2004/07/04

LOCAL GOVERNMENT FINANCE

Lecture note 7

Fiscal decentralization in Asia and Macroeconomic effect of decentralization

1. Introduction

There has been a growing interest in fiscal decentralization in many of developed and developing nations. There seem to be both (i) economic and (ii) political motives behind. Fiscal decentralization can be a way to cope with restructuring public sectors: it is expected to contribute to improving efficiency of public sector. Also, delegating power to a lower level government will make a public policy more accountable and suitable to regional preferences as it is closer to people. It is regarded as a political device of promoting and ensuring democracy. Such a view is especially influential in transition economies. With these economic and political motives for decentralization, fiscal decentralization has been proceeded world wide in the last few decades, while its extents as well as its consequences differ substantially among nations.

Table 1: Key Indicators

Real GDP average annual growth rate % / Gross national income
per capita US $
1999 / Gross domestic saving
% of GDP
1999
1980-90 / 1990-99
China / 10.1 / 10.7 / 780 / 40.1
Indonesia / 6.1 / 4.7 / 600 / 31.6
Japan / 4.0 / 1.3 / 32030 / 27.7
Korea / 9.4 / 5.7 / 8490 / 33.6
Malaysia / 5.3 / 7.3 / 3390 / 47.3
Philippines / 1.0 / 3.2 / 1050 / 19.6
Singapore / 6.7 / 8.0 / 24150 / 51.7
Thailand / 7.6 / 4.7 / 2010 / 33.4
Vietnam / 4.6 / 8.1 / 370 / 23.2

Note: All data are from World Development Indicators (World Bank).

Asia is not an exception of this trend. Over the past three decades, the Asian nations included in this volume experienced remarkable changes in economic and political landscapes. Many of them achieved a notable economic development with a stable political regime as well as with a reasonable degree of equity in income distribution. The surge of Asian economies gained world-wide attention noted as “Asian Miracle”. A role of government in this miracle was important. With a highly centralized fiscal system, central governments in East and South East Asia initiated an economic growth by using various policy instruments. (An exception may be China.) These instruments were designed so as to direct relatively scarce resources for economic growth purpose. An economic growth was not only their policy target, but the central governments were concerned with distribution of income among regions and social classes. The miracle of Asia, however, was shaken in the wake of Asian crisis in 1997, which even turned over a long-standing political regime in nations like Indonesia. Since then, challenges of restructuring economic and political systems have continued, seeking for sustainable economic and social developments. Fiscal decentralization has been regarded as serving to ensure sustainable development and accountability of governments.

Table 2: Share of local expenditure and revenue in total (%)

Expenditure / Revenue / Expenditure / Revenue
China / 71 / 50 / Korea / 65 / 44
India / 51 / 45 / Malaysia / 18 / 12
Indonesia / 23 / 14 / Philippines / 20 / 8
Japan / 70 / 40 / Vietnam / 43 / n.a.

(Source) Hitotsubashi Journal 41(2).

In terms of both expenditure and revenue, China is the most decentralized, while according to the above data, a role of sub-national governments is still minor in Indonesia, Malaysia and Philippines. In Japan and Korea, more expenditure function than revenue raising one is assigned to local governments, and the asymmetry is filled by intergovernmental transfers. With political demand for more local autonomy increasing, there has been institutional reform and/or its attempts in many Asian countries including Japan and Indonesia, which may change traditionally centralized nature of their fiscal systems. (China may be exception of this trend, where there has been a series of attempts of re-centralization and/or establishment of formal and stable intergovernmental relations.)

Given diversity of Asian nations, the way that fiscal decentralization proceeds as well as its consequence should be highly dependent on cultural, geographical and historical nature of each country. In Part I of this course, we overview the experiences of some Asian countries with special reference to the case of Indonesia. Both common and uncommon features of decentralization would be revealed. Also, it would be made clear that fiscal decentralization is not itself panacea. There are many ways of proceeding with decentralization and some of them could be indeed socially harmful, leading to social instability and even to breaking-up of a nation. Of a great importance is proper design of decentralization with clear understanding of its costs and benefits.

In this lecture note, we make brief overview on issues associated with fiscal decentralization with emphasis on (i) its macroeconomic consequences, (ii) political accountability/quality of governments and (iii) income redistribution; the details will be covered in the subsequent lecture notes.

2. Basic Conceptual Frameworks

To begin with, let us clarify what decentralization means. There is no single meaning for "decentralization". It can designate delegation of (a)expenditure responsibilities to lower level governments or decentralization may be on (b)revenue side. Moreover, decentralization on either side involves different scopes and dimensions. On the expenditure side, authority of designing a public policy may be assigned to a different level government than the one that is in charge of implementation. “Political decentralization” accompanies delegation of “authority” of designing a policy, while “administrative” or “managerial” decentralization delegate implementation stage only. Indeed, such separation has been in the practice in Japan, where a centrally designed policy is implemented at the local level. Accordingly, local discretion in public spending is limited.

Turning to the revenue side, local level governments may have an access to tax revenue raised within own jurisdictions without being authorized to determine tax base and/or tax rate. A tax sharing is an example of this, in which tax base and tax rate on it are decided at the central level and a certain portion of revenue accrues to local governments on a derivation (origin) basis. This has been in place in Germany for individual and corporate income taxes and VAT. Alternatively, the local governments possess autonomy to decide own tax rates while definition/choice of tax base is centrally controlled or harmonized with higher-level governments. In Canada, under Tax Collection Agreement, tax bases of individual and corporate income taxes have been coordinated between the federal and participating provincial governments, with a few provinces choosing to opt out. Local level government may have free hand to define and impose taxes as long as constitutional rights such as free mobility of individuals and goods/services are not violated. Constitutionally the states in the U.S. have such a right, although in practice they have harmonized own tax bases with the federal ones largely because of administrative reasons. Finally, tax collection may be centralized or decentralized. The former is the case in Canada, where the federal government collects personal and corporate income taxes on behalf of provinces (it also collect provincial consumption tax (VAT) in Atlantic provinces), while the latter is the practice in Germany where the states are in charge of collecting shared tax revenues. A decentralized tax collection (i.e., the case that sub-national governments administer central taxes and transfer collected revenue to the center) was also the tradition in socialist economies.

Table 3:Definition of Decentralization

(a)Expenditure / (i) Design
(ii) Implementation
(b)Revenue / (i)Cash-flow (Revenue only)
(ii)Authority /
  1. Tax Base
  2. Tax Rate
  3. Tax Collection

Table 4:Variations of Decentralization

Expenditure Side:

Design / Implementation
Japan (Administrative decentralization) / Centralized / Decentralized
Many federal countries / Decentralized

Revenue Side

Cash-flow (Revenue) / Authority
Tax Base / Tax Rate / Tax Collection
Canada / Decentralized / Centralized / Decentralized / Centralized
Germany (Tax Sharing) / Decentralized / Centralized / Decentralized
The U.S. / Decentralized

The above definition applies to “each” public policy and tax. Note that one tax such as income tax and VAT may be shared with the central authority under centralized determination of tax base and tax rate, while another tax may be devolved along with authority of choosing tax rate as well as tax base. A local tax system as a whole involves a mix of taxes to which different degrees of autonomy may be granted. Regarding expenditure, which responsibility -health, education, social assistance and so forth- should be assumed by local governments needs to be carefully examined. Some expenditure responsibility may be exclusive to one level government, while another may be shared between the central and local governments, the latter of which is frequently observed in practice. In addition, since there are more than two tiers of local governments, to which level of local government - state, city or local municipalities- a task should be delegated is the matter of discussion.

Generally speaking, fiscal decentralization involves re-allocation of design, implementation and financing (revenue) responsibilities of different tasks among different level governments;it will mean different things in different contexts. (So try to understand in what context decentralization is being discussed!) During this course, it will be repeatedly addressed that decentralization is not good or bad in nature but its consequence on social well being relies on its “design” that accounts for economic and political costs and benefits associated with each task (say a public service such as education) and/or authority (say control right of tax rate on a certain tax base) being devolved to lower level governments.

3. Normative and Positive Analysis

Analytical tools of economics are applied to (i) explain and (ii) evaluate the status quo, the practice being observed, along with (iii) policy implications based on explanation and evaluation. At this point, it is important to distinguish normative views from positive ones. The former discusses what government (namely the central government) “should” do, while the latter is concerned with what governments (both central and local) “would” do. Such distinction is quite important in the context of local public finance as both political and economic issues are involved. Thus our concern is not only inefficiency and/or inequity arising in market (or private sector) but also political distortions within public sector. Fiscal decentralization or centralization is to cope not only with economic costs of the status quo whatever it is, but also with political costs that are associated with incentives of politicians, bureaucrats at every level of government as well as of special interests and national and regional voters. As stated above, decentralization involves re-allocation of tasks among different levels of government. To evaluate such a re-allocation, not only economic consequences such as changes in resource allocation and income distribution but its incentive effects on governments need to be addressed.

Normative Views

Although the institutional (political) and/or historical backgrounds differ among countries, some universal principles for desirable task allocation among different levels of government should apply. Three functions have been laid out that public sector should undertake: (i)macroeconomic-stabilization, (ii)income redistribution and (iii) provisions of public goods. As we study in more details during the course, the traditional literature contends that the first two should be assigned to the central government, which stems from fundamental constraints on lower level governments: the reason is related to openness of regional economies in terms of mobility and transaction of goods/services and production factors such as labor and capital (Musgrave (1959), Oates (1972)).

Concerning the last function, i.e., provisions of public goods, natures of such goods are of great importance. National public goods like national defense and justice system benefit overall population within a nation, and thus should be the responsibility of the central authority. Lower level governments, on the other hand, is better in providing public goods whose benefits are local (regional) than the central government. This is due to informational advantage of lower level governments concerning regional preferences and needs for such goods. This proposition is known as the Decentralization Theorem (Oates (1972)). Decentralization should allow the local authorities to explore such informational advantage.

Table 5: Function Allocation

Functions / Level of Government
Macroeconomic-stabilization / Central
Income redistribution
Provisions
of public goods / National public goods
Local public goods / Local

At this point, the following two should be noted. First, (a)constraint and/or capacity of government should be distinguished from its (b)incentive. In the present context, possessing better information on local needs is one thing, reflecting it in policy making is another. A mechanism that induces government in charge to act in interest of its electorates needs to be considered, which is addressed in “positive” view. Second, each task (say provision of local infrastructure) involves more or less all three functions: in the case of local infrastructure, it may be used for redistribution and stabilization purposes as well. An important thing is that in assigning tasks among different level governments, care should be taken to minimize functions of redistribution and stabilization at the local level.

As noted above, each task is composed of (i) design, (ii) financing and (iii) implementation. Given that there is a case for decentralized provisions local public services (that is, lower level governments fulfill (i) and (iii)), a question arises how to finance it. The expenditure responsibility may be met by (a)local taxes or own revenue, (b) local bonds and (c) intergovernmental transfers in conditional or unconditional form; the first involves decentralization on revenue side (The last one as well if local governments are granted free hands to issue bonds). In the literature on fiscal federalism, it has been argued that from a “normative standpoint” there is less case for revenue decentralization than decentralization of expenditure responsibility, the latter of which is supported by “Decentralization Theorem”. (And decentralized issuance of local bonds in a decentralized way is claimed to endanger macro-stability as we turn to below.) This reflects the notions that while expenditure responsibility can be delegated to the local level or shared with the central government, the central government is advantageous in raising revenues. This argument, if followed, gives rise to fiscal gap between central and local governments; namely, the central government raises more revenue than it spend, and local government cannot afford own expenditure responsibility only with own revenue. Such a gap needs to be filled by intergovernmental grants.

Table 6: Vertical Fiscal Gap

Local Government / Central Government
Local
Expenditure Responsibility / (a)Local Tax / Central (own) Expenditure / Central Tax and Bonds
(b)Local Bond
(c)Intergovernmental Transfers

Good and Bad Local Tax

From a normative standpoint, while there is a strong case for decentralizing expenditure responsibilities, revenue decentralization can give a rise to significant economic costs. With their open economies, mobility of tax units (e.g., wealthy residents in the case of income tax and firms in the case of business tax) constrains ability of the local governments to raise sufficient revenues for financing their expenditure obligations as increasing tax rates on such mobile factors may lead to outflow of them to other jurisdiction with lower tax burden. In addition, decentralized determination of tax rates levied on mobile factors can distort inter-regional resource allocation. This is because the mobile factors are diverted to regions with lower tax burden rather than to the most productive ones, which hinders economic activities within a nation as a whole. Furthermore, local jurisdictions with different fiscal capacities may end up in different amounts of the local public expenditures, which may be regarded as inequitable. Therefore, in contrast to the expenditure side, a centralized tax system is often preferred from efficiency and equity standpoints, although benefit taxes such as user charges can be delegated to the local level.

While central tax should follow the ability to pay principal due to the central responsibility for income redistribution, benefit principle that requires beneficiaries of local public services to bear tax burden without consideration of income redistribution should apply. Given that, local tax bases should be (i) stable, (ii) immobile and (iii) relatively equally distributed across regions. The first two conditions are to secure stable and adequate provisions of public services that local governments are in charge, while the last is called for from a fiscal equity standpoint. It has been noted that tax bases that fulfill the conditions of “Good local taxes” are not abundant, but limited to land (property) tax and head (poll) tax. Consequently, major taxes such as personal and corporate income taxes are occupied by the central government according to this principle. Fiscal imbalance arises among different levels of government, which must be closed by intergovernmental transfers to close the gap.

Table 7: Own Revenue Sources

Income Taxation / Consumption Taxation / Property Taxation / Non-Tax
Example / Personal Income Tax
Corporate Income Tax (Business Tax) / Value-added Tax,
Sales Tax
Excise Taxes / Land Tax
Natural Resource Tax
Bequest Tax / User Charges
Fees

In this course, we discuss more details of inefficiency and inequity in a decentralized fiscal system as well as normative policy implications to correct them. However, it should be born in mind that in practice, it is often intergovernmental transfers that hinder local autonomy and/or distort incentives of local/central government agents giving rise to inefficiency and inequity. This issue is also covered in the class.

Human Resources

Beside (a) authority and (b) responsibility, it is (c) human resource that should be concerned, especially in the context of developing countries. One of critical assumptions of “Decentralization Theorem” is that local government possesses managerial capability of using own information on local needs, which is often missing in developing and transition economies. A traditionally centralized regime has attracted more “talented”people to the central government. Also local politicians who are ambitious and capable have been seeking for positions in the central politics. These imply that there have not been sufficient human resources or management skills being accumulated at the local level. Then human resources need to be redistributed from central government and/or private sectors or they must be accumulated over periods otherwise. In the case of latter, it is conceivable that policy making/management quality may be deteriorated immediately after decentralization. In addition, as noted in 3.2, local politicians and bureaucrats may be more shortsighted and/or corruptible.