local government

b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 1

Chapter I

Promise and Reality of Fiscal
Decentralization
ROY BAHL
Georgia State University

The growing interest in fiscal decentralization in Asia is part of a worldwide pattern. Most developing countries have long ago placed the strengthening of sub-national government on their development policy agenda. Decentralization is an important part of the transition strategy for many former socialist countries. Some industrialized countries are retuning their fiscal decentralization (Canada) while others have introduced major new initiatives in recent years (Japan,
Spain).
In this paper, we consider the costs and benefits of fiscal decentralization, trends in the practice of decentralizing budgets, and the policy interventions that are being used to strengthen the financial position and the financial autonomy of sub-national governments. In particular, we examine inter-governmental transfers, local taxation and hard budget constraints as key policy issues. The focus in this paper is on developing countries. The process areas that are taken up have to do with comprehensive design of a decentralization reform, sequencing and politics. In outlining these issues, we draw heavily from the experiences of Asian countries, some of which are discussed in detail in the chapters in this volume.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html 1

b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 2

2 ✦ Decentralization Policies in Asian Development

1. The Benefits1
What are the major advantages to be gained from investing more fiscal powers in provincial and local governments? This is the fundamental question that sometimes gets lost in the political debate about fiscal decentralization. The most important benefit is the welfare gain that comes from moving governance closer to the people. This is the economic efficiency argument that drives the thinking of most economists who work on this subject (Oates, 1972). The argument is straightforward. Let us assume that people’s preferences for government services vary, e.g., because of religion, language, ethnic mix, climate, economic base or just because of the inclinations of the local political leadership.
Let us assume further that people have sorted themselves so that those with similar preferences live in the same region. If sub-national governments respond to these preferences in structuring their budgets, decentralization will result in variations in the package of services delivered in different regions. The voters will see to this in a system where there is downward accountability. People will get what they want and so the welfare of the population will be enhanced. Under the same circumstances, but with a centralized system, accountability will be upward to a higher level of government, service provision will be more uniform and people in different regions will get less of the service mix that they want. The more heterogeneous the country, the greater will be the welfare costs of uniformity.
If governance is properly decentralized, two good things can come to pass. One is that there will be more accountability on the part of government officials because they are held responsible for the quality of services delivered to the local population that elected them. If the makeup of local public services is determined by the center and paid for by the center, sub-national government officials can largely escape taking responsibility for the quality of services provided. The other benefit from a well-functioning decentralization is that there will be more willingness on the part of the local population to pay for services, because they get what they want. If one advocates fiscal decentralization, one must believe this story, because it is the primary argument.

1

For more lengthy discussions of the pros and cons of fiscal decentralization, see Bahl and
Linn (1992), Bahl and Wallace (2005), Litvak, Ahmad and Bird (1998), Bird and Vaillancourt
(1998), Tanzi (1995) and Dillinger and Webb (1999).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 3

Promise and Reality of Fiscal Decentralization ✦ 3

True believers can point out that successful fiscal decentralization at once attacks several of the problems that face most developing and many transition countries: economic development, revenue mobilization, innovation in public service delivery, accountability of elected officials, capacity development at the local government level, and grassroots participation in governance.
Whether or not fiscal decentralization actually leads to accountability downward, however, depends on many things. Certainly there must be elections. Local governments must have the power to control their employees, and there must be enough accurate information available so that voters can evaluate the fiscal decisions of their local governments. In many developing countries, one or more of these conditions is not met.
A second important benefit of fiscal decentralization is the promise of increased revenue mobilization. This happens because decentralization can broaden the overall tax base. That is, if sub-national governments are more directly involved in taxation, a greater share of GDP might be reached by the tax system. If this hypothesis is correct, increases in sub-national government tax revenues would not be offset by equal amount reductions in central government tax revenues. In addition, the claim of sub-national governments on central revenues via inter-governmental transfers could be reduced by increased revenue mobilization at the sub-national government level.
The argument behind this hypothesis is that sub-national governments have the potential to reach the traditional income, consumption and wealth tax bases in ways that the central government cannot.
Typically, central governments rely on a combination of company income tax, individual income tax, value added tax, excise and customs duties. All of these taxes, however, have a high entry threshold in most developing countries. Small firms, most individuals and owners of immovable property are “under-represented” in the tax base. In fact, local governments can broaden the overall tax base with a variety of tax instruments and administrative measures, and they do so in many countries. These instruments include payroll taxes, levies on the sales of assets of firms, licenses to operate, betterment charges and various forms of property taxation.2 As we discuss below, few Asian countries have acted on this comparative advantage in local taxation.
2

A good review of local tax practices in developing countries is in Bird (1999).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 4

4 ✦ Decentralization Policies in Asian Development

2. The Costs
There also are significant costs to fiscal decentralization, at least in the eyes of central governments, and perhaps this is why all countries do not embrace this policy route with the same degree of enthusiasm. Heading the list of costs is the loss of macro-economic control.3 Central governments would argue that they should have the flexibility to respond quickly to changes in national economic conditions, for example, to raise taxes, cut expenditures or limit credit in order to deal with a deficit.
There also are external pressures on developing country governments to implement an effective stabilization policy. The pressures from the IMF and World Bank for more austere economic policy to bring about internal or external balance usually requires holding the fiscal deficit to an acceptable level and limiting the level of domestic credit.
In a truly decentralized governance system, stabilization policy is more difficult to implement than in a centralized system. One reason is because the central government may not be able to control sub-national government spending so as to reduce an overall deficit. If the lower levels of government have their own resources — either significant own source revenues or a guaranteed share of national government collections — they may set their own level of spending. In many decentralized countries, sub-national governments do not face a hard budget constraint, and a more serious challenge to macroeconomic stability can emerge. In India, the deficits of the state and local governments rose to
10 percent of GDP in the late 1990s, and in Japan local government debt outstanding increased to a level equivalent to about 40 percent of
GDP in 2003. (Rao, 2008; Ikawa, 2008). In the Philippines, the growing share of national tax collections earmarked for local governments compromised the budgetary position of the central government (Diokno,
2008).
Even when the central government does not tie the vertical share of grants to national tax collections, it may not have the flexibility to cut subsidies (conditional grants) to local governments. These subsidies can be sticky downward, especially in cases where they support public employee salaries or direct payment to needy individuals, and where their champions include powerful central government ministries.
3

More detailed discussions of this may be found in Bahl and Linn (1992), Prud’homme
(1995), Ter-Minassian (1997), Tanzi (1995) and Spahn (1997).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 5

Promise and Reality of Fiscal Decentralization ✦ 5

This “cost” of decentralization to the central government, however, may not be seen the same way when viewed from the provincial or local level. Where central governments are able to control the budgets of subnational governments, for example because the budget law does not clearly define expenditure assignments, central deficits have sometimes been offloaded on to regional and local governments. For example in the
1990s, Russia transferred expenditure responsibility for certain infrastructure and social welfare services to regional governments without assigning additional revenue responsibility (Martinez-Vazquez et al.,
2006).
A second cost of decentralization to the central government is that it could lose some control over infrastructure development in cases where sub-national governments have discretionary spending power.
The net result of fiscal decentralization will be a shift of resources from central governments that have higher rates of capital spending to provincial and local governments who spend at a greater rate on consumption goods and services. Fiscal decentralization, therefore, could lead to a lower overall rate of spending on infrastructure, and national growth could be slowed. Fiscal decentralization also may lead to a shift in the composition of public capital investments. This is because national priorities for capital investment are not the same as sub-national government priorities. The national government is interested in investments in infrastructure that have regional and national benefits, for example, large scale irrigation projects, national (interstate) roads, and power. State (provincial) and local governments will be focused more on capital investments with provincial and local benefits. Central governments have tried to resolve this problem in a number of ways. The most used remedy is to provide local governments with conditional grants that require expenditures in certain functional areas (e.g., local infrastructure) and often carry conditions as to the nature of the expenditures (e.g., construction standards, pay scales etc.). From the point of view of the sub-national governments, however, this approach compromises their discretion and leads to sub optimal uses of resources because of the failure to recognize local conditions. In some countries there is movement away from conditional grants.
Indonesia replaced its Inpres, a large number of specific grants for purposes ranging from re-greening to the construction of public markets,
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 6

6 ✦ Decentralization Policies in Asian Development

with a general purpose grant allocated by formula (Fengler and Hofman,
2008). Japan is moving away from conditional subsides to local taxes
(Saito and Yunoue, 2008).
A third cost of fiscal decentralization is that it is not an inherently equalizing economic development policy. In fact, depending on how the system is designed, there may be little in it for poorer and rural provincial and local governments. Certainly revenue centralization gives a greater potential for equalization. In countries where the claim of subnational governments on the overall tax base is small, the central government can create a larger pool of funds for allocation among local governments on an equalizing basis. However, just because the central government has more funds to allocate, it does not necessarily follow that they will allocate these funds on an equalizing basis. In fact, countries vary widely in terms of the amount of equalization that is done through their grant systems.
If fiscal decentralization takes the path of heavy reliance on own source revenues, a decided advantage is given to sub-national governments with a greater fiscal capacity. This would include, for example, industrial provinces and large cities where there is a larger tax base that is easier to reach and a better chance of developing the administrative capacity to collect taxes. This part of the equation alone suggests less equalization in the system. The way to address this, in a context of fiscal decentralization, is with a program of equalization grants that takes into account the weaker fiscal capacity of some regions/local governments. An example of this approach is South Africa, where the large cities raise over 90 percent of revenues from own sources and rural local governments are supported by an equalizing grant program
(Reschovsky, 2003).
Finally, fiscal decentralization can be costly relative to the benefits gained. Sub-national governments may not have the administrative skills to deliver decentralized services efficiently or to collect taxes efficiently. In both areas, there may be duplication of services with the central government. If local governments attempt to improve their tax collections or service delivery efficiency to central government levels, the cost could be prohibitive. This comparative advantage is what leads many to argue to keep collection of major taxes at the central government level. This argument is reinforced by the observation that subnational governments are often assigned “difficult” taxes, i.e., those that would be costly to collect at any level of government. Compounding
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 7

Promise and Reality of Fiscal Decentralization ✦ 7

these cost problems, according to some, is that decentralization can lead to increased corruption.4
Advocates would argue that fiscal decentralization need not be a more costly system to administer. Part of the tax administration costs could be reduced if tax assignment is more reasonable and if tax base sharing were adopted (central administration but local rate setting)
(McLure, 1997). On the service delivery side, it is true that the learning curve for sub-national governments can be steep, even for functions that are properly assigned to lower-level governments, but eventually the responsiveness to the demands of citizens can outweigh these costs.

3. The Practice
One might test the hypothesis that the benefits of decentralization outweigh the costs, by looking for evidence on the growing fiscal importance of sub-national governments. In fact, countries around the world have moved only slowly toward the adoption of more decentralized inter-governmental fiscal systems. Based on IMF Government Finance
Statistics, which is about the only comparable data source available, one can observe that the sub-national government share of public expenditures has remained at about 13 to 14 percent in developing countries over the last three decades.5 The rather remarkable stability in the subnational government expenditure share (which also holds in OECD countries) is reported in Table 1.
These averages hide a great deal of intercountry variation, which several analysts have tried to explain. Bahl and Wallace (2005) find that the sub-national government expenditure share is significantly higher in countries with a higher per capita GDP a larger population size and
,
4

Empirical work on the relationship between decentralization and corruption is inconclusive. Fisman and Gatti (2002), for example, find that corruption is lower in more decentralized countries; Treisman (2000) finds corruption to be higher in federal than in unitary countries.
5
We measure decentralization here as the sub-national government share of total government expenditure in the country, i.e., sub-national government expenditures as the numerator, and total central plus sub-national government expenditures as the denominator. This is a flawed measure of fiscal decentralization because it does not indicate whether the sub-national government has any significant influence over how the money will be spent.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html 1980s

1990s

Developing
Countries

OECD
Countries

Developing
Countries

OECD
Countries

Transition
Countries

10.44
(43)

18.71
(22)

7.70
(35)

18.75
(22)

9.27
(28)

19.13
(20)

16.59
(14)

13.01
(48)

33.78
(22)

13.24
(43)

32.27
(23)

13.78
(54)

32.41
(23)

26.12
(23)

Note: Sample sizes are in parenthesis.
Source: Government Finance Statistics Yearbook. International Monetary Fund.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html Page 8

Sub-national Government
Tax as a share of Total
Government Tax
Sub-national Government
Expenditure as a share of Total Government
Expenditure

OECD
Countries

5:52 PM

Developing
Countries

11/24/2008

1970s

b675_Chapter-I.qxd

Fiscal Decentralization Indicators

8 ✦ Decentralization Policies in Asian Development

Table 1.

b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 9

Promise and Reality of Fiscal Decentralization ✦ 9

a lower degree of corruption. This more or less matches the findings in other studies.6

4. Policy Instruments for Decentralization
There are many fiscal policy instruments used in implementing decentralization systems. Three of the most important, and controversial, are sub-national government taxes, inter-governmental transfers, and the requirement of a hard budget constraint.

4.1. Sub-national Government Taxes
Voters will hold their elected officials more accountable if local public services are financed to a significant extent from locally imposed taxes and charges, as opposed to the case where financing is primarily by central government transfers. The local tax must be visible to local voters, large enough to impose a noticeable burden, and the burden must not be easily exported to residents outside the jurisdiction.7 Minor taxes and nuisance taxes will not measure up to all of these requirements.
Reliance on own source taxes also has the advantage of imposing fiscal discipline on the sub-national government. A greater share of financing from own sources drives up the tax price of public services and reduces the upward pressure on sub-national government expenditures. Reliance on inter-governmental transfers has the opposite effect.
Significant tax assignment to sub-national governments is common practice in many industrial countries. At one extreme, US State governments and Canadian provinces have almost complete autonomy in choosing any tax base, so long as there is no interference with interstate commerce. In Denmark and Sweden, local taxes account for nearly onehalf of local government spending (Owens and Norregaard, 1991).
Revenues from sub-national government taxes in Switzerland are greater in amount than revenues received from grants. Japan has lagged other industrialized countries in the assignment of taxes to local governments but is now introducing a new inter-governmental reform that shifts local government finance significantly toward local taxation
(Ikawa, 2008).
6
7

For a good literature review, see Letelier (2005).
A tax may be considered “local” if the sub-national government sets the tax rate.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 10

10 ✦ Decentralization Policies in Asian Development

In most developing and transition countries, however, central governments have been reluctant to release taxing powers to sub-national governments. As may be seen in Table 1, the sub-national governments tax share in developing and transition countries is about 10 percent, in comparison to 20 percent in industrialized countries. Public financing in low income countries is still primarily through the transfer system.
One common reason for a low level of taxation by sub-national governments in Asia is the unwillingness of central governments to share the productive tax bases. In many cases the taxes passed down are notoriously difficult to administer. For example, Pakistan gives its provincial governments the right to tax agricultural income, the consumption of services and property transfers. The result is a level of sub-national government tax revenue that is less than one percent of GDP (Bahl et al., 2008).
The Philippines is a country where about one-third of sub-national government revenues are raised from own source revenues. This is a high share by comparison with most developing countries. In addition to the property tax, the principal revenue source is a business tax on gross receipts. However, Philippine local governments are limited in terms of their ability to set tax rates, and tax revenues fall well short of the level of devolved expenditures (Manasan, 2008). Sub-national governments in Cambodia, China and Vietnam all raise less than 5 percent of total revenues from own sources. (Taliercio, 2005).

4.2. Inter-governmental Transfers
No policy instrument of fiscal decentralization is more discussed, or more often reformed, than inter-governmental transfers. On the one hand, transfers are preferred to local taxation in most developing and transition countries because they recognize the superior tax administration capabilities of the central government and because they allow a greater measure of central control. On the other hand, transfers are an object of criticism precisely because of the central control, and in some countries because they discourage local revenue mobilization, are not adequately equalizing, and are not transparent. Reform in most countries has concentrated on finding a middle ground between central control and the goals of decentralization.
In fact, there are many different kinds of inter-governmental transfer systems in use, and each has a potentially different impact on sub-national
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 11

Promise and Reality of Fiscal Decentralization ✦ 11

government finances and a different implication for central control.
Some stimulate local government spending, some are substituted for local government revenue effort, some are equalizing, and some lead to more local government fiscal autonomy than do others. The policy mistake most often made, and perhaps the reason why grant systems remain under review, is that countries often enter into grant design without fully exploring the alternatives available and their differential impacts. One might begin such an analysis by considering that intergovernmental transfers have two dimensions: the size of the divisible pool, and the distribution of this pool among eligible local government units. Some have referred to the divisible pool dimension as having to do with the vertical fiscal balance between the central and subnational governments, and the allocation dimension as having to do with horizontal fiscal balance. Both dimensions must be part of the policy design.8
We present a taxonomy of inter-governmental transfers in Table 2, with the vertical sharing arrangements shown in the column heads and the distribution arrangements shown in the rows. The cells in the table show the different grant types. To demonstrate how different the outcomes of these systems can be, we might focus on the Asian experience with three grant types: unconditional transfers to sub-national governments, conditional grants and ad hoc transfers.
Unconditional grants are probably most consistent with the local autonomy goals of decentralization. If they are based on defined shares of national taxes (see Column 1 of Table 2), they give sub-national governments access to an elastic tax base as well as relative certainty about the amount they will receive. Asian countries vary widely in terms of their vertical sharing arrangements, but the shared tax option — which would tend to be favored by countries with a greater commitment to decentralization — is common. The Philippines, India,
Indonesia and Pakistan all guarantee a percent of national taxes to sub-national governments, but the vertical shares differ. Indonesia shares
25 percent of actual collections, and India allocates about 30 percent of tax collections to the divisible pool for Finance Commission transfers.
The Philippines shares 40 percent of internal revenue collections in the third previous year, and Pakistan shares 41 percent of total tax collections. 8

Bahl and Linn (1992, Chapter 13).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 12

12 ✦ Decentralization Policies in Asian Development

Table 2.

How Should the Grant System be Structured?

Alternative Form of Inter-governmental Grant Programs
Method of Determining the Total Divisible Pool
Method of Allocating the Divisible Pool
Among Eligible
Units

Specified Share of National or
State
Government Tax

Ad Hoc
Decision

Reimbursement of Approved
Expenditures

A

NA

NA

B
C

F
G

NA
K

D

H

NA

Origin of collection of tax
Formula
Total or partial reimbursement of costs
Ad hoc
Source: Bahl and Linn (1992).

With respect to the distribution across eligible local government units, the revenue sharing method may be by origin of collections
(China’s income tax or VAT sharing) or by formula (as is done in
Japan, Taiwan and Philippines). A derivation-based distribution (an
“A” grant in Table 2) favors higher income provinces, where taxable capacity is greatest. Formula-based systems (B grants) can vary greatly in their impacts, depending on the formula used. In fact, formulas usually are driven by what a country wants to achieve with its grant system but are constrained mightily by politics and by “expediency”. This is because data on provincial and local governments are so limited that formulas are driven more by what is available than by the objectives of the grant program. The following are some examples of formulas now in use in Asia.
• In the Philippines, the distribution formula for the IRA grant is based on population size, land area and equal shares. It is transparent but has no explicit element that rewards local revenue mobilization.9
9

Diokno (2008) argues that the increased vertical share of the IRA grant has dampened local government revenue mobilization.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 13

Promise and Reality of Fiscal Decentralization ✦ 13

The equal shares allocation, which favors smaller governments, reflects an equalization concern.
• Pakistan distributes its National Finance Commission awards based on population size. This is transparent, and a compromise solution that has been accepted by the four provinces. This formula recognizes expenditure needs in a broad way, but does not recognize taxable capacity, nor does it provide an incentive for the four provinces to increase tax effort (Bahl et al., 2008).
• Thailand uses a formula based on population size, land area, equal shares and revenues raised. The inclusion of revenues raised (inverse) effectively penalizes increased tax effort. The equal shares element in the formula arguably adds an equalization component.
• The grant component of Finance Commission transfers in India is distributed according to a formula that includes population, land area, the inverse of income level, and measures of both expenditure needs and tax efforts (Rao, 2008). While the system is equalizing, the weight attached to tax effort probably is not large enough to stimulate increased revenue mobilization.
Many have argued that a proper equalization grant would recognize both expenditure needs and fiscal capacity. Japan’s local allocation tax grant attempts to do this, by distributing grants according to the difference between “standard expenditure” and “standard revenues”.
Taiwan more or less follows the same system as Japan. A similar approach is taken in Vietnam, which is judged to result in a strong equalization effect (Hofman and Guerra, 2005). Indonesia distributes grants with a formula that includes measures of both revenue capacity and expenditure needs. None of these programs, however, provides an incentive for increases in revenue mobilization by sub-national governments. The systems in Japan and Vietnam would appear to provide a disincentive. Are inter-governmental transfers in Asia equalizing? If the criteria are whether per capita expenditure disparities are reduced by transfers,
Hofman and Guerra (2005) conclude that there is some degree of equalization in the systems in China, the Philippines, Thailand and Indonesia, and a stronger equalization in Vietnam. Bahl, Wallace and Cyan (2008) find a slight degree of equalization in Pakistan, using the same criteria.
Rao (2008) concludes that India’s finance commission transfers reduce the disparity between higher and lower income states. Saito and Yunoue
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 14

14 ✦ Decentralization Policies in Asian Development

(2008) find evidence of equalization in the Japanese transfer system.
Tokyo, for example, receives a near zero amount from the local allocation tax transfer.
Another form of grant on which countries rely heavily is conditional grants i.e., grants that permit the central government to control the direction of spending, presumably toward functions where there are positive externalities and/or income distribution benefits. The vertical shares are most often determined on an ad hoc basis by the central government, or are based on estimated costs of providing the subsidized service (Columns 2 and 3 of Table 2). The similarity ends there. The distribution of conditional grants among eligible recipients can be done in many ways, though most commonly it is a reimbursement of some standard cost for a prescribed function, for example, teacher’s salaries or payments to indigent families (G or K grants in Table 2).
Conditional grants usually are successful in diverting sub-national government spending toward the target functions. If the target functions lead to higher expenditures that generate spillover benefits, the conditional grants programs can be welfare enhancing. But, conditional grants are roundly criticized by sub-national governments and even by some central governments. Such grants are nationally uniform and do not give adequate consideration to local conditions, and they cause subnational government spending patterns to deviate from local preferences. And, because both the line ministries and the local government ministry lobby for these subsidies, they can contribute significantly to expenditure growth at the local government level.
Asian countries also rely on ad hoc distributions of intergovernmental transfers. These are grants where either the vertical share or the horizontal share is determined by political considerations
(D or H grants in Table 2). From the point of view of parliament, this approach offers a great deal of flexibility (though there is a tendency toward “incremental-ism”), but from a point of view of good public policy this approach is not transparent and lacks objectivity. Nevertheless, there is some degree of ad-hoc-ism in virtually every system of intergovernmental transfers. In Asia, both China and Thailand allocate their grants partially on an ad hoc basis, and this creates budget uncertainty at the local level (Charas and Weist, 2008).
A particularly difficult issue is the coordination of inter-governmental transfer policy among various Ministries and agencies. For example, in
India, transfers are awarded by all of the national Finance Commission,
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 15

Promise and Reality of Fiscal Decentralization ✦ 15

the Planning Commission and the line ministries (Rao, 2008). All have different objectives and all use different allocation methods. Under this arrangement, some offsetting effects would be expected.

4.3. Hard Budget Constraints
Just as important as taxes or grants as instruments for implementing fiscal decentralization is the requirement of a hard budget constraint for sub-national governments. A hard budget constraint means that sub-national governments must balance their budgets without recourse to any end-of-year (ex post) assistance from the central government.
Borrowing can co-exist with a hard budget constraint regime for sub-national governments, but only if the borrowing is used to finance long-lived capital assets, and only if it is understood that central government bailouts on unpaid debt obligations will not occur.
Sub-national governments must have reason to believe that they are
“on their own” in paying for their fiscal decisions. Otherwise, they will game the system by over-borrowing or under-taxing, in order to shift the burden of financing local services on to the center.
An issue here is the definition of a current budget deficit for a subnational government. One working definition is that budget balance exists when all recurrent expenditures are covered by current revenues, with the latter defined to include only tax and non tax revenues of the sub-national government and “regular” inter-governmental transfers.
In the case of developing countries, this measurement is not as easy as it might seem. Accounting systems are often not uniform, and accounting practices may be haphazard, especially for smaller local governments (Sethi, 2004). There also may be problems with the flow of expenditures during a fiscal year and cash balances often play a major role in expenditure policy. Finally, central governments do not always disburse grant entitlements in full and on time. So, a relatively simple concept, as the hard budget constraint seems to be, may not be so simple when it comes to implementation. As Rodden, Eskeland, and Litvack
(2003, p. 4) point out, the question may be less whether the budget constraint is hard or soft, but whether there are weak spots that soften the budget constraint to a point where it alters the expectations and behavior of sub-national governments.
Soft budget constraints may be built into inter-governmental fiscal policy, i.e., sub-national government deficits may be an expected part of
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 16

16 ✦ Decentralization Policies in Asian Development

the system. Some central governments hold to a paternalistic approach to inter-governmental fiscal relations. In these cases, the policy frame implicitly allows for a soft budget constraint. The underlying problem may be an over-assignment of expenditure responsibility relative to revenue access, the failure of the central governments to pay full grant entitlements to sub-national governments, unfunded mandates, or an unconstrained access to credit. Under this scenario, the sub-national government begins the fiscal year with a projected imbalance between expenditures that it plans to make and revenues it expects to receive.
A year-end budget deficit is, in effect, planned, and various forms of deficit grants are a guarantee that local governments may come to depend on. Enemies of the hard budget constraint rule include central government measures such as the following:
• Deficit grants, i.e., year-end grants to cover revenue shortfalls;
• Bailouts on delinquent debt; and
• Direct coverage of year-end shortfalls on certain items of expenditure.
Japan is a case where soft budget constraints on local governments have resulted from open-ended grants from the center, little local government taxing power, access to credit by local governments, and successful political lobbying. The predictable result is that total central government budget revenues have come to be viewed as “a common pool for local governments” (Ihori, 2008, p. 2).

5. Design must be Comprehensive
The crucial first step toward implementing a fiscal decentralization system is to have a plan or blueprint that spells out the strategy and the objectives that are to be achieved. It is crucial that this design address all of the major issues relevant to gaining a sustainable fiscal decentralization. In this connection, inter-governmental fiscal relations must be thought of as a system, and all the pieces in this system must fit together.10 Implementation should begin with the design of a
10
Inter-governmental fiscal relations is a term that refers generally to the division of fiscal powers and responsibilities among levels of government. Fiscal decentralization refers to an inter-governmental system where the balance of power moves more toward the sub-national government sector than has been the case.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 17

Promise and Reality of Fiscal Decentralization ✦ 17

comprehensive system, and should lay out the plan for each element of the system.
A little reflection will lead one quickly to the conclusion that fiscal decentralization involves a lot more than fiscal matters. In fact, the electoral system, the civil service arrangements and the public administration in general are arguably as important as the taxing and spending components.11 Other key pieces are sub-national government revenue raising power, borrowing powers, expenditure assignment, and budgetary discretion. A “one-off” piecemeal reform, encompassing only one element of the system (e.g., revenue sharing), is not likely to lead to success.
Consider the following four examples which are suggestive of the complications that arise when important inter-governmental issues are not part of the design. Indonesia’s big-bang decentralizations of the
2000s did consider both expenditure assignment and revenue assignment, but the planning was done by two different ministries with little coordination (Alm, Aten and Bahl, 2001). While the expenditure decentralization transferred nearly 40 percent of the public expenditure budget and 2 million employees to the local governments, little was done by way of assignment of taxing powers to the local governments.
The result is very much an expenditure decentralization. The
Indonesian decentralization can be rightly thought of as a very successful decentralization that was comprehensive (Fengler and Hofman,
2008), but the failure to address the issue of sub-national government taxation has led to a challenge that is yet to be resolved. A similar imbalance was realized in the Philippines (Manasan, 2008), and in
Thailand (Charas and Weist, 2008).
In India, the National Finance Commission allocates inter-governmental transfers and shared taxes among the states, to try and achieve some budgetary balance. However, a national pay commission sets central government wage levels that roll out to the states and in recent years has dramatically compromised state budget balance. The lesson here is that in cases such as India, the determination of public sector wage rates and employment levels must be part of the decentralization design.
A comprehensive “trinity” reform in Japan is simultaneously considering revision of conditional grants, the general purpose grant and
11

For a good discussion of the need for public administration reform as a component of decentralization, see the discussion of the Vietnam case in Hung (2008).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 18

18 ✦ Decentralization Policies in Asian Development

the assignment of taxing powers to local governments (Ikawa, 2008;
Ihori, 2008). The intent is to design a revenue neutral reform that would impose a hard budget constraint on local governments. While the three major elements of local government finance are part of the package, some analysts have argued that a borrowing strategy should also have been included to insure achieving the hard budget constraint goal.
Finally, there is the interesting case of China, where responsibility for financing social welfare and public pensions has been assigned to local governments. This component of the fiscal decentralization, however, has given local governments an incentive to limit the migration of workers from rural areas (Zuo, 2008). Such actions may not be consistent with China’s interest in increasing domestic demand for its consumer products.

6. Sequencing Fiscal Decentralization12
The success or failure of fiscal decentralization in developing countries depends on implementation as well as on program design. In particular, when introducing decentralization policies and administration, sequencing is a key to the transition to this different approach to governance.
Bahl and Martinez-Vazquez (2006) argue that there is an optimal pattern of sequencing. Even before beginning the implementation, two prerequisites are important for success: a rule of law and an existing de-concentration of public service delivery. The former makes it possible for sub-national governments to protest violations of the decentralization law, even those committed by the central government. The
Supreme Court in the Philippines ruled that the central government could not hold back on the entitlements of local governments to IRA transfers (Diokno, 2008). The latter makes it possible to shift central government employees to local status without having to train a new force of local public employees. It is arguably true that an existing deconcentration was a primary reason why Indonesia’s big-bang expenditure decentralization was accomplished without disruptions in service levels. The sequencing of decentralization should begin with two important steps (Fig. 1). The first is to hold a national debate about decentralization,
12

This section draws heavily from Bahl (forthcoming) and from Bahl and MartinezVazquez (2006).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 19

Promise and Reality of Fiscal Decentralization ✦ 19

Step 6: Monitor, Evaluate, and Retrofit
Step 5: Implement the Decentralization Program
Step 4: Develop the Implementing Regulations

Step 3: Pass the Decentralization Law

Step 2: Do the Policy Design and Develop a White Paper

Step 1: Carry out a National Debate on the Issues Related to Decentralization Policy

The Platform: Deconcentration, Rule of Law, etc.
Source: Bahl and Martinez-Vazquez (2006)

Fig. 1.

Sequencing Fiscal Decentralization: A Normative Approach

possibly in the context of an election or the report of a national commission. The national discussion should culminate in a policy paper on fiscal decentralization that lays out the goals of the program and the strategy for achieving it. These steps make up the road map for the decentralization program. The policy paper would include matters such as the assignment of expenditure responsibilities, the nature of the inter-governmental transfer system, and the revenue-raising powers of sub-national governments. The White paper cannot be limited to policy.
It must address some key issues related to the public administration in the country, such as whether local government administrations will be subordinate to their elected leadership or to higher level line ministries.
Vietnam is currently grappling with this difficult issue of dual subordination (Hung, 2008). Going forward without a policy and general administrative guideline would be tantamount to adopting a “make it up as we go” strategy.
Based on this road map, the decentralization law can be written.
This key document of the program should guide all else that is done in the implementation process. Although the law must contain the key
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 20

20 ✦ Decentralization Policies in Asian Development

features of the program, it must not be too specific because legal drafting cannot accommodate all the realities of administration that will arise, nor can it anticipate changes in the country that will come with development. This process of gradual reformulation underlines the reason why it is dangerous to include too much detail about decentralization in the constitution. Charas and Wiest (2008) discuss the need for two changes in the Thai constitution within a decade.
The administrative phase begins with developing the implementing regulations to accompany the decentralization law. For example, the law may specify an equalizing grant program, but the implementing regulations will specify the exact formula to be used in distribution. The implementing regulations must conform to the decentralization law. If there are no implementing regulations, or if they are not clearly written, policy making could implicitly fall into the hands of bureaucrats.
A final step in the process is to provide for monitoring and evaluation. Fiscal decentralization programs emerge and change over time, sometimes due to poorly formulated policy and sometimes due to the changing needs of the country. It is important to have in place a process for fine-tuning the structure, and in general for measuring the success of the program. There is a mixed record on this count. Indonesia has put in a good data base to track local government finances, but in the
Philippines — 10 years after enactment of its local government code — no monitoring system is in place (Diokno, 2008).
One problem with the stepwise approach to implementing decentralization in the way proposed here is that it makes the process very transparent and vulnerable to criticism. It also requires time, especially at the stage when the program is being formulated. Advocates will argue that this gives the opposition enough time to organize their objections and their constituency. Their approach would be to push ahead before opponents can get organized and to get the law written and adopted. The hope in this strategy, as was the case in the big-bang decentralization in Indonesia, is that once the law is written, there will be no turning back. Democrats, on the other hand, will argue that policies that touch everyone, as decentralization clearly is, should be debated widely before they are adopted.
Some will point out that the kind of approach outlined here is unrealistic, and ignores the politics of getting a fiscal decentralization in place. The main thing, it might be argued, is to “get on with it” while there is an opportunity. Fengler and Hofman, 2008 describe the highly
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 21

Promise and Reality of Fiscal Decentralization ✦ 21

successful Indonesian experience, which started with a drafting of the law, even before a policy design was in place.13 On the other hand, the lack of clarity in expenditure assignment and inadequate tax assignments remain as important constraints to the Indonesian fiscal decentralization. As White and Smoke have observed in tracking the success of decentralization in Asia, a key message is the need for focused attention on policy design and legislative development (White and Smoke,
2005, p. 20).

7. The Politics of Fiscal Decentralization
Fiscal decentralization is about shifting the balance of power, both between levels of government and between ministries in the central government. It also has to do with shifting political risks, e.g., decentralization of taxing powers and hard budget constraints can put subnational government politicians in a more difficult position in terms of their accountability to local voters. Financing sub-national government budgets with inter-governmental transfers is an easier way to go for local politicians and is a concession that many central governments are all too willing to make. The result of all this is that fiscal decentralization has relatively few strong champions. This may be why, to date, fiscal decentralization has been much more rhetoric than action.
Decentralization should be a grassroots movement where voters and elected politicians, including the President, will be the natural champions. Voters see themselves gaining control over their governance, and Presidents count voters. It is a natural alliance. However, if decentralization conflicts with macroeconomic stabilization policy, as was the case in the Philippines in the 1990’s, the President will advocate pulling back transfers to local governments (Diokno, 2008). Or, if the President is too closely aligned with the line ministries, his/her support for fiscal decentralization will be less firm. There also may be a concern that decentralization could strengthen the hand of governors who might oppose Presidential policies or who might become more formidable political rivals. Some have argued that the transfer of the Taiwan
Provincial government to the central administration was a concern with the vote-getting power of the local governor (Shih, 2008).
13

For an interesting summary of how the decentralization policy in Indonesia developed, see Rasyid (2004).

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 22

22 ✦ Decentralization Policies in Asian Development

Parliament (Congress) often will embrace programs that voters embrace, and therefore is a potential champion of decentralization.
However, members of Congress usually are most interested in how programs benefit their own constituency; hence they will be less enthusiastic than policy analysts about the need for transparency and for programs that are designed in the national vs. the local interest.
Especially the more powerful members of Parliament might even favor ad hoc grants in hopes that they might direct more resources to their home districts. In the aftermath of the big-bang decentralization in
Indonesia, Congress was fervent about overriding the results of the new grant formula to hold the grant revenues of each local government harmless at pre-reform levels. Manasan (2008) reported that this reluctance of congressmen toward decentralization in the Philippines was lessened by fraternal relations with local government politicians. She points out that it is not uncommon “to find sites (or provinces) where the mayor (or governor) is the congressman’s wife (or brother/sister/father/son).”
The Ministry of Finance, the keeper of the tools to address instability, will not want to give up control over the major fiscal tools. If this
Ministry is on record as favoring decentralization, it will tend to be a controlled form of decentralization. One might look for the following features in such a program:
• Limited freedom for sub-national governments to set tax rates for any major taxes;
• Strictly controlled borrowing powers, often including central government approval.
• Centrally controlled wage and salary rates for local government employees. Giving sub-national governments a guaranteed share of national taxes compromises the flexibility of the Ministry of Finance, but the popularity of revenue sharing makes it hard to defy this policy. In many cases the MOF has not only accepted a shared tax version of intergovernmental transfers, but it has championed this approach.
Apparently, it is a more palatable policy than the alternative of giving up taxing power to sub-national governments.
The Ministry of Economy or planning could be a significant opponent of fiscal decentralization. This Ministry will be interested in a system that allows central rather than local direction of investment.
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 23

Promise and Reality of Fiscal Decentralization ✦ 23

If investment decisions are decentralized to any significant extent, it could compromise national planning on the distribution of capital expenditures by function and by location.
The line ministries often will oppose decentralization on grounds that seem more paternalistic. Their view is that the local governments (particularly rural local governments) do not have the technical capacity to deliver services or to plan resource allocation; hence there must be strong central direction. Another explanation is that the line ministries stand to lose significant bureaucratic power and some share of the central budget if conditional transfers are replaced with general purpose grants to sub-national governments. Line ministries, if they are persuaded on fiscal decentralization, will be more comfortable with a regime of conditional grants and mandated expenditure requirements.
The Ministry of Local Government (Ministry of Home affairs,
Ministry of Interior) will be an advocate of directing more funds to local government budgets. Usually this Ministry will argue for more control over the expenditures in local government budgets, e.g., determination of wages, hiring and firing of employees, oversight on expenditure processes, etc. It also can be an effective lobbyist for increased central government subsidies to local governments, as has been the case in
Japan.
The sub-national governments generally will favor decentralization, but the rich and poor will have very different views about what is the best version of decentralization. Generally, sub-national governments will argue for increased fiscal discretion on the expenditure side of the budget. On the revenue side of the budget, there is not likely to be too strenuous an argument for more taxing powers, so long as the flow of inter-governmental transfers is adequate. Increased taxation implies incurring the wrath of local voters, especially those with significant tax-paying capacity. The rural local governments will be primarily interested in a redistributive system based on a guaranteed revenue flow.
In fact, all sub-national governments will agree that they are entitled to a greater share of total national tax collections. But when it comes to the distribution of this pool among local governments, considerable disagreement will arise and the politics will force some kind of compromise. The constitution in Pakistan requires that all four provinces agree on the allocation formula. The result, predictably,
DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 24

24 ✦ Decentralization Policies in Asian Development

is that no agreement can be reached and the allocation formula
(an equal per capita sharing) is made by Presidential decision (Bahl et al., 2008).
Finally, some of the external donors and advisors will champion fiscal decentralization. The World Bank, ADB (Usui, 2007) and the InterAmerican Development Bank (1997) see decentralization as part of a development strategy that will lead to a more satisfactory and balanced growth, and promote decentralization as a country-wide strategy.
USAID is also an advocate of decentralization, and is heavily influenced by the democracy aspects. The IMF takes a more cautious and qualified view because of its concern with any policy that might promote fiscal instability. The external advisors play an important catalytic role. When they bring funding as the carrot, their main contribution is that they oftentimes catch the attention of government officials and stimulate the government to begin to look harder at the decentralization issue. But unless the government itself is enthusiastic, the harder look will not lead to meaningful policy reform and in fact will be quickly forgotten when the external support money is gone. The implementation stage is never reached.

8. Conclusion
Fiscal decentralization has not emerged as the groundswell economic development strategy that many expected. The advantages of centralization and the political power of the centralists have been too strong.
But the world has changed, and decentralization is becoming a more irresistible strategy. Its progress may be slowed by politics and by an unstable world economy, as most new policies will be, but its time may have come. Governments around the world are increasingly elected, and sooner than later, a platform of citizen participation in governance; economic development has eroded some of the advantages of fiscal centralization; and the service delivery capabilities of sub-national governments have improved dramatically. Moreover, much of the world has come to see that granting some form of local autonomy is better than separatism as a policy direction. The major roadblock to success now may be poorly conceived decentralization policies. Design must match objectives, and implementation must face up to the many dimensions of decentralization. This paper is an attempt to stimulate that discussion.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 25

Promise and Reality of Fiscal Decentralization ✦ 25

References
Alm, J, R Bahl, and R Aten (2001). Can Indonesia decentralize successfully?
Plans, problems and prospects. Bulletin of Indonesian Economic Studies,
37(1), pp. 83–102.
Bahl, R (2008). Opportunities and Risks of Fiscal Decentralization: A
Developing Country Perspective. In Fiscal Decentralization and Land
Policies, G Ingram and Y-H Hong (eds.), Cambridge, Mass: Lincoln
Institute of Land Policy.
Bahl, R W and J F Linn (1992). Urban public finance in developing countries.
New York: Oxford University Press.
Bahl, R and J. Martinez-Vazquez (2006). Sequencing fiscal decentralization.
World Bank Policy Research Working Paper 3914, Public Sector
Governance Group. Washington, DC: World Bank.
Bahl, R, S Wallace, and M Cyan (2008). Pakistan: Provincial government taxation. ISP Working Paper. Atlanta: International Studies Program,
Georgia State University.
Bahl, R and S. Wallace (2005). Public Financing in Developing and Transition
Countries. Public Budgeting and Finance, Silver Anniversary Issue, pp. 83–98.
Bird, R M and F Vaillancourt (eds.) (1998). Fiscal decentralization in developing countries. Cambridge: Cambridge University Press.
Dillinger, W and S B Webb (1999). Decentralization and fiscal management in
Colombia. Policy Research Working Paper 2122, Latin American and
Caribbean Division. Washington, DC: World Bank.
Fisman, R and R Gatti (2002). Decentralization and corruption: Evidence across countries. Journal of Public Economics, 83, pp. 325–345.
Hofman, B and S C Guerra (2005). Fiscal Disparities in East Asia: How Large and Do they Matter? In East Asia Decentralizes, P Smoke and R White
(eds.), Washington DC: World Bank.
Inter-American Development Bank (1997). Latin America after a decade of reforms. Washington, DC: Inter-American Development Bank.
Letelier, L (2005). Explaining Fiscal Decentralization. Public Finance Review,
33(2), pp. 155–183.
Litvack, J, J Ahmad, and R Bird (1998). Rethinking decentralization.
Washington, DC: World Bank.
Martinez-Vazquez, J, A Timofeev, and J Boex (2006). Reforming Regional-Local
Finance in Russia. Washington DC: World Bank Institute.
McLure, C E (1997). Topics in the Theory of Revenue Assignment: Gaps, Traps and Nuances. In Macroeconomic Dimensions of Public Finance, M Blejer and T T-Minassian (eds.), London: Routledge.
Oates, W E (1972). Fiscal federalism. New York: Harcourt Brace Jovanovich.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html b675_Chapter-I.qxd

11/24/2008

5:52 PM

Page 26

26 ✦ Decentralization Policies in Asian Development

Owens, J and J Norregaard (1991). The Role of Lower Levels of Government:
The Experience of Selected OECD Countries. In Local Government: An
International Perspective, J Owens and G Panella (eds.), Amsterdam:
North Holland, pp. 3–54.
Prud’homme, R (1995). The dangers of decentralization. World Bank Research
Observer, 10(2), pp. 210–226.
Reschovsky, A (2003). Intergovernmental Transfers: The Equitable Share. In
Restructuring Local Government Finance in Developing Countries:
Lessons from South Africa, R Bahl and P Smoke (eds.), Cheltenham, UK:
Edward Elgar Publishing, pp. 173–236.
Rodden, J, G Eskeland, and J Litvack (eds.) (2003). Fiscal decentralization and the challenge of hard budget constraints. Cambridge: The MIT Press.
Sethi, G (ed.) (2004). Fiscal decentralization to rural governments in India.
Washington D.C.: The World Bank.
Spahn, P (1997). Decentralized government and macroeconomic control.
Infrastructure Notes FM-12. Washington, DC: World Bank.
Taliercio, R (2005). Sub-national Own-Source Revenue: Getting Policy and
Administration Right. In East Asia Decentralizes: Making Local
Government Work. Washington: The World Bank, pp. 107–128.
Tanzi, V (1995). Fiscal federalism and decentralization: A review of some efficiency and macroeconomic aspects. In Annual World Bank Conference on
Development Economics, M Bruno and B Pleskovic (eds.), Washington,
DC: World Bank, pp. 295–316.
Ter-Minassian, T (ed.) (1997). Fiscal federalism in theory and practice.
Washington, DC: International Monetary Fund.
Treisman, D (2000). Decentralization and the quality of government. Washington,
DC: International Monetary Fund.
Usui, N (2007). Critical Issues of Fiscal Decentralization. ERD Technical
Note No. 21. Manila: Asian Development Bank.
White, R and P Smoke (2005). East Asia Decentralizes. In East Asia Decentralizes,
P Smoke and R White (eds.), Washington DC: World Bank.

DECENTRALIZATION POLICIES IN ASIAN DEVELOPMENT
© World Scientific Publishing Co. Pte. Ltd. http://www.worldscibooks.com/economics/6888.html