33043 Taxes and Transfers in Japan's Local Public Finances Nobuki Mochida Abstract A strong inter-regional equity bias has been a distinctive feature of the Japanese local public finance system. This paper shows that substantial equalization of revenues per capita is achieved via transfers from the central government and that, over time, this appears to have substantially improved the regional distribution of income: the Gini coefficient of per capita regional income declined from around 0.17 in 1950 to 0.10 in 1990. Now that considerable regional equality has been achieved, a greater concern for the exercise of local preferences is being voiced. World Bank Institute Copyright © 2001 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. May 2001 The World Bank enjoys copyright under protocol 2 of the Universal Copyright Convention. This material may nonetheless be copied for research, educational, or scholarly purposes only in the member countries of The World Bank. Material in this series is subject to revision. The findings, interpretations, and conclusions expressed in this document are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or the members of its Board of Executive Directors or the countries they represent. Taxes and Transfers in Japan's Local Public Finances Nobuki Mochida 2001. 33 pages. Stock No. 37171 Contents Foreword v Basic Framework of the Current System 1 Characteristics of Local Tax System in Japan 8 Intergovernmental Fiscal Transfer and Regional Disparity 13 Implications of the Japanese Experience 17 Endnotes 20 Bibliography 22 Appendix 25 iii Foreword This paper was prepared for a project on Local Government Development in Japan. The project was organized by the World Bank Institute under the auspices of the Program for the Study of Japanese Development Management Experience financed by the Policy and Human Resources Development Trust Fund of the Government of Japan. The principal objectives of this Program are to conduct studies on Japanese and East Asian development management experience and to disseminate the lessons of this experience to developing and transition economies. Typically, the experiences of other countries are also covered in order to ensure that these lessons are placed in the proper context. This comparative method helps identify factors that influence the effectiveness of specific institutional mechanisms, governance structures, and policy reforms in different contexts. A related and equally important objective of the Program is to promote the exchange of ideas among Japanese and non-Japanese scholars, technical experts and policy makers. The papers commissioned for this project cover a number of important issues related to local government development in Japan. These issues include: the process of controlled decentralization; increasing political inclusiveness; redistributive impact of local taxes and transfers; allocation of grants; municipal amalgamation; personnel exchanges; personnel policies; agency-delegated functions; and local policy initiatives. Farrukh Iqbal, Program Manager World Bank Institute v Taxes and Transfers in Japan's Local Public Finances Nobuki Mochida1 University of Tokyo, Japan Japanese local public finances account for roughly 70 percent of general public expenditure and 80 percent of public capital formation. This suggests that, in implementation terms at least, local governments play an important role in the provision of public services. The efficiency and equity characteristics of the Japanese local government system are, therefore, likely to have an important effect on regional and national development. The purpose of this paper is to examine Japan's intergovernmental relations from a fiscal viewpoint. It is divided into four sections. The first section provides an outline of the basic framework of the system. The second provides an analysis of the local tax system. The third provides data on the regional equalization effect of intergovernmental transfers. The fourth and concluding section discusses some lessons of experience to date for other countries as well as for Japan. Basic Framework of the Current System Centralized and Decentralized Aspects The main features of the Japanese system are centralized tax administration, decentralized provision of public services, and dependence of local government on intergovernmental transfers. In other words, intergovernmental fiscal relations are marked by a vertical fiscal imbalance in Japan.1 Table 1 provides comparative data on vertical fiscal imbalance in 10 major countries based on national accounts. It can be seen that the average amount of imbalance is quite large although the degree does vary from country to country. In particular, three countries, Japan, England, and Australia, exhibit the highest levels of imbalance between expenditure responsibilities of local governments and their tax resources. In Sweden and France the imbalance is more moderate, while in federal states, except Australia, the level of imbalance is relatively low. Vertical fiscal imbalance can be high if tax shares of local governments are very low and/or if their expenditure responsibilities are very high. In the case of Japan, the latter is more likely to be the reason. The local tax share of total tax revenues in Japan is 36.5 percent, which is on the high side for OECD countries. On the other hand, as already noted, expenditures channeled through local governments are very high, at 70% of the total. Indeed, this share is the highest among OECD countries, higher even than the 50-60 percent observed in the Scandinavian countries. Despite this high rate of spending through local governments, it 1The author wishes to thank Jun Ma, Jisoon Lee, Farrukh Iqbal, Jørgen Lotz, Dubravko Mihaljek, James Mak, Jhungsoo Park, Michio Muramastu, Steven R. Reed, and Shun'ichi Furukawa for helpful comments. 1 2 Nobuki Mochida would be inaccurate to characterize Japan as a decentralized system because a considerable amount of decision-making authority has tended to rest with central units. In Japan, the central government exercises its discretionary powers on local expenditure through delegation of program implementation responsibilities to local chief executives financed, at least in part, by specific-purpose grants. Table 1: Vertical Fiscal Imbalance in Selected OECD Countries (FY 1992) (%) Vertical fiscal imbalance (a) (b) Expenditure Revenue Dependence on share of Share of Surplus/Deficit Central transfer subnational Subnational Grants as % of Country government government (b) ­ (a) current receipts United States 51.4 45.9 - 5.5 20.9 Canada 65.3 53.2 - 12.1 35.4 Australia 48.6 23.0 - 25.6 40.3 Germany 58.6 47.2 - 11.4 18.5 Denmark 57.3 32.3 - 25.0 45.5 Sweden 47.6 38.7 - 8.9 22.2 United 31.9 0.07 - 31.2 79.8 Kingdom France 31.1 18.0 - 13.1 38.6 Spain 43.6 19.7 - 23.9 46.6 Japan 69.2 36.5 - 32.7 39.6 Average 47.3 27.0 - 20.3 36.0 Source: OECD, National Accounts, 1996; IMF, Government Finance Statistics, 1996. Large Scale Redistribution of Sources of Revenue2 The above-noted imbalance is addressed by intergovernmental transfers. There is large- scale reallocation of revenue through earmarked and general subsidies in Japan. Table 2 shows the distribution of tax share and fiscal transfer between central and local government. In 1993, Taxes and Transfers in Japan's Local Public Finances 3 total tax revenues were ¥90,705 billion, which is divided into national and local taxes. Before fiscal transfers, local taxes account for only 37.0 percent of total revenue. However, a substantial portion of national taxes is transferred to the local governments. Major fiscal transfers are of two broad types. Unconditional transfers are tax-sharing grants on a lump-sum basis financed by the local allocation tax. Conditional grants are matching categorical grants that are known as specific-purpose grant. After reallocating the tax sources among levels of government, the final share of total tax revenue accruing to local governments increases to 52.5 percent. This means that one-fourth of central tax revenue is used at the local level. Table 2: Redistribution of Tax Revenue between National and Local Government (%) 1950 1960 1970 1980 1985 1993 (1) Government Expenditure Net Total/GNP 23.6 18.8 20.3 29.4 27.6 30.0 Net National/GNP 10.4 5.9 5.9 9.9 10.2 10.3 Net Local/GNP 13.2 12.8 14.4 19.5 17.4 19.6 (2) Tax Allocation before Fiscal Transfer National Tax/Total Tax 75.2 70.8 67.5 64.1 62.6 63.0 Local Tax/Total Tax 24.8 29.2 32.5 35.9 37.4 37.0 Income Tax/Total Tax 38.6 21.7 31.2 38.1 39.4 36.9 (3) Fiscal Transfers Transfer as % of general 35.1 47.2 48.7 44.0 38.5 36.0 account Local allocation tax as 17.1 17.8 22.0 18.7 17.8 20.4 % of general account Transfer as % of local 40.8 39.3 37.7 40.8 34.7 31.8 revenues Local allocation tax as 19.9 14.8 17.0 17.3 15.5 18.0 % of local revenues (4) Final Share of tax after Fiscal Transfer of National government 65.2 57.0 51.1 46.0 45.3 47.5 of Local government 34.8 43.0 48.9 54.0 54.7 52.5 Source: Mochida, N. (1993), p. 57. Note: Transfer includes both local allocation tax and specific-purpose grant. Intergovernmental transfers are needed not only to balance the budget at the sub- national level, but also to offset regional inequalities arising from differences in physical and demographic endowments. Given regional gaps in tax revenues and financial needs, some means of fiscal equalization is necessary to provide local public services in poor areas. The 4 Nobuki Mochida most important means devised to handle this problem is the unconditional tax-sharing grant. In Japan, the local allocation tax system plays a key role through the equalization transfer scheme. Although specific-purpose grants, local transfer taxes, and even some local taxes also have the effect of equalization, to some extent the discussion in Japan concerns the local allocation tax because of its key role in the equalization transfer system. A comparison of per capita local tax revenue and per capita revenue from general fiscal sources (that is, local taxes and local allocation tax) of prefectures shows a substantial reduction of disparities. This can be seen in Table 3 which groups prefectures into five categories according to the index of fiscal capacity, which is defined as the basic fiscal capacity divided by basic fiscal need of each local government. A marked difference is observed in per capita prefectural tax revenues among localities in fiscal year 1993, the largest in Tokyo, ¥196,000, the smallest in Okinawa, ¥60,000, corresponding closely to the difference in their economic resources and the per capita income of their inhabitants. Disproportionate local allocation tax is provided to areas with lower resource bases to achieve some degree of equalization. The correlation between per capita prefectural tax revenue and per capita local allocation tax is ­0.71. As a result, per capita revenues from general sources in the area with low tax bases increases considerably. A surprise is that the coefficient of variation in prefectural tax revenue, accounting for 0.2408, differs little with that in general revenue, which accounts for 0.2293. This phenomenon resulted in a reversal of the rank-ordering of disparities among prefectures rather than a deterioration of the equalization effect. The degree of reversal in the relative wealth of prefectures can be measured by the rank-ordering correlation. The rank order correlation between per capita prefectural tax and per capita general revenue is ­0.5195. After fiscal transfer, the prefectures with lower tax capacity, as measured by prefectural tax revenue, had the higher total resources, as measured by general revenue. General resources of Aichi, Osaka, and Kanagawa are only a half those of Tottori, Shimane, and Kochi. It may be assumed that the Japanese equalization system reduces geographical fiscal inequalities quite extensively, although many questions about the mechanism remain unsettled.3 MoHA's Role Two ministries dominate when it comes to local fiscal matters in Japan. One is the Ministry of Finance (MoF) and the other is the Ministry of Home Affairs (MoHA). There is often conflict between the two as the latter plays the role of a counterpower in the central bureaucracy against MoF incursions into local matters. Therefore, MoHA is often to said to be "an opposition party within the national government." Today, MoHA has a secretariat, three bureaus, two departments, and a college. Two of these bureaus, the Local Finance Bureau and the Local Tax Bureau, are intimately connected with local fiscal matters. The former deals with planning and implementation of the local finance system. The local allocation tax is the most important fiscal transfer the bureau has devised. The latter is charged with planning and implementation of the local tax system. Since the taxpayer and tax base are the same for national and local taxes, this bureau establishes the organization for Taxes and Transfers in Japan's Local Public Finances 5 national, prefectural, and municipal levels. Based on the framework provided by the Local Tax Law, each local government prepares its own tax bylaws for its tax administration. Table 3: Fiscal Equalization by Local Allocation Tax (47 Prefectures, FY 1993) Prefectural Local General Per capita (thousand of Yen) tax revenue allocation tax revenue Local (hundred (hundred (hundred Prefectural allocation General Classification million) million) million) tax tax revenue A Aichi 9,552 69 10,017 143 1 150 Osaka 11,369 272 12,232 130 3 140 Kanagawa 9,210 124 9,765 115 2 122 B Shizuoka 4,495 927 5,853 122 25 159 Saitama 6,286 1,563 8,145 98 24 127 Chiba 5,652 1,448 7,408 102 26 133 Hyogo 5,738 2,286 8,400 106 42 155 Kyoto 2,800 1,213 4,221 108 47 162 Tochigi 2,188 1,240 3,621 113 64 187 Ibaragi 3,122 1,742 5,049 110 61 177 Fukuoka 4,363 2,394 7,039 91 50 146 Gunma 2,121 1,203 3,513 108 61 179 Hiroshima 2,930 1,837 4,971 103 64 174 Gifu 2,201 1,546 3,927 107 75 190 Shiga 1,398 1,079 2,574 114 88 211 Mie 1,949 1,469 3,586 109 82 200 Miyagi 2,342 1,661 4,183 104 74 186 C Okayama 1,925 1,752 3,830 100 91 199 Ishikawa 1,290 1,217 2,688 111 105 231 Nagano 2,266 2,116 4,606 105 98 214 Kagawa 1,064 1,072 2,220 104 105 217 Toyama 1,261 1,388 2,750 113 124 246 Fukushima 2,120 2,222 4,549 101 106 216 Nara 1,150 1,348 2,580 84 98 288 Fukui 1,082 1,152 2,317 131 140 281 Yamaguchi 1,505 1,757 3,394 96 112 216 Niigata 2,539 2,828 5,607 103 114 227 Yamanashi 879 1,228 2,202 103 144 258 D Hokkaido 5,205 7,114 12,837 92 126 227 Ehime 1,253 1,854 3,224 83 122 213 Wakayama 941 1,560 2,600 88 145 242 Kumamoto 1,364 2,306 3,816 74 125 207 6 Nobuki Mochida Prefectural Local General Per capita (thousand of Yen) tax revenue allocation tax revenue Local (hundred (hundred (hundred Prefectural allocation General Classification million) million) million) tax tax revenue Oita 1,003 1,865 2,984 81 151 241 E Yamagata 1,007 1,938 3,068 80 154 244 Saga 740 1,476 2,288 84 168 261 Nagasaki 1,095 2,320 3,528 70 148 226 Iwate 1,093 2,489 2,714 77 176 262 Kagoshima 1,200 2,743 4,085 67 153 227 Tokushima 688 1,565 2,319 83 188 279 Miyazaki 815 1,958 2,862 70 167 245 Okinawa 735 1,800 2,605 60 147 213 Akita 919 2,198 3,218 75 179 262 Aomori 1,041 2,471 2,626 70 167 245 Tottori 486 1,356 1,914 79 220 311 Shimane 615 1,848 2,546 79 237 326 Kochi 590 1,863 2,531 71 226 307 F Tokyo 23,191 - 24,447 196 - 206 Average 138,779 80,878 229,456 112 65 186 Source: Ministry of Home Affairs Note 1: general revenue means the sum of prefectural tax, local allocation tax, and local transfer tax. Note 2: 47 prefectures are grouped into 5 categories based on the index of fiscal capacity. A 1.0~, B 0.5~1.0, C 0.4~0.5, D 0.3~0.4, E~0.3 Division of Expenditure Responsibility It is local government that shoulders the responsibility of Japan's domestic administration. Almost all administrative functions closely connected with the daily life of the nation are carried out by local government. As a result, local public finance accounts for approximately two-thirds of the public expenditure burden, on the basis of final disbursement. Table 4 shows how expenditure responsibilities are shared between national and local governments across a range of public services. As can be seen, the central government directly performs relatively few public functions such as national defense, pension-related public welfare expenditure, and expenditure to repay a debt. About 80 percent of the disbursements of the national government's general account are simply transferred to other accounts, with local government gaining the largest share. In contrast, local governments are responsible for a major share of public spending, including that for Taxes and Transfers in Japan's Local Public Finances 7 national land conservation and development, school education, social education, police and fire-defense, social welfare, sanitation, and general administration. The strength of Japan's system is that considerably more public spending takes place at the local level than at the national level. The assignment of expenditure responsibilities is determined by national legislation (such as the Local Finance Law and the Local Autonomy Law) and cannot be altered at the discretion of the central government. The central government has no legal right to issue unfunded mandates to local government. Nevertheless, the national government remains heavily involved in almost every aspect of local public spending. Unlike the American and Canadian systems, there is no clear separation of central and local function. As a result, major programs (education, health, public works) are formulated by national ministries and financed by many specific grants. Therefore, the issue for Japan is not so much to change/enlarge the expenditure assignments themselves, but to redefine responsibilities for designing, implementing, and financing these assignments. Table 4: Expenditure Shares between National and Local Governments (FY1994) Expenditure Category Percent Spent Through (Percent share of total in parenthesis) National Government Local Government General government (13.2) 19 81 National defense (3.3) 19 81 National land conservation (21.2) 21 79 Industry and economy (7.1) 21 79 Education (14.8) 13 87 Social welfare (22.5) 33 67 Pension (1.3) 92 8 Debt service (15.2) 62 38 Other (1.4) 100 0 Average 34.5 65.5 Source: Ministry of Home Affairs, Local Public Finance System, 1996. In this respect, both the reform of agency delegated functions and reduction of the national government disbursement for specific purposes are of great importance. In addition to ADF, specific-purpose grants are distributed on the condition that the recipient follow the directives issued by the national government. If a local government fails to observe national directives, it is asked to refund the disbursement in whole or in part. A basic principle that underlies national government control seems to be uniformity throughout the country. However, detailed conditions attached to grants lead to waste and administrative inefficiency, and they do not sufficiently take local preferences into account. 8 Nobuki Mochida Characteristics of Local Tax System in Japan Tax Assignment4 In Japan, the ratio of total tax burden to GDP of 27­28 percent does not seem as high as those of other major OECD countries. Total tax revenue in fiscal year 1994 amounted to ¥86.5 trillion, 62.4 percent is national taxes and 37.6 percent is local taxes.5 The ratio of local tax of local governments total revenue is 35.2 percent, which is not always low from the viewpoint of international comparison.6 Every local government is authorized by the Local Tax Law to levy and collect several kind of local taxes. Final authority to levy local tax, however, is guaranteed by local ordinances and bylaws enacted by each local assembly. If a local assembly does not establish local ordinances and bylaws, the taxpayer has no obligation to pay taxes to the local government. A good local tax system should satisfy several criteria. The first is revenue response to economic growth. In the long run, it is desirable that local revenue increase and decrease in line with local expenditure needs. Although a buoyant tax base allows windfall revenue gains to local government, this problem can be overcome provided that the long-run local elasticity of the tax base to economic growth is equal to one (Bennett and Krebs 1987, p. 251). It should be noted that unlike United States and the United Kingdom, where local governments rely predominantly on property tax, Japan's local tax system does well in revenue response to economic growth. This is mainly because the major source of local own-revenue is a kind of tax-base sharing that is similar to a surtax on the national income tax base. Approximately 60 percent of prefectural taxes revenue and 40 percent of municipal tax revenues are imposed on the income of individuals and corporations. There is good evidence to show that elasticities of local taxes are higher than unity. Table 5 indicates that elasticity of tax revenue to economic growth is 1.26 and 1.35, respectively, for prefectures and municipalities during 1971­90. Among local taxes, municipal individual inhabitants tax is highest, at 1.74, prefectural individual inhabitants tax accounts for 1.43, and municipal corporation inhabitants tax accounts for 1.42. Contrary to general belief, the responsiveness of property tax, which is called the fixed-assets tax by MoHA, is not less than unity, primarily because of the sharp rise in market value of land in the late 1980s and the assessments made at regular intervals. According to these elasticities, the share of local tax in total tax revenue is relatively high in comparison with other unitary states. The individual inhabitant tax is a burden-sharing tax--all residents are required to share the cost of maintaining local community functions according to their ability to pay. It can be likened to a membership fee for being a part of the local community.7 The inhabitants tax has two forms, each based on a different tax source: per capita and income. The former is a lump-sum component, while the latter is levied on income in a manner similar to the collection of the national individual income tax. The inhabitants' income tax, however, is assessed on the income of a year previous to the income assessed in the national income tax. Generally speaking, the inhabitants' tax is the best choice for raising a large amount of local tax revenue, because it places the responsibility on as many inhabitants as possible to finance local public services. Taxes and Transfers in Japan's Local Public Finances 9 The second criteria is small revenue fluctuations over time. Strong fluctuations in revenue during the business cycle can be regarded positively for a national tax, but is not for local taxes. First, local expenditures are fairly continuous and revenue fluctuation makes planning difficult. Second, local expenditures should not run contrary to national economic policy, although the scope for local authority to pursue a countercyclical budget policy is rather limited. As Table 5 indicates, the fixed-assets tax produces fairly stable revenue. In contrast, corporation inhabitants tax and enterprise tax fluctuate strongly during business cycle, since these taxes are generally imposed on net income, not on sales or turnover. Individual inhabitants tax fluctuates less than corporation inhabitants' tax and enterprise tax. Apparently the instability of enterprise tax revenue is its most serious problem, because of its large share in prefectural tax revenue.8 Introduction of a new tax base such as sales, capital, or value added taxes has been suggested in order to make tax revenue less sensitive to business conditions. The introduction of a local consumption tax in fiscal year 1997 may also be a first step toward revenue stability.9 The third criteria is distribution among local authorities. A local tax system should produce a relatively balanced distribution of revenue among local government in relation to their expenditure needs. Large differences in the tax base between localities may cause many undesirable effects that require intergovernmental fiscal equalization. Tobacco tax levied on the number of cigarettes score highly in balanced distribution of tax revenue among localities. The base of the fixed-assets tax is also evenly distributed throughout the country. It should be stressed that regional disparity in financial capacity has been gradually reduced during postwar, high economic growth era. Table 5 indicates that the Gini coefficient of per capita prefectural tax revenue decreased from 0.2297 in fiscal year 1965 to 0.1113 in fiscal year 1994. However, a marked difference is still observed in per capita prefectural tax revenues among localities in fiscal year 1994, the largest in Tokyo, ¥183,000, the smallest in Okinawa, ¥56,000, corresponding closely to the difference in their economic resources and the per capita income of the inhabitants. The fourth criteria is local fiscal autonomy and fiscal equivalence. The power to determine the tax rate and base allows local variations in fiscal burdens to be sensitive to local preferences, which should encourage fiscal accountability. Despite strict uniformity,10 there are two options available to local government for setting the tax rate and base in Japan. One is that the central government sets fixed tax rate for a number of local taxes, but provides ranges for some others. Each local authority can use the standard tax rate with an upper-limit set by MoHA and MoF. 10 Nobuki Mochida Table 5: Evaluation of Local Tax System Criteria for local tax system Tax revenue in FY 1995 Distribution among Fiscal equivalence localities Revenue Elasticity1 Fluctuation2 Actual range of tax rate FY1965 FY 1994 (number of localities) million % Yen (Gini (Gini below at over coefficient) coefficient) standard standard standard tax rate tax rate tax rate Prefectural tax Inhabitants tax individual 2,662 19.1 0.261 0.1504 0 47 0 1.43 0.88 interest rate 991 7.1 - 0.1760 0 47 0 - - corporation rate 805 5.7 0.3192 0.1374 0 1 46 1.13 1.49 Enterprise tax 4,235 30.4 0.3225 0.1708 0 40 7 1.26 1.28 Prefectural tobacco tax 378 2.7 0.0823 0.0583 0 47 0 0.90 1.52 Light oil delivery tax 1,332 9.5 0.1446 0.1403 0 47 0 1.15 0.99 Automobile tax 1,587 11.4 0.1924 0.0829 0 40 0 1.08 0.86 Real property acquisition tax 787 5.6 0.2540 0.1557 0 47 0 1.27 1.05 Automobile acquisition tax 611 4.3 - 0.1132 0 47 0 1.21 0.77 Others 520 3.7 - - - - - - - Sub-total (A) 13,908 100.0 0.2297 0.1113 - - - 1.26 0.74 Municipal tax Inhabitants tax individual 6,532 33.0 0.2184 0.1661 0 3237 0 1.74 0.75 corporation rate 2,273 11.4 0.3017 0.1611 0 1785 1237 1.42 1.04 Municipal tobacco tax 669 3.3 0.0827 0.0583 0 3237 0 0.90 1.50 Fixed assets tax 8,429 42.6 0.1552 0.1268 0 2944 289 1.40 0.52 City planning tax 1,304 6.5 - 0.4031 - - - 1.49 1.05 Others 559 2.8 - - - - - - - Sub-total (B) 19,766 100.0 0.1694 0.1441 - - - 1.35 0.61 Source: MOHA, Statistical yearbook for local tax (Chihozei ni kansuru sankokeisusiryou), for each year. Notes: 1) Measured as elasticity of tax revenue to growth during 1971-90. 2) Measured as the coefficient of variation during 1971-90. Taxes and Transfers in Japan's Local Public Finances 11 But principle and practice differ. Good evidence for this is presented in Table 5. There are no localities with a tax rate below the standard tax rate, because these localities are prohibited from issuing local bonds by the Local Public Finance Law.11 It is difficult to find any tax competition among local governments in Japan. At the same time, all but one prefecture raised corporate tax over the standard tax rate, but they did not increase personal tax for fear of the electoral consequences. Excess tax revenues levied by local governments over the standard tax rate represent only ¥1,878 hundred-million at the prefectural level and ¥4,751 hundred-million at the municipal level. The former accounts for only 1.3 percent of total prefectural tax revenue, and latter for 2.3 percent of total municipal tax revenue. As a result, almost all localities use a uniform rate for the same tax base. For example, in fiscal year 1996, 2,944 out of 3,233 municipalities applied the same standard tax rate to the property tax base. This suggests that there is strong preference for equal access to public services and equitable sharing of the burden in Japan. The other option is concerned with the imposition of new taxes not listed in the law. Local government is given the authority to propose new taxes, but must seek the approval of the MoHA and MoF. In fiscal year 1996, only 14 prefectures and 21 municipalities were given permission to use nonlisted taxes such as a nuclear fuel tax on nuclear power plants. Local governments in Japan have relatively large receipts from local taxes, but since the flexibility in determining the tax base and rate is strictly limited, it is difficult to see how they can be accountable to their constituents at the margin, as both efficiency and local autonomy require.12 Local Public Finance Program Japan's intergovernmental system is well-designed to enforce fiscal responsibility. The probability of a local government going bankrupt or getting itself into severe financial difficulties is less than in North America or Western Europe. As Reed (1986) points out very clearly, Japan is like France in the sense that the central government takes responsibility for enforcing proper financial practices on local government, while in other countries this responsibility lies more with the local electorate and the banking system. In this regard, attention should be paid to the role of the Local Public Finance Program. The Local Public Finance Program serves as a tool to estimate annual aggregate local revenue sources to cover standardized local spending. MoHA assumes the role of formulating the Local Public Finance Program every year, and it has primary responsibility for ensuring that local governments have enough revenue to balance the program. On the expenditure side, the Local Public Finance Program covers the whole of local governments` standard activities except for special local public enterprise accounts, which are run on an independent profit system, and a few other special accounts. On the revenue side of the program, it covers all the standard local revenue sources such as local taxes, local allocation tax, national disbursement, local loans, fees, and tuition. The most important function of the Local Public Finance Program is to ensure fiscal responsibility, because if the estimated program does not balance in the year, MoHA must propose 12 Nobuki Mochida some measure such as local tax amendments, increases in the Local Allocation Tax, or an increase in local loans. As the following episode indicates, a kind of special measure, such as borrowing from the Fund Management Board of the MoF and a deficit-covering bond issue, is not determined automatically, but is based on arbitrary political negotiation between MoHA and MoF. MoHA is responsible for the Local Public Finance Program, and it negotiates very hard with the MoF in order to secure sources of revenue for local governments. In principle, the tax-sharing ratio of Local Allocation Tax must remain unchanged, even if the total amount of the financial shortage exceeds the legal amount of local allocation tax. The national government is required to raise the tax-sharing ratio if the legal amount of local allocation tax differs "continuously" and "remarkably" from the financial shortage. But this fundamental principle could not be applied strictly in the post-rapid-growth era. In practice, the short-term borrowing from the Fund Management Board of the MoF and issue of deficit-covering local bonds played a key role in local public finance in addition to raising the tax-sharing ratio. This can be found in table 6 which summarizes the "Special Measure concerning Local Public Finance" after the oil crises. The discussion was focused on whether the tax- sharing ratio would be altered based on the provision of local allocation tax law (clause 2, article 6-3). By fiscal year 1984, both short-term borrowing from the Fund Management Board of the MoF and the issue of deficit-covering local bonds played a key role in the local public finance. In fiscal year 1977, while MoHA and the representatives of local authorities claimed a rise in the tax-sharing ratio of 5 percent, MoF has rejected this request because of the huge financial deficit in the national budget. As a result, the following memorandum was confirmed between the minister of finance and the minister of home affairs in 1977: (a) to make up for the amount of financial shortage by increases in both the local allocation tax and deficit-covering local bonds; and (b) to increase the amount of local allocation tax by transferring provisional local grants from the general account and by borrowing from the Fund Management Board of the MoF. In the latter case, it was agreed to redeem a half amount of the principal and the total amount of interest by the burden of the general account of national budget.13 However, revenue of the three national taxes increased steadily in the "bubble economy" of the late 1980s. The amount of financial shortage, therefore, has been reduced quite extensively. In fiscal year 1984, the following new memorandum was confirmed between the two ministers: (a) to suspend borrowing from the Fund Management Board of the MoF as a rule after fiscal year 1984; (b) to redeem half the amount of both principal and interest by the burden of each national and local government; and (c) to transfer a special addition of local allocation tax from the general account of national budget , in place of borrowing from the Fund Management Board of the MoF. Taxes and Transfers in Japan's Local Public Finances 13 Table 6: Special Measure Concerning Local Public Finance (hundred million yen) Fiscal Year 1975-79 1980-84 1985-89 1990-97 1. Amount of financial 28,046 20,593 (100) 16,514 (100) 46,213 (100) shortage (100) 2. Increase in local 15,413 9,757 (47.4) 2,709 (16.4) 24,783 allocation tax (54.9) (53.4) -borrowing from Trust 8,932 (43.3) 900 (5.4) Fund Bureau 14,408 825 (4.0) 1,809 (10.9) 20,812 -other (51.3) (45.6) 1,005 (3.5) 3,971 (8.6) 3. Increase in local bond 12,676 10,836 (52.6) 11,872 (71.9) 21,430 (100) (46.4) 4. Increase in local tax 0 (0) 0 (0) 1,933 (11.7) 0 (0) Note: Figures in parentheses are percentage of the amount of financial shortage. All figures are average per year. Source: Ministry of Home Affairs, 1989.. Intergovernmental Fiscal Transfer and Regional Disparity14 Evolution of Fiscal Equalization The first regular scheme for equalizing local finance was the local distribution tax in 1940, which was carried out in connection with tax reform of central and local governments corresponding to the quasi-war situation.15 The local distribution tax was a kind of national tax, and the proceeds were shared with local units. The funds were distributed among localities without restriction, not by the tax source principle, but by a formula designed to provide equalization. However, the local distribution tax had some defects from the viewpoint of local autonomy. First, the tax-sharing ratio varied in practice from year to year, partially in accordance with the fluctuation in receipts caused by the sensitivity of income taxation, and partially in keeping with the fiscal deficit in national finance. Second, in the distribution tax, the total amount to be given to individual local units was divided into two parts, which were apportioned separately: one according to the need for services, the other according to fiscal capacity, and these parts had no relationship with each other. A big change in the basic structure of the fiscal equalization system was brought about by the U.S. Occupation after World War II. Great stress was placed on the importance of local autonomy in a democratic nation, and the prewar system was completely restructured in order to encourage decentralization. In accordance with the 14 Nobuki Mochida Shoup Recommendation, the distribution tax was converted to the local finance equalization grant in 1950 (chihozaisei heiko-kofukin).16 It is true that the equalization grant was more reasonable than the distribution tax. The equalization grant was computed by means of a formula that contained two parts, the first relating to the measure of the local need for basic services, and second relating to the measure of local financial ability.17 The total financial capacity was then subtracted from total financial need, and the difference served as the basis for computing the grant for each locality. In the case of the equalization grant, the total amount was determined more closely in accordance with the difference between fiscal needs and resources of localities, irrespective of national tax revenue. Four years of experience revealed that it had not worked as well as had been hoped. The aggregate sum of the grant was not paid out of the general funds of national government as computed by the formula, but was determined every year, taking into consideration, among other things, the degree of stringency in national finance. So, every year it gave rise to friction between local and national officials in the determination of the total amount. In view of these considerations, the equalization grant was abolished in 1953, and in its place the local allocation tax (LAT) was introduced in 1954. Legal Framework and Operation18 LAT is governed by the local allocation tax law. This law stipulates that LAT should be based on a uniform formula; the final authority to approve the distribution lies with the National Assembly. According to the law, the MoHA is responsible for the operation (calculating the amount of LAT) of the transfer and for determining modification coefficients.19 Not granting MoHA the final authority to approve the formula and unit costs is an important mechanism to deter any attempt to manipulate distribution. A certain degree of flexibility is also given to MoHA, because it has the authority of determine modification coefficients, which marginally affect the distribution of LAT. In addition, MoHA has the responsibility to collect data, which are used for the calculation of LAT and to put them in order. And each governor is duty-bound to present these data to MoHA, and each mayor is obligated to present these data to the governor. What kind of role do local governments have on operating LAT? In the LAT for prefectural government, all staff are bound only to collect data and present them to MoHA and the Local Autonomy Information Center. Paradoxically, only MoHA is calculating LAT.20 This legal framework ensures that no single locality or senior official effectively influences the distribution of LAT in favor of a particular region without affecting many other regions. Taxes and Transfers in Japan's Local Public Finances 15 Practical Effects of Fiscal Equalization21 Now we proceed to analyze the practical effects of the Japanese system on the general revenue of local bodies. To determine the actual degree of equalization achieved, per capita local allocation tax is added to the per capita local tax in order to obtain a notional total reflecting the area's resources after the addition of the local allocation tax--this is termed general financial resources (GFR). The disparity is then determined, as measured by the Gini coefficient, in the GFR, and it is compared with the initial disparity in local tax per capita. The extent of the improvement (or deterioration) can then be measured as the difference between the Gini coefficient of local tax and that of GFR, divided by the former. This measure can be expressed by the following equation: ø = ( G2 ­ G1 ) / G2 Where G1 stands for the Gini coefficient of GFR, G2 for the Gini coefficient in local tax, and ø denotes the extent of the improvement (I have termed this the equalization coefficient in this chapter). Figure 1 indicates the change in the extent of improvement measured by the Equalization Coefficient. As Figure 1 demonstrates, the extent of improvement has changed drastically every ten years. Figure 1. Extent of Equalization by LAT 0.20 0.8 0.18 GINI coefficient of per capita regional income 0.7 0.16 coefficient of equalization income 0.6 0.14 0.12 0.5 equalization regional of in 0.10 0.4 0.08 0.3 0.06 disparity 0.2 coefficient 0.04 0.02 0.1 0.00 0 1950 65 80 fiscal year 16 Nobuki Mochida Figure 2: Regional distribution of fiscal resources 0.35 per capita local tax 0.30 per capita general revenue 0.25 0.20 coefficient 0.15 GINI 0.10 0.05 0.00 1950 65 80 fiscal year THE FIRST HALF OF THE RAPID-GROWTH ERA (1954 TO 1964). Increase in pre- grant disparity is a feature of this period. The disparity in financial resources among rich and poor local authorities became larger and was maintained at a high level. A large number of young people moved from rural areas to the metropolitan areas such as Tokyo, Osaka, and Nagoya. To deal with this social problem, the political slogan of "Improvement of Regional Disparity" became one of the main national policy goals and was embodied in the National Comprehensive Development Plan, established in October 1962. In line with this national policy guideline, local allocation tax was distributed mainly to the backward districts in inverse proportion to their financial capacities. As a result, local allocation tax served to reduce resource disparities by 70 percent each year. THE SECOND HALF OF THE RAPID GROWTH ERA (1965­74). The reduction of pregnant disparities and the reversal of the rank-ordering is a distinctive characteristic of this period. There was a sharp decrease in the disparities among rich and poor districts. The Gini coefficient of per capita regional income decreased from 0.1248 in fiscal year 1965 to 0.0753 in fiscal year 1975. This improvement in regional disparities was not caused by the success of the National Comprehensive Development Plan, but by the dispersion of factories around the country and the increase in the number of people employed in the local public works. Nevertheless, the distribution of the local allocation tax followed the principle of equalization all the more. As a result, resources disparities actually increased after the equalizing effect of the local allocation tax is taken into Taxes and Transfers in Japan's Local Public Finances 17 account (see figure 2). However, this increase actually resulted in a reversal of the rank- ordering of disparities among prefectures. We should notice that the sharp decline in the equalization coefficients means enforcement of improvement rather than the deterioration of the equalizing effect. THE OIL CRISES AND AFTERWARDS (THE MID-1970S TO THE MID-1980S). Gradual increase in pregnant disparity is a characteristic of this period. During this time, the disparities in the per capita local tax began to increase again as a result of population concentration in the Tokyo metropolitan area caused by the internationalization of financial market. As figure 2 demonstrates, the Gini coefficient of the per capita local tax increased gradually after the oil crises. At the same time, the negative correlation between per capita tax revenue and the per capita local allocation tax became weaker, because of a shortage in the total amount of local allocation tax. As a result of these trends, reversal of the rank-ordering of disparities among prefectures was corrected somewhat. THE BUBBLE ECONOMY AND THEREAFTER (1985 TO THE PRESENT). A sharp reduction of pre-grant disparities and a reversal of rank-ordering is a characteristic of this period. There was a marked decrease in the regional disparities, as figure 2 indicates. The Gini coefficient of local tax declined from 0.19 in fiscal year 1988 to 0.15 in fiscal 1993. As a result, the equalization coefficient has dropped drastically, from 0.4120 in fiscal year 1988 to 0.085 in fiscal 1993. It is noteworthy that there is little difference between pre- grant disparities and the area's resource disparities after the addition of local allocation tax. However, these trends do not mean there was a deterioration of the equalization effect, as mentioned above, but instead a reversal of the rank-ordering of disparities among prefectures. These trends can be explained by both fundamental tax reform and the collapse of the bubble economy. Implications of the Japanese Experience Relevance to Developing and Transitional Economies The defining characteristic of Japan's system of intergovernmental fiscal relation has been the strong collective preference for equal access to public goods. While local autonomy was also a popular objective, especially among progressive or left-wing politicians, equal access to public goods and fair sharing of the burden of financing these goods were viewed as essential for economic and social development. Interregional redistribution was, therefore, the central issue for Japan's system of intergovernmental fiscal relations. A Japanese-style approach, with the assignment of expenditure responsibilities determined by national legislation (such as the local finance law and the local autonomy law), and not alterable at the discretion of the central government, may be a good solution for societies that have substantial regional tensions. The central government must have no legal right to place unfunded mandates on local government, but 18 Nobuki Mochida the central government does exercise its discretionary powers on local expenditure through the delegated function, financed by specific-purpose grants. In developing and transitional economies, income disparity across regions tends to worsen after the initial decentralization efforts. Japan experienced large regional disparities in the early stage of postwar economic development. The Japanese government responded, and attention should be paid to the significant role played by the local allocation tax.22 LAT is distributed according to a uniform formula based on basic financial need and basic financial capacity. The application of the formula contributed to the removal of intense negotiation and lobbying during postwar development. LAT is also paid annually to local governments with basic financial needs that exceed their financial capacity, and it varies directly with local fiscal needs and inversely with local fiscal capacity. Such an approach corrected horizontal fiscal imbalance in Japan. Finally, LAT is not a kind of general grant, but a shared-tax system. An automatic increase in major national taxes was the cause of a continuous increase in the financial pool of the local allocation tax during the rapid growth era. Current Issues in Japanese Local Finance The current system faces considerable challenges in the medium term, given the changing preference of the public with respect to local autonomy. In the 1990s, Japan has faced the second transitional phase since World War II. This means a shift away from a society that emphasizes equal access to public services and equitable sharing of the burden of paying for them, toward a society that gives priority to the individual citizen's expressed preference. Where local governments are unable to set their own tax rates, the concept of local accountability does not function effectively. The current system needs to evolve during the process of fiscal decentralization in order to redefine expenditure responsibilities, gain more flexibility in tax rate setting, and to enhance transparency in the equalization transfer scheme. In Japan, the national government remains heavily involved in almost every aspect of local public spending. Unlike the American and Canadian systems, there is no clear separation of central and local function. As a result, major programs (education, health, public works) are formulated by national ministries and financed by many specific grants. Therefore, the issue for Japan is not so much to change and enlarge the expenditure assignments themselves, but to redefine responsibilities for designing, implementing, and financing these assignments. In this respect, both reexamination of agency-delegated functions and reduction of the national government disbursement for specific purposes are important. Detailed conditions attached to grants lead to waste and administrative inefficiency, and they do not sufficiently take into account local preferences. Japan's local tax system scores highly in revenue response to economic growth. However, the corporation inhabitants tax and the enterprise tax fluctuate greatly during the business cycle, since these taxes are generally imposed on net income, not on sales or turnover. A marked difference is still observed in per capita prefectural tax revenues among localities. Local governments in Japan have relatively large receipts from local Taxes and Transfers in Japan's Local Public Finances 19 taxes, but since the flexibility in determining the tax base and rate is strictly limited, it is difficult to see how they can be accountable to their constituents at the margin, as both efficiency and local autonomy require. Therefore, the issue for Japan is how to make the local tax system more accountable and ensure stability of tax revenue. Local accountability is indispensable not only for taking into account local preference but also for macroeconomic control. We can draw some relevant points from European experience for Japan facing with heavy fiscal deficit. As measured by the real rate of growth in local government spending, both England and Norway are countries where control appear to be easy to apply because their local governments do not have their own taxation of much importance. On the other hand, Denmark and Sweden, who rely on their own taxation and local accountability, did better. This seems to confirm that even when local authorities are free to decide on their own income tax rate, it is through negotiations would it be possible to have the necessary macroeconomic control. After the burst of so called bubble economy at the beginning of 1990, financial shortage of Japan's local finance has been increasing drastically as indicated by Table 6. On this background local governments as a whole had to borrow huge amounts of short- term money from Fiscal Investment and Loan Program of MoF and had to issue deficit- covering local bonds almost every year. The weight of local bond revenue to total local revenues has reached 15.2 percentage which is the worst since the World Was II. In 1998, fiscal deficit of Japan's general government accounted for 4.7 percentage of GDP, of which 1.9 percentage was the share of local government. Of course there are many policy options to reduce fiscal deficit, but it should be stressed that own taxation results in local accountability and that negotiation system is better than formal control through grants and tax ceiling (on this point, see Mochida and Lotz 1999). Given Japan's history of strong collective preference for equal access to public goods, it is unrealistic to imagine that local autonomy will evolve toward a system that will allow substantial regional differences to reemerge. Therefore, a role for LAT will remain. However, the present LAT system is not a complete one, but is still evolving. An effective intergovernmental transfer system, in general, should satisfy several criteria. The first criteria is revenue adequacy. The local allocation tax is not a kind of general grant, but a shared-tax system. An automatic increase in major national taxes was the cause of a continuous increase in the financial pool of the local allocation tax during the rapid growth era. At the same time, total fund of transfer is sensitive to business conditions because a major component of the fund is income-elastic national taxes. During the period of 1970­95, the rate of increase in the financial pool for transfer has changed between ­14.1 percent and 43.5 percent every year. Indeed, both short-term borrowing from the Fund Management Board of the MoF and deficit-covering local bond issues play a key role in filling the gap between total entitlement of local allocation tax and the financial pool of the transfer in the post-rapid-growth era. A future reform necessary for revenue adequacy is to make the financial pool less sensitive to business conditions and more stable.23 The second criteria is local tax effort. Basic financial revenue is measured by using figures of the major tax base and the standard tax rate. To retain incentives for local 20 Nobuki Mochida government to collect their own tax, basic financial revenue is calculated based on the prescribed percentage of the sum of local tax revenues. Regions with high tax effort are not penalized, and regions with low tax effort are not encouraged. As for tax effort, however, the local tax system is a question. Concerning local taxes, the base and rates of general taxes cannot be determined by the independent initiatives of local government under the local tax law. The tax base and the tax rates can be altered by the proposal of both the MoHA and the MoF. This implies that a uniform rate is levied on the same tax base in all prefecture and municipalities. Present local tax systems should be changed into more flexible systems, in which the tax rate is determined at the discretion of local governments. In order to win national taxpayer's confidence, on the other hand, recipient local governments should express their efforts to reduce LAT by enlarging the relevant tax bases or using a non-listed local tax. The third criteria is equity. Because local allocation tax is paid annually to local governments with basic financial needs that exceed their basic financial capacity, it varies directly with local fiscal needs and inversely with local fiscal capacity. Such an approach actually corrected horizontal fiscal imbalance in Japan. Before the 1970s, the transfer system contributed significantly to equality. But after that, as regional fiscal disparities have fallen over time, there has been less inequality to fix through local allocation tax and the intensity of the equalization effect has fallen. Future reform, therefore, should be carried out based not only on equity criteria, but also on efficiency grounds in order to eliminate or reduce "differential net fiscal benefits" which encourage fiscally induced migration.. The fourth criteria is transparency and stability. Local allocation tax is distributed according to a uniform formula based on basic financial need and basic financial capacity. The application of the formula contributed to removing intense negotiation and lobbying during postwar development. However, calculation of the transfer became too complicated for local governments to forecast their own revenue(including the transfer) in order to prepare their budgets. A kind of special measure, such as borrowing from the Fund Management Board of the MoF and deficit-covering bond issues, is not determined automatically, but based on arbitrary political negotiation between MoHA and MoF. Future reform is necessary to strengthen the transparency of the present system. Endnotes 1. Vertical fiscal imbalance is the disparity between revenue means and expenditure needs at various levels of government. This results from the division of expenditure responsibilities and revenue-raising powers between the central and local governments. 2. For the historical development of the equalization scheme in Japan, see Mochida 1993 chapters 4 and 5. 3. It is to be noted that these figures refer merely to the per capita amount of the prefecture. Financial needs for local function are not necessarily proportional to the number of inhabitants. In a sparsely populated district, for example, per capita revenue Taxes and Transfers in Japan's Local Public Finances 21 produces a large figure, in spite of the low level of accomplishment of services. A densely settled district, however, requires fiscal means beyond the average. 4. This section is based on Ishi (1993) and MoHA (1996a, b, c). 5. The major components of total tax revenue are income tax, consumption tax, and property tax. The amount of income tax was ¥50.2 trillion in fiscal 1994, of which 65.3 percent is national tax, 18.7 percent is prefectural tax, and 16.0 percent is municipal tax. The amount of consumption tax is ¥20.7 trillion, of which 76.8 percent is national tax, 19.3 percent is prefectural tax, and 3.9 percent is municipal tax. The amount of property tax is ¥15.6 trillion, of which 34.0 percent is national tax, 4.5 percent is prefectural tax, and 61.5 percent is municipal tax. See MoHA (1996b). 6. Major sources of total annual local government revenues in fiscal year 1993 are local taxes (35.2 percent), local allocation tax (16.2 percent), central government disbursements (14.3 percent), loans (14.0 percent), charges and fees (2.3 percent), and local transfer tax (2.1 percent). 7. The inhabitants' tax is collected by both prefectures and municipalities. In levying this tax, mutual cooperation is established among the municipal, prefectural, and national governments. When the municipal governments levy their inhabitants' tax on individuals, they collect the prefectural inhabitants' tax as well, using the same tax base. Information on taxable income necessary for computing the local inhabitants' tax is given by the national government. 8. In calculating the tax base of the corporate tax at the national level, the prefectural enterprise tax is allowed as a deduction. If a taxpayer has an office within the jurisdiction of two or more prefectures, the tax base is allocated to all prefectures concerned. The allocation is made on the bases of number of employees. 9. For local consumption tax, see MoHA (1996a). 10. All revenue sources are subject to control by the national government under the local tax law. The tax base and rates of major items are legislated by the Diet and can be altered by the proposal of both the MoHA and MoF. 11. Local governments that collect local tax below the standard tax rate cannot apply for permission for debt financing. (See local public finance law, article 5.1.) 12. It should be noted that in the prewar period, the local surtax method, in which a piggyback surtax was applied to the national tax, played a key role, but it was abolished in the postwar era to support local autonomy. Today, each level of local government levies its own taxes, including local income tax, separate from the collection of national taxes. 13. The third point is to carry over special addition of local allocation tax in order to make up for the difference in interest between local bonds placed on the market and those absorbed by the Fund Management Board of the MoF. 14. This section is based on mainly Mochida (1996). There is valuable literature, written in English, concerning the local allocation tax. See, for example, Ito (1967), Yonehara (1987), and Ishi (1993). 15. But we had already had grants as forerunners in the 1930s. Marked territorial inequalities in per capita prefectural tax revenue were seen during the Great Depression. 22 Nobuki Mochida As device to counter depression, the provisional grant (rinji-chihozaisei hokyukin) was introduced. These grants were used for the salaries of primary school teachers and natural disaster rehabilitation, and they were apportioned among rural districts. 16. The 1949 Shoup Mission played a significant role in shaping the style of the tax system in postwar Japan. On this subject, see Ishi (1993, chapter 2). 17. The local need for each item was computed as the number of units of the service, multiplied by the standard cost per unit of the service at an acceptable, but minimal, quantity and quality. The total need for each locality was the sum of the amounts needed for all basic services combined. The financial capacity of each locality was computed as 70 percent of revenues that all regular local taxes would yield, assuming that they were levied at a standard rate with standard levels of assessment and collection. 18. This section is based on mainly Yamauchi (1996). 19. MoHA has four bureaus; one is the Local Finance Bureau. Within the bureau, there are eight divisions. The Local Finance Division and the Local Allocation Tax Division, which belong to the Local Finance Bureau, are in charge of LAT. The former manages the affairs of the special local allocation tax; the latter, the ordinary allocation tax. The Local Allocation Tax Division employs 18 persons, including a director, 2 assistant directors, and 15 staff members. 20. But the MoHA makes its own opportunity to hear wishes of local governments. It holds a conference at least four times a year to make information available to them formally and informally, and the parties sometimes discuss the matter together during the process of calculation. 21. For detailed arguments for this section, see Mochida. (1996, 1993, 1990), Kaizuka and others (1987), and Ishikawa (1995). 22. For the role of LAT in the Welfare State, see Hayashi (1992). 23. To stabilize the financial pool for LAT, Fujita examined a number of methods, including creating a new special account, calculating the tax base with a 5-year moving average, and returning to the equalization grant introduced by the Shoup recommendation. See Fujita (1972, pp. 143­47). Bibliography Benett, R. J., and G. Krebs. 1987. Local Business Taxes in Britain and Germany. Baden- Baden: Nomos Verlagsgessellschaft. Dubravko, M. 1996. "Intergovernmental Relations and Local Public Finance in Japan." IMF Discussion Paper. Washington, D.C. Fujiwara, T. 1996. "The Role of the Ministry of Autonomy in Intergovernmental Fiscal Relations in Japan." Paper presented at the workshop on International Experience in Intergovernmental Fiscal Transfer: the Experience from Japan and Other Countries. March, 1996 Hanoi. Taxes and Transfers in Japan's Local Public Finances 23 Hayashi, T. 1992. Fukushi Kokka no Zaiseigaku (Public Finance of Japanese Welfare States). Tokyo: Yûhikaku. Ishi, H. 1993. "Local Taxation and Intergovernmental Fiscal Relations." In The Japanese Tax System, 2d ed. Oxford, U.K.: Oxford University Press. Ishikawa, T. 1995. "Growth, Human Development, and Economic Policies in Japan: 1955­1993." University of Tokyo, Faculty of Economics, discussion paper series, 95-F-32. Ito, H. 1967. "Equalization of Local Finance in Japan." Finanzarchiv N. F. 26, Heftl. Boston: Little Brown. Kaizuka, K., and others. 1987. Chihôkôfu-zei no Kinô to Hyôka (Local Allocation Tax--Function and its Evaluation). Ministry of Finance, Financial Review, 4. Tokyo. Ma, J. 1996a. "Intergovernmental Fiscal Transfer: A Comparison of Nine Countries." Paper presented to the Workshop on Intergovernmental Fiscal Transfer: The experiences from Japan and other Countries. Hanoi, March 1996. ------. 1996b. "Reforming Intergovernmental Fiscal Relations in China: The Relevance of Japanese Experience." Discussion paper, World Bank Workshop on Local Government and Economic Development in Japan: Lessons for Economies Undergoing Decentralization." August 3-5,1996 Honolulu. MoHA. 1989. Chihokoufuzeiseido Enkakusi (The History of Local Allocation Tax). Tokyo. ______. 1996a. Local Tax Administration in Japan. Tokyo. ------. 1996b. Local Public Finance System. Tokyo. ------. 1996c. Local Public Finance in Japan. Tokyo. Mochida, N. 1990. "Chihôkôfu-zei Seido no Kôzo to Kinô (The Structure and the Function of the Local Allocation Tax Grants)." Keizai to Keizaigaku (Economy and Economics) No. 65: 61-89. ------. 1993. Toshi Zaisei no Kenkyu (Public Finance of Japanese Cities). University of Tokyo Press. ------. 1995. "Balancing Equity and Decentralization." Social Science Japan 5, Newsletter of the Institute of Social Science, University of Tokyo. ------. 1996. "Japan's Local Allocation Tax--An Equalization Transfer Scheme." University of Tokyo, Faculty of Economics, discussion paper 95-F-32. Mochida, N. and J. Lotz. 1999. "Fiscal Federalism in Practice, the Nordic Countries and Japan." The Journal of Economics, Faculty of Economics, University of Tokyo. 64 (4). 24 Nobuki Mochida Reed, S. 1986. Japanese Prefecture and Policymaking. Pittsburgh, PA: University of Pittsburgh Press. Shibata,T ed., 1993. Japan's public sector : how the government is financed , University of Tokyo Press. Tokyo. Shoup Mission.1949. Reports on the Japanese Tax System, The Shoup Mission, Tokyo. Wolman, H., and Page E. 1987. "The Impact of Intergovernmental Grants on Subnational Resources Disparities: A Cross-National Comparison." Public Budgeting & Finance, autumn. 17(3). World Bank. 1995. "A Proposal for the EDI Research Project on Local Government and Economic Development in Japan: Lessons for Economies Undergoing Decentralization." World Bank Economic Development Institute, Washington, DC. Yamauchi, K. 1996. "By Whom, When and How is the Japanese Local Allocation Tax Calculated?" Paper presented the Workshop on Intergovernmental Fiscal Transfers. July 8-10, 1996 Vienna. Yonehara, J. 1987. "Financial Relations Between National and Local Governments." In Shibata T. ed., Public Finance in Japan, University of Tokyo Press. Taxes and Transfers in Japan's Local Public Finances 25 Appendix Computation Formula of Local Allocation Tax LAT has continued to the present with some minor alterations. The framework of the local allocation tax is founded on the former distribution tax, enforced between 1940 and 1949, while retaining the formula used in the equalization grant for the distribution of funds to localities. In the LAT system, the total amount to be distributed to local authorities is a fraction of yields from major national taxes. The present system is a shared tax, in which a share in the proceeds of national taxes is granted to poor localities, without limitation on use. The calculation proceeds as follows. First, the total amount of the local allocation tax is calculated according to the formula below: (1) TT = 0.32* ( NTy + NTc + NTa ) + 0.29* NTv + 0.25* NTt Where TT denotes total financial pool of transfer, NTy is the total yield of personal income tax, NTc is that of corporate income tax, NTa is that of alcohol tax, NTv is 80 percent of consumption tax revenue, NTt is total yield of the tobacco tax. These prescribed percentages of five major national taxes are apportioned among local bodies in proportion to the amount of the difference between need and revenue. This is expressed by following equation. (2) LATi = Ni ­ Ci Where LATi denotes local allocation tax to ith region, Ni is basic financial need of ith region, and Ci is the basic financial capacity of ith region. It is annually paid to local governments whose basic financial needs exceed their basic financial revenues. Rich localities with revenue that exceeds need are neither eligible for the grants nor liable to contribute money for fiscal adjustment, as is the case in some countries. Before calculating basic financial needs, public services for each prefecture and municipality are divided into some service items (gyôsei-kômoku). In the prefectures, there are 24 service items such as police, roads and bridges, and primary school; for each municipality there are 24 service items such as city planning, parks, garbage collection, and so on. Basic financial needs of ith local authority are calculated according to following formula: (3) Ni =?k ( Iik * Uik * Mik ) Where Iik is a measurement unit for service K of ith region, Uik is unit cost for service K of ith region, and Mik is a modification coefficient for service K of ith region. For each local body, according to the formula mentioned above, basic financial needs for each service item are calculated as the number of measurement units by multiplying the unit cost, adjusted by modification coefficients. The total basic need in each locality is the 26 Nobuki Mochida sum of the amounts needed for all service items combined. The first step is to select measurement units. A measurement unit reflects the number or size of the beneficiaries of a particular expenditure. For example, a measurement unit for education is number of teachers, that of police is number of policemen, and that of roads is length of roads. The second step is to determine a unit cost. Unit cost is a kind of net standard cost per measurement unit for each service item. Assuming a certain local body with standard conditions and scale, the unit cost for each service item is calculated based on following formula. In a prefecture, only one fictitious local body with a population of 1.7 million and a land area of 6,500 square kilometers is assumed as a "standard local body"; in a municipality, population is 0.1 million and land area 160 square kilometers. (4) U = (Cg ­ Rs) / S Where U is unit cost, Cg is gross standard cost, Rs is special revenue, and S is a figure of measurement unit. The third step is to determine modification coefficients. The unit cost, however, is uniform throughout the country, and no regard is given to either the unique services nor to the special circumstances of localities. So an exceedingly complex adjustment is made of the unit cost applicable to such services and localities by means of detailed modifiers determined in accordance with their differences. Modification coefficients are currently classified according to eight categories. On the other side, the basic financial revenue of each locality is expressed as a combined total of two types of revenue: (1) 80 percent in the case of prefectures, and 75 percent in municipalities of the sum of the yields of all regular local taxes, assuming that each is levied at the uniform rate or standard rate prescribed in the local tax law, and (2) the sum of revenues from local transfer taxes. This is expressed in the following equation: (5) Ci = G (Bij* tj + LTTi ) Where G is 0.75 (for a municipality) and 0.80 (for a prefecture), Bij is ith region's jth tax base, tj is the standard tax rate on the jth tax base, and LTTi is revenue from the local transfer tax. There are two reason for adopting such prescribed percentages. First, it is impossible to measure completely the basic financial needs of all local governments with a uniform formula. Second, it is necessary to retain incentives for local governments to collect their own taxes. At the same time, all revenues allotted from the local transfer tax are included, mainly because it is collected by the national government and has no relation to the tax collection effort at the local level. The funds available for transfer calculated in advance, however, do not necessarily cover the sum of the entitlement--that is, the aggregate amount of the deficiencies of local governments with basic financial needs that exceed their basic revenues. The method currently used is either to increase the size of the fund or to adjust the size of the entitlement proportionally, according to the size of the fund. First, some special measure has been taken every year without changing the tax-sharing ratio to increase the size of the Taxes and Transfers in Japan's Local Public Finances 27 pool. These special measures, which will be explained later from an historical perspective, can be divided into following five types: Borrowing from a special account of the Fund Management Board of the MoF Carrying forward local allocation tax Cancellation of a local allocation tax cut Transfer of a provisional local grant Special addition/reduction of the local allocation tax. In addition to the above-mentioned special measure, final adjustment is necessary to adjust the size of the entitlement proportionally, according to the size of the fund, by using an adjustment coefficient. The actual amount of ordinary allocation tax granted to a local government is calculated according to following formula: (6) LATi= (Ni­ Ci) ­ _Ni Where LATidenotes local allocation tax to ith region, Ni is the basic financial needs of ith region, Ci is the basic financial capacity of ith region, and _ is the adjustment coefficient.