22 Korea

Teresa Ter-Minassian
Published Date:
September 1997
  • ShareShare
Show Summary Details
Ke-young Chu and John Norregaard 

Korea broke away from a long tradition of a highly centralized system of government and embarked on a new system of political devolution following the nationwide subnational government elections in 1994. Until that time, the subnational governments were purely administrative arms of the national (central) government and their heads were appointed by the President. Korea’s long tradition of a centralized political system had been interrupted only briefly in the early 1960s. The introduction of the new political arrangement gives rise to an entirely new public finance question in Korea: what would be the intergovernmental fiscal arrangement most appropriate for the new political arrangement?

The old framework of highly centralized intergovernmental fiscal relations helped the government pursue centrally formulated policies. The centralized system supported the government’s design and implementation of a series of national economic development programs. It enabled the government to maintain the central control of public investment and the provision of services, including some often delivered by local governments in other countries (such as local public safety and justice and primary and secondary education).

The centralized fiscal system, however, leads to various inefficiencies. It is not suitable for subnational governments, now politically autonomous, to meet the demands for locally differentiated public services; nor does it promote competition among subnational governments for providing efficient services. As the Korean economy grows and becomes more complex, it becomes increasingly difficult for the central government to control subnational expenditures. Moreover, the political devolution is leading to demands for greater subnational fiscal autonomy.

Korea, therefore, has embarked upon a transition from fiscal centralization toward increased subnational fiscal autonomy. The steps taken by the government so far, however, have increased administrative, rather than fiscal, decentralization. Subnational governments still remain the central government’s administrative arms and continue to have only limited autonomy to formulate their own fiscal programs. Their budgetary resources come to a considerable extent from national tax revenues. Subnational governments’ fiscal capacities vary substantially throughout the country, given the regional variation in the local tax base, which reflects, inter alia, the cumulative effects of the past regional allocation of public and private investments.

The limited degree of fiscal decentralization to date implies that more fundamental and wide-ranging issues will need to be addressed in Korea than in other countries, which in general have longer-established systems of intergovernmental fiscal relations.

For Korea, the issues to be considered include the following:

  • What is the appropriate level of decentralization and local fiscal autonomy and how can the present vertical fiscal imbalance be reduced?

  • Which expenditures should be assigned to subnational governments and which should remain under central control?

  • What is the appropriate tax assignment?

  • How should the transfer system be improved and how should the required degree of equalization be achieved without undermining efficiency?

  • What kind of access should subnational governments have to financial markets?

  • What are the institutional implications?

A successful transition requires that both macroeconomic stability be preserved and that the right balance be found between an increase in efficiency through better and more varied public services, a reduction in the existing vertical fiscal imbalances, and a reduction of inequalities in the distribution of fiscal capacities across different regions. The transition is taking place against a difficult background of growing demands for central government investments and services (such as increased public infrastructure, social security, and higher-quality health and education services), the economy’s structural changes, substantial population migration from rural to urban and from urban to suburban areas, and wide disparities in regional fiscal capacities and needs, given the concentration of wealth, income, and public services in urban areas and certain high-growth regions.

The transition will require substantial preparatory work and political negotiations aimed at achieving the necessary political consensus at all levels of government. The process will take time and has to proceed through cooperation between the national and subnational governments, as the latter strengthen their administrative capacities, while establishing secure roots for local democracy.

This chapter focuses on the key options facing policymakers. The following section briefly outlines the main features of the present system.

Present System of Intergovernmental Fiscal Relations Structure of Government and Vertical and Horizontal Fiscal Imbalances

The present structure of government has the features of a unitary system with a degree of administrative decentralization. A three-tiered structure of government comprises the national government and two-tiers of subnational governments—regional governments and local governments. The national government is referred to as the central government (or government), and the subnational governments as local autonomous bodies. Regional governments comprise the Seoul metropolitan government, 5 direct jurisdictional cities (large cities), and 9 provincial governments; local governments comprise (1) 22 autonomous city districts and 33 self-governing wards under the direct jurisdictional cities and (2) 67 cities (small cities) and 137 counties under the provincial governments.1 As in other administratively decentralized unitary systems, each tier has its own responsibilities and some own sources of financing, but the higher tier provides transfers to the lower tier. Seoul and the five large cities have been assigned provincial functions and taxes and also receive transfers from the central government.

A main feature of the system is substantial vertical and horizontal fiscal imbalances.2 Subnational governments, while accounting for over half of the general government expenditure, raise only some 20 percent of their revenue through taxation, financing the rest mainly through central government transfers (Table 1). There is a wide variation in regional incomes and subnational government capacities to mobilize revenues. For example, in 1994, per capita gross regional product (GRP) ranged between W 4.6 million ($5,750) for Taegu and W 8 million ($10,000) for South Kyungsang. With central government transfers, per capita subnational government expenditures varied considerably, ranging between W 0.5 million for Pusan, Inchon, and Kwangju and W 1.2 million for Kangwon (Table 2).3

Table 1.Korea: Consolidated General Government Fiscal Operations, 1995

(In trillions of won)
Ratio to GDP

(In percent)
General government
Tax revenue72.620.8
Other revenue18.15.2
Expenditure and net lending91.926.4
Capital expenditure and net lending36.010.3
Other expenditure55.916.0
Of which:
Local autonomous bodies
Tax revenue13.33.8
Expenditure and net lending44.912.9
Of which: capital spending19.95.7
Memorandum items:
Central government transfers to local autonomous bodies22.66.5
Current transfers17.85.1
Capital transfers4.81.4
Source: Bank of Korea, Economic Statistics Yearbook, 1996; based on the 1995 budget figures.Note: There are some discrepancies between various estimates of government revenue, expenditure, and deficit. For example, in another set of estimates, the budgeted local government deficit is W 4 trillion rather than W 9 trillion, and general government revenue nd expenditure were somewhat higher than reported in Tables 4 and 6.
Source: Bank of Korea, Economic Statistics Yearbook, 1996; based on the 1995 budget figures.Note: There are some discrepancies between various estimates of government revenue, expenditure, and deficit. For example, in another set of estimates, the budgeted local government deficit is W 4 trillion rather than W 9 trillion, and general government revenue nd expenditure were somewhat higher than reported in Tables 4 and 6.
Table 2.Korea: Regional Distribution of Incomes and Budgetary Resources
Per Capita

(In millions)


(In trillions

of won)


(In millions of won)

Gross regional product

(In millions of won)


(In thousands of won)

Large cities
North Chungchong1.
South Chungchong1.
North Cholla2.
South Cholla2.
North Kungsang2.
South Kyungsang4.
Sources: Ministry of Home Affairs (1994); and Seoul Metropolitan Government (1996).
Sources: Ministry of Home Affairs (1994); and Seoul Metropolitan Government (1996).

Expenditure Assignment

Principles Underlying Korean Expenditure Assignment

The Law on Local Autonomy provides a broad system of expenditure assignment in line with the new devolved intergovernmental political arrangement. The principles underlying the distinction made between purely national and purely local policies is broadly in accordance with the theory of intergovernmental fiscal relations. According to the law, functions of the local autonomous bodies are limited to the provision of local public services:4

  • Local public administration (such as local legislation, collection of local taxes, civil defense administration, and firefighting).

  • Support for local economies (for example, promotion of farming, small-scale animal husbandry, and consumer protection) and regional development (for example, urban planning and local roads).

  • Promotion of social services for local residents and local education, sports, culture, and arts (such as the establishment of local kindergartens and primary and secondary schools).

The law prohibits local autonomous bodies from engaging in the provision of pure national public goods or those with strong nationwide externalities:5

  • National security, public order and safety, and foreign relations.

  • National economic policy, including such areas as monetary policy, fiscal policy (including national taxation), pricing policy, management of grain stocks, and external trade policies; economic development programs; industrial standards; postal and rail transportation services; and public services requiring high technology.

Main Features of the Expenditure Assignment

Several features of the expenditure assignment deserve attention:.

The practice of expenditure assignment often deviates from enunciated principles. The legislation provides an expenditure assignment in line with the basic theory of intergovernmental fiscal relations. In practice, however, the system still reflects the old centralized regime. Local autonomous bodies are engaged extensively in the provision of central government functions as its agents.

The central government has considerable influence over the fiscal operations of local autonomous bodies. Because of this influence, the local autonomous bodies’ relatively high share in general government expenditure does not reflect their relative fiscal autonomy. For example, at present, local autonomous bodies account for some 50 percent of general government expenditure if education board expenditures are consolidated into their expenditures. The central government exercises its influence through at least two channels: (1) its largely conditional transfers and subsidies, which account for some 30 percent of general government expenditure if central government transfers to education boards are considered local autonomous bodies’ expenditures,6 and (2) considerable regulatory authority over local autonomous bodies’ operations. Local autonomous bodies have limited authority to introduce new taxes. Moreover, the Ministry of Home Affairs and other central government ministries exert considerable influence on local autonomous bodies’ policy formulation and implementation through a combination of mandates and administrative instructions. The central government imposes price controls on certain locally provided public services or privately provided local services, such as intracity transportation.

Moreover, the law is not specific regarding many national or subnational government functions. There are many mixed national and subnational government functions that are carried out mostly by local autonomous bodies with central government financial support. There is an extensive overlapping of functions between the various levels of government in many areas (Table 3). These features can result in either underprovision or overprovision of public services.

Table 3.Korea: Distribution of and Budgetary Resources for Cities, Provinces, Counties, and Districts, 1996(In trillions of won)
Total ExpenditureLarge Cities and ProvincesSmall Cities, Counties, and Districts
Large cities
North Chungchong1.70.81.0
South Chungchong2.61.31.3
North Cholla2.91.31.7
South Cholla3.11.41.8
North Kyungsang3.31.32.0
South Kyungsang4.31.62.7

A large share of the expenditures of local autonomous bodies, presently more than one-third of general government expenditure, is devoted to capital projects, constraining the amount of resources available for the delivery of current services to the population. The demand for current services is rising, as witnessed by popular pressures on the newly elected local autonomous bodies.

Tax Assignment

Basic Structure of Tax Assignment

The tax system consists of nonoverlapping national and subnational (or local) taxes (Tables 4 and 5).7 The national tax system comprises standard central government taxes, such as the VAT, an income tax, and a corporate tax. The structure of local taxes is shown in Table 6.8 More than two-thirds of the yield from the 15 different local taxes derives from taxes on property, with only 12 percent deriving from taxes on income, largely from the inhabitant tax.

Table 4.Korea: Functional Composition of General Government Expenditure, 1995(In trillions of won)
General GovernmentOf Which: Local Autonomous Bodies
Total expenditure91.944.9
Public administration15.98.9
General administration10.08.0
Public order and safety5.90.9
National defense11.0
Primary and secondary education12.311.3
Tertiary education2.81.0
Social welfare7.92.1
Housing and community development9.85.2
Economic services28.914.4
Agriculture and fishery9.64.0
Mining and manufacturing3.91.2
Transportation and communication4.03.4
Source: Bank of Korea, Economic Statistics Yearbook, 1996; based on the 1995 budget figures.
Source: Bank of Korea, Economic Statistics Yearbook, 1996; based on the 1995 budget figures.
Table 5.Korea: Tax System
National TaxesLocal Taxes
Internal taxesTaxes for provinces and metropolitan cities
Direct taxesAcquisition tax
Income taxRegistration tax
Corporate taxHorse race tax
Tax on the excess increase in land valueCommunity facility tax
Inheritance and gift taxRegional development tax
Assets revaluation tax
Excess profits taxTaxes only for provinces
License tax
Indirect taxes
Value-added taxTaxes only for metropolitan cities
Special excise taxInhabitant tax
Liquor taxAutomobile tax
Telephone taxFarmland tax
Liquor taxTobacco consumption tax
Stamp taxBurchery tax
Securities transaction taxCity planning tax
Transportation tax
Taxes for cities, counties, and districts
Customs dutiesProperty tax
Education taxAggregate land tax
Workshop tax
Special tax for rural developmentTaxes only for cities and counties
Inhabitant tax
Automobile tax
Farmland tax
Tobacco consumption tax
Butchery tax
City planning tax
Taxes only for districts
License tax
Table 6.Korea: Composition of General Government Revenue, 1995

(In trillions of won)
Ratio to GDP

(In percent)
Total revenue and grants90.726.0
Tax revenue72.620.8
Taxes on income and profits21.86.3
Social security contributions5.01.4
Taxes on property10.02.9
Taxes on goods and services27.27.8
Taxes on international trade3.51.0
Other taxes5.11.5
Nontax revenue18.15.2
Memorandum items:
Local autonomous bodies
Tax revenue13.23.8
Of which: property taxes7.92.3
Source: Bank of Korea (1996); based on the 1995 budget figures.
Source: Bank of Korea (1996); based on the 1995 budget figures.

Local tax revenue in 1994 constituted 18.1 percent of total tax revenue, compared with about 10 percent in 1975. Local taxes reached about 4 percent of GDP in 1995, against 1.6 percent in 1975. Despite this strong increase, the Level of local taxes is still fairly modest by international standards.9 During the past two decades, local taxes have increased in importance, mainly through a series of tax reforms; these reforms improved the local tax base through the introduction of new local tax sources and the assignment of previously central taxes to the local level.

Main Features of Tax Assignment

Tax assignment has several notable features.

First, the buoyant taxes belong to the central government; local taxes lack buoyancy.

Second, seven different taxes are levied on property, and although land and buildings are key elements in most of them, there are significant differences in the way tax bases are measured. Most property taxes involving land and buildings are progressive ad valorem taxes,10 presumably because the ownership of property is considered to be closely correlated with income, and thus a good indicator for ability to pay. Although mostly ad valorem—based in principle, there are elements of specific property taxation, particularly in the registration tax, as indicated by fairly infrequent rate adjustments. The present valuation system, based on computerized local cadasters and administered at the city, district, and county levels, is, in principle, based on market values; however, owing to a complex system of coefficients applied to different areas and different uses of land and buildings, current assessed values deviate significantly from market values. According to some estimates, assessed values constitute on average only 30 percent of market values.

Third, the inhabitant tax, a major local income tax, is primarily a source-based tax (that is, its yield accrues to the jurisdiction of the workplace), with little weight on residence-based taxation (the per family element of the tax, which is based on the taxpayer’s place of residence, generates only 5 percent of the total revenue, whereas the source element accounts for the remaining 95 percent). Furthermore, if a person has more than one type of income (for instance, wage and interest income), the source-based tax element may be paid to different jurisdictions, depending on the location of the payment place (in the example, the workplace and the location of the bank, respectively).

Fourth, earmarking is extensively used in the present system of local taxation, including for the facility tax, the regional development tax, the city planning tax, and the workplace tax. An excessive use of ear-marking tends to reduce budgetary flexibility and efficiency, particularly when the existence of an element of benefit taxation is not obvious (for example, the automobile tax).

And, finally, although the present tax legislation provides local autonomous bodies with substantial flexibility with regard to varying tax rates, all local entities have chosen not to utilize this option, using instead the standard tax rates.11

Intergovernmental Transfers

Overview of the Transfer System

Revenues from local taxes presently cover slightly more than one-third of local expenditure, excluding education expenditures. This vertical imbalance is corrected by a system of central government transfers, with specific conditional transfers dominating over general unconditional transfers. Transfers from the central government are a substantial source of revenue for the local autonomous bodies. These transfers, excluding those for education, represent about 70 percent of the estimated total local tax revenue. A parallel system of transfers is used to finance the education boards.

The transfer system is a multipurpose system, which in its design is similar to those used in many developed countries. The formulas for the distribution of the various transfers as well as the provisions for determining their total amount are very detailed. However, the central government retains large discretionary authority with regard to many aspects of the transfer system.

Main Features of the Transfer System

The transfer programs have the following features.

The local shared tax fund is used for the distribution of general purpose transfers to the local governments (Table 7). The global amount of the fund corresponds to 13.27 percent of total internal tax revenues (earmarked taxes, such as the transportation tax, are not included in these revenues). The fund has two components. The general component (comprising 1011 of the total transfer) is distributed according to a complex formula, which takes into account expenditure needs and the revenue capacity of each jurisdiction. The cities of Seoul and Pusan, Kyungki Province, and nine other local governments are excluded from the distribution because their revenue capacities are believed to exceed their expenditure needs. The distribution is made by the Ministry of Home Affairs, which is also responsible for the design of the formula. The special component (comprising 111 of the total transfer) is used for unforeseeable occurrences, such as natural disasters.

Table 7.Korea: Structure of Local Government Tax Revenue, 1995

(In trillions of won)
Share of Total

(In percent)
Taxes on income1.612.4
Inhabitant tax1.410.5
Farmland tax
Taxes on property
Acquisition tax9.168.3
Registration tax2.216.7
Property tax2.921.8
Automobile tax0.54.0
Global land tax1.18.6
City planning tax0.64.7
Community facility tax0.21.5
Taxes on consumption2.317.6
License tax0.21.4
Butchery tax0.2
Horse race tax0.21.4
Tobacco tax1.914.3
Regional development tax0.10.4
Memorandum items:
In percent of total revenue38.9
In percent of GDP3.8
Source: Bank of Korea (1996); based on the 1995 budget figures.
Source: Bank of Korea (1996); based on the 1995 budget figures.

The local transfer fund is used for the distribution of specific transfers, basically for road construction programs, water provision, youth support, and the development of rural areas. Its total amount is determined as 50 percent of the revenue of liquor taxes, plus the entire revenue of the phone tax, 50 percent of the tax on real estate capital gains, and 111 of the special tax on agriculture and fisheries. It is a typical matching grant distributed by the Ministry of Home Affairs. The matching rate, as well as the criteria for the allocation of the fund among various sectors and local governments, is dictated by sectoral laws and other statutes. The transfers from the fund are predominantly allocated to rural areas. Beneficiary governments are free to determine the characteristics of their projects.

National treasury subsidies include a variety of specific transfers distributed by sectoral ministries and are used mainly for financing capital projects. There are no general criteria for the determination of the total amount, which reflects the implementation of sectoral laws and decrees. The laws also determine the criteria for the selection of the projects to be financed and the matching rates. The distribution is made by the relevant sectoral ministries. Beneficiary governments have to comply with strict and detailed criteria regarding the formulation and the implementation of their projects.

Transfers to the education boards are provided for the administration of primary and secondary schools (Table 8). Each education board is composed of 25 individuals, each selected by municipal councils from a list of persons established by the autonomous districts. Their budgets are financed by students’ fees and transfers from the central government and from local autonomous bodies. Transfers from the central government to the education boards replicate those paid to local autonomous bodies. There are three types of transfers: (1) a local education shared tax, with the global amount corresponding to 11.8 percent of the revenue from internal taxes, plus total wages and salaries, to be distributed according to a formula that compares expenditure needs and revenues from transfers and fees; (2) local education transfer funds, an earmarked fund financed by the education tax, to be distributed according to total population; and (3) national subsidies for education, distributed by the Ministry of Education for the construction of school buildings.

Table 8.Korea: Transfers to Local Autonomous Bodies and Education Boards, 1996
Types of transfersAmount

(In billions of won)
To local autonomous bodies12.8
Local shared tax6.1
Local transfer funds2.6
National treasury subsidies4.1
To education boards13.0
Education shared tax8.5
Education transfer fund4.1
Education subsidies0.4
Source: Bank of Korea (1996); based on the 1995 budget figures.
Source: Bank of Korea (1996); based on the 1995 budget figures.

The law also mandates regional governments (provinces and large cities) to transfer specific percentages of their own tax revenues to the cities, counties, or autonomous city districts included in their jurisdictions. The structure of these transfers is an almost perfect duplicate of those provided by the central government to regional governments. Thus, there are general purpose transfers, distributed according to a formula based on a comparison between needs and revenue capacities, and specific purpose transfers for projects selected by the donor government. All these transfers are strictly regulated by the central government.

Local Borrowing

Basic Principles of Local Borrowing

As in most other developed countries, local autonomous bodies in Korea are permitted to borrow, but subject to an elaborate regulatory framework, detailing the conditions for bond financing and other debt instruments. The overall objectives of this regulatory framework are:

  • To limit the aggregate amount of local borrowing.

  • To enhance responsibility on the part of the local autonomous bodies, taking into account that the time horizon of elected officials’ decision making might be considerably shorter than the economic lives of the projects financed by borrowing.

  • To prevent the concentration of economic power in richer areas of the country.

The Ministry of Home Affairs plays a key role in the regulation of local borrowing—except in the case of foreign borrowing, where the Ministry of Finance and Economy is the key player—by issuing detailed regulations and guidelines for local borrowing. Local governments are required to get approval from the Ministry of Home Affairs—as well as approval by the local councils—for all borrowing. Sector ministries also participate in the approval process through an elaborate hearing procedure within their respective areas of interest, and the Ministry of Finance and Economy is also involved in this process. Furthermore, this ministry may itself extend subsidized loans to local governments to support national objectives and projects, although formally, the lending is executed through the relevant sector ministries (such as the Ministry of Transportation or the Ministry of the Environment).

Key Features of the System of Local Borrowing

Two aspects of the system of local borrowing are noteworthy: the implications of central government regulations and compulsory local bond placements.

The regulations concerning borrowing comprise detailed eligibility criteria12 that determine (1) which local governments are allowed to borrow—only those with a history of sound financial policies (for instance, those with no overdue obligations, a low debt-service ratio, and a low deficit); and (2) which projects are eligible to borrow—capital projects, revenue shortfalls due to natural disasters, and others that are especially useful for improving the welfare of residents.13 There is no a priori ceiling set on the overall amount of borrowing, either by the Ministry of Finance and Economy or by any other central authority.

A particular feature of the Korean system is the existence of compulsory bond placements. Through a city ordinance, Seoul and other cities have decided that the beneficiaries of services eligible for local borrowing should also contribute to the financing. Compulsory bond placements, introduced in 1979 initially for “water bonds,” continue, primarily relating to Urban Railroad Bonds, and take place when automobiles are purchased, when licenses or permits are issued, and when bids are made for local government projects. The terms, which are decided solely by the city council, comprise a five-year grace period and an interest rate of 6 percent (compared with the market rate of about 11 percent). In addition to these compulsory bond placements, local governments issue bonds on the international market (in Japan and the United States) on market terms and, in recent years, at a much lower interest rate than the domestic level (even after taking into account exchange rate movements).

As a consequence of the historically determined close integration of functions and responsibilities of governments at different levels in Korea, the risks associated with borrowing are also shared in the sense that, when central government approval has been granted, a loan is automatically considered covered by an effective—if not statutory—state guarantee.

Main Reform Issues

The reform of Korea’s intergovernmental fiscal relations cannot be considered separately from the country’s broader economic goal: sustained high-quality economic growth. The central government faces a major task of maintaining the soundness of macroeconomic policy in an increasingly open international economic environment and meeting the country’s growing demands for high-quality public investment and social programs, while keeping the tax burden at a modest level. The economy’s overall productivity and efficiency must increase. To achieve these objectives, Korea needs to increase public sector productivity, and the public sector must aim its policies at promoting efficiency in the economy’s overall resource use. An essential step toward this goal is to reduce government regulations and to improve the quality of regulations. In pursuing these objectives, Korea needs to reduce the existing vertical and horizontal fiscal imbalances, while meeting the local autonomous bodies’ increasing demand for fiscal autonomy.

It appears inevitable for Korea to decentralize further its intergovernmental fiscal relations. As a matter of fact, the country is already moving in this direction. The present system does not enable Korea to benefit from the potential efficiency gains that can result from a fiscally decentralized system of intergovernmental fiscal relations.14 The move toward a system of increased fiscal decentralization will pose significant challenges for the decision makers. A large measure of transparency, cooperation, and consultation in intergovernmental fiscal relations is also required. The pace of reform will have to take into account the development of local governments’ administrative capacity. During the process of transition to a new system, Korea must address several closely related issues, as a basic prerequisite for the introduction of a well-functioning decentralized system of government, namely:

  • Ensuring that all levels of government play proper market-complementing roles and strengthening macroeconomic management.

  • Allocating expenditures properly among the different tiers of government, including the appropriate level of local government autonomy over their expenditures.

  • Redesigning the system of tax assignment, including the appropriate degree of local taxation autonomy.

  • Defining clearly the appropriate local autonomous bodies’ access to borrowing.

  • Establishing institutional arrangements, in particular, to address externalities and spillover effects and to coordinate intergovernmental fiscal relations.

The rest of this section discusses these key issues.

Role of Public Sector, Vertical and Horizontal Imbalances, and Macroeconomic Management

Role of Local Autonomous Bodies as Public Sector Entities

Facing an increasingly competitive external environment, Korea is searching for ways to improve the efficiency of the economy. The central government is looking for ways to reduce its role, including through substantial deregulation. A danger is that, while the central government pursues policies of deregulation and privatization, local autonomous bodies may expand market-replacing activities for short-term gains in their revenues. This would reduce or offset efficiency gains of central government policy. Without a strict and transparent legal framework aimed at defining the roles of local autonomous bodies, such a danger is real. Local autonomous bodies should be strictly prohibited from engaging in market-replacing activities without clear justification.

Vertical and Horizontal Imbalances

Large vertical fiscal imbalances often reflect the desire of the central authorities to control local government fiscal operations. Excessive transfer financing resulting from large vertical imbalances, however, limits local fiscal autonomy, masks the true costs of the services provided locally, and thereby reduces local political accountability. Vertical fiscal imbalances reduce significantly the potential efficiency gains associated with decentralization. On the other hand, given the substantial inequality in the geographical distribution of the tax base, an assignment of a larger tax base to local autonomous bodies would increase the need for a system of intergovernmental equalization transfers. Moreover, it would imply, with a given overall tax burden, a reduction in the base for central government taxation. While a reduction in the present level of vertical imbalances clearly is warranted, the desired level must balance these conflicting considerations.

To give local autonomous bodies equal opportunities to provide their citizens with adequate public service at comparable “tax prices,” fiscal capacities and “objective” expenditure needs (that is, expenditure requirements resulting from external factors outside the control of local autonomous bodies) should be equalized. To this end, two important questions must be addressed: what is the appropriate degree of equalization of fiscal capacities of local autonomous bodies and what are the implications of alternative equalization options.

The first question is a political one, which has to be resolved through the political process, but the Korean authorities must pay attention to the implications of the equalization for efficiency. Equalization may aim at reducing disparities in the actual provision of current services, or at filling existing disparities in infrastructure, or even at reducing existing gaps in income and wealth levels among the various areas.

The second question has a number of aspects. If an increase in the degree of equalization is found warranted, the Korean authorities could consider a mix of the following approaches: (1) The central government could increase the total amount of equalization transfers. For a given level and assignment of taxation, this would reduce the resources available for other central government activities (including “other transfers” to subnational governments). (2) The central government and local autonomous bodies could introduce a transfer aimed at requiring “wealthier” jurisdictions to make transfers to “poorer” jurisdictions. This would increase horizontal equity without changing the vertical balance. (3) The central government could finance an increase in equalization transfers by increasing national taxes. This would increase the national tax burden unless the increased transfers were used by local autonomous bodies to reduce their taxes correspondingly. This option would, of course, worsen vertical fiscal inequity. (4) Local autonomous bodies could increase their tax efforts, which—if successful—would reduce the degree of vertical imbalance, enabling the central government to redirect its transfers to achieving greater equalization. This would also increase the overall tax burden, unless matched by a reduction in the national taxation.

Macroeconomic Management

In Korea’s present centralized system of government, a lack of macroeconomic control is not an issue. In any “good” system of multilevel government, the central government maintains a significant measure of macroeconomic control based on a variety of tax and expenditure policy instruments. Ideally, however, this policy should be implemented with a minimum of interference in the activities of local autonomous bodies. When interference is necessary, policies should be implemented through general measures such as reductions in either the overall level of borrowing or general grants. Micromanagement through detailed constraints on either tax policies or expenditure policies would run counter to the rationale for decentralization and would seriously hamper local accountability. There are three specific issues:

  • The more own sources of taxation local autonomous bodies are assigned, and the larger the autonomy granted them, the less is generally the degree of central government fiscal control (over expenditures, tax rates, and other tax policy variables). In the present circumstances, however, such a possibility appears rather remote in Korea.

  • The fixed coefficient for dividing the aggregate national tax revenue between the central government and local autonomous bodies might create difficulties for fiscal adjustment at the general government level. For example, the central government’s successful effort to increase revenues from national taxes aimed at correcting deterioration in the general government fiscal balance will be frustrated if local autonomous bodies use the increased transfers to pursue expansionary expenditure policies.

  • The local autonomous bodies face a growing need to finance major regional infrastructure projects (for instance, subway systems in major cities). With a limited tax base, the pressure for local borrowing will increase. Sound management of local borrowing is essential for macroeconomic stability. One approach to promoting local fiscal sustainability would be to allow full autonomy of local borrowing, but subject to strict market discipline, with no central government financial backup. The conditions for such an approach to be effective (see Chapter 7) do not appear to be fulfilled in Korea at the present time, and therefore should be considered only as a long-term goal. In the short to medium term, the approach of continued central government control of local borrowing would appear to be necessary to ensure adequate macroeconomic discipline.

Expenditure Assignment

In a decentralized system, the central government has a fundamental role as the provider of national public goods. This role includes pursuing sound macroeconomic policies, setting national standards and basic priorities, and monitoring their implementation and financing. Excessively detailed regulations, however, may be counterproductive. The design of expenditure assignment should aim at the efficient provision of public services and a balanced promotion of national and regional economic objectives. A successful reform of the present system of expenditure assignment would need to address the following weaknesses of the system.

Overlapping of Responsibilities

While the overlapping of responsibilities of different levels of government is unavoidable in practice, and the drawing of boundaries between levels of government is difficult, Korea has an overly complex system of overlapping functions (particularly in the areas of transportation, roads, and regional development). Excessive overlapping reduces accountability and responsiveness in the provision of services. To reduce such overlapping, each level of government should be assigned as clearly defined as possible a sphere of autonomy. The key objectives of this assignment should be simplicity, transparency, and administrative efficiency.

Local Autonomous Bodies’ Function as Central Government Agents

In many sectors, local autonomous bodies are acting mainly as agents of the central government, such as in the case of economic development and infrastructure investment. As previously mentioned, this situation is typical of highly centralized states. It may ensure uniformity of service provision across the various areas of a country, but excessive constraints and regulations may impede accountability for local governments by giving rise to claims by local autonomous bodies that inappropriate policies are the result of decisions imposed by the central government. Particularly problematic is the practice of mandating local autonomous bodies to carry out central government functions without providing resources. The practice of imposing unfunded mandates on local autonomous bodies reduces accountability, as well as promoting the emergence of fiscal imbalances at the subnational level.

Externalities and Spillover Effects

Expenditure assignment must be designed to address externalities and spillover effects.

  • National public goods can have negative regional externalities. For example, the Korean government recently has met strong local opposition to its plan to build a nuclear power plant. Its construction of a train system connecting Seoul and a new international airport has been delayed as a result of the lack of cooperation of some local autonomous bodies.

  • Local public goods can have both positive and negative cross-border spillover effects. Regional public goods can also have negative cross-border spillover effects. The growth of urban areas and greater mobility have increased externalities relating to certain basic services such as mass transit, water provision, and environmental protection, especially in large metropolitan areas. For example, Seoul’s subway system benefits not only Seoul’s residents, but also those who commute to Seoul from adjoining regions. New mechanisms are needed for the sharing of responsibilities for these services among affected communities.


While earmarking has its merits (for instance, protecting certain programs against arbitrary allocation of resources and instability in the provision of services), its extensive use in Korea—for transportation, road construction, and education—may result in disadvantages, such as nonadaptation of policies to local needs and excessive rigidity in the supply of services. Moreover, earmarking reduces the political accountability of elected assemblies, since the budget they approve may concern only a limited fraction of the services that are provided to their citizens.

Tax Assignment

In principle, the types of tax bases presently allocated to local autonomous bodies are in conformity with the theory, since local autonomous bodies are given autonomy to set the rates of a number of taxes. However, the local taxes are excessive in number, complicated, and inefficient. They are largely based on property, with only little emphasis on income and sales taxation. The basic objective of the reform measures should be to establish a simple and transparent local tax system that provides local autonomous bodies with sufficient and buoyant tax sources, offers the required level of local autonomy, and at the same time enhances efficiency and fairness. Since property taxes will remain the major revenue source for the foreseeable future, reform efforts should focus on this source.

Efficiency and Transparency of Local Taxes

A proliferation of local taxes, each with a relatively modest yield, has created a complex and not adequately transparent system of local taxation. This has adverse effects on the perceptibility of the true tax burden by local taxpayers and voters, and hampers the price signal effect of the taxation system that is—as described earlier—key to the functioning of a decentralized fiscal system. But even in a situation where local governments are intended mainly to carry out agent functions, the present system is unnecessarily complex, with attendant administrative and compliance costs. The fairly low revenue buoyancy results from the extensive use of property taxes, based on property values that have not been fully indexed to market prices. This has led to the proliferation of new taxes, resulting in excessive complexity of the present system. The inadequate transparency has been accentuated by the significant volatility in the revenue structure over time as a result of the frequent changes in the system, although the reforms have been motivated by a desire to improve the system and to provide subnational governments with the necessary means of satisfying growing needs. The present system is characterized by variations in the set of taxes that have been assigned to the different types of local governments (large cities, provinces, small cities and counties, and districts), even for entities at the same “level,” such as provinces and large cities. To the extent that local autonomous bodies are carrying out similar tasks, it could be argued that they should also have access to the same sources of taxes, to maximize transparency.


Although the system formally allows for variations in tax rates by local autonomous bodies, the de facto autonomy is very limited owing to the almost complete absence of incentives to deviate from standard rates. Because of strict central control on expenditures, there is very little scope for local autonomous bodies to provide their populations with services at varying levels of “quality” and “quantity” in accordance with local preferences.

Power of Local Autonomous Bodies to Grant Tax Exemptions

Most countries grant exemptions in well-defined cases to, for example, government buildings under the property tax. However, the fairly extensive power of local autonomous bodies in Korea to grant discretionary reductions and exemptions in a large number of specific cases, particularly for property taxation, involves a risk of local governments engaging in tax competition to attract businesses and economic activity, which involves loss of revenue and complicates tax administration.

Complexity of Property Taxes

Additional structural problems of the present system of property taxes include the following:

  • There is a relatively high reliance on transaction taxes as opposed to ownership taxes (about 56 percent of property tax revenues derive from transaction taxes), creating locking-in effects that hamper the efficient functioning of property markets.

  • To the extent that property taxes are levied on business property, an issue of perceptibility arises since businesses to some extent can shift the tax burden backward to labor and other inputs, or forward to consumers via higher prices; thus, it is uncertain who ultimately bears the burden of this element of the property tax. Furthermore, there is little rationale for using progressive rates for businesses (a “large” corporation may be owned by numerous individuals with large disparities in personal income levels).15

  • Infrequent valuations necessitate large changes when made, adding to taxpayer resentment.

  • Values used for taxation purposes differ significantly from market values. This might not in itself distort taxation, assuming that the ratio between valuation and market value is constant across areas and uses. However, this condition is not likely to be satisfied, in which case the present system introduces a distortion, in addition to resulting in losses of potential revenue.

Basing Income Taxes on Sources

As regards income taxes, the main problem in the present system relates to the source-based nature of the inhabitant tax. This reduces transparency and results in lack of alignment of expenditure responsibilities and tax financing. Possibly accentuating this is the fact that the inhabitant tax is also levied on businesses (including corporations), part of which may be borne by labor and other factor inputs, whereas another part is borne by the consumers of the products, further severing the relation between those who pay and those who benefit from the services.


Korea uses a fairly sophisticated formula that takes into account expenditure needs and revenue capacity, similar to the systems in force in most other developed countries.16 The system can be improved in a number of respects.

Complexity of the Formula

The formula for allocating equalization transfers should be simpler and should reflect adequately, and in a more transparent manner, major expenditure needs (such as higher unit costs in some regions). Needs in the present system are evaluated by multiplying, for each of selected categories of expenditure, a need indicator by a unit cost estimated for that category. A distinct unit cost is used for each category of local autonomous body (for example, Seoul, other large cities, and provinces), and calculated on the basis of statistical techniques. While conceptually appropriate, a proper understanding of the calculations, however, is frequently beyond the reach of many, especially the small, local autonomous bodies. Such ambiguities eventually will lead to political resentment.

Objectives of Transfers

Another problem is related to the more general one of the inadequacy between goals and instruments. Transfers are currently used in Korea to meet a number of objectives: to correct the vertical fiscal imbalance between the central government and local autonomous bodies; to equalize expenditure needs and revenue capacity between the various local jurisdictions; to foster national and regional development; to constrain subnational governments’ decision-making processes according to existing national priorities; and, finally, to increase efficiency in the provision of services and revenue-raising capacity. While most of these goals are similar to those of other countries’ transfer systems, the list of objectives seems somewhat overstretched compared with the means, and has contributed to the complexity of the distribution formula.

Use of Indicators

The formula uses a large number of indicators, but two have played a dominant role: total population and the ceiling on the number of civil servants. Total population is widely used around the world, but is a quite generic indicator of expenditure. When used for a large share of the expenditure, as in the present case, it amounts to implying that expenditure needs are nearly equal on a per capita basis across the various areas. The use of the ceiling on civil servants may also pose some problems, even if there is a ceiling on their total number for each local autonomous body. The ceiling on civil servants is largely an indicator of present capacity, rather than a measure of need. Thus, local autonomous bodies may have no incentive to reduce the number of staff to increase efficiency. The same considerations apply to the size of administrative buildings. In general, expenditure needs should be connected with specific characteristics of the targeted beneficiary population.

According to efficiency and equity criteria, expenditure needs should be compared not with effective tax revenues but with tax capacity, that is, with the revenue that could be raised by applying standard tax rates to the effective tax bases. This is only partly achieved with the present formula based on estimated tax revenue and a partial adjustment between this estimate and actual tax collections in the following year. In a given year, if a local autonomous body increases its actual tax collections, by improving tax administration or by increasing tax rates, its revenue projections are subsequently increased, thus reducing central government transfers and partly nullifying the revenue effort.

Capital and Current Expenditures

The present system is geared to provide strong incentives to infrastructure investments. Legal mandates operate in the same direction. This reflects the growth orientation of the national policy. However, the building of infrastructure will rapidly increase maintenance expenditure and the provision of related current services. Local governments are presently eager to use the funds provided by the central government for capital projects. They may, however, face shortages of funds for current expenditure in the not-too-distant future.

Use of Transfers

An important issue is the balance of the transfer system between specific and general transfers. Currently more than half of total transfers are specific transfers. Their sectoral allocation is entirely determined by the central government. While the implementation of national priorities is a legitimate goal for a transfer system, excessive constraints on the use of the funds may be counterproductive (see Chapter 4). A change of balance is therefore needed from specific to general purpose transfers, in order to strengthen local autonomy and scale back central intervention in local priorities unless there is a clear element of externalities or merit goods.

Cyclical Nature and Complexity of the Equalization Transfers

The total amount of transfers should be determined in a way that reduces the procyclical effect on local fiscal operations, for example, the total amount might continue to be determined as a fixed share of internal tax revenue, but with a floor and a ceiling. Another mechanism may be devised to smooth cyclical fluctuations in transfers.

Local Borrowing

A market-oriented system of local borrowing (for example, at market interest rates), with simpler eligibility criteria, has to be established, but within an overall ceiling set in accordance with macroeconomic stabilization objectives. The present system of borrowing has generally been successful in achieving its objectives of regulating local borrowing and in directing borrowing toward high-priority areas. There are, however, a number of potential problem areas, discussed below, that have to be addressed in the reform process. In particular, these are related to the allocation mechanism, the overall ceiling, the eligibility criteria, and the system of bond placements.

Market Orientation of Allocation Mechanism

The present system of allocating credit is not market oriented, except that debt ratios are used as one of the eligibility criteria. Increased reliance on market forces, by letting financial markets evaluate the creditworthiness of subnational governments and pricing loans accordingly, may significantly improve the allocation of credit and also impose borrowing discipline on local governments. This would require, however, that financial markets be fairly well developed and timely and reliable information be available to them on the local government finances (see Chapter 7 for a comprehensive discussion of these issues).

Overall Ceiling

Apparently, no a priori decision is made concerning the appropriate level of the overall borrowing volume, consistent with the prevailing macroeconomic targets. If the degree of local autonomy in coming years is gradually increased, the authorities will probably be confronted with stronger demands for increased local borrowing, which may well raise the question of measures to control the global volume of borrowing.

The argument that there is a need for an overall ceiling, in addition to the existing project-oriented regulations, for local borrowing is based on practical considerations. In a world of perfect capital markets where accurate foresight prevails and the externalities of local capital projects are fully taken into account, project-oriented regulations on borrowing would be sufficient to ensure the sustainability of local borrowing. In practice, however, capital market conditions are volatile, financing costs fluctuate, and externalities may not necessarily be fully taken into account. An overall ceiling for local borrowing, based on macroeconomic conditions, would help ensure the consistency between local government operations and a sustainable macroeconomic framework.

Eligibility Criteria

The complexity of the present eligibility criteria is a key issue. Associated with a possible move away from the detailed control prevalent under the agent function and toward more local autonomy, regulations should set out a general and transparent framework within which greater autonomy is provided to the local bodies in the management of their borrowing operations.

Method of Bond Placements and Use of Proceeds

The system of compulsory bond placements involves an implicit fiscal cost on the consumers and firms affected. The fiscal costs take on three forms: the forced saving implied by the system, the grace period involved, and the lower-than-market interest rates offered. Thus, the “cheap” financing is achieved by imposing a nontransparent fiscal burden on the entities purchasing the bonds, a cost that under normal conditions would be covered by explicit taxation to finance credit on market terms. The cheap credit, furthermore, involves the risk of a “money-illusion” on the part of local autonomous bodies, which may induce them to incur excessive borrowing to finance wasteful or unsustainable spending.

Institutional Arrangements for Coordination

To translate the principles discussed in preceding sections into practice, sound legal and other institutional arrangements should be established.

Role of Local Autonomous Bodies as Public Entities

The Law on Local Autonomy prohibits local autonomous bodies from conducting the functions of the central government, but leaves room for bodies to engage in activities that may be carried out more efficiently by private sector enterprises. It would be helpful for the law to state clearly that local autonomous bodies should not engage in commercial activities.

Expenditure Assignment

There is a clear need to initiate a process to establish a more transparent and consistent legal framework for the assignment of functions between the central government and local autonomous bodies, which necessarily will be a prolonged process and will require regular consultations between the central government and local autonomous bodies. The present assignment of responsibilities does not seem to derive from a single and well-defined legal framework, be it the Constitution or an act of legislation. It is rather the result of the accretion over the years of a number of sectoral laws and decrees. In this situation, officials may find it extremely difficult to understand the exact range of their responsibilities. It would be helpful to introduce a comprehensive legal framework for expenditure and tax assignments as a basis for decentralization.

Nationwide Coordinating Mechanisms

Without appropriate and effective coordinating arrangements, adverse regional externalities can result in the underprovision of a national public good. This calls for a coordinating mechanism to ensure that harmonization of national and local spending priorities takes place in a coherent way and is preceded by technical analysis and political discussions. This could be activated through a two-tier approach: (1) a political forum would decide on the framework and set the priorities, for example, through the parliament or through regular meetings between the different levels of government; and (2) a technical forum would provide the means of implementing the required measures. This latter may take the form of a technical committee with representatives from the central government and local autonomous bodies, supplemented by external experts. It may either work under the auspices of the relevant ministry or take the form of a body independent of existing political and administrative structures. While the majority of developed nations have bodies of this nature, countries have chosen very different approaches with regard to the composition and specific objectives of the coordinating bodies.

Regional Coordinating Mechanisms

Cross-border spillover effects of local public goods call for regional cooperation and coordination. A public good may provide benefits to two or three provinces. In these cases, two or more local autonomous bodies can establish supraprovincial authorities such as metropolitan area authorities or regional environmental corporations.

Main Conclusions

Korea’s intergovernmental fiscal relations are now in transition. Although the central government continues to play a key role in the fiscal operations of local autonomous bodies through regulations and the allocation of centrally controlled financial resources, politically independent local autonomous bodies are beginning to explore possibilities for increasing the provision of local public services. At this juncture, Korea faces a number of issues:

  • The present intergovernmental fiscal relations are based on a number of complex and often overlapping individual laws, regulations, and practices, rather than on a systematic legal framework that establishes a transparent and streamlined assignment of functions between the central government and local autonomous bodies. The excessive overlapping of responsibilities can result in over- or underprovision. There seems to be an inadequate operational arrangement for effective conflict resolution between the central government and local autonomous bodies.

  • There are areas where functions, authority, financing arrangements, and accountability are not adequately aligned. The financing arrangements do not sufficiently take into account the spillover effects of public goods provision between different tiers of government. Also, the financing arrangements do not adequately take into account the costs of carrying out central government functions delegated to local autonomous bodies.

  • Intergovernmental fiscal relations do not sufficiently distinguish between two objectives—an improvement of horizontal equity and compensation for performing central government functions delegated to local autonomous bodies. These result in efficiency losses. The inadequate distinction between the two objectives diminishes the effectiveness of central government transfers in promoting national policy objectives.

  • Central government control and regulations of local autonomous bodies in some areas remain extensive and may be judged excessive, particularly in such areas as budget formulation and taxation. Local autonomous bodies are still, to a large extent, considered agents of the central government. At the same time, local autonomous bodies are not exploiting potential policy flexibility (for example, flexible tax rates), and have excessive power to grant tax exemptions.

  • Local taxes are too numerous, complicated in structure, and not sufficiently buoyant with respect to income. There is excessive reliance on property taxation, which, in turn, relies too much on transaction taxes, rather than on ownership taxes, with the valuation of property outdated.

  • The present system of central government transfers to local autonomous bodies is aimed at achieving too many goals, and relies on an excessively complex allocation scheme, while not adequately taking into account some fiscal needs—such as those arising from changing demographic trends. The determination of the total amount of transfers tends to have a procyclical impact on the fiscal operations of local autonomous bodies. The present system of transfers focuses excessively on capital projects. This would later lead to high demands for recurrent expenditures on operations and maintenance.

  • Local borrowing, largely based on nonmarket principles, such as complicated administrative allocation and compulsory bond placements, can lead to excessive, poorly timed borrowing, as well as to the inefficient use of borrowed resources. The lack of an overall ceiling on local borrowing could potentially have adverse macroeconomic implications.

The authors would like to acknowledge Giorgio Brosio’s input into the analysis, as well as into the formulation of many ideas in this chapter. They also thank Teresa Ter-Minassian for valuable suggestions.

Korea’s Constitution has only two articles dealing with local autonomy. Article 117 defines briefly the functions (providing services relating to residents’ welfare and managing properties) and authority (establishing ordinances within the framework of relevant laws) of local autonomous bodies. Article 118 states that the local autonomous bodies should have their own legislative bodies. The Constitution does not provide criteria for the distribution of responsibilities either among provinces, small cities, and counties or between the large cities and their autonomous districts. The basic legislation for the establishment and operations of local autonomous bodies is provided in the Law on Local Autonomy and the Ordinance for the Implementation of the Law on Local Autonomy.

See Lee (1992) for a discussion of local public finance in Korea before the political devolution.

See Jun (1992) for a description of the evolution of the Korean intergovernmental fiscal relations between 1960 and 1992.

Article 9, Law on Local Autonomy.

Article 11.

While most of the education board expenditures are provided by the central government, Seoul and Pusan contribute partly to the operations of the education boards in their respective areas.

See Oh (1992) for a discussion of the local tax system.

The overall tax-to-GDP ratio of about 21 percent is also considerably below the ratio, even in low-tax member countries of the Organization for Economic Cooperation and Development (OECD) (for example, Australia, Japan, and the United States, where tax ratios are about 30 percent). With unweighted averages of the ratio of central government taxes to total taxes equal to about 43 percent in federal OECD countries and about 63 percent in unitary OECD countries, the large majority of these countries have allocated a significantly larger share of total tax revenues to subnational governments than has Korea.

This holds for the property tax, the automobile tax (only indirectly related to value), the global land tax, and the community facility tax. In contrast, the city planning tax on land and buildings has a flat standard rate. The acquisition tax and the registration tax are mainly flat rate taxes, but with numerous different rates applied to different types of assets.

Variations in tax rates are permitted for the following taxes, with the allowed range of variation around the standard rate shown in parentheses: inhabitant tax (plus and minus 50 percent); automobile tax (plus 50 percent); livestock tax (minus 1 percent); city planning tax (plus 50 percent); regional development tax (plus and minus 50 percent); and community facilities tax (plus 1 percent).

Less strict criteria apply to public enterprises.

This criterion is nor defined clearly.

See Tanzi (1996) for a discussion of fiscal decentralization versus administrative decentralization of intergovernmental fiscal relations.

Furthermore, progressive property taxes on businesses create an incentive for splitting of businesses.

In fact, it can be shown that the technique applied in Korea of calculating transfers by comparing “standard” revenue with “standard” expenditures for individual subnational governments is identical, in principle, to a system that applies separate equalization schemes to tax capacities and to expenditure needs, respectively.


    Other Resources Citing This Publication