EAST ASIA & PACIFIC REGI 178 0 World Bank Discussion Papers China and Mongolia Department China: Reforming Intergovernmental Fiscal Relations Ramgopal Agarwala FILE COPY Recent World Bank Discussion Papers No. 118 Improving the Pe!formance of Soviet Enterprises. John Nellis No. 119 Public Enterprise Reform: Lessons from the Past and Issues for the Future. Ahmed Galal No. 120 The Information Technology Revolution and Economic Development. Nagy K. Hanna No. 121 Promoting Rural Cooperatives in Developing Countries: The Case of Sub-Saharan Afiica. Avishay Braverman, J. Luis Guasch, Monika Huppi, and Lorenz Poh1meier No. 122 Performance Evaluation for Public Enterprises. Leroy P. Jones No. 123 Urban Housing Reform in China: An Economic Analysis. George S. Tolley No. 124 The New Fiscal Federalism in Brazil. Anwar Shah No. 125 Housing Reform in Socialist Economies. Bertrand Renaud No. 126 Agricultural Technology in Sub-Saharan Afiica: A Workshop on Research Issues. Suzanne Gnaegy and Jock R. Anderson, editors No. 127 Using Indigenous Knowledge in Agricultural Development. D. Michael Warren No. 128 Research on Irrgation and Drainage Technologies: Fifteen Years of World Bank Experience. Raed Safadi and Herv6 Plusquellec No. 129 Rent Control in Developing Countries. Stephen Malpezzi and Gwendolyn Ball No. 130 Patterns of Direct Foreign Investment in China. Zafar Shah Khan No. 131 A New View ofEconomic Growth: Four Lectures. Maurice FG. Scott No. 132 Adjusting Educational Policies: Conserving Resources While Raising School Quality. Bruce Fuller and Aklilu Habte, editors No. 133 Letting Girls Learn: Promising Approaches in Primary and Secondary Education. Barbara Herz, K. Subbarao, Masooma Habib, and Laura Raney No. 134 Forest Economics and Policy Analysis: An Overview. William F. Hyde and David H. Newman, with a contribution by Roger A. Sedjo No. 135 A Strategy for Fisheries Development. Eduardo Loayza, in collaboration with Lucian M. Sprague No. 136 Strengthening Public Service Accountability: A Conceptual Framework. Samuel Paul No. 137 Deferred Cost Recovery for Higher Education: Student Loan Programs in Developing Countries. Douglas Albrecht and Adrian Ziderman No. 138 Coal Pricing in China: Issues and Reform Strategy. Yves Albouy No. 139 Portfolio Performance of Selected Social Security Institutes in Latin America. Carmelo Mesa-Lago No. 140 Social Security and Prospects for Equity in Latin America. Carmelo Mesa-Lago No. 141 China's Foregn Trade and Comparative Advantage: Prospects, Problems, and Policy Implications. Alexander J. Yeats No. 142 Restructuring Socialist Industry: Poland's Experience in 1990. Homi J. Kharas No. 143 China: Industrial Policies for an Economy in Transition. Inderjit Singh No. 144 Reforming Prices: The Experience of China, Hungary, and Poland. Anand Rajaram No. 145 Developing Mongolia. Shahid Yusuf and Shahidjaved Burki No. 146 Sino-Japanese Economic Relationships: Trade, Direct Investment, and Future Strategy. Shuichi Ono No. 147 The Effects of Economic Policies on African Agriculture: From Past Harm to Future Hope. William K. Jaeger (Continued on the inside back cover.) China: Reforming Intergovernmental Fiscal Relations  認〕， Copyright 0 1992 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing September 1992 Discussion Papers present results of country analysis or research that is circulated to encourage discussion and comment within the development community. To present these results with the least possible delay, the typescript of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply die expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of publications, which contains an alphabetical title list (with fiffl ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'l&na, 75116 Paris, France. ISSN: 0259-21OX Ramgopal Agarwala is economic adviser in the World Bank's Office of the Regional Vice President, East Asia and Pacific Regional Office. Library of Congress Cataloging-in-Publication Data Agarwala, Ramgopal. China : reforming intergovernmental fiscal relations Ramgopal Agarwala. p. cm. - (World Bank discussion papers 178) Includes bibliographical references. ISBN 0-8213-2223-0 1. Intergovernmental fiscal relations-China. 1. Tide. II. Series. HJ1405.A38 1992 336.51--dc2O 92-23554 CIP -v- Abstract Since 1979, the Chinese authorities have made a sustained effort at introducing market-oriented reforms and the results have been impressive in terms of growth and poverty reduction. However, reform of fiscal system has lagged behind. In particular, the system of intergovernmental fiscal relations has been unstable, non-transparent, and inequitable. As the economy moves toward greater market orientation, a rule-based system of intergovernmental fiscal relations will be essential. The paper reviews the theory and experience of multilevel fiscal systems in various countries and makes some suggestions for China. Its basic conclusion is that while recognizing the specificities of Chinese conditions, the Chinese authorities cannot but follow some basic principles of multilevel fiscal systems. Among these are the following. First, international experience in countries big and small, federal and unitary, indicates that the Central Government must have effective control over the most important sources of tax revenue, not only in terms of tax law and policy, but also in terms of administration, collection, and allocation of revenue. Second, taxes should not be looked upon mainly as providers of revenue; they must also be seen as tools of policy that, whether intended or not, affect the allocation of resources and their efficient utilization, interregional and interpersonal distribution, and the level of aggregate demand. Third, while for the sake of efficiency and equity most of the major taxes have to be collected centrally, a considerable degree of decentralization is required on the expenditure sidle because closeness to the beneficiary helps efficiency. There is thus a need for designing a system for transfer of resources from the center to the localities, but it needs to be done in a way that is transparent and equitable and gives incentives to the localities to maximize their efforts for revenue mobilization. The paper details the practices followed by Japan, India, Canada and Germany as examples on which China can draw in designing its own system of intergovernmental fiscal relations. - v - Acknowledgments This report is based on the findings of a mission that visited China during May 6-18, 1991. In addition to meetings and discussions in Beijing, the mission visited Changsha (Hunan) and Guangzhou (Guangdong). Discussions on the green cover draft were held with the Chinese authorities in Beijing during February 17-20, 1992. The mission comprised Ramgopal Agarwala (mission leader), Raja Chelliah (consultant), Toshihiro Fujiwara, Yuan Wang, Heng-Fu Zou, and Arigapudi Premchand (IMF). The mission was supported in Beijing by Peter Harrold. The host agency for the mission was the Department of Budget Management, Ministry of Finance. Editorial assistance was provided by Caroline K. McEuen. The report draws heavily on recent reports by the IMF and World Bank on the related areas; in particular, on the following Bank reports: China Revenue Mobilization and Tax Policy (1990), China: Financial Sector Policies and Institutional Develpment (1990), and China: Central-Local Fiscal Relations (1991, Draft). - vn - CONTENTS Executive Summary .. . ix I. The Setting ... . ... 1 II. Evolution of the Budgetary Process Since 1980 .. ........ 4 Revenue-Sharing Framework .......................... 4 Framework for Expenditure Control .................. ...... 6 Budget Formulation and Implementation in the Provinces ................................ .. 7 Issues of Control ................................ 9 III. Consequences of the Present Intergovernmental Fiscal System ............................... 11 Adverse Tendencies: Ex Ante ........................ ..11 MixedResults: ExPost ........................... 13 Costs of Ad Hoc Adjustments ........................ 16 IV. Analytical Issues In Constructing a System of Intergovernmental Fiscal Relations ....... ..... 19 Analytical Issues and Case Studies ......................... 19 SeparationofTaxPowers .......................... 24 Restructuring Intergovernmental Fiscal Relations in China: The Long-Term Perspective ................ 27 Grants-in-Aid .................................. 33 Bureaucratic Interface between Center and Localities .................................. 34. V. Transitional Arrangements for Long-Term Reform in the Intergovernmental Fiscal System ... . .......... 35 VI. Concluding Remarks ............................... 317 Appendix: Intergovernmental Fiscal Arrangements In Germany, Canada, India, and Japan .......... o. ............ 38 Statistical Appendix .......................... . . . ..65 References ..................................... 75 - viii - Text Tables 1.1 China: Macroeconomic Indicators ...................... . 3 11.1 China: Center-Local Shares in Revenue, 1990 ..................... 6 111.1 China: Major Economic Parameters Related to the Fiscal System ............................... 18 n1.2 China: Interprovincial Disparities ........................ 18 IV. 1 China: Estimated Percentage Shares of Current Taxes in Total Tax Revenue, 1990 ......................... 30 Appendix A Tables A. 1 Germany: Allocation of Shareable Taxes ................... 39 A.2 Germany: Expenditures by Function and Level of Government, 1974 . . 40 A.3 India: Trends in Vertical Fiscal Imbalance .................. 54 A.4 Japan: Major Prefectural Taxes, FY1989 ................... 57 List of Tables in Statistical Appendix SA. 1 The Revenue-Sharing System in 1980 ...................... 66 SA.2 The Revenue-Sharing System in 1985 ...................... 67 SA.3 The Revenue-Sharing System in 1991 ...................... 68 SA.4 China: Government Finance, 1985-90 .......................... 69 SA.5 Center-Local Shares in Revenue Expenditures ................. ... 70 SA.6 Trends in Investment and Consumption ..................... 71 SA.7 Extrabudgetary Funds ............................... 72 SA.8 Revenue and Expenditure by Province Per Capita .............. 73 SA.9 Tax Revenue and Contracted Revenue Sharing by Province ........ ... 74 - ix - Executive Summary The Setting 1. Since 1979 the Chinese authorities have made a sustained effort to move from a highly centralized planned economy to a partially decentralized system that would combine some key features of the market economy and planning within the framework of dominant public ownership coexisting with other forms of ownership. Among the important characteristics of this reform process are: greater outward orientation in trade and investment; gradual movement toward market determination of prices, except for some key commodities; allowing effective individual control of small units of production in rural and urban areas; and, as a proxy for privatization, using decentralization to sectoral agencies and localities for introducing incentives and market principles. The reform process has been endogenous, with emphasis on incorporating the Chinese characteristics, and has been guided by the principles of gradualism and experimentation. 2. The reform program has, on the whole, been successful. The annual rate of growth of GDP over the last 12 years of reform has been about 9 percent, which is significantly higher than the rate China experienced before reform and higher than that of other countries that are either in the process of transition from socialism to markets or in mixed economies under adjustment programs. During this period, the investment rate has been maintained, on average, at over 35 percent of GDP and the inflation rate at below 10 percent per year. Naturally there have been ups and downs over this period, but China -- unlike other socialist countries in transition -- has managed to avoid a decline in GNP in any year over this period and any prolonged inflationary spiral. There was an incipient inflationary process in 1987-88, but the situation was brought under control; in 1991, the economy seemed to be poised for a growth rate in the range of 6-7 percent, inflation below 5 percent, and a healthy balance of payments situation. It is clear that if the past record of good performance is to be maintained, the momentum of reforms must also be maintained. The Chinese authorities seem to agree with that view and indications are that they are now preparing to initiate a new wave of reforms, including price and fiscal reforms. A well functioning fiscal system is essential if markets are to replace controls. Evolution of the Budgetary Process since 1980 3. It is in the context of the overall reforms that this report reviews the issue of intergovernmental fiscal relations in China. Decentralization of fiscal authority to localities and enterprises was an important part of the reform process. Before 1980 the Chinese fiscal system was characterized by "centralized revenue collection and centralized fiscal transfers"; all taxes and profits were remitted to the Central Government and then transferred back to the provinces according to expenditure needs approved by the center. In 1980 a new system of fiscal contract was introduced, which in China was called "eating out of separate kitchens." The basic framework of revenue sharing was to designate three types of taxes: central fixed revenue (those that accrue to the center), local fixed revenue (those that accrue to the localities), and shared revenues (those that are shared between the center and localities according to some broad rules). The task of defining the sharing formula, however, has proven difficult, and there have been as many as three systems tried since 1980. For the 1980-84 period, revenues from the shared taxes were allocated to provide 80 percent to the center and the remaining 20 percent to local governments. In 1985 the sharing system was redesigned to address the problem of surplus and deficit provinces. The provinces whose local fixed revenue exceeded their expenditure needs were required to transfer excess revenue to the Central Government, and the provinces whose local fixed revenue could not cover their expenditure were permitted to retain a certain proportion of shared revenue. The system was changed again in 1988. On the basis of negotiations conducted with various provinces, six different systems of revenue sharing were agreed upon. Although the exact formulas differed, the basic principle was to agree on a contract between center and provinces for sharing the tax revenues, with a view to giving maximum incentives to provinces for revenue collection by allowing them to retain a relatively high share of marginal revenues beyond the levels agreed. However, these formulas have not been strictly adhered to, and many ad hoc adjustments have been made over the years. 4. The system of expenditure responsibilities between the center and provinces is not based on any clear-cut principles of allocation and there is a considerable degree of overlap in their expenditure responsibilities. The annual provincial budget is formulated on the basis of economic plans prepared by the State Planning Commission (SPC) and the guidelines provided by the Central Ministry of Finance. These guidelines indicate the acceptable level of expenditure for a province and take into account historical trends, results of the current year's budget so far, and policy requirements for the next year. The Provincial Bureau of Finance reviews these guidelines, adjusts the amounts according to local conditions, and may augment expenditures if more resources are likely to be available to the province. The two essential requirements are that the local budgets must be consistent with the national plan and should not have a deficit. Within these parameters, the Provincial Bureau of Finance conducts its review and consolidation operations and prepares a budget that is reviewed by a Standing Committee of the Provincial People's Congress and later by the Congress itself. 5. The framework of controls in intergovernmental fiscal relations in China is typical for a centrally planned economy; the policy and regulatory controls are more important than the process controls that are generally associated with financial management. The control framework is also unique in that it reflects the unitary structure of the Government as well as the guiding philosophy of unified leadership and decentralized management. For the important component of expenditure relating to investment, the final decisions on investment by the enterprises, sectoral ministries, and localities are made by the State Planning Council on the basis of proposals made by the decentralized units, which in turn are based on the guidelines provided by the Central Government agencies. 6. In principle the Central Government can exercise tight control over the finances of the provinces, but in practice there are several factors that reduce the effectiveness of controls. First, there seems to be a considerable degree of overlap with regard to the responsibilities of different levels of government. Second, although the framework of control specifies that provincial budgets are to be formulated in accordance with central guidance, it appears that in - xi - practice there are several "twilight zones" that encourage bargaining. Finally, the system is characterized by the importance of contracts at all levels rather than by the governance by rules. 7. This report does not review the issue of reforming the system of expenditure responsibilities or the system of budgetary controls. Instead, it concentrates on a review of the system of revenue sharing between the center and provinces. The report implicitly argues that the proposed system is in principle consistent with a widely varying degree of final expenditure share between center and provinces. Consequences of the Present System of Revenue Sharing 8. The Chinese economy still consists largely of state-owned enterprises, which account for about 75 percent of the state revenues. In many key issues such as determination of investment programs, wage policies, and subsidy policies, the central authorities play a key role. In this situation, the issue of resource sharing is mainly one of administrative decentralization, and some form of contracting between the state and enterprises and between the center and provincial governments is unavoidable. Because of this, the essential delinking of functions -- between the state as tax collector and the state as owner of assets -- has not taken place. The forms of contract that have been adopted in China in recent years aim at giving maximum incentives to the enterprises and provincial governments by minimizing the share of revenue collected over and above the contracted tax amount. Accordingly, the contracts are often fixed in nominal amounts; even when they allow for growth rates in tax revenues, these rates are moderate and generally fixed in nominal terms. Analysis has shown that, in principle, this system has some inherent adverse tendencies in terms of revenue collection and resource allocation. Among these are the following: (a) Direct tax revenue increases tend to be less than proportionate with respect to GDP and to be procyclical rather than acting as an automatic stabilizer as they do in most market economies with progressive tax structures. (b) The revenue-sharing formula creates an inherent tendency for the revenue going to the center to decline relative to local revenue and to GDP. (c) The fiscal system is likely to lead to increasing central fiscal deficits. (d) The decline in the center's share of revenue runs the risk that investment in infrastructure and other key projects of national significance will be neglected. (e) The fiscal system has allowed a rapid growth of extra-budgetary funds that escape the discipline of review and approval by national authorities and may thus lead to inefficient resource allocation. (f) The revenue-sharing system encourages local protectionism and thwarts national efforts toward industrial restructuring and the creation of an integrated market. - Xii - (g) The revenue-sharing system conflicts with equity objectives, favoring better-off provinces over poorer ones. 9. Although the fiscal system suffers from the inherent tendencies noted above, the bottomline results in China in recent years seem to depart significantly from those that could be expected on the basis of ex ante analysis. For example, over the past three years the revenue- sharing system has been in operation, the tax-GDP ratio seems to have stabilized around 18 percent despite some deceleration of GDP growth rate. The center's share of total revenue also seems to have increased marginally rather than to have declined during this period. The budget deficit seem to be less than 2 percent of GDP even after incorporating large subsidies on consumer goods and to loss-making enterprises. The problem seems to be one of of expenditure rationalization rather than tax mobilization, since the tax-GDP ratio is in line with other countries at a comparable stage of development. The rate of investment including investment in infrastructure, seems to have increased during the past three years. The growth of extra- budgetary funds has indeed been rapid, but it is difficult to demonstrate that this is not a desirable form of decentralization. Similarly, local protectionism, although a serious problem, can be only partially corrected through the fiscal system and is more closely connected with the overall system of interregional trade rules and enterprise reforms. Finally, the figures on per capita Government expenditure among the provinces show no correlation between per capita expenditure and per capita revenue generation. 10. These results may be partly due to inadequacies of theoretical analysis to capture all the complexities of the fiscal system. However, it seems that the main reason why the inherent weaknesses of the fiscal system are not revealed in the bottomline figures is that the rules are not implemented rigidly. Various ad hoc adjustments are made whenever considered appropriate by the central and provincial authorities. It appears that despite decentralization, the central authorities -- the governmental and party apparatus -- do have substantial control over the credit and fiscal system of the country, particularly when there is consensus at the top of the hierarchy. However, although the product is not unimpressive, there are serious problems with the process. The fiscal system that is developing has been somewhat unstable and is widely regarded as having elements of inefficiency and inequity. This process-related problems may, over the long term, hurt the performance of the economy, and the Chinese authorities are rightly concerned about reforming the fiscal system. 11. The Government has started an experiment in tax assignment in two provinces and in six cities, and it is interested in designing an appropriate nationwide system of revenue sharing between different levels of government to be implemented during the Ninth Five-Year Plan (1996-2000). Analytical Issues. Case Studies, and Suggestions for Reform 12. The report reviews the analytical literature as well as the experience of several countries in the matter of intergovernmental fiscal relations and makes some suggestions for the Chinese authorities to consider. These suggestions relate to long-term perspectives as well as to some transitional arrangements. The report concludes that there are some serious flaws in the design of the experiments launched by the Chinese authorities for reforming - xiii - intergovernmental fiscal relations, and it proposes a long-term system of revenue sharing between the center and provinces, as well as an alternative approach to an experimental phase. 13. While recognizing the specificities of Chinese conditions, the Chinese authorities cannot but follow some basic principles of multilevel fiscal systems. Three important postulates need to be emphasized. First, international experience in countries big and small, federal and unitary, indicates that the Central Government must have effective control over the. most important sources of tax revenue, not only in terms of tax law and policy, but also in terms of administration, collection, and allocation of revenue. Second, taxes should not be looked upon mainly as providers of revenue; they must also be seen as tools of policy that, whether intended or not, affect the allocation of resources and their efficient utilization, interregional and interpersonal distribution, and the level of aggregate demand. Third, "marketization" of enterprises cannot be achieved by administrative decentralization alone; for efficient resource allocation, hard budget constraints on enterprises are essential, and the enterprises must be brought under a rule-based tax system. 14. A basic problem in multilevel fiscal systems arises from the fact that, for the sake of efficiency and equity, most of the major taxes have to be collected centrally; on the expenditure side, in contrast, closeness to the beneficiary helps efficiency and may require a considerable degree of decentralization. There is thus a need for designing a system for the transfer of resources from the center to the localities, but it needs to be done in a way that is transparent and equitable and gives incentives to the localities to maximize their efforts for revenue mobilization. The system should also minimize overlapping in expenditure responsibilities. 15. The major considerations to be kept in view in structuring intergovernmental fiscal relations are as follows. First, the revenue flows to the different levels of government must be adequate to enable them to fulfill their expenditure responsibilities, given the desired size of the total public sector. Second, the Central Government's capacity to pursue effectively policies of macroeconomic management must be ensured. Third, the system must be conducive to the efficient use of resources in the economy and must promote fiscal discipline. Fourth, while allowing for local autonomy and promoting initiative at the local level, it must be possible for the center to effect transfers to the poorer regions so as to reduce interregional disparities; i.e. the center must have some surplus of revenues over its own expenditure needs. Fifth, in the division of tax resources among the provincial governments, weights must be assigned to both autonomy and incentive at the provincial level as well as to interregional equity. There are four alternative ways in which tax revenues can be apportioned among different levels of government: (a) The center levies and collects all the revenues and gives a portion of the revenues to provincial governments as grants. (b) The center levies and collects all taxes but shares the proceeds of some or all taxes with the provinces on the basis of a formula or a set of formulas; - xiv - (c) The center levies and collects most of the more important taxes, but the provinces have autonomy to levy and collect other taxes; in addition, the provinces get a share of one or more central taxes and/or receive grants from the center. (d) The center and the provinces enjoy more or less concurrent tax powers, so that the provinces can raise on their own most of the revenues they need; they may, however, also receive some grants from the center. In practice the models may be modified (e.g., instead of sharing the proceeds of a tax, the center might share the base of a given tax with the provinces); again, there could be a mixing of the elements of the different models (generally, the provinces or local governments are allowed to levy and collect some taxes). In choosing among the models or in selecting elements from the models to build a mixed system, the major considerations set forth earlier must be applied. However, since center-provincial relations will ultimately have to be based on a political consensus, in practice the various criteria such as equity, efficiency, and scope for autonomous action of the provincial governments will have to be balanced against one another. Also, since taxes are involved, administrative considerations and the costs of compliance have to be given a high weight. If the autonomy of the provinces or states is given very high weight, the fourth model -- of concurrent tax powers -- will be chosen. It is the system that might be favored in a federation with a pronounced accent on the decentralization of power. Apart from its nonapplicability to a unitary country such as China, experience has shown that it is also a messy system, with overlapping tax powers, high cost of compliance, and potential conflict among the provinces regarding the right to tax on the basis of residence versus origin. Some of the important taxes are not capable of being efficiently levied and collected by subnational governments. 16. The basic principles of assignment of tax powers suggest the following: (a) Taxes with procyclical trends should be assigned to the Central Government (b) Taxes that are to be levied on a progressive basis, such as income taxes, should be levied and collected by the center. (c) Taxes on production, such as excises on manufactures and the value-added tax (VAT), should be assigned to the center. (d) Taxes with bases that are unequally distributed among the provinces should be assigned to the center (for example, taxes on natural resources). (e) Taxes on international trade should be assigned to the center. 17. Once these principles are accepted, it is clear that the center would be raising the major part of the taxes, leaving small taxes (such as taxes on immobile localized bases, taxes on consumption, and user charges) to the localities. The inherent strength of these principles is indicated by the review, made in the Appendix to the report, of the experiences of Germany, Canada, India, and Japan. Despite historical differences, differences in size, and governmental - xv - diversities in these countries -- and despite different starting points -- most of them have ended up with the center collecting all the major taxes including income tax and VAT. It is clear that a system of concurrent taxes should be avoided. Among modem federations, only Switzerland and the United States (and to some extent, Canada) continue to operate systems based on concurrent tax powers. In the United States, however, over time the most important tax, has become predominantly a federal tax: the yield of the federal tax on individual incomes and corporate profits far outweighs the yield of the state income taxes. Effectively, a separation in the use of tax powers has come into existence, with the U.S. Government relying predominantly on income taxation and the states on sales taxes, although concurrent tax powers are being used, resulting in complications and inefficiency. 18. The total revenue could, of course, be shared between the center and provinces at different rates. The share of the tax pool going to the center can be different for different taxes, and the principles of distribution of the total pool among provinces can give different weights to collection, population, and other needs. The main report and the Appendix review in some detail how these principles have been effectively implemented in different ways in the four countries considered. The analysis also suggests a formula for illustrative purposes based on central authorities' judgment regarding an appropriate sharing of revenue between center and provinces. The formula for tax sharing would need to be supplemented by some formula for grants, of which there are many working examples in different countries. 19. The reforms suggested cannot of course be introduced in isolation or overnight. They have to go hand in hand with reforms in pricing and enterprises. The report also accepts the advisability -- in fact, the superiority -- of the gradual and experimental approach to reform that has been adopted by the Chinese authorities. However, the experiments in tax reform started in certain provinces seem to have some intrinsic problems. Tax-sharing formulas are by their nature national in character, and effective experimentation in this area is not possible for selected provinces. An alternative is to experiment with certain types of taxes at the national level, rather than with all taxes at the regional level. In this context, it is also important to emphasize that the transitional arrangements should not conflict with the long-term objective. As such, it is probably a mistake to assign the income tax to provinces, even in an experimental stage, as seems to have been done by the Chinese authorities. Rather, the income tax could be selected for experimental collection by the center, with a certain percentage distributed to provinces in accordance with an agreed formula. 20. Reform in center-local fiscal relations should become a part of the new wave of reforms that seems to be beginning in China. These fiscal reforms can reinforce, and also be strengthened by, other reforms. 21. Introduction of any such reforms can only follow a process of consensus building undertaken with the provincial authorities. There is a need for agreement on principles as well as on the parameters for resource distribution. The provincial authorities must be convinced that their expenditure capacities will not be unduly adversely affected by surrendering some of their authority to collect revenue. There has to be an agreement on the basic expenditure responsibilities of the local governments before a revenue sharing formula can be implemented. It may also be possible to apply revenue sharing formula initially to incremental revenue only - xvi - leaving the distribution pattern of base year revenue unchanged. Detailed statistics on tax collection and expenditures at different levels are needed so that simulation exercises can be made showing the implications of various sharing formulas for central and local finance. A seminar involving Chinese authorities (both central and provincial) and international experts on the subject would be helpful for building consensus about the principles of resource sharing. This could be followed up by the simulation exercises illustrating the implications of various revenue sharing formulas, which also would help in consensus building and eventual agreement at the political level. The process could be completed within the Eight Five-Year Plan (1991- 96), as envisaged by the Chinese authorities. I. THE SETTIN 1.1 The search for the correct balance between centralization and decentralization is one of those endeavors in human societies where there is no once-and-for-all solution. The search is ongoing: periods of tilt in one direction need to be compensated for; a tilt in the other direction again creates the need for correction. Chinese society has been going through such phases for centuries. Similar cycles have been repeated in recent years since 1949, and they have been analyzed in several documents on recent Chinese history. 1.2 Since 1979 a sustained effort has been under way to introduce a system combining market and plan within the context of dominant public ownership. There are now questions in many quarters of China about whether the process of decentralization has gone too far in the area of fiscal and financial management. It is evident that the issue of central-local resource sharing is an intensely political one, and over the years the Chinese authorities would have to bring about a greater degree of transparency in the whole range of center-province relations. However, this report focuses on the narrower issue of the fiscal responsibilities of the center and in the provinces, while acknowledging that center-province relations involve broader issues that cannot be solved through fiscal policy alone. It examines critically the prevailing notions on center-province fiscal relations and makes suggestions for the future. 1.3 A review of the system of fiscal reforms must be done in the context of overall reform programs and their macroeconomic consequences. The reform program was initiated in late 1978. Although there have been many changes in the content of the reform programs, the basic objective is to find a mixture of plan and market that could deliver growth with stability and equity. Among the important characteristics of the reform are: allowing effective individual control (and ownership) of small units in rural and urban areas; greater outward orientation in trade and investment; moving toward market determination of prices, except for some key prices; and using decentralization to sectoral agencies and localities as a proxy for privatization for introducing incentives and marketization. The reform process has been endogenous and guided by the principles of gradualism and experimentation. 1.4 This process of reform has, on the whole, been successful. The annual rate of growth of GDP over the past 12 years of reform has been about 8.8 percent, which is significantly higher than the rate China experienced before reform and higher than that of other countries either in the process of transition from socialism to markets or in mixed economies under adjustment programs. Naturally, there have been ups and downs in the growth rate, but even the most severe recent downturn in 1989 was associated with a growth rate of 4 percent. During this period, the investment rate has been maintained on average at over 35 percent of GDP; the investment rate in infrastructure and key industries has also been maintained at a high rate (Table 1.1). 1.5 The management of macroeconomic balance has, however, proven more difficult. Since 1978 China has experienced three episodes of macroeconomic instability. Following each period of accelerated reform (as in 1979, 1984 and 1986), there were increased budget deficits, rapidly accelerated growth rates of money and credit, and the emergence of external trade -2- deficits. The process of correcting price distortions can be expected to be associated with some inflation, and the inflation rates in the first two episodes of instability were in the single digits. The budget and balance of payments deficits were also at moderate and manageable levels. The rates of inflation in 1979 and 1984 were high by past Chinese standards but not unusual in the best of cases of adjustment and relative price reform. The episode of inflation in 1987-88 was more serious, with inflation rates touching 20 percent, a rapid increase in imports, and deterioration in the foreign exchange position. In 1989 and 1990, however, the Central Government managed to regain control of the macroeconomic situation. Inflation in 1990 was moderate, at the 3 percent level, and the balance of payments showed a surplus, leading to a comfortable reserve position. The rate of growth of overall credit and money supply has been rapid, but the inflation rate is still moderate (Table 1.1). 1.6 The record of the reform era is also creditable with respect to regional income distribution. For a country of China's size and diversity, it is natural to expect some significant differences in income levels among provinces. During the Maoist era, despite socialist rhetoric on equality, income differences among provinces appear to have increased. During the reform era, the growth rates of special economic zones and coastal provinces seem to have accelerated much faster than those of other provinces. There are many contrasting examples such as the spectacular growth of Guangdong, a coastal province, and the relatively slow growth of neighboring Hunan province. There are many pockets of poverty, and in some cases the relative position of the poor and of some poor provinces is getting worse. Yet it is impressive that, for the country as a whole and by overall measures of inequality, provincial income differences seem to have narrowed rather than increased during the reform period. However, the outlook for the 1990s is one of increasing inequalities of income amongst provinces. 1.7 It is in this context of overall reform that the reforms in the fiscal system have to be seen. The record of performance lends credibility to the Chinese process of reform, which emphasizes experimentation and gradualism. However, if the past record of good performance is to be maintained, the momentum of reforms must be maintained. The Chinese authorities fully recognize this point and in fact seem to be in the process of initiating a new wave of reforms. Recently, there have been adjustments in the exchange rate and in the prices of some key commodities, as well as social security and housing reforms -- all geared to facilitate effective enterprise reform. A well functioning fiscal system is necessary if markets are to replace controls and the Government has already initiated reforms in this area. Table I.1: China: Macroeconomic Indicators Item 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1. Growth rates (percent per year) Real GDP 7.6 7.7 4.5 8.7 10.1 14.5 13.0 8.4 11.2 11.2 4.6 5.6 7.0 Money and quasi-money (M2) 9.7 24.1 19.7 13.1 19.2 42.4 17.1 29.3 24.2 20.9 18.4 28.0 26.4 Of which: Currency 26.3 29.1 14.5 10.9 20.7 49.4 24.7 23.3 19.4 46.7 9.8 12.8 20.4 Net domestic assets 31.1 22.1 33.9 21.5 20.7 18.3 24.1 25.0 Retail price index 2.0 6.0 2.4 1.9 1.5 2.8 8.8 6.0 7.3 18.5 17.8 2.1 2.9 II. Ratios to Current GDP (percentages) INVESTMENT-GDP RATIOS (current prices) GDUGDP 34.9 32.2 29.2 29.7 30.4 32.3 40.5 40.8 39.3 39.5 38.4 36.6 35.7 Fixed Investment/GDP 25.3 24.0 20.1 23.7 24.7 26.4 29.7 31.1 32.1 32.0 25.9 25.1 26.7 Current account balance to GDP 0.0 0.3 1.1 2.4 1.6 1.0 -3.9 -2.8 0.0 -1.0 -1.1 3.2 3.3 (after off. grants) FISCAL RATIOS Budget deficit -5.2 -3.3 -1.2 -1.4 -1.7 -1.5 -0.5 -1.9 -2.2 -2.4 -2.4 -2.1 -2.5 Tax revenue 30.8 28.7 28.4 26.4 26.5 25.3 25.6 23.2 20.5 18.4 18.5 17.2 Current expenditure 21.6 22.2 22.4 21.8 21.8 19.8 19.6 19.4 18.3 16.9 18.1 17.4 16.9 Capital expenditure 15.2 10.6 7.8 6.8 7.3 8.0 7.5 7.8 6.7 5.4 4.7 4.6 4.2 Source: China: Statistical Yearbooks -4- II. EVOLUTION OF THE BUDGETARY PROCESS SINCE 1980 2.1 As a part of overall economic reform efforts, the fiscal system has been going through a process of decentralization. Before 1980, the Chinese fiscal system was characterized by "centralized revenue collection and centralized fiscal transfers": all taxes and profits were remitted to the Central Government and then transferred back to the provinces according to expenditure needs approved by the center. In 1980 a new system of fiscal contract was introduced, which in China was called "eating out of separate kitchens." The exact arrangements for expenditure and revenue allocations between the center and provinces have been experimental. The framework for revenue sharing has been particularly unstable, with as many as three systems tried since 1980. Revenue-Sharing Framework 2.2 The basic framework of revenue sharing is provided by a system that designates three types of taxes: central fixed revenue (those that accrue to the center), local fixed revenue (those that accrue to the localities), and shared revenues (those that are shared between the center and localities according to some broad rules). The taxes in each category were changed in 1980, 1985, and 1988; Tables SA. 1-SA.3 show the situation for these three years (Table SA.3 updates 1988-90 data for 1991). The task of defining the sharing formula has proven difficult. For the 1980-84 period, revenues from the shared taxes were allocated to provide 80 percent to the center and the remaining 20 percent to local governments. This created a lack of incentive for local governments to collect the shared revenues effectively and undesirable surpluses and deficits for particular provinces as a result of the uniform sharing formula. In 1985 the sharing system was redesigned to address the problem of surplus and deficit provinces. The provinces whose local fixed revenue exceeded their expenditure needs (surplus provinces) were required to transfer excess revenue to the Central Government. Those provinces whose local fixed revenue could not cover their expenditures were permitted to retain a certain proportion of shared revenue. If the local fixed revenue and the shared revenue could not meet their expenditure needs, the deficit provinces would receive a fixed transfer from the Central Government. This system was good for meeting the expenditure needs of different provinces, but it still did not create incentives for provinces to mount resource mobilization efforts. 2.3 The system introduced in 1988, for a three-year period, was intended to increase the incentive for revenue collection by the provinces. Negotiations were conducted with various provinces to design a system best-suited to their circumstances. As a result, six systems emerged (Types A-F in Table SA.3). For 16 deficit provinces, transfers were fixed in nominal amounts, so that any additional efforts for revenue mobilization and expenditure savings would accrue to them (Type F). For three provinces -- Shanghai, Shandong, and Heilongjiang (whose transfers to the center were regarded as a relatively high proportion of their total revenue) -- transfers to the center were fixed in nominal terms, so as to give them the full benefit of additional resource mobilization for shared taxes (Type E). For two provinces, Guangdong and Hunan, where the center's share in the revenue was small, a high rate of growth of revenue to the center was provided but was also fixed in nominal terms, thus leaving to the provinces all -5- the revenue above the specified growth rates (Type D). For six provinces, the earlier system of sharing at certain percentages, either uniform or variable, was retained (Types B and C) though for two of them the marginal rate of remittance was higher than the average. For ten provinces, the formula was a mixture of the above two systems: a fixed rate of sharing up to a certain rate of growth of revenue, and 100 percent retention by the province of any revenue above the specified growth rate (Type A). 2.4 The basic objective in designing the revenue-sharing formulas was to create incentive for the provinces to maximize resource mobilization. However, these formulas have not been strictly adhered to. In 1990, for example, Hunan Province gave the center much less than the stipulated revenue on the ground that centrally mandated agricultural price, subsidy price, and wage policies have created expenditure pressures on provincial budgets that did not enable the provinces to fulfill their contracts. In contrast, Guangdong Province, which was doing well, was "forced" by the center to give transfers larger than stipulated, in part by transfer of ownership of some "state-owned" enterprises from province to the center. Moreover, the center continues to exert a crucial role in resource allocation through its control of the mechanism for expenditure allocation. 2.5 The Central Government has a very small tax collection machinery of its own; taxes are mostly collected by local authorities. The amount of tax passed on to the center consists of two parts: first the part accruing to the center directly and the second, accruing to the center as part of the revenue sharing agreement. The exact figures on these taxes by type and accrual have not yet been rovided. The available figures presented in Table 11.1 suggest that in 1990 localities transferred about 50 billion yuan to the center (about half of the tax revenue accruing directly to the center) which was more than offset by the 60 billion yuan transferred from the center to the localities. -6- Table 11.1: China: Center-Local Shares in Revenue, 1990 (Yuan billion) Central Local Revenue Total Amount % Share Amount t Share Domeet.i2 307.2 117.2 38.2 19. 01 Tax 299.9 86.2 28.7 213.7 71.3 Bonds 17.0 17.0 100.0 0.0 0.0 Construction/Energy Tax 21.5 17.3 80.5 4.2 19.5 Other 17.2 12.4 72.3 4.7 27.3 Ent. Revenue 4.9 3.2 65.2 1.7 34.8 Ent. Losses (65.8) (27.1) 41.2 (38.7) 58.8 Regulatory Fund 12.5 8.2 65.2 4.4 35.2 Foreian Exchanae Loans 16.4 100.0 0 TOTAL 323.7 13. 13 190.0 5. Memo item: Transfers from Localities to Center 48.2 from Center to Localities 59.6 Source: Ministry of Finance, Government of China Framework for Expenditure Control 2.6 The state budget is divided into two parts, central and local. The central budget includes the unit budgets of all central departments and the financial budgets of key enterprises under the direct control of the Central Government. The local budget includes the provincial, county, and municipal budgets. The respective expenditure responsibilities, insofar as they are limited to the budget, are the following. Central expenditure responsibilities include capital construction and funds for upgrading central enterprises and outlays on their new product development; national defense, foreign aid and external relations; agriculture, forestry, and water conservation at the central level; industrial, transportation, and commercial operations; education, culture, health and social services; outlays on centrally obligated price subsidies; geological surveys; and financing of public debt. Local expenditure responsibilities include basic construction and technical improvement of locally owned enterprises and their new product development; rural production assistance and agricultural development and water conservation at the local level; urban maintenance and construction; education, health, culture, and social services; social welfare and pensions; administrative expenditures; and a range of price subsidies. 2.7 As noted above, under the system that came into effect in 1980, and later in terms of the contract budget system in operation during 1988-90, 16 provinces that cannot balance their -7- revenue and expenditure receive quota subsidies from the Central Government. In addition, local government levels receive special subsidies to cover expenditures for natural disaster relief and outlays for the development of backward areas. Further, local governments also receive settlement subsidies at the end of the year to compensate for slippages when the actual budgetary outcome differs from estimates, and when such slippages are attributable to changes in ownership of local enterprises between center and provinces. The Central Government is empowered to borrow both internally and externally, whereas local budgets are expected to be in balance, local governments and cannot borrow locally to finance any part of their budgetary operations. Budget Formulation and Implementation in the Provinces 2.8 The annual provincial budget is formulated on the basis of the economic plans prepared by the State Planning Commission (SPC) and the guidelines provided by the Central Ministry of Finance. These guidelines indicate the acceptable level of expenditure for the province and take into account historical trends, results of the current year's budget so far, and policy requirements for the next year. The Provincial Bureau of Finance reviews these guidelines, adjusts the amounts according to local conditions, and may augment expenditures if more resources are likely to be available to the province. The two essential requirements are that the local budgets must be consistent with the national plan and should not have a deficit. Within these parameters, the Provincial Bureau of Finance conducts its review and consolidation operations and prepares a budget that is reviewed by a Standing Committee of the Provincial People's Congress and later by the Congress itself. 2.9 During budget implementation, the Bureau of Finance in Central Government is responsible for the release and management of funds. Funds for outlays other than basic construction and technical improvement of enterprises (including new product development) -- categories equivalent, in a broad sense, to capital outlays -- are released on a monthly basis. For outlays on basic construction and technical improvement and similar categories, funds are released in relation to the actual progress of work in specified projects and with reference to the availability of resources. To the extent that there are revenue shortfalls, the burden of adjustment falls on outlays on these two categories. The Bureau of Finance is responsible for monitoring budget transactions and for review of the monthly, quarterly, and annual accounts before they are sent to the Central Ministry of Finance. The Provincial Finance Bureaus also hold regular meetings with other officials at the county and township levels to review progress in the implementation of the budget. 2.10 The framework of controls in intergovernmental fiscal relations in China is typical for a centrally planned economy; policy and regulatory controls are more important than the process controls that are generally associated with financial management. The control framework is also unique in that it reflects the unitary structure of the Government as well as its guiding philosophy of unified leadership and decentralized management. By the same token, the control framework is also different from what is found in several federal-type systems in which central governments tend to exercise controls that are directly related, and frequently proportional, to the amounts of money transferred from central to provincial or state (as they are called in some cases) governments. -8- 2.11 An important segment of expenditures relates to investment. Investment planning in China has four components: (a) the "capital construction plan" in the state sector; (b) the "technology promotion plan" in the state sector; (c) fixed investment in the whole economy (including collective and individual sectors); and (d) commercial housing investment. The basic principles guiding investment decision are two: (i) achieving balances -- between production and consumption, between agriculture and industry, between heavy and light industry, and between the coastal and inland provinces, and (ii) key national projects whose justification lies in their impact on other sectors and on the economy as a whole. 2.12 The procedure for investment planning can be summarized as "two ups and two downs." The steps are as follows: (a) Guided by the state's general investment principles, the production ministries and localities initiate and submit to the SPC (State Planning Commission) investment projects in capital construction, technology innovation in the state sector, and investment in other sectors. (b) The SPC reviews existing and new projects and decides on the level of investment and sources of financing in consultation with the Ministry of Finance, the Central Bank, and the sectoral ministries concerned with raw material production. With initial approval from the State Council, the SPC delivers to the sectoral ministries and localities the key investment control figures in investment levels in different ministries and regions, investment in large and medium projects, and investment levels in construction, technology innovation and other categories of investment. (c) On the basis of the investment control figures, the sectoral ministries and localities propose "formal investment plans" and submit them to the SPC. (d) Finally, the SPC makes an overall balance in the national investment plan on the basis of local investment plans and submits it to the People's Congress for approval. 2.13 The SPC thus exerts strong control over project approval. According to the current regulations, all "production" investment projects above a certain limit, no matter what their sources of finance, should be approved by the SPC and/or the State Council. The SPC approves investment projects above 130 million (for energy, transportation, and raw materials, the limit is 150 million). For projects above X200 million, the State Council's approval is required. Projects below T1 million can be approved by the administrative unit designated by the provincial governments; projects between 1 million and Y10 million are approved by the provincial governments; those between 110 million and Y30 million can be approved by the provincial government, but all files relating to the project should be submitted to the SPC after provincial approval. 2.14 The process of investment allocation and approval is an important part of intergovernmental resource sharing. First, because a substantial part of investment outlays is, as noted previously, in the central sector, the location of projects, given their possible benefits to the local economy, becomes a matter of importance. Provinces tend to vie with each other -9- to garner greater share of these outlays. Also, given the magnitude of the outlays, they become an important instrument for redressing interprovincial economic disparities. Second, the terms and conditions of the loans re-lent by the Central Government acquire some importance in relation to the financial capacity of the province. These loans are, however, of recent origin, and their full impact on provincial finances may not be felt for some time to come. Issues of control 2.15 In principle, there are a number of controls exercised by the Central Government over the finances of the provinces, and therefore the center has the ability to pursue and implement appropriate macroeconomic policies. In practice, however, there are several factors that would appear to have reduced the effectiveness of controls. 2.16 First, although the enumeration of tasks and responsibilities of the provincial governments may give the impression of a neat and precise division of labor, in practice there seems to be a considerable degree of overlap. Each level appears to be engaged in all types of economic activity. A more organized fiscal relationship between the center and the provinces should show the areas that are the exclusive reserve of the two levels, the areas where both levels of governments have a role, and the areas where provinces would function as implementation agents of the Central Government. Pending such a specification, experience illustrates that there have been different perceptions (hence different degrees) of involvement in various activities, with consequential impact on resources and expenditures. These different perceptions may also have contributed to the pre-eminence of bargaining in intergovernmental fiscal management. 2.17 Second, although the framework of controls specifies that provincial budgets are to be formulated in accordance with central guidance, in practice it appears that there are several "twilight zones" that encourage bargaining. The guidance is general and is based on the accounts of a previous base year; as such it may not fully reflect the needs of each province. The absence of data on the level of services to be provided by provinces, and on regional differences in costs, exacerbates these issues. In addition, previous experience also indicates that the avoidance of deficits is not enforceable. In fact, many provinces may have been incurring deficits in the form of accumulated arrears. In regard to investment projects, the provinces have a good deal of freedom to shift outlays from one project to another. Hence, although the Central Government nominally has powers (under the aegis of the SPC) to impose embargoes or physical limitations on projects, these powers represent an ultimate step that, while reflecting the failure of policy, would not appear to have any restraining effect on the provinces. All this may have been rendered possible because of the very limited role that the Central Government plays in the budget implementation phase. Its role is confined, by and large, to the consolidation of accounts submitted by the Provincial Finance Bureaus for onward submission to the People's Congress. 2.18 Third, the process permits the dominance of contracts, rather than the governance by rules appropriate for a market environment. Admittedly, not every procedural aspect can be specified by law, but there is clear need for a comprehensive framework of rules. In the - 10 - existing system, there are at least four points where bargaining is specifically indicated: (a) the determination of additional contributions to the center, (b) transfers to the provinces to meet additional expenditure requirements stemming from natural calamities and related factors, (c) credit from the banidng sector, and (d) the determination of quotas for the subscription of state bonds. These aspects have crucial impact on intergovernmental fiscal relations in that they determine the budgetary outcome of the national budget. - 11 - 1I. CONSEQUENCES OF THE PRESENT INTERGOVERNMENTAL FISCAL SYSTEM 3.1 In analyzing the nature and consequences of the tax and revenue-sharing system, it is important to view the system in the context of the special characteristics of the Chinese economy. The Chinese economy still consists largely of state-owned enterprises, which account for about 75 percent of state revenues. As noted in Chapter II, the investment programs of state-owned enterprises are largely a part of a centrally approved program. Many of the major items of expenditure of the state-owned enterprises, such as wages, are also largely determined by central policies. In this situation, an essential delinking of functions -- between the state as tax collector and the state as an owner -- has not taken place. Once the investment program for an enterprise is approved, it is not critical for the managers whether the funds are from profits or from some other state resources. In this case, some form of contracting between the state and enterprises is unavoidable, and the usual role of taxation as an allocator of resources is changed. The issue then becomes one of administrative decentralization. The purpose of introducing a "tax system" at this stage may thus be seen as preparation for the future, when enterprise reform in the true sense might take place. For the time being, the tax system has its usual meaning only for private and collective enterprises and for foreign enterprises, and for three categories the system is applied in the usual fashion. This chapter discusses some of the adverse inherent tendencies of the present system of taxation and revenue sharing as it applies to the state-owned enterprises. It goes on to note how many of these inherent tendencies are muffled in practice because of ad hoc departures from the system, and what the costs of such arbitrariness are. Adverse Tendencies: Ex Ante 3.2 Analysts in and outside China have noted that, to the extent that the tax and revenue- sharing system is implemented, it would have several adverse consequences for macroeconomic stability, efficiency, and equity. Some of these arguments elaborated in various documents, including those by the Bank and the IMF, are noted below. (a) The fiscal system has an inherent tendency for the tax generation to be inelastic with respect to GDP and to be procyclical. 3.3 As in the case of revenue sharing between the center and provinces, tax collection between the state and enterprises is largely governed by contracts, not only for income taxes but also often for indirect taxes. These contracts are sometimes fixed in nominal amounts; sometimes they allow for certain moderate growth rates of tax revenues, again in nominal terms. The objective of this approach was to create an incentive for enterprises to increase their profits by making marginal tax rates low or zero. In an economy that is growing rapidly, this will mean that tax revenue is not growing as fast as income; i.e. an inelastic tax system. Moreover, the system will introduce a procyclical element in tax generation: when income growth slows down or becomes negative, the tax-GDP ratio will increase; when it accelerates, the tax-GDP ratio will decline. This is exactly the opposite of what happens in normal market economies, where the tax system generally works as an automatic stabilizer. In these economies, because - 12 - of the progressivity of the tax structure, when income growth accelerates, the Government gets a larger share of income in the form of tax revenue, which reduces the expenditure capacity of the public and thus contributes to slowing down aggregate expenditure and income in the economy. When an economic slowdown occurs, the reverse ensues. (b) The fiscal system is likely to lead to increasing fiscal deficits. 3.4 If the tendency noted in (a) above operates, the tax-GDP ratio will decline over time. However, government expenditure is likely to increase in line with GDP, particularly with the increasing fiscalization of social expenditures such as social security. There would therefore be a danger of increasing fiscal deficits over time. (c) Given the revenue-sharing formula, revenue to the center is likely to decline relative to local revenue and GDP. 3.5 As noted in Table SA.3 for some of the major contributors to central revenue, tax contribution to the center is fixed in nominal terms for three years. On the reasonable assumption of some growth in taxes accruing to the provinces, the center's share will go down in these cases. For other provinces, the marginal share of contribution to the center is lower than the average, and, for growth of revenue above a certain moderate rate, all revenues are retained by the provinces. This situation is also likely to lead to a fall in the center's share of revenue. An illustrative calculation -- made by the mission on the assumption of 10 percent growth in shared revenues -- shows an unambiguous decline in the center's share if the contract sharing formulas are adhered to (Tables SA.9). (d) The decline in the center's share of revenue is likely to lead to a decline in investment in infrastructure and in other key projects of national significance. 3.6 For many large projects in infrastructure and key industries, investments are likely to be undertaken only by the center, in part because the resources required for these projects may be beyond the capacity of individual provinces and partly because the benefits of many such projects will not accrue only to the individual provinces. Therefore, if the center's share of tax revenues goes down and the center has to maintain its essential current expenditure, it is these key investment projects that are likely to suffer. This will be particularly unfortunate because the task of building adequate infrastructure is already a serious constraint on the Chinese economy. (e) The tax system has allowed a rapid growth of extra-budgetary funds that escape the discipline of review and approval by the national authorities and that may thus lead to inefficient resource allocation. 3.7 The rapid growth of extra-budgetary revenues has been a particularly striking feature of the tax reform system during the 1980s. By 1990, the size of such revenues nearly equals that of state revenue. They accrue mostly to state-owned enterprises and their supervisory agencies. However, various other government departments -- at all levels -- also impose vanous "levies" without any legislative authorization and use various arbitrary techniques for the - 13 - realization of these levies. Such levies are totally contrary to all principles of sound fiscal management. Not only do they have unwarranted and unknown effects on enterprise behavior, they may also create temptations for unproductive use of resources by public authorities at various levels. (f) The revenue-sharing system encourages local protectionism and thwarts the national efforts toward industrial restructuring and creation of an integrated market. 3.8 In the current revenue-sharing system, the provinces keep a major part of additional revenues generated in their territories. They thus have an incentive to encourage location of industries in their area for the revenues that the industries will generate, irrespective of the viability or optimality of the location of these industries from a national point of view. The provinces are thus tempted to prevent movement of raw materials from their territories to other provinces and instead start factories for processing them. Ironically, this applies with particular force in industries such as tobacco or alcohol, for which tax rates on products are high, and works against the national policy of discouraging production of such goods. Market integration is a national policy objective for maximizing allocative efficiency. However, the revenue- sharing system works against that objective and contributes to a duplication of investments and fragmentation of the market. (g) The revenue sharing system conflicts with equity objectives, favoring better off provinces over poorer ones. 3.9 As noted in Table SA.3 the deficit provinces have their grants fixed in nominal terms for three years. This would imply that, over time, their command over these resources would decline in real and per capita terms. This situation is to be contrasted with that of richer provinces -- especially coastal provinces that, under the period of liberalization, have experienced rapid growth of incomes and revenues -- which will retain a higher percentage of their revenues as their income grows. Mixed Results: Ex Post 3.10 The inherent tendencies of the fiscal system noted above are indeed there. However, as noted in Table III. 1, the bottomline is a mixed one, with many of the results departing significantly from those that could be expected on the basis of an ex ante analysis of the system. 3.11 The tax-GDP ratio has declined over time, from over 30 percent in 1980 to about 18 percent in 1990. However, this decline has been associated with an increasing retention of profits by the enterprises and can be regarded as the intended decentralization. Over the past three years (1988-90), when the revenue-sharing system has been in operation, the tax-GDP ratio seems to have stabilized at about 18 percent, which is not totally out of line with international experience; nor (for reasons noted below) can it be regarded as demonstrably inadequate for meeting essential state expenditures. Equally important is that, even though these three years saw major ups and downs in both growth and inflation rates in the economy, the tax-GDP ratio was stable. Thus the inherent tendency of the tax system to be procyclical does not seem to be operating in practice; the tax-GDP ratio seems to have been cycle neutral during this period. - 14 - 3.12 Nor is it clear whether the tax revenues can be regarded as obviously inadequate or conducive to unsustainable budget deficits. As noted in Table 111.1, the budget deficits, following the definition in the IMF's Government Finance Statistics (GFS), are less than 2 percent of GDP. During 1988-90 they were declining, from 2.6 percent of GDP in 1988 to 1.5 percent of GDP in 1990. Moreover, these figures incorporate large subsidies on consumer goods and to loss-making enterprises. These subsidies amount to a large figure of 6 percent of GDP (in 1990). Rather than increasing tax mobilization to finance these subsidies, the Government should be aiming at (and has in fact been) reducing these expenditures. It should be emphasized that according to the available statistics, in China the Government runs a surplus on current budget account (see Table 1.1), and there is no unsustainable rise in the overall rate of consumption or in reliance on foreign resources. 3.13 It has been argued that the published figures on deficits are only the "tip of the iceberg." The Government uses the banking system to create credit outside the budgetary system on a large scale, and these credits should appropriately be regarded as deficit financing. In a largely state-owned economy, the distinction between borrowing by the Government and borrowing by state-owned enterprises is difficult, and deficit financing by the public sector is indeed much larger than shown in the budget. However, whether deficit financing suggests inadequate revenue or excessive expenditure is not clear. The unexpected credit expansion to finance inventories of agricultural products due to bumper crops (as in 1990) cannot be attributed to revenue inadequacy. Nor should the tax system be used to validate the losses of enterprises. To a large extent it is rationalization of expenditure, rather than tax mobilization, that is the issue in China for tackling the problem of excessive credit expansion in China in recent years. 3.14 Looking ahead, it has been argued that in China the Government should increasingly provide social security services now provided largely by enterprises. This will help in improving labor mobility and will facilitate the restructuring of enterprises. To the extent that these programs are implemented, Government revenue will increase; however, these programs should be funded from own sources, and social security contributions shoulid go hand-in-hand with fiscalization of the social security expenditures. In fact, the experience of social security systems in the developed countries warns strongly against relying on general tax revenues for funding social security programs. 3.15 Contrary to a priori expectations, the center's share in total revenue has not declined during 1988-90. In fact, it increased from 36.0 percent in 1988 to 37.5 percent in 1989 -- in part because the center obtains the major part of its revenue from taxes directly accruing to it. The shared pool of revenue is in fact only a small part of the total revenue. The share of the center's revenue contributed by the localities was only about 30 percent (?) in 1989-90 and was approximately at that level in 1985-87, before the new revenue-sharing contracts were initiated. Ad-hoc changes in the revenue-sharing formulas that the center enforces no doubt contribute to ex post results being different from ex ante expectations. 3.16 Nor is it clear whether the revenue-sharing system can be blamed for inadequate investment, particularly in key sectors. The rate of investment financed by the budget has of course been declining, but that is the intended result of decentralization. The overall investment rate continues to be high. As noted in Table SA.6, there was some significant decline in - 15 - investment in infrastructure (transport and communications), from 2 percent of GDP in 1985 to 1 percent in 1989, but in 1990 these investments increased sharply to 1.6 percent of GDP, double the level in 1981. (By international comparison, these figures seem low.) In key sectors such as energy and heavy industries, the investment rate has not declined significantly in recent years; in fact, for 1990, combined investment in these two sectors accounted for 7.7 percent of GDP, higher than in 1981 (5.6 percent of GDP). Thus the ex post figures on investment in total or in key sectors do not suggest that the revenue-sharing system has resulted in any major declines in these areas. Given the system of investment approval described in Chapter II, it is not surprising that the Government can control the level and composition of investment when it wants to. 3.17 Growth of extra-budgetary funds has indeed been rapid, and they now account for more than 90 percent of state revenue. This undoubtedly reduces the control of the Central Government over resource allocation, but that is the objective and consequence of decentralization. After all, when an economy moves away from state ownership, more and more resources become "off-budget." The key question is whether the decentralization mechanism chosen by the Chinese authorities is achieving the objective of marketization and efficiency. The evidence is circumstantial. The central financial authorities are naturally concerned about loss of their control, but the local authorities, sectoral ministries, and enterprises claim that the growth of extra-budgetary funds have allowed them greater flexibility and speed in decisionmaking and thus have improved efficiency. Figures indicate that about two-thirds of extra-budgetary expenditures go to investment and contributions to the Government budget. A large part of these funds are for depreciation, which are perhaps undervalued. The amounts spent on "Bonus" and *Welfare" account for about 13 percent. The item called "Others," which may hide unproductive expenditures, however, has increased from about 7 percent in 1983 to 12 percent in 1987. There is no basis to conclude that the efficiency of use of extra-budgetary resources is lower than that of budgetary resources. The term "extra-budgetary" may be a misnomer with an inappropriate, pejorative connotation' It would probably be counterproductive to reverse the trend and increase the size of the Government budget. Clearer rules of the game and the enhancement of the interests of decentralized units in the long-term profitability of enterprises are probably more promising directions, with Government playing some overall supervisory role. 3.18 Local protectionism is a growing problem in China. It is reported that individual provinces restrict sales of their raw material to other provinces. In some cases they have developed their own processing plants, duplicating investment facilities elsewhere. Plants for cotton textiles, tobacco, and alcohol are usually mentioned. There are examples of one province (Hunan) restricting the supply of its raw material (silk) to another locality (Shanghai), preferring instead to export it abroad (Hong Kong); the deficit locality (Shanghai) must then import the raw material from abroad (Hong Kong). There is a large (and excessive) number of automobile plants producing on an inefficient scale. Road blocks and other restrictions on interstate trade are apparently commonplace. .1/ Perhaps an alternative term, "public entity funds" would be more appropriate for the major part of the current "extrabudgetary funds." - 16 - 3.19 The problem of local protectionism in China is indeed serious, but the revenue- sharing system is only one -- and probably not the most important -- contributing factor. Whatever the revenue-sharing system, a locality will gain some positive marginal benefit (in terms of revenue-profits and employment) from locating a plant in its territory. The key question is whether these benefits will outweigh the costs to the locality and the nation. An effective check on setting up inefficient enterprises can be provided only if enterprises have to pay the true economic cost of capital, face hard budget constraints, and are not allowed to impose arbitrary restrictions on interprovincial trade. The experiences of other large federal systems (such as the United States and India) show that regulations against restrictions on interstate trade are necessary irrespective of revenue-sharing system. Market integration is essential for efficiency, as well as for equity among provinces, and should be an important item on the agenda of reforms in China. As will be discussed in Chapter IV, the revenue-sharing system should incorporate elements that encourages inter-provincial mobility and equity, but it also requires important changes in the enterprise system, the credit system, and the regulatory framework for interprovincial trade. 3.20 Finally, the issue of inter-provincial disparities in public expenditures merits analysis. Table SA.8 gives the per capita government revenues and per capita expenditures of the provincial governments. The provinces are arranged in ascending order of per capita revenue collection and are grouped into three classes -- "low," "medium," and "high" per capita revenues. In the low group, per capita expenditures are higher than per capita revenue collections, whereas in the high group, the reverse is true (except for Shanxi and Heilongjiang). The extent of equalization is brought out by the comparison of group averages (leaving out the extreme observations for Tibet and Shanghai at either end), as shown in Table 111.2. The gaps between group average per capita expenditures are much narrower than those between average per capita revenues. If one leaves out the municipalities of Beijing, Shanghai, and Tianjin because of their special expenditure needs, the group averages of per capita expenditures are remarkably close to each other. It is not clear what the mechanisms of transfer or the factors behind reallocations are. But the ex post results for per capita expenditures do suggest a remarkable degree of equalization among the groups, even though there are large differences among individual provinces, these do not seem to be correlated with their revenue generation capacity. Costs of Ad Hoc Adjustments 3.21 One reason that inherent weaknesses of the fiscal system are not revealed in the bottomline figures is that in practice the rules are not implemented. Various ad hoc adjustments are made whenever considered appropriate by the central and provincial authorities. From the experience of the 1980s it appears that despite decentralization, the central authorities -- the governmental and party apparatus -- do have substantial control over the credit and fiscal system of the country, particularly when there is consensus at the top of the hierarchy. However, the system that is developing is one of ad hoc directions and bargaining. Even though the product is not unimpressive, there are serious problems with the Rrocess. 3.22 The system of revenue sharing has been unstable. The instability is indicated by the fact that, as regards center-province fiscal sharing (often a constitutional matter for most - 17 - countries), there have been as many as three revenue-sharing systems since 1980. There were proposals for a new system beginning in 1991, in the context of the Eighth Five-Year Plan, but the changes have been postponed for the time being. In addition, of course, the rules of the game can be (and are) changed within the contract period. Such instability creates uncertainty about the availability of revenue to the center and provinces and is a serious hindrance to medium- and long-term programming for efficient resource allocation. 3.23 The system of intergovernmental fiscal relations is not uniform. Different provinces are given different formulas of revenue sharing with no clear rationale for choosing one formula over another. The system also lacks transparency: it is not clear to the parties involved what their fiscal share will be in revenue and expenditures. 3.24 In a contract economy there is a preMium on the bargaining skill of the participants involved. Improving productivity and mobilization of resources becomes a lesser priority than skillful bargaining and exercise of leverage. Reliance on such skills, however admirable they may be in themselves, is not conducive to efficiency in the economy. 3.25 Finally, among the parties involved, there seems to be a pervading sense of unfairness. The central authorities argue that their share of revenue is declining and want to make amends by various ad hoc levies. The provinces that are doing well (such as Guangdong) feel aggrieved that the center imposes arbitrary levies on them. The less well-to-do provinces (such as Hunan) complain about unanticipated costs from central policies and find ways of not delivering on their contracts. Provinces receiving special treatment, such as Hainan, manage to run unanticipated and unapproved deficits in their budgets. Rich provinces such as Shanghai complain of an unfairly high tax share and of unfair treatment in expenditure allocation by the center. Some of these pressures and sentiments are unavoidable in any large or federal state. However, the combination of instability, nontransparency, and unfairness in the system is a potential danger to the efficiency of the economy and the cohesiveness of the polity. 3.26 In view of the relatively good performance of the Chinese economy of during the reform period, there is a risk of complacency about the contract economy. Experience shows that, although inherent systemic weaknesses can be patched up for some time, they cannot be suppressed forever. Problems of efficient resource allocation in a non-market, command economy were suppressed for a long time but they did come to the surface eventually. The risks of short-term maximization and of decapitalization in labor-run enterprises, as in Yugoslavia, were pointed out by analysts; although not evident immediately, these problems did become real after a while. The Chinese authorities may be well advised to consider the inherent weaknesses of their present fiscal system and move toward its rationalization befor these weaknesses become seriously debilitating for the economy and the country. In fact, the Government has started an experiment in tax assignment in five provinces and one city and is interested in designing an appropriate nationwide system of revenue sharing between different levels of government to be implemented during the Ninth Five-Year Plan (1986-2000). - 18 - Table 111.1: China: Major Economic Parameters Related to the Fiscal System Parameter 1980 1985 1987 1988 1989 1990 1. Tax revenue (% of GDP) 25.6 20.4 18.4 18.5 18.4 2. Overall budgetary deficits (% of GDP) (GFS format) . 0.5 2.0 2.6 1.8 1.5 3. Subsidies (% of GDP) 5.9 5.9 5.4 6.1 6.1 4. Center's tax revenue Y (% of total tax revenue) 19.4 37.0 35.3 36.0 37.5 5. Investment (capital construction in key sectors (% of GDP) - Transport & communicationy 0.8 2.0 1.7 1.5 1.0 1.6 - EnergyX 2.0 2.4 3.0 2.9 2.8 3.1 - Heavy industriesy 3.6 4.5 5.2 4.9 4.4 4.6 6. Extra budgetary funds (% of state revenues) 67.0 79.0 81.2 of which: Investment (% of total) 61.7 60.0 /1 Format as defined in the IMF's Government Finance Statistics. /2 Figures available are those for "revenue;" need data on tax revenue only. 13 1981. Table 111.2: China: Interprovincial Disparities Group Average Per Capita Revenue Per Capita Expenditure Bottom third ("low")y 55 113 Middle third ("medium") 73 117 Top third ("high")II 112 108 [Note: See Table SA. 8 for province-by-province data.] /1 Leaving out the extreme observations for Tibet. /2 Leaving out the municipalities of Shanghai, Beijing, and Tianjin. - 19 - IV. ANALYTICAL ISSUES IN CONSTRUCTING A SYSTEM OF INTERGOVERNMENTAL FISCAL RELATIONS 4.1 The foregoing description of the present state of public finance in China in general, and of the complex structure of intergovernmental fiscal relations in particular, clearly indicates that fiscal reform has to be an indispensable and integral part of the process of overall economic reform. Similarly, recent macroeconomic indicators suggest that the time may be propitious for a new wave of reforms. 4.2 Short-term or transitional arrangements might have to make allowances for traditional thinking on the subject and the specificities of Chinese conditions. However, as regards the long-term perspective and solutions, the Chinese authorities cannot but follow the basic principles of multilevel fiscal systems. Three important postulates need to be emphasized. First, international experience in countries big and small, federal and unitary, indicates that the Central Government must have effective control over the most important sources of tax revenue, not only in terms of tax law and policy, but also in terms of administration, collection, and allocation of revenue. It is possible that, under appropriate conditions and safeguards, the sub- national governments could also make use of the same bases, but the major part of the revenue from taxes with nationwide bases would have to be collected by the Central Government. Second, taxes should not be looked upon mainly as providers of revenue; they must also be seen as tools of policy that, which whether intended or not, affect the allocation of resources and their efficient utilization, interregional and interpersonal distribution, and the level of aggregate demand. Third, marketization of enterprises cannot be achieved by administrative decentralization alone; for efficient resource allocation, hard budget constraints on enterprises are essential, and the enterprises must be brought under a rule-based tax system. In order to suggest directions for reforming revenue sharing system in China, this report draws on analytical work in the area of public finance as well as case studies of countries with particular relevance to China. The countries selected are Canada, Germany, Japan and India. Needless to say, all of these countries have their own problems in revenue-sharing and none can be regarded as a "model" for China. Analytical Issues and Case Studies 4.3 A basic problem in a multilevel fiscal system is to fashion the division of expenditure responsibilities and revenue flows so that the layered system, while allowing the desired degree of expenditure and revenue decentralization, would enable the public sector to pursue the goals of fiscal policy effectively. Accordingly, the major considerations to be kept in view in structuring intergovernmental fiscal relations are as follows. First, the revenue flows to the different levels of government must be adequate to enable them to fulfill their expenditure responsibilities, given the desired size of the total public sector. Second, the Central I/ For more detailed description of revenue-sharing systems in these countries, see Appendix. - 20 - Government's capacity to pursue effectively policies of macroeconomic management must be ensured. Third, the system must be conducive to the efficient use of resources in the economy and must promote fiscal discipline. Fourth, while allowing for local autonomy and promoting initiative at the local level, it must be possible for the center to effect transfers to the poorer regions so as to reduce interregional disparities; i.e. the center must have some surplus of revenues over its own expenditure needs. Fifth, in the division of tax resources among the provincial governments, weights must be assigned to both autonomy and incentive at the provincial level as well as to interregional equity. 4.4 In the discussion that follows, the division of expenditure responsibilities among the central, provincial, and local' governments that obtains now is taken as given, with the proviso that the stabilization and distribution functions should be responsibilities of the Central Government. At present there is a large degree of duplication in expenditure responsibility at different levels of government and this situation needs to be corrected. However, this report does not review that subject. Attention is concentrated on the manner of making revenue resources (taxes, mainly) available to the center and subnational governments and on the principles determining the relative amounts that different provinces should be given or enabled to raise. 4.5 There are four alternative ways in which tax revenues can be apportioned among different levels of government:!' (a) The center levies and collects all the revenues and gives a portion of the revenues to provincial governments as grants. (b) The center levies and collects all taxes but shares the proceeds of some or all taxes with the provinces on the basis of a formula or a set of formulas; (c) The center levies and collects most of the more important taxes, but the provinces have autonomy to levy and collect other taxes; in addition, the provinces get a share of one or more central taxes and/or receive grants from the center. (d) The center and the provinces enjoy more or less concurrent tax powers, so that the provinces can raise on their own most of the revenues they need; they may, however, also receive some grants from the center. 4.6 In practice the models may be modified (e.g., instead of sharing the proceeds of a tax, the center might share the base of a given tax with the provinces); again, there could be a I/ In this report, the term "local" is used to refer to subnational governments below the provincial level, whereas in official documents the term is used to refer to all subnational governments. 4/ For the sake of simplicity, only the center and the provincial governments are considered. - 21 - mixing of the elements of the different models (generally, the provinces or local governments are allowed to levy and collect some taxes). 4.7 In choosing among the models or in selecting elements from the models to build a mixed system, the major considerations set forth earlier must be applied. However, since center-provincial relations will ultimately have to be based on a political consensus, in practice the various criteria such as equity, efficiency, and scope for autonomous action of the provincial governments will have to be balanced against one another. Also, since taxes are involved, administrative considerations and the costs of compliance have to be given a high weight. 4.8 If the autonomy of the provinces or states is given very high weight, the fourth model -- of concurrent tax powers -- will be chosen. It is the system that might be favored in a federation with a pronounced accent on the decentralization of power. Apart from its nonapplicability to a unitary country such as China, experience has shown that it is also a messy system, with overlapping tax powers, high cost of compliance, and potential conflict among the provinces regarding the right to tax on the basis of residence versus origin. As will be argued later, some of the important taxes are not capable of being efficiently levied and collected by subnational governments. Among modern federations, only Switzerland and the United States (and to some extent, Canada) continue to operate systems based on concurrent tax powers. In the United States, however, over time the most important tax, has become predominantly a federal tax: the yield of the federal tax on individual incomes and corporate profits far outweighs the yield of the state income taxes. Effectively, a separation in the use of tax powers has come into existence, with the U.S. Government relying predominantly on income taxation and the states on sales taxes, although concurrent tax powers are being used, resulting in complications and inefficiency. 4.9 A more clear-cut trend toward centralization of tax collection can be seen in Australia and Canada, two federations whose constitutions had provided for some degree of concurrent tax powers. In Australia, the Constitution gave the Commonwealth (federal) Government unlimited taxing powers subject to the condition that it could not discriminate between states or parts of states, and the states were given concurrent powers with the federal government over all taxes except customs and excise duties. However, through the pressure of emerging circumstances and for other reasons, the collection of taxes has become highly centralized, with the federal government collecting nearly 80 percent of all tax revenue. During World War II, the federal government passed legislation to become the sole authority to impose taxes on income for the duration of the war and a year thereafter, and the states were compensated by tax reimbursement grants. This system has continued ever since. 'Proposals for allowing the states to levy income taxes with a simultaneous partial withdrawal of the federal government from the income tax field have been rejected on several grounds, the important ones being: (a) allowing the states access to income taxation would make the task of economic management more difficult; (b) the states would face budgetary problems because of fluctuations in income tax collections; and (c) the per capita yield of income tax would vary between states, calling for additional measures of equalization. Thus, the income tax has become firmly embedded in the Australian fiscal system as a federal tax. - 22 - 4.10 The framers of the Canadian Constitution wished to create a strong central government and granted to the federal government unlimited powers of taxation. The provinces could levy direct taxes within their boundaries. Originally, the federal government did not levy the income tax, but the course of time, it did become an important source of federal revenue. Although some of the provincial governments levy their own personal and corporate income taxes, the majority get their taxes -- levied as a percentage of the federal tax base -- collected by the federal government. The major part of the income tax revenue goes to the federal government. The provinces levy a retail sales tax on consumption, but the federal value-added tax (VAT), recently levied, will soon predominate over the provincial sales taxes. Thus the growing responsibilities of the federal government have led to its imposing the two most important taxes with wide bases, the income taxes and the VAT. 4.11 Although the Constitution of Germany gives the federal government and the states concurrent legislative process in respect of most taxes, it also stipulates that once the federal government occupies a tax field, the powers of the states in relation to that tax will cease. The federal government levies both income taxes (personal and corporate) and the VAT -- these being the most important and productive taxes. The states are assured of autonomy by a constitutional provision regarding the sharing of proceeds from the income and corporate taxes with the states and local authorities; the percentage shares going to the different levels of government are specified in the constitution itself. The proceeds of the VAT are also shared between the federal government and the states; the percentage shares determined by agreement. Apart from these two major shared taxes, some taxes belong entirely to the federal government (customs excise duties, except on beer, and surcharges on income taxes) and others belong to, and are levied by, the states (taxes on wealth, beer, motor vehicles, and inheritance). The German system combines the advantage of uniform rates in respect of taxes with nationwide bases with financial autonomy for the states. There is clear separation of tax powers, centralization of the levy of the major taxes, and substantial sharing of the yield of the most productive taxes. 4.12 Like the framers of Germany's constitution, the framers of India's constitution also learned from the experience of the older federations with a system of concurrent tax powers. Impressed by the difficulties and confusion created by such a system as well as by the need to give to the Central Government control over important taxes (productive taxes with nationwide bases), the framers of the Indian constitution provided for separation of tax powers, centralization of the levy and collection of the most productive taxes with nationwide bases, and a flexible system of sharing of tax proceeds. The fiscal autonomy of the states was assured through (a) the assignment of some substantial tax powers to the states, (b) assignment of the entire net proceeds of some taxes (which could be levied by the center) to the states, (c) sharing of the proceeds of some central taxes, and (d) grants-in-aid to be recommended by a Finance Commission. Customs, excises (except on liquor), all taxes on nonagricultural income and nonagricultural wealth and inheritance, taxes on railway freights and fares, and the sales tax on interstate trade were the important taxes given to the center. The states were given sales taxes on intrastate sales, taxes on agricultural income, and land and inheritance, the urban property tax, a motor vehicles tax, a tax on bus passengers and freight, a profession tax, and taxes on amusements. Income taxes on the expanding nonagricultural base, excises (apart from those on - 23 - liquor) that affect all industries, and excises that could be converted into a VAT at the manufacturing stage are given to the Central Government under the Indian constitution. This has been done because the first two (i.e. income tax and excises) are potentially the most productive sources of revenue and because central control of all the three taxes (along with customs) is needed to preserve the unity of the common market. 4.13 The four countries considered above all have federal or quasi-federal constitutions. Japan, in contrast, is a unitary state like China, although much smaller in size. Japan has a robust and active system of local government. In terms of net expenditures (i.e., excluding intergovernmental transfers), local governments account for about two-thirds of public expenditure. The Central Government's direct expenditure amounts to only one-third of the total, but the Central Government collects about 64 percent of the revenue. As in most unitary countries, important taxes such as income taxes and the VAT are levied and collected by the Central Government. 4.14 The experiences of the major countries that have multilevel fiscal systems and the evolution over time of the different facets of fiscal federalism yield valuable lessons and guidelines for the future. Although the practices vary, the trends are clear: the broad pattern that has emerged seems to be that the Central Government has control over the most productive taxes, particularly the income taxes and the VAT; that the provinces or states also enjoy some substantive tax powers, subject to restrictions; and that there is linkage between the central and provincial fiscal systems through tax sharing and/or grants, with differences in the relative importance placed on tax sharing (or tax-base sharing) and on grants. 4.15 A system of concurrent tax powers should be avoided. The best solution to the problem of division of revenues would be a clear separation of tax powers. Although it would be desirable to make available to the subnational governments some substantial taxing powers, the most productive taxes and those which have the potential to affect the allocation and efficient use of resources in the national economy should be levied and collected by the Central Government. This is necessary to avoid undesirable effects on the economy, to give the center effective control over the level of taxation in the economy, and to make possible the creation of a surplus of revenues over its own expenditure needs that the center can use to make equalizing transfers to the poorer provinces or states. The provinces could be given the freedom to tax their own respective residents to a greater or lesser extent, but they should not be endowed with the power to levy taxes that would distort the allocation of resources in the economy or fragment the common market. If the separation of tax powers is worked out with the above considerations in mind, it would probably turn out that the revenues that the provinces can raise on their own would fall short of their expenditure needs, given the normal degree of expenditure decentralization in large countries. This would give rise to the familiar vertical fiscal gap, which must be filled by the sharing of one or more of the centrally levied taxes or by grants-in-aid from the center. This is borne out by the inter-governmental fiscal arrangements that have - 24 - evolved in Australia, Canada, Germany, India, and Japan.11 Grants would be needed also to tackle horizontal fiscal imbalance in countries where there are significant interprovincial disparities. Separation of Tax Powers 4.16 The broad pattern of separation of tax powers that would be desirable has been indicated above. It is necessary to discuss briefly the principles of assignment of tax powers between the center and the provinces. The following principles may be stipulated: (a) Taxes with procyclical trends in yield should be assigned to the Central Government; taxes assigned to the provincial governments should be, as far as possible, cyclically stable. (b) Taxes that are to be levied on a progressive basis for redistributive purposes should be levied and collected by the center -- not only because the degree of progression has to be uniform throughout the country, but also because the base must be global (e. g., income from all geographical and occupational sources must be included in the base of a progressive personal income tax). (c) Taxes on production, such as excises on manufactures and the VAT, should be assigned to the center; nonuniform product taxes levied by provinces would lead to distortions in the locational allocation of resources and in the pattern of factor use. (d) Taxes with bases that are quite unequally distributed among the provinces should be assigned to the center; (e.g., a tax on natural resources). (e) Taxes on immobile and localized bases can be assigned to the subnational governments. (f) The provinces could be assigned the power to levy a tax on the consumption of their residents; i.e., a destination-based product tax can be assigned to the provinces (a retail sales tax is one such tax). (g) A supplement to personal income tax could be assigned to the provinces, with a ceiling to be prescribed by the center. (h) The center as well as the provinces may levy user charges for services supplied. I/ For more details on intergovernmental fiscal relations in Germany, Canada, India and Japan, see the Appendix. 6/ The kinds of taxes that could be assigned to local bodies will not be discussed. - 25 - 4.17 The above can be called the economic principles of tax assignment.2' To them should be added a criterion derived from administrative requirements. In assigning a tax to one level or another, the cost of collection and the administrative feasibility of determining the base should be taken into account. (If a tax has a base that extends beyond the boundaries of one province, it will be difficult for any one province to determine that tax base accurately.) 4.18 Income taxes -- personal and corporate (or on enterprises) -- should be assigned to the center. It is generally desirable that the personal income tax should be levied on a progressive basis, and for this purpose the base of the tax should be defined comprehensively and income from all sources should be aggregated to form the base. In a unitary state, presumably it would also be desirable that the degree of progressivity should be uniform throughout the country. If the provinces are given the power to levy their own income taxes, the rates and the degree of progressivity could vary from province to province. This could be avoided if the center levies one common tax and lets the provinces collect it and retain the respective collections. However, as stated above, the desired degree of progressivity and equal treatment of assessees with equal taxable capacity can be achieved only if the incomes of a taxable entity (an individual, a partnership, or a firm) from all occupational and geographical sources can be aggregated. Any given province would find it difficult, perhaps impossible, to capture adequately incomes arising from outside its jurisdiction. In China, the personal income tax is not important yet, and individuals and private firms may not now be deriving income from places other than the province of domicile. But in the future, with greater private activity and growing ownership of businesses and stocks of enterprises, better-off individuals and growing businesses would begin to derive incomes from more than one province. Hence, in order to implement an equitable personal income tax, it is necessary for the center to levy and collect it. In China, the intention is that the center would continue to levy the personal income tax. The provinces would not derive any advantage by having the power to collect it, if the tax rules are to be applied strictly and the proceeds are to be shared. Considerations of efficiency and economy in collection also warrant collection by the center. 4.19 There are even greater objections to assigning the tax on profits of enterprises to the provinces. Joint enterprises, foreign enterprises, and even public enterprises would in the course of time have productive activities in more than one province and sales throughout China. It would obviously be erroneous to assume that all the profits of an enterprise are generated where the head office is situatedY As more and more prices are freed and as household savings are invested in enterprises, either directly or through financial intermediaries, the tax on the profits I/ For a detailed discussion of the principles of assignment, see R. A. Musgrave, "Who Should Tax, Where and What?" In Tax Assignment in Federal Countries, ed. Charles E. McLure (Australian University Press, Canberra, 1983). B/ Investible resources would also be coming to an enterprise from provinces other than the one of domicile. Already, there are province-province joint enterprises (e.g., in Guangdong). See Charles E. McLure, Tax Assignment in Federal Countries, (Australian University Press, Canberra, 1983) for a discussion of different aspects of the problem of assigning the corporate profits tax in a federal system. - 26 - of enterprises will have an impact analogous to that of a corporate profits tax in a market system. The incidence of the tax will fall partly on the suppliers of capital and partly on the users of products. That being so, the province where the enterprise is situated cannot claim exclusive right to appropriate the tax on the profits of the enterprise because the "ultimate" contributors to the tax will not all be resident in that province. Sharing the tax with a province does not carry that implication, but assigning the tax to the provinces would. This is not just an academic point. The assignment of the tax on enterprise profits to the provinces would mean that the richer and more industrially developed provinces would gain proportionately more, and the differences in per capita revenues from the income tax would become large among provinces. Such differences can be avoided if the center levies the enterprise income tax and then shares a part, even a good part, of the proceeds with the provinces. In the formula for the distribution of shareable proceeds among the provinces, collections in a province could be given a substantial weight. 4.20 The tax on the profits of enterprises would increase in importance with the fast growth of industry and commerce that China is planning to maintain. Its impact on the economy would increase accordingly. It is a fiscal policy tool that the center should be handling. An efficient pattern of industrialization would not result if in the implementation of the tax the provinces offer their respective enterprises differing concessions. 4.21 Finally, the income tax yield fluctuates with changes in the level of economic activity and income. As mentioned earlier, taxes with cyclically unstable yield should be assigned to the center. 4.22 It is generally argued that origin-based product taxes, such as excises on manufactures, should not be assigned to subnational governments because differential tax rates would cause distortions in the location of industries. The same reasoning will apply to a VAT levied by provinces at varying rates. Here again, if the center determines the rate(s) and the base, uniformity can be ensured. However, there will be difficulties in working out the scheme for refunding input taxes when transactions take place between firms situated in different provinces. 4.23 It was argued earlier that taxes with bases that are unequally distributed should not be assigned to provinces, since that could result in unequal distribution of revenues. This argument applies to the VAT as well as to product taxes because industrial activity and the value of industrial output are unequally distributed among the provinces. 4.24 In China as in most developing countries, the major part of the revenue comes from internal indirect taxes on commodities. If the VAT or a general product tax is assigned to the provinces, the center will be short of its revenue requirements. For this reason, the Chinese authorities wish to assign to the center the VAT and the product taxes in the experiments in tax assignment they wish to carry out in a few provinces. - 27 - 4.25 Taxes on natural resources should be assigned to the center for the reason that such resources tend to be distributed quite unequally among the provinces. Part of the revenues from these taxes could be distributed to the provinces from which the resources are extracted. 4.26 Destination-based product taxes, such as a selective or retail sales tax, can be assigned to the provinces. Similarly, taxes on consumer services such as entertainment can be given to the provinces. 4.27 Because taxes on localized and immobile bases can be given to subnational governments, taxes on land and real estate can be assigned to the provincial governments, which may share them with, or delegate them to, lower-level governments. 4.28 Business taxes in the form of license fees, as well as a tax on gasoline (within limits) as a proxy charge for road use, may be assigned to the provinces. Similarly, taxes on vehicles could also be given to the provinces, which may share the proceeds with lower-level governments. Restructuring Intergovernmental FIscal Relations in China: The Long-Term Perspetye 4.29 As noted above, this report does not consider the issue of reforms in expenditure responsibilities of center and provinces (see Chapter II). To start with, taking the present division of expenditure responsibilities as given, this report discusses the principles of assignment of taxes, tax sharing, and grants from the center.2 It should be emphasized that if there is a different division of expenditure responsibilities, the final resource sharing will, of course, change accordingly, but the principles discussed here would still apply. 4.30 In contrast to any other country of comparable size with multilevel finance, in China all taxes, except for a few minor ones, are levied by the center, (i.e. tax bases and rates are determined by the center). Only the center can introduce new taxes or abolish any existing ones. The provincial and local governments have no legal taxing powers. Again unlike any other comparable country, almost all the taxes in China are collected mostly at the county and city levels by officers of the local tax bureaus under the supervision of the Provincial Tax Bureaus. The center itself has only a small tax-collecting machinery. The revenue collections are effectively under the control of the provincial governments. The collected revenues are "shared up" from the provinces to the center.A Apart from China, only in Germany are the centrally (federally) levied taxes collected by any other than Central Government tax officials: in Germany the federal taxes are collected by the tax officials of the state governments, but under 2/ In a planned economy, one other aspect may be considered, - the principles of allocating of capital funds among the provinces for financing investment in infrastructure. JQ/ However, nonshared revenues - tax and non-tax - taken together, would seem to form the major part of central revenues. - 28 - the supervision of federal officers; also, the collections go to the federal government and then are "shared down" with the state and local governments.!" 4.31 As already indicated, the Central Government in China is able to undertake macroeconomic management yet has only loose control over how (and how effectively) the tax laws are applied. Instead of distributing a share of its own taxes to the provinces on the basis of well-defined formulas, the center is at the receiving end. It gets a good part of its tax revenues from the provinces on the basis of contracts that are based more on bargaining than on principles applied across the board and that tend ex ante to reduce the share of the center in total revenues as national income grows. 4.32 If all prices are controlled -- i.e., if input and output prices are given to the enterprises and they are expected to produce only the targeted outputs -- then the enterprises profit margins are predetermined. It can then be argued that any tax levied on the enterprises would only cut profit margins and that in this sense the taxes are "borne" by the enterprises themselves.L' This argument assumes that in fixing prices some adjustment was not made to ensure a certain "normal" rate of profit. In any case, now that prices are being increasingly freed and enterprises can sell above-quota output at market or quasi-market ("guided") prices,-1 product and business taxes and the VAT will be shifted forward to an increasing extent. Under conditions of price determination (largely by the market), the incidence of the tax on enterprise profits will fall mostly on the suppliers of capital and partly on the users of commodities.L' The suppliers or owners of capital would be the shareholders of joint enterprises and those who supply capital to public enterprises (the taxpayers who supply funds via the government and household savers who do so via the banks or government bonds). Thus, those who bear the taxes (and from whom the contributions to the treasury ultimately come) will be spread across the country. Therefore, the taxes collected from enterprises situated in a province cannot be considered to be mainly the contribution by the residents of that province, except when the tax is destination based (e.g., a retail sales tax). 4.33 When the incidence and other effects of taxes affect the people of all provinces, it is right that the center should determine the structure of the taxes. However, as experience has demonstrated, delegating tax collection to the provinces and letting them entrust that task to the counties and cities actually means that the tax laws or rules will not be properly applied. The 11! R. J. May, Fiscal Federalism and Fiscal Adjustment, (Clarendon Press, Oxford, 1969), p. 111. 2/ In China, indirect taxes are not added to price. Formally, the tax is to be paid out of the proceeds of sale at the prices already fixed, i.e., the tax is "price-inclusive." J/ It is estimated, in respect of retail prices, that "free market" prices, guided prices, and state- fixed prices account for 45 percent, 25 percent, and 30 percent, respectively, of total retail sales. IMF, People's Republic of China: Background Papers (Washington D.C., January 1991). 14/ Ultimately, the real burden of taxation will rest on individuals, not institutions. - 29 - position is made worse when the provinces are also permitted to offer tax preferences and concessions. Then not only are the tax bases eroded -- but equally important -- the incidence of the taxes is arbitrarily affected. Again, revenue contracts with enterprises make the whole system indeterminate in terms of allocative effects and incidence.L' Finally, the system of revenue contracts, on the basis of which the center obtains its share of revenues from the provinces, tends, if strictly applied, to reduce the income-elasticity of Central Government revenues. If these deficiencies are to be removed, it follows that reforms in fiscal system must be considered. The approach of the authorities 4.34 The Chinese authorities themselves are contemplating a shift from the tax contract system to a system of tax sharing based on rules. According to current plans, during the next five years beginning June 1991, while the tax-contracting system will be continued with possible improvements, an experiment in tax assignment or tax separation will be conducted in two provinces and six cities. If the experiments prove to be successful, a full-fledged system of tax assignment would be implemented in the Ninth Five-Year Plan period (1996-2000). The cities and provinces participating in the tax sharing experiment are: Shenyang, Dalian, Tianjin, Qingdao, Wuhan, Chongqing, Zhejiang and Xinjiang. Local taxes are: 13 small taxes assigned as local fixed revenue plus agriculture tax, local enterprise income tax, collective enterprise income tax, private enterprise income tax, salt tax, foreign enterprise income tax and joint venture income tax (if the partner is local government). Central taxes are: custom duty, product tax and VAT collected by customs, central SOEs income tax and profits, income tax from railways, and income tax from the banking system. The shared taxes are: product tax, VAT and industrial and commercial tax, business tax and resources tax. For shared taxes, the sharing ratio for the six municipalities and Zhejiang Province is 50 percent and Xinjiang can retain 80 percent of the shared taxes, since it is the only minority region participating in te experiment. 4.35 The relative shares of the different taxes, levied now, in total tax revenue are estimated (1990 budget) as shown in Table IV. 1: 1J/ Although legally the provincial governments are not to deviate from the course of collecting indirect taxes according to the tax code, negotiations and bargaining do play a part in respect of these taxes also. See IMF, People's Republic of China: Background Papers (Washington, D.C., 1991). - 30 - Table IV. 1: China: Estimated Percentage Shares of Current Taxes in Total Tax Revenue, 1990 Tax Percent of Total Taxes on income and profits 23.4 Taxes on goods and services, 52.5 of which: Product tax (17.5) Value-added tax (15.0) Business tax (20.0) Salt tax 0.2 Taxes on international trade (customs) 6.5 Others (including taxes on extra budgetary construction (7.5%) 17.3 TOTAL 99.9 Source: Based on data provided by Ministry of Finance. 4.36 The intention seems to be to secure for the center 60 percent of total tax revenue. The product tax, the VAT, the salt tax, the taxes on international trade, and the taxes on extra- budgetary construction and funds would give the center around 47 percent of total tax revenue. Some share in the revenue from the business tax would raise the center's share in total tax revenue to 55 percent. If non-tax revenues are also taken into account, the center's share in total revenue (excluding receipts from borrowing) would perhaps add up to 60 percent. Of the total revenues accruing to the center, 16.7 percent (i.e., an amount equal to 10 percent of total revenues, according to current thinking) would be given to the needy provinces as grants. This would mean that the direct expenditures of the center and the provinces would be largely equal. Tax assigmnent 4.37 This report accepts, for illustrative purposes, the figure of 60 percent as the center's initial share. This is in line with the proposition, enunciated earlier in this chapter, that the center should be enabled to have a surplus of revenues over its own expenditure needs. The fiscal reforms and changes that the mission would like to recommend, however, represent a fundamentally different approach to the problem. The mission is of the opinion that taxes on income and profits (excluding the agricultural tax) should be assigned to the center, along with the product tax and the VAT. - 31 - 4.38 The provinces will be assigned a new tax in the form of retail sales tax, which will replace the business tax on wholesale and retail tradeA' The provinces may be permitted to levy the tax at rates of their choice falling within the band of, say, 4-6 percent. They also could be assigned the business tax on services (excluding important activities such as electricity, insurance, and banldng and with rates determined by the center), the agricultural tax, the real estate tax, and the other minor taxes. The taxes on extra-budgetary funds and extra-budgetary construction, so long as they exist, can be assigned to the center, along with business taxes on important activities. The provinces will retain the collections from the taxes assigned to them and may also be permitted to determine the rates of their taxes other than the retail sales tax and the business tax. 4.39 The proposal that the center should levy and collect the taxes on income and profits as well as the product tax and VAT is designed to ensure that the system will become based on rules and that the tax law will be properly applied. A full shift to the rule-based system could be made fairly easily in the case of indirect taxes. In the case of income taxes, the switch to a rule-based system would require some changes and simplification in the tax provisions. It would be necessary to reduce the rate of tax simultaneously with the withdrawal of all but the essential exemptions and deductions. The rate reduction is also required if enterprises are to be prohibited from deducting debt repayment from taxable profits. Tax sharing 4.40 The assignment of the most important taxes to the center would call for the sharing of part of the proceeds of the central taxes with the provinces. The major questions to be considered are: (a) what are the taxes to be shared, (b) what proportion of the designated tax collections should be shared, and (c) what should be the basis for distributing the shareable pool among the provinces? 4.41 Taxes and proportions to be shared. It would be best to confine tax sharing to the three important taxes with wide bases: taxes on income and profits, the product tax, and the VAT. The proportions to be shared must be determined with two considerations in mind: incentives for tax mobilization on the part of provinces must be preserved, and the amounts going to the center and provinces should be such as to help reach the 60:40 shares in total tax revenues that have been envisaged by the central authorities. Accepting this ratio for illustrative purposes, the following percentages are suggested: Cena Provinces Taxes on income and profits 50 50 VAT and product tax 60 40 j_/ This means that in the course of time, the VAT could be extended down to the wholesale level. - 32 - If these proportions are applied, the provinces will obtain from the shared taxes around 28 percent of total tax revenues. The retail sales tax, the tax on some business activities, and the minor taxes will give the provinces another 12 percent of the total, raising their share to about 40 percent of the total.' 4.42 Because the provinces will obtain as much as 50 percent of revenues from the income and profit taxes, they will have sufficient incentive to encourage and enable their enterprises to maximize their profits. A substantial degree of price reform is also needed to persuade the provinces to let their enterprises pursue commercial goals. 4.43 Basis for distribution. In the present system, the shares retained by the provinces are related to collections, but the differing terms of revenue contracts introduce several modifications. Therefore, it is difficult to infer the implicit weights given to different relevant factors such as collection, population, and level of development. However, it is clear that the differences in the contract terms were introduced in part to take care of differing circumstances. When one is looking for a uniform formula to be applied to all provinces, in some way "the needs aspect" should be taken into account. The respective collections in the provinces cannot be the only base for distribution, for two important reasons. First, revenue collections are concentrated in a few provinces; as much as 41 percent of revenues is collected in only five provinces. Distribution on the basis of collections alone would result in large inequalities in per capita revenues from shared taxes. Second, as argued earlier, collections in a province do not represent only contributions by its residents (although they do reflect residents' contributions to some extent). Also, indirect taxes levied on firms represent to some degree charges for public services that have been rendered and have gone as inputs into production. There is a case for giving substantial weight to collections in the distribution formula,' and population is a good and simple indicator of need. It would be desirable to base the distribution on both. 4.44 It is recommended that about 70 percent of the provincial share of collections from the income and profits taxes be distributed on the basis of collections in the province and about 30 percent on the basis of the province's population. In the case of shared indirect taxes, the distribution formula should give equal weights to population and collection. 4.45 In Germany, where tax sharing plays an important role, population and collection are given weights in the distribution formula. In India, where tax sharing is also important, population, backwardness (as indicated by relative levels of per capita income), and collections (only to a minor extent) are given weights. In Japan, the local allocation tax is given to localities on the basis of a formula that is analogous to formulas applied to determine the distribution of grants-in-aid: the attempt is to cover the gap between basic financial need (BFN) 17/ These figures are derived from the percentage shares in total revenue of the different existing taxes indicated in Table IV.1. They are in the nature of tentative suggestions. jj/ Giving substantial weight to collection would also implicitly reward better performance. - 33 - and basic financial revenue (BFR), these being determined somewhat normatively. In recommending the distribution of shareable taxes using population and collection as weights, which is in line with practice in other countries, the intention is to introduce incentives for initiative and to reward performance, and at the same time to link distribution to needs as indicated by population size. By itself, the last will not result in the desired degree of equalization; hence the need for grants-in-aid (see below). Qualifying statements 4.46 In making these suggestions for tax sharing, it has been assumed that the accrual of tax revenues to the center and the provinces, after tax sharing, should be in the ratio of 60:40. Several qualifications are necessary. First, if this ratio is to apply to the shares in total receipts including borrowing, the central share in tax revenue could be less than 60 percent. Correspondingly, the proportions for sharing suggested in respect of income and product taxes would have to be altered. Second, the sharing proportions will also be influenced by the shares of the central and subnational governments in direct expenditures, which it would be necessary to maintain. At present the share of the subnational governments is around 65 percent. If this proportion is to be maintained, tax sharing should yield more than 40 percent of revenues to the provinces -- otherwise, the transfers needed through grants-in-aid would become unduly large. Third, the formulas for the distribution of shared taxes among the provinces must be considered tentative. In the absence of the required data, it has been possible to estimate the relative shares of the different provinces in total revenues that would accrue to them if the suggested weights are applied. The final weights to be given to collection and population would have to be decided after studying the impact of the suggested formulas. The mission believes that the basic reasoning advanced would remain valid, but that the exact figures might have to be changed somewhat. Meanwhile, every province will be protected by the rule recommended for the initial period: that none should receive less, in real terms, than it is getting when the changeover takes place. Grants-in-Aid 4.47 As shown in Chapter III (Table 11.2; see also Table SA.8), the existing system accomplishes a substantial degree of equalization of per capita expenditures among provinces with high, medium, and low per capita revenues. However, the scheme of equalization is based not on formula but on informed judgment. In view of the very different conditions prevailing in the poor provinces of southwest and northwest China, it is clear that variables cannot be introduced into a general formula to take adequate care of their requirements. Therefore, for these provinces, special schemes of assistance or grants-in-aid will have to be continued. However, for the counting as a whole, it would be desirable to move toward a scheme of formula-based grants. 4.48 Grants can be of different types, depending on the objectives to be attained. Equalizing block grants would be needed to supplement tax sharing. Tax sharing is primarily intended to address the problem of a vertical fiscal gap; to the extent that an equalizing element (such as the criterion of population) is introduced into the distribution formula, tax sharing serve - 34 - also to narrow horizontal imbalances somewhat. Equalizing block grants would directly attack the problem of fiscal inequalities. Citizens living in some areas should not be disadvantaged because multilevel finance creates fiscal inequalities among jurisdictions. These inequalities arise because taxable capacities are usually unequal, whereas per capita standardized expenditure needs for providing the essential services would be the same (apart from special cost disabilities, which can be taken care of separately). Block grants are essentially intended to equalize, fully or partly, these differences in fiscal or taxable capacities. The taxable capacity of a province can be defined as the amount of potential per capita tax revenue that a province can raise by applying a standard tax structure (that represents the average tax structure for all the provinces) plus the per capita tax share it receives from the center to its own tax bases. Its deficiency in (excess of) taxable capacity is the difference between its own taxable capacity and the per capita yield of the standard tax structure applied to all the provinces plus the per capita average tax share for all provinces. The amount of the equalizing grant that a province should get is equal to its deficiency in taxable capacity multiplied by its population. Given the competing claims on the center's own revenues, it may not be possible for it to cover the entire deficiency in provinces' taxable capacities, but the aim must be to cover as much of it as possible. 4.49 Though block grants are not supposed to be earmarked -- so that the provinces may enjoy freedom in deciding for what purposes the grants should be spent -- it would be legitimate to stipulate that the grants should be spent only on specified essential or basic services. The grants would thus become partially conditional; however, the provinces will have the freedom to distribute the grants among the specified services as they consider best. In addition, the provinces may be given specific grants, tied to particular services, in order to take care of special problems or to promote certain national priorities. Such grants seem not to be of much importance in China now when the provincial budgets are formulated as part of the national budget. Bureaucratic Interface between Center and Localities 4.50 It seems desirable for the Central Government to consider establishing in the central bureaucracy a mechanism to assist in devising and implementing the rules for center-local fiscal relations and protecting the interests of localities. This could be seen as an instrument for realizing the "principle of giving full scope to the initiation and enthusiasm of the local authorities under the unified leadership of the central authority," as stated in Article 3 of the Chinese Constitution. The precedent for this type of setup exists in Japan (see Appendix), where the Ministry of Autonomy plays an important role in designing and implementing the local and intergovernmental system. Japan uses bureaucratic balance of power between the Ministry of Autonomy, which is backed by the constitutional principle of local autonomy and other central ministries to "help keep local and central relations balanced." The Japanese system seems to provide effective institutional protection for local governments in keeping intergovernmental rules working properly, and it deserves careful study for its adaptation in China. - 35 - V. TRANSITIONAL ARRANGEMENTS FOR LONG-TERM REFORM IN THE INTERGOVERNMENTAL FISCAL SYSTEM 5.1 For many legitimate reasons, the reforms suggested in this report cannot be introduced in isolation or overnight. Reforms in the pricing of inputs and outputs will have major consequences for the profits and losses of various enterprises, hence, for the revenues and profits earned by different provinces. Social security and housing reforms may increase the fiscal responsibilities of government at various levels. The degree of opening up to foreign trade and investment on the provinces will make a big difference in their revenue base. The degree of market integration and mobility of labor and capital that is practiced in the economy will affect the revenue potential of different provinces. The nature of financial sector reforms will have an influence on how much the enterprises in different provinces can rely on credit to finance their needs. For all these reasons, a rule-based tax and revenue-sharing system can only be introduced in tandem with other reforms. It is also clear that medium-term plans have been formulated by the provinces and the center on the basis of revenue expected from the current system. Any sudden changes that can disrupt this situation should be avoided. In any case, in the quasi-federal system that China is in practice, changes in center-province fiscal relations can only come on the basis of consensus, which will take time to build. 5.2 The Chinese process of reform has been distinguished by its emphasis on experience and experimentation to supplement and test a priori theoretical analysis. In view of the success achieved in the Chinese programs thus far, there is something to be said for this approach. However, tax-sharing formulas are by their nature national in character, and effective experimentation in this respect is not possible for selected provinces. An alternative approach would be to experiment with certain types of taxes at the national level. In this context, it is important to emphasize that the transitional arrangements should not conflict with the long-term objective. As such, it is probably a mistake to assign the income tax to provinces, even in an experimental stage. Rather, the income tax could be selected for experimental collection by the center, with a certain percentage distributed according to agreed formula and along the lines discussed in this report. Provision could be made, that under the transitional arrangements, no province would receive less revenue in real terms from the income tax than it received in 1991 under the present system. Reform of the income tax and contract system should be initiated on a national basis; among the elements of these reforms could be the following: (a) All new enterprise contracts should be formulated on an after-tax basis, so that the role of the state as tax collector and as owner of enterprise capital can be separated. (b) A unified tax system, including tax rates and the basis for calculating taxable income, should apply to all domestically owned enterprises, regardless of their form of ownership, and should accrue to the Central Government. (c) Interest payments, but not loan repayments, should be allowed as a deductible item for tax purposes. - 36 - (d) If the above is done, it may be appropriate to consider reducing the income tax rate to the 30-40 percentage range. (e) To bring about greater discipline in credit creation, it is suggested that macro plans for credit creation and currency issuance be submitted to the People's Congress for review and approval, as is the case for the fiscal budget, and that the credit plans should be strictly adhered to once approved. Deviations from the plans would require legislative approval, as does the budget. To reduce the pressure of local authorities on the local branches of financial institutions, the branches should be organized on a regional, rather than provincial basis and the managers of these branches sho-ld be appointed by the Central Bank's Head Office and be rotated among the regions. - 37 - VL CONCLUDING REMARSK 6.1 Reform in center-local fiscal relations should become a part of the new wave of reforms that seems to be beginning in China. These fiscal reforms can reinforce, and also be strengthened by, other reforms. 6.2 Introduction of any such reforms can only follow a process of consensus building undertaken with the provincial authorities. There is a need for agreement on principles as well as on the parameters for resource distribution. The provincial authorities must be convinced that their expenditure capacities will not be unduly adversely affected by surrendering some of their authority to collect revenue. There has to be an agreement on the basic expenditure responsibilities of the local governments before a revenue sharing formula can be implemented. It may also be possible to apply revenue sharing formula initially to incremental revenue only leaving the distribution pattern of base year revenue unchanged. Detailed statistics on tax collection and expenditures at different levels are needed so that simulation exercises can be made showing the implications of various sharing formulas for central and local finance. A seminar involving Chinese authorities (both central and provincial) and international experts on the subject would be helpful for building consensus about the principles of resource sharing. This could be followed up by the simulation exercises illustrating the implications of various revenue sharing formulas, which also would help in consensus building and eventual agreement at the political level. The process could be completed within the Eight Five-Year Plan (1991- 96), as envisaged by the Chinese authorities. - 38 - APPENDIX; INTERGOVERNMENTAL FISCAL ARRANGEMENTS IN GERMANY. CANADA, INDIA, AND JAPAN A. 1 The following major aspects of intergovernmental fiscal arrangements will be covered in this appendix: (a) the assignment of expenditure responsibilities, what degree of state/provincial autonomy it reflects; and whether it is conducive to macroeconomic management; (b) the assignment of tax powers, are the tax powers concurrent or separate? If the latter, what is the degree of tax or revenue centralization? The limitations on the taxing powers of the states with a view of safeguarding the interests of the national economy, evaluation of the system of tax assignment; (c) the degree of vertical fiscal imbalance (d) how fiscal imbalance and horizontal inequalities are sought to be tackled and narrowed; the system of grants-in-aid: types and principles Germany A.2 In the three federal countries (Canada, Germany, India) and one unitary state (Japan) considered here, the national constitutions are strongly biased in favor of the center as regards the distributioi, of power. The constitution of Germany -- the Basic Law -- has endowed the federal government with predominant power for both expenditures and taxation. The four distinctive characteristics of the German federal fiscal system are: centralization of taxing powers; tax sharing as the major way of resource transfer; a wide area of concurrent powers with the proviso that if the federal government enters a field, the states will cease to have power to take action in that area; and a mechanism for interstate transfers to effect equalization of fiscal capacity. A.3 The German constitution makes provision for concurrent federal and state legislation to cover fields not specifically assigned to the federal government, such as defense and foreign affairs. However, the states can legislate in a concurrent field only if the center does not do so. These provisions have enabled the federal government to occupy the most important tax fields: the income tax, the corporate profits tax, and the general sales tax (VAT) apart from customs and other minor levies such as those on alcohol. Thus, the federal government collects about 63 percent of total tax revenues collected by the federal, state, and local governments. The state governments' own taxes are few: those on property, beer, automobile use, and estates (the mainstays of the local governments are the business tax and the property tax). The financial autonomy of the state governments is ensured through a system of tax sharing, that is based 1/ J. S. H. Hunter, Federalism and Fiscal Balance (Australian National University Press, Canberra, 1977). - 39 - partly on a constitutional prescription and partly on an agreement between the federal government and the states.2 That is to say, the states are provided with sufficient resources to fulfill the functions assigned to them without the need for large grants from the federal government. A large vertical fiscal gap is created through the centralization of the levy of taxes, and it is virtually closed by tax sharing A.4 The proportions by which the shareable taxes are allocated among the three levels of government are shown in Table A. 1. Table A. 1: Germany: Allocation of Shareable Taxes Share (percent of total) Federal Local Tax Government States Governments Income tax 42.5 42.5 15.0 Corporate profits tax 50.0 50.0 - Value added tax 67.5 32.5 - Business tax 12.5 12.5 75.0 Other taxes 65.0 7.5 7.5 Source: Bird, Richard M. (1986): Federal Finance in Comparative Perspective, Canadian Tax Foundation, Toronto. Of these taxes, the income tax, the VAT, and the corporate profits tax account for 73 percent of all tax revenues. A. 5 Tax sharing, as noted earlier, is used primarily for closing the vertical fiscal gap, but it is also partially used for equalization. Whereas the states' share of income tax revenues is distributed among them primarily on the basis of the estimated source of the revenues, the states' share of VAT proceeds is distributed thus: (a) a 75 percent on an equal per capita base, and (b) as much of the remaining 25 percent as needed to be allocated in such a way as to bring the tax capacity of the financially weak states to 92 percent of the national average. In addition, the Federal Government gives a small percentage of its share of VAT (around 1.5 percent) to 2/ The provision for sharing the taxes on income is incorporated in the constitution, whereas that for sharing the VAT is the subject of an agreement. Since the terms of the agreement can be changed only with the consent of the Bundesrat (Council of States), wherein the states are powerful, the interests of the states are safeguarded; at the same time, a degree of flexibility is introduced by having an agreement outside the constitution. - 40 - the five weak states. The formula for distributing these supplementary transfers among these five states has been politically negotiated. A.6 Another interesting feature of the federal fiscal system in Germany is that there is no clear-cut division of expenditure responsibilities. True, all of the concurrent fields not formally occupied by the Federal Government go to the statesY However, the Federal Government sets policies in a large number of areas, although in most of these the administration of the policies is left to the states. Therefore, one finds few expenditure functions that are exclusive to only one level of government. As Table A.2 shows, every level of government is represented in most fields of activity listed; only defense, foreign affairs, and social security are almost entirely federal, and only sewerage and other community matters are largely local. There is, however, considerable expenditure decentralization: the share of the Federal Government in total expenditure is only around 43 percent, and that of the subnational governments is about 57 percent. Table A.2: Germany: Expenditures by Function and by Level of Government, 1974 As Percentage Share of expenditure by Function of Total Federal State Local General administration 6.2 20.0 38.4 41.3 Foreign affairs 2.0 99.4 0.6 - Defense 9.8 100.0 - - Security 3.0 9.3 66.3 24.4 Education 15.1 6.5 73.1 20.4 Social security 6.1 98.4 1.6 - Social assistance 6.5 23.7 24.3 51.9 Other social affairs 7.0 68.5 25.7 5.9 Hospitals 4.3 8.3 28.9 63.0 Other health 2.1 6.2 24.0 70.0 Housing 1.5 19.7 64.2 16.0 Sewerage 1.5 0.6 8.0 91.3 Other community 1.8 1.7 11.4 86.9 Roads 5.6 39.2 29.2 31.6 Agriculture 1.6 45.4 47.3 7.2 Other 2.3 50.2 41.0 8.8 Total 100.0 43.4 34.9 21.6 Note: May not add to totals owning to rounding. Source: Bird, Richard M., (1986); original source, Zimmerman, Horst (1981) 2/ In Germany, Canada and India, local government affairs are entirely the responsibility of the states/provinces. - 41 - A.7 Given the overlapping sharing of expenditure functions, the German federal system has developed the mechanism of "joint tasks" to coordinate and plan activities in several fields where the Federal Government wishes to influence policy. In these policy areas -- e.g., educational plans, university education and agricultural structural policy -- the federal and state governments interact in planning and spending.Y It has been agreed that "all administration would remain the responsibility of the states with only the overall framework planned as being a joint responsibility."i' Furthermore, as a rule, decisions in the joint planning committees require the agreement of a majority of state representatives and of the representative of the Federal Government. Germany has developed a unique type of "cooperative federalism" in the framework of which national policy is formulated and implemented in consultation with (and with the cooperation of) the state governments. A.8 The fourth distinctive characteristic of the German intergovernmental fiscal arrangements is the system of direct interstate transfers for equalization. The idea is to bring the per capita revenues of all the states at least up to 95 percent of average per capita revenues. For this purpose, the gap between "tax potential" and "tax need" is worked out. Tax potential is taken to be the actual per capita state tax revenue (including shared revenues) plus 50 percent of standardized local taxes in a state; tax need (a proxy for expenditure need) is the average per capita tax revenue of all states. Before working out the gap, an adjustment is carried out by multiplying by a weighing variable consisting of a population size element, a density element, and a city-state element. Positive and negative gaps (deficits and surpluses) emerge as percentages, which are again weighted and the amounts of interstate transfers worked out. The methodology may be subject to some criticism because actual tax revenues are used, as are several judgmental weights, but what is of interest is that Germany attempts to accomplish through horizontal transfers what other countries achieve through federal transfers. A.9 It remains to consider how conditions and provisions conducive to sound macroeconomic management have been ensured in the German federal fiscal system. Observers have pointed out that the German Federal Government exercises tight control over the economy. It controls monetary policy and has effective control over all important taxes. To strengthen further the powers of the government to take effective action to maintain macroeconomic equilibrium, the Economic Stability and Growth Law was enacted as far as back as 1967. This law opened up the way for the coordination of budgetary policies of the Federal Government and of the states. Under the Law, the Trade Cycle Council was set up. Although it is an advisory body, because of its composition its recommendations are rarely rejected. If the Council determines that the capital market is getting overloaded, it can recommend that the state governments and municipalities place limits on the amounts they plan to borrow. (Such measures must be approved by the Council of States before implementation.) The Trade Cycle 4i The Federal Government gives grants to support the financing of joint-task programs. 5/ Bird, Richard M. (1986), p. 89, Federal Finance in Comparative Perspective, Canadian Tax Foundation, Toronto. g/ Each level of government is represented on it. - 42 - Council can also advise on the reactivation of funds frozen with the Central Bank. The Federal Government can require (with the approval of the Council of States) that both the federal and state governments should increase the funds frozen with the Central Bank. Finally, the Law authorizes the Federal Government to alter the rates of income and corporate taxes by a maximum of 10 percent (in either direction) for a period of one year in order to tackle macroeconomic disequilibrium. A. 10 To conclude, intergovernmental fiscal arrangements in Germany give the Federal Government overall control over the economy as well as over policy planning in regard to public expenditures. Although centralization of the levy of the important taxes makes for an efficient tax system and places an important economic instrument in the hands of the center, the large tax- sharing arrangement strengthens the financial autonomy of the states without dependence on grants. The federal principle is respected and implemented through joint planning of policy and effective decentralization of the administration of states' functions, joint tasks, and even the collection of shared federal taxes, albeit under federal supervision. Canada A.11- Compared with Germany, Canada has much greater decentralization of the power to levy taxes, but about the same degree of decentralization of expenditures in terms of shares of the Federal Government and of the subnational governments. Hence, the relative size of the vertical fiscal gap is not as large as in other federal countries such as Australia, Germany and India. However, federal transfers do play an important role in the Canadian fiscal system, accounting for about 20 percent of provincial revenues and 16 percent of provincial and local revenues taken together. A. 12 Although the framers of the Canadian Constitution wanted a strong central government and hence endowed it not only with unlimited powers of taxation but also with the power to reserve or disallow provincial legislation, in practice the federal principle has been respected, and the provinces enjoy a good deal of autonomy and jealously guard their rights. The presence of French-speaking Quebec has, of course, meant that considerable accommodation has had to be shown by the Federal Government, but the other provinces have also striven to maintain financial autonomy and are rather vociferous in their complaints about the impact of federal policies on their budgets. At the same time, since there are large and small provinces and also financially strong and weak provinces, equalizing and compensatory action by the Federal Government has been called for and forms an important part of intergovernmental fiscal arrangements in Canada. A. 13 The Canadian Constitution has given the provinces significant responsibilities. They have exclusive right to 15 subjects enumerated in the Constitution, which, as in most other federal constitutions, cover social sectors like health and education as well as infrastructural sectors like transport and communications. (The provinces delegate some of their powers to the local governments.) The expenditure responsibilities of the Federal Government are also enumerated, and in addition, it also enjoys the residuary powers. There is no significant overlapping of expenditure responsibilities, but the Federal Government has chosen to support - 43 - and promote provincial expenditures in key social sectors through the giving of grants for specific purposes. A. 14 As noted in the main text, the Canadian provinces were given the right to levy direct taxes. Since the Federal Government was empowered to levy any tax, income taxation has been a subject for concurrent action. When the provinces had their right to levy a "direct" tax on consumers in the form of a retail sales tax upheld in court, the power to levy a sales tax also became a concurrent power, and both governmental levels were levying a sales tax. Now that the Federal Government has decided to levy the Goods and Services Tax, there will be clear overlapping in the indirect tax field. A. 15 Until the time of World War II, the Federal Government did not make much use of income taxes. When it started exploiting them in a big way, it entered into tax rental agreements with the provinces, by which the provinces received compensation payments in return for their voluntary surrender of tax powers. In 1957 there was a switch from the tax rental system to the system of tax sharing. According to this arrangement, the payments to the agreeing provinces were based on the yields in the provinces from "standard" rates of income tax (and also succession duties). Provinces that out of the system were offered abatements; i.e., their residents received abatements in their tax payments so that the province could tax them independently. From 1962 onward, all the provinces began to levy their own income taxes, and the Federal Government correspondingly withdrew from the field via increased abatements. The provinces had to accept the tax bases as defined in federal law, and the procedure was to express the provincial income tax as a percentage of the basic federal tax. The Federal Government offered to act as the provinces' collecting agent. All provinces except Quebec entered into collection agreements with respect to the personal income tax, and all provinces except Quebec and Ontario entered into collection agreements with respect to the corporation income tax. In time, the connection between the provincial tax rates and federal abatements was removed in respect of the personal income tax. As regards the corporate income tax, the Federal Government provides a tax credit of 10 percent of corporate taxable income earned in the provinces to make room for the provinces to levy their own taxes. A. 16 As of now, all the provinces have their own independent income taxes. The Federal Government collects the tax on behalf of the provinces if the tax bases of the provincial taxes are defined in the same way as the federal income tax base. Quebec, Alberta, and Ontario administer their own corporate income taxes, and Quebec alone administers its own personal income tax. With for these exceptions, the provincial income taxes are collected by the Federal Government. A. 17 It is clear that, under these arrangements, it is not possible to have a uniform or even rational income tax system. Each province has its own system of tax credits and incentive provisions. There is acute tax competition, with the intention of luring investors into one's own province. Where taxes are administered independently by the provinces, the assessees have to submit returns to two authorities and have their taxes on more or less the same bases assessed twice. The provinces do have some additional maneuverability because they can levy their own taxes, but given the decision of the Federal Government not to offer any further reduction in the -44- rates of its own taxes, the provincial tax rates have to be kept low. The major part of the income tax revenue in any case is raised through the federal tax. A. 18 The Canadian system of sharing the income tax base (in respect of the majority of the provinces) comes somewhere between the operation of concurrent tax powers and tax yield sharing on the basis of collection. The element of autonomy varies among the systems, but in all three cases a province gets its revenue from the collections within its jurisdiction. Therefore, the richer provinces gain proportionately more. A. 19 The handicap that the poorer provinces have in raising revenues (because of the lower value of tax bases) is redressed to some extent by the system of equalization grants, which are unconditionalY These are intended to make up for any deficiency in taxable capacity -- to ensure "that all provinces are able to provide their citizens a reasonably comparable level of basic services without resorting to unduly burdensome taxation."' These are enabling grants; the provinces may not use them to increase the provision of the intended basic services. The system of equalization payments involves bringing the revenues of a province from 33 different sources up to the level of the representative average standard. This representative average standard is based on the average per capita revenue from each source in five "representative provinces" -- Quebec, Ontario, Manitoba, Saskatchewan, and British Columbia. A national average tax rate for each revenue source is calculated by dividing the revenue collected from that source in all provinces by the value of the relevant tax base for all provinces. For each province, the per capita yield for a revenue source is equal to the product of the national average tax rate applied to that province's tax base divided by its population.2' The per capita deficiency in taxable capacity is the sum of the differences between the average per capita revenue of the different sources in the "standard" provinces (i.e., the representative average standards) and the per capita yield of the revenue sources in that province. Positive per capita deficiency (if any) multiplied by the population of a province gives that province's entitlement to equalization payment. However, equalization payments are subject to an overriding limit on total federal equalization payments. A.20 The federal-provincial fiscal arrangements also provide for what are called stabilization payments. If the total estimated revenue of a province in any one year at the previous year's tax rates falls below the revenue of the previous year, the Federal Government makes a stabilization payment to make up for the shortfall. However, there is a maximum limit on per capita stabilization payments; any excess of entitlement over the maximum is given in the form of a loan. 7/ These grants have been established constitutionally. I/ Federal Provincial Tax Structure Committee (1966); quoted in R. A. Musgrave, Peggy Musgrave, and Richard M. Bird, Public Finance in Theory and Practice, First Canadian edition (New York: McGraw-Hill Ryerson Ltd., 1991), p. 510. 9/ Canadian Tax Foundation, Provincial and Municipal Finances. 1989 (Toronto, 1990), p. 13:6. - 45 - A.21 Far more important, in quantitative terms, than equalization grants are the specific- purpose grants. These now largely consist of the Established Program Financing (EPF) arrangements and the Canada Assistance Plan. The EPF has evolved from the early post-war period, when the Federal Government began to use its increased revenues to establish several grant programs in education and health. In 1977 the long-standing specific-purpose transfers, which were in the nature of matching grants with an equalization element and were given to finance three "established programs" (hospital insurance, medicare, and post-secondary education) were combined and transformed into EPF arrangements, which abolished the detailed federal supervision of these provincial spending programs. The connection with the cost of the programs through the matching provision was also severed. These changes were introduced because, on the one hand, the provinces complained about having to submit their health and education expenditure to federal scrutiny, since health and education were clearly areas of provincial responsibility; on the other hand, the Federal Government did not find it convenient to have a good part of its expenditure being determined by the volume of provincial expenditure in the concerned fields. The agreement was that per capita EPF transfers to a province would increase each year in accordance with per capita GNP growth over the 1975-76 base. Although the link with the amounts spent by the provinces has been removed and federal control and accountability have been minimized, there are still contentious issues regarding minimum standards and accountability. The transfers have not become completely unconditional. A.22 The other major component of specific transfers is the Canada Assistance Plan, under which the Federal Government pays 50 percent of the costs of provincial expenditure on a wide array of social welfare programs. In addition, there are some minor grant programs in areas such as resource conservation, transportation and communication, and industrial development. Together, the total of specific cash grants far exceed the general-purpose grants. A.23 From the above description it should be clear that federal-provincial fiscal relations in Canada are not conducted largely on the basis of fixed provisions in the constitution. The fiscal arrangements have evolved over time and are the subject of recurrent negotiations. While this procedure imparts a degree of flexibility, the absence of finality in almost every aspect generates pressures for changes. The budget of the Federal Government and those of the provinces are linked through the sharing of the income tax base and through the EPF and other specific grants. Changes in the terms of these arrangements or unilateral action could seriously affect the budgetary position of one party or the other. Recently, the Western Finance Ministers' Reporti-0' (prepared at the behest of the Western Premiers) has voiced serious complaints: Adequacy and affordability have been considered only in relation to the federal need to reduce expenditure growth - the dollar commitments provinces have to make to maintain costly social programs have been ignored. Flexibility, efficiency, stability and predictability, (f considered at all, have been considered in federal terms. For the provinces, fiscal relations through the decade have been marred by inflexibility, inefficiency, instability and unpredictability. Accountability n0/ Report of the Western Finance Ministers, Economic and Fiscal Development and Federal: Provincial Fiscal Relations in Canada (Lloydminster, Saskatchewan, 1990), p. 20. - 46 - to the public has been destroyed in a "numbers war' between the federal and provincial governments. Lower fiscal capacity and the equalization ceiling make equalization-recipient provinces especiaUy vulnerable to cuts in other federal programs and transfers which shift responsibilities to the provinces. In the end, the fundamental partnership among the governments in Canada has been severely damaged. A.24 These feelings have been aroused through the experience of budgetary difficulties. If there is a large or rising federal budgetary deficit, part of the burden is felt by the provinces. They have complained that the federal control over the income tax system has been exercised with less than appropriate flexibility for the provinces. The introduction of the Goods and Services Tax has been held to be an invasion of a tax field traditionally occupied by the provinces. Other major complaints are: (a) the Federal Government has been gradually lowering the limit on the rate of growth of EPF transfers; (b) similarly, the ceiling on the equalization payments as a percentage of GNP was lowered; and (c) the Federal Government has demanded that the provinces increase their share of financing cost-sharing programs such as crop insurance and agricultural assistance. A.25 It would appear that federal action to contain its budgetary deficit has its fallout on the provincial budgets. The intergovernmental fiscal arrangements as they have evolved do not have the mechanism to mediate a situation of macroeconomic disequilibrium with adequate sharing of costs. A.26 To conclude, although the Canadian Constitution is heavily biased toward the Federal Government, in practice intergovernmental relations, as they evolved, have been built on the federal principle. The Federal Government, in response to the wishes of the general population, has respected that principle and has permitted provincial autonomy to grow. However, it would seem that informal arrangements and a high degree of flexibility have led to constantly shifting relations and conflict. The concurrent income tax powers have not only led to dissatisfaction, but also to complexities and difficulties in building a rational tax system. A satisfactory solution has yet to be found. The introduction of the Goods and Services Tax at the federal level will call for tax harmonization in the indirect tax field. Finally, despite concurrent tax powers in the income tax and sales tax fields, federal transfers form a substantial proportion of provincial revenues because of specific-purpose grants to promote and support expenditures essentially in the fields assigned to the provinces. Thus in Canada, as in several other federations, although for good reasons certain expenditure responsibilities have been assigned to the subnational governments, the Federal Government has considered it necessary or desirable to enter those areas. - 47 - Ia A.27 The Constitution of the Republic of India can be described as quasi-federal in character because it provides for a federal structure with strong unitary features.A The states do enjoy a substantial degree of autonomy within the area of responsibility granted to them by the constitution. However, this freedom of action is circumscribed by several provisions intended to ensure either that the national interest and fundamental rights will not be jeopardized or that the national viewpoint will prevail in important matters. A.28 As in the cases of Canada and Germany, local government affairs are entirely within the states' sphere. In fact, the local governments do not have a constitutional status. A.29 The constitution assigns to the states several major responsibilities relating not merely to the basic functions of government, such as the maintenance of law and order and courts, but also to widely ranging crucial functions in the social sectors, agriculture, infrastructure (including water management), trade within the state, and overall development of the state economy. The constitution provides three lists of subjects: the Union list, the states list, and the concurrent list. Briefly, all matters relating to defense, currency, banking, foreign affairs (including economic relations with foreign entities), matters affecting the country as a whole, and those relating to interstate relations are placed in the exclusive domain of the Union Government (the center). Parliament can decide by law which industries should fall within the sphere of the center. As pointed out above, matters closely connected with the life and welfare of the people and several economic subjects -- particularly those relating to land, water, and agriculture -- are placed on the states' list. A.30 A large number of subjects, however, are placed in the concurrent list. These include important civil matters such as law, marriage, succession, administration of justice, trusts and civil procedure, and several economic subjects such as economic and social planning, social security, trade unions, and electricity. Education, which was a state subject in the original constitution, was made a concurrent subject through a constitutional amendment. Thus, although there is a good deal of expenditure decentralization, the device of the concurrent list ensures that the center will prevail in a wide area, if such a result is found necessary. If there is a central law in an area, it will prevail over any state law in that area, unless the center makes an exception. A.31 Economic planning in practice, with the socialist bias toward public ownership, has considerably strengthened the center vis-a-vis the states. To the control acquired for purposes of planning (such as power to license industries, exchange control, import control, and approval of foreign collaboration) has been added the power of nationalization. Through nationalization, 11/ This section is based largely on R.J. Chelliah, "Intergovernmental Fiscal Relations and Macroeconomic Management in India" (paper submitted to Senior Policy Seminar on Inter- Governmental Fiscal Relations and Macro-Economic Management in Large Countries (EDI) in New Delhi in 1991); (mimeo; World Bank, 1990). - 48 - the center has gained direct control over the entire financial system because it owns the most important financial institutions. Since the Reserve Bank of India, which is justifiably placed on the Union list, has also been nationalized, the center has overwhelming control over investible resources in the economy. A.32 The assignment of tax powers in the Indian Constitution seems to have been based on two fundamental principles. The first is that the assignment should avoid giving concurrent tax powers to the center and the states: any one tax should be assigned to only one level of government. The second is that the most important taxes, which have economy-wide implications or which can be collected most efficiently and economically only by the Central Government, should be assigned to the center. Using the second criterion, the center was given the power to levy income tax on nonagricultural income (individual as well as corporate), all excise taxes on production except those on alcoholic liquor, and custom duties. The center thus had within its jurisdiction the most productive sources of revenue with wide bases. The states have been assigned tax revenues related to land and agriculture (land revenue, agricultural income tax, and other duties on agricultural land), sales taxes except sales tax on interstate trade, excise on alcoholic liquor, taxation of inland transport except railways, property tax, and the entry tax (the last two have generally been delegated to local authorities). Other taxes, such as the tax on railway fares and estate and succession duties on nonagricultural property, were placed on the Union List, but their entire proceeds, if levied, were to be assigned to the states. A.33 The basic features, strengths, and weaknesses of tax assignment under the Indian Constitution may be summarized as follows: (a) Taxes that have a nationwide base and/or that can be most economically and efficiently collected by the central authority and those that could lead to misallocation of resources if left to the states were assigned to the center. Under these categories come income taxes and the excise taxes on production. For understandable reasons, foreign trade taxes were also given to the center. (b) Although the power to tax production was given to the center (except for the tax on alcoholic liquor), sales taxation, which is essentially a consumption tax, was given to the states. The production tax is levied on the basis of origin, a consumption tax is levied on the principle of destination. The constitution makers seem to have based the assignment on the intention that the states could levy the sales tax largely as a tax on the consumption of their respective residents. Certain other taxes on consumption, under the category of luxury taxes (taxes on entertainment, amusements, etc.), were also assigned to the states. (c) Taxes with local bases were assigned to the states. (d) To augment the revenues of the states, the proceeds of some taxes that for efficiency and administrative reasons could only be levied and collected by the center were assigned to the states. - 49 - (e) Although the intention of the constitution seems to have been to avoid overlapping tax powers, the division of tax powers contained did not succeed in avoiding economic overlapping. As the Central Government expanded the coverage of Union excise duties to make them applicable to almost all manufactured products, the excises became more and more a manufacturers' sales tax. The Union excises interacted with the states' sales taxes, with the excise tax paid forming part of the sales tax base and the sales tax paid in turn becoming part of the excise tax base in subsequent stages of production.12' (f) Because adequate restrictions were not placed on the power of states to levy sales taxes, different forms of sales taxation evolved in the states. The sales tax became in part a tax on production and enabled different states to export their taxes to residents of other states. (g) There was no provision in the constitutional assignment of tax powers to ensure that indirect taxes levied by the center and the states would not lead to cascading and distortions in production (through unintended changes in relative factor prices). (h) It has generally been argued by economists that taxes that can and are to be levied on a progressive basis should be assigned to the Central Government.!" This rule was violated in India insofar as the power to levy taxes on income and capital was split between the center and the states on the basis of nonagricultural and agricultural origin. This has led to distortions and avoidance in respect of the nonagricultural income tax and wealth tax levied by the center, as well as to serious violation of the principle of horizontal equity. It has also been argued that taxation of natural resources could best be left with the Central Government. The Indian Constitution gave the power to levy taxes on mineral rights to the states, "subject to any limitations imposed by Parliament by law relating to mineral development." The arrangement flowing from this provision has led to considerable friction between the center and the states, with the states resenting the center's reluctance to raise royalties on mineral rights, in the interest of safeguarding the national economy. (i) Because some taxes, such as those on electricity and road transport were assigned to the states without proper safeguards or restrictions, the states began to tax inputs, including transport of commodities. This has resulted in unjustified and unintended increases in the costs of production. Avoidance of tax exportion by individual states was not an important criterion in the constitutional scheme. 12/ Chelliah, R.J. (1980); Government of India, Final Report of the Indirect Taxation Enquiry Committee (New Delhi, 1978). 1J/ Ra. A. Musgrave, "Who Should Tax, Where and What?", ed. Charles E. McLure, [ax Assignment in Federal Countries (Australian University Press, Canberra, 1983). - 50 - A.34 With the most important and productive taxes having been allotted to the center, the constitutional arrangement for the decentralized provision of important public services by the states would naturally lead to vertical fiscal imbalance. Anticipating such an imbalance, the constitution provided for the sharing of two of the important central taxes with the states. Provision was also made for giving grants-in-aid to states that would be in need of assistance. Elaborate arrangements were stipulated for effecting intergovernmental transfers from the center to the states. A.35 The constitution provides that taxes on the nonagricultural income of noncorporate entities shall be distributed between the center and the states. In this case, the sharing is mandatory. The constitution also provides that the yield of Union excise duties may be shared with the states if Parliament so provides. Parliament has also been authorized to provide for grants-in-aid to such states as it may determine to be in need of assistance. A.36 In view of the above provisions, questions arise about how the shares of the states in the proceeds of the designated divisible taxes should be determined and how the divisible pool should be distributed among the states. The amounts of general grants-in-aid of revenues to be given to the different states have also to be determined. The constitutional provision is that the President, in respect of income tax, and the Parliament, in respect of Union excise duty and grants-in-aid, shall be the final authorities to take decisions on the respective matters. However, in order to assure the states that these important decisions concerning transfers of funds would be taken on a fair and impartial basis, the constitution provides that the President should appoint a Finance Commission, which would make recommendations to the President on (a) the distribution between the center and the states of proceeds of taxes that are to be, or may be, divided between them; (b) the principles that should govern the grants-in-aid of revenues to the states and (c) any other matters that may be referred to the Commission by the President in the interests of sound finance. A.37 The institution of a Finance Commission with duties as mentioned above was an innovation in the ordering of intergovernmental relations in modern federations. By introducing this, the constitution imparted considerable flexibility in the arrangements regarding the distribution of resources between the center and the states. Also, the interposition of an independent body like the Finance Commission has ensured that the executive branch of the Central Government will not take decisions on matters affecting the distribution of revenues purely at its discretion but is required to act on the recommendations of the Finance Commission and report on such decisions to Parliament. At the same time, since the Finance Commission is not authorized to take final decisions but only to make recommendations, the President (that is, the Central Government) can amend or reject any part of the recommendations if the national interest so requires. In the course of time, it has become the convention for the Central Government to accept all or most of the Finance Commission's major recommendations. A.38 In the view of some observers, the establishment of the Finance Commission under the constitution is not an unmixed blessing. The constitution has not provided guidelines to the Commission about how it should proceed in respect of its duties either in relation to distribution of taxes or in relation to grants-in-aid. Although the Finance Commission has striven to act impartially and to keep the national interest in mind in making its recommendations, the - 51 - approaches of the Commission have varied over time and have not led to the evolution of a set of principles firmly grounded on economic analysis or empirical research. A.39 The general approach of the Finance Commission, from the Second through the Eighth Commissions, has been to project, for the next five years, revenues and nonplan expenditures on the revenue account of the states, starting with the base year figures, in order to work out the surplus or deficit of the revenue account for those years. The projections were carried out on "reasonable assumptions," with past trends and expected growth rates of the economy kept in view. The total volume of transfer of resources through tax sharing (devolution) was presumably determined on the basis of leaving only a few states with deficits after devolution. Deficit-filling grants were given to those few states. Devolution became the main vehicle of central transfers. This inevitably meant that some states were left with projected large surpluses on the nonplan revenue account areas, while others were left with zero deficits. A.40 The formula for distribution of tax shares adopted by the Eighth Commission gave very little weight to collection. A weight of 10 percent, only in respect of the income tax share, was given to collection. Otherwise, equal weights were given to population and to the (inverse) of per capita income. The Ninth Commission (the most recent one) gave some weight to the number of disadvantaged and poor groups and correspondingly reduced the weight given to per capita income. In addition to the per capita income weight, population is used as a scale factor, and hence the total (direct and indirect) weight given to population is very high (around 80 percent). A.41 As stated earlier, the quinquennial review by a Finance Commission provided for in the constitution has served to introduce flexibility in the arrangement for the division of financial resources between the center and the states. However, the reopening of the pattern of distribution every five years has led to expectations of larger transfers and increasing pressures on the Finance Commission to recommend devolution of tax revenues from all states, rich and poor alike. It could be argued that the expenditure needs of the state governments, as assessed in the light of their constitutional responsibilities, have proved to be much greater than what could have been supported on the basis of the percentage of shareable taxes going to the states at the time of the appointment of the First Finance Commission and that, therefore, the later Commissions had to make the needed upward adjustments. However, although the states' share of central taxes had been raised from 50 percent of income tax only in 1950-51 to 85 percent of income tax and 45 percent of Union excise duties in the period 1985-86 to 1989-90, the financial position of most of the states was unsound, and there were continuing demands for further increases in the percentage shares of the states. Central finances began to be destabilized (to be sure, there were other causes, too) with the states' finances being unsound. With successive Finance Commissions raising the volume of transfers substantially to fill the deficit gaps in the normal revenue budgets of the states, the center had to raise its own tax revenues or borrow more to finance increased expenditures by the states, whose respective residents might not have (probably would not have) voted for such increased expenditures had they been told that - 52 - they would bear the concomitant tax burden.!'In other words, the nexus between expenditure decisions and the related raising of tax revenues was seriously weakened. It would be generally agreed that, in the interests of efficiency and economy (fiscal discipline), such a link must be preserved, at least at the margin, in a federal financial system. A.42 The Indian Constitution had envisaged that grants-in-aid of revenues to the states would mainly flow on the basis of the recommendations of the Finance Commission (Article 275), although another article (Article 282) had stipulated that "The Union or a State may make grants for any purpose." With the establishment of the Planning Commission, grants and loans began to be given by the center to the states for helping them to finance their (public sector) plans. The Planning Commission, which functioned directly under the Prime Minister (as its Chairman) and was kept somewhat outside the regular governmental structure, coordinated the plans of the center and of the states and produced a national development plan. The plans of the states were "approved" by the Planning Commission, and the plan transfers to the states were made through (or on the recommendations of) the Planning Commission. In the course of time, the transfers to the major states became formula-based, but the Planning Commission still retains some discretion in the matter particularly in respect of the allocation of plan funds to the hill states (called the Special Category States). The plan transfers consist of grants and loans, and the grants are in the nature of block grants. In addition, conditional grants are given for the implementation of nationally important development or welfare programs under the rubric of central sector and centrally-sponsored schemes. A.43 The total volume of plan assistance to be given to the states for the five-year and the annual plan is decided jointly by the Ministry of Finance and the Planning Commission. Of the total amount, a certain portion is pre-empted -- currently about 30 percent -- and made available to the Special Category States. The rest of the assistance is apportioned among the other states on the basis of a formula agreed upon with them. The formula has been changed from time to time. For the Eighth Plan, now being implemented, it has been proposed that the allocation will be based on the following weights: population, 55 percent; lower level of per capita income, 25 percent; better fiscal management, 5 percent; and special problems, 15 percent. Assistance is provided in the proportion of 70 percent loan and 30 percent grant for the major states; for the Special Category States the proportion is 90 percent grant and 10 percent loan. A.44 A certain portion of the central plan transfers is earmarked for particular sectors. However, since the allocation is basically left to the states and earmarking is done after the allocation is decided upon, the grants and loans involved are not conditional in the sense of being 14/ This is analogous to the so-called "flypaper effect," whereby general- purpose grants seem to have stimulated increased expenditures on the part of states governments in the United States which have been explained as being due to misperception by the residents of the respective states. For a discussion of the flypaper effect, see Mieszkowski and Oakland, Fial Federalism and Grants-in-Aid (The Urban Institute, Washington, D.C., 1979). - 53 - intended as stimulants for expenditures. Rather, earmarking is an attempt to prevent the diversion of expenditure to unplanned activities-15' A.45 During the 1980s the Central Government became attuned to the problems created by the "gap-filling" approach of the Finance Commission and the additional debt burdens caused by the inability of the state governments to repay the loans they had taken from the center. It was also felt that the various Finance Commissions had not made sufficient allowances for differences in revenue capacity and in the pursuit of economy in expenditure. In accordance with this perception, the terms of reference given to the Ninth Finance Commission (1987-89) were formulated on lines quite different from the traditional approach. The terms of reference required the Ninth Commission to adopt a normative approach to the assessment of revenues and expenditures of the states and the center. On capital account side, the Ninth Commission was not asked, as many of the previous Commissions had been to suggest measures to cover the "gaps" created by the debt liabilities of the states; instead, the Ninth Commission was asked to review the debt position of the states and to suggest corrective measures.!' The Ninth Commission itself was gravely troubled by the alarming deficits and the growing indebtedness of the center and the states; the approach it applied estimated a state's relative revenue capacity and expenditure needs (in terms of public services) on the basis of norms. The normative approach did not reward relative expenditure extravagance, but neither did it stand in the way of a state's raising more resources if it wanted to do so. On the capital side, the Ninth Commission chose to make the terms of central loans to the states more reasonable rather than to make recommendations for any substantial rescheduling or write-offs of existing debt.!-' A.46 Much remains to be done in developing a system of central revenue transfers that would be conducive to the efficient and economical use of resources, that would satisfy generally accepted norms of interstate equity, and that would not militate against the maintenance of macroeconomic equilibrium. The flow of grants outside the recommendations of the Finance Commission gives rise to the problem of coordination of different flows. The overlapping of the functions of the Finance and Planning Commissions gives rise to problems that have yet to be fully sorted out. Again, that the Planning Commission continues to approve increases in expenditure within a five-year horizon, even if the "planned" increases in revenue expenditure lead to revenue deficits (or increases in deficits), will continue as a source of disequilibrium in the budgets. A.47 Table A.3 presents the trends in vertical fiscal imbalance in India. The share of the states in revenue expenditure is around 56 percent. Their share in combined total expenditure (on the revenue and capital accounts) has been somewhat lower than that in revenue expenditure; .1/ For an appraisal of the system of transfers, see M.G. Rao and R.J. Chelliah, Surygy o Research on Fiscal Federalism in India (Processed), (National Institute of Public Finance and Policy, New Delhi, 1990). j&/ M. G. Rao and R. J. Chelliah, Survey of Research on Fiscal Federalism in India (Processed), (National Institute of Public Finance and Policy, New Delhi, 1990). 17/ Ibid. - 54 - in recent years it has hovered around 51-53 percent of the total. The second column in the table shows the ratio of states' own revenues to the states' expenditure (revenue account): this share has fallen from around 63 percent in the early 1960s to around 55 percent in the late 1980s. Similarly, the ratio of states' own total receipts (revenue and capital accounts) to states' own total expenditure has fallen gradually (though not steadily) from 58-60 percent in the 1960s and 1970s to 50-55 percent in the 1980s. Currently, states' own receipts finance around 50 percent of their expenditures. Table A.3: INDIA: Trends in Vertical Fiscal Imbalance (in Percent of total) Ratio of states' Ratio of states' expenditure own revenue to Ratio of states' Year to total combined states' expendi- own receipts expenditure ture (revenue (revenue and capital (revenue and account) accounts) to state's capital accounts) total expenditure (1) (2) (3) 1960-61 56.76 63.86 57.57 1965-66 53.33 63.46 54.97 1970-71 53.87 60.57 58.24 1975-76 47.55 70.39 60.29 1980-81 55.97 60.07 51.39 1982-83 54.57 64.26 52.86 1983-84 52.78 62.13 53.66 1984-85 51.82 56.47 53.74 1985-86 52.63 57.69 54.10 1986-87 50.23 56.40 53.04 1987-88 53.02 55.08 55.09 1988-89 51.98 54.05 47.55 1989-90 53.06 53.91 50.25 Source: Government of India, Ministry of Finance (Annual) Indian Economic statistics, New Delhi A.48 As may be expected given the constitutional division of tax powers, there is a high degree of centralization of tax collection: the Central Government raises around 67 percent of total tax revenues. Through tax sharing (devolution), the share of the states in the accral of tax revenues is raised to around 50 percent. A.49 Intergovernmental fiscal arrangements in India cannot be said to obstruct sound macroeconomic management. It is true, as pointed out earlier, that the readiness of the early Finance Commissions to fill in the (modified) deficits in the state budgets promoted expenditure extravagance on the part of the states and also led to a severe deterioration in the budgetary balance of the Central Government. The basic constitutional provisions, however, do not - 55 - militate against the power of the center to pursue sound macroeconomic policies. The center has exclusive control over all matters that pertain to monetary stability and balance of payments equilibrium. Currency, coinage, and banking are central matters, and only the center can contract foreign loans. The states can borrow domestically, but they need the center's permission if they are indebted to it. Because all of the states are (mainly through obtaining plan loans), their market borrowings are allocated as part of the pattern of plan financing by the Planning Commission. Since 1985, the states have not been allowed to run up overdrafts with the Reserve Bank of India beyond the prescribed ways -- and -- means limits. Thus the states cannot have fiscal deficits in excess of what has been permitted, except for their enterprises owing money to suppliers. It cannot be denied, however, that the states significantly contribute to overall fiscal imbalance by letting their major enterprises incur huge and continuing losses. A.50 To conclude, there is considerable expenditure decentralization in the Indian quasi- federation, coupled with centralization of collection of tax revenues (the share of the center being 67 percent of the total). The resulting, relatively large, vertical fiscal imbalance is closed by tax sharing and grants. Until recently, the design of transfers has not been conducive to the maintenance of fiscal discipline at the state level. The central transfers are equalizing to the extent that population is given the highest weight in the formula for the distribution of transfers among the states. However, the greater relative importance of tax sharing in total current transfers to the states leaves less room for equalizing grants targeted toward the financially weaker states. The very low weight given to collection in the distribution of the divisible pool of income tax revenues indicates the relative negligence of rewarding better economic performance as well as basing "entitlement" on contribution. The concentration of economic power in the hands of the center taken together with central control over the magnitude of the fiscal deficit of the public sector, gives the Central Government the necessary tools to implement policies for macroeconomic management. However, the center's unlimited access to Reserve Bank of India credit and a rather unsound design of transfers (including tax sharing) linked to deficit filling have been factors leading to macroeconomic disequilibrium. Japan A.51 Japan is the only unitary country studied hereAl Japanese local governments operate on the constitutional principle of local autonomy. They have a wide range of responsibilities and account for about 62 percent of total public expenditure in Japan. In fiscal terms, however, the Central Government levies important taxes -- such as the income tax, the corporation tax, the consumption tax (VAT), etc. -- and collects more than 64 percent of total tax revenue. More than 45 percent of this amount is then transferred to local governments, chiefly through an unconditional, formula-based transfer called the local allocation tax (about 52 percent of the total transfer amount) and through specific-purpose grants (about 40 percent of the total transfer). Another feature of the Japanese intergovernmental fiscal arrangements is the institutional protection accorded local governments within the central bureaucracy. The 18/ This section is based largely on T. Fujiwara, "Intergovernmental Fiscal Relations in Japan" (mimeo; World Bank, 1991) forthcoming. - 56 - Ministry of Autonomy is an advocate of local governments within the Central Government and is a built-in counterbalance to central power. Assignment of Tax Powers A.52 Japan has a two-tier system of local government: at the prefectural level and the municipal level. Japan is divided into 47 prefectures, which are then divided into municipal entities. Each level of government -- Central Government, prefectural government, and municipal-level government -- has its own tax power. Tax powers are clearly separated for each level of government. A.53 At the local level, each government levies and collects taxes according to by-laws passed by its local assembly in accordance with the Local Tax Law. The Local Tax Law provides the framework for local governments to act as local tax authorities through these by- laws. In other words, it lays down the local tax arrangement in line with intergovernmental fiscal considerations. A.54 The Local Tax Law stipulates the specific taxes that each level of local government can levy. In many cases, standard rates are given at which local governments are expected to levy the taxes. In some cases, ceiling tax rates are also given. Rates are fixed for taxes such as prefectural and municipal tobacco taxes and a light-oil delivery tax, which assume the smooth circulation of the goods nationwide. Local governments can set up their own new taxes with the Autonomy Minister's permission, which is given unless the proposed tax is designed irrationally in respect of the overall tax system and national economic policy. A.55 According to the Local Tax Law, prefectural governments have the following nine ordinary taxes (i.e., taxes on sources of general revenue): prefectural inhabitants tax, enterprise tax, real property acquisition tax, prefectural tobacco tax, golf course tax, special local consumption tax, automobile tax, mine-lot tax, and hunter registration tax. In addition to these ordinary taxes, prefectures levy an automobile acquisition tax earmarked for roads, a light-oil delivery tax earmarked for roads, and a hunting tax earmarked for preservation of birds and beasts and for expenditure on hunting administration. Major prefectural tax revenues are as shown in Table A.4. The table suggests that prefectural tax bases are linked mainly to corporate and individual income and economic activities and not to property. - 57 - Table A.4: Japan: Maior Prefectural Tax Revenues (FY19891 Amount Share Taxes (billion yen) (percent) Enterprise tax 6,548 44.4 Prefectural inhabitants tax 4,337 29.4 Automobile tax 1,196 8.1 Light-oil delivery tax 766 5.2 Real property acquisition tax 631 4.3 Automobile acquisition tax 578 3.9 Prefectural tobacco tax 318 2.2 A.56 Municipalities have the following five ordinary taxes: municipal inhabitants tax, fixed assets tax, light motor vehicle tax, municipal tobacco tax, mineral product tax, and special landholding tax. In addition to these ordinary taxes, cities in highly urbanized areas (such as the greater Tokyo and greater Osaka areas) and cities with populations over 300,000 levy a business office tax, which is earmarked for improving the urban environment. Municipalities where there are mineral springs levy a spa tax that is earmarked for the preservation of the mineral spring sources, promotion of sanitation facilities, fire protection facilities, and tourism. If they need them, municipalities can also levy a city planning tax and some other taxes. Major municipal tax revenues are as follows: Taxes Anounl Shr (billion yen) (percent) Municipal inhabitants tax 9,275 54.4 Fixed assets tax 5,688 33.4 City planning tax 904 5.3 Municipal tobacco tax 565 3.3 Thus, municipal tax bases combine elements of individual and corporate income, property, and consumption taxes. Vertical Fiscal Imbalance A.57 As regards vertical fiscal imbalance, the allocation of Japanese intergovernmental tax revenue is heavily shifted toward the center. The Central Government keeps hold of major taxes such as the income tax, the corporation tax, the consumption tax (VAT), the liquor tax, etc. - 58 - A.58 Here, the vertical imbalance will be measured in terms of actual transfers from the center to local governments. In 1989, the Central Government spent 170,851 billion, of which 126,202 billion (38 percent) was transferred to local governments in one form or another. Thus, the national government spent only 144,649 billion (38 percent) of total government spending to cover its own policy needs. A.59 On the other side, local governments spent X72,729 billion in 1989, where 11,072 billion was paid to the Central Government as burden sharing. Thus, local governments spent X71,657 billion (62 percent of total government spending) to cover local policy needs. This means that, although local governments spent 62 percent of total government spending, 37 percent of their spending came from the central government transfers. A.60 As regards central and local tax collection in 1989, the Central Government collected more than 64 percent of the total tax revenue, whereas local governments collected less than 36 percent. This ratio is almost the reverse of the one for central and local direct spending. Assignment of Functional Responsibilities A.61 The Local Autonomy Law gives a general idea about the division of functional responsibilities between central and subnational governments in Japan. The idea is that local governments perform a wide range of administrative and community affairs necessary for the local public's interest, unless otherwise provided for in national laws or in Cabinet orders duly authorized by laws. That is, local authorities can perform any role necessary to solve local public problem unless explicitly denied by law. Although diplomatic negotiations with foreign countries, monetary system control, national defense, the postal service, judicial administration, etc. are reserved for the Central Government, other fields are in principle open to local governments. Thus, generally speaking, Japanese local governments can potentially handle a very wide variety of local policy matters, and this explains the high local expenditure share mentioned in the preceding paragraphs. Local authorities' main restrictions in relation to the center come from laws regulating local governments' activities, central subsidies earmarked for specific services with centrally set conditions, the amount of local finances at their discretion, and both local and national political support for what local governments want to do. A.62 As regards the relations between municipal-level and prefectural government, the Local Autonomy Law stipulates municipalities as the basic local public bodies. Municipalities are assumed to have responsibilities for local policy and administrative affairs unless these are designated as more suitable for prefectural governments. In contrast, prefectural governments are assumed to be comprehensive local public bodies comprising municipalities. Prefectural governments are expected to cover those affairs which cover a wider area, or require uniformity in performance, or relate to the liaison and coordination of municipalities, or exceed the scale of work that ordinary municipalities can be expected to handle efficiently. A.63 These are just general guidelines. The Local Autonomy Law also states that prefectural governments and municipalities must pay attention that their services do not overlap. This clause is in line with the idea of efficiency as one of the principles underlying Japan's local self-government system. In practice, there are many individual laws that stipulate the level of - 59 - local government to which particular competence and responsibility belong. These laws solve most of the possible overlaps of functional responsibilities. A.64 Generally speaking, municipal-level governments are responsible for primary and junior high school education; regular solid waste collection and disposal; water supply and sewerage; local environment improvement with regard to noise, vibration, and odor; city planning projects of local interest; fire defense, etc. Big cities are responsible for health services and ports. A.65 In general, prefectural governments are responsible for high school education; industrial waste disposal; construction and maintenance of regional trunk sewerage pipes; health services, anti-pollution measures and improvement of the environment; large-scale or basic city planning projects; police; etc. A.66 Each level of government is responsible for roads, rivers, and housing according to the types and purposes of the particular program. A.67 There are other examples of coordination between governments. If a capital city's government has built a museum, for example, the prefecture encompassing the city would not plan a similar museum, or it would plan one at a location not close to the city museum. This kind of coordination works well in Japan because it is backed by the traditional value placed on social harmony. Assignment of Expenditure Responsibilities A.68 Strict rules for intergovernmental expenditure responsibilities are also a salient feature of Japanese intergovernmental fiscal relations. These rules were originally designed to protect local finance from bureaucratic pressure from the center to engage in unreasonable burden sharing. The Local Finance Law sets forth these rules; for example: (a) Local governments bear all the costs of handling their own responsibilities. (b) National government bears a share of local governmental responsibilities in which the national government also has certain interest (such as education, livelihood protection, public health, prevention of infectious disease, etc.); of local governments' civil engineering works that have to be undertaken according to national investment plans (such as roads, rivers, ports, public housing construction, etc.); and of similar activities of shared interest. (c) National government may not transfer its burden to local governments that do not have responsibilities for that task. A.69 These rules are important because they give local budgetary expenditure stability and predictability. They also protect local governments from arbitrary burden-sharing requirements that might otherwise be imposed by the central bureaucracy. These rules also help to develop local governments' trust in the Central Government. -60- Correcting Fiscal Imbalance A.70 How the fiscal imbalance between the central and local governments, mentioned earlier, resolved? Broadly speaking, it is solved by central transfers. There are two types of transfers: transfers in the form of general funds, and specific-purpose grants-in-aid. The transfer of general funds is called the local allocation tax system. This system encourages the development of local autonomy in Japan. Specific-purpose grants also play an important role in the implementation of national policy by mobilizing local governments' high administrative capacity, by controlling their activities in line with central policies, and by securing local cooperation for the center's policy objectives. A.71 The local allocation tax system makes central transfers to local governments to supplement their tax revenues. The total amount of the local allocation tax is distributed to local governments to alleviate the fiscal disparity among them. Local governments handle the amounts they get as general revenue; they can use the allocated amount at their discretion, just as they would other local tax revenues. This system was established in 1954, after evolutionary steps taken since 1936. It has proven to be very successful. A.72 The local allocation tax is taken as a ratio of certain major national tax revenues. Originally, these sources were three major national taxes -- the income tax, corporation tax, and liquor tax. The local allocation ratio was initially set as low as 20 percent, but it was gradually raised (almost every year) until it reached 32 percent in 1966. Recent tax reform, which introduced the consumption tax, added 24 percent of the central government's consumption tax revenue and 25 percent of the national tobacco tax to the sources of the local allocation tax. A.73 The local allocation tax is intended to provide local governments with stable and predictable financial circumstances so that they may rely on their own investment and other business plans. In accordance with this purpose, the system has two objectives: to guarantee a certain basic level of general revenue to all local governments and to achieve an equitable distribution of general revenue among all local governments. A.74 It would be ideal if all local governments could get sufficient and equitable revenues from local taxes only. However, experience has shown that it is very difficult to design a system that can attain this ideal solely by means of a combination of local taxes, since tax bases are unevenly distributed. This difficulty is the root of the need for central transfers to local entities. A.75 The local allocation tax is distributed to those local governments that have shortfalls in general revenue needed to cover a basic level of administrative need. That basic level of need, calculated in money terms according to a formula for each local government, is known as the "basic financial need." In other words, the "basic financial need" is the amount of general revenue necessary for each local government to cover a standard level of administrative services and facilities. The "basic financial revenue," also calculated according to a formula for each local government, is the amount of taxes (i.e., general revenue) to be collected by each local government. The difference between basic financial need and basic financial revenue for - 61 - is a local government's "deficiency." This constitutes the amount of the local allocation tax that it can claim before adjustment for the total source amount available. Because the total of the local allocation tax is given first as the sum of major national taxes multiplied by certain ratios, the sum of the "deficiencies" is usually larger than the total amount at hand for allocation. Adjustment is made by subtracting a proportional amount from each local government's deficiency. Normally this adjustment is not substantial. If, on the contrary, a local government's basic financial revenue exceeds its basic financial need, then that local authority is better off, and its residents are already enjoying a higher level of administrative services. A.76 Basic financial need for a local government is calculated as the sum of: {unit cost x value of indicator x modification coefficient} for all public service items such as road and bridge expenses, primary school expenses, police expenses, etc. The indicators are indexes or factors that appropriately represent local financial needs. For example, the number of police officers is an indicator of police expenses. The number of teachers, students, and classes is used as an indicator of educational expenses. A.77 Basic financial revenue is calculated basically as the sum of general taxes (i.e., local taxes without earmarking). One feature of this calculation is worth noting. Tax revenues are calculated by standard tax rates. This works as an incentive for fiscal effort on the part of local governments. If tax revenues were summed up by actual tax rates above the standard rates, local governments would stop levying taxes at higher rates. Levying higher taxes is a 1painful decision for governors, local leaders, and members of local assemblies. No government would dare to choose unpopular but necessary higher tax rates only to see the additional tax revenues completely subtracted from the national transfer. A.78 Another important feature here is that basic financial revenue is limited to sum up only to 75 percent (80 percent for prefectures) of total local tax revenues. This gives some financial leeway or scope for local governments to cover administrative services specific to local demand. If it were not for this measure, local finance would be virtually stifled. Further, from a technical point of view, it is impossible to account for all the financial needs completely. The limit to the summing up of basic financial revenue also works as an incentive for local governments to continue their fiscal efforts. This measure, therefore, has great importance in making the local allocation tax system workable and sustainable. A.79 By the way, the Japanese system does not allow the Central Government to confiscate part of local revenue from wealthy local authorities for the center or other local governments to use. The Chinese shared-tax system, on the contrary, virtually enables the Central Government to confiscate tax revenue from wealthy provinces. Grants-in-Aid A.80 In contrast to the local allocation tax system, specific-purpose grants work as the most powerful and important fiscal channel for pursuing national policy objectives through local governments. It is the most important means for many Central Government agencies to secure - 62 - local governments' cooperative performance in accordance with the agency's instructions. According to the FY1990 local finance program, national disbursement -- or national grants in the broad sense -- amounts to more than 110 trillion. The recent share of central grants in local revenue is about 18 percent for prefectures and 9 percent for municipalities. Although this share has been gradually decreasing for the past ten years, the amount of total local expenditure tied to earmarked grants is huge. A.81 Grants have various functions for the national government: for example, to attain a standard level of administrative services nationwide; to enable large-scale investment on important projects of both national and local interest in accordance with national plans; to give financial assistance to local governments for disaster relief or other special financial needs; and to disseminate new projects that national government advocates. A.82 At the same time, problems are often pointed out that concern the inefficiency of subsidies and the Central Government's irrational and meticulous intervention in local administrative activities that are subsidized by the center. Some grants are based on unrealistically cheap cost estimates and eat up local funds that could be spent on local government's own projects. The criticism was once made that houses were constructed before sewer pipes had been connected because the grants involved had been made separately to each division in charge. Rigid conditionality put on subsidies is also a problem. Several years ago it was found that some local community centers had more than two entrances, with different nameplates. This happened because local governments used more than two grants with similar purposes to build what they needed; each grant, controlled by separate central agencies, demanded its own entrance bearing a plaque inscribed with the grant's title. Sometimes local governments take grants they do not really need in order to keep good relations with officials in charge so that they can get grants when they do need them. Further, the abundance of grants seems to have brewed a tendency among local leaders and other officials to look and ask for national grants whenever they embark on a new task. New local leaders are sometimes surprised when they realize how often they have to visit Tokyo and how many days a year they have to spend there visiting ministries and members of the Diet asking for the government's support and subsidies to solve local problems they face. But soon they get used to this routine, and some of them feel uneasy if they have not meet or talked with government people in Tokyo for some time. A.83 The Ministry of Autonomy is making its own effort, from the viewpoint of enhancing local autonomy and improving administrative efficiency, to check whether central subsidies are reasonably designed, are based on reasonable cost estimates, contain rational and reasonable conditionality for local governments, or can be abolished because the program can be left to local money and initiative after many years of repetitions. But the Ministry does not have the final power to revise the subsidies. Moreover, local governments are hesitant to point out problems caused by subsidies for fear that ministries may retaliate in time of real need. Further, downsizing the grants faces very strong resistance from ministries, local governments, and politicians. Ministries are afraid of losing grant-related personnel; local government departments are afraid of losing their easy access to stable work budgets; and Diet members do not want to lose a means of showing their presence to their constituents. - 63 - A.84 Despite all these problems and criticisms, it seems difficult to argue that central grants are totally unnecessary and to show a way to completely eliminate them. To do so would require proposing another system whereby national government could reasonably get local government's cooperation in pursuing national policy objectives. For the Central Government, grants-in-aid are an easy and powerful tool for controlling local governments. In general, greater attention can be paid to seeking intergovernmental fiscal circumstances whereby local governments have greater discretion in choosing central grants they really need. Macroeconomic Management A. 85 It is important to consider how local finance relates to macroeconomic management in Japan, or how macroeconomic management by the Central Government works through local budgets. A.86 Although each local government makes and implements its own budget independently, the Central Government can estimate the aggregate local spending that is needed for macroeconomic control of the whole economy. Before each fiscal year, the Central Government estimates both central and local tax revenues. While the Ministry of Finance prepares the budget and the fiscal investment and loan program of the Central Government, the Ministry of Autonomy prepares the local public finance program and the local loan program. Although these programs do not pay attention to how much money is spent and borrowed by each government, they do give an aggregate fiscal framework for local governments. The growth rate of the local public finance program gives local leaders a good guideline for how much revenue they can expect and how much money they can spend during the next year unless they experience unexpected, disastrous problems. A.87 The local public finance program was originally a tool to estimate aggregate annual sources of local revenue to cover standardized estimates of total local spending. The Ministry of Autonomy has primary responsibility to ensure that local governments' revenue sources are sufficient to balance the program. The local public finance program covers the whole of local governments' standard activities except local public enterprise special accounts, which are run on an independent profit system, and a few other special accounts. On the revenue side of the program, all the standard sources of local revenue -- such as local taxes, the local allocation tax, national grants-in-aid, local loan and fees, etc. -- are considered. If the estimated program is expected not to balance for the year, the Ministry of Autonomy has to propose measures (such as local tax amendments, an increase in the local allocation tax, an increase of local loans, curtailment of local administrative costs, and so forth) to restore financial balance. A.88 The local loan program is formulated each fiscal year as part of the Finance Ministry's fiscal investment and loan program, which is the plan for allocating huge amounts of government funds for policy purposesA' Loan amounts in this programs are scheduled for 19/ The sources of the funds are deposits with the Postal Savings Bank, reserves for welfare pension insurance funds and national pension insurance funds, etc. -64- various types of projects, considering local governments' requests, together with the allocation plan of various funds available for local loans. A.89 A local loan is used as an exceptional or complementary source of local revenue in the light of maintaining sound long-term local finance. To attain equitable allocation of funds for local governments, the Autonomy Minister's permission is needed for local loans. Procedures for obtaining permission have been simplified over the years. The Autonomy Minister gives permission to prefectures and very big cities but delegates permission for other municipalities to governors. Permission is given automatically for loans needed for national grant projects. Many other permissions are given as bloc allocations. A.90 The local public finance program and the local loan program were originally devised to estimate the amount necessary to be transferred to local government as general transfer funds and for allocating government and other funds available for various local projects. But they also serve, together with the level of specific-purpose grants in the central budget, as instruments for intergovernmental transmission of macroeconomic management by the Central Government. Protecting Local Autonomy A.91 Finally, it is important to mention at greater length an important institutional feature of Japanese intergovernmental relations. The Ministry of Autonomy is charged with designing and implementing a system of domestic governance and intergovernmental relations that is based on the principle of local autonomy, including technical support for local needs as a whole. More specifically, it is in charge of the local government system, local public enterprise system, local tax system, local finance system, local public service system, the election system, local development policy, the training of local officials, and so forth. A.92 The Ministry of Autonomy's function in the central bureaucracy is that it stands in for local governments as their sole advocate in intergovernmental relations and intergovernmental fiscal relations. This function is important because local governments have very little formal power over legislative processes in the national Diet, where law-making power resides. At the same time, local governments must be protected from government agencies that generally seek more control over local governments. Under these circumstances, the Ministry of Autonomy works as an institutional protector of local governments as well as an institutional liaison between the center and subnational governments. But laws and rules do not work by themselves. For example, the intergovernmental expenditure rules will not be implemented properly by the Central Government if there is no policeman in charge. The Ministry of Autonomy is the central bureaucratic interface for local governments. It is an important weight to maintain the institutional balance of power between the national government and local governments. It is also, as mentioned, a policeman for intergovernmental fiscal rules. The Ministry works to win as much tax sources as feasible for local governments. It requires other ministries to obey the intergovernmental fiscal rules. The local allocation tax system was devised by the Ministry. A.93 To conclude, the Japanese intergovernmental fiscal system is working well, with the local allocation tax system and the Ministry of Autonomy functioning as mechanisms for fiscal equalization and institutional balance, despite powerful and abundant specific-purpose grants as the main lever of central control. - 65 - STATISTICAL APPENDIX Table No. ]n SA.1 The Revenue-Sharing System in 1980 ..................... 66 SA.2 The Revenue-Sharing System in 1985 ..................... 67 SA.3 The Revenue-Sharing System in 1991 ..................... 68 SA.4 China: Government Finance, 1985-90 .................... 69 SA.5 Center-Local Shares in Revenue Expenditures ................ ... 70 SA.6 Trends in Investment and Consumption .................... 71 SA.7 Extrabudgetary Funds............................... 72 SA.8 Revenue and Expenditure by Province Per Capita ............. 73 SA.9 Tax Revenue and Contracted Revenue Sharing by Province ........ 74 - 66 - Table SA.1: The Revenue-Sharing System in 1980 Central Fixed Revenue: - Profit from centrally owned enterprises 1/ - Customs duty and industrial and commercial tax collected by customs - Industrial and commercial tax from rail Local Fixed Revenue: - Profit from locally owned enterprises - Salt tax - Agriculture tax - Business income tax - Industrial and commercial tax Shared Revenue at Fixed Ratio: In 1980, some of the locally owned enterprises were taken back to the central government, and the revenue from these enterprises was shared by central and local governments - 80 percent of it to the center and the remaining 20 percent to the local governments Adjusted Revenue: 2/ - Industrial and commercial tax 2/ Source: China, Ministry of Finance. I/ This is to be compared with the traditional socialist system of 'eating from one big pot.' 2/ Before 1982, there was no enterprise income tax in China, and the state-owned enterprises remitted all their profits to the Central Government. I/ Adjusted revenue is shared between central and local government at various ratios depending on the overall budget revenue situation of each province. - 67 - Table SA.2: The Revenue-Sharing System in 1985 I. "Fixed Central Government Revenues": 1. Income and adjustment tax of all Central Government enterprises 2. Business tax from railroads, bank and insurance company headquarters 3. Profit remittances by all centrally owned enterprises 4. Price subsidies paid to producers of grain, cotton, and oil (treated as a negative revenue of the Central Government) 5. Fuel oil special tax 6. Income taxes, sales taxes and royalties from offshore oil activities of foreign companies and joint ventures 7. Treasury bond income 8. 70% of the three sales taxes collected from enterprises owned by the Ministry of Industry, the Ministry of Power, SINOPEC, and the China Nonferrous Metals Company 9. All customs duty and all VAT and product taxes collected at customs 10. Tobacco tax and business tax on tobacco 11. Product tax on liquor and tobacco II. The "Local Fixed Revenues" (1985-87): 1. Income tax and adjustment tax of locally-owned enterprises 2. Income tax from collectively owned enterprises (ICIT) 3. Agriculture tax 4. Rural market trading tax levied on private sector traders 5. Local government grain trading loss (a negative tax) 6. Fines for delinquent taxes 7. The urban maintenance and construction tax (UMCT) 1/ 8. Housing tax 2/ 9. Vehicle utilization tax 10. 30% of the sales tax revenues collected from enterprises owned by the Ministry of Power, SINOPEC, and the China Nonferrous Metals Company 11. Wage bonus tax from state-owned enterprises 12. Self-employed entrepreneurs tax 13. Slaughter tax 14. Cattle-trading tax 15. Contract tax III. Taxes Shared between Central and Local Goverunents: 1. All sales taxes (VAT, business, and product) revenues from all enterprises, except those expressly excluded as described above under 1.6, 1.9, 1.10 2. Natural resource taxes 3. Construction tax 4. Salt tax 5. Industrial and commercial tax, and income tax, levied on foreign and joint venture enterprises 6. Energy and transportation fund tax 7. Personal Income tax Source: "China Revenue Mobilization and Tax Policy," Bax 3.2, (World Bank, 1990), p. 86. (Figures updated) / The UMCT is set at 7 percent of total sales tax liability for municipalities (5 percent for towns and 1 percent everywhere else. Z/ Private, owner-occupied housing and government buildings are exampt, and the Housing Bureau pays at a preferential rate of 12 percent. Payments by enterprises are deductiblefrom adjustment tax liability. On foreigners, it is called the "real estate tax" and is equal to 18 percent of rental value or 1.2 percent of capital value. Land is not taxed, only buildings. Table SA3: The Revenue-Sharinm System in 1991 Type A: Type B: Type C: Type 0: Type E: Type F: Contracted Income Increase Uniform rate Variable rate of Increased amount contracted Fixed quota Contracted Contracted Local retention of remittance(%) to go to State to quota transfers rate of rate retention Basic MarginaL Initial amount Annual rate (Y 100 M.) to increase percent) (percent) remit- remit- to state of growth (%) provinces (percent) tance tance (Y 100 m.) (Y 100 M.) Beijing 4.0 50.0 -- -- -- -- -- -- Hebel 4.5 70.0 -- -- -- -- -- -- -- Liaoning 3.5 58.3 -- -- -- -- -- Shenyang 4.0 30.3 -- -- -- -- -- -- Harbin 5.0 45.0 -- -- -- -- -- -- Jiangsu 5.0 41.0 -- -- -- -- Zhejiang 6.5 61.5 -- -- -- -- -- -- -- Ningbo 5.3 27.9 -- -- -- -- -- -- -- Henan 5.0 80.0 -- -- -- -- -- -- -- Chongqing 4.0 33.5 - -- - -- -- -- Tianjin -- -- 46.5 -- -- -- -- Shanxi -- -- 87.6 -- -- -- Anhui -- -- 77.5 -- -- ** -- -- Dalian -- -- -- 27.7 27.3 -- -- -- Qingdao -- -- -- 16.0 34.0 -- -- -- -- Wuhan -- -- -- 17.0 25.0 -- Guangdong -- -- -- -- -- 14.1 9.0 -- -- Hunan -- -- -- -- 8.0 7.0 -- -- Shanghai -- -- -- -- -- -- -- 105.0 -o Shandong -- -- -- -- -- -- -- 2.9 HeiLongjiang -- -- -. -- -- -- -- 3.0 - Jilin -- -- -- -- -- -- -- -- 1.1 Jiangxi -- -- -- -- - -- -- 0.5 Shaanxi -- -- -- -- -- -- -- -- 1.2 Gansu -- -- -- -- -- -- -- * 1.3 Fujian -- -- -- -- -- -- -- -- 0.5 Inner Mongolia-- -- -- -- -- -- -- -- 21.6 Guangxi -- -- -- -- -- -- -- -- 8.6 Tibet -- -- -- -- -- -- -- -- 9.1 Ningxia - -- -- -- -- -- - - 5.9 Xinjiang -- -- -- -- -- -- ** -- 17.6 Guizhou -- -- -- -- -- -- -- -- 9.0 Yunnan -- -- -- -- -- -- -- -- 7.1 Qinghai -- -- -- -- -- -- -- -- 7.4 Hainan ------ ------1.4 Source: "China-Local Revenue Sharing Cdntracts, 1988-1990n (IMF), Table 111.2. Figures updated There are five main types of contracts h"der which transfers are effected from the provinces to the central government: (a) Type A (adopted by ten localities): based on 1987 revenue, and localities retain a specified proportion ranging from 28 percent to 80 percent of any revenue that is within a certain percentage growth ranging from 3.5 to 6.5 percent) from the previous year. Any revenue in excess of this growth rate is retained entirely by the province; (b) Type B (three localities): a specific proportion of all revenue is remitted to the center; (c) Type C (three Localities): a certain proportion is retained up to a quota, and then a different (usually higher) proportion of revenue is retained in excess of the quota; (d) Type D (two localities): a specified amount is contracted to go to the center in the initial year; in subsequent years the amount was to increase at a contracted rate (7 percent or 9 percent); this is a variant of Type A, contracting a specific amount rather than a proportion; (e) Type E (three localities): a specific amount is contracted to be transferred to the center with no annual adjustment for high growth. - 69 - Table SA.4: China: Government Finance, 1985-90 (IMFIGFS format) (percent of GDP) Overall Subsidies in Government Government Current Government State Government Surplus/ Current Account Year Revenue Tax Expenditure Deficit Expenditure Surplus 1985 26.7 25.6 27.2 -0.5 5.9 7.0 1986 25.2 23.2 27.1 -1.9 6.0 5.8 1987 22.7 20.5 24.9 -2.2 5.9 4.5 1988 20.0 18.3 22.3 -2.4 5.4 3.0 1989 20.8 18.9 22.8 -2.4 6.1 2.4 1990 19.8 17.7 21.9 -2.1 5.4 2.5 1991 (budget) 18.6 17.6 21.1 -2.5 4.5 1.8 Source: IMF, "People's Republic of China - Background Papers,* January 1991. Note: Calculated in IMP's Government Finance Statistics (GFS) format. - 70 - Table 9A.5: Center-Local Shares in Revenue Expenditures Center's Sharing (%) in State Budget Adjusted by On official excluding Year definition borrowings 1980 19.4 19.4 1981 20.7 15.0 1982 23.0 16.7 1983 29.8 25.0 1984 - 34.9 31.4 1985 37.0 33.4 1986 39.7 36.7 1987 35.3 30.3 1988 36.0 29.8 1989 37.5 32.3 Source: Christine Wallich, "China: Draft Report on Center-Local Fiscal Relations" (World Bank, 1990). (Data needed from the Chinese authorities to revise the table). - 71 - Table SA.6: Trends in Investment and Consumption Capital Construction in State-owned Units Growth Growth % of GDP GDI/ Rate of Rate of Transportation GDP Personal Public and Heavy Year (%) Consumption Consumption Energy Communication Industry I (annual % in real terms) 1979 34.9 4.4 21.2 1980 32.2 12.7 4.8 1981 29.2 8.6 2.6 2.0 0.8 3.6 1982 29.7 5.5 9.7 2.0 1.1 4.1 1983 30.4 9.7 8.6 2.2 1.3 4.2 1984 32.3 10.0 22.7 2.4 1.6 4.3 1985 40.5 3.0 3.0 2.4 2.0 4.5 1986 40.8 4.2 12.5 2.8 1.9 4.6 1987 39.3 11.6 9.7 3.0 1.7 5.2 1988 39.5 12.7 5.2 2.9 1.5 4.9 1989 38.4 6.4 13.1 2.8 1.0 4.4 1990 36.6 5.1 9.5 3.1 1.2 4.7 1991 35.7 7.8 7.8 3.2 1.6 5.0 Source: China: Statistical Yearbook 1991, p. 133, Statistical Abstract 1992 (in Chinese), p. 26 and World Bank staff estimates. Note: GDI - gross domestic investment - 72 - Table SA.7: Extrabudgetary Funds Extrabudgetary Funds (billion yuan) Allocated to Investment, Total Major Repair, Contribution (billion % of and Simple to Government Year yuan) GDP Construction Funds 1983 96.8 16.7 50.0 7.0 1984 118.8 17.1 58.7 10.1 1985 153.0 17.9 72.6 12.3 1986 173.7 17.9 75.5 12.6 1987 202.9 17.9 94.7 15.4 1988 227.0 16.2 Source: China: Statistical Yearbook, 1990 - 73 - Table SA.8: Revenue and Expenditure by Province Per Capita Per Capita Per Capita Province Revenue /1 Expenditure /2 LOW Tibet 40 517 Guizhou 49 80 Guangxi 52 77 Jiangxi 56 81 Sichuan 57 63 Anhui 58 66 Qinghai 60 248 Xingjiang 54 178 Average for Group /3 55 113 MEDIUM Henan 63 64 Shaanxi 68 92 Ningxia 70 237 Hunan 70 71 Inner Mongolia 72 170 Yunnan 80 108 Gansu 81 118 Hebei 81 75 Average for Group /3 73 117 HIGH Shandong 88 67 Fujian 92 113 Shanxi 95 135 Heilongiiang 98 135 Hubei 102 88 Guangdong 112 107 Jiangsu 135 81 Zhejiang 145 93 Liaoning 231 154 Average for Group /3 122 108 Tianjin 597 334 Beijing 609 344 Shanghai 1492 346 Source: China, Ministry of Finance /1 Yuan, 1985. /2 Yuan, 1985. /3 Leaving out the extreme observations for Tibet and municipalities of Shanghai, Beijing, and Tianjin. Table SA.9: TAX REVENUE AND CONTRACTED REVENUE-SHARING BY PROVINCE Tax Revenue Revenue Revenue Contract Formula (%) /2 Contract Amount 13 in the Remittance 10% Growth Remittance Total Remittance Rate of Marginal Initial Rate of Initial in the Ini- of Tax of 10% Tax on Province Revenue /1 Weight Rate Increase Remittance Amount Increase Year tial Year Revenue Revenue Growth Beijing 6,361 4.34 50.0 4.0 6,361 3,180.50 6997.1 3307.72 Tianjin 5,587 3.81 53.5 5,587 2,989.05 6145.7 3287.95 Hebel 5,762 3.93 30.0 4.5 5,762 1,728.60 6338.2 1806.39 Shanxi 3,361 2.29 12.4 3,361 416.76 3697.1 458.44 Inner Mongolia 1,943 1.33 (1,840) Liaoning 6,134 4.18 41.7 4.5 6,134 2,557.88 6747.4 2672.98 Shenyang 2,510 1.71 30.3 4.0 2,510 760.61 2761.3 791.03 Dalian 2,156 1.47 27.7 27.3 1 Jilin 3,752 2.56 (110) Hellongliang 4.132 2.82 300 4,132 300.00 4545.2 300.00 Harbin 1,244 0.85 45.0 5.0 1,244 559.79 1368.4 587.78 Shanghai 16,513 11.26 10,500 16,513 10,500.00 18164.3 10500.00 Jiangsu 10,717 7.31 59.0 5.0 10,717 6,323.03 11788.7 6639.18 Zhejiang 6,703 4.57 38.5 6.5 6,703 2,580.66 7373.3 2748.40 Ningbo 993 0.68 27.9 5.3 993 277.10 1092.5 291.78 Anhui 3,884 2.65 22.5 3,884 873.90 4272.4 961.29 Fujian 3,316 2.26 (50) _ Jiangxi 2,821 1.92 (50) Shandong 7,279 4.96 1 290 7,279 290.00 8006.9 290.00 Henan 6,315 4.31 20.0 5.0 6,315 1,263.00 6946.5 1326.15 Hubel 3,617 2.47 Wuhan 2,918 1.99 * 17.0 25.0 Hunan 5,438 3.71 800 7.0 5,438 800.00 5981.8 856.00 Guangdong 9,588 6.54 1,410 9.0 9,588 1,410.00 10546.8 1536.90 Guangxi 3,054 2.08 1 (610) Sichuan 5,680 3.87 Chongqing 1,906 1.30 33.5 4.0 1,906 638.51 2096.6 664.05 Guizhou 2,194 1.47 (740) - Yunnan 3,749 2.56 (670) Tibet (4) 0.00 (900) Shaanxi 2,818 1.92 (120) Gansu 2,258 1.54 (130) Oinghai 401 0.27 1 (660) 1 Ningxia 411 0.28 (530), XingjIang 1,158 0.79 (1,530) - ------ Total 146,630 100.00 104,427.40 37.449.37 114,870.14 39,02.04 Share of the Center 35.86% 33.97% Source: China, Miaistry of Finance /1 Million yuan in 1987. /2 1988-1990. /3 Million yuan from 1988 to 1990 - 75 - Richard M. Bird, Federal Finance in Comparative Perspective. (1986) Canadian Tax Foundation, Toronto, p. 89. Canadian Tax Foundation, Provincial and Municipal Finances. 1989 (Toronto, 1990), p. 13:6. R.J. Chelliah (1980); Government of India, Final Report of the Indirect Taxation Enquiry Committee (New Delhi, 1978). R.J. Chelliah, "Intergovernmental Fiscal Relations and Macroeconomic Management in India" (paper submitted to Senior Policy Seminar on Inter-Governmental Fiscal Relations and Macro-Economic Management in Large Countries (EDI) in New Delhi in 1991); (mimeo; World Bank, 1990). China-Local Revenue Sharing Contracts, 1988-1990 (IMF), Table M.2 China Revenue Mobilization and Tax Policy, Box 3.2 (World Bank, 1990), p. 86. China, Statistical Yearbook, 1990. T. Fujiwara, "Intergovernmental Fiscal Relations in Japan" (mimeo; World Bank, 1991) forthcoming. J. S. H. Hunter, Federalism and Fiscal Balance (Australian National University Press, Canberra, 1977). IMF, People's Republic of China: Background Papers (Washington D.C., January 1991). R. J. May, Fiscal Federalism and Fiscal Adjustment, (Clarendon Press, Oxford, 1969), p. 111. Charles E. McLure, Tax Assignment in Federal Countries, (Australian University Press, Canberra, 1983). Mieszkowski and Oakland, Fiscal Federalism and Grants-in-Aid (The Urban Institute, Washington, D.C., 1979). R. A. Musgrave, "Who Should Tax, Where and What?" In Tax Assignment in Federal Countries, ed. Charles E. McLure (Australian University Press, Canberra, 1983) - 76 - R. A. Musgrave, Peggy Musgrave, and Richard M. Bird, Public Finance in Theory and Practice, First Canadian edition (New York: McGraw-Hill Ryerson Ltd., 1991), p. 510. M. G. Rao and R. J. Chelliah, Survey of Research on Fiscal Federalism in India (Processed), (National Institute of Public Finance and Policy, New Delhi, 1990). Report of the Western Finance Ministers, Economic and Fiscal Development and Federal- Provincial Fiscal Relations in Canada (Lloydminster, Saskatchewan, 1990), p. 20. Christine Wallich, China: Draft Report on Center-Local Fiscal Relations (World Bank, 1990). Distributors of World Bank Publications ARGENTINA EL SALVADOR JAPAN SOUTH ARICA, BOTSWANA Cadols Hirlsch, SRL Fusadm Esern Book Servk e Forsigletie Galeria Guemes Alarn Dr. Manuul En oque Araujo #3530 Hongo 3-home, Bunkyo-ku 113 Oxford University Pres Flodda 165, 4th Floor-Ofi. 453/465 Ed~fco SISA. ler. PlN Tokyo Soutlhern Arica 1333 Buenos Aires San Salvador 011 PO. Box 1141 KENYA Cape Town 8000 AUSTRALIA, PAPUA NEW GUINEA, FINLAND Africa Scok Service (,.A.) Ltd. FIJI, SOLOMON ISLANDS, Akateemninen IGrjakauppa Quaran House, Mfwngano Street For sukmcrpoont: VANUATU, AND WESTERN SAMOA P.0. Box 128 P.0. Box 45245 International Subseiption Service D.A. Bo & Journal& ~F-00101 Helnk 10 Nabi P.O. BOx 41095 648 Whitehorse Road Crag M ~icham3132 FRANCE KOREA, REPUBLICOF johamesburg2024 Vktoria World Bank Pubicadoms Pan Korm Book Corporation 66. avenue d'léna P.0. BK 101. Kwangwhamun SPAIN AUSTRIA 75116Paris Seoul Mundi-Prensa Lbros, S.A. carold and Co. Ca~o37 raben31 CERMANY MALAYSIA 28001 Madrid A-1011 Wien UNO.Verlag University of Malaya Cooperative Poppel~dorfer Ale 55 Bookshop, Limted LhrerLa InternacionalAEDOS BANGLADESH D-5300 Bonn1 PO. Box 1127, Ja~a Pantai bru Con~ell de Cent,391 Micr Indugtries Development 597W Kuala Lumpur 08009 Barcelona Asdance~ociety(MIDAS) CUATEMALA Huse5, Road 16 Libreril Pledra Sant MøXCO SRI LANKA AND THE MALDIVES Dhanmondi R/Area Sa. Calle 7-55 INPDTEC Laks House Bookshop Dhaka 1209 Zona Apartado~Potal22-60 P.O. Box24 Gua~emala City 1406011alpan, MexDY .P. 100,51r Chittampalam A. B~t~ oes Gardiner Mawatha 156, Nur Ahmed Sarak HONGKONG.MACAO NErHERLANDS Colombo 2 Chi g4000 Asia 2000 Ltd. De Undoomn/InOr-Publikates 46.48 Wyndham Street P.O. Box202 SWEDEN 76, K.D.A. Avenue WinningCentre 7480 AE Haak~bgen For singe fiti: Kulna 9100 2nd Floor Fritzes Packboksforetaget Central Hang Kong NEW ZEALAND Regeringsgatan 12, Box 16356 BELGIUM EISCONZ Ltd. 5-10327 Stockholm Jean Da Lannoy INDIA Private Mail Bag99914 Av. du Roi 202 Allied Publishers Private Ltd. New Market For ~uhsriØin onfen: 1060 Brusels 751 Mount Road Auckland Wennergren-W~lliams AB Madras-600002 5ox30004 CANADA NIGERIA S-104 25 Sckholm Le Diffuseur Bmewk o Univerity Pros Limited C.P. 5,1501ruc Amphre 15 J.N. Herdde Marg Three Crowns Building Jericho SWITZERLAND Boudr~«ile,Qubec Ballard Esktte Private Mal sag 509s For win=g d: 1485 A Bombay- 400028 nban LibrariePayot 1, rua de Bourg CHINA 13/14Aaf All ad NORWAY CH 1002 Lausnne China Finanial & Economc New Delhi -110 002 Narvum informaon Center Publ~hing House Bock Departent For sucrpto uuiuu: 8, Da Fo 51 Dong Jie 17 ChitharanjanAvenne P.O. Box6125 Ettertad Librairle Payot Beiling Calutta-70002 N-0602 0lo6 Service da Abonnementa Case postl 3312 COLOMBIA Jayadeva Hostel hullding PAKISTAN CH 1002 Lausanne Infoenlace Ltda. 5th Main Røad Gandhinagar Mire Book Agency Apartdo Aereo34270 Dangalor-560 009 65,sShnhrah--Quaid-Aram TANZANIA Bogot D.. P.. Box No. 729 Oxford Univrsty Preas 3.5-1129 Ka~hguda Crom ad Lahoe54000 P.. BOx 5299 COTE DIVOIRE Hyderabad-500027 Mak+~eRuad Centre d'Edition et de Dffuon PERU Dars Afrcaine (CEDA) Prarth~na Plata, 2nd Flor Editoral Darrollo SA 04&.P.541 Ner Thakore aug, Nvrangpurt Apartdo3024 THAILAND Abidjan 04 Plateau Ahmedabad -31)009 Lia 1 Central Departmnt Store 306Silem Road CYPRUS Patiala Hous= PHILIPPINES Bangkok Cyprus College Baok~a 16-Å Ashok Marg Intamational Book Center 6, Diogene Street, Engomi Lucknow-226001 FIMt Floor, Fildpinas Lfe Building TRINIDAD & TOBACO, ANTICUA PO. Box 2006 Ayala Avenue,Makati BARBUDA, BARBADOS, NIcoeia C~ntral Baaar Road Me~roManila DOMINICA. GRENADA, GUYANA, 60 Stjal Napr JAMAICA, MONTSERRAT, ST. DENMARK Nagpur440010 POLAND KiTT &NEVIS,ST.LUCIA, Samfundsutteratur ORPAN ST. VINCENT & GRENADINES Rosanoerns AMl111 INDONESIA Palac Kultury i Nauki Systematic Studie Unit DK-1970 Prederikberg C Pt. Indira Limiled 00-901 Warmwa *9 Watt Street JLSmRatulangi37 Cpe DOMINICAN REPUBLIC P.O. Box181 PORTUGAL Trinidad,Westindles Edltorm Taller, C. par A. Jakarta Pusat Lvraria Portugal Restauración Isabell Catlca 309 Rua Do Carmo ?074 UNITED KINGDOM ApartdodeCorres2190Z-1 ISRAEL 1200 Lisbo Micrinfo Ltd. Santo Domingo Yomot Uterature Ltd. P.0. BOx 3 P.. BOx 56055 SAUDI ARABIA, QATAR Alton, Hampshire CU34 2PC EGYPT, ARAB REPUBLIC OF Tel Aviv 61560 Jlarir ook Ste England Al Ahram Israel P.O. Box3196 Al Gala Stret Rlyadh 11471 VENZELA Cairo ITALY Librera del Ente Lisa Comn,Mieaua Sanni SPA SINCAPORIL TAIWAN, Aptdo. 60.37 The Middle Est Ob*erver Via Dum Di Calabri 1/1 MYANMALBRUNEI Caraca1060-A 41, Shaf Stret Caeml Pasta 552 Information Pubhlcations Cairo 50125 Pirenze Private, Ltd. 02.06 lt FL, Pe-u Industrial øld& 24 New Industrial Road Singapore 1953 u Recent World Bank Discussion Papers (continued) No. 148 The Sectoral Foundations of China's Development. Shahid Javed Burki and Shahid Yusuf, editors No. 149 The Consulting Profession in Developing Countries: A Strategy for Development. Syed S. Kirmani and Warren C. Baum No. 150 Successful Rural Finance Institutions. Jacob Yaron No. 151 Transport Development in Southern China. Clell G. Harral, editor, and Peter Cook and Edward Holland, principal contributors No. 152 The Urban Environment and Population Relocation. Michael M. Cernea No. 153 Funding Mechanisms for Higher Education: Financing for Stability, Efficiency, and Responsiveness. Douglas Albrecht and Adrian Ziderman No. 154 Earnings, Occupational Choice, and Mobility in Segmented Labor Markets of India. Shahidur R. Khandker No. 155 Managing External Debt in Developing Countries: Proceedings ofa joint Seminar,Jeddah, May 1990. Thomas M. Klein, editor No. 156 Developing Agricultural Extension for Women Farmers. Katrine A. Saito and Daphne Spurling No. 157 Awakening the Market: Viet Nam's Economic Transition. D. M. Leipziger No. 158 Wage Policy during the Transition to a Market Economy: Poland 1990-91. Fabrizio Coricelli and Ana Revenga, editors No. 159 International Trade and the Environment. Patrick Low, editor No. 160 International Migration and International Trade. Sharon Stanton Russell and Michael S. Teitelbaum No. 161 Civil Service Reform and the World Bank. Barbara Nunberg andJohn Nellis No. 162 Rural Enterprise Development in China, 1986-90. Anthony J. Ody No. 163 The Balance between Public and Private Sector Activities in the Delivery of Livestock Services. Dina L. Umali, Gershon Feder, and Cornelis de Haan No. 165 Fisheries Development, Fisheries Management, and Externalities. Richard S. Johnston No. 166 The Building Blocks of Participation: Testing Bottom-up Planning. Michael M. Cernea No. 167 Seed System Development: The Appropriate Roles of the Private and Public Sectors. Steven Jaffee andJitendra Srivastava No. 168 Environmental Management and Urban Vulnerability. Alcira Kreimer and Mohan Munasinghe, editors No. 169 Common Property Resources: A Missing Dimension of Development Strategies. N. S. Jodha No. 170 A Chinese Province as a Reform Experiment: The Case ofHainan. Paul M. Cadario, Kazuko Ogawa, and Yin-Kann Wen No. 171 Issues for Infrastructure Management in the 1990s. Arturo Israel No. 172 Japanese National Railways Privatization Study: The Experience ofjapan and Lessons for Developing Countries. Koichiro Fukui No. 173 The Livestock Sector in Eastern Europe: Constraints and Opportunities. Cornelis de Haan, Tjaart Schillhorn Van Veen, and Karen Brooks No. 174 Assessing Development Finance Institutions: A Public Interest Analysis. Jacob Yaron No. 175 Resource Management and Pastoral Institution Building in the West African Sahel. Nadarajah Shanmugaratnam, Trond Vedeld, Anne Mossige, and Mette Bovin No. 176 Public and Private Sector Roles in Agricultural Research: Theory and Experience. Dina L. Umali No. 177 The Regulatory Impediments to the Private Industrial Sector Development in Asia: A Comparative Study. Deena Khatkhate The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'I6na Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100,japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: wui 64145WORLDBANK Telex: 640651 Facsimile: (3) 3214-3657 RcA 248423 WORLDBK Telex: 26838 Cable Address: iNTBAFRAD WASHINGTONDC ISBN 0-8213-2223-0