6 Budgetary and Financial Management

Teresa Ter-Minassian
Published Date:
September 1997
  • ShareShare
Show Summary Details
Barry Potter

Whether in a federation or in a unitary state, lower-tier governments play an important role in delivering public policies and services. Their provision of services has an economic impact on overall activity, on the allocation of resources, and on the pursuit of equity objectives. Their financing has an impact on the levels and types of taxation, with allocative and distributional consequences. Their borrowing has an influence on monetary conditions and interest rates. The importance of secondary tiers of government is recognized in the IMF’s financial programming exercises that, whenever possible,1 focus on the financial surplus or deficit of the wider general government sector.

Yet this important macroeconomic dimension of local government activities poses many difficulties for policymakers, particularly when determining annual expenditure budgets and borrowing, both for central and local governments. The institutional arrangements for the determination and conduct of public policies and services need to take appropriate account of macroeconomic goals. The main issues include the following:

  • What direction, guidance, or controls—recognizing the range of relationships between the central and local government under different fiscal federalism models—should be given to lower-tier governments, in advance of their budget setting?
  • How are the budgets of local governments to be taken into account by the central government, when drawing up its own budget?
  • What coordination can be put in place to ensure that the respective policies of each tier contribute both to central and local resource allocation and distributional goals?
  • What monitoring requirements are needed to ensure that the central government can take into account developments in the delivery of local government budgets, in deciding on any in-year fiscal action?
  • How can grant systems, in their size, structure, and payment patterns, best accommodate the need for financial planning and cash management at both levels of government?
  • How can the respective plans of both tiers for borrowing best be coordinated to avoid unnecessary borrowing in aggregate or excessive pressures on financial markets at certain times?
  • What role, particularly in a unitary state,2 should the government play in checking not only the aggregate levels of spending by local governments but also the efficiency and effectiveness of that expenditure?

This chapter focuses on the institutional arrangements in intergovernmental relations necessary to meet macroeconomic policy goals. First, some macroeconomic objectives are suggested for the overall conduct of local government budgets and their coordination with central government objectives. Second, some specific approaches are recommended for budget preparation, distinguishing between a federal and unitary (that is, a more agency-type) relationship between local and central government. Third, the need for monitoring of budget implementation and the modalities of corrective fiscal action within the year are considered. Fourth, payment of grants and issues of coordination of cash management and borrowing plans are discussed. Finally, the increasingly important role of audit, inspection, and evaluation in securing objectives on resource allocation and efficient delivery of public services in local governments is discussed.

Public Expenditure Management Objectives for Local Governments

The principal macroeconomic objectives in budget setting and implementation (from a general government perspective) may be described as follows:

  • To ensure that aggregate spending and borrowing are consistent with government macroeconomic objectives for inflation, economic growth, and fiscal and monetary stabilization.
  • To enable the delivery of public services in an effective way, which meets the allocative and distributional objectives underlying the spending plans.
  • To bring about the delivery of services consistent with these plans (and the related budget authorizations and relevant financial regulations), while retaining the ability to modify the plans in the light of evolving fiscal circumstances and in particular revenues available, within the year.
  • To ensure that the above objectives are mutually achieved while pursuing effective cash management and minimizing borrowing costs.
  • To put in place audit, inspection, and evaluation procedures in order to promote an efficient and effective delivery of public services.

Budget Preparation: Aggregate Expenditure Control

Budget preparation procedures must take into account several macroeconomic dimensions, including (1) the impact of local government budgets on the economy; (2) the interaction between central and local government budgets—for example where some grant expenditure is effectively open-ended; and (3) the need and scope for some coordination between central and local government policies. This last point is particularly important if certain policy objectives are to be secured—for example, where expenditure assignments are such that both tiers of government have policies aimed at the same client group.

Irrespective of the relationship between central and local governments, a precondition for the respective harmonization of budget proposals in this area is adequate information on the public policies, spending plans, revenue projections, and borrowing requirements of each tier of government. For less-developed economies, securing adequate and timely information can be a significant problem. Indeed, lags and inaccuracies, different expenditure (and revenue) classifications, policy misunderstandings (for example, who will pay for a particular service), the absence of a global budget—such that there are important offbudget transactions—can be serious drawbacks to the coordination of budget preparation.

To some extent, the significance of the information problem depends on the local government model in operation. Under a federal arrangement, the central government’s main aim is usually to achieve overall macroeconomic control principally through borrowing limits and controls on subnational governments (see Chapter 7). Even then, however, particularly where expenditure assignments have any element of overlap and where central government policies affect local budgets, coordination of policies is highly desirable. Information within the year is essential for monitoring macroeconomic development. Under the unitary model, however, the need for information and coordination, though different as described below, may be just as important. Information requirements vary between the autonomy model—typical of a federal structure as in Germany, the United States, and Canada, for example—and the centralized agency approach, as applied in recent years in the United Kingdom, for example, and described below.

In principle, under either a federal or unitary model, the appropriate fiscal policy from a macroeconomic perspective—that is designed to meet the objectives set out earlier—would have to meet a formidable set of requirements:

  • Adequate coordination of spending, revenue raising, and borrowing plans in aggregate to ensure macroeconomic stabilization.
  • Agreement on the allocation of resources among spending programs so that, to the extent possible, they reinforce rather than contradict the allocative and redistributional policy goals of each tier.
  • Mutually agreed and accurate forecasts of expenditures, revenues, the timing of cash inflows and outflows, and hence borrowing needs.
  • Coordination of the timing and amounts of borrowing to avoid undue pressures on the financial markets.

In reality, such a comprehensive approach is difficult to achieve in full. Yet, in a number of countries, the budget preparation procedures incorporate several features designed to meet all or some of these objectives. The approach differs between the federal, more autonomous, models and the unitary, more direct, control arrangements.

The practical experience of federal systems illustrates the importance of having medium-term financial plans that cover all levels of government in the formulation of fiscal policy and plans. For example, in Australia, before the Premiers’ Council decides on the overall level of commonwealth (central government) transfers to the states, the Treasury prepares a memorandum for the Cabinet that outlines fiscal developments and prospects in the state or local sectors. These are prepared in the context of forward estimates of outlays for general purpose assistance and the commonwealth’s broader objectives for public sector spending and borrowing. This is usually considered alongside updated forward estimates and projections of the economic outlook for Australia as a whole.

In Germany the budget is planned within the framework of a rolling five-year financial plan (covering the current year, the budget year, and three subsequent years). These plans are prepared both by the federal government and by each of the Länder. The Financial Planning Council then ensures that these plans are coordinated and that they are consistent with the aggregate fiscal target. This Council is chaired by the Federal Minister of Finance, with finance ministers from each of the Länder and representatives of municipalities as members. The work of this Council is complemented by the Committee on Public Borrowing, with similar representation, which coordinates public borrowing and ensures that demands on the market are not excessive or destabilizing. The division of labor is clear: the Council determines the overall level of borrowing consistent with the aggregate fiscal targets, whereas the Committee decides on the phasing of public borrowing. The importance of borrowing controls under this federal system is clear.

The result is a Federal Financial Plan comprising a set of compatible financial plans for the lower levels of government. Based on this financial plan, budget preparation consists of the allocation of funds to line items in the budget of each budget organization in cash terms. Supplementary allocations are rare, and the budget does not include an explicit reserve. This requires a strict cash limit approach to fiscal planning.

With the more centralized government systems under unitary models, the legal powers of the central over local government can make the operation of coordination easier to achieve. Several countries have set up formal procedures and institutions for discussions or negotiations among different levels of government. The impact of these arrangements differs markedly among countries. In the United Kingdom, a Consultative Council on Local Government Finance, established in 1975, acts mainly as a formal notification mechanism for decisions formulated at central government level. The most important negotiations on aggregate local government expenditure and central government grants take place between the Treasury and the sponsoring service ministries for local expenditures. In reality, the Consultative Council has become increasingly redundant with the greater centralization embodied in the post-1990 arrangements for capping local taxes in the United Kingdom (see Chapter 14). While the formal budget plans presented to the U.K. Parliament typically underestimate the level of local government spending—to suggest that the level of local taxes could and should be lower—it is widely understood that the reserve allows for the “overspending” between the notional budgets assumed for presentational reasons and those that the authorities will set (and spend).

Some countries such as Denmark and Sweden have more transparent and effective arrangements, conducting regular negotiations between the ministries of finance and the central organizations representing the regions and the communes. Agreements are reached on targets for the local and regional expenditures. A similar institutional apparatus has been developed in Belgium.

Some countries have also experimented with procedures for specialized negotiations and agreements between central government and individual local units or clusters of such units. Since the tax bases exploited by different levels of government are often the same or related, there are usually strict rules to circumscribe the volume and methods of raising local revenues that must be negotiated with subnational governments. In the early 1970s, the French government began to make planning contracts with large cities (communautés urbaines), and in 1974–75 this practice was extended to contracts with middle-sized towns and rural areas (contrats de pays). Under such agreements, the central government committed itself to support a multiyear program of local operations and to pay a fixed overall grant.

Against that background, it may be possible to suggest certain requirements for the necessary harmonization of budget plans from a macroeconomic perspective.

  • First, the central and local governments need to work on common assumptions for the budget year about the key macroeconomic variables—including prices (and particularly important relative price changes), wages, and the exchange rate. The central government is best placed to formulate these, within the context of its own macroeconomic forecasting function. The central forecasts (and any projected regional variations on prices or wages) need to be circulated to lower-tier authorities in good time for budget preparation.
  • Second, a common revenue and expenditure classification is necessary. Most countries have adopted the IMF’s classification for Government Finance Statistics for central government programs and it is highly desirable that this classification be extended to state and local government expenditures. A common financial year (often set by law) is essential.
  • Third, budget forecasts covering both central and local government revenues, expenditures, and borrowing need to be made well in advance of the budget year if the benefits of coordination from a macroeconomic perspective are to be secured. In addition to guidelines on macroeconomic indicators, the central government should ensure that subnational authorities are given, in good time, an indication of affordability constraints—whether directly for the unitary model with information on expenditures and revenues or indirectly with guidance on borrowing.
  • Fourth, coordination of the budget plans in aggregate is essential for macroeconomic purposes and should ideally be set within a medium-term rolling framework. The form and extent, however, will vary according to the vertical structure of government and in many instances the balance of political and legal powers, as discussed below.
    • —For the federal systems, the German and Australian approaches are good examples of the types of coordination that are necessary. Although institutional formats will inevitably vary, and the focus will typically be on borrowing control, the advantages of securing expenditure and revenue data also for macroeconomic monitoring, with the consequent creation of a capacity for fiscal action within the year, are significant.
    • —For the unitary model, usually the legal or administrative authority can be relied upon to limit directly or indirectly the expenditure and revenue levels of the lower-tier authorities. But timing is important, if each tier’s plans are to take into account the spending plans of the others. In addition, care needs to be taken to avoid the emergence of off-budget transactions or other means of circumventing central controls. Moreover, as the U.K. example suggests, this coordination may not always be achieved with the degree of transparency that is desirable.
    • —By contrast, poor coordination is likely to have important economic costs. These include (1) unnecessarily high interest rates (created by excessive or poorly timed borrowing); (2) expensive off-budget transactions (when local government seeks to circumvent central government controls); (3) disruption of expenditure plans (where emergency cutbacks become necessary to maintain overall fiscal stability); and (4) poor resource allocation (because of duplication of effort or gaps in provision between different tiers of government). At worst, fiscal stabilization itself may be threatened.
  • Fifth, the coordinated budget projections should be accurate and reliable (in their timing) not least to achieve efficient financial planning. Central and local governments must still be able to exercise flexibility within the year. Expenditures and revenue projections should be kept up to date and, ideally, policies should thus remain stable during the year insofar as possible. The interactions between central and local government expenditures—for example with open-ended specific grants or central government programs directly managed by local governments—should be taken into account in preparing consolidated budget projections.
  • Sixth, given that from a macroeconomic perspective even the best-laid plans are subject to exogenous shocks, there is a case for including contingency reserves within both the local and central government budgets. For the local government budget, such reserves should be used to cope with the normal variations, for example, in demand-led programs. For the central government, the contingency reserve must be large enough to accommodate or offset variations from plans in the local government sector, from a macroeconomic perspective. In some respects, this poses problems. There is a moral hazard problem if some share of the contingency reserve becomes perceived as “belonging” to the local or subnational authorities; it is no longer a genuine contingency reserve but a reflection of nontransparent budget preparation practices.

Budget Preparation: Resource Allocation Among Programs

The other major aspect of budget preparation where some coordination is desirable from a macroeconomic perspective lies in the pursuit of allocative and distributional goals under both federal and unitary models. Central government can be involved in the allocative decisions of lower-tier authorities: (1) to prescribe uniform or minimum standards; (2) to internalize the externalities associated with small areas providing services, with spillover benefits to others; (3) to require local governments to undertake policies that are of national rather than local importance; and (4) to administer central policies where regional organization has efficiency benefits. The precise mechanisms for this central dimension in the delivery of second-tier authority policies will vary, encompassing specific or matching grants, regulation, joint initiatives, and so on.

The essential requirements for efficient resource allocation among expenditure programs were discussed in Chapter 2. In principle, unambiguous, stable expenditure assignments that avoid duplication are to be desired. In practice, however, firm lines of demarcation can be difficult to draw. It is important to distinguish among three aspects: policy, implementation, and financing.

  • Policy responsibilities can often be fairly clearly defined. Even here, however, detailed policy issues may be a source of controversy. Stability is generally advantageous but can also be difficult to achieve, not least as political fashion changes. In the United States the trend has been to give new mandates to the states. In the United Kingdom, on the other hand, the recent history has been of service responsibilities being taken from local government back to the center.
  • Implementation can also be a source of difficulty because the line between a policy and an implementation issue may well be hard to draw. Dual implementation, where both tiers are offering a different but related welfare service to the same client groups, may mean that problems of coordination are inevitable. Moreover, as individual welfare cases progress, the service responsibility for the individual may pass from central to local government or vice versa.
  • Finally, financing, particularly the award of specific matching or conditional grants, inevitably requires close coordination.

Coordination in service delivery (implementation) in such cases is best achieved at the local level, that is, at the level of the individual lower-tier authority. The need for policy and financing coordination, however, has to be addressed at the budget preparation stage—under any model of local government. Such coordination is often best achieved service by service. Many countries—Australia, Germany, and the United Kingdom, for example—have consultative groups of different kinds for individual services. Their goals include (1) interpretation and elucidation of expenditure assignments; (2) the derivation of detailed service policies and minimum or target service standards that help reinforce their respective policy goals; (3) the drawing up of practical guidance on service implementation; (4) the phasing of expenditures over years (mainly for capital projects) or within the year; (5) discussion of the precise rules for the award and triggering of specific or matching grants; and (6) technical issues, for example, on the recurrent costs of new projects aided by specific grants or of proposed new expenditure assignments.

Such consultative groups can play an important role in helping each tier meet its respective service policy objectives. Yet there are limits to how far such consultation may lead to total accord on policy, implementation, and financing issues. It is inevitable in democracies with more than one tier of government that the political color of central and local government will differ from time to time. The ability of one to enforce or accommodate the policies of the other will depend on the perceived political balance and on the actual (legal) powers of each tier. Views on both resource allocation—say schools versus hospitals—and on redistribution objectives may differ. Not all will accept the primacy of the central government on matters of income distribution that are normally associated with the centralized income taxation and social benefits.

It is also possible that, in the case of specific or matching grants, there may be a less than transparent mechanism for handling grant applications. Indeed, specific grants in themselves raise awkward issues for budget and financial management. First, so long as there is any conditionally in the terms of the grant, there will be uncertainty for budget preparation about the timing and the amounts of receipt. Second, compared with block grants, specific grants can be inefficient in theory, influencing resource allocation decisions of local governments to reflect perceived national priorities rather than local. The implicit resource diversion can result in a net welfare loss. Third, and more practically, there are the administrative costs involved in setting up monitoring and payments of grants. Fourth, to trigger grant receipt on a timely basis requires investment by both central and local government in efficient information systems. But, at worst, the triggering becomes automatic (partly because both sides have incentives to minimize the information checks), leading to inefficient use or misuse of specific grants.

Ideally, therefore, the main objective on both policy and financing, while trying to achieve coordination to the extent possible, must be to ensure that the policy intentions and financing plans of each tier, as reflected in their respective budgets, are accurate, achievable, and—in aggregate—consistent with macroeconomic stability. This also means that the policy pattern should be predictable; that policy changes should be generally avoided within the year, unless they have been taken into account in the budget-setting process; that the intentions on size and timing of specific grants should be as transparent as possible; and that—given some policies are open-ended, for example, for natural disaster expenses—there should be some contingency reserve.

For developing countries, however, this list is likely to represent a rather demanding requirement. The initial focus should be on efficient and accurate budget preparation by all tiers, and putting in place effective budget constraints on the lower-tier authorities, which prevent excessive fiscal deficits, the buildup of arrears, or other off-budget finance. As discussed in the next section, this also means that careful attention must be paid to the implementation of the budget during the year, if the scope for remedial action within the year is to be sustained.

Budget Reporting and Actions Within the Year

In many respects, the requirements for good budget implementation in the local government sector are the same as those for the central government. Starting from clear and well-prepared budget plans, the objective must be to implement the plans in the most efficient manner, with the maximum economy on day-to-day and longer-term borrowing and safeguards to ensure that budget appropriations are held to and not exceeded. This requires (1) clear budget authorization processes internally; (2) efficient payment systems; (3) a suitable accounting framework; and (4) an information system that can cover all the stages of the budget process from budget preparation through commitment, verification, payment authorization, the payment itself, and the accounting for the transaction. Also important are regular, timely, and accurate reporting, built on good information systems, and timeliness and efficiency in both financial planning and the associated borrowing program.

So far as the macroeconomic dimension is concerned, apart from the general need for local authorities to remain within their budgets and manage their budget implementation process in an efficient way, the two main concerns must be the reporting framework and cash planning. The latter, as well as its associated borrowing implications, is especially important where the main financial control over second-tier authorities is exercised through limits on borrowing.

The principal requirements from the reporting system are for timely, consistent, and accurate reporting of the budget implementation. As noted earlier, the revenue and expenditure classification—the latter both in terms of the functional and economic codes—should be consistent with the central government classification so as to allow easy consolidation of the accounts of the whole general government. In many countries, however, it has been possible to go further and ensure that the entire accounting framework is either common between central and local governments (such as in Brazil) or that local government accounting requirements are set by central government law or regulation, as in Germany and the United Kingdom. Many developed countries require their local government authorities to report monthly to the central government on expenditures, revenues, and on borrowing—discussed below. However, such information may not be illuminating, if it cannot be compared with expected patterns to date. Thus, many countries maintain a profile of planned spending—drawn from the aggregate of local governments’ own plans and actual patterns of spending in the past. This enables the central government to spot individual localities in trouble, while also allowing trends across subnational authorities to be analyzed.

Whether it is possible and desirable to go further, for example, to monitor expenditure commitments and even verification, will depend on the degree of sophistication of the economy and on the vertical structure. In federal countries, such micromanagement of the affairs of state governments might well be thought unacceptable and unnecessary. In more centralized models, it might appear a logical component of financial planning—particularly in developing countries where there may be concerns about arrears or off-budget financing. For the more advanced economies such monitoring may not be necessary; for the less advanced, including many of the countries of the former Soviet Union at this time, it may well be essential, if macroeconomic problems associated with arrears and off-budget financing are to be avoided. Indeed, in such countries a basic policy dilemma has to be faced: the economic welfare gains available in principle from decentralization can be overwhelmed if there is a loss of control over the macroeconomic position. Despite the political attractions of deepening democracy, the rate of progress may have to be attuned to the rate of institutional improvements in information reporting and budget control.

Although the hope will be that the budget will be delivered as planned, in practice, adverse exogenous economic developments may damage revenue collection, while bad weather or unexpected natural disasters can give rise to additional expenditures in subnational authorities (as well as in the central government). Changes in central government policies, perhaps in response to an exogenous shock, may also have an impact on the delivery of local budgets. In general, it follows from the prime responsibility of the central government to undertake macroeconomic policy coordination that it should have the responsibility for deciding on the need for any countercyclical or other adjustment within the year. The action may be confined only to the central government sector, where it can often be implemented more quickly, sometimes with greater certainty of economic impact. However, the central government action may be partly frustrated by the subnational government reactions. For example, cutting central grants may lead local governments to continue financing a favored project out of reserves or to accumulate arrears. In other cases, such as wage reductions or hiring freezes, it may be necessary or desirable for the second-tier authorities to shoulder part of the fiscal adjustment within the year. Without an adequate reporting system, the need for and targeting of such remedial action can represent a serious obstacle to the efficient conduct of macroeconomic policies.

It is also important, however, that actions within the year to address emerging fiscal problems be appropriate in nature. All attempts should be eschewed that pass the problems of the center to other tiers of government by transferring expenditure assignments, without prior agreement on the policies to be followed and how the resources are to be found. Such “unfunded mandates” should be avoided either within the year or in the annual budget preparation exercise. Also to be avoided are sudden withdrawals or reductions in the amount of general or specific grants of the type attempted in Russia in 1995. The result of these actions will be some delayed or postponed expenditures. But they are also likely to lead to payment arrears, to higher borrowing often from local banks or from suppliers, to poor resource allocation as other plans are disrupted to take on board the new responsibility, and to the growth of extrabudgetary accounts or other creative accounting devices. All such devices evade expenditure control but usually only temporarily: the consequent misallocation of resources can be expensive to resolve later.

Where local governments are required by the central government to take action within the year, it is often best applied on the same basis as for central government—“burden-sharing.” That may include hiring freezes, postponement or delay in commencing new capital projects, or reductions in nonwage current expenditures. Once there is agreement on such measures, then it may be sensible to reduce general purpose transfers as the mechanism through which local government will share in the overall fiscal adjustment. In the central government more targeted measures within the year, such as cuts in specific grants, are best avoided except in the context of needing to absorb some unexpected addition that is service specific. An example would be a mandatory but unanticipated increase in teachers’ salaries following a court decision or commission of enquiry.

Cash Planning and Borrowing

In the local government sector, just as for central government, efficient financial or cash management and economical borrowing require careful financial planning. Local authorities must be able to project the flows of revenues, both tax and nontax (including central government grants) and cash expenditures and hence the financial deficits and borrowing requirements. At first sight the requirements are clear: for the revenues it is the timing of receipts, and for expenditures it is the timing of cash payments that will determine the borrowing need—setting aside any refinancing. But the interaction between the central and local government sectors poses a number of difficulties:

  • Where there is tax sharing, the question of local versus central collection and the speed with which revenues are passed on to, or from, the central government are important.
  • The forms of grants—general, specific, and matching—and the need for local governments to demonstrate compliance with the rules determine the timing of the grant receipts in local governments.
  • In general, flows of tax revenues and grants from one tier to the other involve improving the cash position of one tier while worsening that of the other. This has led to the development of central treasuries in some countries (such as Italy) with a global consolidated account covering large parts of the general government sector.
  • When lower-tier governments borrow from a central government source, there is a need to carefully plan access to markets, and perhaps the need to coordinate with other borrowers has to be factored in. This complicates the interaction between identified cash shortages and borrowing. Many local governments do not have the automatic or quasi-automatic access to financial markets that central governments typically enjoy. However, unfettered access creates other problems as discussed in Chapter 7.

Efficient arrangements for the administration of local tax collection and tax sharing are described in Chapter 5. For grants, the main issues are type and modalities of payment as discussed below.

  • In the case of general grants, the normal principle is of payment of grants scheduled according to a profile of estimated expenditures. While this can mean a simple payment of general grants in 12 equal monthly installments, in most countries such grants take into account seasonal patterns of past expenditures, so that profiles over the 12 payments may not lead to equal installments.
  • For current expenditures on specific grants, the arrangements may be similar to general grants—for example, when the local government is carrying out a current service essentially on an agency basis. This might be the case for a welfare or education service that is administered by the local government but under central government direction. Others—particularly those that are one-off—may be more difficult. For example, specific grants may be paid to local governments after severe weather damage. Here expenditures will clearly be incurred before the grant is received, and that needs to be factored into cash planning.
  • For capital expenditure, arrangements for specific grants again vary. Where receipt of the grant is unconditional, the payments can be linked to the phasing of expenditures, again in line with the seasonal pattern of payments. In some instances, it may be possible to link the payment of the grant to the securing of borrowing for the project. It is also possible to pay such grants only on conclusion of the project. This enhances the financial position of the central government, but only at the cost of increasing the borrowing need of the local authorities. Greater complexities still may arise where the central government is acting as an intermediary for grants received from foreign donors.
  • For matching grants the difficulty is that the requirement to satisfy some conditions of central government, whether for a current or capital grant, injects some uncertainty into the timing of the receipt of the grant. As noted earlier, it is thus in the interests of the local government to put in place efficient reporting systems, which will enable it to trigger a request for the grant payment as soon as warranted.

There are several procedures that would seem to indicate good practice in this area. First, the general principle for all types of grants is that, if neither central or local government is to be given a relative unwarranted financial benefit, grants should be paid as close to the point of financial need as possible. That is, the payments should flow to the local government, just before the local tier needs to start making payment to third parties. For general grants that means phasing according to past patterns of expenditure; for others it means as close a match to the payment as can be achieved. Second, if cash planning is to be efficient, the pattern of such grants across the year should be as predictable as possible. This requires that the rules for the payment of grants should be transparent and unambiguous. This is a further reason why openended grants or those with complicated criteria or conditionality should be avoided. Many countries have found this transparency is best achieved through consultative forums. Third, even with transparent rules, however, there is some unpredictability about timing, and sometimes about amount, not least in the case of matching grants. This needs to be taken into account when making cash plans and establishing borrowing requirements.

As noted, the extensive use of both tax sharing and grants in intergovernmental fiscal arrangements means there is an inevitable interrelationship between the cash positions of the respective tiers. All payments of grants initially worsen the financial position of the center, while improving that of the local tier. In some countries the solution has been some form of consolidated bank account that covers both the central and local government sectors—either wholly or in part, for example by including for the local sector only the grant element. This approach has been followed in Italy. The theoretical advantages of consolidation are clear: the general government sector’s borrowing requirements are minimized, and to the extent that the center can borrow more cheaply than local government, there is a financial gain available. In practice, however, there has been a risk of the center putting its financial needs before those of the local government. In Italy, in the past, this has led to local governments being forced to borrow from the commercial banks when resources available to them from the central treasury account were insufficient.

Particularly under federal systems, it is desirable for states to borrow in their own name: the terms on which they can borrow signals their creditworthiness to financial markets. By contrast, global borrowing on behalf of states or local governments (as in India) in principle allows some individual local governments to borrow at a rate implicitly subsidized by the more creditworthy. Also, many federal systems would resist, on political as well as legal grounds, any such consolidated approach. Geographical reasons may rule out the approach in other instances.

The alternative is for lower-tier authorities to manage their own finances. The requirements are much the same as for central government:

  • Preparing cash expenditure and receipts (excluding borrowing) forecasts by month over the year.
  • If the resulting pattern of financing needs is considered undesirable, seeing what can be achieved by rephasing expenditures or accelerating revenues (including grant entitlements).
  • Determining a month-by-month borrowing plan, taking account of any refinancing needs.
  • Monitoring actual borrowing requirements against the plan.
  • Implementing the plan, subject to variation in the light of outturns.

However efficiently this process is conducted by the individual local authority, it means that in the absence of a consolidated approach there can be several local authorities approaching the market at the same time. The central government and even state enterprises may also need to borrow then. In large part, the solution to avoiding the risk of undue financial pressures on the capital market lies in coordination, and perhaps control of access to borrowing, by the central government. The gain from a coordinated approach includes reducing the risk of paying a premium on borrowing because of congestion in financial markets at certain times.

Borrowing constraints thus play an important role in the macroeconomic management of local government expenditure. Indeed, for federal-type systems they represent the main mechanism whereby central control is exercised over states. The main issues related to borrowing policies are discussed in Chapter 7. Here, the main issues relate to the financial management of borrowing—the information requirements covering both stocks and flows, the monitoring systems, and the access to different borrowing “windows,” including public and private, domestic and external.

Whatever the borrowing policies (whether market, rules-based, or administrative controls), there are important requirements from a financial management point of view. Many countries monitor the flows of borrowing of individual authorities. On the one hand, this is important for microeconomic purposes as in India—checking that individual authorities are not incurring a level of debt that threatens their solvency. On the other hand, it enables a check on the aggregate position. Ideally, such information should come from monthly reporting of revenues and expenditures by each local government. Alternatively, the central government can rely on “below the line” information from the banking sector, though authorization for the release of such information by the banking sector may need to be established by law. Nonetheless, in less-developed economies this may be the simplest way of securing the necessary information flows.

But even the more developed economies have found that information on debt flows may be insufficient. First, creative accounting and off-budget financing may lead to more expenditure than revealed by the borrowing data. Second, particularly in federal systems, lower-tier governments can be major holders of assets—including financial assets. Many of the problems in recent years have been on the asset side of the accounts—for example, Orange County in the United States and the Western Isles authority in the United Kingdom, where unwise investments eventually resulted in large additional borrowing needs.

One solution is to regard these matters as the inevitable result of market imperfection, leaving the local government to resolve the problem. But this in turn leads to the issue of whether the central government is to be regarded as the guarantor of all subnational government loans. In the past financial markets have often regarded them as such. The moral hazard danger for the central government is that such an approach may ultimately mean it has to “bail out” the authority concerned, as, for instance, in Brazil. An alternative would be to require better information flows on the stocks of debt (the debt register) and of assets held. This may well be politically more acceptable in unitary states, but even for federal countries, it could be required by law with the assessment of financial soundness left to the auditor.

Also, the sources of borrowing have important financial management and macroeconomic implications. Where there is a special “window” for local government borrowing, as in Italy and the United Kingdom, for example, it is easier for the government to coordinate the total borrowing effort. Borrowing is also usually cheaper in these circumstances. Where there is market borrowing, terms and types of borrowing will reflect both the assessment of individual authority’s creditworthiness and the degree to which central government is seen as the ultimate guarantor. This last dimension means that the wholly unfettered access to the capital market is unlikely to be prudent—as past country experiences might suggest—and that some coordination of borrowing programs is required.

As regards the forms of borrowing—bonds, loans, and so on—there has been a growing tendency to define the acceptable forms more specifically. One reason for this development was the desire to prevent creative accounting that is outside the normal borrowing controls. A second motive was to tap different sectors of the capital market, essentially for financial or money market development reasons, including the structure of interest rates. A third motive was to discourage long-term borrowing to meet current or short-term expenditure needs.

Even when there is little formal coordination of borrowing, benefits can be gained from the exchange of information. Thus, it is in the interests of all parties that the borrowing plans of the subnational authorities should be known to the center. Combined with the regular flows of information on the monthly finances, whether from an above the line or below the line source, this will enable the center to monitor the financial market implications of borrowing plans closely.3 For the industrial countries, particularly under federal systems, there may be a case for requiring not only information on debt registers but also on financial assets held, even though there may be no powers to challenge asset-holding ventures. Possible responses to the expected emergence of pressures on the financial markets would include temporary intervention in the money markets so as to ease borrowing conditions or to advise against the proposed timing of a major bond issue.

Audit, Inspection, and Evaluation

Under most fiscal federalism arrangements, the central government plays a major role in financing local services. But it also frequently does more: setting minimum service standards, paying specific grants to secure central government service objectives, requiring local governments to act as an agent in implementing central government policies, and coordinating the services it delivers with those of local government. Ensuring that such services, and the standards at which they are being delivered, meet policy objectives and overall macroeconomic requirements leads central government to play an important role in audit, evaluation, and inspection.

There has been increasing recognition in recent years that these functions are important. Audit of the traditional propriety or financial types represents the safeguard through which not only the central government, but both local political representatives and the local electorate, can hold the authority accountable for financial performance. Inspection, typically by employees of the central service departments or ministries, enables the center to be sure that mandated service standards are being achieved and to set objective tests of relative service performance. Evaluation, or value-for-money audit, has been the main growth area. It can be seen as combining audit and inspection, by examining not only what financial inputs went into service provision but also (1) what outputs were achieved, (2) what standards of service were provided, and (3) what impact those services had in meeting the policy objective; in short, the economy, efficiency, and effectiveness of services. Moreover, by conducting this exercise on a comparative basis, best practices can be discerned and publicized and other local governments can be encouraged to adopt the same approach. In principle, such evaluations can make an important contribution to improving overall economic performance.

As in public expenditure management more generally, the greatest progress in this area has perhaps been achieved in Australia, New Zealand, and the United Kingdom Some general conclusions may be drawn from their experience.

  • The external audit body should be independent of the central government and the local authorities. Ideally it should be attached to the parliament—as for example in Australia where the audit body is appointed by each of the state parliaments. If created by central government, it should have an arm’s length relationship from the line ministries and the ministry of finance—as for example with the Audit Commission in the United Kingdom. In short, the audit body needs to be seen as independent of both central and local government.
  • The audit body must have “teeth”—the power, through the parliament or the central government, to have its recommendations responded to. Legal sanction (including disbarring members of the local government executive) can be an important ultimate weapon.
  • Particularly as local authorities make progress in their accounting formats, often being ahead of central government practices in areas such as accrual accounting, there should be scope for subcontracting the internal propriety audit work to the private sector. This can also help to disseminate private sector accounting practices within the public sector. Also, the external audit body can often employ private sector experts on a contractual basis to advantage.
  • Regularity in audit is important. While most industrial countries have annual propriety audits, this practice is by no means universal in developing countries.
  • Inspection can be selective rather than universal and is best directed by the sponsoring ministries. But it should form part of a collaborative effort with value-for-money audit to improve service standards. Increasingly, inspection is by reference to set standards rather than regulations (outputs not inputs) and in part carried out by testing, such as in education.
  • It is usually beyond the resources of the external audit body to carry out evaluations of all services every year. Again selectivity is important. The use of structured samples can be important to ensure that the results of any such studies will be representative and universally accepted. Value-for-money studies can be used to identify and then publicize the best practices in service delivery.
  • There needs to be a feedback mechanism from evaluation to budget preparation. Thus having publicized best practices in one year, both the external auditors and the service ministry inspectors need to question the local government response in improving service delivery practices in the next. This forms part of the budget cycle in Germany, for example. The authorities for their part need to be feeding in the changes necessary to move toward “best practice” in their budget preparation plans for future years.


This chapter has suggested appropriate institutional arrangements for achieving efficient budget and financial management, under a unitary as well as a federal structure of government. Considerable emphasis has been laid on coordination of both budget policy and implementation and on information sharing, again both for the preparation and implementation phases of the budget. While ideally applying to both the revenue and expenditure sides of the accounts, particular emphasis has been laid on the need for coordination and information sharing on local borrowing plans as a minimum requirement. This latter aspect is particularly important under federal arrangements where there is often reliance on borrowing control as the major instrument of macroeconomic management.

It might be considered that the proposals for greater coordination and information are too idealistic or that they interfere with the actual or perceived policy rights of the states, particularly under federal models. The requirements for both coordination and information sharing in developing countries may also need to be reduced somewhat to cover only the minimum information needs in budget implementation. In reality, however, certainly for the industrial countries and increasingly for the developing countries, the case for such coordination and information sharing is strengthened by a number of developments.

First, there is a general move toward more decentralization of functions because theory suggests a welfare gain from devolving choice closer to the local level. Yet without adequate coordination and control, the macroeconomic consequences may impose costs that over-whelm any resource allocation or efficiency gains. Second, there are efficiency gains to be secured, not least in allocative terms, from cooperation in the policy, implementation, and financing aspects of local and related central services. Third, the closer integration and internationalization of capital markets means that the local and central government sectors need to be aware of their interaction in dealing with the capital markets. That in turn requires more focus on financial planning and on the interaction between central and local finances.

This is an area, however, where there has been relatively little research. In particular, the modal requirements for “joint” financial planning, the case for greater scrutiny of balance sheet developments, and the role of central government actual or perceived guarantees for lowertier government debts are all areas worthy of further study.


In practice, poor information about the fiscal position of local governments sometimes requires limiting the programming exercise to the central government.


In federal states, central government control is primarily focused on a design of transfers and control over borrowing. However, even in federal states the central government may exert a degree of control on allocative and distributional aspects of local government through special purpose grants.


Symmetry might suggest that the plans of central government should also be shown to the states, for example, where both tiers can borrow directly. In practice, it may be simpler for the central government to advise states that a planned borrowing might be timed differently.

    Other Resources Citing This Publication