4 Intergovernmental Transfers

Teresa Ter-Minassian
Published Date:
September 1997
  • ShareShare
Show Summary Details
Ehtisham Ahmad and Jon Craig

The revenue and expenditure assignments discussed in the preceding chapters normally give rise to vertical and horizontal imbalances within a nation’s intergovernmental finances. A vertical imbalance occurs when the own revenues and expenditures of various levels of government within a federation are unequal. A horizontal imbalance occurs when the own fiscal capacities of various subnational governments of the same level differ. These imbalances must be resolved through a variety of transfer and borrowing mechanisms in order to allow the various levels of government to perform their allotted tasks within a national policy framework.

There are two basic avenues of transferring resources from one level of government to another: sharing of revenues or a system of grants. Revenue sharing can take several forms: tax bases can be shared, or taxes can be pooled and then shared. Tax administration can be seen separately from tax assignments, and revenues collected at one level of government, typically but not exclusively at the central government level, could be assigned in part or entirely to other levels of governments.

A grants system could involve conditional or unconditional transfers. In turn, these transfers can be open ended, or subject to caps. Moreover, some conditional grants may require matching elements by recipient governments. Each alternative approach is examined in this chapter. The choice of transfer mechanism depends on the particular mix of objectives of the policymakers.

Subnational governments play an increasing role in providing infrastructure and lumpy capital outlays in some countries. Such investments may be financed by capital transfers, and/or net advances from the center which may itself borrow to fund such payments. In some cases, subnational governments may be given direct access to capital markets. Unless such markets are highly developed, and other conditions exist that permit market discipline to be maintained on such borrowing, rules may need to be developed to ensure that the scope of such borrowing is consistent with the overall objectives of national macroeconomic policy.

Vertical and Horizontal Imbalances

The existence of a vertical fiscal gap at different levels of government, arising out of own revenue and own expenditure assignments, provides a basic rationale for a system of transfers and borrowing arrangements. In addition, national governments may wish to ensure that citizens in different regions and localities have access to a certain modicum of publicly provided services. Differential subcentral capacities are said to constitute horizontal imbalances. Differential fiscal capacities leading to horizontal imbalances constitute another rationale for the system of transfers between different levels of government. These objectives give rise to different combinations of grants.

Vertical Balances

The vertical fiscal gap is generated by the expenditure and revenue assignments. However, individual policy choices also play a significant role in determining the resulting ex post vertical gap. If a lower level of government chooses to increase spending or not raise assigned taxes, the vertical gap would increase. Thus, if transfers were designed solely to close the vertical gap, there would be little incentive for the lower levels of government to raise own account revenues or restrict or manage expenditures efficiently. Unless there are objective criteria for the determination of transfers, “gap filling” to finance subnational deficits is likely to lead to macroeconomic difficulties as well as indeterminate “bargaining” between the center and lower levels.

Since vertical balances tend to favor the central government, the size of the transfers to subnational levels of government often may be a function of macroeconomic stabilization concerns.

Table 1 shows the vertical fiscal gap for different levels of government in a number of federal and unitary countries’ governments. Three measures are shown: the vertical fiscal balance as measured by the difference between the own account transactions at the central or all sub-central levels of government; the vertical current balance, which measures the balance on own account current transactions only; and a vertical capital balance, which shows the extent to which capital spending is financed from own resources and provides an initial insight into the “need to borrow” by different levels of government.

Table 1.Vertical Current Balances1(Ratio of own source revenues to own source current expenditures)
Level of Government
CentralState or regionalLocal
Overall balanceCurrent balanceCapital balanceOverall balanceCurrent balanceCapital balanceOverall balanceCurrent balanceCapital balance
Federal countries
United States0.930.97−0.401.241.412.940.660.75−2.05
Unitary governments
United Kingdom1.−3.16
Source: International Monetary Fund, Government Finance Statistics Yearbook, 1993.Note: n.a. = not available.

The data show average ratios over selected periods for each country. The periods chosen are Australia, 1987-91; Austria, 1987-91; Brazil, 1982-91; Canada, 1985-89 (excluding 1987 for capital balance); Denmark, 1987-91; France, 1988-92; Germany, 1983-91; India, 1985-92; Netherlands, 1988-92; Spain, 1987-90; Sweden, 1988-92; United Kingdom, 1985-92; and United States, 1987-92.

Source: International Monetary Fund, Government Finance Statistics Yearbook, 1993.Note: n.a. = not available.

The data show average ratios over selected periods for each country. The periods chosen are Australia, 1987-91; Austria, 1987-91; Brazil, 1982-91; Canada, 1985-89 (excluding 1987 for capital balance); Denmark, 1987-91; France, 1988-92; Germany, 1983-91; India, 1985-92; Netherlands, 1988-92; Spain, 1987-90; Sweden, 1988-92; United Kingdom, 1985-92; and United States, 1987-92.

Horizontal Imbalances

In practice, very few countries measure the horizontal imbalances, or fiscal capacities, of their regional governments in a systematic manner. The federations that undertake a comprehensive review of horizontal balances are Australia, Canada, and Germany.1 An analysis of relative fiscal capacities of the different states in the United States was also prepared in the past, but is not used in determining transfers to states.2 Among the unitary nations, Denmark and the United Kingdom evaluate the fiscal capacities of their local government authorities in determining grants.

The horizontal imbalances in fiscal capacity may be addressed by equalization transfers from the center (as in Australia, Canada, and Denmark), or between regions (as in Germany). While some countries do not use an explicit “equalization framework,” redistributional elements are often introduced into special purpose or conditional grants in order to achieve equity objectives, such as in Indonesia. However, in the absence of an overall framework for determining and evaluating grants, it is often difficult to ascertain whether conditionality of use, with numerous and diverse special purpose grants, is actually inequality decreasing or increasing (see Ahmad, 1997). Empirical evidence from countries as diverse as Australia, Argentina, Canada, and China suggests that there may be substantial differences across regions within the country in terms of revenue bases, as well as cost of provision of services, and that these may necessitate the use of a consistent framework for evaluating and determining grants.

Policy Options

Three different policy responses can be made to this link between vertical and horizontal balances:

  • Correct each imbalance by separate policy measures. The vertical imbalance at each level is resolved by tax-sharing or grant arrangements. Horizontal imbalances are then resolved by payments from regions with higher fiscal capacity to poorer regions. This is the approach used in Germany.

  • Implement an integrated system of equalization grants. The vertical and horizontal balances are dealt with simultaneously through a system of grants, including equalization payments and special purpose grants. This is the Australian and Canadian approach.

  • Correct only the vertical imbalance and ignore horizontal balances. As under the first option, vertical balances are resolved by tax sharing and grants, but no action is taken to correct horizontal imbalances. Capital and labor migration then responds, not only to earned income differentials, but also to the regional net fiscal benefits (net benefit received from government expenditure and of taxes paid). There may be, however, special purpose grants servicing central government objectives, which may also reduce horizontal imbalances at least in some functional areas. This is broadly the approach in the United States.

The Rationale for Transfer and Borrowing Mechanisms

If the sole concern of central and subnational governments is to fill the vertical fiscal gap created by imbalances in revenue and expenditure assignments, this could be achieved either by sharing revenue from the major taxes on a “derivation” basis (that is, shared in proportion to the revenue collected in each region), or by “gap-filling” unconditional grants. Both approaches may exacerbate horizontal imbalances, and the latter also generates undesirable incentives for recipient governments that could weaken macroeconomic controls.

In practice, both the central and subnational governments normally have diverse objectives to be met through a system of transfers, and these different objectives may need to be met through a combination of policy tools. Typical central government goals include the following:

  • Ensure that the overall fiscal stabilization objectives for the national economy are met.

  • Provide an acceptable degree of equity between individuals in different regions.

  • Encourage the efficient use of resources across the nation.

These goals and the associated policy responses are discussed below.


Aside from variations in its own account revenue, expenditure, and borrowing operations, the central government has two weapons at its disposal in this area: (1) varying the level of transfers and/or revenue shares to subnational governments, and (2) adjusting subnational borrowing, where the subnational governments have separate, but controlled access to the capital markets.3 In some countries (for example, Canada), the latter option is not available to central governments. But, even in such cases, its relative size and influence may provide the central government with some leverage in persuading subnational governments to participate in coordinated stabilization efforts.

Since economic downturns often have a varying impact on different geographical areas within a nation, the central government may be expected to adjust both the level and geographical composition of its transfers to subnational governments in an anticyclical manner to assist the stabilization task. To the extent that the central government maintains an ongoing vertical current surplus with subnational governments, it can also draw on these resources to boost, or reduce, aggregate spending.


The term “equity” is often raised in the political debate on intergovernmental finances, especially in federations, and has a number of interpretations. First, there is the need to examine whether each region has the fiscal capacity to deliver an equivalent level of public services to its citizens. Second, there is a question as to which level of government should have primary responsibility for income redistribution.

Horizontal Equity Issues

Some federations (for example, Australia and Canada), but not all, have adopted extensive equalization arrangements. The arguments for such arrangements can be understood at both an intuitive and theoretical level. At the intuitive level, it can be argued that the creation of a national market place, or “economic space,” brings with it large benefits in terms of the production, distribution, and administrative scale economies, including common fiscal and monetary policies, common education, health, production and labor standards, and speedy communication and transport facilities. These benefits can be seen as being over and above those that would be available to any individual region, and a “fair” distribution of these benefits is often seen as the glue that bonds together the nation-state.4

The public finance literature approaches equalization issues from a different viewpoint. The term “horizontal equalization” traditionally related to governments providing equal treatment of individuals who are equally well off. Application of this concept involved an examination of the so-called fiscal residuum (see above).

The concept of horizontal equalization was used by Buchanan (1950) to justify equalizing measures between regions. The problem can be summarized as follows. Even if a central government treats equals equally, while at the same time each regional government treats individuals within its boundaries equally, the overall impact of public policies within a federation is likely to violate the equal treatment of equals principle. The difficulty stems from that fact that even if two communities, one rich and one poor, have identical levels of public goods provision, the wealthier of the two communities will, ceteris paribus, be able to meet its revenue requirements with a lower level of tax rates. This follows because, for a given amount of revenue for each resident, lower tax rates are required in a community with the higher level of per capita income. Therefore, from the standpoint of the federal system, equals are not treated as equals.

The issue can be handled in two ways. First, it can be ignored. Some argue that such an overall equity goal “is not a prime goal in a federation.” Others consider that the horizontal equity goal is inconsistent with regional diversity, which is a virtue of federal systems. But Buchanan and others disagree. They call for measures to compensate for such inequality by geographically discriminating tax rates at the central government level, to equalize the total tax bill of individuals, or to make equalizing grants to communities.

These simple theoretical concepts have been the subject of considerable debate. One major criticism has been that the original models used by Buchanan assumed that the total spending of each government is distributed over the relevant citizens in proportion to an assumed (equal) sharing basis and the resulting money amount is taken as a measure of the total individual benefit. But this approach is probably only relevant in the context of universally provided welfare benefits. In the case of public goods, such as education services, it becomes merely a cost imputation exercise and cannot serve as a measure of the individual benefits received.

The equalization concept therefore involves many practical measurement difficulties, since particular individuals may choose not to avail themselves of a particular service or avoid payment of a tax or charge by refraining from the activity concerned. For these reasons, even if two regional governments applied the same revenue-raising and expenditure-provision policies, the fiscal residuum of two equally well off individuals would not necessarily be identical. This situation would continue to apply even if the term “individual” was redefined into some concept of family unit or household. The existence of companies and other organizations further complicates application of the principle. An attempt to tackle the application of the “equal treatment of individuals” would also raise a host of practical measurement difficulties, including measures to determine equality between citizens of different regions. All these factors may combine to make the strict “equal fiscal treatment of equals” a difficult objective in either a federation or a unitary state. However, in some countries, increasing interest is being accorded to direct provision of transfers to individuals, for example, through vouchers financed by the central government. These vouchers often allow public services to be provided by private sector contractors. Such contracting out has led to a reevaluation of the role of the state as a service provider, in many countries, without necessarily changing the nature of the public good.

Nevertheless, it can be argued that a central government can achieve an approximation of the equal treatment of equals objective through transfers to different subnational governments. The aim then is to provide each subnational government with the ability to make available a uniform set of public services at a comparable revenue effort.

Three different practical options might be considered:

  • The regional government’s capacity to provide regional services without having to impose higher taxes and charges than other regions (or regional budget capacity equalization).

  • Comparability of the regional government’s capacity to provide equally well-off individuals with equal fiscal residuals (or “individual-based” regional government capacity equalization).

  • Uniformity across regions in actual standards of service provision and in actual taxes and charges applied (termed performance-based regional government equalization).

Strictly speaking, the third choice might come closer to the desired theoretical objective. However, its achievement would require the central government not only to make the difficult judgments on relative fiscal needs of different regions, but also to impose detailed and binding conditions on the grants paid to regional governments to ensure that those chosen services and taxation norms are implemented. The second choice avoids the conditionality problem but also encounters major measurement problems.

Of the three objectives, the first method may be more acceptable to many democratically elected governments. While it still faces difficulties in quantifying judgments about publicly provided services and taxes, it avoids the subjective measurement of individual differences implicit in the second choice. The equalization of subnational fiscal capacities is likely to be consistent with the actual abilities of governments to achieve the desired objective of equal treatment of equals.

Each of these approaches has strong advocates, and individual preferences are colored by cultural and historical differences. In the United States, for example, there has been an emphasis on performance equalization, at least in respect of specific purpose transfers for public goods such as education. This was also the case in the former Soviet Union and other centrally planned economies, where absolute expenditure norms were used as a basis for regional allocation of funds. By contrast, Canada, Australia, Germany, Denmark, and a number of other countries have favored systems involving varying degrees of capacity equalization. Australia and Canada employ equalization grants from the center, which have the effect of helping to close the vertical gap, while at the same time at least partially correcting for the horizontal imbalances.5 Another alternative approach involves transfers between subnational levels of government to close horizontal gaps, as employed in Germany.

It is important to recognize that the equalization approach does not necessarily imply full equalization of living standards (as, for example, measured by GDP per head), since private goods are not directly affected by such grant structures, and because subnational capital investment programs are generally managed separately. Even in respect of public programs, inequalities may remain, because different subnational governments may have different priorities. Some may choose to tax at a higher level and use the funds raised to provide a higher standard of public good, whereas others may have a preference for lower taxes and smaller governments.

Income Redistribution in a Regional Context

It is often argued that the role of redistributing incomes within a multilevel state should reside primarily with the central government. For some, this judgment reflects a view that common standards of redistribution ought to be applied, regardless of where people reside. A contrary view is that the degree of redistribution is essentially a question of taste and that preferences will differ between regions. Horizontal equalization of the type discussed above may provide each region with a capacity to provide some redistribution (which it may choose to use or not). However, there are clearly limits to such regional income distribution arrangements, since large differentials will tend to be thwarted by fiscally induced migration.

Nonetheless, the expenditure and taxation goals of subnational governments will frequently have redistributive consequences, even where the decisions are based on efficiency criteria. The extent to which subnational governments explicitly engage in redistribution across lower tiers depends on the availability of resources, as well as the need. In the United States, for example, within-state variations in incomes and service levels often exceed across-state variations. Thus, there is a role for state-level redistribution, while the federal government focuses on minimum standards. Moreover, the role of the federal government is being reconsidered, with greater autonomy allocated to the states.


As noted previously, theory suggests that, in the absence of externalities, expenditures should be matched to revenue sources, at the level of government at which the benefit is generated. Unfortunately, the adoption of simple guidelines is complicated by the existence of “externalities” in resource allocation, the so-called spillovers. The term spillover refers to instances where the benefits of a service provided by a subnational government spill beyond its own borders to benefit those not contributing to the cost of providing such services. Often-quoted examples are the cost of control of air or water pollution and the cost of educating students who relocate upon graduation to other regions. In each case, the total value of spending by a subnational government on the activities concerned should not necessarily be fully recouped from the residents of the region. In such circumstances, regional governments may consider only benefits to their own citizens and underprovide public services, hence yielding a case for supplementary central government grants for the given purpose.

“Flypaper Effect”

The so-called flypaper effect is an example of the difficulty of analyzing behavior in a multilevel government where the direct link between taxpayer or voter and the service provided is broken. The concept is based on empirical evidence that grants (or revenue shares) provided from one level of government to another tend to “stick” with the recipient government and be used for service provision, and will not be passed on to taxpayers in the form of lower taxes. As a result, the grant leads to a higher level of service provision than would be the case if the payment was made directly to individuals. To illustrate, take the following example. Region A receives a grant of $1,000,000, unrestricted and without matching requirements, which is used to increase expenditure by $800,000 and to provide tax reductions of $200,000. If the grant had been made directly to the individuals of region A, they may have voted to use only, say, $300,000 to increase services, but $700,000 to reduce taxes. The overprovision, or flypaper effect, is then $400,000. In the case of matching grants, the flypaper effect is measured by comparing the level of the public service if the matching rate were applied to the budget and that which would result if matching rates were applied to individual taxpayers. Empirical estimates suggest that the magnitude of the flypaper effect may be considerable in some countries.6

Minimum Standards

In practice, many central governments wish to exert some influence over the minimum standards of service provided at subnational level, which affect both efficiency and equity objectives. These programs are advocated for a variety of reasons, such as:

  • The national interest, which is seen as being served by achieving some commonality of standards in particular functions (for example, primary school training or roads) mainly provided by subnational government.

  • Greater harmony between the programs of subnational governments.

  • The assurance of social objectives, such as a politically acceptable minimum living standard. This standard varies from country to country, covering programs such as Aid to Families with Dependent Children in the United States and access to public services, such as housing, in the United Kingdom and Australia.

Such transfers may have some spillover effect—such as implementation of national standards and/or performance guidelines that benefit business expansion. Furthermore, there may also be a desire to use such programs to promote a more equal treatment of the citizens of the country.

Of course, central government intervention is not the only possible response to spillovers, nor need it be the most efficient solution. One alternative, seemingly gaining attraction is for subnational governments to work together to operate regional services, such as railways, which yield benefits beyond their respective economic borders.

Conflicts and Complementarity in Objectives for Transfers

The various objectives outlined above for intergovernmental transfers may not be mutually consistent. Some of the issues that may arise are discussed below.

Stabilization and Equity

One example of conflicting objectives is that between the goals of stabilization and equity. This conflict can arise where revenue or expenditure actions taken by the central government to address stabilization concerns have a differential impact on the fiscal capacities of the subnational governments. For example, an income tax imposed by a central government may act to reduce the income and/or consumption tax bases available to subnational governments in a manner that differentially reduces their individual capacities to raise revenue. Similarly, expenditure cutbacks at the center may affect the expenditure needs of subnational governments in a differential manner.

Equity and Efficiency?

Conflicts may also arise between the efficiency and equity goals. Thus, the use of grants to relieve perceived efficiency problems (for example, conditional grants for education requiring the achievement of minimum standards within certain regions) may conflict with fiscal equity goals. This occurs because such grant arrangements ignore the overall fiscal capacity of a region. It is conceivable, therefore, that a region with a relatively high revenue-raising capacity may receive a relatively large education grant simply because it had chosen in the past to spend relatively less on education and, most probably, enjoy a commensurately lower level of taxation. It is possible to reconcile this potential conflict by designing the grants system within an overall equalization framework (see Appendix).

There has been considerable debate about whether efforts to equalize through intergovernmental grants detract or add to the overall efficiency of the economy. There are two strands to this argument. The theoretical debate on this subject7 centers on the possibility that, in the absence of equalizing transfers, residents in one region will move to another that exhibits higher net fiscal benefits, and whether this “fiscally induced” migration, in turn, creates economic efficiency costs making the nation as a whole worse off.8

Boadway and Hobson (1993) provide an intuitive example within the Canadian context,9 to demonstrate that a free migration equilibrium may be inefficient, when regions have different degrees of access to source-based taxes. This case is important in Canada where the provinces have unequal access to natural resources taxes. Moreover, while the free migration equilibrium will be efficient if residence-based benefits are levied to finance the public provision of private goods, the chances are that such outcomes will not be consistent with the redistributive goals of provincial governments, particularly where regions have to impose taxes on an ability to pay basis. Thus a proportional income tax imposed to finance publicly provided public goods will be redistributive when individuals are heterogeneous with respect to their ability to earn income, and the free migration equilibrium will be inefficient if average per capita incomes differ across regions. Within this model, therefore, efficiency requires that the average per capita income tax revenue be equalized across regions.

Within the model postulated, efficient free migration equilibrium would require either a rearrangement of tax assignments between levels of government to better align types of revenue sources with the types of expenditure to be financed by subnational governments or a system of intergovernmental equalization transfers.

The “efficiency arguments” for equalization are not universally accepted. For example, Swan and Garvey (1992) challenge the view that net fiscal benefits create migration equilibria. In mounting a case against equalization, and especially the original analysis of Buchanan (1950), they suggest that a free migration equilibrium is efficient, since the migration process itself would lead to an equalization of incomes. Thus, all individuals would be equally treated from a public policy perspective.

Alternatives to equalization payments by central governments are also discussed in the literature. Cornes and Sandler (1986) draw on the theory of clubs to argue that regions could impose migration taxes on intending migrants. Myers (1990) also investigated the possibility that regions might undertake voluntary lump-sum transfers among themselves in order to deter migration. His results, which assume that migration is costless, suggest that a system of voluntary transfers would suffice and that no central government equalization transfers are required. Hercowitz and Pines (1991) continued this theme by examining the conditions under which a region will voluntarily transfer resources to another in the case where mobility is costly. They showed that, in this case, the interregional transfers will not be socially optimal and that central government transfers will still be required.

A number of empirical studies have also been undertaken in Canada to try to determine whether migrational efficiency distortions are potentially important.10 These studies confirm that (1) differential net fiscal benefits do exist; (2) resource endowments are a significant contributing factor to these differential net fiscal benefits; (3) inefficient migration does occur in response to these net fiscal benefits; and (4) equalization slows inefficient migration. However, while these studies support the expected direction of change in the relevant variables, they do not provide much guidance on whether the magnitude of change is significant, or that the changes have had a major impact on national welfare.

Ultimately, the issue of whether equalization affects efficiency and growth depends on empirical investigation.

Transfer Options

The main transfer mechanisms used to tackle the various goals of government can be grouped into the following categories: (1) conditional transfers (or specific purpose payments) and (2) unconditional transfers.

Conditional transfers consist of matching grants, nonmatching grants for specific purposes, and block grants. Each of these may be implemented with or without redistribution criteria, which could be either open- or closed-ended.

Unconditional transfers consist of revenue-sharing arrangements, with or without redistribution criteria, and general purpose grants, which may be open-ended or subject to caps. In most countries, a combination of grant mechanisms will be used depending on the objectives of the central government.

General Choices

If grant mechanisms are employed, four levels of choice can be identified.

First, there is a basic choice as to whether the transfers should be made on a conditional or unconditional basis.

In many countries, the central government imposes conditions on some transfers on the use of the funds, and/or the performance achieved in the programs as a means of increasing central government influence over spending, which is primarily the responsibility of regional or local governments but which may also have an impact on matters of vital national concern (for example, air or water pollution control). Similarly, the central government may wish to leave the primary administrative responsibility for certain functions at the regional or local level, but seek to attain national minimum standards in these functions because of a national public concern (for example, on access to health or education programs). The extent of conditions imposed is likely to vary. At one extreme, the regional government may simply be reduced to acting as an agent of the central government. At the other extreme, the conditions may be limited to such matters as information supply, leaving the subnational government with ample scope for innovation and experimentation.

An unconditional grant will often yield a greater increase in utility in the recipient jurisdiction than will a conditional grant. This follows because unconditional grants simply increase community income without altering subnational government spending priorities, which themselves are dictated by local preferences.

The main justification for conditional grants over unconditional grants, therefore, must be that local decision making fails to produce the socially optimal outcome, as in the case of interjurisdictional spillovers discussed earlier. However, many developing countries have relatively weak expenditure management capabilities at the subnational level, and the proliferation of conditionality and performance criteria for special purpose grants is likely to generate confusion and pro forma fulfillment of the needed criteria. Thus, unless they possess the ability to monitor and manage the conditionality for grants, central governments would do better to simplify the design and conditionality of special purpose grants, and to supplement these by lump-sum transfers (which would be seen as own resources by recipient governments).

Second, within the category of conditional transfers, there is also a choice as to whether the central government should require subnational governments to undertake some matching of funding of programs by lower level governments.

Matching conditional grants generally alter local priorities to take account of the central government’s spending preferences. They are particularly effective where spillovers are thought to exist. There is extensive literature on this subject. Boadway and Hobson (1993), for example, demonstrate that the optimal matching rate is the rate that will induce the recipient government to provide the socially optimal level of public service—that is, the level for which the marginal social cost is just equal to the marginal social benefit.

Another potential difficulty in administering matching arrangements is that different regions may be able to exert different leverage on the size of the overall grant, because of their different fiscal capacities. In order to limit the ability of any one region to influence the size of the grant, the amount of grant may be tied to total expenditure summed across all provinces. The grant is then tied to expenditures in all provinces, the cost to provincial taxpayers of an additional dollar of expenditure is $(1 − s/n), where n denotes the number of regions. That is, the additional dollar of expenditure generates a transfer of $s from the central government but the transfer is spread over all regions, some what reducing the leverage of particular provinces as well as the disincentive to use resources effectively.

Third, there is a choice as to whether there is to be some redistribution in the transfer mechanism or whether the transfers will be simply made on an equal per capita basis to each member of the defined population in each region. While formal equalization is usually restricted only to general purpose transfer systems, some element of redistribution between geographic regions is often built into conditional grants (for example, for grants to poorer regions where education or health needs are greater). However, in the absence of an overall framework for evaluating grants, it is not obvious that separately formulated conditional grants (or revenue sharing) with redistributional factors will actually be inequality decreasing or redistributive in an aggregate sense.

Finally, within both conditional and unconditional transfer mechanisms, there is a choice of whether the grants should be open-ended or subject to some limit. Open-ended matching grants encourage local governments with an incentive to internalize identified spillovers and provide the required level of services. At the same time, stabilization considerations often lead central governments to regard such arrangements with some skepticism, and limits to grants are considered desirable on macroeconomic grounds.

Tax and Revenue Sharing

If revenue-sharing or tax-sharing approach is followed, the choices are normally reduced. For example, transfers of shares of major tax collections (such as VAT) are generally made without conditions since they are not related to specific expenditure functions. However, this need not always be the case. For example, excise taxes collected on fuel consumption are often earmarked for conditional transfers for specific road construction. Similarly, most tax-sharing arrangements do not require matching by subnational governments. Most are open-ended in nature, although stabilization policy requirements may induce central governments to impose floors and caps on the amounts transferred.

Generally, tax-sharing arrangements are made on a derivation basis, although formula-based distributions are to be found in many countries. As described above, it is not evident a priori that the overall effect of separately determined redistributive formulas will decrease overall inequality in the absence of a consistent framework for the evaluation of transfers.

Where different sharing ratios are established for different taxes, there may be an incentive for the administering authority to place more effort on taxes that provide greater benefits to its immediate level of government (see Chapter 5 on Tax Administration). India, China, and Russia have experienced such difficulties. There is thus a possibility for “strategic games” between different levels of government that might cloud the transparency of the tax system.11

Alternatively, a separate “fund” or “pool” may be established, and resources may be distributed automatically on a “formula” basis. The disadvantage with such an arrangement is that the automatic sharing of all revenues could complicate a stabilization package, where it is generally assumed that lower levels would spend all additional resources. Thus, the federal government may have to make a larger own account fiscal adjustment than might otherwise be necessary.

The advantage in tax sharing is that lower levels of governments share in productivity gains leading to enhanced tax collections without having to petition the central government for additional resources (and ensuring that subnational governments have a vested interest in an efficiently functioning economy). This automaticity is not generally assured through a grant system. Moreover, those favoring the tax-sharing approach also often point to the added accountability in public decision making that may be expected to accrue, given that such resources are seen as own revenues of the recipient governments.


Unconditional equalization programs are in use in many parts of the world. These may be based solely on differences in revenue (or tax) capacities across subnational levels of government, as in Canada. This approach implicitly assumes that there are no significant differences in cost in the provision of public services across provinces, or those that exist are taken care of by special purpose transfers (see Clark, 1997).

An alternative formulation is based on both the assessment of revenue capacities, as well as an explicit incorporation of expenditure needs (such as in Denmark and Australia). This more general formulation requires more data than the revenue-capacity only option (see Rye and Searle, 1997; and Craig, Chapter 8). It is important that the factors chosen for the estimation of expenditure needs be independent of the actions of individual provinces. If this criterion is ignored, then there is a danger that the process could be manipulated by recipient governments. In this case, the system could degenerate into a variant of gap filling (see Ahmad and Thomas, 1997). Note that even if an elaborate formulation of expenditure needs is adopted, the resulting equalization grant is still untied and the recipient governments could choose to spend it as they wish. For instance, a relatively low level of public services could be provided, with lower-than-standard tax rates, or higher-than-average services with a higher tax effort. This would not affect the amount of the equalization payment, which is lump sum. Provided that the recipient governments are able to vary tax rates and effort, a lumpsum transfer should result in greater accountability, as the policy choices by a spending government would be reflected in the resulting tax rates and burdens.

As with special purpose grants, a choice has to be made whether the equalization estimation should determine (1) the relative amounts going to different provinces; and in addition (2) the total amount to be made available for this purpose.

It is clear that the relative amounts going to each province should be determined by the equalization exercise. This could be based on transfers only to poorer provinces, organized by the central government, of transfers directly from richer to poorer provinces organized on a cooperative basis (as in Germany). Both variants have the advantage that the redistribution is clear and transparent. However, not all countries have the political cohesion that would permit a duplication of the voluntary transfers as in Germany. In addition, the visible redistribution may lead to strong opposition to transfers on the part of the better-off provinces (many of which also have unmet expenditure and investment requirements). Thus, equalization transfers going from the center to all provinces may have certain political advantages. This organizational choice will vary from country to country, depending on political economy realities.

The issue of the amount to be made available under equalization is important. Even in countries that conduct elaborate estimations for the basis of equalization, the amounts for this purpose may be less than that for special-purpose transfers (as in Australia). In some cases, the specification of the formula also determines the amount of the grant. And if this is more than can be justified given macroeconomic constraints, countries such as Canada have resorted to adjusting the formula or imposing caps (see Clark, 1997). The Australian formulation focuses on the determination of relativities for each state, which are then applied to the amounts to be transferred, determined exogenously by the government (see Chapter 8). This arrangement provides greater flexibility and objectivity to the estimation procedures.

Another decision that has to be taken relates to the extent to which special purpose grants are to be incorporated into the equalization process. On their own, numerous special purpose grants may be inequality increasing. Incorporation of such transfers within the equalization process provides a framework for such grants.

A description of a general grant determination process, based on both revenue capacities and expenditure needs, is presented in the Appendix. This draws largely on the Australian model.

Capital Grants

Many countries make extensive use of capital grant systems to finance public investment programs by subnational governments. This is especially the case in countries that do not have well-developed capital markets or where the weak financial position of subnational governments does not permit them to access such markets directly. While most of these grants are for specific purposes or projects, a question arises whether capital needs should be included in an equalization program. This may be of importance when some regions of a developing country lack the basic infrastructure (such as school buildings) necessary for the provision of key public services deemed to be relevant for the equalization exercise. In this context, it would be important to examine the methods of assessing needs, the appropriate mix of grants and advances, as well as the nature of payment arrangements.

As noted earlier, capital needs are normally excluded from equalization exercises because of the difficulties of measuring and assessing relative needs in different regions. Ideally, one might compile a list of projects and conduct cost-benefit analyses on each project. The projects might then be funded in descending order of merit, with a specific capital grant for each functional area, although there are considerable practical difficulties involved in implementing this theoretically appealing approach.

In particular, a difficult area is to decide on the extent to which project financing is to be based on a grant and the portion that should be financed through capital markets. Proxy and sometimes arbitrary measures must therefore be used to allocate funds. In some areas, such as roads and housing, some approximate stock measures may be derived that can serve as a crude basis for evaluation (but see earlier discussion of the varying possible approaches to equalization of capital needs).

These difficulties suggest that a prudent approach to formulating capital grants may argue for providing separate financing for large infrastructure projects (for example, regional airports) and some capital investments of a repetitive nature (for example, rural and district roads and low-income housing), but to leave other smaller enabling investments to be financed by block grants or the general purpose equalization grant (if the relevant activities had been considered in assessing the factors for such a grant).

It must be recognized that capital projects have a long life and that at least a portion of their benefits will be enjoyed by future generations. Moreover, for many large infrastructure projects user charges can be implemented to finance major portions of the total cost. Both these factors make equalization grants an inappropriate tool for financing such capital projects. Consideration should then be given to financing all or part of the projects by advances from the central government at market interest rates that at least cover the center’s own cost of borrowing.

In practice, central governments often employ a mix of special purpose grants and advances for large capital projects. The split of advances to grants must again be made on a pragmatic basis, reflecting judgments of the circumstances surrounding each type of expenditure.

Planning and Implementing Transfer Arrangements

The design and monitoring of intergovernment transfers is a complex task as it involves balancing many competing objectives across a wide range of activities. Some general principles emerge from country experiences:

(1) Stabilization concerns should predominate. The central government must preserve its capacity to manage the economy. Failure here may jeopardize other goals being pursued by the central and subnational governments.

(2) Developing a macroeconomic framework. Individual transfer arrangements should not be negotiated in a vacuum. Ideally, they should be considered in the context of jointly prepared medium-term rolling fiscal plans covering projections of the aggregate revenues and expenditures of all levels of government at least three years ahead.

(3) Arrangements should contain some flexibility. While the central government cannot avoid some degree of commitment of medium-term resources, it also needs room to vary overall levels of commitment. Open-ended commitments and variable grants or tax shares with floors or minimal guarantees need to be scrutinized carefully. Indexation arrangements—especially those that might be susceptible to actions by the grant recipient—also need to be handled with care. And while some medium-term funding indications may be unavoidable, a reasonably large component of the total grant should be left for final determination in the annual budget context. So-called sunset clauses should apply to ensure that programs are regularly reviewed.

(4) Objectives must be clearly spelled out and be capable of being monitored. The central and subnational governments must be particularly cautious of overlap and duplication in the provision of services. When central governments set conditions on the use of transfers they should have a good idea of how to monitor their objectives, what can be realistically achieved, and what sanctions can be applied in case of nonperformance. Matching requirements in many specific areas may simply undermine the capacity of the subnational government to manage crucial local services.

(5) Interrelationships need to be taken into account. While individual ministries will naturally focus on achieving specific objectives in any program, it is important that the ministry of finance monitor the overall impact of transfers. The consistency between conditional transfers and unconditional grants is important.

(6) Simplicity is important. The initial presumption should be that the expenditure needs can be met by the subnational own revenues, including shared taxes, and unconditional grants. The case for a special purpose grant should be clearly spelled out, with simple conditionality and reporting procedures.

(7) Examine alternatives. Not all services need to be provided by the state (at any level). In some cases, payments can be made to companies or persons via a voucher or some other mechanism to facilitate access to services such as education. Such measures may be particularly important where the national government seeks to encourage efficiency and diversity in the supply of services.

Concluding Remarks

Most countries need a combination of grants, ranging from special purpose grants, to correct for spillovers and other conditional grants to meet the policy objectives of the central government, to other conditional or unconditional grants, to meet vertical or horizontal imbalances. Each combination of grants will likely have different macroeconomic implications and could affect the incentives of recipient governments to raise their own revenues at the margin or to control own expenditures. A strong lesson from both the theoretical literature and the experiences of other countries (see Ahmad, 1997) is that the design of grants matters for macroeconomic stabilization, as well as efficiency and distributional objectives.

Gap-filling grants to meet the deficits of subnational governments are pernicious and should be avoided to the extent possible, to minimize the danger of fiscal irresponsibility.

Central governments will continue to rely on special purpose grants for a variety of reasons. However, the objectives should be stated clearly, and the conditionality defined in a manner that can be monitored and enforced. It is all too common to find the greatest reliance on special purpose transfers in countries that are unable to monitor effectively the usage of these transfers. Under these circumstances, such grants can induce inefficient resource use, as well as corrupt practices, because of the lack of accountability engendered. In addition, complex conditionality can lead to administrative paralysis at lower levels of government. Moreover, a multitude of special purpose grants bestowed by different agencies can lead to undesirable outcomes, such as an overall increase in horizontal imbalances. Thus, a simple design of special purpose grants, within a consistent overall framework, is to be strongly recommended.

Unconditional equalization grants, provided that these are lump sum and are not influenced by the actions of a recipient government, can be an effective vehicle for financing decentralized expenditures in a manner that encourages accountable resource use. Difficult choices need to be made in relation to the extent and formulation of such transfers, and these need to take into account data and institutional constraints that are to be found in particular contexts.

Appendix. Implementation of Equalization Grants Systems

Subnational governments differ in their fiscal capacities because some can raise more money from their available tax base than others and because the need for and the cost of providing certain services differs among regions.

A number of geographically large countries have felt a need to provide some intergovernment fiscal mechanisms to provide a degree of equity between regions over time. Specifically, full equalization would require that each region should have the capacity to provide the same standard of public services as the other regions, provided it makes the same effort to raise revenue from its own sources and conducts its affairs with an average level of operational efficiency.

If a country anticipated the introduction of a system of grants based on horizontal equalization, a decision would be needed as to the basis for equalization in the medium term: (1) whether to restrict the process to revenue capacities, or (2) whether to include both revenue capacities and expenditure needs.

If the decision is the former, then the revenue capacity equalization could be introduced fairly quickly in most countries. However, in large and diverse countries, there may be a preference to also allow for differences in expenditure needs, particularly if there are substantial differences in the cost of provision of such services and in access to such services.

Whatever is done and no matter how long full implementation might take, the long-term objective must be kept in mind at each stage of the system design, and it is therefore appropriate at the start to set the principles on which the system is to be based. Some considerations in designing the process and objectives of the system might be as follows:

  • Ensure that, as far as possible, the grants do not simply fill fiscal gaps in subnational government budgets. Thus, the recipients of grants should not be able to influence their grant share by their decisions—thereby preventing policy-induced disincentives.

  • Achieve an implementable system without imposing too great a burden on government in either the collection or processing of data.

  • Involve the subnational governments (especially at the state or provincial level) in the design of the system to achieve a degree of political consensus. This is likely to involve a gradual process in which sharp changes in provincial activity levels are avoided.

With the above considerations, a needs-based capacity equalization may have to be introduced in line with the development of databases and the choice of variables to be included in the exercise. The initial priority might be to introduce revenue equalization as soon as practicable but, since most countries have special purpose grants, these should be examined for inconsistencies with the equalization process and objectives—this too can be effected early on. It should be noted that an equalization process based on needs does not imply that the central government determines what lower levels of government must do. Rather, this should be seen as providing subnational levels of government with the capacity to provide a standard of services, and they may choose to do otherwise. The important aspect is that the equalization grant should be lump sum, thus ensuring that recipient governments have the incentive to use the resources wisely, and to manage their expenditures efficiently.

The steps needed to introduce the full revenue capacity and expenditure needs framework in an administratively manageable fashion are outlined below.

The First Step: The Scope of the Equalization Budget

Ultimately, the objective of an equalization system is to give provincial governments the capacity to provide equal levels of public services if they make equivalent tax efforts. This does not ensure equality in the actual provision of services to each person. An important first step in system design, however, must be to identify the range of public services that a provincial government should have the capacity to equalize.

The initial question would be whether to include both capital and recurrent expenditures and revenue sources within the range of functions to be considered. Typically the equalization process focuses on current expenditures. This however is less defensible in developing countries where the provision of current public services is constrained by the absence of appropriate infrastructure, such as school buildings.

On the other hand, there are major problems with measuring capital needs. Three approaches may be considered:

Flow equalization. This approach would just examine the capacity of each region to deliver a standard increment in own-financed capital stocks in any time period.

Stock equalization without memory of past accumulation. This approach would simply examine the capacity of a region to deliver a standard stock of own-financed capital in any time period without reference to past investment.

Stock equalization with memory of past accumulation. This approach would evaluate a region’s capacity to provide own-financed capital over the long-term, making each region bear the consequences of its own past actions.

Aside from the immense statistical difficulty of measuring regional capital stocks, each of the approaches has particular interpretation problems, which may prove insuperable. For example, poorer regions may object strongly to the first approach, on the grounds that the deficiency of the initial stock was due to insufficient transfers. The second approach could be opposed by states that had made a consistent effort to boost their capital stock, and may be penalized by this arrangement. The third approach may find more support among constantly good performers, but would not be favored by regions that had consistently ignored their capital needs.

Two parallel systems could be implemented:

  • A system of special purpose capital transfers enabling the central government to target specific infrastructure needs and projects.

  • An equalization system for untied transfers to assist in financing recurrent services. It may also be possible to include small capital investments within the equalization formulation on a needs-based approach, such as local public works and rural roads.

Of course, this approach would not necessarily mean that capital needs would be ignored in the equalization exercise. Capital needs may enter in various ways into the assessments. For example, regions may be assessed on their relative debt charges that would take at least some account of the accumulation in capital stock contributing to the present level of debt payments.

Once the range of recurrent services to be covered in the equalization exercise has been decided, this would be related to the own-revenue capacities of the provinces, along with the equalization transfers from the central government. The overall transfers would incorporate an estimated inflation adjustment, and only in very exceptional circumstances would there be ex post adjustments (for example, for major population shifts). The basic relationship is:

Cost of standardized recurrent services = provincial sources of recurrent revenue (at standardized levels) available to finance these functions + transfers from the central government (both untied and special purpose).

In an important way, therefore, the provincial capacity to link revenue sources to specific expenditure functions has an influence on the scope of the expenditure budget. If provincial revenues are very largely uncommitted, a matching of standardized revenue sources with standardized expenditure functions is impossible. Thus, if only one function is chosen to represent expenditure needs (such as education) and is related to total available own revenues, there would be a large standard budget surplus (Bi), and a danger that the factors for education might misrepresent overall expenditure needs. It would thus be preferable for the standard budget to include as many relevant provincial services as possible. Methods of achieving this (with and without extensive data) are discussed below.

It is seldom possible to completely separate expenditure assignments among various levels of government, given that in each expenditure category there are responsibilities for policy, financing, and administration (see Chapter 2). In Australia, universities are constitutionally assigned to the states. However, all the financing comes from the commonwealth government, which provides funds for this purpose to the states. In policy terms, the universities are fairly independent. For equalization purposes, the universities are not treated as being a state expenditure function, even though the rest of the education sector is included. The definition of a “state-type public service” in the Australian standard budget covers items mainly administered by the states, even if such items are also provided by other levels of government, communities, or public sector enterprises.

In theory, the equalization objective should include all aspects of service provision regardless of the provider. It should not matter how the service provision is arranged or financed. The grants to the provinces can still be determined in such a way as to equalize capacity for them to ensure that, in total, there could be equal provision in each province.

It may be that, in at least some functions, the combined level of activity of the nonprovincial government sources can be assumed to satisfy the same proportion of the eligible group in all provinces. If this assumption holds, then the nonprovincial activity would not affect relativities if excluded from the standard budget and the assessment of disabilities.

If the level of activity of the subprovinces and enterprises is not assumed equal in all provinces, then there are two possible options for proceeding. Either the standard budget, or the expenditure assessment, can be adjusted.

The first option would require the inclusion of the following items in the standard budget:

  • The revenue or expenditure in the provincial government accounts.

  • The expenditure on the standard budget items included in sub-provincial and enterprise accounts.

  • The revenue received into subprovincial and enterprise accounts used to finance the standard budget expenditure items (this would normally be equal to expenditure but may be financed from higher levels of government).

The second option of adjusting expenditure assessment approaches is discussed below.

The Second Step: The Structure of the Standard Budget

This step involves the organization of items to be included in the standard budget in a manner that best assists the assessment process.

On the revenue side of the budget, it is probable that a small number of major sources of revenue should be assessed separately. For the others, the task involves distinguishing some common elements that might contribute to differences in revenue capacities and, where they exist, grouping the taxes into one category. For example, although taxes might differ on different types of gambling, it might be decided that income levels are the major source of capacity differences and that all gambling taxes might best be combined into one category.

Similarly, looking at the expenditure functions of government, it might be that some of the small functions such as the registration of births, deaths, and marriages; registration of motor vehicles; and other administrative services could be combined into one single category labeled, say, administrative services. The IMF’s Government Finance Statistics classifications provide a good guide to appropriate groupings.

Experience with equalization suggests that there is a tendency for separate categories to be developed in the early assessment stages, because they have attributes (especially availability of some specific data) that appear to lend themselves to separate analysis. On closer inspection, it is often found either that their weight in the total budget is so small as to make separate analysis unwarranted or that they would be better grouped with other items that have common underlying, revenue, demand, or cost influences.

The Third Step: Deriving Standardized Revenues

There are three options for determining the standardized revenue for any source of revenue to be included in the standard budget.

Undertake an “Active” Assessment

This involves the determination of the revenue base using objective data. This may be based on the actual revenue base being accessed by the provinces (such as payrolls) or some overall indicator of economic activity that can be used as an objective measure (such as state gross product on total consumption spending). An important consideration in this task is that the data used should not be simply a reflection of the actual revenues; rather, standard tax policies should be applied to an “independent” provincial tax base to reflect relative revenue capacities (for example, if there are a number of exemptions from the payroll tax base permitted by one province, then an attempt should be made to assess what the base would be if those exemptions had not been made). This difficulty points to the use, where possible, of nationally consistent databases.

Undertake an Equal Per Capita Assessment

This approach is best used when analysis suggests that there is no measurable difference in the per capita capacity of different provinces to raise revenue from a particular revenue source. For example, an equal per capita assessment may be appropriate if provinces collect a poll tax of $100 for each household and demographic data reveal that the number of persons in each household is broadly equivalent in all provinces. Any differences in amounts collected may then be seen to be essentially due to differences in the efficiency of tax administrations, or to policy differences such as particular exemptions granted to some households in some provinces. These differences would imply a need for the provinces concerned to address the efficiency or policy problems leading to lower collections, or make a case to the assessment authority that it suffered an unavoidable disability in the area concerned.

Undertake an Actual Per Capita Assessment

This approach could be taken when analysis suggests that there are few differences in the tax bases, tax rates, or efficiency of collection among provinces in a particular item. The actual collections could then be taken as a sufficiently accurate measure of underlying differences in tax capacity and, therefore, used as a measure of standardized revenue in the overall assessment. An example of successful use of this approach may be a tax item that is collected by the central government (such as VAT) and then shared with the province on a derivation (original source) basis. In that instance, it could be reasonably assumed that the policy and efficiency of the collection authority are broadly the same across all provinces.

There is a danger that the use of actual per capita assessments could be gap filling in that they discourage efficient tax collection practices. It is important, therefore, in any decision to use this approach, to ensure that no significant policy or efficiency differences exist. Revenue trends in any areas assessed by this method should be monitored closely to identify any attempts at grant-share manipulation and, if such action occurs, early steps will be necessary to vary the assessment approach.

The Fourth Step: Deriving Standardized Expenditures In this area, four alternative approaches are available:

The Factor Assessment Method

This method is used when there is sufficient confidence in an ability to identify provincial differences in the relative underlying demand and/or cost per unit of service provision. The process involves the identification of relevant disability factors and then quantifying their relative influence on the service in question. For example, in the area of primary education, the number of people in the age group of 5 years to 12 years might be seen as a relative measure of demand for the service, and differences in cost due to energy consumption to heat or cool schools might be seen as causing differences in the unit cost of providing education to a student. The Australian experience has been that demand influences are usually easier to quantify than cost influences, either because of the existence of better data or because of greater ease in getting agreement between recipient governments on their relative impact.

A further consideration is that demand factors are likely to be specific to individual expenditure categories in the standard budget—for example, different age or sex—and demographic influences are likely to apply to health and education services. However, cost factors tend to have a more common impact across functional expenditure categories. For example, Australian experience suggests that cost factors, such as variations caused by diseconomies of scale, variations in population density, and dispersion of the population, may have a broadly similar effect on many budget functions. In that context, studies of these broad cost factors, while time-consuming in themselves, may yield benefits for all aspects of distributional policy, including the allocations used in specific purpose grant programs.

Undertake an Assessment Based on a Current Distribution of Specific Purpose or Tied Grants

This approach may be appropriate where there are insufficient data to apply the factor assessment method but there is a specific purpose grant scheme in operation that contains a distribution system that the parties judge to be fair and appropriate (perhaps because of past studies of underlying demand and cost factors). This assessment process simply compares the individual province’s per capita share of the specific purpose payment with the national average per capita grant, and uses the relationship between the two as a measure of the global disability that should be applied to the function.

There are two benefits to this approach. First, it does not detrimentally influence the relations between the line ministry currently responsible for the specific purpose grant allocation and those charged with making the equalization grant (in some countries, an independent grants commission, and, in others, the ministry of finance). Second, it does not detrimentally influence the budget flexibility of the provinces by overriding the distribution of the specific purpose payment to which they have become accustomed, without an assessment having been made.

Undertake an Equal Per Capita Assessment—as with Revenue

The assumption is that provinces have no relative demand differences or variations in unit costs of providing services. Standardized expenditure for each province is thus equal to the standard level of expenditure.

Undertake an Actual Per Capita Assessment

Similar in principle to the revenue case mentioned above, the assumption here is that actual costs are an accurate measure of overall relative need. Actual expenditures can therefore be used as an accurate measure of standardized expenditure.

The problem of different contributions to total service provision being made by the nonprovincial suppliers could be overcome in the assessment process. This would avoid the possible difficulty of including subprovincial and enterprise financial transactions in the standard budget. If it was known, or could be estimated that nonprovincial supply of a service across the nation reaches, say, 20 percent of the total relevant population, this could be seen as a standard level of provision. Thus, a province in which more than 20 percent of the eligible population is being served by nonprovincial suppliers would have an advantage. It could be assessed as having a “nonprovincial supplier” factor of less than unity (that is, a relatively lower-than-standard disability). A province in which the nonprovincial suppliers were serving less than 20 percent of the eligible population would have a factor of more than 1.000 (that is, a relatively higher-than-standard disability).

Once such factors are applied, any change in the distribution of services between provincial and other suppliers would need to be closely monitored to ensure that provinces do not use policy decisions to change the balance in an attempt to influence their assessments of relative needs.

This approach may not be adequate for longer-term assessment if provinces are able to change policies in a manner that can affect the assessments. However, it may be acceptable in the short run if the inclusion of the financial activity of nonprovincial suppliers is judged to be too difficult to include in the standard budget.

A further aspect to this solution is that it is effectively excluded when equal per capita assessments are undertaken. This arises because the equal per capita assessments, if nonprovincial expenditure is excluded from the standard, assume an equal provision of nonprovincial supply. It may be that the equal per capita assessments need to be divided into two groups: those in which it is desired to assess a non-provincial supplier factor in which case that becomes the only factor; and those where a true equal per capita assessment remains.

A nonprovincial supplier allowance, however, is automatically included in actual per capita assessments. This will not present difficulties because any policy action by a province to reduce its expenditure at the expense of a nonprovincial supplier will automatically result in the province being assessed as having a reduced standardized expenditure.

A further consideration to the full equalization process in the environment of a developing country is the realization that some provinces—the ones where infrastructure is less developed—will receive grants to provide recurrent services that cannot be provided because of the absence of the necessary infrastructure. Unless an adjustment or other arrangements are made, these provinces will be able to either raise less than their standardized level of taxation, or provide above-standard services in areas unaffected by their poor infrastructure development. For example, even if there are no school buildings, the provinces would be given funds to employ teachers. These funds could not be spent on teachers, and may be used for other purposes or to reduce taxation as the province sees fit.

Under these conditions, the central government might encourage the provinces to use this part of their equalization grant for infrastructure development to ensure longer-term equalization. Such use of recurrent resources for capital purposes would, along with any funds received for capital investment, enable a province to achieve a standard level of service more rapidly.

However, the simple formulas used in the equalization exercises in Germany are not based on detailed region-by-region evaluations of fiscal capacity as in Australia and Canada.

Such controls may spring from constitutional or legal restraints but often simply reflect a broader recognition by subnational governments of the need for such constraints.

Of course, the separate regional entities may still enjoy some of the benefits via some sort of confederation by participating in a free trade zone (for example, the European Union and the North American Free Trade Agreement), but those benefits may be less substantial than those of a full federation.

Even in Australia and Canada, it can be argued that the systems employed do not provide full equalization, since they cover only a selected portion of the overall state or provincial budgets.

The equity case for equalization has often been criticized as setting up a disincentive for labor migration out of poorer regions. In the presence of net fiscal benefit differentials, however, such migrations may go too far—so-called fiscally induced migration—and result in an inefficient allocation of labor. To prevent this, full equalization of net fiscal benefits is called for.

The argument is based on a model that distinguishes, on the one hand, between publicly provided private goods (such as health, education, and welfare expenditures, which are assumed to vary proportionately with the population of the region) and pure public goods (where significant economies of scale may exist) and, on the other hand, between source-based taxes and resident-based benefit taxes. Efficiency requires that residence-based taxes should be used to finance the provision of publicly provided private goods, while pure public goods should be financed by source-based taxes.

Revenue sharing could be determined on a tax-by-tax basis or some pooling arrangement. Under the first choice, a portion x of specific tax (for example, VAT), may accrue to the central government, and the residual (100 − x) to a subnational government in which the revenue was generated. A different share may be used for the enterprise income tax, and so on. Under the pool approach, all shared tax revenues (VAT, enterprise profits tax, and so on) are paid into one fund and then shared according to an established formula. If x percent of the pool is retained by the central government and the rest allocated on a derivation basis (that is, according to the region in which the revenue is generated), then this is equivalent to the x: (100 − x) ratio applied to each shared tax.


    AhmadEhtishamJonCraig andDubravkoMihaljek1995“Implementing and Managing Grants—Institutional and Informational Requirements,” in Reforming China’s Public Financesed. by EhtishamGaoQiang andVitoTanzi (Washington: International Monetary Fund).

    AhmadEhtisham andRaviThomas1997“Types of Transfers—A General Formulation,” in Financing Decentralized Expenditures: An International Comparison of Grantsed. by EhtishamAhmad (Cheltenham, England; Brookfield, Vermont: Edward Elgar).

    AhmadEhtishamed.1997Financing Decentralized Expenditures: An International Comparison of Grants (Cheltenham, England; Brookfield, Vermont: Edward Elgar).

    AhmadEhtishamGaoQiang andVitoTanzieds.1995Reforming China’s Public Finances (Washington: International Monetary Fund).

    AlbonR.P.1990“The Efficiency Implications of Locational-Based Factors in Commonwealth Grants Commission Determinations,”Appendix to the Victorian Submission to theInquiry into Issues of Methodology Victoria’s Response to the Discussion Papers Inquiry into Grants Commission Methodology.

    Australian Commonwealth Grants Commission1993Report on General Revenue Grant Relativities (Canberra: Australian Government Publishing Service).

    BoadwayRobin1985“Federal-Provincial Transfers in Canada: A Critical Review of the Existing Arrangement,” in M.KrasnickK.Norrie andR.Simeon (research cordinators) Federalism and the Economic Union: Royal Commission on the Economic Union and Development Prospects for CanadaVol. 65 (Ottawa: Canadian Government Publishing Centre).

    BoadwayRobin andFlattersFrank1982a“Efficiency and Equalization Payments in a Federal System of Government: A Synthesis and Extension of Recent Results,”Canadian Journal of EconomicsVol. 15 (November) pp. 61333.

    BoadwayRobin andFlattersFrank1982b“Equalization in a Federal State,”Economic Council of Canada.

    BoadwayRobin andHobsonPaul A.R.1993Intergovernmental Fiscal Relations in CanadaTax Paper No. 96 (Toronto: Canadian Tax Foundation).

    BrennanGeoffrey andJames M.Buchanan1980The Power to Tax: Analytic Foundations of a Fiscal Constitution (Cambridge, England: Cambridge University Press).

    BrosioGiorgio1992“The Balance Sheet of the Australian Federation: Some Tentative Estimates,”Working Paper No. 24 (Canberra: Federalism Research Centre, Australian National University).

    BuchananJames M.1950“Federalism and Fiscal Equity,”American Economic ReviewVol. 40 (September) pp. 58399.

    BuchananJames M. andCharles J.Goetz and1972“Efficiency Limits and Fiscal Mobility: An Assessment of the Tiebout Model,”Journal of Public EconomicsVol. 1.

    BuchananJames M. andRichard E.Wagner1970“An Efficiency Basis for Federal Fiscal Equalization,” in The Analysis of Public Outputed. by JuliusMargolis (New York: National Bureau of Economic Research).

    ClarkDouglas H.1997“The Fiscal Transfer System in Canada,” in Financing Decentralized Expenditures: An International Comparison of Grantsed. by EhtishamAhmad (Cheltenham, England; Brookfield, Vermont: Edward Elgar).

    Commission of the European Communities1997Report of the Study Group on the Role of Public Finance in European Integration (MacDougall Report) Vol. 1; General Report (Brussels: Office for Official Publications of the European Communities).

    Commonwealth Grants Commission1989Report on Issues in Fiscal EqualizationVols. I and II (Canberra: Australian Government Publishing Service).

    CornesRichard andToddSandler1986The Theory of Externalities, Public Goods, and Club Goods (New York: Cambridge University Press).

    CourcheneT.1970“Interprovincial Migration and Economic Adjustment,”Canadian Journal of EconomicsVol. 3 pp. 55076.

    CourcheneT.1984Equalization Payments: Past Present and Future (Toronto: Ontario Economic Council).

    Economic Council of Canada1982“Financing Confederation: Today and Tomorrow” (Ottawa: Supply and Services).

    FlattersFrankVernonHenderson andPeterMieszkowski1974“Public Goods, Efficiency and Regional Fiscal Equalization,”Journal of Public EconomicsVol. 3 (May) pp. 99112.

    HercowitzZvi andDavidPines1991“Migration with Fiscal Externalities,”Journal of Public EconomicsVol. 46 (November) pp. 16380.

    KingDavid N.1984Fiscal Tiers: The Economics of Multi-Level Government (London; Boston: Allen & Unwin).

    MathewsR.L.1975“Fiscal Equalization in Australia: The Methodology of the Grants Commission,”FinanzarchivVol. 34No. 1 pp. 6685.

    MathewsR.L.1982“Federalism in Retreat: The Abandonment of Tax Sharing and Fiscal Equalization” (Canberra: Centre for Federal Financial Relations, Australian National University).

    MathewsR.L. and others1981Australian Federalism (Canberra: Centre for Research on Federal Financial Relations, Australian National University).

    MathewsR.L.ed.1974Intergovernmental Relations in Australia (Sydney: Angus & Robertson).

    MaxwellJudith andCarolinePestieau1980Economic Realities of Contemporary Confederation (Montreal: CD. Howe Research Institute).

    MusgraveRichard A.Peggy B.Musgrave andRichard M.Bird1984Public Finance in Theory and Practice (Toronto: McGraw Hill).

    MyersGordon M.1990“Optimality, Free Mobility and the Regional Authority in a Federation,”Journal of Public EconomicsVol. 43 (October).

    NorrieK.H. andM.B.Percy1984“Province Building and Industrial Structure in a Small Open Economy,” in Economic Adjustment and Public Policy in Canada.ed. by Douglas D.Purvis (Kingston, Ontario: John Deutsch Institute for the Study of Economic Policy, Queens University).

    OatesWallace E.1972Fiscal Federalism (New York: Harcourt, Brace, Jovanovich).

    OatesWallace E.R.M.Schwab and1988“Economic Competition Among Jurisdictions: Efficiency Enhancing or Distortion Inducing?”Journal of Public EconomicsVol. 35 (April) pp. 33354.

    PetcheyJ.D.1992 1997“The Welfare Effects of Fiscal Equalization in a Federal Economy with Different Factor Endowments,”Working Paper (Canberra: Federalism Research Centre, Australian National University).

    RyeC. Richard andBobSearle1997“Expenditure Needs: Institutions and Data,” in Financing Decentralized Expenditures: An International Comparison of Grantsed. by EhtishamAhmad (Cheltenham, England; Brookfield, Vermont: Edward Elgar).

    ShapiroP. andJ.D.Petchey1992“Guns That Point East Can Point North and South: The Economics of Centralized Federalism,”working paper presented in a seminar at the University of AdelaideSeptember.

    SwanP. andG.Garvey1992“The Equity and Efficiency Implications of Fiscal Equalization,”draft working paperAustralian Graduate School of Management, University of New South Wales.

    TanziVito1995“Basic Issues of Decentralization and Tax Assignment,” in Reforming China’s Public Financesed. by EhtishamAhmadGaoQiang andVitoTanzi (Washington: International Monetary Fund).

    TreschRichardW.Public Finance: A Normative Theory (Piano, Texas: Business Publications).

    TieboutC.M.1956“A Pure Theory of Local Expenditures,”Journal of Political EconomyVol. 64 (October) pp. 41624.

    United States Advisory Commission on Intergovernmental Fiscal Relations1995“Measures of State and Local Fiscal Capacity” (Washington).

    WalshCliff andJeffrey D.Petchey1992“Fiscal Federalism: An Overview of Issues and a Discussion of Their Relevance to the European Community,”Discussion Paper No. 12 (Canherra: Federalism Research Centre, Australian National University).

    WatsonWilliam G.1986“An Estimate of the Welfare Gain from Fiscal Equalization,”Canadian Journal of EconomicsVol. 19 (May) pp. 298308.

    WinerS. andD.Gauthier1982Internal Migration and Fiscal Structure: An Econometric Study of the Determinants of Interprovincial Migration in Canada (Ottawa: Economic Council of Canada).

    Other Resources Citing This Publication