Japan: Selected Issues

This Selected Issues paper analyzes Japan’s medium- and long-term fiscal challenges. It discusses the initiatives that will be necessary to cope with Japan’s medium- and long-term fiscal challenges. It describes medium- and long-term projections of the fiscal position, and assesses the size of the consolidation measures needed to restore long-term fiscal sustainability. The paper explores possible options for consolidation and offers a representative package of such measures. The paper also describes the projected baseline path for the long-term fiscal balance, incorporating the long-term pension reform plan formulated in 1994.


This Selected Issues paper analyzes Japan’s medium- and long-term fiscal challenges. It discusses the initiatives that will be necessary to cope with Japan’s medium- and long-term fiscal challenges. It describes medium- and long-term projections of the fiscal position, and assesses the size of the consolidation measures needed to restore long-term fiscal sustainability. The paper explores possible options for consolidation and offers a representative package of such measures. The paper also describes the projected baseline path for the long-term fiscal balance, incorporating the long-term pension reform plan formulated in 1994.

IV. Intergovernmental Relations and Local Public Finance in Japan 1/

1. Introduction

After almost half a century of status quo, decentralization has become a major issue on Japan’s political and economic agenda in recent years. Soon after his inauguration, Prime Minister Hashimoto outlined four challenges that Japan must tackle: deregulation, administrative reform, devolution of authority, and moving the capital. These challenges are deeply interrelated. The local public sector, which presently accounts for about 16 percent of gross domestic product and employs over 3 million people, is one of the most regulated sectors in the Japanese economy. The Central Government exercises very detailed and stringent controls over the whole of the revenues of local governments and a wide range of local expenditures. As a result, the degree of effective revenue and expenditure centralization in Japan is among the highest in industrialized countries.

There seems to be broad agreement in both the government and the private sector that the rigid system of local public finance and intergovernmental relations needs to be reformed, but consensus on the direction of reforms has yet to emerge. Meanwhile, pressures for decentralization stemming from the need for fiscal consolidation and the process of deregulation and structural reform, as well as from the wave of decentralization efforts in other industrial countries, are rising. In theory, fiscal decentralization is likely to improve welfare because central government control interferes with the efficient supply of local public goods and services, and may impede the incentive for local governments to act responsibly. Local governments in Japan also show a high degree of “grant dependency,” which in the fiscal federalism literature has been associated with an inefficient size of the public sector and distortions in the allocation and financing of public goods.

This chapter analyzes the system of local public finance and intergovernmental relations in Japan with a view to assessing the strengths and weaknesses of this system and identifying, in broad terms, possible approaches to fiscal decentralization. Section 2 provides a historical overview of the Japanese system of local government. Section 3 describes the key constitutional, socio-political, and organizational aspects of the system that has emerged after the Second World War. The basic structure of the present system of intergovernmental relations, including the assignment of expenditure and revenue functions and the grants system, is analyzed in Section 4. Section 5 summarizes the views of Japanese policymakers and public finance experts on the problems of local public finances. Section 6 draws on an extensive theoretical and empirical literature to identify the main areas for reform of local public finances and intergovernmental relations in Japan. Finally, Section 7 concludes with an outline of possible approaches to fiscal decentralization.

2. Historical background

Prior to World War II, local governments in Japan had little or no independence and were generally treated as administrative arms of the national government. 1/ There was no basic constitutional recognition of local autonomy, and the legal existence of local governments rested on a few basic statutes. Governors of prefectures were local agents of the national government appointed by the Emperor and responsible to him. All local governments were supervised by the powerful Ministry of Home Affairs on most matters, and by other ministries (e.g., Transport, Education, Agriculture) where relevant.

Under a complex set of nationally established rules, local governments were responsible for sharing the costs of all national government programs, while having little or nothing to say about the policies which defined them, or even the manner of their implementation. Municipalities were fully subordinated to prefectural governors in all but the most minor local issues. The electoral base was very narrow, being primarily male land or property owners. 2/ Political leadership tended to be an elite of minor nobility and landowners who viewed themselves not as public servants but as officials of the crown.

The process of urbanization fostered by the national government from the 1880s up to World War II began inevitably to force the pace of urbanization. As millions of people left the countryside, the old-style concepts of paternalistic local government proved inadequate to meet the burgeoning demands for housing, sanitation, health care, transportation, electric power, and other elements of modern infrastructure. Local authorities did not have the revenue resources to meet these growing needs, and they soon realized that they had to force the national government to pay for its own demands. At the same time, a broader base of educated people began to insist on a greater political voice in their own affairs, and were able to exert more direct influence on the political elite which ran the national government. These trends were interrupted by the Second World War, but they were a powerful influence toward democratization which established a climate for the rethinking of Japanese government.

3. Intergovernmental relations in modern Japan

a. Constitutional aspects

The new Constitution of 1947 emphasized the basic principle of democratic government and fundamental rights, drawing heavily from European and American traditions. It established for the first time the principle of a government of the people, in contrast to a government loyal to the Emperor. Japan committed itself to a structure that carried representative government down to the prefectural level and cities, towns, and villages. The Constitution itself gave a recognition to these bodies equal to that of the national government, and it further stipulated that:

  • Regulations concerning organization and operations of local public entities shall be fixed in accordance with the principle of local autonomy (Article 92).

The Local Autonomy Law (1947) defined a two-tiered structure of local government consisting of prefectures and municipalities. While each level has equal status under the Constitution, national policies have continued to be paramount, and the national government retains a superior position:

  • Each local public body shall, in addition to its own community affairs, and the affairs required by national law or by cabinet order, perform other administrative affairs within its area insofar as such affairs are not reserved to the State (i.e., the national government) (Article 12).

Local authority in Japan is thus conveyed in a form directly opposite to the concept of residual states’ rights in federations—that is, specific authorities not given to the national government are reserved to the states. In contrast, the Local Autonomy Law makes specific designations of roles that the national government authorizes local governments to exercise.

In addition to its authorizing role, the Local Autonomy Law defines the authorities of local legislative bodies and local officials, establishes many uniform rules which must be obeyed, and final approvals which are retained by the national government. 1/ The Local Autonomy Law also cites some 550 other national laws that define an estimated 70-80 percent of all the activities of local governments. 2/ The system which emerged is very “public administration” oriented, in the sense that it contains many elements that also define how the administration or management of public functions is to be carried out (Bingman (1989)).

One of the more striking features of Japan’s system in this regard is the mandamus proceedings. The relevant minister in the national government can directly order a prefectural governor—who is an elected official—to carry out certain actions, and if the governor does not obey such orders, he or she can be removed from office by the national minister, subject to certain legal appeals. 1/

b. Socio-political aspects

The defining characteristic of Japan’s system of intergovernmental relations is the strong collective preference for equal access to public goods. The Japanese people and government were willing to commit themselves after World War II to the evolution of autonomy for their newly defined structure of local government. However, they would not do so in a pattern that would reject their history and tradition of strong central government and allow for substantial regional differences to emerge. Equal access to public goods and fair sharing of the burden to finance these goods were viewed as essential for economic and social development. Hence, local governments were willing to sacrifice autonomy to maintain regional equity.

To implement these principles, local autonomy would have to function within a framework of uniform structure and ground-rules defined by the central government. Interregional redistribution—the extent to which populations of different regions are supplied with the same level of public services and subjected to same burdens of taxation—is, therefore, the central issue for Japan’s system of intergovernmental fiscal relations. In this respect, Japan is similar to Germany, where the constitution guarantees “equality of living standards” to all Germans, and unlike the United States, where the federal government pursues a largely “hands-off” regional policy.

c. Organization

There are currently 47 prefectures and 3,281 municipalities in Japan. Forty-three prefectures and four equivalent bodies (Tokyo, Osaka, Kyoto, and Hokkaido) correspond to U.S. states, while three large cities have been given special authorities equivalent to prefectures. Municipalities include cities, “designated” cities (those with populations of 500,000 or more), towns, and villages. Currently there are 10 designated cities, 663 cities, 1,994 towns, and 577 villages. In addition, there are several thousand “special” local public entities, including the 23 wards of metropolitan Tokyo and over 7,000 functional associations of public bodies such as property wards and unions of local public enterprises.

The basic structure and authorities of local governments have not been defined by those governments themselves, but by a series of national laws that mandate essentially the same institutional setup at all three levels. Each prefecture has a bicameral legislative assembly, whose members are elected for four-year terms. In general, all activities of the prefecture are under the direct authority of the Governor as chief executive. Governors submit a high proportion of new legislation each year (including the budget) and have veto powers on bills. Governors appoint and supervise all public employees, and organize and manage the departments of the government, including ownership and management of all public facilities. They may also let contracts or establish and supervise public enterprises to carry out the public’s business. The assembly may pass a resolution of non- confidence by vote of two thirds of the membership. Governors, in turn may dissolve the assembly.

Municipal assemblies have basically the same structure as prefectural assemblies, and mayors as chief executives have similar duties and powers as governors of prefectures.

4. Basic structure of intergovernmental finances

a. Vertical fiscal imbalance

Financial relations between the national and local governments in Japan are marked by a substantial vertical fiscal imbalance—i.e., revenues generally exceed expenditures at the central level and fall short at the local level. The central government receives tax revenue that typically exceeds by 15 percent the amount that it spends for its activities (see tabulation below). By contrast, the tax revenue of local governments amounts to less than 40 percent of the funds necessary to perform their functions.

Ratio of Own Tax Revenue to Own Expenditure 1/

(In percent)

article image
Source: Ministry of Home Affairs (1994).

Own expenditure does not include transfers to local (central) governments. Own revenue includes all revenue collected by the central government (i.e., before revenue sharing with local governments).

To fill the financing gap at the local level, the central government transfers part of its “excess” tax revenue to local governments; these transfers amount to about 35 percent of local revenue (6 1/2 percent of GDP). Prefectures and municipalities raise the remaining funds through nontax revenue and borrowing.

b. Division of expenditure responsibilities

The vertical fiscal imbalance results from the division of expenditure responsibilities and revenue raising powers between the central and local governments. In Japan, the central government directly performs relatively few public functions, including: defense, national law and order, judiciary affairs, national transportation systems, postal service, national hospitals and medical facilities, and institutions of higher education and research. Local governments are responsible for a major share of public spending, including on education and culture, infrastructure, health and welfare, and public law and order. On average, central government expenditure accounts for about 10 percent of GDP, and local expenditure for about 18 percent of GDP (see tabulation below and Table IV. 1 in the appendix to this chapter).

Table IV.1.

Japan: Central and Local Government Expenditure, FY 1992

article image
Source: Ministry of Home Affairs (1995).

Shares of Central and Local Expenditure 1/

(In percent of total expenditure)

article image
Sources: Ministry of Home Affairs (1994); and national accounts.

Including expenditure financed by transfers from central/local governments.

GDP shares in all tabulations in this chapter are calculated with respect to fiscal year GDP.

At the local level, municipalities are presently responsible for slightly over half of total expenditure (see tabulation below). The relative share of municipal spending increased substantially in the 1970s and remained relatively stable at about 51 percent of local spending since.

Shares of Prefectural and Municipal Expenditure 1/

(In percent of local expenditure)

article image
Sources: Jichi Sogo Center (1993); and national accounts.

Including expenditure financed by transfers from higher-level governments; expenditure by prefectures excludes transfers to municipalities.

The average is taken over 1970, 1975, 1980, and 1985-90.

Regarding the nature of local expenditure, about two thirds is current spending and the rest is capital spending. The largest expenditure functions of local governments are education and culture (26 percent of local spending), infrastructure (21 percent), health and welfare (17 percent), and law and order (9 percent). 1/

Although considerably more public spending takes place at the local level than at the national level, the national government remains heavily involved in almost every aspect of local public spending. Unlike in most unitary states and federations, the Local Autonomy Law does not clearly distinguish between the responsibilities of the central government and those of local governments. Rather, the Law provides that prefectures perform those functions which “require uniformity in performance, cover a wide geographical area, are deemed too extensive for management by municipalities, or require efforts to coordinate two or more cities, towns, or villages.” Major programs (education, health, infrastructure) are in practice formulated by national ministries and financed directly or indirectly by the central government. National ministries also retain numerous bureaucratic authorities with respect to local governments. 1/ Local governments in principle have large administrative responsibilities—over planning, construction and maintenance of facilities, regulation and zoning, management, inspection, and monitoring of compliance—but in practice national direction can be so detailed and so intrusive that it can usurp the authority of local officials and leave them with the job—but not the power—to satisfy public needs.

c. Revenue assignment

As noted above, the main sources of local revenue are local taxes (on average, 37 percent of local revenue), transfers from the central government (35 percent), nontax revenue (20 percent), and bonds (9 percent). Excluding transfers, 16cal governments’ own revenue accounts for about 42 percent of consolidated government revenue (11 percent of GDP) (see tabulation below). Local governments have increased their share in consolidated revenue since the early 1980s.

Shares of Central and Local Revenue 1/

(In percent of total revenue)

article image
Sources: Jichi Sogo Center (1993); and national accounts.

Includes tax revenue, nontax revenue, and borrowing; local revenue does not include transfers from the central government.

The average is taken over 1970, 1975, 1980, and 1985-90.

At the local level, prefectures raise about 55 percent of revenue (9 1/2 percent of GDP) (see tabulation below). While the share of local taxes in revenues of prefectures and municipalities is about the same (35 percent of prefectural/municipal revenue), municipalities raise relatively more funds from nontax sources (20 percent of total revenue, compared with 15 percent for prefectures) and local bonds (10 percent vs. 7 1/2 percent for prefectures). Prefectures receive more transfers from the central government (42 percent vs. 36 percent for municipalities). Details of the assignment of major national and local taxes are provided in the appendix to this chapter.

Shares of Prefectural and Municipal Revenue 1/

(In percent of local revenue)

article image
Sources: Jichi Sogo Center (1993); and national accounts.

Includes transfers from the central government. Municipal revenue does not include transfers from prefectures.

The average is taken over 1970, 1975, 1980, and 1985-90.

Although local taxes represent one of the most important revenue sources for local governments, the tax bases and tax rates cannot be determined by the independent initiative of local governments. The Local Tax Law—which was, significantly, enacted by the National Diet—prescribes in great detail the taxes that can be imposed by local governments, defines the tax base in each case, and determines the tax rates. While local governments are given some flexibility with certain taxes, any discretion that local governments may have must be exercised with the approval or informal agreement of the central government.

d. Grants and fiscal equalization

To transfer its “excess” revenue to local governments, the central government uses revenue sharing grants, referred to in Japan as the “local allocation tax,” and a large number of specific purpose grants. The local allocation tax accounts for 50-60 percent of all transfers to local governments (15-20 percent of local revenue). Central government disbursements account for 45-60 percent of central government transfers (up to 15 percent of local revenue), while local transfer taxes account for 3-7 percent of all transfers (up to 2 percent of local revenue). 1/ Details of the grants system are discussed in the appendix to this chapter.

A major purpose of the local allocation tax is to equalize the fiscal capacity of local governments. The Japanese equalization model takes into account the inherent differences in both revenue raising capacities and expenditure needs of local governments, approach is very similar to General Revenue Assistance in Australia. A prefecture or municipality that enjoys a high revenue raising capacity and has a low cost of providing services is considered fiscally strong, and vice versa. The index of fiscal capacity and, hence, the amount of equalization payment depends on the ratio of “basic financial needs” over “basic financial revenue.” Where basic revenue exceeds basic financial needs (this is generally the case in large metropolitan areas), the local government receives no transfer but is entitled to retain its excess. By contrast, low-capacity governments (generally those situated in rural areas) receive large per capita equalization payments.

Basic financial need is not equivalent to the actual expenditure of a local government, but rather represents a standardized amount deemed necessary to provide public services at the level prescribed by the national government. The basic financial need for each public service is calculated according to the following formula:

Basic financial need = Unit of measurement × Unit cost × Cost differential

The total financial need of a local government is the sum of the basic financial needs for all expenditure items. The unit of measurement in the above formula is generally local population, or length and area of roads or other public facilities. Unit costs for public services are determined each year by the national ministries, taking into account price changes and changes in the demand for services. Finally, cost differentials depend on factors such as population density and growth, climate, area and geography, degree of urbanization and industrial diversification, etc. 2/ These factors reflect the differential costs of providing standard services that arise due to unavoidable circumstances in which local governments operate.

Basic financial revenue is defined as general revenue that can be appropriated to meet the basic financial need. It is calculated as the sum of the local transfer tax and a prescribed percentage of the “standardized” local revenue—80 percent of such revenue for prefectures, and 75 percent for municipalities. 1/ Standardized revenue is the local tax revenue that can be obtained under the “standard” tax rate determined by the central government. If a local government levies taxes at lower rates and experiences a revenue shortfall, its basic financial revenue will not be considered inadequate (such a government may, therefore, receive no transfer). Conversely, if a local government levies taxes at higher than standard rates, it will not be penalized—the amount of transfer will be calculated as if the local government applied the standard rate.

The equalization impact of the local allocation tax is quite strong. For example, in FY 1992, per capita local tax revenue in Tokyo (the wealthiest prefecture) was ¥262,000, but only ¥61,000 in Okinawa, the poorest prefecture. After the transfer of the local allocation tax, however, per capita revenue in Okinawa rose to ¥217,000.

The main objective of specific purpose grants is to achieve uniformity in the provision of local public services. Unlike the local allocation tax, most of the national specific purpose grants are allocated among local governments at the discretion of the central government; there are only a few formula-based grants. Specific purpose grants are mainly designed as cost-sharing programs, i.e., they subsidize a certain percentage of the standard cost prescribed by the national government. The rate of subsidy differs from program to program, depending on the financial burden of the program and the interest the central government has in the program. Almost 35 percent of the total amount of specific purpose grants are subsidies for public construction projects.

The system of specific purpose grants creates a close link between corresponding departments of national and local governments. For example, an officer of in the public works department of a local government may come under the control of an officer of the Ministry of Construction, and that control may be more direct than that exerted by the head of the local government body itself.

e. Borrowing

Local borrowing accounts for about 9 percent of local revenue. The approach to local public finance in Japan has regarded borrowing as a source of funds to be resorted to only in case of special financial need (Yonehara (1981), Ishihara (1993)). Because it gives rise to future financial claims against the local government, borrowing is generally limited to capital outlays. Local governments are prohibited from raising loans without the approval of the central government. The Local Loan Program, drawn up by the national government each year, lays down the rules governing approvals for local borrowing. This program is drawn up at the same time as Local Public Finance Program, which determines local governments’ shares of the local allocation tax and specific purpose grants.

The main creditors of local governments are the central government, public enterprise loan funds, and the private sector. The central government funds, which provided finance for about 60 percent of local government borrowing in the late 1980s, were provided through the Fiscal Investment and Loan Program (FILP), i.e., indirectly from Post Office savings accounts, Post Office Life Insurance, and Annuity Systems (Ministry of Home Affairs (1995)). The Japan Finance Corporation for Municipal Enterprises’ Fund (JFM) lends both short-term and long-term funds to local authorities which run public enterprises. The JFM financed about 12 percent of local borrowing in the late 1980s, mainly by issuing its own bonds, which are held by either securities companies and other financial institutions, or organizations with special links to local governments (e.g., the Local Government Employees Mutual Aid Corporation).

As for private funds, which financed about 28 percent of local borrowing in the late 1980s, a number of large local governments (including Tokyo Metropolis and Osaka Prefecture and City) sold their bonds in the open bond market, or floated loans in foreign countries. Most local authorities, however, borrowed long-term funds from commercial banks, insurance companies, and agricultural cooperatives (sometimes without issuing marketable bonds), as well as from the central government and the JFM. The FILP and the JFM lend to local government at a lower rate and for a longer term than private financial institutions. Consequently, local governments see more advantage in borrowing from these sources. This is one of the reasons why an allocation of funds by the national government is necessary in connection with local loans (discussed below).

In recent years, there has been an increase in the use of local loans to assist in the implementation of fiscal policy. Given the relative size of the local public sector, the central government must rely on local governments to implement its fiscal policy. It is impracticable to alter the local tax system or the local allocation tax to compensate for changes arising out of the business cycle. The local tax system and the proportion of national taxes allocated to local governments are fixed by statutes, which can be amended only by the Diet. The size of local borrowing, on the other hand, can be changed by executive action, and such changes are accompanied by an easing or tightening of fiscal policy. In view of this flexibility, the proportion of local loans in local revenue tends to increase in years when the national government adopts expansionary fiscal policy, and vice versa.

f. Local public finance program

The involvement of the central government in the determination of local revenue and expenditure makes the budgets of the national and local governments in Japan much more closely linked than is the case in other unitary states. But the way that the vertical fiscal imbalance is resolved through grants and borrowing links the national and local budgets even more closely together, into what is effectively a single budget entity.

The whole complex set of financial relationships between the central and local governments is captured in detail in what is known as the Local Public Finance Program. It is assembled by the Ministry of Home Affairs, but the real power resides in the Ministry of Finance and the Cabinet. Each year, the national government makes an official estimate of the revenues and expenditures of local governments. On the revenue side, this estimate covers revenues from local taxes and nontax revenues, local allocation tax, local transfer taxes, the national specific purpose grants, and local loans. On the expenditure side, the estimate covers wages and salaries, expenditure on goods and services, capital outlays, interest payments, and subsidies to public corporations. These estimate are then combined with an array of national government policies and programs such as the Fiscal Investment and Loan Program in order to produce, for each local expenditure program, an “acceptable” income-expenditure plan. The most important feature of this process is how revenues and expenditures are balanced through the local allocation tax and local borrowing.

The Local Public Finance Program is really a budget, and from a public choice perspective the program is subject to all of the vagaries of the budget process. But when the program is approved by the Cabinet, it locks in local governments in terms of their own local taxes, their local loan programs, their employment and general administrative expenses, and most importantly, the levels of funding for almost all public programs.

5. Reform initiatives and views on fiscal decentralization in Japan

a. Government reform initiatives

The last major restructuring of Japanese local finances took place in the early 1950s, partly as a result of recommendations by the Shoup Commission. 1/ In the mid-1950s, the local public finance system nearly collapsed under the burden of heavy borrowing to finance the rapidly rising local budget deficit, which prompted the central government to make a special temporary grant and create the system of fiscal equalization grants in 1954. Owing to buoyant economic growth and more stable revenues, the financial situation of local governments improved considerably between the mid-1950s and mid-1970s. Since then, however, both the central and local governments have been faced with declining revenue buoyancies and expanding expenditure requirements.

Most recently, the recession of 1992-95 has led to a sharp slowdown in the growth of central and local revenues, while at the same time giving rise to the need for fiscal expenditures to stimulate the economy. As a temporary measure, the Central Government allowed local governments to issue far more honds than in “ordinary” years, and encouraged them at the same time to increase their capital spending to offset weak private spending. To place the additional bonds, local governments had to tap private sector funds, hence raising their indebtedness. The reform of local public finances has thus become intertwined with, and dependent upon the progress of, overall fiscal consolidation.

Partly in response to the growing strains of local public finance, in 1993 the Diet passed a resolution calling for greater decentralization, and in 1995, the Murayama coalition government enacted the Decentralization Promotion Law, which obliged the Government to draw up a five-year decentralization plan and set up a committee in the Prime Minister’s Office to advise on and monitor the decentralization process. While it represents an important step in political and administrative terms, the Decentralization Promotion Law remains rather vague on the overall objectives of decentralization and the methods for their implementation. Recent pronouncements by the Government suggest that future reform initiatives are likely to focus on administrative reforms, in particular, further rationalization of expenditures. Such reforms may also be necessary in order to mobilize public support for the increase in the consumption tax rate from 3 percent to 5 percent in 1997.

Other recent reform initiatives include: promotion of so-called hometown development projects (locally designed and implemented regional development projects); greater involvement of local governments in the design of social welfare programs that correspond to local needs and circumstances; and the debate on the relocation of the capital, which has highlighted the need to devolve authority to the local level and enhance the transparency, simplicity, and efficiency of government. 1/

b. Views on fiscal decentralization in Japan

Despite the slow progress toward fiscal decentralization, there has been a lively public debate on the reform of intergovernmental fiscal relations. In 1979, the Local Government System Research Council, an advisory organ of the Prime Minister’s Office, proposed decentralization for the first time in its thirty-year history. The council noted that the Ministry of Home Affairs and the Ministry of Finance shared a common interest, although from different standpoints (Shindo (1984)). The Ministry of Home Affairs was keen on promoting decentralization in order to respond to the growing demands by the public for a more responsible and flexible local government that would be able to address such issues as urban congestion and environmental problems. At the same time, the Ministry of Finance, facing the urgent need to reduce the administrative expenses of national ministries, was searching for government organs and agencies to which national administrative affairs and projects could be transferred. In this context, the Ministry of Finance did not oppose decentralization initiatives.

More recently, the Ministry of Home Affairs has argued that the most significant theme for both central and local governments for the foreseeable future would be obtaining the cooperation and understanding of the general public for higher taxes (Ministry of Home Affairs (1984)). In the view of this Ministry, the level of Japan’s public services compared favorably with that of major other countries; however, the average Japanese citizen’s tax burden was still among the lowest among developed countries. The Ministry therefore proposed measures to increase local tax revenue from existing taxes, reduce tax incentives for large corporations and capital income, and increase the reliance on user charges.

Opinions about fiscal decentralization have differed sharply at the local government level. Financially stronger governments have favored decentralization as a way of mobilizing larger resources for their growing expenditures. They have demanded greater flexibility to determine both revenues and expenditures, and have generally been in favor of greater freedom of initiative in policymaking. Financially weaker local governments have been concerned that any shift in the regime might decrease the local allocation tax to which they would be entitled in the future. Therefore, they were in favor of reforms that would result in either larger transfers from the central government or smaller expenditure responsibilities for local governments, noting that, for the public bodies lacking the ability to raise revenues, decentralization may mean the “freedom to go bankrupt.” Meanwhile, the Central Government, reflecting the resistance of the Japanese public to horizontal inequality, was concerned that fiscal decentralization might increase regional fiscal disparities, and eventually increase the need for resource transfers to local governments.

In contrast to the government, Japanese industry and financial institutions have generally associated decentralization with smaller, more efficient government, emphasizing the need to reduce the tax burden and improve public infrastructure. Opinions about fiscal decentralization have been also sharply divided among public finance experts. While many have pointed to the need to restore local self-government as a means of enhancing government responsiveness to citizens’ needs and improving efficiency of resource allocation, others have questioned the emphasis on individual preferences and local accountability, given the widely shared ideas of fairness in Japan. 1/ Finally, the opposition parties recently raised decentralization in political debate, arguing that the clear division of responsibilities between different levels of government was essential for reinvigorating the Japanese society and economy (Ozawa (1996)).

6. Theoretical perspectives on fiscal decentralization

This section examines selected aspects of the theoretical and empirical literature on intergovernmental fiscal relations that are potentially relevant for reform of local public finances in Japan. The section first reviews arguments for decentralization in unitary states, and then discusses recent empirical evidence on the effects of expenditure and revenue decentralization, and the impact of various grant programs.

a. Assignment of fiscal policy functions

The traditional argument in favor of decentralization builds on the standard welfare theory, which provides a rationale for decentralized resource allocation (competitive equilibrium is Pareto efficient) as well as government intervention (existence of market failures). The assumption that the government acts as a single agent is relaxed by assuming that several levels of government may be created in order to carry out efficient interventions. Questions that arise are, then, whether government intervention should be centralized or decentralized; and if decentralized, at what level the intervention should take place. For the assignment of major fiscal policy functions—stabilization, redistribution, and resource allocation—Musgrave (1959) derived a strong normative conclusion: stabilization (including growth promotion) and redistribution can be properly assigned only to the central level, while allocative functions can be performed at the local level.

The case for centralized stabilization policy would be hard to question in the case of Japan in view of its unitary structure of government and highly integrated national economy. Similarly, the case for centralized redistribution policy in Japan is relatively straightforward, given the high degree of homogeneity in preferences for core public services, the mix of taxes, and the degree of inter-personal and inter-regional redistribution that has evolved over time. 1/

The economic rationale for decentralization of the allocation function is based on observations about the spatial characteristics of public goods—from local public goods such as playgrounds and public libraries to national public goods such as defense and foreign affairs—heterogeneity of and imperfect information about citizens’ preferences, immobility or imperfect mobility of production factors, and the political economy and other benefits of interjurisdictional competition. The fiscal federalism literature has argued that, under these conditions, a decentralized government can supply public goods demanded by the citizens more efficiently (Oates (1972)). Whereas a centralized government is constrained to adopt policies that apply more or less uniformly across the entire population, a decentralized government can vary the contents of tax-cum-services packages across jurisdictions in response to citizens’ preferences. Even where the benefits of localized provision of public goods spill out to other jurisdictions (e.g., in the case of provision of public roads and education), the costs of coordination (subsidies from a central fisc, or direct negotiation between jurisdictions) are presumed to be smaller than would be the efficiency losses from having a higher level of government provide the service uniformly across all jurisdictions.

From a political perspective, it should be noted that the benefits of decentralization should not be viewed as an argument in favor of federalism, but rather as the formalization of an essentially economic criterion against which alternative models of government may be judged; indeed, the decentralization thesis is entirely compatible with a unitary system of government. The decentralization thesis does not specify the procedures or mechanisms of a federal system—the implicit argument in much of the traditional fiscal federalism literature is that mobility of voters/taxpayers, together with competition among local governments for tax bases and voters/citizens, is sufficient to implement the idealized outcome characterized by the decentralization thesis (Hamlin (1991)). However, there can be no guarantee that these procedural mechanisms will implement the efficient solution, whether in federations or in unitary states.

As discussed in previous sections, the allocation function in Japan remains highly centralized. To better illustrate the case for decentralization of this function, the following subsection discusses in more detail the theoretical underpinnings of the decentralization hypothesis and the empirical evidence on its effects.

b. Decentralization of expenditure functions

One of the clearest statements of the decentralization hypothesis was provided by Brennan and Buchanan (1980): the greater the extent to which taxes and expenditures are decentralized, the smaller should be, ceteris paribus, the overall involvement of the government in the economy. The first line of argument in support of this hypothesis is based on the premise of greater political competition between various fiscal jurisdictions (Tiebout (1956)). Thus, in a centralized setting, the monopoly power to extract resources through tax legislation is less challenged than within a decentralized, multilayer government, where citizens’ access to comparative political “shopping” introduces a dimension of contestability on the political markets. This argument is believed to be particularly relevant at the local level (municipalities, counties) because of the influence of alternative tax-cum-services packages on the locational choice of residents and businesses.

The second line of argument focusses on the status of the budget constraint for different levels of government. In general, all levels of government are involved in allocation, but the opportunity for deficit spending through public borrowing and seignorage creates a softer budget constraint for the central government than for lower levels of government. In a representative democracy, it is expected that the legislative representatives, after election, will articulate the fiscal preferences of their voters (including special interest groups) to safeguard their reelection. Under a soft budget constraint, the proportion of budgetary means available for discretionary use can be generally expected to be higher, and the political decisionmakers should, therefore, be able to give in more easily to the demands of interest groups and bureaucrats, as generally only a part of the tax bill has to be presented to voters/taxpayers. Therefore, one may expect that, other things being equal, the size of the public sector will be larger in a centralized setting.

There has been quite strong empirical support for the decentralization hypothesis. Moesen (1993) showed in a recent study commissioned by the European Union that extends earlier work in this area (see, e.g., Saunders (1988), and Moesen and Van Rompuy (1990)), that an increase in the ratio of central government revenue to GNP by 1 percent raises the overall size of the public sector in industrial countries by 0.18 percent of GNP, while an increase in the ratio of central government expenditure to GNP of 1 percent raises the size of general government expenditure by 0.22 percent of GNP. 1/ As expected, the degree of centralization is also found to have a strong positive effect on expenditure for transfers and subsidies, a finding consistent with the decentralization hypothesis.

The staff’s estimates of the relationship between historical changes in the degree of centralization and the size of the public sector in Japan yielded the following equations:


where the dependent variable, TEG, is total government expenditure (central plus local) in percent of GDP; LDV is five-period moving average of the dependent variable; CENR is a measure of revenue centralization, i.e., total revenue of central government (before transfers to local governments) in percent of total revenue (central plus local); and CENE is a measure of expenditure centralization, i.e., total expenditure of central government (including transfers to local governments) in percent of total expenditure (central plus local); t-statistics are shown in parentheses. Both equations were estimated by OLS on the FY 1970-93 (settlement) data. 1/

The above regressions suggest that a 1 percent increase in the revenue centralization ratio increases the size of Japan’s public sector by 0.52 percent of GDP, while a 1 percent increase in the expenditure centralization ratio increases the size of the public sector by 0.72 percent of GDP. These results indicate that the tendency of general government to expand as a result of higher revenue and expenditure centralization is much stronger in Japan than in other industrial countries.

Subsidiary regressions, where the above explanatory variables are regressed on the transfers-to-GDP ratio, indicate that these expansionary effects operate to an important degree through transfers to local governments: a 1 percent increase in the revenue centralization ratio increases the size of transfers to local governments by 0.2 percent of GDP; while a 1 percent increase in the expenditure centralization ratio increases the size of these transfers by 0.36 percent of GDP. Again, these effects are considerably stronger than those obtained by Hoesen (1993) and Moesen and Van Rompuy (1990) for industrial countries other than Japan. 2/

c. Revenue decentralization

Starting from the three-way classification of fiscal policy functions, the traditional fiscal federalism theory developed the following tax assignment criteria (Musgrave (1983):

(1) The central government should control taxes suitable for macroeconomic stabilization, and lower-level governments taxes that are cyclically stable;

(2) The central government should control redistributional taxes with progressive tax schedules;

(3) Personal taxes with progressive rates should be levied by the jurisdictions most capable of implementing a tax on a global base;

(4) Lower-level governments should control tax bases that are relatively immobile between jurisdictions;

(5) Tax bases distributed highly unequally between jurisdictions should be centralized; and

(6) All levels of government may apply benefit taxes and user charges.

These principles imply the following assignment of taxes:

  • Central government—integrated income tax, consumption tax, natural resource tax, user charges;

  • Intermediate-level governments—resident and nonresident income tax, destination-based consumption tax, natural resource tax, user charges;

  • Local governments—property tax, payroll tax, user charges.

For dual assignment of taxes (income, consumption, natural resources), Musgrave suggested overlapping rather than separate tax bases. Given that income taxes and broad-based consumption taxes are by far the most revenue elastic, this set of rules produces a high degree of tax centralization.

The public choice literature, which is generally concerned with the potential of governments for revenue maximizing behavior, emphasized costs of tax harmonization and benefits of competition for access to tax bases, and suggested more decentralized tax assignment criteria (Brennan and Buchanan (1980)). Subsequent research has tended to reduce the validity of the case for tax centralization, partly because of a more favorable academic perception of the competitive process, and partly because more sophisticated empirical studies suggested that the quantitative importance of the potentially distorting effects from unharmonized tax structures—i.e., reduced tax neutrality, increased administrative and compliance costs, and tax exporting—was exaggerated in the traditional literature. In particular, the empirical work has shown that differences in taxes were generally too small an element in business costs (or individuals’ locational decisions); that tax differentials reflecting differential quality and costs in local public services were not necessarily perceived as inefficient; and that subnational tax policymakers often had real fear of the potential effects of tax competition and, hence, were not keen on creating tax havens. 1/ Of particular interest for Japan may be the evidence that the allocative distortions arising from (local) income tax competition are of minor importance: in a detailed study of the Swiss data, Feld, Kirchgassner, and Pommerehne (1995) found that the proximity of tax havens had no impact on the spatial distribution of high-income taxpayers.

d. Intergovernmental grants and fiscal equalization

The fiscal federalism literature has identified three broad purposes for intergovernmental grants in federal systems: to subsidize, specific categories of sub-national expenditure where there are spillover benefits to those outside the jurisdiction; to equalize fiscal capacity between sub-national jurisdictions; and to reap the benefits of centralized collection of some major tax revenues (King (1984)). The first purpose gives rise to a case for specific-purpose, open-ended, matching grants; the second purpose gives rise to a case for lump-sum unconditional grants; and the third purpose requires unconditional revenue sharing grants.

Intergovernmental grants are seen in the traditional fiscal federalism literature as a reasonable substitute for fully independent subnational financing because they strengthen the central government’s control over macroeconomic and redistribution policies, while permitting decentralization of decisions about expenditure patterns and priorities (Walsh (1990)). However, these benefits must be weighed against the costs associated with grants—the loss of decisionmaking independence by subnational governments, the blurring of responsibility for the levels and quality of local public services, and the potential threat of intervention by higher-level governments for paternalistic and/or politically motivated reasons. An extensive empirical literature has examined the costs and benefits of alternative grant programs. Among the findings which may be potentially relevant for Japan are the following.

  • Economic principles and stated program objectives often provide a less satisfactory explanation of the structure of grant programs and distribution of grants (even in the case of fiscal equalization grants) than do political variables (see, e.g., Inman (1988), and Grossman (1989)).

  • Lump-sum grants from the central government stimulate much more grantee spending in the long run than predicted by the theory. 1/ The estimates by Gramlich (1977) and virtually all subsequent studies (see, e.g., a review in Inman (1979)) suggest that a dollar received as a grant generates 40 cents of public spending, while an additional dollar of private income increases public spending by only 10 cents. This phenomenon has been referred to as the “flypaper effect,” because the money seems to “stick” in the sector where it initially “hits.”

  • While decentralization of expenditure decisions has been generally associated with a smaller public sector size, the presence of substantial intergovernmental grants (i.e., lack of commensurate decentralization of revenue raising powers) has been associated with larger public sector size (Grossman (1989)). 1/ As indicated above, there is some evidence that the large grant program in Japan may have this effect.

  • Grants from the central government (both lump-sum and matching) to support regional programs induce reductions in financial support by intermediate level governments for the same as well as complementary programs (Boadway, Pestieau, and Wildasin (1989); Wildasin (1990)).

7. Options for reform

One of the most distinctive features of local public finance in Japan is the strict control that the central government exercises over local governments. As discussed above, the central government retains more or less controlling power not only over the whole of the revenues of local governments, but also over the whole range of local expenditure. In the case of construction works, for example, even the brands of construction materials and parts are sometimes specified by the central government. Such detailed control is bound to interfere with the efficient supply of local public goods and impede flexible adaptation of local governments to changing economic conditions and demands of the public. Therefore, the need for fiscal decentralization is largely self-evident; the real issue is what functions should be devolved to local governments and how to approach the task of decentralization.

The analysis in the preceding chapters suggests that the main problems with local public finances in Japan are excessively tight expenditure controls, lack of freedom of local governments to determine local tax rates, and a relatively high dependence of local governments on grants and borrowing. Possible approaches to reforms in these areas are as follows.

a. Relaxing expenditure controls

It seems fair to say that the devolution of powers to local governments that occurred after the Second World War has been largely aimed at achieving the goals of the central government (i.e., to improve national welfare as a whole). The alternative would, therefore, seem to be freeing local governments from central controls (i.e., improving local welfare). As discussed above, economic analysis, like political theory, does provide a strong rationale for the establishment of local governments that are responsive to the wishes of their citizens instead of being simply the agencies of the central government. So long as there are local variations in preferences and costs, there are clearly efficiency gains from carrying out public sector activities in as decentralized a fashion as possible. The only services that should be provided centrally are those for which there are no differences in demands in different localities, where there are substantial spillovers between jurisdictions that cannot be handled in some other way (e.g., by negotiation or by grant design), or for which the additional costs of local administration are sufficiently higher to outweigh its advantages.

The assignment of expenditure responsibilities in Japan formally conforms, to a large extent, to these principles. Therefore, the issue is not so much to reshuffle the assignments themselves, as to reshuffle responsibilities for designing, implementing, and financing these assignments. One concern that is often expressed in this context is that local governments would allow the level and quality of public services to deteriorate below the standard considered desirable. This may be to a large extent a misplaced concern. If the service in question is of national importance (e.g., health, education, research), or one in which there is a strong national interest in maintaining standards (e.g., poverty alleviation or some other distributional goal), such a service should remain nationally funded (at least in part) and its implementation monitored. If it is not a matter of national interest (e.g., local roads or sewage system), why should the national government be concerned? If the local electorate does not like what the local government does—or does not do—the government can be voted out of office at the next election. As frequently emphasized in the public choice literature, the freedom to make mistakes and to bear the consequences of one’s mistakes is an important component of local autonomy (Bird (1994)).

b. Managing tax diversity

Concerning the local revenue system, prima facie there seems to be a strong case for allowing local governments more freedom in determining local taxes. In principle, local governments should not only have access to those revenue sources that they are best equipped to exploit—such as residential property taxes, income taxes, and user charges for local services—but they should also be both encouraged and permitted to exploit these sources without undue central supervision. In Japan, local governments have relatively large receipts from what appear to be local taxes, but since they can neither set the tax rates nor determine the tax bases, it is difficult to see how they can be accountable to their constituents at the margin, as both efficiency and local autonomy require. 1/

As noted above, the empirical evidence does not support the view that tax competition and tax exporting create serious distortions in practice. On the other hand, tax diversity takes account of different circumstances in which local governments operate and different preferences over the mix of taxes and public goods. Large urban centers, which have a strong revenue base, usually also have much larger expenditure needs. Uniformity of the local tax code may thus prevent them from raising additional revenue that they need—and are able to raise. Presently, although the central government allows local governments to vary the tax rate, the degree of flexibility is small: prefectures can raise the standard tax rate of the personal income tax up to 1.2 times the standard rate, and municipalities up to 1.5 times (up to 1.1 times and 1.2 times, respectively, in the case of the enterprise tax); but they cannot levy a tax rate lower than the standard rate (see appendix to this chapter). Some flexibility in setting the consumption tax rate would also seem warranted. The actual tax practice of such countries as the United States, Canada, and Switzerland indicates that a substantial degree of tax diversity can be permitted without interfering with the mobility of goods and factors. 1/ It should also be noted that most elements of the existing tax system—including the assignment of taxes, harmonized tax bases, centralized collection and revenue sharing of certain taxes—can and should be preserved while allowing for the greater freedom of the local governments to determine the tax rates. Finally, local governments should be encouraged to rely on user charges and fees for providing public services as much as possible.

c. Reducing dependency on grants and borrowing

Local governments raise about 50 percent of revenue from local taxes and nontax sources, while another 20 percent comes from the local allocation tax, a revenue-sharing grant which is essentially local governments’ “own” revenue (albeit collected and distributed by the central government). Of the remainder, about 17 percent are specific-purpose disbursements by the central government, designed to cover the costs of providing certain functions on behalf of the central government, and another 2 percent are earmarked revenue-sharing grants. Hence, if revenue-sharing grants are treated as local governments’ own revenue, the level of grant dependency is about 17 percent of local revenue, and the overall revenue deficiency, i.e., the amount raised through local borrowing, amounts to about 10 percent of local revenue. These ratios are, indeed, high—about 2 1/4 percent and 1 1/4 percent of GDP, respectively.

It is largely a matter of structural fiscal reform how to reduce the dependency of local governments on grant money and borrowing. The fiscal federalism theory cannot offer much guidance on this issue, except noting that promoting local autonomy is likely to help reduce the inefficient size of the local public sector. It should be recognized, however, that the need for large local borrowing arises to a certain extent because of the way the central government conducts macro fiscal policy. Given the flexibility that such arrangement allows the policymakers, it is not entirely clear that the revenue gap should be closed completely through, for example, larger permanent transfers from the central government. Another point to note is that local borrowing is used almost exclusively to finance capital projects. Hence, to the extent that these projects are cost-effective and raise the long-run productivity of social capital, they do not necessarily have to be financed by taxes on the current generation of taxpayers.

As for the design of specific-purpose grants and the modalities of borrowing, a number of improvements could be achieved. The main problem with specific purpose grants, recognized for a long time by the Japanese experts (see, e.g., Yonehara (1981), and Jichi Sogo Center (1993)) is that, although (or perhaps because) they contain very detailed provisions, they do not sufficiently take into account local circumstances and needs. As noted in the preceding sections, the detailed nature of the these grants and their overall large volume can lead to much inefficiency and frustration for both the central and local governments. The number of specific purpose grants should, therefore, be reduced to a few most important expenditure areas. Other expenditures presently financed by specific-purpose grants could be either devolved to local governments (and, hence, financed from their general revenue resources, including the local allocation tax), or, to the extent that they remain a central government responsibility, financed by grants-in-aid to local authorities, which would encourage local governments to take a more active role in implementing the activities deemed important by the national government.

From the perspective of local authorities, the main problems concerning local borrowing are that procedures for loan applications are too complicated and time consuming, and that in conducting countercyclical policy through variations in local borrowing, the central government does not take into account the burden of future interest payments by local governments. If local governments are to become more autonomous in their spending and taxation decisions, local borrowing should be facilitated through simplification of procedures for raising local loans. Given that financial markets in Japan are highly developed, there is no reason why the central government should be involved in almost every step of local loan raising. While it may be desirable to set overall limits on local borrowing, such limits could be set on the stock of debt at the local level rather than annual deficits, so as to allow local governments greater flexibility in responding to regional economic shocks. To the extent that local borrowing is mandated by the central government, the central government should also bear the additional cost of such loans through, for example, larger general-purpose transfers to local governments.

Given how tightly most aspects of local public finance in Japan are presently controlled, fiscal decentralization is not likely to produce exactly the pattern the central government—or local governments—would choose to implement if left to their own devices, except in the unlikely case that their goals precisely coincided. Conflicts between the central and local governments as to what should be decentralized are, therefore, inevitable. It should be recognized, however, that if accessibility, local responsibility, and the effectiveness of government are to be improved, it is of fundamental importance that governments that are responsible for expenditure decisions should be responsible for raising the revenue to fund them, and should have control over—and responsibility for—revenue sources adequate to enable them to do so.

APPENDIX: Summary of Local Public Finance in Japan

This appendix summarizes the main statistical data on local public finance in Japan and describes in more detail the local tax system and intergovernmental grants that are discussed in Chapter IV. For comparative purposes, data on local expenditure and revenue in major industrial countries are also presented. The appendix concludes with a description of variables and sources of data for the regressions presented in Chapter IV.

1. Expenditures

The following table illustrates the scale of public expenditure and intergovernmental transfers effected by the central and local governments in FY 1992.

On a net basis, the central government was responsible for about one-third of total expenditure (10 percent of GDP), and local governments for two-thirds (19 percent of GDP) in FY 1992. Transfers from the central to local governments consisted of the local allocation tax (52 percent), local transfer taxes (6 percent), and specific-purpose grants (42 percent). Transfers from local governments to the central government were payments for certain public works carried out by the national government. Total transfers amounted to about 7 percent of GDP in FY 1992.

A standard classification of government expenditure by function in the IMF’s Government Finance Statistics framework is not available for Japan’s local governments. The following table represents an attempt to compile such data on the basis of expenditure categories used in Japanese budgetary statistics. The entries for the central government are based on the Initial Budget for FY 1992; the entries for local governments are based on the settlement data. Both sets of data refer to gross expenditure (detailed breakdown on a net expenditure basis is not available). In view of these data restrictions, the expenditure shares presented in the table below, as well as in the Tables IV.4 and IV.5 that contain international comparisons of expenditure shares, should be regarded as illustrative only.

Table IV.2.

Japan: Expenditure by Function and Level of Government, FY 1992

article image
Sources: Staff estimates, based on Ministry of Finance (1993) and Ministry of Home Affairs ((1995), (1994)).

Based on the Initial Budget for FY 1992; gross expenditure basis.

Based on settlement data for FY 1992; gross expenditure basis.

For central government, includes miscellaneous expenditure; for local governments, includes local assembly, general administration, fire protection, and police.

For central government, includes public health, government-managed health insurance, seamen’s insurance, and subsidies to National Health Insurance.

For central government, includes public assistance, social welfare, measures for the unemployed, employees’ pension insurance, national pension, “pensions and others,” children’s allowance, other social insurance, and housing expenditure from the public works budget.

For central government, includes public works (except housing), economic cooperation, and measures for small businesses, energy, and foodstuff control; for local governments, includes public works and measures for employment, agriculture, and trade and industry.

For central government, includes local allocation tax and contingency; for local governments, includes disaster relief and miscellaneous.

The following table presents a breakdown of expenditure by function at the level of prefectures and municipalities for FY 1990. The breakdown covers expenditure financed by central government transfers, as well as prefectural transfers to municipal governments.

Table IV.3.

Japan: Local Expenditure by Function, FY 1990 1/

article image
Source: Jichi Sogo Center (1993).

Based on settlement data for FY 1990; gross expenditure basis.

Includes local assembly, general administration, fire protection, and police.

Includes public works, employment and industrial relations, agriculture, forestry, and fisheries, and trade and industry.

Includes disaster relief and miscellaneous.

On the whole, prefectures had a slightly higher share in (gross) expenditure than municipalities in FY 1990 (52 percent vs. 48 percent of total local expenditure). Prefectures carried out a higher proportion of expenditure on education and economic affairs than municipalities, while municipalities carried out a higher proportion of expenditure on social security, welfare, and health. Economic affairs were the main expenditure item for both prefectures and municipalities; education, general public services and public safety, and social security and welfare were the other major spending categories. Debt service accounted for a relatively high proportion of both prefectural and municipal spending.

Table IV.4.

Japan: International Comparison of Expenditure Shares by Level of Government and Function 1/

(In percent of expenditure on each function)

article image
Source: IMF (1990); and staff estimates based on Ministry of Finance (1993) and Ministry of Home Affairs (1995).

Data for Japan are staff estimates, based on Table IV.2. Data refer to the following years: FY 1992 for Japan; 1985 for France; 1988 for the United Kingdom and the United States; and 1987 for Germany.

To compare the composition of expenditure at the central and local levels of government in Japan with other countries, data from Table IV.2 were used along with the data from the IMF’s Government Finance Statistics Yearbook for two unitary states—France and the United Kingdom—and two federations—the United States and Germany. As the data for Japan were not compiled on the same basis, these comparisons only illustrate an approximate order of magnitude of central and local expenditure shares.

In interpreting the above data, it is important to note that, while the proportion of local government spending in Japan is relatively high for both total expenditure and individual spending categories (such as general administration and safety, education, health, and economic affairs), local governments have little effective control over the scope of spending that takes place at the local level. As discussed in Section 4, major programs (education, health, public works) are in practice formulated by national ministries and financed directly or indirectly by the central government. Local governments in Japan execute these expenditure programs under detailed national direction, and, therefore, the degree of effective centralization of expenditure functions is higher than suggested by the above figures.

The following table compares the composition of expenditure at each level of government in Japan and five industrial countries. As in the preceding table, these comparisons are only an illustration of expenditure shares at each level of government, given that local government data for Japan are not available in the GFS format.

Table IV.5.

Japan: International Comparison of Expenditure Shares by Function and Level of Government 1/

(In percent of central/local expenditure)

article image
Sources: IMF (1990); and staff estimates based on Ministry of Finance (1993); Ministry of Home Affairs (1995).

Data for Japan are staff estimates, based on Table IV.2. Data refer to the following years: FY 1992 for Japan; 1985 for France; 1988 for the United Kingdom and the United States; and 1987 for Germany.

For Japan, includes debt service, local allocation tax grants, and contingency.

Compared with other industrial countries, local governments in Japan spend a relatively higher proportion of their budgets on general public services, safety, and economic affairs, and a relatively lower proportion on social security, welfare, and health. Also, the education spending as a proportion of local budgets is lower in Japan than in the United Kingdom, the United States, and Germany. At the central level, the government in Japan spends about the same proportion on administration, defense, and education as the central government in France. The share of health and social security spending in Japan’s central government budget is, however, lower than in other industrial countries, while the share of expenditure on economic affairs (public infrastructure, agriculture, energy) is higher.

2. Revenues

Total revenue of local governments consists of local taxes, central government transfers of revenue, nontax revenue, and local borrowing. For FY 1990, the main revenue sources were distributed between prefectures and municipalities as follows:

Table IV.6.

Japan: Local Government Revenue, FY 1990

(In percent of total)

article image
Source: Jichi Sogo Center (1993).

Transfers from prefectures to municipalities are netted out from the total.

Prefectures and municipalities thus raised roughly the same proportion of their revenue from local taxes, charges and fees, and local bonds. Central government transfers represent a higher proportion of prefectural revenue because prefectures carry out more functions on behalf of the central government than municipalities. On the other hand, municipalities raise more revenue from nontax sources such as income from properties, shares, reimbursements, and contributions.

Major national and local taxes are assigned to the three levels of government as follows:

Local taxes are generally levied on identical tax base as national taxes. The following are some representative tax rates. 2/

Prefectural taxes

  • Prefectural inhabitants tax

  • article image

  • article image

  • Prefectural enterprise tax

  • article image

  • Real property acquisition tax

    • Standard tax rate: 4 percent of the appraised value of land or house at the time of acquisition. Special deductions apply for certain residential properties.

  • Municipal taxes

  • Municipal inhabitants tax

    article image

  • Municipal property tax

    • Standard tax rate: 1.4 percent (up to 2.1 percent maximum) of the cadastral value of land, houses, and depreciable assets. Special deductions and exemptions apply for certain residential properties.

    • Local governments are also assigned a large number of minor taxes, whose yield is very small (0.1 percent of local tax revenue):

  • Minor prefectural taxes: Local entertainment tax, meals and hotel tax, mine lot tax, hunter’s registration tax, fixed assets tax, nonlegal ordinary tax, hunting tax, water utilization and land benefit tax.

  • Minor municipal taxes: Electricity tax, gas tax, mineral product tax, timber delivery tax, special landholding tax, nonlegal ordinary tax, spa tax, business office tax, city planning tax, water utilization and land benefit tax, common facilities tax, land development tax, national health insurance tax.

The following table describes how the tax revenue was allocated among the three levels of government in FY 1990:

Table IV.7.

Japan: Central and Local Government Taxes, FY 1990

article image
Sources: Ministry of Home Affairs (1995); Jichi Sogo Center (1993); and staff estimates.

Includes local allocation tax, local transfer tax, and specific-purpose disbursements to local governments.

Contributions for central government projects.

Includes municipal taxes collected by prefectures.

Includes prefectural share of the local allocation tax, local transfer tax, and specific-purpose grants.

Includes municipal share of the local allocation tax, local transfer tax, and specific-purpose grants.

The breakdown of prefectural and municipal taxes in FY 1992 was as follows:

Table IV.8.

Japan: Local Tax Collections by Source, FY 1992

(In percent of total)

article image
Source: Ministry of Home Affairs (1995).

For prefectures, includes real property acquisition tax and prefectural fixed assets tax; for municipalities, includes municipal fixed assets tax and special land holding tax.

On the whole, the share of municipalities in local taxes was higher than that of prefectures in FY 1992 (57 percent vs. 43 percent). Municipalities raised a higher proportion of their tax revenue from personal income and property taxes, while prefectures raised more tax revenue from corporate income and automobile taxes. Taxes raised by local governments amounted to 8 percent of GDP in FY 1992. As discussed below, this represents a relatively high proportion of total revenue by international standards.

Unlike the expenditure data, the revenue data compiled in Japan are fully comparable internationally. The following table presents the shares of three main revenue sources—taxes, nontax revenue and grants—at each level of government in Japan and five industrial countries.

Table IV.9.

Japan: International Comparison of Revenue Assignment by Level of Government 1/

article image
Sources: OECD (1990); and Jichi Sogo Center (1993).

Data for Japan are for FY 1990; data for United States are for 1987; all other data are for 1988.

The share of Japan’s local governments in total tax revenue is about one-third, the same as in France but significantly higher than in the United Kingdom and Italy. In terms of the share in total nontax revenue, Japanese local governments are comparable with states and local governments in the United States and Germany. Central governments in all industrial countries shown (except Italy) raise about 88 percent of their revenue from taxes. The share of grants in local revenue is roughly the same in Japan, France, and Italy. Finally, in terms of the GDP shares, local taxes in Japan amounted to about 7 3/4 percent of GDP in FY 1990, about one-third lower than in France (10 1/2 percent of GDP in 1988), but higher than in the United Kingdom and Italy. The proportion of nontax revenue raised by local governments in Japan is highest among the countries shown (5 3/4 percent of GDP). Interestingly, grants to sub-national levels accounted for about 5 percent of GDP in all the countries shown except Italy.

3. Intergovernmental transfers

Japan’s central government makes three basic types of grants to local rnments: governments:

  • Unconditional, general purpose revenue sharing grants, referred to in Japan as the “local allocation tax.” These grants are essentially mandatory transfers of revenue from a number of national taxes (in whole or in part) to local governments. Specifically, the local allocation tax consists of:

    • 32 percent of the yield of the income tax, corporation tax, and liquor tax;

    • 24 percent of the yield of the consumption tax; 1/ and

    • 25 percent of the yield of the tobacco tax.

    • The local allocation tax is divided into the ordinary allocation tax (94 percent of the total) and special allocation tax (6 percent of the total), which is designed to compensate for shortfalls in the ordinary allocation tax. Local governments can spend the local allocation tax for any purpose, but the amounts transferred are designed so as equalize the fiscal capacity of local governments (discussed below).

  • Specific purpose grants, i.e., transfers designed to finance (in whole or in part) the expenses related to specific local expenditure programs. The main objective of specific purpose grants is to achieve uniformity in the provision of local public services. Specific purpose grants are of two types:

    • Central government disbursements,” further divided into:

      • (i) payments for “agency delegated tasks,” i.e., for tasks which are essentially central government responsibility but their execution is entrusted to local governments (e.g., administration of health insurance and the provision of national pensions);

      • (ii) central government’s obligatory share in certain local expenses (e.g., salaries of public school teachers, contributions to construction work and disaster relief); and

      • (iii) grants-in-aid to local authorities, i.e., incentive subsidies used to encourage local governments to perform the activities deemed important by the national government;

    • Local transfer taxes,” which are essentially specific purpose, revenue-sharing grants, i.e., taxes levied and collected by the national government on behalf of prefectural and municipal governments (e.g., vehicle tonnage and certain fuel taxes) and earmarked, for the most part, for specific purposes (e.g., local expenditure on roads).

4. Regression variables and data sources

The variables used in the regressions reported in Chapter IV (Section 6) were defined as follows:

TEG = the sum of: central government expenditure on general account

  • (settled), minus expenditure on transfers (local allocation tax, local transfer taxes, and specific-purpose grants), plus local government expenditure in local public finance program; divided by fiscal year GDP. This variable represents general government net expenditure.

CENR = central government revenue on general account (settled), divided by

  • the sum of: central government revenue on general account (settled), minus transfers to local governments (local allocation tax, local transfer taxes, and specific-purpose grants), plus local government revenue in local public finance program. This variable measures the revenue centralization ratio.

CENE = central government expenditure on general account (settled), divided

  • by the sum of: central government expenditure on general account (settled), minus expenditure on transfers (local allocation tax, local transfer taxes, and specific-purpose grants), plus local government expenditure in local public finance program. This variable measures the expenditure centralization ratio.

All variables were obtained from Nomura Research Institute (1994) database.

The following subsidiary regressions were referred to in Chapter IV (Section 6):


where the dependent variable, TRANS, is total central government transfers to local governments (local allocation tax, local transfer taxes, and specific-purpose grants) in percent of fiscal year GDP. Parameter estimates in the first equation are statistically significant at the 10 percent level (except the intercept); parameter estimates in the second equation are statistically significant at the 5 percent level.


  • Bingman, Charles F., Japanese Government Leadership and Management (New York: MacMillan, 1989).

  • Bird, Richard M,Threading the Fiscal Labyrinth: Some Issues in Fiscal Decentralization,National Tax Journal. Vol. 46, 1994, 20727.

    • Search Google Scholar
    • Export Citation
  • Boadway, Robin, and Frank Flatters,Efficiency and Equalization Payments in a Federal System of Government: A Synthesis and Extension of Recent Results,Canadian Journal of Economics. Vol. 15, 1982, 61333.

    • Search Google Scholar
    • Export Citation
  • Boadway, Robin, P. Pestieau, and David Wildasin,Tax Transfers and the Voluntary Provision of Public Goods,Journal of Public Economics. Vol. 39, 1989.

    • Search Google Scholar
    • Export Citation
  • Brennan, Geoffrey, and James M. Buchanan, The Power to tax: Analytic Foundations of a Fiscal Constitution (Cambridge: Cambridge University Press, 1980).

    • Search Google Scholar
    • Export Citation
  • Feld, Lars, Gebhard Kirchgässner, and Werner Pommerehne,Tax Harmonization and Tax Competition at the Local Level: “Is There an Inevitable 1 Collapse of the Welfare State?Paper presented at the 51st Congress of the international Institute of Public Finance, Lisbon, August 1995.

    • Search Google Scholar
    • Export Citation
  • Gramlich, Edward M.,Intergovernmental Grants: A Review of the Empirical Literature,” in The Political Economy of Fiscal Federalism, ed. by Wallace Oates (Lexington: Heath-Lexington, 1977).

    • Search Google Scholar
    • Export Citation
  • Groenewegen, Peter,Taxation and Decentralization: A Reconsideration of the Costs and Benefits of a Decentralized Tax System,” in Taxation and Fiscal Federalism: Essavs in Honour of Russell Matthews, ed. by Geoffrey Brennan, Bhajan S. Grewal, and Peter Groenewegen (Sydney: Australian National University, 1988), 87115.

    • Search Google Scholar
  • Grossman, P.J.,Fiscal Decentralization and Public Sector Size,Public Choice. Vol. 62, 1989.

  • Hamlin, Alan P.,Decentralization, Competition and the Efficiency of Federalism,Economic Record. Vol. 67, 1991, 193204.

  • Hashiyama, Reijiro,Whither the New Capital,Look Japan. June 1996, 1415.

  • Inman, Robert P.,Federal Assistance and Local Services in the United States: The Evolution of a New Fiscal Order,” in Fiscal Federalism: Quantitative Studies, ed. by Harvey Rosen (Chicago: University of Chicago Press, 1988).

    • Search Google Scholar
    • Export Citation
  • Inman, Robert P.,The Fiscal Performance of Local Governments: An Interpretative Review,” in Current Issues in Urban Economics, ed. by Peter Mieszkowski and Mahlon Straszheim (Baltimore: Johns Hopkins University Press, 1979), 270321.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, Government Finance Statistics Yearbook 1990 (Washington: International Monetary Fund, 1990).

  • Ishihara, Nobuo,The Local Public Finance System,” in Japan’s Public Sector: How the Government is Financed, ed. by Tokue Shibata (Tokyo: University of Tokyo Press, 1993), 14166.

    • Search Google Scholar
    • Export Citation
  • Jichi Sogo Center, The Situation of Local Public Finance in Japan (Tokyo: Jichi Sogo Center, 1993).

  • Jichi Sogo Center, Statistical Abstract of Japanese Local Public Finance (Tokyo: Jichi Sogo Center, 1989).

    • Search Google Scholar
  • King, David N., Fiscal Tiers: The Economics of Multi-Level Government (London: George Allen and Unwin, 1984).

    • Search Google Scholar
  • Ministry of Home Affairs, General Affairs Office, Source Material on the Bill on the Promotion of Regional Decentralization (in Japanese) (Tokyo: Ministry of Home Affairs, 1995).

    • Search Google Scholar
    • Export Citation
  • Ministry of Home Affairs, Local Autonomy College, Local Public Finance in Japan (Tokyo: Ministry of Home Affairs, 1995).

  • Ministry of Home Affairs, The Japanese Local Public Finance System (Tokyo: Ministry of Home Affairs, 1994).

  • Ministry of Home Affairs, “Local Administration and Finance,” in Public Administration in Japan, ed. by Kiyoaki Tsuji (Tokyo: University of Tokyo Press, 1984), 87107.

    • Search Google Scholar
  • Ministry of Finance, Tax Bureau, An Outline of Japanese Taxes. 1995 (Tokyo: Ministry of Finance, 1996).

  • Ministry of Finance, Budget Bureau, The Japanese Budjget in Brief (Tokyo: Ministry of Finance, 1994).

    • Search Google Scholar
  • Mochida, Nobuki,Balancing Equity and Decentralization,Social Science Japan. November 1995, 1011.

  • Mochida, Nobuki,The Role of Local Government Expenditures in Pre-War Japan,Annals of the Institute of Social Science. Vol. 26, March 1985.

    • Search Google Scholar
    • Export Citation
  • Moesen, Wim A.,Community Public Finance in the Perspective of EMU: Assignment Rules, the Status of the Budget Constraint, and Young Fiscal Federalism in Belgium,” in Commission of the European Communities, The Economics of Community Public Finance. European Economy. Vol. 5, 1993.

    • Search Google Scholar
    • Export Citation
  • Moesen, Wim A., and Paul Van Rompuy,The Growth of Government Size and Fiscal Decentralization,” in Public Finance With Several Layers of Government. ed. by Rémy Prud’homme (Brussels: Foundation Journal of Public Finance, 1990), 11324.

    • Search Google Scholar
    • Export Citation
  • Musgrave, Richard A.,Who Should Tax, Where, and What?” in Tax Assignment in Federal Countries, ed. by Charles E. McLure, Jr. (Canberra: The Australian National University, 1983).

    • Search Google Scholar
    • Export Citation
  • Musgrave, Richard A., The Theory of Public Finance (New York: McGraw-Hill, 1959).

  • Nomura Research Institute, NRI/E: Japanese Economic. Business, and Financial Database. Database Manual (Tokyo: Nomura Research Institute, December 1984).

    • Search Google Scholar
    • Export Citation
  • Oates, Wallace E., Fiscal Federalism (New York: Harcourt, Brace, Jovanovich, 1972).

  • Organisation for Economic Cooperation and Development, Revenue Statistics 1965-89 (Paris: OECD, 1990).

  • Ozawa, Ichiro:Reforming Japan: The Third Opening,” Economist. March 9, 1996, 2123.

  • Saunders, Peter,Explaining Differences in Public Expenditure: An Empirical Study,Public Finance. Vol. 43, 1988, 27194.

  • Shindo, Muneyuki,Relations Between National and Local Government,” in Public Administration in Japan, ed. by Kiyoaki Tsuji (Tokyo: University of Tokyo Press, 1984), 109120.

    • Search Google Scholar
    • Export Citation
  • Tiebout, Charles,A Pure Theory of Local Expenditures,Journal of Political Economy. Vol. 64, 1956, 41624.

  • Walsh, Cliff, Fiscal Federalism: An Overview of Issues and a Discussion of their Relevance to the European Community (Canberra: Federalism Research Centre, The Australian National University, 1990).

    • Search Google Scholar
    • Export Citation
  • Wildasin, David E.,Budgetary Pressures in the EEC: A Fiscal Federalism Perspective,American Economic Review. Vol. 80, 1990.

  • Yonehara, Junshichiro,Financial Relations Between the National and Local Governments,” in Japan’s Public Sector: How the Government is Financed, ed. by Tokue Shibata (Tokyo: University of Tokyo Press, 1993), 16778.

    • Search Google Scholar
    • Export Citation
  • Yonehara, Junshichiro, Local Public Finance in Japan (Canberra: Centre for Research on Federal Financial Relations, The Australian National University, 1981).

    • Search Google Scholar
    • Export Citation

This chapter was prepared by Dubravko Mihaljek.


This section draws on Bingman (1989). For a more detailed analysis, see Mochida (1985).


Universal suffrage for males was achieved in 1927, and for females in 1947.


For example, the law defines the powers and authorities of chief executives, including the number, name, and responsibilities of all major sub-divisions of local governments.


These laws are not just major statutes such as the City Planning Law or the Environmental Preservation Law, but also such acts as the Law for School Meals for Night Courses of Senior High Schools, or the Laundry Law, which sets forth detailed procedures for licensing, registration, and examination of laundries and their workers (Bingman (1989), p. 52).


National ministers are themselves elected officials: the Constitution requires that the Prime Minister and at least half of the Cabinet come from either house of the Diet, In practice, almost all of them are Diet members, primarily from the House of Representatives, whose 511 members are elected for four-year terms.


For a breakdown of expenditure by function and level of government, see Tables IV.2-IV.4 in the appendix to this chapter.


These authorities include: the right to demand reports and evaluations of performance; the right to conduct investigations, inspections, and audits of local government activities; and the right to impose various corrective measures for inadequate performance or violation of national regulations (Bingman (1989)).


Local governments also make (small) payments for specific purposes to the central government, mainly toward improvements for public works carried out by the national government.


For details, see Yonehara ((1981), (1993)).


The full amount of standardized revenue is not used in this formula in order to provide incentives to local governments to enlarge their tax bases, and also to take into account special local needs that require nonstandard amounts of revenue.


In 1949, Japan’s Central Government invited Professor Carl Shoup of Columbia University to head a mission for the investigation of the Japanese tax system. The recommendations of this mission were considered in the drafting of the Local Autonomy Law, the Local Tax Law, and the legislation on central government grants.


See Ministry of Home Affairs (1994), and Hashiyama (1996).


For a review of the academic debate on decentralization, see Mochida (1995).


The standard argument in the literature is that, if local governments were responsible for major redistributive programs and labor was perfectly mobile, jurisdictions with more generous programs would attract those who benefit from such programs (generally the poor) and repel those who are primarily net contributors (i.e., the higher-income taxpayers). Failure to introduce fiscal systems that broadly aim for horizontal equity can also induce flows of population which lead to the inefficient use of mobile resources (Boadway and Flatters (1982)).


The estimated equations were: TEG=0.0017+0.7655LDV+0.1782CENR,andTEG=-0.0403+0.7767LDV+0.2248CENE, where: TEG is total expenditure of the general government (in percent of GNP), LDV is a lagged dependent variable (the lags extend to the pre-sample period, so as to capture the historical ratio of public expenditure to GDP as a social and cultural preference variable); CENR is total revenue of central government (in percent of total revenue of general government); and CENE is total expenditure of central government (in percent of total expenditure of general government). The sample included five federations and ten European unitary countries during the 1980s; the estimation results were quite satisfactory statistically.


For a detailed description of variables and sources of data, see appendix to this chapter.


Transfers include the local allocation tax, local transfer taxes, and specific-purpose grants to local governments. For details of these regressions, see appendix to this chapter.


For a survey of empirical studies, see Groenewegen (1989).


The theory predicts that an increase in the grant—equivalent to an outward shift of grantee’s budget constraint—should have more or less the same effect on the grantee’s spending as an increase in the grantee’s income brought about by, for example, lower taxes.


One explanation of this phenomenon has been that grants reflect the existence of a largely implicit revenue raising cartel—governments have generally an interest in minimizing the degree of competition over tax bases because it enables them to blur responsibility for fiscal decisions and reduces the pressure for creativity in policymaking (Brennan and Buchanan (1980)).


As the economic role of local government is to provide to local residents those public services for which they are willing to pay, accountability is essentially the public sector equivalent of the “bottom line” in the private sector.


In the United States, for example, 43 states have their own personal income tax, 45 states have their own corporate income tax, and 45 states have their own sales tax. It would be hard to deny that, despite this tax diversity, there is a single, highly integrated domestic market.


Twenty percent of the consumption tax is transferred directly to local governments.


This percentage is applied to the central government’s share of the consumption tax (i.e., to the 80 percent of the yield of the tax). Local governments get an additional 20 percent of the total yield of the consumption tax as a direct transfer; this portion, however, falls outside the scope of the local allocation tax.

Japan: Selected Issues
Author: International Monetary Fund