "earmarking", i.e. by raising a levy to pay for a specific service which is popular, such as education. Any taxes levied by single-purpose bodies, such as water authorities or school boards, will automatically assume this character. A direct connection between the scale of taxation and service expenditure also stimulates accountability. On the other hand earmarking can be counterproductive. Earmarking can lead to an undesirable broadening of taxes which is costly not only in time and manpower but also in the taxpayer's patience. A multiplicity of taxes makes it even more difficult to maintain an equitable incidence upon the community. 3.1.5 Tax Administration by Subnational Governments The criteria already discussed concern the effectiveness of any tax source, whether national, local or regional. It is also necessary to consider whether a particular tax can be effectively exploited - levied, assessed, or collected - by subnational government as opposed to central government. Several issues arise. Local and Regional Imposition of Taxes and Tariffs The first issue is whether subnational governments have sufficient political will to levy a tax effectively and equitably. This relates to decisions both to impose a tax and to vary its coverage or tariffs. The "development from below" theory argues that people will more willingly pay tax to a local than to a national government, because they can see the benefits in the amenities and development which accrue to their locality. The lower the level of government, however, the closer is those who impose and assess taxes to those who pay them. This can result in corruption behaviour of local politicians and civil servants, effort not to carry out politically unpopular steps (i.e. to increase taxes when it is needed), or on the contrary to seek popularity by reducing taxes. It is necessary to point out that allocating a particular tax source for imposition or administration by a regional authority raises the question whether the general problems of its political acceptability are more acute at this level of government. Regional Variations in Liabilities and Tariffs Freedom to determine the coverage, assessment methods, and tariffs of their own taxes clearly enhances the financial discretion and flexibility of subnational governments. This raises the question whether the administration of a tax considered for devolution can cope with variations in liabilities and rates between regions. The problem arises chiefly where individuals or organizations are paying taxes to more than one region. The degree of difficulty involved clearly depends upon the number of regions, the extent of the variations, and the technical sophistication of the tax administration. A related question is whether regional and local tax variations have distorting effects on the economy and how harmful these might be. But those distortions are not unacceptable in themselves; it is part of the rationale of decentralization that regional communities should share in deciding upon levels of public expenditure in the light of their economic implications. The degree of distortion and the severity of its consequences for overall management of the economy are what matter. Regional Assessment and Collection The next question is whether regional government has the administrative capacity to administer a tax effectively. Responsibility for assessing or collecting a tax or charge does not necessarily coincide with effectiveness. A central government agency may well undertake or assist in the assessment or collection of a regional or local tax. Conversely, regional authorities may assess or collect revenue on behalf of the central government. The allocation of responsibility for assessment and collection depends upon a number of factors. One is the degree of professional skill required and its availability at any particular level of government. Tax administration may also require a widely diffused network of assessors and collectors with physical proximity to a scattered population and access to taxpayers' records and pockets. This is particularly important to any largescale direct taxation in Third World countries. Three other factors relate to the characteristics of the taxes themselves. There are problems in taxing the income of individuals or corporations on a regional basis if these derive from more than one region. Centralized assessment may also be desirable if there is a need for national uniformity. There are strong advantages in joint assessment and collection of any levies imposed by two or more levels of government on a common tax base, as with the federal and provincial income taxes in Canada or the central and local income taxes in Scandinavia. Regional Tax Enforcement Assessment and collection have to be supported by an effective system of enforcement. Delay in paying a tax or charge is often penalized by a percentage surcharge on the amount outstanding. Sanctions for known failure to pay altogether take varying forms including: 1) criminal proceedings involving a liability for arrest and imprisonment; 2) civil proceedings analogous to those for the recovery of a private debt, leading to seizure and sale of property;

3) 4) 5)

direct seizure and sale of property; withholding of public services, e.g. cutting off water supplies, telephones, or electricity, eviction from public housing, refusal of schooling or medical treatment; or complete inaction.

Criminal proceedings, withholding services, or direct seizure of property are usually effective encouragement to payment. They all involve administrative energy and impartiality and a fair degree of political will. Assignment of Tax Proceeds to Regions The first issue is whether it is clear to which authority a particular tax liability is due. This is rarely a problem with property taxes since a plot or building is clearly located within a particular jurisdiction. The most difficult question arises over where the liability attaches to operations carried on in several regions - for example, taxes on corporate incomes where profits may be derived from operations in several locations, or sales tax imposed on a company with branches in several regions.

Frequently, the point at which a tax can be collected differs from that at which it is effectively paid. This problem may arise where income tax is collected by an employer whose headquarters is in one region, from the wages of an employee who lives (and possibly works also) in another one. Whether a collecting region should retain tax proceeds depends upon their purpose. It is usually held that the area in which a person lives (and a tax is intended to finance the infrastructure or utilities) should get the tax. Any discrepancy between the region where a tax is collected and the region to which it is due requires a system of remittance to overcome misallocation. Remittance requires an efficient administrative process; it also needs dependable sense of fairness on the part of the authority which should be giving up the money. Relationships between Tax Liabilities Regional administration of taxes raises questions whether tax payments to one level of government affects the payer's liabilities to another level. There are three alternative patterns. In the first case, liabilities to central government and different levels of subnational government are entirely separate and no allowance is made by any authority for any parallel taxation by another. The second situation is where taxes paid to other authorities are deductible from the total net income subject to income or profits taxation. Businesses can almost invariably count property and other regional taxes as expenses deductible from their income before it is assessed for corporate income or profits tax. Individual income tax payers are not usually so fortunate. (For example both income and sales tax payments to state and local governments are deductible from personal income for federal taxation in the United States). The third case is where payment of tax to one authority earns credit against liability to another, i.e. where the amount of tax paid to authority A actually reduces that payable to authority B by an equal amount. It is not widespread and there are few examples (local authority taxes on business and personal incomes are credited against liability for state income tax in some American states, notably New York). Economic Efficiency Taxation has a dual purpose: to provide money for public purposes and to influence economic behaviour. Taxes affect the cost of individual decisions - a property tax affects the profitability of building and renting a house, a sales tax affects the cost of buying a shirt, an entertainment tax affects the cost of going to the cinema. Economic efficiency criteria are generally more important in appraising national than local taxes. There are two reasons. Firstly it is usually central governments who are principally responsible for overall economic management and who use taxes to manipulate economic behavior. Secondly the scale, of local and regional taxes is usually insufficient to make a significant difference to people's choices. But it must be nevertheless of concern whether local and regional taxes will have a significant effect - harmful or helpful - on local and regional economic behavior. 3.2 Property Tax Taxes on property are the most common form of direct revenue for local and regional authorities. Land is taxed by urban local governments in nearly all countries, buildings in most. The clear physical connection between a piece of land or a building and its locality, or between property and the local physical services to it, gives such taxation an obviously local character. They are normally general purpose revenues contributing to a range of urban services, particularly the physical infrastructure such as roads, drainage, refuse collection etc. Property tax valuations, however, are also used frequently as a base for special taxes or charges to meet the costs of individual services such as water supplies or street lighting. Taxes on land and/or building are normally imposed on all kinds of property -commercial, industrial, institutional and residential. There are usually some exemptions, often including places of worship and charitable institutions. Government properties are customarily exempt, but a grant is often paid to local authorities in lieu of the tax. Some countries exempt small properties below a stipulated value; this may be a legal exemption or simply an administrative choice which ignores squatter hutments etc. Property taxes are normally assessed by applying a tariff (a percentage) to a valuation. The tariffs may be prescribed by law or left to the discretion of the local authority. Such discretion may again be restricted by statutory minimum or maximum levels. The tariff in any one locality is normally a uniform percentage; however, it may differentiate between types of property, e.g. being heavier on industrial or commercial than on residential properties, or higher on rented building than on those occupied by the owner. Some countries also impose progressive tariffs, the percentage rising with the value of individual properties. 3.2.1 The Basis of Valuation Valuation for tax purposes may be based on:

1) 2) 3)

annual rental values, i.e. the potential income to the owner from renting a property (whether it is actually rented or not); capital improved site values, i.e. the potential market sale price of the land or the land and its improvements (meaning the buildings, infrastructure and amenities developed on the land); unimproved site values, i.e. the potential market sale price of the land as if it were vacant, disregarding any actual development on it.

1) Annual Rental Values Rental values2 are normally calculated on the basis of "a fair rent" after deducting a standard percentage representing the notional cost of repairs and maintenance. I n theory, rental values are closely linked to income and therefore provide a fair and readily explained basis for taxation. This assumes however that there is a substantial and fair rental market, and that the valuers have clear evidence of the prevailing rents levels. Rental values may be much harder both to calculate and to justify where: the private rental market is relatively small (e.g. in Britain)

rent levels are subject to violent fluctuation or severe distortion by conditions of scarcity

ineffective rent control legislation causes a substantial disparity between the official rents used for purposes of valuation and those really paid by tenants. The most obvious way of assessing rental values is to extrapolate from a large-scale of actual rents paid for different types of property in different localities. Where such evidence is difficult to obtain, rental valuations are often based upon a notional percentage return upon capital values. In such cases the distinction between annual rental and capital value taxation becomes somewhat artificial. 2) Capital Values Assessment of capital improved site values may be based upon one of three methods of calculation:

a comparison of actual sales of different types of property in different localities

a capitalisation of rental incomes (assuming a current market rate of return upon investment of capital but allowing for depreciation) current costs of the land plus current replacement costs of the buildings. The choice is clearly dictated by the availability of evidence. The third alternative may be preferable where types of development in a given location are very heterogeneous, making it difficult to isolate the impact of different factors on value, it may be difficult to apply, however, in highly built-up areas where there is a little land changing hands. 3) Unimproved Site Values In some countries, capital valuations are based upon "unimproved site values", i.e. the value of the land and its best possible development, not its current state of development. This system has the notional advantage of encouraging owners to develop their land since they will not incur any more tax in doing so. Indeed the taxation penalizes any underdevelopment. It also ensures that local authorities do not lose revenue through any lack of potential development. These potential advantages may not be realized in practice for a number of reasons: property tax liability is only one, and rarely the most important of the factors which influence the investment decisions of land owner Unimproved site valuation requires evidence of "the best potential development" of a particular plot. This is hard to establish; The incentive element can only work on the assumption that owner either develops his land or vacates it. Effective application of this principle would lead to the concentration of desirably located property in the hands of those with the greatest capital. The application of unimproved site valuation is often modified in practice to offset the difficulties and potential hardship which have been described. This means assessing property below the value of its "best potential development" or keeping tariffs at a low, effective rate. Either method, however, restricts the real incidence of property tax on the more expensive developments, leading to a considerable loss of potential revenue. Some countries have attempted to find reconciliation between the advantages and disadvantages of this system by imposing a dual tariff on the value of a site and of its improvements, with a heavier weighting on the site value. The Process of Valuation The valuation process starts with collection of comparative information on rents or sale prices. The most common method is extrapolation from comparative sales. Ideally, such calculations should comprise a matrix including the following considerations:

size of plots, with weightings for depth and frontage topography (particularly for industrial/commercial uses) actual land use or permitted land use under planning legislation the form of land tenure

standards of construction, including types of materials, age, facilities (lifts, sanitation, heating, garages, etc.), quality of finish and number of floors (which may be treated as an increasing or reducing factor according to planning policy) location, including infrastructure, amenities (or, negatively, nuisances) and general neighborhood quality.

Such a matrix provides a set of coefficients which are then applied to the square metre age of the land and/or the buildings. The range and degrees of variations taken into consideration (and, thereby, the accuracy of the whole valuation) must, however, depend upon the technical capacity of the valuation staff. It is normally desirable to compare evidence on current values from a variety of sources such as official records of land transfers, owners" declarations, estate agents and brokers, and local administrators. Valuations clearly get out of date rapidly and a regular system of general reassessment should operate to allow both for inflation and for changes in the relative values of different properties due to improvement or decline of neighborhoods. Most tax systems provide for regular reassessment at intervals of up to ten years. A few countries attempt to maintain the real value of the property tax base by some form of indexation. 3.2.2 Appraisal with the General Criteria In large-sized towns, revenues from property taxes are frequently substantial, stable and predictable. A World Bank survey of fourteen major Third World cities showed that property taxes in the mid-1970s represented over half of local government revenue in four of them and over one-third in five more. The tax base does not fluctuate in face of short-term changes in the economy and revenues are immune from immediate cyclical phenomena. To this extent, property tax is valuable part of urban finance. The only exceptions relates to a few European countries, Czech Republic and the Scandinavian block in particular, where property taxes has declined below 5 percent of income of local governments. See the Figure 3.2. The elasticity of property taxes is poor in relation to growth in prices, population and incomes. Real increases in rents or market values are only reflected in the growth of the tax base when revaluations are carried out - often an infrequent process, inefficiently executed. In between revaluations, revenues can only be kept in line with difficulty. Automatic indexation of taxable improves buoyancy, but this practice does not obviate

the need for regular revaluation to maintain equity. Population increase does give rise to increase in the tax base, but this is rarely proportionate with the rate of population growth since much of this is absorbed in overcrowding existing buildings or illegal construction. Property taxes are crudely equitable in that there is usually a rough correlation between the income of individuals and businesses and the value of the premises they occupy. There are, however, a number of factors which impair equity. Market forces may result in considerable differences in rental or capital values which depend upon location rather then the quality of the premises or the incomes of the occupants; The rental or purchase costs of housing or business premises upon which the tax is based, may represent a higher percentage burden on the income of the poor than on the rich; Assessment methods may include an inequitable bias; Differences between business firms in the size and value of the premises they occupy may reflect the type of business more than its profitability (selling cements needs a bigger shop than selling jewellery). The great administrative merit of property taxation is that liability is, or should be obvious. One cannot conceal a plot or a building. Even so, considerable effort is necessary to keep cadastral surveys and tax rolls up to date. Industrial surveys are expensive in themselves. The biggest administrative problem with property taxation is assessment. T h i s requires technical skill which is often in short supply, though pooling valuers at national level or using consultants may help. Secondly, the process often involves a high degree of personal judgment and of contact between assessors and taxpayers which offers wide temptation to collusion in under-valuation. Simpler systems which limit site visits to physical measurement and apply standard formulae, perhaps by computer, may be more effective. Collection of property taxes can also prove difficult if liability attaches to elusive owners rather than occupants, if records are not kept out to date and if sanctions are not punitive or strictly enforced. Property taxes are highly sensitive politically. There are two reasons. Firstly, they have to be collected directly from the taxpayers, and are therefore even more overtly seen as a burden than income taxes which may be deducted by employers. Secondly, any increase in tax, whether by tariff revision or revaluation, is normally dependent upon a deliberate political decision by ministers, members of Parliament, mayors or councilors. Property taxes are usually levied by a locally elected assembly. Individual property owners or occupiers are likely to be the major part of the electorate. Property taxation is often debated in terms of its incentive effects on land use and constructions. As already mentioned, unimproved site value taxation is frequently recommended - and occasionally adopted - to encourage high value development and, conversely, penalise underdevelopment. It is likely that the supposed incentive and disincentive effects of property taxation in most third world cities are considerably exaggerated, simply because of the low ratio of the tax to the other costs and incomes which influence an owner's choice over development. 3.3 Taxes on Income Income tax is almost invariably a major national government tax source,jealously guarded by the central treasury. The sharing of this resource by assignment or dual taxation is sometimes prescribed by federal constitution. There are a number of ways in which local and regional governments in some unitary countries, also, derive revenue from the taxation of income. Some related to the national tax system and some to separate tax bases. Participation by subnational government in income taxation takes a number of forms. Three relate specifically to the national systems by duplication, surcharge, or sharing. Two other types - taxes on occupations and mass personal taxes - operate on completely or largely separate bases. Separate Local and Regional Income Taxes In a few cases, subnational government are permitted by law to impose a tax on income which is levied, assessed and collected by themselves, and which is parallel to, but distinct from, the national income tax. An entirely separate tax administration is used and calculation of the local liability is not related to the national system of rates, personnel allowances, or exemptions. Local authorities in Finland have this power. However, the main example is the United States where the Federal and State Governments have concurrent powers of taxing income; in several states, the cities and countries are also permitted to levy a separate income tax, so that in New York, for example, a citizen will be paying a separately assessed and collected tax on income to three levels of government. Surcharge on National Income Tax A more common system for subnational governments is to impose a surcharge on the national income tax, i.e. a percentage addition to the tax liability as assessed and collected by the central government. By agreement with the federal government, provincial governments in Canada levy tax in this manner. Local authorities in Japan and a number of European countries - Belgium, Denmark, Italy, Norway, Sweden and Switzerland have this power. A rate of surcharge maybe: fixed by law, as in Switzerland; fixed by each local authority within a range prescribed by law (as in Belgium, Italy and Norway); prescribed by each local authority without limit as in Denmark, and by each of three levels of local government in Sweden. The surcharge is normally a flat percentage rate charged on taxable income (i.e. net of personal allowances, etc), but a progressive tariff operates in Switzerland. Sharing of National Income Tax In several countries the central government assigns to subnational government part of the national income tax. In Germany income tax is split three ways between federal, state and local governments. In the Czech Republic income tax is part of the general pool of national revenues from which regional and local government are assigned a share. A point of variations is whether each authority's share is a percentage of the tax derived from within region, or whether the regional share is pooled and distributed by some criterion unrelated to derivation. In Czech Republic, for example, a national pool is distributed according to the population and other formula factors, without regard to derivation.

Most tax sharing arrangements cover major taxes such as personal income taxes, corporate income taxes or value added taxes (see Figure 3.1). Their high yield makes them attractive for subnational governments. The procedure for changing the sharing formula is mostly laid down in laws on tax sharing, fiscal equalization or the like. For the OECD countries, decisions on the tax sharing arrangements seem to be taken at the parliamentary level; in some countries the share is defined in the constitution and adjustments require a qualified majority in parliament. Consultation of subnational governments is quite frequent, but their explicit consent for adjustments is needed in some federal countries only. The frequency and regularity of formula adjustment varies across countries, from irregular to never, but it appears that tax sharing arrangements are a comparatively stable item in national fiscal policy. Finally, some countries redistribute tax revenue from affluent to poorer jurisdictions; hence those countries combine tax sharing and fiscal equalization in one single arrangement. A Taxation of Corporate and Business Income Whether local and regional taxation of income applies to both individuals and companies varies widely. The separate income tax levied by Finnish local governments and the regional surcharges upon national tax in Canada, Italy, Japan, Norway, Sweden and Switzerland all apply to individual incomes and corporate profits alike. American state governments impose personal income and corporation taxes under separate legislation. The three-way sharing of German income taxes between federal, state and local authorities applies uniformly to both personal and corporate taxation. Assignment of domestic revenues to regional and local governments in Czech Republic includes corporation tax as well as VAT. In other countries (Denmark, Belgium), on the other hand, local surcharges are confined to personal income taxes. There are several examples of local taxes on the total turnover, profits, or receipts of business enterprises, which are not related specifically to corporate income taxes. Taxes (often a flat rate) on individual types of enterprise such as shops, slaughterhouses, restaurants, and hotels are extremely common. These are sometimes imposed as a "licence fee" with the twin objectives of control and revenue raising: they effectively constitute a tax if, as is so often the case, the size of fee considerably exceeds the costs of inspection and administration. In certain countries, such as France and India, a tax is imposed upon the pursuit of a particular profession (e.g. medicine or the law), payable to local authorities. The tax may consist of a percentage applied to payroll or turnover, or alternatively a fixed rate for each profession. In the latter case, the tax operates very much like a licensing fee. Country Tax type shared PIT, CIT, property tax, VAT PIT, CIT, VAT PIT, CIT CIT PIT, CIT, VAT Transaction and specific service taxes VAT, excise duties PIT Figure 3.1 Tax sharing arrangements Procedure for Frequency of Horizontal formula changes formula changes equalisation objective Parliament, Law on Every four years yes Fiscal Equalisation Irregularly yes Government, Law of Tax Assignment Government, Law Very rarely no on Tax Sharing Government, Law no on Tax Sharing Both Parliaments 13 changes since yes (Bundestag and 1970 Bundesrat) Rarely no Central government Parliament Parliament, Law on Fiscal Equalisation Rarely Never since 1959 no yes

Austria Czech Republic Denmark Finland Germany Greece

Spain Switzerland

Note: PIT = Personal Income Tax, CIT = Corporate Income Tax, VAT = Value Added Tax Source: Blchliger, H., King, D. (2006). Fiscal Autonomy of Sub-central Government,OECD
Working Paper, No. 2, p. 17 Mass Personal Taxes In some third world countries, subnational governments have levied a general personal tax on the lower segments of income which fall below the point at which liability for national income tax begins. This has operated in countries where only a relatively small minority pays national income tax, partly because most incomes do not reach the lower exemption level and partly because the national income tax administration is not ready to cover a mass of self-employed smallholders, farmers, small traders, etc. Regional personal taxes normally attempted to cover the entire adult population and have been characterised by: a basic minimum rate of tax for which every able bodied adult is liable, regardless of income; a fairly low maximum tax rate acting as a cut-off point below the incidence of national income tax; a series of "stepped" rates between the minimum and maximum applying to ranges of income; the use of crude, "imputed" or "presumptive" income criteria to estimate earnings from common sources of livelihood such as cultivation, livestock ownership, trading, etc. The regional personal taxes have operated mainly in tropical Africa where traditions of communal land ownership and shifting cultivation have made land taxation an unsuitable base for mass taxation. 3.3.1 Appraisal with the General Criteria

On a nationwide level, income tax is potentially an extremely lucrative source of revenue, able to provide substantial and increasing yields in a fair and progressive manner. It is clearly much easier to realise its potential in a society in which the majority of people are wage employees than in one where farmers, small traders, self-employments and casual labourers predominate. National income taxation is far more buoyant, certainly so far as inflation and economic growth are concerned, and this would benefit local authorities which exploit it through surcharging or tax sharing. Income tax normally responds well to inflation. Whether it is responsive to population growth depends upon the structure of employment and the extent to which income taxation extends to the informal economy. The adequacy and elasticity of income tax as a regional and local revenue clearly depends upon the nature and the extent of subnational government participation in it. Surcharges upon, or assignments of national income taxation clearly derive full benefit from its scale and growth. Income taxation is potentially the fairest form and can most readily be made progressive. The problem, however, is to achieve horizontal equity, i.e. to achieve equal taxation of people with equal incomes but from different sources. Assessing salary earners is easy, assessing the self employed more difficult. Two devices are common - self assessment or the use of presumptive income criteria. Income taxation is relatively simple to administer where the majority of taxpayers are wage and salary earners of subnational governments, if employers are required to deduct at source. Assessment of and collection from the self employed is far more difficult and collection costs can be high. Income taxation is sensitive insofar as it is also a direct tax, but collection at source mitigates this to some extent, and increases in payment take place automatically as incomes rise. Taxation of the self employed, however, is more sensitive since much argument may arise in the assessment process, and effective collection may depend upon a degree of "muscle" which may be perceived as harassment. "Regionalizing" income taxation raises major issues of assignment. In terms of personal taxation, the question arises whether tax should go to a man's area of work, residence, or origin. The generally accepted principle is that personal income taxation should accrue to the authority of the region in which man lives, because services to the family as education and health make the biggest call upon regional government expenditure. If regional authorities are directly elected, it is also right that taxation should "go with representation". Problem arises where taxes are collected in different region from that in which the taxpayer resides. Changes in residence also cause the problem. The most difficult question arises over taxes on corporate incomes, where profits may be derived from operations in several locations. I t i s usually accepted thatany regional taxation of business profits should accrue to the region where the business take place, rather than that in which the owners reside, in view of the contribution of local infrastructure to its operations. Countries with local and regional taxes on corporate profits have developed formulae for assigning these to different regions but the process is complex. A second set of administrative problems surrounds regional variation in assessment rules or tax rates. Differences in rates of local income taxation impose more difficulty where deduction is made at source. Employers, companies paying dividends etc, may be severely inconvenienced in complying with different tax rates, though this is not impossible given the increased use of data processing equipment. 3.4 Taxes on Expenditure Some definitions of this category include taxes on consumption, production, the right to trade, possessions, occupations and advertisement. A t t h e subnational level this category comprises a wide range of local sales taxes, royalties, entertainments taxes, vehicle taxes etc which are often described as "indirect taxes" because they can usually (but not invariably) be passed on by the payer to someone else, normally the end user. As with income taxes, subnational governments may have access to this tax field by direct levy, by surcharge upon a national level tax, or by assignment of, and share it. Sales of Goods General taxes on sales of goods are normally levied by central governments, or by the state level in a federation. A sales tax is very rarely a local authority levy. There area few exceptions. About 3000 local authorities in the United States impose a surcharge on state sales tax, and Madras Municipal Corporation receives part of a similar surcharge. In the Netherlands and Philippines municipalities receive a share of national sales tax, and Brazilian municipalities receive 20% of value added tax (Davey, 1983). A local surcharge on petroleum sales taxes operates in some cities in Japan and both North and South America. A few individual goods attract specific taxation on sales apart from general taxes, usually at much higher rates. These are mainly taxes on liquor, petroleum fuels and tobacco. These are rarely regional levies. Sale of Services There is a variety of local and regional taxes on sales of individual services. Most common taxes are on: entertainment, including cinemas, theatres, sports, fairs and festivals (normally by a percentage surcharge on the price of admission); hotels and restaurants or night clubs, taxes on gambling and casinos,

public utility bills, e.g. electricity, gas and telephones.

Taxes on Production Excise taxes on manufactured goods are customarily national revenues, and geographical concentrations of industry make them generally unsuitable for regional taxation. Some division is encountered in federal states, however (American states, in Brazil, in Pakistan). Royalties are frequently imposed upon the extraction of natural resources such as minerals or timber. Ownership of natural resources, particularly those underground, is usually vested in the nation and central governments are anxious both to assert this right and to distribute the benefit from royalties more widely than the area of derivation. As a result, royalties are normally national revenues, but some percentage may be assigned to authorities in the region of origin, usually small in the case of minerals and larger in that of timber.

Movement of Goods The majority of municipalities in India, Nepal and Pakistan (and formerly in Bangladesh) impose a tax, known by an old French title "octroi" upon the value of goods entering a town for use, consumption or sale. It is collected by the equivalent of customs posts on the borders of cities and towns. Taxation of Vehicles Vehicle taxation is described separately for convenience and because of the substantial revenue involved. Almost all countries impose a substantial annual licensing tax on vehicles which considerably exceeds any registration costs. National governments often claim these revenues because of their size and the argument that vehicles do not stay in one region. Regional authorities often have an equally valid claim upon them insofar as they have to meet the costs attributable to motoring - road construction and maintenance, and parking provision, in particular. A number of countries recognise this and assign vehicle taxes to local authorities. In some countries taxes on motor vehicles are reserved to state or provincial authorities. A few city governments also receive a substantial tax on the initial registration of vehicles or on subsequent changes of ownership. 3.4.1 Appraisal with the General Criteria The major types of tax described above, such as those on sales, motorized vehicles, extraction and production, normally yield very substantial revenues. Where they are allotted to regional government, they frequently constitute its largest revenue source. Elasticity is generally high, since the volume of business taxed should automatically reflect growth in population or personal incomes. Where such takes are imposed ad valorem (e.g. as a percentage of the value of the goods or services involved), they respond automatically to inflation. Taxes on expenditure are perhaps the most buoyant of all, particularly if tariffs are expressed in percentage of prices and not on a unit price basis. The volume of business taxed (or of traffic in the case of vehicle taxation) should automatically reflect growth in population or personnel incomes. Local tax sources such as octroi in India and Pakistan have proved extremely elastic. The equity of taxes on expenditure is difficult to determine because of uncertainty as to who really bears the tax. Insofar as such taxes are passed on to the consumer, it can be argued that there is a rough correlation with ability pay - that equity is preserved since those with higher disposable incomes buy more taxable goods and services. This is reflected in the axiom that expenditure is a reflection of ability to pay. Evidence suggests that expenditure taxes are highly regressive where they extend to mass consumption goods such as food and clothing, since purchase of taxed goods takes a higher proportion of the income of the poor than that of the rich. This applies particularly to taxes on liquor and tobacco. This rarely constitutes a major criticism of local expenditure taxes such as octroi since their incidence on final price is normally insignificant compared with that of national levies. Taxes on expenditure raise very large revenues with very little cost or difficulty where they are collected trough large marketing or manufacturing organisations, or imposed (as with vehicles taxation) through forms of licensing which the public have a strong incentive to obtain. These levies are much more costly and difficult to apply to commercial transactions conducted through small business, markets and the "informal sector". A tax on the movement of goods such as octroi, is relatively simple to enforce insofar as the taxed goods "come to the collector", though this does impose some inconvenience on transporters (often much exaggerated by the transport lobby) and dangers of corruption. Taxes on expenditure are usually less sensitive politically since they are normally indirect and consequently attract less political sensitivity. They may however encounter opposition where they are seen to contribute to price rises which cause discontent, or where an attempt is made to impose high rates on luxury items which affect political and business elites. The issue of "regionalizing" of taxes on expenditure concerns consigning tax yields to the correct region. Sales taxation may be collected, for example at a point of import or manufacture but effectively passed on to a buyer in another locality. Similarly towns which are major retail centres may derive substantial revenue from people who live in neighbouring jurisdictions, though that may not be wholly unfair). The other question is whether it is feasible for regions to vary the coverage and tariffs of their taxes on expenditure. Since these can lead to significant variations in prices between jurisdictions and, therefore to movements in trade; again, however, one can argue that local authority should be free to make a decision in the light of all its consequences, beneficial or adverse.

Conclusions

Local and regional taxes should be classified into three groups according to the object of taxation - taxes on property, taxes on income, and taxes on goods and services. This chapter concerns all types of tax sources of local and regional revenues, whether or not they levy or administer it, or determine its tariffs and tax base. These taxes include - taxes which subnational governments impose by their own legislation, taxes levied under national legislation but with tariffs determined by regional or local authorities, taxes which subnational governments assess and/or collect, and taxes which are levied and administered by central government but whose proceeds are given to, shared with, o r surcharged by subnational governments. General criteria for appraising the potential of taxes as subnational revenue are - adequacy and elasticity, equity, administrative feasibility, and political acceptability. These concern all taxes whether they are centrally or regionally levied. The last criterion concerns the suitability of a tax for imposition and administration by local and regional rather than central government. The most obvious requirement for tax sources is that they should be adequate to meet the cost of the services which they are intended to finance. Small multiple levies which neither can nor do yield substantial revenues are expensive in money, time and public patience. It is usually desirable to concentrate on taxes which can raise a substantial portion of the cost of the services to which they are devoted. Because of rising costs of services (due to inflation, growing population, rising standards of services), it is desirable that tax sources should exhibit elasticity - their capacity to yield additional revenue should respond to the increasing demands for government expenditure and that the tax base should grow automatically. The criterion of equity represents a concept of social justice - the burden of maintaining public expenditure should be borne by sections of the community in proportion to their wealth and their ability to contribute. In terms of local and regional taxation the question of equity has to be seen in three dimensions: vertical equity, horizontal equity and geographical dimension of equity. The equity of taxes is not always compatible with the other criteria governing their effectiveness.

Administrative capacity required for assessing and collecting different revenue sources and taxes is other consideration. Tax sources vary in the amount of skill, integrity, and determination required in their administration. In many developing countries assessment and collection of taxes (particularly on property and income) is very time consuming and administrative costs are very high. Political acceptability of the taxes depends upon the sensitivity and visibility of the tax and the occurrence of discretionary decisions. Taxes are usually less sensitive politically if they are indirect (concealed) and do not involve too many overt political decisions to rise the rate of tax. In additional to general factors, tax administration by subnational governments raises specific questions of feasibility. Some relate to the availability of administrative skills and capacity to administer a tax effectively (including the system of tax enforcement). Another issue is whether it is clear to which authority a particular tax liability is due; whether the feasible point for collecting a tax is necessarily the location where it is effectively paid; whether local variation in tax rates or assessment rules is required; whether tax payments to one level of government affects the taxpayer's liabilities to another level. The most widespread base for local and regional taxation is property - the value either of land alone, or of land and buildings. Taxes are normally imposed on all kinds of property -commercial, industrial, institutional and residential. The clear connection between such property and a given locality or the supply of physical services such as roads and drains, give it an obvious "local" character. Property taxes are normally assessed by applying a tariff (a percentage) to a valuation. Valuation may be based upon annual rental values, capital improved site values, unimproved site values. The most common method of valuation process is extrapolation of information from comparative sales which consider the various characteristic of property and set a matrix of coefficients which are then applied to the square metre age of the land and/or the buildings. Revenues from property taxes are frequently substantial, stable and predictable. T h e y a r e highly politically visible, a r e inelastic in relation to inflation, population growth or rising incomes, causing pressure on public spending. These weaknesses have led to a widespread decline in the percentage contribution made by property taxation to the support of urban services. In many European countries this tax is quickly losing its role as a significant source of urban revenue. Usually income tax is the major national tax source. Participation by subnational government in income taxation takes a number of forms. Three relate specifically to the national systems by duplication, surcharge, o r sharing. Two other types - taxes on occupations and mass personal taxes - operate on completely or largely separate bases. A more common system for subnational governments is to impose a surcharge on the national income tax, i.e. a percentage addition to the tax liability as assessed and collected by the central government. Local or regional taxes on corporate and personal income - b y surcharge - are major sources of revenue for subnational governments in some European countries (especially in Scandinavia), North America, Japan, and South Korea. In several countries the central government assigns to subnational government part of the national income tax. Variations are whether each authority's share is a percentage of the tax derived from within region (e.g. Germany), or whether the regional share is pooled and distributed by some criterion unrelated to derivation (the Czech Republic). Income taxation is potentially the most lucrative source of revenue, characterized by a high degree of elasticity and equity. The very attractiveness of income taxation creates obvious problems of competition with the central government. Any form of regional income taxation beyond a straight percentage of national tax yields, presents strong administrative challenges, particularly in a rural economy dominated by self-employment. At the subnational level the category of taxes on expenditure comprises a wide range of local sales taxes, royalties, entertainments taxes, taxes on hotels and restaurants, public utility bills, vehicle taxes. Subnational governments may have access to this tax field by direct levy, by surcharge upon a national level tax, or by assignment of, and share it. Taxation of goods and services is a lucrative form of revenue upon which reliance is particularly strong in countries where more direct taxation faces critical problem of assessment and collection. It is valuable revenue source for regional governments (state and provinces) in federal countries; in unitary states local authorities get some revenues through assigned shares or surcharging. There are several cities which draw major revenues from taxes on goods, services and general business turnover. More widespread are taxes like those on entertainments.

Taxes on goods and services are very substantial revenues, w i t h a v e r y high elasticity, t h e equity is difficult to determine because of uncertainty as to who really bears the tax, expenditure taxes raise very large revenues with very little cost, and are usually less sensitive politically since they are normally indirect, and consequently attract less attention. Regional administration of the major sales taxes and excise taxes creates problems such as differences between the points of effective payment and collection, strain upon administrative capacity, and the effects of tariffs variations upon price levels. These may only be tolerable when they are operated by very large regional units such as American state governments.

4. Charging - General Principles and Practices

This chapter:

introduces the general theoretical principles for charging describes some problems associated with distinguishing whether particular service should be financed through taxes or charges describes the range of services for which user charges are imposed, discusses variations in the extent of direct cost recovery (charging full costs of the service, charging below and above the costs) sets and appraises the charges with general criteria as adequacy, elasticity, equity, administrative capacity and political acceptability attempts to submit some general remarks on charging at local and regional level (categories of services, significance of charges for local revenues). Part at least of the financial resources of local authorities shall derive from local taxes and charges of which, within the limits of statute, they have the power to determine the rate. (European Charter of Local Self-Government: Article 9) 4.1 Theoretical Perspectives for Charging

Some public services are financed by general taxation and others by direct charging to the consumer. In the one case, every taxpayer has to contribute regardless of whether the service is immediately available to him and how far he uses it; in the other, payment depends upon the service being provided and the use made of it. In practice, this distinction is not always so evident. First, some imposed charges may greatly exceed the cost of the services rendered, the surplus being appropriated to a general revenue pool (as opposed to extending the service or subsidizing other users). This is typical of many charges for regulatory purposes - licensing fees, for example. Such "charges" are effectively taxes. Secondly, a service may only be partially financed by charges, the balance coming from specific or hidden subsidies from general tax revenue. Which type of financing - taxing or charging - is appropriate to a particular function? What distinguishes a service funded by taxes from one charged to the consumer? The conventional answer is to distinguish between "public" and "private" goods. A public good is a service which benefits everyone collectively and indiscriminately, like defence or disease control. To treat a man for smallpox is in the interests of everyone, benefiting not only the patient but those whom he might infect. Such services deserve obligatory financing by every taxpayer in relation to his means and not to his consumption (which may, in any case, be unquantifiable). According to a Bailey (1999, p. 126) pure public goods have to be provided free at the point of use because their non-excludability means that payment cannot be secured from "free riders" and nobody's use of service is rival with that of other users. In contrast, pure private goods are both rival and excludable and charges should fully reflect the costs of their provision if allocative efficiency is to be secured. In the other worlds (Davey, 1985) a service is a private good if one man's consumption benefits him alone and not his neighbours. Whether a householder has electricity supply does not affect his neighbours, and there is no reason for them to contribute to its cost; a direct charge is appropriate. The distinction is not always easy to apply in practice. There are several reasons. First, the boundaries between public and private goods are difficult to define. Very few local and regional government services qualify as public goods, most being categorised as mixed goods with positive externality or merit goods features. These characteristics justify only partial subsidy. For services generating positive externalities or spillovers, the amount of subsidy should be directly related to the monetary value of the benefits given upon others b y the consumer's use of the service. In respect of merit good characteristics, the value of the subsidy should be directly related to the degree of undervaluation (in terms of willingness to pay) of a service by myopic consumers. Secondly there may be services which are unquestionably public goods, but where some element of direct charging is found to be required to impose discipline upon their use. Fees for medical treatment and drugs or water rates are frequent examples. Charges force people to be careful in consuming expensive or scarce resources. Finally, there may be pressure to recover the costs of a service through charges rather than taxes, simply because they are easier to collect. In the previous chapter we have underlined the difficulty of achieving adequate and equitable taxation, particularly in Third World countries with a predominantly non-wage population. There is a strong temptation to rely heavily on charges which people have to pay to get a service they demand, and may indeed be willing to pay at a higher level than obligatory taxes. The low tendency for direct taxation of many poorer countries greatly contrasts with the high proportions of essentially low incomes spent on school fees, water cartage, and medical treatment. Indeed, there are those who argue (controversially) that charging is essentially democratic because it reinforces popular choice: people can choose what they will pay for and what they will not, and the pattern of public expenditure is, at least partially, directed by their choice. In practice, direct consumer charges are usually imposed for one or more of the following reasons (Davey, p. 88):

1)

Whether a service is a public or private good, it may not be possible to provide it to everyone and it is therefore unfair to charge its cost to those taxpayers who cannot get it. This is one reason for charging for water supplies, or for education in advance of universal provision. (This argument may not apply where a service is financed through property taxation, and the availability or absence of services is reflected in the valuation.) A service may involve an expensive or scarce resource and public consumption needs to be disciplined. This, again, is often a reason for charging for water supplies (particularly through metered systems) or for the prescription of medical drugs. There may be widespread variations in individual consumption which relate, at least partially, to choice rather than need. The use of recreational facilities is an example. A service may be utilized for profitable commercial operations as well as the satisfaction of individual domestic needs. Water, electricity, refuse disposal, posts, and telephones are all used extensively by industry, for example. (This may lead to charging all consumers, or only those in commerce and industry.)

2) 3) 4)

5) Charging may test out the directions and scale of public demand for a service, where essential needs or types and standards of provision cannot be unequivocally prescribed. A case can be made out for almost any form of public expenditure; willingness to pay directly for it is a more significant test of public preference. Figure 4.1 provides a decision tree choice about whether to charge for particular local services.

4.1.1 The Extent and Variety of Charging There are such international variations in the extent and methods of public user charging that it is impossible to easily categorise and distinguish charging systems. There is not a single form of charge that, like property tax, would be typical for local government. There are variations whether to charge for a service at all, variations in the extent of direct cost recovery, and variations in the dependence of particular public authorities upon revenue from charging. Taking the first dimension, there are some services for which direct user charging (with or without subsidy) is almost universal. Piped domestic and industrial water supplies are normally billed through metered consumption rates, through a rate based on property valuation or the diameter of the mains connection pipe, or through sale from public water points. Public transport costs are recovered, partially at least, from passenger or freight fares (though subject to increasing degrees of subsidization from general revenues in developed countries). Postal and telephone services are normally sold per unit of service, though there are exceptional cases of basing telephone charges upon property valuation. Gas and electricity are also basically charged according to volume consumed, though unit costs often decline as the amount used increases. Public housing tenants almost invariably pay rents (or tenant purchase instalments) except where free housing is provided to public servants as part of their conditions of service. Some form of entrance fee is normally charged for use of certain recreational facilities such as municipal swimming pools, golf courses, or theatres, though not for others such as parks. Provision of public utility services such as water, gas, electricity, and telephones is also usually subject to initial connection fees and sometimes to a basic standing charge, regardless of consumption. There is a far greater range of services for which user charges are imposed in some cases or places, but not others. The following are prominent examples. Education Most countries now provide free tuition within their public school system at primary level, though the costs of "extras", like uniforms, meals, and textbooks, may still be a significant burden for poor families. A minority still charge for primary tuition. The position at secondary level is not dissimilar, but with more countries charging and at significantly higher rates; these are frequently counterbalanced b y greater provision of exemptions or scholarships for the poorer student. In tertiary education, charging may be regarded as the rule; what varies is the extent of subsidization and of scholarship provision. An alternative concept in higher education is charging the costs to a loan account which the student discharges over his subsequent career. Parental contribution is frequently sought for the capital costs of school building through "self-help" community effort. Roads (and related infrastructure) Construction and maintenance of highways is typically financed from general taxation. Internationally, there are a variety of specific user charges for roads. Trunk roads with limited access are often operated as toll roads. Initial construction of estate or neighbourhood roads, together with drainage and street lighting, is frequently charged to the landowners with frontages through initial service charges or, more generally, to local taxpayers. These costs may alternatively be recovered from the developers and residents b y public acquisition, servicing, and resale of land - a widespread practice by urban development authorities or new town corporations - or by the variant of land readjustment. Costs of city centre roads may be recovered indirectly through parking fees or area licences. Maintenance of local roads and paths in rural village or urban neighbourhoods may depend upon obligatory communal labour or some monetary commutation.

Medical Services Generalizations are difficult, but two basic models can be discerned. In the first, public hospitals and clinics provide a free basic service, while charging for preferential accommodation or priority treatment for non-emergency cases in hospitals. A substantially free basic service may, nevertheless, be associated with a low flat rate fee for treatment and/or drugs, subject to exemption for children, the old or destitute. The other model is full cost charging by hospitals and clinics - whether publicly or privately run - but with a national medical insurance scheme, or a range of private insurance schemes to cover patients' costs; some form of protection is usually given to those with low incomes and inadequate insurance cover b y means-tested exemption or the graduation of charges. Environmental Health As a demonstrably public good, public health services are customarily tax borne. There are exceptions. Sewerage or refuse collection is frequently charged to those served, occasionally b y specific fees, but usually in the case of households by a rate graduated according to property valuation. Removal of industrial refuse is usually charged specifically according to volume and effluent strength. Where communal sanitation is provided, users may be expected to maintain it collectively. Fire Services Urban authorities occasionally charge their citizens (or their insurers) for putting fires out. 4.2 Cost Estimation The basis of charging is cost recovery. A policy may permit charging less than the full cost, or may seek the recovery of more than the full cost, involving respectively a contribution from or to general revenue. Issues of subsidization or surplus generation will b e discussed in later sections. The first necessity under a charging policy is to define and calculate the full costs of the service under consideration. This concerns three issues. 1) The first issue is what expenditures to attribute as costs to a particular service . Where does one draw the boundary between the costs of a particular service and those of general public services in the locality or the general administrative overheads of the regional authority concerned? For example, in calculating the costs of municipal housing for the assessment of rent, should one include the expense of off-site services (local roads, drainage, street lighting, etc.), administrative overheads, or social and community services (e.g. schools, community halls), or even administrative apparatus (e.g. housing management department and the time of managers, architects, engineers etc.) in the estate? An intermediate approach is to attribute only the marginal costs of administration, i.e. those increases in overheads caused specifically b y the expansion of a service. The community services on an estate might be regarded as part of the general services provided by regional government to public and private housing tenants and owner-occupiers alike. If this is so, there would be no justification for including them in the housing costs. The cost "boundaries" issue arises most sharply in respect of car parking. The full costs of a car park might be regarded as confined to the provision and management of a parking site. It can b e argued, however, that the costs of parking include providing access to the general area, thereby including the whole of the road and traffic management system in the vicinity. 2) The second issue is whether the costs are calculated according to the actual expense of a particular unit of service or on a pooled average basis. T h e actual costs of an identical level of service may vary considerably. Amortization of capital costs will vary with the age of the capital assets because of inflation and because of fluctuations in the interest rates attached to any loans involved. Location affects costs. Water and electricity will cost more to supply to outlying areas because of transmission costs - the length of mains, pumping, leakage, etc. Low density populations consume less than high density, but the fixed costs of provision will not fluctuate proportionately. A half-empty bus serving a sparsely populated area will cost almost as much to run as a crowded one in the city centre; depreciation and labour costs may be almost constant and fuel consumption will not be proportionately lower. There are conflicting arguments. If chargeable services are seen as purely private goods, and charging as a market pricing instrument, each unit of service should be charged according to its own marginal cost. A consumer is only provided with a service if he is prepared to pay the true cost of providing it to him. This encourages rational use of a service and an optimal location of settlement. The opposite argument is that the circumstances which vary the cost of a service are not necessarily of the consumer's making; in so far as the service might meet a basic human need, he should not be penalized if these costs are above average in his case. For example if the poor live on the outskirts of a town, distance from work will increase their essential expenditure in any case. Their burden should not be increased further by charging above-average unit costs for essential services. This issue is difficult to resolve. Two questions, however, affect the balance to be struck between the arguments for marginal cost and for equalized charging. The first is the extent to which the service (or some minimum provision of it) is meeting an essential human need. The second is the degree to which individual consumers choose the circumstances - particularly the location - which affect the cost of the service they use. Both questions really involve the degree of consumer choice - the extent to which use of a service at above-average cost is selfimposed, like buying an expensive shirt. 3) The third issue in cost estimation is whether capital costs are included and on what basis. There are many examples of services which are meant to be self-financing, but only the operating and maintenance costs are charged to the consumer. The capital costs have been met out of general public revenues or from loans which have been fully discharged. Many long-standing water supply and sanitation systems fall in this category. Where debt charges are still current, these would be included in the chargeable costs of a service unless it is being deliberately subsidized. There are arguments, however, for including capital costs in the estimation of charges regardless of the fact whether the authority administering the service is currently discharging them or not. Capital investment is regarded as having an opportunity cost; it could have been used on some other form of public expenditure or left in the taxpayer's pocket. Seen in this light, investing capital in a particular service can only be justified if it earns a rate of return comparable with alternative forms of public or private use. The public's willingness to buy a service at the resulting level of charging is comparable to its readiness to buy goods or services from a commercial operator using the same amount of capital;

it is the essential market test of viability. It is even argued that this test of comparison can only be fully met if charges include the equivalent of the taxation which a private operator would have faced. Charging should also include the amortization of capital assets at their current rather than historical value, i.e. the cost of acquiring or constructing them now. Marginal cost pricing is a further variation of this approach, aiming to charge all consumption at the unit cost of meeting any additional demand; if extra demand would necessitate new capital expenditure, its unit costs would be reflected in the price of the whole existing supply. Such pricing is advocated on the grounds that every consumer must be faced with the full cost implications of increasing demand. It is mainly urged in relation to utilities like water and electricity where increases in consumption can flow from the decisions of existing as well as potential consumers. Including provision for future capital requirements through depreciation charges, current cost accounting, marginal cost pricing, or similar approaches is obviously clear in theory. In practice there are a variety of practices in cost estimation. Chargeable costs may include:

Operation and maintenance only

amortization of capital costs on "soft" terms, i.e. interest free or on submarket interest rates, and possibly with moratoria on repayment or prolonged loan durations amortization of capital costs at full commercial rates of interest depreciation of the value of capital assets over their estimated life a commercial rate of return on the value of the investment in capital assets

the marginal costs of investment (i.e. the unit capital costs of providing more than the present supply) pertaining to all units above the level of current supply. In calculating the value of capital assets for assessing depreciation or rates of return, valuation may be based on historical costs

current market or replacement costs.

A two-part tariff may be adopted with a fixed charge for the capital installation costs, and a variable charge based on consumption. These variations reflect the changing fashions of economic thought, the ideological stances of socialist or market-oriented governments, and ultimately what the consumer will tolerate - or what the politicians think he will tolerate. 4.2.1 Charging Below Costs The grounds for imposing charges would normally suggest that they should be based upon the total costs of the services under provision (as we discussed in previous part). As we have seen, the theoretical justification for charging involves some compromises. These may result in charges being fixed below full cost level and subsidized from general revenue. There are four major reasons why this may happen: 1) The first arises where a service is basically a public good - it is provided because of its collective benefit - but a charge has to be imposed to discipline consumption. The charge would then have to be set at a level calculated to deter waste but to permit the minimum essential level of consumption by all income groups. Medical prescription charges or water standpipe rates may fall into this category.

2)

The second case for subsidization occurs where a service is mixed good - partly a private and partly a public good - where it primarily benefits the individual user, but its consumption needs to be encouraged for some public saving or benefit. The most conspicuous example is where rail or bus fares are subsidized to encourage people to use public rather than private transport, as a means of reducing traffic congestion and road expenditure.

3) 4)

Thirdly, services which are wholly private goods may nevertheless be subsidized if these are in popular demand and the authorities are reluctant to face the public with their full cost. This often applies to the provision of recreational facilities, from swimming pools to opera houses. It may be argued that having a fit population or a prestigious orchestra is a collective benefit. Lastly, and most significantly, a private good may be subsidized because it is regarded as a basic human need and the lower income groups, at least, could not be expected to meet its full cost. One has to be careful in giving examples simply because the conception of "basic needs" is highly subjective and relative to the general standard of living. Subsidization can lead to inefficiencies - either because resources are expended on provision which is not cost effective or because it benefits all consumers whether they need the subsidy or not. To resolve it, reduced charging should only apply to certain categories of user: lower income groups, children, old people, etc. Or, secondly, subsidization should apply only to a minimum level of consumption, above which market pricing should apply. The first 40 litres of water supplied to a person per day might be charged, for example, at a low rate below cost; consumption above this level would be priced at the full marginal cost. This is known by the World Bank as the "lifeline" approach (Davey, 1983). Adoption of either approach depends essentially upon administrative feasibility and political pressure. The "lifeline" approach also depends upon being able to measure consumption adequately. In other conditions eligibility for subsidy is less easy to determine. 4.2.2 Charging above Costs In some cases charges may be based upon recovering more than the full cost of a service, i.e. upon making a profit. There are three cases at least where this may arise. The first is where charges are imposed for a basically regulatory purpose involving little direct cost. Licensing fees or parking charges are examples. Secondly, charges may be imposed at an above-cost level to reinforce their disciplinary effect upon consumption. Telephone charges may be graduated according to the time of day to discourage congestion at the peak business hours. Parking fees or area licensing fees may be imposed at punitive levels, to keep the private car out of city centres.

Finally, a service may be in heavy demand and people willing to pay highly for it because of its importance or popularity and short supply. This may particularly operate where the same service is subject also to costly private enterprise provision. Charging above cost could be justified in such circumstances if the surplus revenue is given back into extending the service so that a wider number have access to it. A good example of this relates to the provision of public housing in many Third World countries, where inadequate supply gives rise to extortionate rents in the private sector, and where the tenant of a publicly owned house maybe regarded as relatively privileged. 4.3 Appraisal with the General Criteria 4.3.1 Adequacy and Elasticity Many texts dealing with local and regional government finance either ignore charges or deal with them only after considering intergovernmental grants and subnational taxation. Charges are often perceived as the financing instrument of last resort. The extent and practice of charging are subject to such variation that generalizations about the scale of its contribution to local and regional government revenues would be meaningless. These range from relatively important ones (such as those related to the use of cars) to relatively insignificant ones (such as licences fees for dogs). The better application of charges for activities that require some use of social services (such as education, health, and commercial activities) can provide important resources that come in the form of benefits received and are therefore consistent with the basic principles behind decentralization. Regarding to elasticity, charging should be responsive to growth in population and incomes, since these are generally reflected (often more than proportionately) in the growing demand for, and consumption of a service. The response is dependent upon the accessibility of capital to expand the service to meet a growth in demand. There is an associated problem that population growth, particularly in large urban sectors, is so often associated with diseconomies of scale - the higher pumping costs and greater leakage involved in supplying outlying areas with water, for example. Charges tend to be unresponsive to inflation. They are almost invariably based upon fixed tariffs per unit of service and a discretionary decision is needed to increase them when costs rise. Increasing house rents, water charges, and bus fares is unpopular and they frequently fall far behind the rate of inflation. 4.3.2 Equity

1) 2) 3)

Charging has traditionally been considered as regressive - that is impacting disproportionately on the poor. There are three reasons. First, it falls upon consumption, which may be dictated more by essential needs than income levels. Secondly, subsidization often benefits middle and higher income groups more than the poor. Thirdly, because the capital costs of installation have to be recovered regardless of the level of consumption and do not vary greatly according to that level, many tariffs are based upon a declining unit cost, i.e. the more water or electricity used, the cheaper it gets.

Charging is not usually seen as an instrument of redistribution per se. It is an inefficient tool for this purpose precisely because consumption is not related proportionately to income. In recent years this traditional view has changed, there is a growing concern to limit its regressive tendencies, and even to use it as a positively redistributive mechanism. This tendency implies from following reasons. The first is the effect which regressive incidence may have upon the access of the poor to chargeable services regarded as basic needs. The second is the failure of the taxation system in many countries to shift resources from the rich to the poor; if charges are easier to exact than taxes as a general means of financing public expenditure, this may extend to their utility in redistribution. There are various methods by which charging can be made less regressive or positively redistributive. The first is the widespread basing of charges upon property values; water and sanitation are often priced on this basis. The second means is the use of differential tariffs with higher rates upon certain classes of user - usually commercial and industrial; this is of course a questionable instrument if the incidence falls on the consumer rather than the shareholder. A third is a progressive tariff which imposes increasing unit prices as consumption rises. If ability to pay is really of concern, there are two ways to address it. Targeted subsidies defined through means-testing with reduced tariffs or exemptions for the old and the poor (rarely an easy process to administer which allow account to be taken of the user's income and therefore ability to pay). Cross subsidies designed to charge the rich a higher price so as to subsidise poorer customers. Water supply schemes financed by the World Bank have been widely characterized by cross-subsidization, progressive tariffs on larger domestic or commercial consumers reducing or eliminating the charge for supply to public standpipes and to households consuming small volumes only. 4.3.3 Administrative Capacity Charges are theoretically easy to assess and collect. They are easy to assess because liability is based upon measurable levels of consumption; easy to collect because people only get what they pay for. If rents are unpaid, the tenant is evicted; water, electricity, or telephones are cut off if bills are unpaid; people can only enter the swimming pool or the theatre through the turnstile. In practice, there are some difficulties. The first is technical - the problems of controlling illegal water connections and meter bypassing, for example, or of collecting fares from a crowded bus with passengers inside and outside. A second concerns the political will to implement sanctions. Evictions rarely win votes and are pilloried by the media; politicians may intervene to stop disconnections of water or electricity supplies to their supporters. The third problem is integrity. Tax liabilities are fixed, but those for charges vary with consumption; it is harder to check what collectors should have received. 4.3.4 Political Acceptability Most charging is accepted in principle. In so far as they are directly related to a specific tangible service and consumption includes an element of choice, charges are understood and met with reasonable willingness. Nevertheless, the level of charging is most sensitive politically. Most charges have to be paid out of pocket; many relate to what people regard as daily necessities - education, transport, water, housing, and the like. Increases require specific political decisions and are unpopular.

Moreover, the need to raise charges arises particularly in times of inflation when governments are under pressure not to increase prices themselves. Representative bodies are often even less willing to raise bus fares, house rents, or water rates than taxes. As a result of failure to increase charges through political inertia, services either get worse or require increasing subsidy. Alternatively, increased charges due to rising costs may result in decreased consumption. Since capital costs and overheads may be constant, this gives rise in turn to increasing unit costs. A vicious circle of declining use and rising charges is created. The previous section discussed charging generally. In practice many, if not most of the directly charged services are provided in fact by regional authorities. Therefore, we will now focus on charging as it is carried out by regional or local government. We will also analyse some general practices and experience. There are no special problems attached to the administration of charging by local and regional government. The performance of a service is normally located clearly in a given region and assignment does not create difficulty. Regional and local variation in tariffs may cause political dissatisfaction, but no impossible administrative problems. The relative closeness of local and regional representative bodies to their electorate may, however, sharpen their willingness to increase tariffs or enforce compliance. The weight of revenues from charges varies country to country and according to the local government tier, but is generally more substantial for municipalities than for the other tiers since they are entities that are most often responsible for community services. Except main categories as environmental services (such as water, sanitation, and solid waste management and fire services, environmental health) and social services (such as primary health care, social care and primary education) they include other type of local services that do not fit into former: such as markets, bus and taxi parks, housing, planning/building permits and host of "minor fees and charges". Markets are in competition with other retail outlets. They also tend to cater for the needs of poorer customers in many towns. These two factors constrain greater exploitation of municipal market charges, although for some small municipalities markets may represent the only significant local revenue source. The emphasis is usually on improving the collection of existing fees. Bus and taxi parks can offer a significant source of revenue for local councils. They rise many of the same issues as markets, although they may be more vulnerable to cartels of transport operators. Contracting-out the revenue collection is a common practice, often to an association of transport operators. Planning and building fees are connected with land development (for recovering the costs of new infrastructure from benefiting property owners through charges on the increased value of land. Many local governments are permitted to apply a kind of development or impact fee, usually to finance improved access to a development. Charges and fee revenues provide local governments that are allowed to set their fees independently with some financial flexibility. However, some rates are fixed by the central government or regulated in many countries. (In Hungary and Czech Republic, for example, operating fees for water, household waste and heating can be set independently, whereas education, social welfare and public health fees cannot.). In addition in several new EU member countries, fees are controlled by regulating bodies. The international evidence is that charges yield substantially less revenue than grants or local taxes and there is no significant tendency for charges to increasingly replace the other income sources.

Conclusions
Charges are payments for services provided directly to the consumer, unlike taxes where there is no direct relationship between payment and benefit. According to the theory of public finance, charging is appropriate where the benefits are essentially private, consumption is rival, non payer could be excluded (private good). By contrast, where the benefits are mainly public, non-rival and non-excludable then it is appropriate to fund the service from taxation (public good). However, the distinction is not always so evident in practice. Very few local and regional government services qualify as public goods, most being categorised as mixed goods with positive externality or merit goods features. These services should be partly subsidized. The other problem concerns those who are too poor to pay full costs of the service (especially if it is regarded as a basic human need) - the solution is not to abandon charging for the service. Charging for services can be justified where a service is basically a public good but a charge has to be imposed to discipline consumption. In the international context there are wide variations in the extent and methods of charging (variations in charging for a service at all, in the extent of direct cost recovery and variations in the dependence of particular public authorities upon revenue from charging). Some broad categories of services for which charges are levied are: water supply, public transport, postal and telephone services, gas and electricity, public housing, recreational facilities. The other categories of services for which user charges are imposed in some cases or places, but not others are: education, roads and relating infrastructure, medical services, environmental health, and fire services. Charging should provide the resources to maintain and expand the service. The charge should be set to recover the full costs of the service. Some issues have to be considered when defining the costs of the service: the extent of costs to attribute as to a particular service, calculation of the costs according to the actual expense or on a pooled average basis and whether the capital costs are included and on what basis. The conventional recommendation is that the charge should be set to recover additional cost of providing service to that particular consumer (referred to by economists as marginal cost pricing), or through depreciation charges, current cost accounting or similar approaches. In some cases charges should be fixed below full cost level and subsidized from general revenue: in case of public good charge should ration demand and prevent waste of service, or if service delivers a mixture of public and private benefits (mixed good), in case of popular and demanded services (recreational facilities) accepted by authorities to be subsidized, and with private goods regarded as a basic human need. In some cases charges may be imposed at an above-cost level of a service. These charges are imposed for a basically regulatory purpose, to reinforce their disciplinary effect upon consumption, or if heavy demand for service exists and people willing to pay highly for it because of its importance or popularity and short supply and the surplus revenue (from charges) is given back into extending the service so that a wider number have access to it. The range of charges and the scale of contribution to subnational governments varies country to country. According to the evidence charges yield substantially less revenue than grants or subnational taxes. Charges are elastic to growth in population and incomes, and tend to be unresponsive to

inflation. Although not inherently or traditionally an instrument of redistribution, charging can be used for this purpose where there is political will and no suitable fiscal alternative. Charges are easy to assess and collect, they are relatively acceptable, but much charging is highly sensitive politically. Revenue may fall below the levels needed to operate a service effectively, particularly in times of inflation, because of political reluctance to increase tariffs or enforce sanctions. The weight of revenues from charges varies according to the local government tier, but is generally more substantial for municipalities than for the other tiers. Broad categories of services for which charges are levied (depending on the functional responsibilities assigned to local government) are: environmental services, social services, local services.

5. Intergovernmental Transfers
This chapter: introduces the main purposes for the payment of intergovernmental transfers

discusses two dimensions of the design of intergovernmental transfers and describes twelve potential grants types based on this classification summarises a typology of intergovernmental grants, their advantages and disadvantages discusses the social and economic rationale for financial equalization both the expenditure needs and revenue capacity gives empirical data and examples of country experience with intergovernmental transfers.

Local authorities' financial resources shall be commensurate with the responsibilities provided for by the constitution and the law. The protection of financially weaker local authorities calls for the institution of financial equalisation procedures or equivalent measures which are designed to correct the effects of the unequal distribution of potential sources of finance and of the financial burden they must support. Such procedures or measures shall not diminish the discretion local authorities may exercise within their own sphere of responsibility. As far as possible, grants to local authorities shall not be earmarked for the financing of specific projects. The provision of grants shall not remove the basic freedom of local authorities to exercise policy discretion within their own jurisdiction. (European Charter of Local Self-Government: Article 9) 5.1 The Purposes for Transfers Chapters 3 and 4 demonstrated the case for user-charges and regional and local taxes. However, revenues from those two sources may be insufficient to finance subnational government expenditures. If subnational governments are responsible for providing a significant proportion of public services, central government should compensate their need for resources beyond those left to their own exploitation. These allocations from the central government are often described as intergovernmental transfers. They are of considerable importance to most systems of regional and local government, and this importance is growing. This chapter deals with different forms of central government allocations to subnational governments - with tax sharing3 (There is often a debate about what is an intergovernmental transfer. Grants to lower-level governments are clearly intergovernmental transfer. The confusion comes in the case of taxes. If the local government can control either the rate or tax base, it is a tax. If the rate and base are determined by the higher-level government, and revenue collections are assigned to the local government, it is a transfer ) and grants, subsidies, contributions or subventions (the titles vary, but their nature is the same - a transfer funds from the budget and accounts of the central government to those of a self-accounting subnational government). There are many different kinds of intergovernmental transfer systems, but they have many different types of impact on subnational government finance. Some stimulate local spending, some are substituted for local and regional revenue effort, some are equalizing, and some lead to more local government fiscal autonomy than others. Allocations by central government to subnational governments have varying purposes which affect their form and scope: 1) Financing wholly or partly the cost of services or development programmes which are of national significance, i.e. which are regarded as consistent with national level interests, policies, and targets; these include contributing to spillovers - expenditures by regional authorities which confer benefits beyond their boundaries. If the central government did not pay grants to regional authorities there would be low levels of provision of those services, or regional and local tax levels would be sub optimally high. 2) Encouraging effort by regional authorities to develop programmes and services in line with national policy. 3) Stimulating growth in regional economies, both to contribute to national growth and to reduce inter-regional disparities. 4) Controlling subnational expenditure to ensure compliance with national policies or standards. 5) Securing an equal, or more equal, standard of services or development. 6) Compensate to subnational governments with a low fiscal capacity or relatively high expenditure needs in order to avoid fiscal stress. Inequalities in the abilities of subnational governments to raise revenues and in their need to spend, each per head of population, may result in severe fiscal stress as affluent groups vote with their feet and exit poor needy authorities to relocate in prosperous authorities, and a businesses do likewise. 7) Assisting regions to cope with emergencies. Central governments may be trying to restore the local and regional economies with adverse industrial structures, on the grounds of both equity and efficiency. High levels of local taxation may lead to cumulative decline as selective out migration of population and out migration of business activity occurs. These objectives are not mutually exclusive. In practice, systems of intergovernmental transfers can combine control with equalization. 5.2 The Design of the Intergovernmental Transfer System Intergovernmental transfers have to dimensions: the size of the divisible pool, and the distribution of this pool among eligible government units. Some have referred to the divisible pool dimension as having to do with the vertical balance between the central and subnational governments, and the allocation dimension as having to do with horizontal fiscal balance.

Bahl and Linn (1992) have developed taxonomy of grant system that takes both of these dimensions into account (see Figure 5.1). First, consider the determination of the size of the total amount to be distributed in a given year (the columns in Figure 5.1). The international practice suggest three basic approaches: a specified share of national (or state) government tax revenues, ad hoc decision (such as an annual appropriation voted by parliament), or reimbursement of approved expenditures. Once the amount of the distributive pool is determined, allocations among local governments are typically made in some combination of four ways: by returning a share to the jurisdiction from which the taxes were collected, that is a derivation principle; by formula; on ad hoc basis; or by reimbursing costs. This two-way classification gives taxonomy of twelve potential grant types; eight of which are more or less common in developing and other countries. First, consider the shared tax column in Figure 5.1. A type A transfer is shared tax on a derivation basis, i.e., the subnational government is allowed to keep a specific share of what is collected within its boundaries. This is the approach to tax sharing used in a many countries. A type B transfer is based on a share of a national tax, but the distribution among local governments is made by formula - according to some need criteria which do not relate directly to the origins of the tax (on the population, land area etc.) This approach is used in the Czech Republic. A type C grant differs in that the distribution is on the basis of project costs. For example a fixed percentage of a national tax may be distributed among local government on the basis of the (approved) costs of constructing public works projects, or the costs of teacher's salaries, etc. Figure 5.1 Alternative forms of intergovernmental grant programs Method of determining the total divisible pool G Method of allocating the Specified share of Ad hoc Reimbursement of divisible pool among eligible units national or state decision approved expenditures government tax Origin of collection of the tax A L not applicable Formula B F not applicable C G K Total or partial reimbursement of costs Ad hoc decision D H not applicable The second column in Figure 5.1 are ad hoc grants, i.e. the central government decides on the total distribution on a year-to-year basis. Even within this category, however, there are major differences depending on the horizontal allocation method used.

A type H grant is completely centralized, with the central government making all decisions about who gets the money and how much is given to each recipient. The type L, F,andG grants allow the centre to make annual political decisions about the total distribution, but there is some objectivity in the distributions among local governments units. The type G grant would be conditional, but the other three grants would be for general purpose. The more strings tied to a grant, the more subject it is to control by higher-level governments.

Column three are the costs reimbursement grants, where the line ministry will decide on both the amount of funds necessary to carry out the work, and which local projects live up to central standards. These would take the form of conditional grants. Examples are infrastructure grants where construction standards specified, teacher's salaries grants where salary amounts are pre-determined, and subsidies to individuals which are fixed in terms of eligibility and payment amount by higher-level governments. From these taxonomy it is apparent that some of the transfers systems are decidedly more decentralizing then others. For example, type A transfer is counter equalizing in that it will favour the rich local governments with the stronger tax bases, whereas the formula grant (type B) could be distributed towards those with weaker tax bases. The ad hoc column is the most centralizing because it allows the centre maximum flexibility in deciding how much to distribute to the subnational governments sector each year. The type C, G,andK grants are also centralising because they give the central government ministries significant control over how their money is to be spent and allows for construction and service delivery standards to be laid down. 5.2.1Tax Sharing or Grants? Both tax sharing arrangements and intergovernmental grants provide resources to subnational governments. Drawing the dividing line between the two fiscal arrangements proves sometimes difficult. If taxes are levied and collected by central government, assigning a share to subnational government may appear little different from paying grants. This is particularly so if both the tariffs and the proportion assigned to the regions are entirely at central government discretion. There are differences between tax sharing and grants which generally work in favour of regional government. First the assignment of a tax share may in theory be as discretionary as the calculation of a grant. In practice, it would be unusual for central government to vary an assignment to the disadvantage of regional authorities without consultation and agreement. Secondly, if the shared tax is elastic, subnational governments receive automatic growth in revenue; increases in grants, on the other hand, are purely discretionary and considerable pressure is often necessary to secure adjustments in formulae to reflect the rising costs of local and regional functions. Tax sharing does convey a slightly different political nuance to grants. The idea of sharing taxes is consistent with an image of partnership between levels of government in the performance of public functions. Grants are inherently more patronizing; reliance upon them can give a totally misleading impression of poverty and encourage a beggar-type posture by subnational government. One weakness of tax sharing is its vulnerability to irresponsible administration by central government. Delay in paying over their share can impose major cash flow problems on regional authorities. Another disadvantage is the inability of subnational government to vary the rate of a shared tax to finance marginal changes in expenditure. From this point of view the discretionary surcharge on national revenue - "piggy backing" - is much preferable to fixed shares of a centrally determined tax rate. 5.2.2 A Typology of Grants

A typology of grants is provided in Figure 5.2. It is assumed that the grants are paid by central government (the grantor or donor) to local or regional government (the grantee) in order to stimulate the provision of services. Specific grants are paid in respect of specific needs and can only be spent on that service. They are therefore also referred to as conditional, categorical or earmarked grants. They may be used to finance services which local authorities provide on behalf of central government, to encourage provision of services with substantial spillover effects, or to finance a minimum standard of service required by central government. General grants can be used to finance the broad range of services provided by local authorities as long as their activities are not illegal. They are therefore also referred to as unconditional or block grants. They are often used to achieve horizontal and vertical equity through financial equalization schemes. Both general and specific grants are sometimes paid to subnational governments, for example, where a specific grant is paid in respect of municipal housing and a general grant is paid in respect of education, personal social services, leisure and recreation and so on. Specific grants can be lump-sum (that is, of fixed amount) or matching (where the grant forms a given percentage of the local government's spending, for example, 35 per cent). However, there may be an upper limit on the amount of grant that can be levered by the local government, in which case it is a closed-ended grant. If there is no such limit it is open-ended.

Similarly, general grants may be lump-sum or effort-related. While a matching specific grant matches expenditure on the specified grantaided service, the effort-related general grant is related to the revenue effort of the local authority. Revenue effort is usually measured in terms of tax effort: the greater the revenue raised from local taxes, the more grant the local authority receives. Again, grant may be open-ended or closed-ended, according to whether or not central government wishes to avoid too stimulatory an impact on local government expenditure. Article 9 of the European Charter of Local Self-Government, reproduced at the start of this chapter, clearly expresses a preference of general grants because they are not earmarked to specific projects. However, it express no view as to whether they should be lump-sum or effort-related. A related issue is whether grants to subnational governments for specific purposes should be on a matching basis, dependent upon some ratio of contribution from the recipient's direct revenues. There are two major criticisms of matching grants. The first is that they benefit rich areas more than poor; by their ability to spend more from their own resources, wealthy areas can earn a larger central matching contribution, worsening existing disparities in levels of service and development. This is a valid criticism of many matching grants, but they do not have to operate in this way. The United States government varies the ratio of many of its matching grants to states according to the average per capita income of each state. High income states get a lower proportional contribution from the federal government and vice versa. The second criticism is that matching grants distort the pattern and priorities of regional expenditure. Regional authorities shift funds within their budget to attract greater central contribution. They encourage a given level of expenditure on a service, not necessarily a set standard of achievement. 5.3 Financial Equalization 5.3.1 Equalization of Local and Regional Expenditure Needs Differences in the need to spend per head of population reflect differences between subnational governments in their demographic, socioeconomic, geographic and other structures. The variables affecting expenditure needs are wide-ranging, including the proportion of population of school age and retirement age, the proportion of households headed by unemployment or single parents, urban density or rural sparsity, demographic and economic growth or decline, and so on. For example, local governments with high proportions of children and elderly people in their populations, the two groups to which the most expensive services are directed (school education and social services), would have to spend more per head of total population than governments with low proportions of those two groups, even if they have spent the same amount per child or per elderly person as those other governments. There are two broad approaches to the definition of regional expenditure needs for the purposes of central government allocations. 1) Allocations can be based upon the estimates put forward by each recipient authority, subject to such amendment as central government wishes to impose. 2) Allocations to individual authorities may be based upon objective criteria for measuring their needs which are not related to their own budget proposals. These needs may be measured in terms of

a) b)

the functions performed by authorities, or of their general capacity for expenditure.

1) Assessing needs based on historic costs The first approach just categorized is most commonly used for calculating budget votes, and grants to individual authorities. Qualifying expenditure is calculated on the basis of estimates submitted by each authority, comprising the cost of services at their existing level, plus

increases argued "on their merits".

These elements are based in turn upon the historic growth of expenditure in each region. The problem is that existing levels of service expenditure may (and frequently do) contain considerable disparities between regions which reflect historic development rather than an objective assessment or comparison of need; they usually favour the more central or economically prosperous areas. 2) Assessing needs upon objective criteria The second approach is to assess the expenditure needs of individual authorities by use of objective criteria, which do not relate to their actual levels of current expenditure. The problem lies in finding forms of measurement. For specific allocations, need will be based upon the costs of a particular service, project, or programme and some measurement of demand for it. An education subsidy may be based, for example, upon unit costs for types of schooling, multiplied by the relevant age-group population of each area. General grants allocations can be calculated by aggregating the costs of each individual function using standard criteria based on unit costs. This has been attempted (with considerable difficulty) in calculating "grant-related expenditure" for the new British block grant to local authorities. This approach poses technical problems of measurement and data collection. The alternative is to use some objective criteria for calculating a desirable overall level of regional expenditure, either on all functions, or on the range of functions to which the allocation is devoted. General grants or tax shares are frequently based upon such broad criteria which define need in terms of an overall capacity to spend, rather than individual functional costs. The most common basis of allocation is population (sometimes weighted by the proportion of children and old people to reflect heavy local authority commitment on education and social service expenditure). Area is a second factor frequently appearing in distributive formulae. A third factor often taken into consideration in distributing general grants is relative income or wealth. Equalization schemes and the calculations would be enormously complex and suffer severe data problems. A more manageable model necessarily has to exclude some indicators of expenditure needs. Judgement therefore has to be exercised about which expenditure needs factors should be included or excluded. This often causes that the process loses much of its objectivity. The other major problem is of deciding the relative weights of needs indicators, since these crucially influence the subsequent distribution of grant between different types of local government (e.g. between urban and non-urban). 5.3.2 Equalization of Local and Regional Revenue Capacity Variations of subnational governments to raise their tax revenues per head of population largely reflect the uneven distribution of economic activity, some authorities having relatively large business sectors. In order to raise the same amount of their tax revenues per head, rural authorities would have to levy much higher tax rate (both incomes and properties) than city authorities, for example. It creates horizontal inequity, in that people in similar circumstances living in different regions will have to pay differing amount of local or regional tax for the same standard of service. This is the economic rationale for the payment of intergovernmental grants to authorities with inadequate local taxable resources per capita and /or with high per capita expenditure needs. Equalization of the differences in the potential per capita revenues is often related to the amount of direct revenue which regional authorities raise, or might be expected to raise. It requires a measurement of the subnational tax base or "fiscal capacity" and "fiscal effort". Such compensation would normally assume that regions made a "standard fiscal effort", i.e. that tax rates, the precision of their assessments, and the effectiveness of their collections represented a standard effort to exploit their tax base, however restricted it might be. The other possible aim for relating allocations to regional taxation is to stimulate and reward regional fiscal effort. Some standard level of revenue exploitation may be established, central allocations being increased when it is exceeded or reduced when it is not achieved. The basic problem is to define and calculate regional "fiscal capacity". If a grant is calculated purely on a region's own estimate of revenue or its actual tax performance, the effect would be to penalize effort and reward laxity. Equity and incentive can only be achieved by using a national objective assessment of revenue potential. Calculation of revenue potential is therefore more accurate if related directly to the revenues which regional authorities can exploit. Where they rely upon charging, comparisons can be made of revenue collected per unit of service. Where taxes or assessment methods vary, particularly at regional discretion, it is much more difficult to relate an assessment of revenue potential to the actual sources exploited. In such cases, measurement may only be possible by some general comparison of regional economies. Regional per capita incomes are often used for this purpose. Occasionally revenue equalization involves transfers between regional authorities themselves (so called "Robin Hood" approach). This may be used to reduce disparities between the core city and the periphery in metropolitan cities. What degree of revenue equalization should be taken depends upon several factors: The first is economics - the scale of resources which a nation can afford to devote to equalization. This is a problem not only of national poverty or wealth, but also of the degree of disparity to be bridged. The second is a question of political priorities. How far is a country committed ideologically to equality of service and development opportunity?

The third consideration is the ability of the less developed areas to absorb an equal level of public investment. The relative poverty of these regions may be reflected in low administrative capacity.

Conclusions

Since the revenues from taxes and charges in most countries exceed the subnational government expenditures, compensation of this fiscal gap is needed from the central government. These central government allocations to subnational governments are described as intergovernmental transfers. They may have different forms - tax sharing, and grants, subsidies, contributions or subventions. Most common purposes for intergovernmental transfers include: financing the costs of services which are of national significance, encouraging effort, stimulating growth of subnational governments, securing an equal standard of services, compensating low fiscal capacity or high expenditure needs to subnational governments, controlling subnational expenditure, assisting regions to cope with emergencies. According to the size of divisible pool (referred as the vertical balance between the central and subnational governments) and the distribution of this pool among eligible government units (the horizontal fiscal balance), one can consider a taxonomy of twelve potential grant types. Most common types are shared taxes distributed on a derivation basis or by formula; grants based on projects costs; grants based on more or less discretionary decision of central government. Both tax sharing (levied and collected by central government) and grants are discretionary payments, but there are some differences: central governments usually make consultations and make agreement with subnational governments about tax assignment, the shared taxes are elastic contrary to grants, sharing taxes is consistent with an image of partnership between levels of government in the performance of public functions. On the other hand, tax sharing is vulnerable to irresponsible administration by central government. The main types of grants are specific (conditional, categorical or earmarked) grants are paid in respect of specific purpose and can only be spent on that service; general (unconditional or block) grants can be used to finance the broad range of services provided by local authorities. Specific grants can be lump-sum (that is, of fixed amount) or matching (where the grant forms a given percentage of the local government's spending) which should be closed-ended or open-ended grants. Similarly, general grants may be lump-sum or effort-related. According to the evidence in OECD countries earmarked grants account for a larger portion than non-earmarked grants (a ratio of roughly 60 to 40 percent). Around a third of all earmarked grants are matching and around three quarters of all earmarked grants are mandatory. 72 percent of central governments in both federal and unitary countries provide the overwhelming part of grants to local governments. In 86 percent of federal countries the central level is the main provider for states and regions. The state government is the main source for local governments in majority of federal countries. Differences between subnational governments in their demographic, socioeconomic, geographic and other structures affecting expenditure needs and rationalize equalization of expenditure needs. The definition of regional expenditure needs for purposes of central government allocations is based upon estimates submitted by each recipient authority or upon objective criteria. There are three approaches of expenditure estimates. The first approach is based upon the historic costs and growth of expenditure in each authority, the second approach uses as criteria for allocations the costs o f a particular service, project, or programme and some measurement of demand for it (for specific grants); aggregates the costs of each individual function using standard criteria based on unit costs (general grants); or calculates the overall regional expenditure capacity based on population, area, relative income or others criteria (for general grants or tax shares). The existence of regions with inadequate taxable resources (regions whose potential direct revenue is limited by the relative smallness of their tax base), regions with a below standard "fiscal capacity" are economic rationale for equalization of differences in regional revenue capacity. It requires a measurement of the subnational tax base or "fiscal capacity" and "fiscal effort". The problems of equalization schemes and the calculations are technical - they may be complex and suffer severe data problems; problem is of deciding what indicators include, judgment of the relative weights of needs indicators; financial equalizations reflect political preference. Timing the calculation of allocations to subnational governments is also important. The degree of financial equalization between subnational governments has long varied greatly between countries through world. Most of the countries equalize per capita taxable resources; only few adopt the "Robin Hood" approach (for example Germany). In developing countries, objective criteria for intergovernmental transfers are largely non-existent.

6. Borrowing
This chapter:

specifies purposes of borrowing by subnational governments and discuss the arguments both for and against loan financing describes different characteristics of loans as duration, repayment terms, interest rates and security gives some examples of country experience with borrowing regulation and assessment of borrowing capacity

summarises sources and methods of borrowing classifies four main approaches to control of subnational borrowing briefly characterises a nature of central credit institutions, and their sources of capital

For the purpose of borrowing for capital investment, local authorities shall have access to the national capital market within the limits of the law. (European Charter of Local Self-Government: Article 9) 6.1 The Purposes for Borrowing How far subnational governments should borrow is a highly controversial issue. There are several arguments in favour of loan finance (Davey, 1983). Borrowing accelerates development since it frees investment from the limits of current revenue. To depend solely upon the excess of tax and operating revenues over operating expenses would often limit long-term development by regional government. The additional investment through borrowing may well generate extra current revenue directly or indirectly. Although borrowing normally involves interest charges as well as capital repayment, the burden of these is often cut by inflation; indeed, if the rate

of inflation exceeds the interest rate, it is cheaper in real terms to borrow money. The real burden of interest charges may also be discounted by the accelerated benefits of the project financed by loan. This is often assumed in case of expenditure on infrastructure such as roads or water supplies - they will lead to an expansion in the economy which will automatically increase current revenues and finance the debt. These arguments are open to question on three counts. Does the particular investment financed by loan actually lead to economic growth? How long does it take for such effects to materialize? Does such economic growth increase the specific revenues which the borrowing authority does or can exploit? An increase in industrial production may only be reflected in rising yields of taxes which may not benefit local government income if they are monopolized by central government. There are other counter-arguments on extensive borrowing by subnational government (Dafflon, 2002). Taxpayers are not faced immediately with the full cost of loan financed projects; this can weaken financial discipline and accountability. Low priority or extravagant proposals might not get the hard scrutiny they would receive if they involved raising taxes or economising on other expenditure. Excessive borrowing, particularly at high interest rates, can build up a heavy burden of debt servicing; irresponsible leaders may win cheap popularity or benefits for their supporters by waste investment, which exceeds any reasonable expectation of increases in revenue. 1) to fund short-term cash flow deficits;

2) 3) 4) 5) 1) 2)

to finance deficits in annual budgets covering operating expenses and debt charges; to purchase plant and equipment with a medium-term life; to finance investment which is expected to earn income; or to pay for long-term capital development.

Borrowing to meet short-term cash flow deficits is quite common throughout the world, which causes usually by uneven patterns of revenue collection. Bank overdrafts are the normal form of such borrowing. Deficit financing of annual current budgets is common for central governments but rarely encouraged or permitted in the case of subnational governments. The recent history of New York is a good example, the city having to borrow extensively to pay its employees and to redeem its maturing debts, including those to its own pension funds. It does happen, by default, e. i. when subnational governments fail to clear off "short-term" overdraft.

3) 4) 5)

The purchase of plant and equipment poses problems in so far as it involves "lumpy" expenditure - substantial outlays on replacement costs at irregular intervals. Borrowing loans for purchase spread over the estimated life of the equipment is a common solution. The concept and practice of loan finance for "self-liquidating" investment is widely accepted. This is expenditure upon operations which earn a direct return on capital covering both debt charges and operating costs. Regional development authorities or utility corporation type are frequently capitalized by loan finance which they are expected to invest in income-earning activities. The use of loan finance for capital development of services or infrastructure without a direct revenue return varies widely.

6.2 Sources and Methods of Borrowing There are several common sources and methods of subnational governments borrowing: Loans from higher levels of government (normally from central government, but state or provincial governments may lend extensively to local authorities in federal countries). Loans from international agencies such as the World Bank, African, Asian, or Latin American Development Banks, or bilateral aid donors - usually made to central government and on-lent. Loans from a central credit bank or loans fund for regional authorities. The issue of interest-bearing bonds or stock, normally with fixed dates of redemption. Loans or overdrafts from commercial or public savings banks.

Mortgages on an authority's physical assets. Internal borrowing from reserve funds such as renewals funds for plant and equipment. Hire purchase or rental leasing of equipment. Contractor finance for construction projects.

The borrowing systems among countries are variable and generalization is difficult, but four dominant systems can be described. The first represents countries where central government is the major source of loan finance. There are countries where the weakness of capital market or lack of creditworthiness makes local and regional authorities dependent upon central government or a central credit institution for loan finance. In contrast, there are countries where subnational governments rely almost exclusively on borrowing direct from the financial market. State governments and local authorities in the United States, for example, raise loan finance almost entirely through bond issues; smaller authorities often join together to float a corporate loan through a bank or a broker. In a third group of countries, a central credit institution operates to provide loan finance for subnational governments, but the latter still obtains the bulk of their loans from the financial market. For example, in UK the Public Works Loans Board provides loans for British local governments. Its scale and government backing allow it to borrow and on-lend at rates marginally below those at which local government can borrow direct. Finally, there are numerous countries where a centralized credit institution is the exclusive or predominant source of loan finance for subnational governments. The Netherlands and Belgian municipal banks, Venezuela's Fundacomun, and Turkey's Hier Bank are examples of institutions playing this role. 6.3 Loans Conditions The loans are different according to their conditions which have to be considered in decision by subnational government to borrow. There are four main characteristics of loan conditions: duration

repayment terms interest rates security

6.3.1 Duration of Loans The duration of loans may be short, medium and long-term, which may be anything from twenty-four hours to sixty years. Local and regional authorities frequently engage in daily transactions on the financial market for short-term cash flow management. For capital investment purposes, authorities frequently seek to equate the duration of a loan to the "life of the project", where a capital asset has a predictable life expectancy. Alternatively, the duration is deduced from the level of debt servicing which can be recovered annually through taxes and charges. In many cases, authorities operate a consolidated loans fund which borrows comprehensively from the market for all loan finance purposes and then "loans out" to individual projects. Ultimately, the duration of loans depends upon the conditions on the market and the interests of the lender. A governmental or international loans fund will often seek to "turn over" its lending as rapidly as possible. A private or institutional investor will be influenced by the difference in the prevailing market rates of return on short, medium and long-term lending, and the timing of any liabilities which will have to be met by the funds invested. There are several alternative methods of repaying loans. The annuity method involves equal instalments combining repayment of capital and interest; the notional interest element will be greater in the earlier instalments, whereas the later instalments will be largely discharging the capital debt. In contrast, interest only may be paid during the life of the loan, the capital being repaid in full at maturity - the normal system with bond issues. This may be accompanied by the operation of a sinking fund to which the borrowing authority pays regular instalments so that the capital to be repaid is accumulated over the life of the loan. In between there are two "loading" methods combining repayments of interest and capital by instalment. "Front end loading" involves repaying equal instalments of capital at regular intervals together with interest on the loan outstanding; the total instalment reduces progressively as the interest elements decrease. By the "rear end loading" method, instalments of capital and interest increase progressively over the life of the loan, sometimes after an initial moratorium or grace period. Rear end loading has two potential benefits: it can enhance the effect of inflation in reducing the real burden of repayment, and it can be helpful in providing a breathing space for a project to develop revenue returns capable of meeting the debt services. 6.3.3 Interest Rates Interest rates on borrowings from the capital market will clearly be dictated by prevailing market trends. Normally, the longer the loan, the higher the rate of interest, though this rule does not always prevail. In the case of subnational government borrowing the market rate of interest may be affected by the fact that subnational borrowing may carry some tax exemption, which increase its rate of return compared with fully taxed investments. A governmental or international aid lender may consider charging rates of interest below market level. In such cases, a distinction is often made between "hard" loans on market terms for directly productive investments expected to earn a commercial rate of return, and "soft" loans for social or infrastructural expenditure which earn no direct income. "Softness" may consist of low or zero interest rates, a moratorium on repayment, an exceptional duration, or some combination of these concessions. 6.3.4 Security of Loans Lenders normally require some security against default on repayment. This may comprise a right to foreclose on the borrower's assets such as investments, land, buildings, or equipment, a guarantee by the central government or a lien over certain current revenues (i.e. a right to first claim upon income). Where the lender is the central government or a credit bank or loans fund under its control, the simplest method may well be to secure loans by a lien over central government grants to the borrower, enabling deduction of outstanding instalments at source. 6.4 Control of Subnational Government Borrowing There has been growing concern in some countries about contribution of subnational borrowing to public deficit and debt. In the last decades, public sector debt has shown a rising trend in relation to GDP in a wide range of countries, both federal and unitary. While these deficits have been recorded in most frequently at the central government level - sometimes because of gap-filling transfers to subnational governments - in several countries, especially but not exclusively federations, deficits and debt have emerged and risen over time at the regional (state or provincial) and local levels as well. It is also argued that excessive borrowing and accumulating of the subnational debt can affect adversely macroeconomic management and stabilisation fiscal policy (Ter-Minassian 1997, Tanzi 1995, Dafflon 2002). A question that arises in this context is to what extent the growth of subnational debt may be promoted, or at least facilitated, by controls and limits on subnational government borrowing. A number of approaches to limit the growth of local and regional government debt has been put forward. These different approaches reflect different constitutional and institutional frameworks, and also depend on the country's financial and economic situation. In general they can be grouped into four broad categories (Ter-Minassian and Craig, 1997):

1) 3) 4)

Reliance on market discipline

2)

Cooperation by different levels of government in the design and implementation on debt controls

Rules-based controls and Administrative control.

1) Reliance on market discipline In practice, several stringent conditions need to be satisfied for financial markets to secure effective discipline on subnational borrowing. First, financial markets should be free and open, so much so that the government must not be put in a privileged position with respect to private borrowers. Second, adequate information on the government's outstanding debt and repayment capacity should be available to potential lenders, who must also be aware of the government's off-balance-sheet operations. Third, lenders should be sure that no chance of bailout exists in case of local government default. Fourth, the borrower should adequately react to market signals before reaching the point of exclusion from the credit market. These are indeed stringent conditions which are unlikely to be realized in the majority of countries. Recognition of these conditions may be a major rason why sole reliance on market discipline to secure fiscal discipline and limit subnational government borrowing is not usual. Only a limited number of countries use this approach, Canada and United States for example. 2) Cooperation by different levels of government Closest to sole reliance on market discipline is an approach where limits on subnational government borrowing are arrived at through a negotiation process involving the central and subnational government levels. The cooperative approach has a clear advantage in promoting dialogue and exchange of informations across various government levels. 3) Rules-based controls A number of countries, both federal and unitary, have relied on approaches to the control of subnational government borrowing that are based on standing rules, specified in the constitution or in laws. Some of these rules set limits on the absolute level of indebtedness of subnational authority; others specify the purpose (typically for investment projects); or maximum allowed service ratio. Rules-based approaches have the advantage of transparency. On the other hand, rules-based approaches lack flexibility and often end up fostering the development of behaviour and practices aimed at circumventing the rules. The direct control may take alternative forms, including the setting of annual (or more frequent) limits on the overall debt of individual subnational jurisdictions (or some of its components, such as external borrowing); the review and authorization of individual borrowing operations; the monitoring of the local government's financial operations and the centralisation of all government borrowing. Control powers generally encompass not only the ex ante authorisation of proposed borrowing, but also the ex post monitoring, on a more or less detailed and timely basis, of the local government's financial operations Direct central government controls are more common in unitary states than in federations and go against the spirit of fiscal federalism. Effectively and timely monitored aggregate limits on the overall debt of individual governments, based on market-type criteria, like maximum ratios of debt service to revenues, would seem preferable to either centralised borrowing or pre-approval of individual borrowing operations. Figure 6.1 presents an overview of subnational borrowing control in selected countries. Figure 6.1 Subnational borrowing controls in selected countries
Country Market Cooperative Rules-Based Administrative Discipline Control Control Control overs. domes. overs. domes. overs. domes. overs. domes. Borrowing Prohibited overs. domes.

Australia Austria Belgium Brasilia Canada Denmark Estonia Finland France Germany Greece Hungary India Ireland Italy Japan Latvia Lithuania Netherlands Poland Portugal Slovenia Spain Sweden Switzerland

United Kingdom United States

6.4.1 Borrowing Capacity How much can local and regional government afford to borrow? A decision by a regional authority to borrow, or by higher levels to permit it to do so, has to be based upon some assessment of its capacity to service the debt. It can be set by central government or be self-imposed by subnational governments, and it can refer to specific purposes. If the loan is "self-liquidating", i.e. for a direct income-earning investment, the concern would normally be only for the internal viability of the scheme and its potential rate of return. For projects which are not directly revenue-earning, capacity to borrow is often judged by some mathematical ratio - of percentage of total outstanding debt to the total revenue, or the volume of debt contracted in any one year's as a percentage of annual revenue, or on amounts of new borrowing , or annual debt service as a percentage of the previous year's revenue or some specifies source of revenue, for example. Where subnational borrowing capacity is assessed on the basis of revenue forecasts, estimates of borrowing capacity are highly susceptible to assumptions about future revenue growth. An example of basic principle how to estimate the debt service capacity may be following: Recurrent Revenues (including taxes, charges and revenue sharing allocations) - projected according to current trends minus recurrent expenditure (projected according to current trends minus existing debt service (projected according to repayment schedules) minus outstanding commitments on revenue funded capital investments equals resources available for new debt service. 6.4.2 Country Experience Local governments are free to borrow in the form of loans or bond issues in all European Union countries. Within the OECD countries with the exceptions of Australia, Canada, Spain (states) and Switzerland, a higher level of government typically imposes borrowing constrains. The requirement of prior approval from higher levels of government is also quite widespread, including permission to borrow in foreign currency as for example in Mexico, Turkey, and Greece. The need for prior approval on a project-by-project basis is gradually being relaxed in OECD countries, such as Mexico, which abandoned such a system in 2000. In Japan and Korea the formal requirement to obtain permission from a higher level of government is being relaxed. In Norway and Spain, prior authorisation can be imposed when subnational governments breach agreed deficits or the proposed borrowing is substantial. A few countries apply limits on borrowing for specific purposes. For example, in Spain, local authorities can borrow up to 30 percent of current revenues to cover short-term liquidity needs, while long-term borrowing is restricted to capital investment. No constraints on access to borrowing are applied in the Czech Republic, Finland, the Netherlands, and Japan. Figure 6.2 shows some examples of borrowing constrain and other conditions in OECD countries. Borrowing can only be used to finance capital expenditure in the vast majority of countries. In the most restrictive cases, borrowing may not be allowed at all (as in Denmark, or in Korea and Spain for current expenditure). In some countries, it can also be used to finance operating expenditure (the Czech Republic, Hungary and Poland) or to finance equity investments in local public companies or non-profit organizations (Belgium, for example). The choice of the lending institutions is usually unrestricted, with only rare exceptions. Most European countries have established prudential rules to avoid excessive debt on the part of local governments. In several countries (Spain, Austria, Belgium, Portugal, Italy and Poland) the rules have recently been consolidated through internal stability pacts introduced within the framework of the implementation of the Maastricht Treaty. The outstanding debt of each local government is notably capped in Austria (the amount varies from Land to Land), Spain (within the framework of the measures taken to comply with the Maastricht Treaty, borrowing approval for the autonomous communities is subject to zero growth in their outstanding debt), Estonia (outstanding debt minus earmarked grants restricted to 60 percent of annual net income), Lithuania (outstanding debt minus earmarked grants capped at 35 percent of annual revenue) and Portugal. In Poland municipal individual debt must not exceed 60 percent of its forecast income for the current year.

Prohibited

Denmark local Korea local (current)

Figure 6.2 Borrowing constrains - access conditions Prior approval is Restricted to No restriction on access to required certain purposes borrowing Canada local Germany local Canada state Japan (capital) Korea (capital Spain local (capital) Norway local Spain local (capital) Portugal local Canada local France local Hungary local Italy state and local Slovak Republic Czech republic local France local Netherlands local Japan local (current) Poland local

Imposed

Turkey local Greece local Ireland local Luxembourg local Mexico local United Kingdom local

Negotiated binding Self imposed

Spain region (current)

Spain region (capital) Switzerland state Canada state

*Note: In the Netherlands, only local governments with balanced budget can borrow and only in Euros Annual debt service ceiling (interest and repayment of capital) have local governments in Spain, Estonia, Hungary, Lithuania, Poland, Portugal, the Czech Republic (30 percent of local current income, for example) and Slovenia. The ceiling for the amount of new loans exists for example in Denmark and Portugal. 6.5 Central Credit Institutions for Subnational Government Now, we will pay special attention to central institutions specifically providing loan finance for local and regional government, because of their growing number and distribution and their importance. The central function of providing a source of long or medium-term credit to subnational local and regional governments is common to all these institutions. Their organizational form varies, however, some operate as a bank and others as a loans board, a trust, or a cooperative society. The former may provide the full range of banking services to subnational government. Their control and management also varies. Some institutions, like the British Public Works Loans Board, or the Japanese Finance Corporation of Local Public Enterprises, are constituted and supervised directly by the central government. Some have been instituted and subscribed directly by the subnational governments themselves, the Municipal Credit Bank of Belgium, for example. Others are jointly owned and controlled by central and regional government. The Bank of Netherlands Municipalities is an incorporated company, its capital subscribed by central and local governments. Share capital - whether from central or local government - is important in determining control of the institution, and in guaranteeing the funds invested in it by the capital market. It is rarely a major source of the funds actually lent by these institutions. These are obtained from a variety of sources: Capital obtained from central government, it takes several forms, including grants, loans, share capital and initial capitalisation. Local government itself. Contributions come from initial share capital or the deposit of reserves and pension funds. Several countries have required local authorities to deposit in the fund the proceeds of a particular tax or a fixed percentage of total revenue.

Capital market, obtained from share capital, deposits, the issue of bonds. Personal savings bank. The example is the federation of communal savings banks in Germany which deposit their surplus balances in banks at state level. Operating profits are another source of capital where loan funds "revolve", the initial capital being increased by interest charges after paying administrative costs. Lastly, international agencies such as the World Bank or bilateral donors occasionally provide assistance to local development through such funds.

Some national institutions provide finance to both local authorities and other agencies for programmes within a specific sector. They may cater extensively to the needs of local government for capital, but not exclusively. Central credit institutions have several strong advantages as sources of loan finance for urban authorities. Most countries, at whatever stages of development, have a substantial capital market comprising private and institutional investors such as pension funds, trusts and insurance societies, which are looking for secure outlets for investment with guaranteed rates of return and fixed redemptions. In some cases, these institutions provide not only loan finance, but technical assistance with the design and execution of development projects as well.

Conclusions

In economic theory there are several arguments in favour of loan finance by subnational governments as well as counter-arguments. Borrowing enables to accelerate investment beyond the limit of current revenue. There is a question if an expansion of regional economy due to additional

investments leads to generation of extra current revenues for financing the debt. Extensive borrowing weakens the connection between the taxpayer and the full costs of loan financed project; it can build up a heavy burden of debt servicing; and it can weaken financial discipline and accountability of subnational governments.

Subnational governments borrow money for funding short-term cash flow deficits, for financing deficits in annual budgets, for purchasing plants and equipment, for financing investment which is expected to earn income, for financing long-term capital development. Subnational governments may borrow from higher levels of government, from international agencies, from a central credit bank, from issuing interest-bearing bonds or stock, from commercial or public savings banks, from internal borrowing from reserve funds, they may hire purchase or rental leasing of equipment, mortgages on an authority s physical assets. In decision to borrow, subnational governments have to consider the main characteristics of loans: duration (short, medium and long-term loans), repayment terms (annuity method; method when only interest is paid during the life of the loan, the capital is repaid in full at maturity - system with bond issues; methods combining repayments of interest and capital by instalment - "front end loading", "rear end loading" methods); interest rates (may differ in case of "hard" loans - based on market terms and "soft" loans with low or zero interest rates) and security (right to foreclose on the borrower's assets, a guarantee by the central government, or a lien over certain current revenues). Approaches to limit the growth of subnational government debt can be grouped into four broad categories: sole reliance on market discipline, cooperation by different levels of government in the design and implementation on debt controls, rules-based controls and administrative control. Federations have moved towards more cooperative approaches, while direct controls are more common in unitary states. Sole reliance on market is limited by stringent conditions which have to be satisfied.

An assessment of debt service capacity can be based on potential rate of return (in the case of income-earning investment) and in the case of not revenue-earning projects by some ratio - percentage of total outstanding debt to the total revenue, or the volume of debt contracted in any one year's as a percentage of annual revenue, or on amounts of new borrowing, or annual debt service as a percentage of the previous year's revenue or some specifies source of revenue.

In all European Union countries, the local governments are free to borrow, however the requirement of prior approval from higher levels of government is quite widespread. Within the OECD countries with the exceptions of Australia, Canada, Spain (states) and Switzerland, a higher level of government typically imposes borrowing constrains. Most European countries have recently established the rules through internal stability pacts introduced within the framework of the implementation of the Maastricht Treaty (Spain, Austria, Belgium, Portugal, Italy and Poland). No constraints on access to borrowing are applied in the Czech Republic, Finland, the Netherlands, and Japan. Borrowing can only be used to finance capital expenditure in the vast majority of countries. The choice of the lending institutions is usually unrestricted. Central credit institutions have become relatively common mechanism for providing a source of long or medium-term credit to subnational local and regional governments. Their organizational form varies, as well as sources of capital. (Capital obtained from central government, from local government itself, from capital market, personal savings bank, operating profits and international agencies).

7. Budgeting
This chapter: explains principles and processes of subnational authorities budget creation, budget roles and dimensions

describes the budget structure - the recurrent and capital budget characterises the types of budget - line-item budgeting, performance budgeting, programme performance budgeting system and devolved budget discusses single and multiple agency budgets describes budget preparation process and budgetary responsibilities explains revenue and expenditure estimation step by step introduces approaches to balancing a budget.

7.1 The Role of the Budget This chapter explains principles and processes in budget creation within subnational authorities. Budgeting is one of the main planning tools in any organisation. It involves estimating the time, amount and costs of resources required to achieve a specific objective. In the private sector the budget may be adjusted regularly during the financial year depending on income from sales. In the public sector, income is much less variable during the year as taxes and grants tend to be fixed annually in advance. Income from fees and charges is a much smaller proportion of income, and even here it is common for tariff levels to be varied during a financial year. Budgeting is generally seen as a means of limiting expenditure and ensuring compliance with approved allocations, making it difficult to vary expenditure during the course of the year. Expenditures that have not been authorised in the budget are simply not permitted. In this context, approval of the budget has become one of the most significant ways for politicians to influence the overall direction and objectives of the organisation. The annual budget occupies the central place in the management process of virtually all systems of subnational government. The budget has several roles: it is a plan for keeping the local and regional authorities solvent, for ensuring that expenditure is covered by o existing reserves o revenue which can be realistically expected o loan which can be obtained and repaid it establishes priorities for the authority's services the budget allocates resources among the different activities of the local authority and determines the levels and directions of work to be undertaken during the budgetary period it determines, so far as the law provides any discretion, levels of taxes, fees and charges to be collected by the local authority during the forthcoming year

the budget provides the legal authorisation for expenditure during the budgetary period it provides comprehensive information upon the financial position and plans of the authority.

The budget as such has three dimensions: 1) a policy making role in choosing how to allocate resources between major outputs in terms of services, infrastructure etc., i.e. in deciding what the authority can hope to achieve 2) a management role in allocating resources to particular agencies to obtain the inputs required to achieve the outputs 3) a control role in giving legal authority to the local authority personnel to collect and spend money, and in prescribing exactly who can spend how much on what. 7.1.1 Budget Structure - Capital and Recurrent Budget Budget should be structured logically to reflect the nature and scope of revenue and expenditures. A traditional form of budgeting normally contains separate estimates of capital and recurrent revenue/expenditure. The capital budget is normally concerned with creation of long term assets - construction of new roads, schools, water treatment plants, etc. The recurrent budget is concerned with the regular operation of services (including the salaries, pension contributions etc. of personnel, the purchase of short life equipment, costs of routine repair, and maintenance) and the servicing (repayments of capital and interest ) of long term debt. In this traditional form of budget, regular revenue - taxes, charges, fees, grants in aid -is credited to the recurrent budget. The only revenue credited direct to the capital budget will be loans or grants specifically for capital project, or revenues from the sale of capital assets -land, buildings etc. If recurrent revenue exceeds recurrent expenditure, a transfer may be made from recurrent to capital budgets so that the surplus can be devoted to capital spending. There are also certain notable "grey areas" between capital and recurrent expenditure. Debt service is sometimes shown as capital expenditure; it should however be included in the recurrent budget if it is a charge against recurrent tax, charging or grant revenues. Vehicles and equipment with a medium term life may also appear in either budget (Davey et all, 1995). Most local governments are used to a single budget document on line-item budgeting. These budgets are normally divided into recurrent (or revenue) and capital. The capital budget will be further divided, with separate heads for revenue capital (that is, capital funded from current revenue), loans and capital grants. In some cases there may be, in addition, an extraordinary budget, with separate heads for deposits, advances and funds. However, there are risks that such an extraordinary budget could be misused to hide certain financial operations, resulting in reduced transparency of the overall budget. 7.1.2 Types of Recurrent Budget and Budget Classification One of the main weaknesses in subnational budgeting is the emphasis on inputs, with little explicit link to outputs and their measurement. The traditional form of subnational budget is normally sub-divided according to the organisational breakdown of the authority, i.e. into a hierarchy of departments/directorates, sections, sub-sections etc, so that the amount allocated to each unit and sub-unit to collect or spend is clearly stated. Recurrent expenditure is further sub-divided into line-items by object of expenditure, e.g. personnel costs, travel, utilities, supplies etc. In the absence of any linked output or outcome indicators it is difficult to measure spending performance and hence fulfilment of policy objective. The only test of budget performance then is whether the money was spent. Performance budgeting (performance review) represents a shift in the focus of budgeting from inputs and work processes to outputs (and, in some cases to outcomes). An output is the quantifiable or observable result of the activity (number of patients treated, for example) and the outcome is the overall and result (improved health status of the population). Performance budgeting is generally characterised by the use of certain mechanism that require organisations to apply measurable targets to the budgetary process. However, the term is considered by some to be imprecise as it does not make a distinction between "outputs" and "outcomes", nor between evidence of internal performance and the effects of external factors. As a consequence other terms have been suggested, including "result-focused budgeting" and "budgeting for results". An example of performance budgeting approach is Programme Performance Budgeting Systems (PPBS), which classify expenditures by objective rather than organisation. Proponents of PPBS argue (see for example Hepworth, 1990) that the traditional budget is concerned with inputs rather than outputs; it does not allow the decision makers to apply a true sense of priority to the allocation of funds because it is not clear how much is spent on particular overall purposes, e.g. promotion of public health may be financed through a range of departments responsible for environmental health, water supplies and sanitation. A full application of PPBS brings together expenditures for the programmes serving major objectives to which a number of operating units may contribute. There are a number of implications that arise from the introduction of some measurements of performance budgeting.

1) 2) 3)

First, departmental managers will need to consider what level of output and outcome they could expect for different levels of spending, forcing departmental managers to consider the nature and costs of outputs. Second, output or performance budgeting may require changes in the classification of the budget, moving form a strict departmental classification to activity-oriented classification requiring greater inter-departmental collaboration. For example, several departments in an organisation might provide service for the elderly. Considering the budgets for these together might suggest alternative ways of working. Third, such a change will require the publication of targets or standards for service outputs alongside the budget. This, in turn, allows subsequent monitoring, auditing and accountability mechanisms to operate more successfully.

PPBS formats have proved complex to install since budgeting across departmental boundaries present control difficulties, defining objectives can cause controversy, and some expenditures may serve more than one objective. A compromise approach more widely adopted is to stick to conventional budgetary sub-divisions by organisations, but

to preface each department's estimates with a statement of objectives and output targets (e.g. provide sewerage connection for 70 percent of households),

to attach performance targets to the vote for each activity (e.g. improve the percentage of refuse lorries in daily running order to 80 percent, to incorporate workload measures to show how variations in the funds allocated relate to variations in the work performed e.g. Activity Year Estimated expenditure Households to be services Refuse collection 2008 100 000 EUR 570 Refuse collection 2009 150 000 EUR 600

A related problem is that of overheads. Although it is may be clear that department's whole expenditure is incurred on specific activities with specified objectives, workloads, targets, etc, these will also create costs for central departments providing administrative and technical support (personnel, accounting, internal audit, store keeping etc). Even department's own central management may divide its attention between a numbers of major "outputs". Internal recharging is often used to overcome this problem so that all expenditure in the budget is charged to activities with a direct output of service to the public.

A more radical response to the problem of lack of output-orientations is to devolve or delegate control of the budget to the manager(s) of the relevant unit (Blore, Devas and Slater, 2004). In the strongest version of this approach the unit manager would receive a "one-line" budget and would be able to set his or her own detailed line-item budget to achieve the policy objectives prescribed by the organisation. For example, a local government could set a one-line budget for each of its schools, giving the head teachers of the schools authority to identify the detailed pattern of spending. Naturally there would be a limit on what the money could be spent on, and also requirements for reporting, monitoring and accountability. This would allow managers to use their specialist knowledge of needs and opportunities to set spending allocations in the most appropriate way to achieve the required results. Devolved budgeting can help to bring accountability closer to those that matter, the service users, and it also relieves central decision-maker of the burden of detailed negotiations with service units over allocations, allowing them to focus on policy-setting. However, a fully devolved budgeting process requires a mechanism for setting the one-line budget for each devolved unit, which may be difficult in the absence of detailed knowledge about the service costs. Devolved budgeting also reduces policy-maker's flexibility to adjust the budget in the light of changed circumstances. A devolved budgeting is likely to require increased budgeting and book-keeping skills in each individual unit, and effective audit and accountability mechanism for unit managers, both for the use of money and for the achievement of specified performance targets. 7.1.3 Budget Duration Local authorities traditionally budget for one year. This may well be the maximum period for which revenue and service cots can be predicted with sufficient certainty to make firm decisions carrying legal authority for execution. There, are, however a number of drawbacks to annual budgeting. Major capital development can rarely be completed within twelve months, so the budget should include at least implied commitment to the full cost over two or more years, and also to the extra recurrent costs of staffing supplies etc which may well follow completion. Borrowing also involves an obligation to service a debt over the life of the loan. To solve this problem, it is normal to show the full costs of a capital scheme in the budget, while only authorizing finally the amount to be spent on it over the budget year. An estimate of the future operating costs should also be included in the memorandum accompanying the budget, so that the full costs implications are shown. This should help the authority to see whether the scheme is affordable and also what trade-offs may exist between construction and maintenance costs. Very often apparent savings in capital costs may have expensive consequences in recurrent operation. Whether authority can really afford the long term consequences of decisions will only be fully apparent from a multi-year forecast of overall revenue and expenditure. This is often attempted on a rolling basis (Hepworth, 1990), i.e. the forecast always looks x years ahead (normally three years); each year the forecast is revised to take into account inflation rates and other unpredictable changes, and rolled forward one further year. 7.1.4 Single or Multiple Agency Budgets Some local authority activities may be funded, wholly or in part, from revenues which are specifically collected for that particular purpose. It may also be the intention that such activities are "self-financing", i.e. paid for entirely by the users, so that any charge on the general tax payer should be limited and therefore clearly shown and monitored. For such purposes, budgets must clearly show the relationship between the expenditure, its related revenue sources and any subsidy from general revenue. To achieve this degree of budgetary isolation or commercial flexibility, local authorities often set up a subsidiary agency or enterprise to undertake a self financing activity such as a water supply, markets or public transport service, with their own accounts balances and budgets. Any subsidy from general revenue will be shown as a revenue item in the undertaking's budget and as an expenditure item in the parent authority's main budget.

Municipal activities with specific revenue sources may be too small or too many to justify full separation from the main budget. A compromise approach may be to operate "special accounts" or "undertakings" only for one or two major self financing services such as water supplies, but to show tied revenues resources as "appropriations in aid" with the relevant expenditure within the major budget e.g. Operation of sports centres Less admissions fees and hire charges Net general fund expenditure 70 000 EUR 25 000 EUR 45 000 EUR

There are further refinements to the relation of expenditure to the specific revenue sources. The authorization of expenditure can be tied to the amount of actual revenue collected. In the sports centre example given above, the amount of expenditure authorised by the budget could be one of the following options:

a) b) c)

70 000 EUR regardless of how much revenue is actually collected; 45 000 EUR plus the actual amount of revenue collected up to a maximum of 70 000 EUR or 45 000 EUR plus the actual amount of revenue collected, regardless of the amount.

Clearly options (b) and (c) provide the operating department with greater incentives to market a service and collect revenue efficiently, and also to vary expenditure in response to consumer demand. Option (a) would only be sensible where the service was a significant "public good". 7.1.5 Preparation Process and Budgetary Responsibility Budget preparation will normally follow a prescribed process and calendar covering strategic goal setting, revenue and expenditure forecasting, budget debate and approval. Budget preparation normally involves a series of stages: Submission of first draft estimates of revenue and expenditure in respect of each organisational division and sub-division. Scrutiny of departmental submissions by the central budget staff and initial discussion with the departments. Analysis will focus initially upon the accuracy of the estimates in terms of costing and legality, but will also seek to establish a sense of justification and priority for proposed increases as well as the possibility of economies. Formulation of a draft comprehensive budget with options over revenue increases, expenditure increases/reductions etc. Executive decision on the final draft of the budget.

Enactment of the budget by the authority's governing body. Approval of the budget, if required by law, by a central government ministry or other supervisory body.

Responsibilities for these various stages vary according to the statutory and internal management framework of the local or regional authority. Possibilities are: Submission of first draft estimates: preparation may be done by the operating departments, or by central budgeting/accounting staff in consultation with them. Council committees may also discuss and approve drafts for services under their supervision at this stage. Scrutiny of the drafts: normally conducted by the chief executive (whether an elected mayor or an appointed City manager/Town Clerk) or under a chief finance officer (treasurer, director of finance etc). Formulation of a draft comprehensive budget with options: normally the duty of the central budget staff. Various possible options can be assessed. Decision on the final draft of the budget: either by the Chief Executive, or the Finance Committee, or by a "Cabinet" of senior political leaders. Enactment of the budget: normally by an elected council. Approval of the budget: by a central or regional government minister, or by a provincial governor or commissioner as required by law. A supervisory authority may also scrutinise and comment on the budget before it is enacted. In some countries there is a wider public participation in the budget process. It involves shifting from a traditional exclusive method of budgeting by the executive to an inclusive method that empowers citizens to play a direct role in the planning and allocation of municipal resources. This process is called participatory budgeting (Blore, Devas and Slater, 2004). The concept of participatory budgeting implies a broad representation of people and interests in decision-making about the allocation and disbursement of resources for local delivery. Business firms, voluntary and charitable organisations, neighbourhood associations etc will lobby the Council either for direct assistance or for extra expenditure on particular services. There may even be formal public hearings at which local authority officials explain major budget proposals and members of the public have the opportunity to question them and voice opinions. The format of the budget that will be presented publicly should include user-friendly graphs, pie charts and other visual devices to convey information to citizens on where funds come from and how they are spent. 7.2 Revenue Estimation Revenue forecasting is often a weakest aspect of the subnational budget and poor revenue forecasting can seriously undermine the budget as a tool of financial management. In order to improve the revenue estimates it may be useful to concentrate analysis on the few larger items of revenue and the key variables (inflation, housing, trade, etc) that relate to such items. Revenue estimates should be itemised under individual taxes, charges, fees, loans, grants etc. Comparisons over time should be shown, including in respect of each item, actual receipts in the previous year, and probable receipts during the current year, and estimated receipts in the forthcoming year for which the budget is being formulated. Actual revenue 2008 Probable revenue 2009 Estimated revenue 2010 7.2.1 The Existing Revenue Base The first stage is to estimate revenue from existing sources on existing conditions, i.e. according to the existing tariffs, the existing rules about liability and assessment and the existing methods of collection. These estimates will normally be based upon projections of recent

experience. It is important, however, to analyse each revenue item with regard to the specific factors which cause variations in yield from year to year. Tax yield may vary for example: with changes in the specific tax base, e.g. the number and size of buildings bearing a property tax, or the number of vehicles subject to a registration tax, or with fluctuations in the general economy, influencing the yield of revenues such as a local sales tax. Charging revenues will vary with the volume of service offered or demanded by the public. Once the causes of variation are understood, their impact upon revenue yields in the forthcoming year can be predicted. Above all, it is important to be realistic and to avoid the temptation to support authorisation of increased expenditure, simply by inflating revenue estimates beyond reasonable expectation. It is then necessary to estimate what further increases in revenue could be obtained by specific changes in the existing revenue sources. Such changes could be of several kinds: Widening the scope of the tax or charge and making more people or transactions liable to the payment, e.g. by removing existing exemptions. Changing the assessment methods so that some or all payers become liable for a higher rate of tax or charge. Increasing the tariffs, i.e. the rates of tax or charge.

Changing collection methods to reduce evasion or accelerate payment.

7.2.2 Changing the Revenue Base The criteria for evaluating any proposals for specific changes are those already discussed in Chapter 3 and Chapter 4. As a first step it is necessary to identify the underlying purpose of each revenue source, whether it is meant to be A contribution to general expenditure or a recovery of the cost of particular service. If a recovery of a particular cost, whether it is meant to be a straight recovery of the cost, a recovery with profits, a recovery less an element of subsidy to all users, or a recovery less a subsidy to certain classes of user. A "flat" levy or redistributive, i.e. fall more heavily upon the rich than the poor. A second step is to appraise the recent performance of each revenue source. Two major questions to be asked are In the case of a charge, have yields kept pace with the costs of the service for which it is levied? In the case of a tax, have yields grown at the same rate as the base upon which it is charged, e.g. have property tax yields kept pace with the growth of property and its value, have sales tax yields kept pace with the growth in local earnings, has a personal tax revenue grown in proportion to local population? A further step in appraising the potential for changes is to compare revenue yields with those in other comparable authorities. In the case of taxes, a per capita comparison may provide a quick indication of any striking contrast in performance; in the case of a charge, the comparison would have to be in terms of revenue per unit of service. Where another authority appears to achieve considerably higher yields, it would then be necessary to examine differences in coverage, tariffs, collection methods etc, to see where the greatest possibility of improvement lies. Finally consideration should be given to the possibility of introducing new sources of revenue, whether taxes or charges. Clearly there are statutory, political and administrative obstacles to new revenues; these are usually greater in respect of new taxes than new charges. Improving an existing source is usually easier than introducing a new one. This stage of revenue estimation will provide an initial estimate of the range of revenue expectations to be compared with the expenditure estimates at a later stage. The minimum will be the expectation from existing sources at existing levels to which can be added the yields of different possible increases, ranked in order of their desirability and feasibility. 7.3 Expenditure Estimation As discussed in part 7.1.2 of this chapter, it is normal to classify expenditure according to the departmental organisation and then by the different activities undertaken within such divisions. In conventional budgeting, expenditure on an activity is further broken down into costs items such as salaries and wages, transport costs, telephone charges etc. Detailed costing of inputs will clearly be needed at the initial estimate stage to support the proposed expenditure on each activity. The alternative possibility of classifying budgets by major objectives has already been discussed above. An intermediate approach to this problem is that of "recharging". This practice relates to departments and sections, e.g. personnel, motor transport pools, stores, which exist to service other departments rather than to provide a direct service to the public. A budget is prepared for such a section, but its expenditure is then recharged to the functional departments in proportion to their use of its services; these departmental charges are shown as revenue to the service department, cancelling out its expenditure, so that the cost of its work is shown as a net charge upon the different functions of the organisation. Clearly, such a process enables the decision maker to see the real costs of the functions of the organisation, e.g. the real costs of education comprise not only the expenditure of the education department, but also the costs of the central accounting staff who have paid the teachers' salaries and the personnel staff who keep the teachers' employment records. But recharging complicates the budgeting and accounting process and how far it can be practised depends upon the administrative capacity of the organisation. 7.3.1 The Existing Expenditure Base Most government budgeting, central or subnational, tends to be on an incremental basis, i.e. existing costs are accepted as a base to which amounts are added for inflation, expansion in demand or improvements. The weakness of "incrementalism"is that it accepts the historic costs of a service as given (Davey et all. 1995; Blore, Devas and Slater, 2004, Hepworth, 1990) and does not examine such question as: Is the service provided by the expenditure still justifiable or of sufficient priority to merit its current share of resources? Is the service fairly distributed among its consumers? Geographical inequity is often prevalent in municipal services, such as roads, parks etc.

Is the service being performed efficiently and economically?

The alternative is to use a zero base approach (Wildavsky 1989, Hepworth 1990). This basically would ignore current provision and estimate expenditure from scratch according to the following steps: decide what kind of service is needed (e.g. refuse collection)

set a standard of provision for the service (e.g. twice weekly collection) estimate the unit cost of the service at this standard (e.g. 10 EUR per ton of refuse collected) estimate the quantity of service needed to meet the standards (e.g. removal of one hundred tons of refuse per day) estimate the total expenditure needed to perform the service, multiplying the quantity of service by the unit costs (e.g. one hundred tons of refuse per day at 10 EUR per ton = 100 x 10 x 365 = 3 650 EUR) There are obvious difficulties in applying the zero based budgeting (ZBB) to the whole of the budget. As Blore, Devas and Slater (2004) noted ZBB can lead to an enormous amount of work scrutinising alternatives for each item of spending each year. Some services are much easier to define in terms of standards, units of provision and unit costs than others. Assessments of needs, standards, and priorities are basically subjective; to define them all from scratch would lead to endless inconclusive debate, whereas definition may be much easier to agree in a particular case, e.g. when a particular service is clearly perceived as substandard, or when a choice of priority has to be made between a very limited range of "hard" options. One compromise would be to carry out periodic through reviews of particular services, especially those with potential for efficiency gains, to identify alternative ways of delivering service. Nevertheless the existing expenditure base should not be accepted without question. Possibilities of savings through increased efficiency need to be examined. Unit costing is a particularly valuable instrument in such examination, i.e. the expenditure per unit of service. Various comparisons of unit cost can be made:

Between local or regional authority's services and those of other comparable authorities.

Between the costs of services in different parts of the city, e.g. between different schools, different road maintenance gangs or refuse collection units. Between the costs of different departments incurring a similar type of expense, e.g. costs per mile of lorries. Between the costs of the same departments and service over time, e.g. costs per ton of refuse collection over the last five years (after discounting inflation). Some services may not be readily reducible to units of provision because output is not easily quantifiable or there is not uniformity in the type of service provided. In such a case a comparison (between authorities or over time) of expenditure per capita may be a proxy measurement. Comparisons of unit costs (or expenditure per capita) do not themselves prove anything. They simply indicate the areas of expenditure which need more detailed examination.(For example if costs per unit have been rising overtime, is it due to improving quality or to decreasing efficiency?) Close analysis of all existing expenditure may not be possible in one budget preparation. Selectivity may be necessary. Simple comparisons of expenditure per capita at constant prices on each service either over time or between comparable authorities should provide an initial basis for selectivity. It should highlight expenditure fields where is apparent scope for substantial economy through improvements in efficiency or there is apparent under-financing and need for increased investment. The existing expenditure base therefore needs some examination in terms of

7.3.2 Increases in Expenditure

its potential savings and greater efficiency any lessening public demand for the service, or any desirable redistribution of costs to achieve a more equitable provision of service.

Transfers of responsibility to other agencies may be another ground for savings.

Once the expenditure base has been determined, a number of potential increases have to be examined. The first set to be considered are those arising from inflation. In the absence of better information, these may be deduced from a general rate of inflation prevailing in the economy or in the organisation. However, it would be more accurate to look at each type of cost, e.g. wages, building costs, fuel costs, general supplies, and project an individual rate of inflation for each of these according to the indices available. The next set of increases to be considered are those arising from prior commitments by the subnational authority, i.e. those increases which are unavoidable. These may result from previous decisions; for example, if capital development is in progress, increases in expenditure may be committed in the following year under construction contracts, to repay part of the loan or interest upon it, and to meet the operating expenses of the completed project. Alternatively, additional expenditure may be imposed upon an authority because of the rising public demand for service which it has an inescapable obligation to provide, e.g. emptying the dustbins on a new housing estate. It is important to ensure, however, that the increases in expenditure proposed are in proportion to the increased amount of work involved, and take advantage where possible of economies in scale. In this respect use should again be made of unit costing. The final sets of increases to examine are uncommitted proposals for improvement or expansion of services, i.e. those involving expenditure which the authority still has the opportunity to reject or postpone. Uncommitted proposals for improvement/expansion of services have to be considered within an objective analysis of public service needs. Needs are affected by Demographic trends, arising both from natural growth and migration, leading not only to changes in total population, but also to its age distribution. In many cities, for example, the rate of increase of school aged children may be faster than that of the population as a whole. Economic trends leading to expansion or decline in particular types of business and employment with its consequent demands upon infrastructure. Physical trends, leading to growths of urban settlement in particular localities, but also to the decay of existing settlement in others.

Where major changes in provision are required to meet such needs, they cannot be achieved within a single year budget. A medium term plan of both capital and recurrent expenditure is required so that realistic targets can be set for achieving required improvements and expansions over time. Such plans may not be necessary for all service -simply for those where major deficiencies exist and/or significant changes in need can be anticipated. Medium term forecasts of the expenditure needed for major changes in service provision may well result in demands for funding at an unattainable level. Only medium term forecasting, e.g. over three to five years, can provide the basis for such judgement and for assessing the measures needed to overcome the problem. Solutions may include: A fundamental reassessment of the standards of service provision, with substitution of lower cost solutions. Partial privatisation, i.e. leaving certain types of service to be provided by the private sector. A renegotiation of financial relations with central government, leading to central government undertaking more responsibilities itself, or assigning additional grand aid, or assigning additional revenue sources, and revising statutory limitations on local and regional taxing and charging powers. To summarise, this stage of expenditure estimation will produce: A base aggregate expenditure consisting of current expenditure minus transfers of responsibility, minus reductions in workload, minus other economies, plus inflationary increases, plus committed increases in workload/obligation.

A set ofproposals for uncommitted increases, preferably in rank order ofpriority. This together with the initial revenue estimates described in previous part of this chapter provides the basis for the further stages devoted to balancing the recurrent budget. 7.4 Capital Budgeting Capital expenditure proposals are usually only approved after more rigorous screening. Capital expenditure budgeting involves a variation in format. Expenditure is based more upon the phasing and progress of a single project over a number of years ,rather than fixed allocations for fixed time periods. The estimates should therefore include details of the following: The total original estimate for the project. The total revised estimate, taking into accounts subsequent amendments and increases in cost. Total expenditure up to the end of the previous year.

Total anticipated expenditure by the end of the current year. Anticipated expenditure during the forthcoming year (for which the budget has been formulated). Estimated unexpended balance of the project at the end of the forthcoming year.

The approval of recurrent expenditure is ultimately dependent upon the reconciliation of overall expenditure with overall revenue. Capital expenditure, by contrast, is normally financed on a project by project basis. Revenue from recurrent budget surpluses, bond or stock issues or sale of capital assets may be generally available for capital expenditure, but otherwise a source of funding is generated for each scheme. Sources of capital finance include (and they have been discussed in chapter 6):

revenue surplus external loans internal borrowing stock or bonds issues sales of assets hire purchase/contractor finance.

The capital budget should clearly state the source of funds for each project. Authorisation will then depend upon the realisation of the revenue source, e.g. a loan funded scheme can only be commenced once the loan has been obtained etc. The exception relates to those projects financed from revenue surplus, i.e. from funds already at the authority's disposal at the time of budgeting. It is not always easy to predict exactly how much capital expenditure on a project will be completed within a financial year. Normally unexpected balances will be automatically devoted in the following year's budget, assuming that these are supported by unexpected balances of the loans, grants, reserves by which the scheme is financed. A central element of the capital budget and the multi-year capital investment plan is the financial analysis aimed at minimizing the financial effect of investment project(s) on the subnational budget, while maximising the impact or returns of the project in addressing community needs. Methods of appraising capital expenditure proposals are not discussed here. They involve techniques of project appraisal which are the subject of separate publications and exercises. 7.5 Balancing the Budget The first question in balancing a budget is whether to estimate for a surplus, breakeven or deficit. This is partly determined by the state of the subnational authority's reserves. Reserves are needed: to provide a cushion against a large unexpected demand for expenditure or shortfall in revenue;

to insure against year to year fluctuation in revenue (often associated with climatic variations or economic conditions); to provide working balances where expenditure tends to run ahead of revenue in time;

to accumulate resources to meet a large future commitment such as major renewal of plant and buildings.

Reserves must be reviewed at the beginning of the budgetary process; if they are considered inadequate, the authority should seek to replenish them by budgeting for a surplus of revenue over expenditure during the year. Alternatively the authority can afford a deficit on its budget to the extent that its reserves are above the safely level. One option is to cover a budgetary deficit by external borrowing. There are (controversial) economic arguments for this practice, but it runs the risk of encouraging a level of expenditure and indebtedness which eventually outstrips the authority's capacity to service its loans and retain the confidence of potential lenders. Self Financing Expenditure Another issue is to distinguish self-financing from general fund expenditure.Items of expenditure which are supposed to be directly recoverable should be compared with the appropriate revenue estimates and any net "profit" or "subsidy" carried forward to the general revenue or expenditure totals respectively. If the expected revenue does not cover the expenditure on such items, possible cuts in cost or increases in charges should be examined to see if the subsidy from general revenue can be eliminated or reduced. 7.5.1 Alternative Approaches to Balancing a Budget There are three optional approaches to balancing a budget which can be considered (Davey at all., 1995): The first approach is to decide upon a total expenditure ceiling and direct each department to budget within a share of this, allotted in advance. The aggregate expenditure would be calculated after examining the state of reserves and revenue estimates, deciding firmly at what level revenue charges and taxes were to be fixed. The breakdown of this total into departmental allocations might be based upon a flat average increase over the current year's total. It is better, however, if this could incorporate some variation based upon priorities; these weightings might be based upon a generalised statement of priorities or, more satisfactorily, on agreed long range forecasts of expenditure. The second approach is to call for revenue and expenditure estimates independently and then reconcile them stage by stage. At the fist stage the lowest priority, uncommitted expenditure increases would be cut until aggregate expenditure was supportable by the outer limits of possible revenue collections. At the second stage the merits of the higher priority uncommitted increases would be compared with those revenue increases which were regarded as feasible but not necessarily desirable. Hopefully a balance could then be struck. In a tight situation however the feasible revenue increases might only just cover committed expenditure necessitating the elimination of all uncommitted proposals. If the budget could still not be balanced, economies in committed expenditure would have to be achieved; the normal approach would be to look for reductions in overhead expenditure first, but ultimately cuts in output and service might have to be considered. A third and intermediate approach is to establish an expenditure ceiling for each department after the first stage of reviewing its estimates, i.e. cutting out the lowest priority expenditure, then setting revenue levels and expenditure aggregates and passing the responsibility back to departments to achieve the further reductions necessary. This may be necessary where it is politically or administratively impossible for the financial mangers to determinate all the real opportunities for economy themselves. Departmental ceiling established at this stage can be weighted more accurately to take account of the varying needs and commitments revealed by the first set of estimates, if there are no longer term forecasts as a guide. Arbitrary cuts in overheads, freezing of staff vacancies, etc. may be the only effective way of achieving economy. However, where cuts in output and service are required, these need to be specified to be effective. Otherwise the effect of the economy can by ignored or avoided by such means as postponing work or payments or bills, but leaving a full level or commitment for further years.

Conclusions

Proper budgeting and financial management must adequately control the total level of revenue and expenditure; appropriately allocate public resources among sectors and programs; and ensure that governmental authorities operate as efficiently as possible. These issues represent three dimensions of a budget: budget as a policy making tool,a management role and a control role. Budget of subnational governments is a planning tool for keeping the subnational authorities solvent, establishes their priorities for services, allocates resources among different activities, determines the level of taxes, fees and charges, provides authorisation for expenditure, and provides information to the public. A traditional form of budgeting consists of separate estimates of recurrent revenue and expenditures and capital spending. The recurrent budget is concerned with the regular operation of services, including the salaries, pension contributions etc. of personnel, the purchase of short life equipment, costs of routine repair and maintenance, and the servicing of long term debt (principal and interest). The capital budget is largely concerned with the creation of long-term assets (roads, pipes, schools, water treatment pants).

Three types of recurrent budgets are used in practice: the line-item budget, the performance budget and the program performance budget. Traditionally subnational budgets are planned and approved on a detailed line-item basis. Both the program and performance budget represent a shift in the focus of budgeting from inputs and work processes to outputs or outcomes. They apply measurable targets to the budgetary process. A budget may be a combination of more than one type of a budget.

The other approach which responses the problem of lack of output orientation is devolved budgeting. It devolves or delegates control of the budget to the manager(s) of the relevant unit. After receiving a one-line budget for each devolved units, each manager is responsible to set own spending allocations in the most appropriate way to achieve the required results. Devolved budgeting can help to bring accountability closer to the service users.

Subnational authorities may in some cases separate revenue sources and expenditures from the main budget. They often set up a subsidiary agency or enterprise to undertake a self financing activity with their own account balances and budgets to clearly show the relationship between the expenditure, its related revenue sources and any subsidies. A compromise approach may be to operate "special accounts" or "undertakings" only for one or two major self financing services.

Budget preparation involves a series of stages: submission of first draft estimates of revenue and expenditure, scrutiny ofdepartmental submissions by the central budget staff and initial discussion wit the departments, formulation o f a draft comprehensive budget,

executive decision on the final draft, enactment of the budget by the authority's governing body and approval of the budget, if required by law. In some countries also citizens and public play a direct role in budgetary process and decision-making about the resource allocation - so called participatory budgeting.

To make sure budget serves properly as a tool of financial management, it is necessary to focus on a well-prepared revenue estimate. The estimation should be based upon projections of revenue from existing sources on existing conditions including the estimates of further increases in revenue which could de obtained. Last but not least it is important to consider the potential for changes of revenue with evaluating the recent performance of each revenue source, with comparing revenue with other comparable authority or with introducing new sources of revenue. The estimate of expenditure is traditionally based on an incremental basis existing costs are accepted as a base to which amounts are added for inflation, expansion in demand or improvements. It accepts the historic costs of a service as given without examining the authorization of the costs or service itself. The alternative approach is zero based budgeting which ignore current provision and estimate expenditure afresh each year with no presumption that the past spending patterns are appropriate. ZBB can lead to an enormous amount of work scrutinising alternatives for each item. In the expenditure estimate process it is desirable to examine possibilities of savings through increasing efficiency by comparing unit costs (or expenditures per capita) o f services between comparable authorities, between costs of services in different part of city, between the costs of the same departments and service over the time. The following stage provides more detailed analysis of costs comparison on selected spending areas. Capital expenditure is normally financed on a project by project basis. Sources for financing capital expenditure are: revenue surplus, external loans, stock and bond issues, or sales of assets. The capital budget should clearly state the source of funds for each project. The budget may be estimated with surplus, breakeven or deficit. Three optional approaches to balancing a budget can be considered: approach based upon establishing total expenditure ceiling and breakdown it to the each department. The second approach is to prepare revenue and expenditure estimates independently and then reconciles them stage by stage. The intermediate approach is to establish an expenditure ceiling for each department after the first stage of reviewing its estimates.

8. Budgetary and Financial Control

This chapter:

discusses the purposes for budget control, the responsibility for control within voting budget and budgetary adjustment characterises the key concepts to be considered in monitoring revenue and expenditures and the cases for budgetary correction

describes the main principles and problems of revenue and expenditure control explains the necessity of cash-flow management sets up other aspects of financial control not directly related to the execution of the budget

8.1 Budgetary Control The previous chapter was focused on budget creation. However, no budget is meaningful without adequate implementation. Therefore, this chapter is dedicated to the issue of the implementation of the budget and its control. There are other aspects of financial control such as control of staff establishment and emoluments, control over the purchase of goods and services, and other miscellaneous financial controls. These issues will be dealt with in the second part of this chapter. An intrinsic part of a successful implementation is also an efficient audit, both internal and external. Although audit as such is not elaborated in this book, we should remember it is essential for efficient control of financial management of local and regional authorities. 8.1.1 Purposes of Budgetary Control Budgetary control has a number of purposes: to ensure that the authority stays solvent, that expenditure does not exceed its revenues and reserves

to ensure that revenue is collected and money spent legally, i.e. that it conforms with the limits authorised through the approval of the budget to ensure that the members of the public meet their legal obligations to pay for public services through taxes and charges to ensure that money is spent efficiently to ensure that money is handled honestly.

Responsibility Budgetary control can only be effective if personal responsibility for collecting each revenue source and incurring expenditure under each budget item is clearly identified. After a budget has been approved, it is normal to issue some form of warrant to those authorised to incur expenditure, specifying the items of expenditure under their control and the approved provision under each.

1)

The degree of delegation of authority to incur expenditure varies between regional and local government systems and between individual or local authorities. There are roughly three patterns: The Chief Executive (the Mayor, Commissioner, City Manager etc) or the Financial Manager (Treasurer, Director of Finance, Chief Accountant etc) has to authorise all expenditure and makes all payments; operating departments have to submit requisitions to him or her when they want to hire staff, order supplies etc.

2) 3)

The operating departments authorise expenditure within budgetary limits, but the central treasury makes all payments. The operating departments authorise expenditure within budgetary limits and make payments.

A variation (1) or (3) may require the approval of the Chief Executive, Financial Manager or a Committee for items of expenditure above stated amounts of money.

Much depends clearly upon the size of the authority and the ability and trustworthiness of its staff. Generally speaking, it is better to delegate authority to incur expenditure to operating departments within a voted budget since: it helps the department to plan its work

it avoids delay in undertaking the departments work it places responsibility squarely on the department to keep the scale of its commitments within the money voted to it.

Centralised control can lead to irresponsibility since the operating departments feel no compunction to be careful about the requisitions they submit, and the chief executive or treasurer lacks knowledge and time to scrutinise the bids he or she authorises. The same principles of delegation can be applied within operating departments. In a larger authority departments may have zonal branches which effectively supervise operations such as road maintenance or refuse collection in their particular areas of the city. There may be other clear operational sub-units such as schools, sports centres or large markets. There are considerable advantages in delegating responsibility for the control of expenditure to sub-units so that they can plan their operations and be held fully responsible for the efficient use of resources. Budgetary Adjustments Some variation in the budget is inevitable during the course of the year. It may be necessary for a spending department to seek an absolute increase in the amount of money allocated to it; this is usually known as a "supplementary estimate". Alternatively it may be possible to reallocate money so that overspending on one item is off-set by under spending on another; this is known as "virement". Supplementary estimates almost invariably require approval by the authority's Chief Executive or Treasurer and often by its Council since they affect the overall budget and financial prospects. But discretion is often given for departments to authorise virements, perhaps within fixed financial limits. This discretion again speeds up work and encourages departments to accept responsibility for matching unexpected commitments in one direction with economies in another. 8.1.2 Principles of Revenue Control Assessment of the amount of taxes, fees or charges to be paid by individuals must be conducted strictly and impartially. The aim is to charge the exact amount due to the authority - neither more (to swell the collections) nor less (out of favouritism or leniency). The assessor is usually acting in a judicial rather than a money raising capacity, i.e. he or she is applying a given set of rules to the situation of the individual payer. Random checks are necessary to verify the work of the individual assessors and also the information supplied to them by the people being assessed. Many taxes or charges include some provision for remission or exemption. These often benefit old people, school children or the physically handicapped, and sometimes those in lower income groups. Again the rules must be strictly and impartially applied. Those responsible for granting exemptions must be aware of their underlying purpose; for example poverty due to a physical handicap may be a good reason for exempting a tax payer, but the physical handicap itself may not be a sufficient cause. Again it is necessary to make random checks on the award of exemptions and remissions. Progress in collecting revenue must be continuously monitored. A strict timetable must be set for each tax or charge stipulating the period within which payment is due, intervals at which reminders are automatically sent to defaulters, and the date at which legal proceedings or other sanctions (e.g. cutting off a water supply) will be taken, again automatically. In the case of taxes and charges for which people have a regular continuing liability (e.g. rents), it is necessary to keep an account in respect of each payer showing the amounts due, amounts paid, and any outstanding balances due to the authority. These must be kept up to date if payment is to be enforced strictly. Apart from instituting a regular phased procedure applying to each individual liability, general progress checks are needed. Comparisons of the total amount or the percentage collected with a similar stage of previous years are helpful guides. If collections are falling significantly behind normal experience it will be important to verify the reasons as quickly as possible. The reasons may be a decline in administrative effectiveness requiring a general tightening up. Alternatively it may be due to an external economic disturbance which means adjusting the budgetary estimations as a whole. 8.1.3 Expenditure Control Whoever is responsible for authorising expenditure on a particular budget item must have regular and up-to date information on the total amount spent, and therefore the balance still available for commitment. This means that expenditure must be brought promptly to account in the authority's ledgers and statements of spending to date must be provided regularly and promptly to those controlling expenditure on each item. Although an operating department may keep its own records of the expenditure it has authorised, it is important that this be reconciled regularly with the records in central accounts. It is, for example, important that the departments should be aware of charges that are made to their votes without their prior authorisation such as central overhead charges, debt servicing, salaries or internal recharging. Unless the local authority is operating a full accrual system of accounting (which is rare), ledger accounts may only reflect payments which have actually been made. The total of such payments is not, however, an adequate basis for calculating the amount "unspent" and available for fresh commitment. This is due to the usual time-lag between a decision to incur expenditure, e.g. by ordering supplies, and the consequent cash payment. It is therefore necessary for those authorising expenditure to keep an accurate record of the commitments they have undertaken, so that they know the true unspent balance available to them. The cost of goods or services, for example, should be recorded as expenditure the moment they are ordered, rather than when the bill is paid, since the order effectively uses the money involved. Before payments are made, they must be checked to ensure that no errors or fraud are involved. For example, wage and salary sheets need to be checked against establishment records, official wage levels and, often, against the physical presence of the labour force. Payments for goods need checking against agreed contract prices and stores records verifying their receipt and location. Various procedures are used in different authorities, e.g.: Payments have to be authorised by two people, one from the spending department and one from the treasury (possibly the internal audit section). Cheques have to be signed by two people.

 Wage and salary payments are made by a different section from that preparing the wage sheets, etc. Expenditure may be within approved limits and honest, but still wasteful. It is the duty of the Treasury (usually with the help of internal auditors) to check on the efficiency of the authority's expenditure. Example of the sorts of issuer that often need to be investigate are: Comparing costs of direct building and repairs with the use of contractors.
Examining the running costs of, and utilisations of vehicles. Comparing the costs of alternative means of reproducing documents - printing, photocopying, etc.

8.1.4 Monitoring Revenue and Expenditure Revenue and expenditure levels under each item need to be reviewed regularly to ensure that they conform to legal authority and that solvency is maintained. One can only judge progress at any point in the budget year, however, against an expectation of what levels of revenue and expenditure should be at that stage if the budget is being implemented "according to plan". In some cases one would expect an even pattern of collection or spending throughout the year. This might apply, for example, to staff salaries (assuming all salary increases date from the beginning of the year), or to collection of rents on permanently leased shops or market stalls. But in many cases such patterns are either seasonal or lumpy. People may be required to pay property taxes or licence fees within so many months; market fee income may be higher during a harvesting season or surrounding an annual festival. Annual or bi-annual payments such as insurance premia or debt service charges may lead to particularly heavy expenditure in certain months. Budgetary progress can only be effectively monitored if such irregularities in the expected pattern of revenue collection and expenditure are charted. This means drawing up a profile of collection and expenditure at the beginning of the budget year. A sample profile for road maintenance and total approved expenditure 2 million EUR might be as follows (see Figure 8.1). The amount actually spent at the end of any particular month can then be compared with the profile to see whether any abnormality is really occurring. Taking the example, if expenditure at the end of six months was around one million EUR, i.e. 50 percent of the vote, it might easily be regarded as running at a "safe" level. But comparison with the profile shows that such a level creates serious dangers of over-spending since more than 50 percent is required in the second half of the year to cope with the exceptional commitments in September and October. Figure 8.1 An example of expenditure profile for road maintenance (in thousand EUR) Month Expected expenditure Total Expenditure to date % o f total expenditure Jan 120 120 Feb 120 240 Mar 260 500 Apr 120 620 May 120 740 June 120 860 July 130 990 120 1120 Sept 270 1390 Oct 350 1740 Nov 130 1870 Dec 130 2000

12

25

31

37

43

49,5

56

69,5

87

93,5

100

Regular up-to-date reports of budgetary progress are required to ensure that no serious deviations from the budget are taking place, or that corrective action is taken. A variance analysis will provide senior executives with figures whose significance can be grasped quickly. This compares actual collections and current expenditure at a particular date with the level forecast for that stage of the year, and then shows a simple plus or minus variation on the forecast. 8.1.5 Budgetary Correction Regular use of profiles and variance reports, as just described, should indicate items of revenue where serious shortfalls are likely, or items of expenditure where overspending is likely if current trends are sustained. When such trends become apparent, immediate investigation is needed to determine the cause and see what corrective action, if any, is needed. Investigation may throw up problems which can be readily corrected, for example delays in invoicing for revenue due, or excessive ordering of supplies. In such cases it should be sufficient to warn the department responsible of the need for corrective action. The important element is being aware of the trend early enough in the budget year for such action to be effective. On the other hand investigation of revenue shortfalls or overspending may reveal causes which are unavoidable. Market fee collection may be depressed because of a poor harvest. Expenditure on drugs may be running high because of an epidemic. Two questions then arise. First, in terms of overspending, is the need for immediate budgetary adjustment to authorize extra provision, preferably by virement, otherwise by supplementary estimate. A second, and more important question, is whether the trends in actual revenue collection and expenditure suggest that the budget is fundamentally off course, and that overall expenditure is likely to exceed revenue. If such a deficit does seem likely, the next question is whether it can be cover by reserve balances, or is going to land the authority in serious debt. If a serious deficit is in prospect which cannot be covered by reserves, major corrective action is clearly necessary. The important thing is to be aware of the danger early enough to take action. Various steps can be considered, e.g. postponing the start of capital development projects (though this may only help if they are financed by general revenue, not from some specific grant or loan); freezing staff vacancies;

subjecting all purchases of supplies and equipment to central scrutiny and approving only urgent needs; restrictions in service provisions, e.g. earlier closing hours for libraries or parks.

None of these steps are desirable, but they may be essential if the financial crisis is to be averted. It is advisable to put as much of the burden as possible on the operating departments to make the economies, since they are likely to know best where potential savings lie. 8.2 Management of Cash-Flow The above mentioned methods are also important for cash-flow management. Even when an authority budgets to balance revenue and expenditure over the course of a period (usually one year), there will always be temporary imbalances between cash coming in and going out. At some periods cash receipts from taxes sales, recoveries of debts etc. may be well above the average rate and cash surpluses will build up temporarily. At other times revenue may lag behind expenditure, or abnormally large lump sum payments may be due (e.g. on contracts or external debt service) and temporary cash shortages may occur. It is important that the treasurer for sees and plans his cash flow. On the one hand it is undesirable to accumulate large cash surpluses in safes, current accounts etc. because they may pose a security risk and could earn substantial interest if invested. On the other hand the authority must have arrangements to cover its obligations during temporary periods of cash shortage by bank overdrafts, bond issues or other forms or borrowing. It is advisable therefore to draw up a month by month (or even week by week) estimate of cash receipts and outgoings which will be based upon the budget, but will allow for seasonal variations in the flow of income and expenditure gained both from past experience and from knowledge of when exceptional lump sum transactions are due. This will enable the treasurer to see how long any bank surpluses are likely to last and invest accordingly and alternatively to arrange borrowing in advance to cover temporary cash deficits. It is sometimes thought to be wrong in principle to have a bank overdraft. This is not the case. If a cash deficit is confined to a few months of the year, it will pay to invest working balances and incur a temporary overdraft, if the interest received over the full course of the year exceeds the interest paid on an overdraft for a shorter period. The treasurer is also concerned with the day to day cash balances held within the authority. Since they are at risk and are also unproductive in terms of interest, regulations are required to ensure that sums kept in hand are at a minimum and any excess promptly banked. The authority should stipulate the maximum sum which can be held at any cash office at a time. Cash balances represent a temptation and there must be strict precautions against fraud. These normally include the following:

daily reconciliation between receipts and payments and the balance in hand, checked by an official other than the cashier (and not the same official each day); a reconciliation at least monthly, between receipts, payments, cash balances and the bank statements; issue of receipts for all cash income, with counterfoils in case sums which are not at a fixed level; frequent surprise checks on cash books and balances; opening of mail by two people and immediate recording of any cash remitted.

8.3 Miscellaneous Financial Control This part deals with a number of aspects of financial control which are not directly related to the execution of the budget. 8.3.1 Staff Establishments and Emoluments

Most local and regional authorities operate tight control over staff establishments and emoluments. The budgets prescribe not only the overall amounts to be spent upon salaries and wages but also the numbers of staff to be employed in each post and grade and the salary scales, allowances etc., appropriate to each. Decisions on the appointment, grading promotion etc. of individuals to posts are also hedged by divisions of responsibility and other safeguards. Controls in this area are particularly tight because it is especially vulnerable to corruption and extravagance. The treasurer is responsible for ensuring that departments keep to the budgetary levels and scales and the establishments procedures governing staff employment and emoluments. It is also necessary to keep a careful watch upon the utilisation of manpower. Salaries and wages are normally the largest element of expenditure and therefore most liable to waste of public funds. 8.3.2 Purchase of Goods and Services Control is necessary over the purchase of goods and services to ensure that the subnational authority secures the best available balance of quality and cheapness, reaps the potential savings of bulk purchase and curbs corruption on the part of its members and staff. For these reasons, it is normal to "tender" for most large scale purchase of goods or services i.e. to invite competitive bids from a range of potential supplies. Such procedures usually govern: supplies of particular materials which are regularly acquired over the course of a year (e.g. cement, stationary, petrol);

undertaking any works (e.g. building or road construction) or providing any service (e.g. architectural design or vehicle repairs) costing more than a fixed amount of money; providing insurance cover; selling assets (e.g. land, buildings etc) other than those regularly available to the public at fixed prices or charges.

It is normal to publish a notice inviting tenders. This must have wide publicity and provide adequate notice. (There is danger of collusions so that only one supplier favoured by the officials gets the opportunity to submit a tender). It is normal to require submission of bills of quantity in respect of large construction contracts, and samples in the case of supplies of goods. Tenders should also specify the period within which the goods can be delivered or the works and services completed. Authorities usually establish special bodies e.g. Boards or Committees, to award contracts to tenders. The customary rule is that the contracts for purchases should be given to the applicant quoting the lowest price and for sales to the applicant offering the highest price. There may be reasons for deviating from this norm particularly where there are doubts over quality or reliability. It is important that the reasons for not accepting the lowest or highest tender, as the case may be, are clearly recorded. Awards of tenders must be widely publicised within the authority with details of prices, specifications and maximum periods for delivery. This is necessary to ensure that all departments know where they should obtain goods and services and on what conditions.

Contracts may have to provide for variation in the amount of work to be done or goods to be supplied, the length of time involved or the price to be charged. This applies particularly to large scale construction contracts, or the supply of goods subject, for example, to tax changes. Any provision for variations in price should, however, be geared to review at fixed intervals and related to official indices. Contracts frequently include penalty clauses covering delays in performance. Construction contracts usually provide that a percentage of the contract sum be withheld for a period (e.g. 6 months to a year) as a guarantee for repair of any defects in the work which may appear after completion. Full tender procedure may well be inappropriate for small, irregular purchases of goods or services. There may still be a requirement that the officials concerned should get more than one quotation (usually three) before deciding on the purchase. 8.3.3 Maintenance of Stores and Equipment The treasurer has a responsibility to ensure that stores and equipment are properly looked after. This involves adequate storage, maintenance of proper records showing their use and location, and proper security from theft and damage. Inspection is required regularly to check that stores and equipment can be located and are in good condition. A distinction has to be made in this respect between durable and disposable goods, though some audit checks are needed on the rate of consumption of disposable commodities. Advance, bulk purchasing of goods held regularly and in quantity has clear advantages. It reduces delay in obtaining supplies and should earn some reduction in unit cost. There are disadvantages however. Cash is converted into stores which could otherwise be earning or defraying interest. Stores held over along period can deteriorate or become obsolescent, and the storekeepers, watchmen etc., and space all cost money. A balance has to be struck between these considerations and quantities ordered should relate to a high expectation of use over a reasonably short period of time. Procedures are necessary for inspecting equipment and recommending disposal of that which has reached the end of its useful or economical life. The treasurer is responsible for deciding upon and arranging insurance cover of the authority's assets and liabilities. This may include: insurance of property (e.g. buildings, plant, vehicles, cash etc) against loss, theft, damage, or destruction, and insurance against legal liability to employees or third parties arising from accident or negligence.

Insurance cover may be mandatory or discretionary. In some cases the law requires the authority to obtain insurance cover e.g. for its vehicles. In other cases the treasurer will have to weight up the costs of the cover against the risks. 8.3.5 Reserve Funds and Investments The purpose of reserve funds have been set out in Chapter 7.2. Reserve funds may be held: as working balances

as longer term "insurance" against unexpected expenditure or revenue shortfalls to provide for renewal of plant and equipment to provide retirement benefits for employees to accumulate funds for future capital investment

The level of these funds clearly needs regular review, particularly when drawing up annual budgets. Care must be taken in interpreting the calculation of these funds shown in the balance sheets. There is a danger that the value of the reserves is inflated in the balance sheet because the "assets" making up the reserves include either debt to the organisation which are unlikely to be recovered, or buildings, equipment etc, which are not really saleable. Reserves are normally backed by a money surplus which is invested. Public sector investments are normally in bank deposit accounts - suitable for day to day, very short term fluctuations in balances

treasury bonds - suitable for balances available for short periods (normally 3-6 months) longer term gilt-edged securities - suitable for long term reserves, superannuation funds, renewals funds etc.

Generally speaking the longer the term of the investment, the higher the rate of interest. It is therefore necessary to select the type of investment with strict attention to the length of time before it is likely to be cashed in. 8.4 Fiscal Capacity Framework From the subnational government perspective, many things beyond the control of local officials have major effects on local and regional finances. These include policies by higher levels of government, the overall state of economy, demographic and climatic changes which do change demands for local services or changes in intergovernmental fiscal relationships. Within the context of the chapter we outlined previously, subnational governments must have sound fiscal policies from the perspective of enhanced financial condition . The programs that are developed and implemented rely on resources that must be absorbed by the government and managed well in order to be effective. In addition, continuous monitoring, improvement, and adjustment of policies will help keep of their fiscal management capacity. Local governments must step back periodically and evaluate their fiscal condition and apply what they learn from these "checkups" to future activities such as raising taxes, cutting spending, attracting intergovernmental revenues, or other actions. Figure 8.3 illustrates a Honadle diagram (Honadle, Costa, Cigler, 2004, p. 13) describing what it means conceptually for a local and regional government to have fiscal management capacity, or to manage their fiscal condition better.

Conclusions
Major objectives of budgetary control are: to ensure the authority solvency, legality of revenues collection and money spending, paying for public services through taxes and charges, efficiency of spending money, and honesty in handling of the money. Effective budgetary control requires clear identification of personal responsibility for collecting each revenue source and incurring expenditure under each budget item. In practice there exists three patterns of responsibility delegation - financial manager authorises all expenditure and make all payments (centralised control), the operating departments authorise expenditure within budgetary limits, but the central treasury makes all payments, the operating departments authorise expenditure within budgetary limits and make payments. The recommended approach is to delegate authority to incur expenditure to operating department within a voted budget. Revenue control is based on strict and impartial assessment of the amount of taxes, fees or charges which are expected to be paid by individuals, on appropriate applying of rules for providing remission or exemption in taxes and charges, on continuously monitoring of revenue collection from individual payers (liability) as well as on general checking of the total amount. For efficient expenditure control it is necessary to have regular and up-to date information on the total amount spent. Seeing that local and regional authorities do not usually operate in a full accrual system of accounting, it is necessary to keep an accurate record of the commitments they have undertaken, so that they know the true unspent balance available to them. Before payments are made, they must be checked to ensure that no errors or fraud is involved. Monitoring revenue and expenditure level need to be reviewed regularly to ensure that solvency is maintained. Effective monitoring of budgetary progress need to draw up an expected pattern of revenue collection or spending throughout the year including all irregularities (e.g. market fee income, debt service charges) and its comparison with the amount actually spent at any point in the budget year. This variance analysis shows serious deviations from the budget and shows to financial managers that corrective action is required. If profiles and variations reports indicate shortfalls in revenues or overspending of expenditure, the investigation is needed to determine what action of budgetary correction is necessary. In the case when overall expenditure is likely to exceed revenue and the deficit occurs, it is a question how it can be covered (from reserves, or other corrective action). For example by postponing the start of capital projects, freezing staff vacancies, approving only urgent needs, restrictions in service provisions. Management of cash-flow is important tool for financial manager to estimate and plan the temporary imbalances between cash receipts and outgoings which are based upon the budget and include seasonal variations in the flow of income and expenditure. It is important to avoid undesirable accumulation of large cash surpluses; on the other hand the authority must have arrangements to cover its obligations during temporary periods of cash shortage by bank overdrafts, bond issues or other forms or borrowing. Among the other aspects of financial control there are the following: control over staff establishment and emoluments, control over the purchase of goods and services to ensure the best available balance of quality and cheapness mostly through tenders. The financial manager has also the responsibility for proper maintenance of stores and equipment as adequate storage, and proper security from theft and damage, and for insurance cover of the authority's assets and liabilities. The financial manager is usually in charge of the reserve funds and of the investment the money surplus in financial markets.