UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trench warfare on the tax fields : fiscal sociology and Japan’s centralized tax state Dewit, Andrew Pieter James 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1997-195708.pdf [ 14.44MB ]
JSON: 831-1.0088005.json
JSON-LD: 831-1.0088005-ld.json
RDF/XML (Pretty): 831-1.0088005-rdf.xml
RDF/JSON: 831-1.0088005-rdf.json
Turtle: 831-1.0088005-turtle.txt
N-Triples: 831-1.0088005-rdf-ntriples.txt
Original Record: 831-1.0088005-source.json
Full Text

Full Text

T R E N C H W A R F A R E ON THE TAX FIELDS: FISCAL S O C I O L O G Y AND J A P A N ' S . C E N T R A L I Z E D TAX S T A T E Andrew Pieter James Dewit -B.A., Simon Fraser University, 1988 M A . , York University, 1989 A THESIS SUBMITTED IN PARTIAL FULFILLMENT O F THE REQUIREMENTS FOR THE D E G R E E O F DOCTOR O F PHILOSOPHY in , T H E F A C U L T Y O F G R A D U A T E STUDIES (Department of Political Science) We accept this thesis as conforming to the required standard THE UNIVERSITY O F BRITISH COLUMBIA January 1997 (c) Andrew Pieter James DeWit, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract According to the World Bank, over 80 countries currently plan some form of fiscal decentralization. This group includes Japan, a highly centralized tax state, which in 1995 put in place a law and associated high-profile institutions to promote fiscal and administrative decentralization. But the process in Japan has quickly become bogged down. This dissertation asks why there are hurdles confronting fiscal decentralization in Japan. My research uses the fiscal sociology approach and highlights bureaucratic interests in Japan's intergovernmental tax regime, one that is especially interesting in comparison with other advanced industrial countries. Fully 70 percent of all government spending in Japan is done by subnational levels of government, and 51 percent of the central state's current operating expenditures are transfers downward that serve to support that high rate of local spending. And though Japan is a centralized state, over a third of its taxation is collected locallly. In consequence, Japan is quite anomalous when set alongside a representrative sample of federal and unitary states. No other country among the nine O E C D nations used, for comparative purposes, in this dissertation combines such high fiscal transfers with heavy levels of subnational spending and taxation. i i And no other country gives close control over a large terrain of subnational taxation to a central-state agency (the Ministry of Home Affairs). The ministry's vast bureaucratic turf brings it into conflict with the Ministry of Finance, as both seek to maintain or expand their fiscal jurisdiction. They are thus not interested in fiscal decentralization in large part because they are busy fighting "trench wars" with each other on these tax fields. This dissertation hence undertakes a detailed analysis of intergovernmental fiscal history in Japan, focusing in particular on the Ministry of Home Affairs' ambiguous role in the Japanese state. The case studies of inter-bureaucratic fiscal politics include the Local Allocation Tax (a large general subsidy) as well as an array of taxes on the fields of income, assets and consumption. i n Table of Contents Abstract ii-iii Table of Contents iv-vi List of Tables and Figures vii Preface viii-xif* List of Abbreviations xii! Chapter 1: Introduction 1 Fiscal Sociology and Fiscal Politics 3 Fiscal Decentralization and Japan's 1940 System 7 Comparative Tax States 1 3 Japan's Taxes and Fiscal Transfers 26 Chapter 2: Institutions and Fiscal Decentralization 31 Potential Technical Constraints on Decentralization 35 Degrees of Decentralization 43 Decentralization as a Comparative Variable 48 Public Finance and Local Autonomy in Japan 53 Political Science and Local Autonomy in Japan 55 Financial Dependence and Local Autonomy 57 Lesson's From Minobe's Fiscal War 60 Refuting the "Vertical Control Model" 63 Institutionalizing the Local System 67 Chapter 3: Institutions and Japan's 1940 Fiscal System 71 The Prewar Tax System 71 The 1940 Tax Reform 84 Enhanced Tax Control 90 The Postwar Reforms 91 Fiscal Decentralization under SCAP 92 The Dodge Line Shock and the Shoup Report 96 The Deconstruction of Shoup 100 Fiscal Control for Equity as Uniformity 101 The Redistribute System 106 MOHA's Tax Controls 109 iv Table of Contents, cont. Chapter 4: Interests in Japan's Contemporary Fiscal Politics 114 MOHA's Organizational Mission 114 Politicians and Tax Politics"! 22 Intergovernmental Tax Politics 127 Local Politicians and Fiscal Decentralization 130 Other Interests 135 Other Central Agencies 137 Chapter 5: Controlling Redistribution: The Local Allocation Tax 140 Recentralizing Redistribution 140 Conflicts with MOF143 Allocating the Rules of Distribution 147 The Politics of Allocation Tax Increases 153 MOF-MOHA Compromise on the Special Account 156 General Subsidies and Fiscal Dependence 158 Proposals for Reform 162 Prospects for Politicization 163 Chapter 6: The Income Tax Field and Corporate Taxes167 1940 and the Modern Income Tax 167 Japan's Postwar Income Tax Cutsl 73 Postwar Japan's Special Tax Measures177 MOF's Initiation of Special Measures178 The Role of MOHA and Local Taxation 187 Comparisons of Current Business Tax Regimes 193 Corporate Taxation and the Fiscal Crisis 197 Corporate Taxes and Intergovernmental Fiscal Politics 199 The MOF-MOHA Stalemate over Corporate Income Taxes 201 Chapter 7: The Property Tax Field 208 Prewar Politics and the Land Tax 208 Fiscal Democracy and the FAT 212 The Current Property-Tax Regime 216 The Land-Price Spiral and the Tax Regime 217 v Table of Contents, cont. Skirmishing Within the Tax Advisory Commission 220 The MOF Invasion of MOHA's Domain 223 MOHA's Counterattack 221 The Post-Bubble Property Tax Revolt 231 Chapter 8: The Consumption Tax Field 238 A Tax Hike Foretold 238 Indirect Taxation Before the Consumption Tax 240 Intergovernmental Fiscal Politics 244 The Local Consumption Tax 247 The Return of the Consumption Tax Issue 251 Getting The Tax on the GTAC Agenda 252 The Debate in the GTAC 256 Party Politics and the Compromise 263 Chapter 9: Conclusions 273 Institutions and Intergovernmental Tax Politics 277 Prospects for Fiscal Decentralization 279 Appendix: Institutions and Tax Politics 286 Levels of Analysis 286 The Return of the State 288 The New Institutionalisms 293 Institutions and Wartime Tax Centralization 298 Institutions and the Modern Tax State 301 The Principal-Agent Approach 301 Modified Pluralism and Policymaking in Japan 305 Bureaucratic Rationality 308 Historical Institutionalism 313 Critiques of Historical Institutionalism 317 Opting for Eclecticism 318 References 320 vi Tables Table 1.1: Government Fixed Capital Formation, 1994/ 21 Table 2.1: Tax Structures at National and Local Levels, 1994/41 Table 3.1: Ratio of Local to National Direct Tax, Selected Prefectures, 1934/79 Table 3.2: Tax Rates in Selected Prefectures, 1935/ 80 Table 3.3: National Government Expenditures and National Income in Japan, 1936-1945/82 Table 3.4: Change in the Structure of Local Finance, Percentage of Revenues/86 Table 3.5: Change in Tokyo Metropolitan Government Revenues,1989-1993/ 113 Table 4.1: Changes in the Japanese Tax Mix, 1955-1992/ 122 Table 6.1: National Taxes By Principal Items, 1940-1945/169 Table 6.2: National Taxes By Principal Items, Selected Years 1946-90/171 Table 6.3: Taxes on Corporate Income as Percentage of GDP/172 Table 6.4: Estimated Annual Tax Changes in Japan, by Type, 1950-74/183 Table 6.5: Corporate Income Taxation at National and Local Levels for Selected Countries as Percentage of GDP, 1991/192 Table 6.6: Effective Corporate Tax Rates Among 4 Countries, Selected Years/193 Table 6.7: Toyota Corporation's Tax Position, Various Countries/195 Table 6.8: The Tax Position for Matsushita Electronics, Various Countries/196 Table 6.9: The Tax Position For Representative Steel Companies, Various Countries/197 Table 7.1: Land Tax Revenues, Selected Years 1880-1947/212 Table 7.2: Land Tax Revenues by Level of Government, 1930-1944/ 215 Table 7.3: Taxes on Property as Percentage of GDP, 1965-94/ 229 Table 8.1: Consumption Taxes as Percent of GDP/251 Figures Figure 1.1: Local Expenditures, Revenues and Fiscal Transfers, 1994 /15 Figure 1.2: Government Sector Spending as a Percentage of GDP, 1994/ 22 vn Preface Being Canadian, I most certainly did not head for Tokyo in the winter of 1993 to research Japan's intergovernmental tax politics. As Donald Smi ley 1 once wryly observed, Canadian fiscal federalism is a field that "I once tried to comprehend but which, I am now convinced, is so complicated that one should either cultivate it as a full-time specialty or leave it alone entirely." Federal-provincial fights over fiscal resources are, as is disturbingly clear in the erosion of Ottawa's capacity to bind the country, of great pith and moment in the long run; but the micro-politics of the short run are even more excruciatingly complicated and boring than Canadian debates on the Constitution. Thus I would have thought, in 1993, that only masochism would motivate one to study similar processes elsewhere. But on an especially cold and wet Sunday in October of 1993,1 was fortunate to read a newspaper piece on taxation that quoted Jinno Naohiko, an expert on Japanese fiscal history at Tokyo University's Department of Economics. His comments bore the unmistakeable flavour of fiscal sociology, the once obscure but now resurgent approach to the tax state pioneered by the likes of Joseph Schumpeter and added to by other such luminaries as Richard Musgrave and James O'Connor. Though I was in fact researching only a few 1 Donald V. Smiley, The Federal Condition in Canada. Toronto: McGraw-Hill, 1987, p. x i . v i i i dozen metres from the Economics building, I think it unlikely that I would otherwise have met Professor Jinno, since his work focuses on the intergovernmental fiscal regime. S o I am immensely glad I bought that copy of the A//'/c/ce/when I had nothing better to do, read the article, and wrote the letter introducing myself and explaining my delighted surprise at reading his remarks. Professor Jinno's boundless patience and often hilarious tales of Japanese intergovernmental tax politics kindled my interest in a way that no amount of reading ever could have. And now, oddly enough, I find myself fascinated by intergovernmental fiscal politics anywhere, even in Canada. His is the singular gift of real teaching, and I can only hope to return a fraction of what I have received by similarly motivating my own students. There are others I want to thank for helping me through this project. Lonny Carl i le, the co-chair of my committee, arranged my stay at Tokyo University's Institute of Socia l Science (Shaken) and stuck by this work, giving me invaluable advice, though he had no obligation to do so having taken up a new position in Hawaii. I have also benefited very much from the close readings, heavy criticism, and supportive comment the other members of my committee, Philip Resnick and Alan Siaroff, gave to my work. I should also like to express my gratitude to John C. Campbel l , David Edgington, Nina Halpern and Masaru Kohno for their instructive and insightful comments. Shibagaki Kazuo and Shibuya Horoshi of Shaken sponsored my two and a half year stay ix there, providing me with useful introductions and assistance throughout. Other Shaken faculty, including Hiwatari Nobuhiro, Kudo Akira, and Fujiwara Yukichi, helped me at crucial points in my research and related activities. And the Shaken librarians as well as those in the Economics Department were most helpful as well as patient with my inexcusable habit of wanting to borrow things a minute or two after closing. After Shaken, I went to the Australia-Japan Research Centre at Canberra's Australian National University. I am fortunate that I did, and want to thank the director, Peter Drysdale, and the staff and students, as I benefited from giving seminars in the department and discussing ideas with them. I was also lucky to be permitted access to the literature in the nearby Federalism Research Centre. The Centre itself is now defunct, apparently the victim of an unappreciated critical perspective, but its works on federalism and fiscal history remain a gold mine of international academic exchange. A similar institution clearly needs to be established, and protected from the politicians. Others to whom I owe many thanks for advice and encouragement are Junko Kato, Tsuyoshi Kawasaki, David Leheny, Fujita Kuniko, Hayden Lesbirel, Jonathan Lewis, Steven Reed, and Sven Steinmo. My interviewees, including personnel from the ministries of Finance and Home Affairs, were also most helpful and patient with my questions, and I owe them much gratitude. Specia l thanks to two people in particular who, some years ago, got me interested in fiscal sociology and fiscal politics. Isabella Bakker's graduate courses on State Finance introduced me to an unforgettable literature that, it turns out, is perfectly suited to the fiscal history of Japan. Also, Neil Brooks gave me much more of his time than I deserved, and for me made fiscal politics something you study not only because it is of course of fundamental and unparalleled importance, but much more than that because at the end of the day it is just so damned funny. I also want to thank my wife, Mie Takahashi, whose patience, support, and presence are my anchor. This is for her and for our wonderful little daughter Monika, whose birth at the end of 1996 was just the jump-start needed to get this dissertation pruned and in the mail. As to finances, this work was supported by fellowship monies from the University of British Columbia and from Canada 's Socia l Sc ience and Humanities Research Counci l . Other sources I dare not mention, lest some hard-hearted tax official read this far. Of course, the shortcomings, errors, and so forth in the present work are my own fault. XI Abbreviations CIT Corporate Income Tax CRD Chihou Roku Dantai FAT Fixed Assets Tax LAT Local Allocation Tax GTAC Government Tax Advisory Commission LCT Local Consumption Tax LDP Liberal Democratic Party LDPTAC Liberal Democratic Party Tax Advisory Commission LVT Land Value Tax LSAC Local System Advisory Commission MITI Ministry of International Trade and Industry MOF Ministry of Finance MOHA Ministry of Home Affairs PA Principal-Agent PIT Personal Income Tax SCAP Supreme Commander Allied Pacific VAT Value Added Tax Note that books and articles written in Japanese and by Japanese are cited, as is customary in Japan, with the author's family name first, but materials written in English by Japanese and by non-Japanese with Japanese names are cited as the name appears on the item in question. X l l Chapte r 1 Introduction Decentralization...is viewed as the "third major reform" in modem Japanese history following the Meiji Restoration and the postwar reforms..The Ministry of Home Affairs is committed to seeing that this project is a success. Ministry of Home Affairs Decentralization is all the rage in contemporary Japanese politics. By apt coincidence, this trend waxes in the twilight of a century of aggressive centralization, which fostered Japan's singularly goal-oriented political economy and propelled it past an astonishing array of milestones. Japanese reformists have long advocated devolution for more habitable and democratic communities, and lately assert that ageing and internationalization make more diversified local services imperative. But the novel aspect of the 1990s is that even the conservative establishment, the heirs of the developmentalist state, openly blame their illustrious legacy for the stalled economy and stifling uniformity that typify the nation's enervation at century's end. 1 Hence many observers believe the 1995 Law to Promote Decentralization (Bunken Suishin 1 A good recent example of progressive argument is that of Social Democratic Party of Japan politician Iwamoto Hisato, Chihoubunken de wa nakereba, Nihon wa dame ni nam [Without Decentralization, Japan is Done For], Tokyo: Diamond, 1996. A review of the broad range of arguments for 1 Hou) heralds -- to recycle a hoary cliche -- the sunset of Japan's centralized state. But beneath the flood of supportive discourse, evidence from the pragmatic world of Japanese politics suggests that numerous hurdles confront decentralization and especially its sine qua non, fiscal decentralization. 2 The 1980s campaign of "administrative reform" looked to conservative decentralist ideals and thus largely cut fiscal resources rather than devolve control over them. 3 Moreover, the 1995 enactment of the Law to Promote Decentralization, the structure of its associated deliberative committee, and the committee's continuing research on the various facets of decentralization have all been subjected to incessant rear-guard action by central-state bureaucrats and their political allies. Fiscal decentralization is an especially difficult area,"because the powerful ministries of Finance (MOF) and Home Affairs (MOHA) oppose decentralization can be found in the May, 1996 Japan Research Re vie w special edition "Chihou Shuuken Tokushuu" (Local Centralization). 2 Note the d iscussion in Sakamoto Chuuji, "Seifu Aida Zaisei Kankei no Kaikaku to Bunkenka" [Decentralization and Reform of the Intergovernmental Fiscal System], Okayama Daigaku Keizai Gakkai Zasshi 26 (2) 1994. Also, Hashimoto Touru argues that most f iscal reforms in the postwar period detracted from local autonomy; see, "Chihouzeisei no Genjou to Mondai" [The Current Structure and Problems of Local Taxes], Osaka Gakuin Daigaku Keizai Ronshu, Vol. 9, No. 1, April 1995. In a survey of 47 prefectural governors by the Asatii newspaper, 22 governors identified f inances as a major obstacle confronting decentralization and 20 foresaw problems due to the opposition of central-state bureaucrats who would lose their authority; in "Zaigen, kanryou o kenen: bunkenka 47 chiji ni kiku" [Worries about f inances and bureaucrats: 47 governors are heard on the topic of "decentralization], May 21, 1995, p. 1. 3 Hi ro ta Masao, "Chihou Gyoukaku Taikou no Sakuteijoukyou to Mondai" [Current Planning and Issues in Local Administrative Reform], in ToshiMondai, Vol. 87, No. 3, March 1996, especial ly pp. 39-40. Kimura Osamu's calculations show that the policy of cutting specif ic subsidies saw only 11.5 percent of 3.423 trillion yen shifted over to general subsidies. The vast bulk of the cuts therefore "effectively imposed additional burdens on local resources"; see "Chihoubunken to Chihouzei," in Osaka Shiritsudaigaku Keizaigaku Zasshi, Vol. 96, No. 3-4, March 1996. 2 decentralized taxation and are not, at present, eager to confront other ministries and their allied politicans over cuts in targeted subsidies. 4 The continuing lack of progress belies the ostensible consensus on decentralization, and strongly suggests that the topic merits detailed study. Fiscal Sociology and Fiscal Politics The gap between rhetoric and action gives ground enough to study the politics of fiscal decentralization in Japan, but a larger reason is the unparalleled importance of fiscal politics. The budget of the modern state is a massive fulcrum that shapes economy and society, thus affecting the lives of all citizens, whether rich, middle-class or poor. Moreover, the state's taxes and transfers produce almost diametrically opposed politics. Most people would readily pass the buck when it comes to paying the "the price of civilization," but -- in an ironic display of moral hazard -- even the most loyal acolytes of laissez-faire find a good appetite at the public trough. S o spending is usually welcomed, 4 According to a member of the committee, who must remain anonymous, opposition to serious reform greatly attenuated its March 29, 1996 mid-term recommendations. The report avoided specif ic goals concerning f iscal decentralization, whereas an earlier draft proposal, leaked to the Mainichi Shinbun , contained much stronger recommendations; see "Kokko hojofutankin, zeizaigen kaikaku yoso youshi" [The main points for reform of taxes and subsidies], March 23, 1996, p. 2). The mid-term report and comments on it can be found in "Zeizaigen nado, guttaian shimezu: chihou bunken suishinin chuukan houkoku" [The mid-term report of the Committee for Promoting Decentralization: No concrete proposals for f inances and other areas], Mainichi Shinbun, March 30, 1996, p. 2. See also "Bunya betsuni guttaian: chihou bunken chuukan houkoku, kongo no giron ni nayami" [Mid-term report on decentralization: the debate is bogged down], in Mainichi Shinbun, March 16, 1996, p. 3, Sano Masato, "Chihoubunkenin no chuukan houkoku: Juumin jichi no hasso ga kihaku" [Mid-term report on decentralization: Dilution of the ideal of popular democracy], Nihon Keizai Shinbun, April 1, 1996, p. 7, and Aoki Shisei , "Chihou bunken suishinin, hiaringu shuuryou: shouchou no teikou, izen netsuyoku" [Hearings wrap up for Committee on Decentralization: central agency opposition strong as ever], in Nihon Keizai Shinbun, June 24, 1996, p. 31. 3 and often heartily encouraged, at least by its recipients. Yet woe betide the government that would enact or increase a levy deemed by the many or enough of the powerful to be onerous or unfair. Taxes are therefore politically fascinating because they must be imposed, lest the state collapse and the social order, too, for want of the state; but perhaps more than any other instrument of government, taxes can mobilize overpowering opposition, whether from corporate boardrooms or otherwise politically atomized citizens. Fiscal instruments influence the state as well. The most spectacular expression of this influence arises when opposition to taxes brings down governments or forces sweeping changes in bureaucratic agencies and their cherished traditions. On the other hand, political coalitions' strong support of expenditures bolsters the bureaucratic and political institutions that manage them. A relatively unhindered capacity to extract levies from fertile tax fields will further bind this reciprocal pattern of supportive interests. Control over plentiful fiscal resources can thus advance some state agencies, strengthening them like well-trained muscles or, where fiscal discipline is lacking, merely enlarging them like corpulent limbs, often at the expense of atrophy elsewhere. Fiscal instruments are particularly germane to research on the politics of intergovernmental divisions within the state, as the flow of resources carries more explanatory force, for example, than the often merely nominal provisions of constitutions and the nomenclature that confidently distinguishes federal from unitary forms of government. 4 At a higher level of abstraction, such pluralist and institutional fiscal politics manifest themselves within a tax regime. Rudolph Goldscheid, the founder of fiscal sociology, sought to express this concept as he wrote in 1917, pondering the fiscal collapse of the Austro-Hungarian Empire. Goldscheid focused on "the degree to which the characteristic features of the State as such depend on the evolution of public finances." Joseph Schumpeter soon followed Goldscheid's lead with a masterful account of the "crisis of the tax state,"5 noting that: The public finances are one of the best starting points for an investigation of society, especially though not exclusively of its political life. The full fruitfulness of this approach is seen particularly at those turning points, or better epochs, during which existing forms begin to die off and to change into something new, and which always involve a crisis of the old fiscal methods. Close attention to fiscal issues, from the various levels of analysis, can therefore reveal progressively deeper currents of interest beneath the froth that constitutes so much of politics. One would expect, then, that economists' high regard for the importance of public finance has flowed copiously over disciplinary boundaries and reached a proportionate level especially in the field of political science. But taxation in particular remains a marginal area of 5 Goldscheid 's work is in German and was written in 1917; it has partially been translated as "A Sociological Approach to Problems in Public Finance," Classics in the Theory of Public Finance. Ed, by Richard Musgrave and Allan Peacock. New York: St. Martin's Press , 1958. Schumpeter's contribution dates from1918, and in translation is "The Crisis of the Tax State," International Economic Papers, 1954. 5 political study, leaving a vast terrain for fruitful and necessary research. The politics of intergovernmental fiscal relations are even more neglected than fiscal politics in general, as the work that does exist largely takes the central state's tax regime as the point of departure. This is in spite of the seemingly inexorable increase in the role of subnational governments within vastly different fiscal regimes. However, the flourishing literature in the once-dormant interdisciplinary space occupied by fiscal sociology promises to expand our understanding of the socio-political effects of taxes and transfers.6 The concept of the fiscal regime has returned, for some scholars because the contemporary period is -- like Goldscheid's era of wartime fiscal exhaustion ~ one of transition to a new and as yet unknown terrain of state finance. 7 Intergovernmental fiscal relations fits well within the ambit of concern of this resurgent school of thought, which highlights the historical crisis that is the common backdrop to the great variation among intergovernmental fiscal regimes. Fiscal history provides an essential context for understanding contemporary issues and institutions8 and 6 An excellent review of f iscal sociology is found in John L. Campbell, "The State and Fiscal Sociology," in the Annual Review of Sociology 19, 1993. Examples of new work in the field are Jinno Naohiko, "NihongataZei,.Zaisei Shisutemu" [The Japanese Fiscal System], in Gendai Nihon KeizaiShisutemu no Genryuu[The Sources of the Contemporary Japanese Economic System]. Ed. by Okazak i Tetsuji and Okuno Masahiro. Tokyo: Nihon Keizai Shinbunsha, 1993, Yahya M. Sadowski, The Heirs of Pharoah: The Fiscal Sociology of Democratization in the Arab World. Washington: The Brookings Institution, forthcoming, and the articles in Funding the Modern American State, 1941-1995. Ed . by W. Elliott Brownlee. New York: Woodrow Wilson Centre Press, 1996. 7 This is the core argument of Funding the Modern American State, 1941-1995. Ed . by W. Elliott Brownlee. op. cit. 8 By "institutions," I mean meso-level organizations and f iscal instruments. This contrasts with the rational choice micro-oriented definition of institutions, which according to Randall Calvert, is "any of the rules of 6 the roles they play; and it can also, as we will see, unearth clues for reading the emergence of a true fiscal crisis in the current regime. Fiscal Decentralization and Japan's 1940 System Of course, proposals to devolve power over the public purse have in recent years emerged in many countries. 9 Fiscal decentralization's appeal seems largely to stem, in structural terms, from economic globalization's erosive effects on the role of the central state 1 0 and, in political terms, from the ironic confluence between left radical paeans to localism and conservative Utopias of decentralized self-reliance. Thus French Socialists in the early 1980s and America's ginger group of Gingrich Republicans in 1994 were both zealous advocates of fiscal decentralization, with the former seeking to empower the people and the latter ardent to disempower the federal state. the game by which individuals in society find themselves confronted when contemplating action." See Randall Calvert, "The rational choice theory of social institutions: cooperation, coordination, and communication," Modern Political Economy: Old Topics, New Directions. Ed . by Jeffrey S. Banks and Eric A . Hanushek. Cambridge: Cambridge University Press , 1995, p. 217. Also, note Tatebayashi Masahiko, "Gouriteki Sentaku Seidoron to Nihon Seij ikenkyuu" [Rational Choice Institutionalism and Research on Japanese Polit ics], Kyoto Daigaku Hougaku Ronshuu, Vol. 137, No. 3, June 1995, especial ly pp. 68-9; and Masahiko Aoki , 'Towards a Comparative Institutional Analys is : Motivations and Some Tentative Theorizing," The Japanese Economic Review, Vol. 47, No. 1, March 1996. 9 The World Bank program division on f iscal decentralization notes, on its World Wide Web home page, that "eighty countries [have] a mandate to restructure intergovernmental f iscal relations." The URL is http://gopher.worldbank.org./html/prddr/prdhome/issues.html. On North America, see Jeffrey Simpson, "Decentralization is the rage everywhere, despite its ironies," Globe and Mail, August 21, 1996 (np). 1 0 Susan Strange d iscusses several trends that are undermining the role of the national state, including its f iscal preeminence, in The Retreat fo the State: The Diffusion of Power in the World Economy. Cambridge: Cambridge University Press, 1996. 7 Even in the relatively fluid polities of North America, though, fiscal decentralization generally receives more lip service than concrete action. S o perhaps Japan's lack of progress simply reflects a general tendency to inertia in the fiscal state, with an added degree of ossification from Japan's alleged cultural aversion to change. But this supposition overlooks the immense institutional differences in intergovernmental fiscal regimes. The Japanese case is particularly interesting in this respect, and worth an extended study, because of its towering and unique hurdle of inter-ministerial tax competition. 1 1 M O F a n d MOHA split jurisdiction over, respectively, Japan's national and local tax bases. Entrenched on these vast tax fields, the ministries naturally oppose changes that would detract from their authority. And because Diet politicians have few incentives to reform the fiscal regime, bureaucratic interests dominate the agenda in this important area of policymaking. My argument unfolds through depicting the institutions and fiscal politics of Japan's "1940 fiscal system." Fiscal sociology work on Japan's intergovernmental tax-transfer system shows that it derives more from wartime reforms than, as is commonly believed, those of the postwar Occupation. Most countries exhibit institutional legacies from the war, in fact, but an expanding " T h e s e confl icts are generally referred to as zeigen arasoi, or "struggles over tax sources," as in "Kouteishisanzei age: kuni to chihou no zeigen arasoi utsusu" [Increasing the Fixed-Assets Tax: it reflects the centre-local fight over taxes], Nihon Keizai Shinbun, December12, 1991, p. 3. Interestingly enough, there is little written in English or Japanese about the tax turf wars between MOF and MOHA, pe r se , as opposed to the less institutionally accurate depiction of the conflicts as being between the central and subnational levels of government. Concerning the rather common phenomenon of bureaucratic turf wars in Japan, see Chalmers Johnson, "Telecom Wars: How Japan Makes Policy for 8 literature suggests Japan is marked to an unusually high degree by transwar phenomena. As the term "transwar" implies, these studies trace the origins of much of the contemporary political economy back to innovations made under the press of the Pacific War and the earlier invasion of Ch ina . 1 2 Much of this work on Japan's transwar institutions can and has been critiqued on empirical grounds, and I make no attempt to defend it all here. 1 31 do not, in any case, argue that the 1940 fiscal system is in all respects the same as the structure adopted on the eve of the Pacific War. The wartime regime itself underwent various adjustments even before Japan's defeat in the summer of 1945 and the prompt imposition of American notions of fiscal democracy. Similar to the manner in which other students of fiscal sociology discuss regimes as frameworks, my point is that in Japan centralized fiscal control and the general structure of fiscal instruments survived U.S. Occupation reforms or were reconstructed during the postwar "reverse course" rollback of the Occupation's doings. High Technology," Politics and Productivity: How Japan's Development Strategy Really Works. Ed . by Chalmers Johnson, Laura Tyson, and John Zysman. Ballinger Publishing, 1988. 1 2 The f iscal sociology account is Jinno Naohiko, "Nihongata Zei, Zaisei Shisutemu" [The Japanese Fiscal System], op. cit. Other transwar studies include Chalmers Johnson, MITI and the Japanese Miracle. Stanford: Stanford University Press , 1982, John Dower, "The Useful War," Daedalus, Summer, 1989, and contributions mGendai Nihon Keizai Shisutemu no Genryuu [The Sources of the Contemporary Japanese Economic System], op. cit. andNoguchi Yukio, 1940-Nen Taisei: Saraba "Senji Keizai" [The 1940Sys tem: Goodbye to the Wartime Economy]. Tokyo: Toyo Keizai Shimbunsha, 1995. On wartime learning in Germany, note Simon Reich, The Fruits of Fascism: Postwar Prosperity in Historical Perspective. Ithaca: Cornell University Press, 1990. 1 3 A critique of the 1940 sys tem argument can be found in Nagashima Osamu, "Sengo Nihon Keizaikenkyuu no Shinchouryuu: 40-Nen Taiseiron oChuushin toshite"[The New Trend in Research on the Postwar Japanese Economy: t h e Focus on the 1940 System], Keizai Kagaku Tsushin, No. 80, February, 1996. 9 Fiscal sociology thus offers a persuasive account of the crisis of Japan's prewar tax state and its resolution through fiscal centralization. But among Schumpeter's many insights in political economy was that the state's fiscal structure is as potent an independent variable as the social forces and institutions that seek to recast it.14 My aim is therefore to build on the fiscal sociology research done on Japan through a meso-level, historical institutionalist study of the 1940 fiscal system and its contemporary tax politics. Taking my cue from Schumpeter, I examine how the preferences of Japanese bureaucratic and political institutions shape and are, in turn, shaped by the intergovernmental tax structure.1 5 Thus two major claims are made here that challenge much existing political science work on Japan. The first argument is that studies of intergovernmental relations have been rather too quick to discard lessons from the older orthodoxy's rigid models of centralized dominance over local policymaking. Certainly a major contribution of the recent critical literature and its aggressively empirical approach has been to show that local governments in Japan, as elsewhere, have evolved various strategies for expanding their own policymaking autonomy within the context of a high degree of formal 1 4 See Joseph Schumpeter, "The Crisis of the Tax State," op. cit. 1 5 By comparison, note the "democratic-institutionalist" approach used by students of the US f iscal regime in order to express the influences of a broad complex of ideas (esp. "ability to pay"), socia l forces, and governmental institutions. See the chapters in Funding the Modern American State, 1941-1995. Ed. by W. Elliott Brownlee. op. cit., especially Brownlee's d iscussion of methodologies in "Reflections on the history of taxation." 10 centralization. But in its close embrace of a pluralist methodology, it has unduly downplayed the significance of the central state's fiscal controls. The centre's formal and informal powers are of primary importance in shaping broad swathes of fiscal politics and policymaking in Japan, areas that bear directly on the prospects for fiscal decentralization. My case studies show that MOF and MOHA's tax competition induces them to impede moves toward local fiscal autonomy. They compete because they represent separate levels of the state in a condition - a universal one, though of varying intensity - where fiscal resources are generally scarce. In consequence, they strive to maintain, and if possible augment, their potent levers of administrative power, for complacency could mean effectively ceding them to the other ministry or to the local authorities. Inter-ministerial competition, the product of friction between fiscal jurisdictions, reduces flexibility in policymaking and conditions in large measure the choices available to local governments. A corollary of the above is my second major claim: Studying Japan's inter-bureaucratic tax competition reveals a large area of bureaucratic discretion. Tax competition has serious economic and political consequences for national and subnational tax policymaking. To take one currently very active area of contention, discussed in detail in chapter 7, Japan underwent a property-tax revolt in the mid-1990s largely because of MOF and M O H A ' s struggle over asset taxation. This assertion of bureaucratic discretion challenges, for 11 example, recent rational choice work on Japan which dismisses, or at least severely circumscribes, the autonomous role of state agencies in policymaking. 1 6 Though my account centres on bureaucratic activism and preferences, I do not suggest that these alone explain outcomes in Japan's intergovernmental tax politics. Institutions matter -- in this case, bureaucratic institutions in particular do because their conflicts thunder across a vast fiscal terrain, affecting an array of taxes on the fields of income, consumption, and assets. But fiscal sociology and historical institutionalism play no favourites when it comes to selecting independent variables, and instead caution the analyst to consider a wide range of interests and contingent phenomena. Thus I would emphasize that my argument is not another partisan challenge from the "statist" side of the highly polarized and ultimately pointless debate on whether bureaucrats or politicians run Japan. Parties and politicians are powerful actors in contemporary Japan, but it is a non sequitur to insist that bureaucratic interests and influence are therefore irrelevant. The reverse equation is equally illogical, as presuming bureaucratic power does not require that one also assume elected politicians are thus enfeebled, barely able even to wield a rubber stamp. In researching intergovernmental fiscal politics such gross methodological reductionism is of little value, and thus I have opted for eclectic In order to avoid a long digression from the main trend of argument in this dissertation, I have incorporated an extensive discussion of these works in the Appendix. 12 and nuanced approaches. 1 7 Comparative Tax States Japan's intergovernmental fiscal structure is huge, and includes over half of 1994's 86.5 trillion yen in total government tax revenues (i.e., exclusive of social security contributions 1 8). In 1994 local taxes collected 32.5 trillion yen, or 37.6 percent of total tax revenues, and over half of the central state's current expenditures were transfers to the local authorities, 1 9 via the Local Allocation Tax (LAT), specific subsidies, and other means such as the Consumption Transfer Tax. 2 0 The 31.9 trillion yen in total transfers from the central to the local governments saw 16.8 trillion yen delivered through the LAT and 13.6 trillion yen via specific subsidies. 2 1 If we add local tax revenues and those from 1 7 Examples include journalistic assertions of bureaucratic dominance, the "Japan Inc." model, Marxian theories of "state monopoly capitalism, and the effortless post hoc claims of political control from the rational choice principal-agent school . By contrast, note Peter Evans 'de fence of "the eclectic messy center" of comparative politics; in Atul Kohli et al., "The Role of Theory in Comparative Poli t ics," World Politics 48, October 1995, p. 2. See also J o h n C Campbell's comments on the rational choice versus statist debate in Japanese politics in Social Science Japan, No.8, 1997. 1 8 Japan col lects 35 percent of its tax revenues as social security contributions, and revenue stat ist ics make it clear that contributions are increasingly important aspects of the f iscal state in most O E C D countries. But social security issues are outside the intergovernmental ambit of concern of this dissertation. 1 9 Note that in Japanese usage, the "local authorities" or jichitairefers collectively to Japan's 2, 586 towns and vil lages, 655 cit ies, 47 prefectures, and Tokyo's 23 wards. I will adopt this usage to refer to subnational levels of government in other states as well. 2 0 In 1994, total expenditures were just below 1,414 trillion yen, but because of the prolonged Heisei Recess ion and related income-tax cuts over 50 trillion of the necessary revenues came from public debt. See "Chihouzei Kankei Shiryou Handobukku, Heisei 8-nen, 10-gatsu" [Handbook of Materials on Local Taxation, October 1996], Tokyo: Chihou Zaimukyoukai, 1996, p. 4. 2 1 Tax stat ist ics can be found in the M O F Tax Bureau annual An Outline of Japanese Taxes, Tokyo: Tax Bureau, Ministry of Finance. The 1994 data are calculated from See "Chihouzei Kankei Shiryou 13 the LAT together, along with the 1.9 trillion yen from the Local Transfer Tax, the result for 1994 is 52.7 trillion yen. Using the O E C D ' s 1994 exchange rate of 102.2 yen to the American dollar, we can see that Japan's total local tax and Allocation Tax funds amounted to about 517 billion U.S dollars, a sum larger than most countries' entire annual economic output. Hence, even though Japan's total tax burden in 1994 was only 27.8 percent of G D P , as compared to the 1994 O E C D average of 38.4 percent, the country's enormous economy delivered ample revenues. Indeed, Japan's tax revenues for 1994 were, when measured in American dollars, equivalent to 19.1 percent of total O E C D tax revenues. 2 2 The Japanese fiscal state's atypical features are displayed in Figure 1.1. The figure shows that 70 percent of Japan's government-sector spending takes place at the local level, but that over one-third of overall taxation is collected by the local authorities. This rate of local spending is extremely high, even in comparison with the federal states, as the American, Australian, Canadian, and West German examples indicate. In addition, the "fiscal transfer" component of Figure 1.1 measures the percentage of the central state's operating expenditures that is transferred to subnational governments as specific and general subsidies. In the Japanese case, 50.7 percent of the central state's operating expenditures are a fiscal transfer to the local Handobukku, Heisei 8-nen, 10-gatsu" [Handbook of Materials on Local Taxation, October 1996], op. cit. p. 54, which also describes the Local Transfer Tax. 14 authorities. This rate is nearly twice the U.K.'s 25.8 percent, which records the next highest fiscal transfer. In other words, Japan has a comparatively massive downward flow of funds in order to provide the resources for an unusually heavy rate of local expenditures. Figure 1.1: Local Expenditures. Revenues and Fiscal Transfers, 1994 80 i 1 Note: US data is 1993, "Austral"=Australia. Sources: OECD: National Accounts, Detailed Tables, Volume 2, 1982-1994. Paris: 1996, OECD: Revenue Statistics, 1965-1995. Paris: 1996. Refer to tables on pp. 75 and 90 \v\OECD Revenue Statistics, 1965-1995, Paris, O E C D Publications, 1996. 15 As is also clear from the figure, Japan's percentage of subnational taxation to total taxation, excluding social security contributions, resembles that of the federal states and Sweden more than the unitary states of France, Italy, and the U.K., whose percentages are quite low. Moreover, using O E C D figures for 1994 that include social security contributions, Japan's 24.4 percent level of local taxation as a proportion of total tax revenues is certainly well under the 31.5 percent average of all O E C D federal states and the 27 percent average for the Scandinavian unitary states, including Denmark, Finland, Norway, and Sweden. But the rate of local taxation for all O E C D unitary states was 11.8 percent, which Japan exceeded by more than double. I would repeat that overall tax burdens among these states vary considerably. Japan is on the extreme low end of total tax burdens among the O E C D countries, as in 1994 all levels of government in Japan taxed at a combined rate of 27.8 percent of G D P whereas the O E C D average was 38.4 percent. The US tax burden of 27.6 percent was close to Japan's, but other countries were generally much higher, with Australia taxing at 29.9 percent, Canada 36.1 percent, France 44.1 percent, W. Germany 39.3 percent, Italy 41.7 percent, the UK 34.1 percent, and Sweden 51.0 percent. The issue here, however, is not total levels of taxation, but the relative flow of tax revenues and transfers to support local government spending, which is inordinately high in Japan. Moreover, Japan's economy is so large that even the public sector's comparatively low tax bite represents vast sums of money with commensurate 16 political significance. In gross fiscal terms, then, most federal states resemble Japan to some extent in having a large scale of local spending that is in part supported by fiscal transfers from the federal government. But the point to be emphasized here is that none of the federal states approaches Japan's level of fiscal transfer. Moreover, with the exception of Australia (whose case will be examined in more detail below), subnational governments in the federal systems included in Figure 1.1 enjoy a comparatively large sphere of fiscal autonomy. Local taxes and debt flotation in the U S , for example, are the prerogative of subnational governments in a system of clearly delineated powers; and so an agency like centralized Japan's MOHA, which oversees local debt and taxes, simply has no scope to exist. Sweden, a unitary state, also has fairly high rates of local spending -- the product of concentrating a generous outlay of universal services at the local level. But compared to Japan, Sweden has no intermediary institution such as MOHA and has a much stronger tradition of local fiscal democracy. One notable sign of this tradition was the replacement of the Swedish local property tax base with the more buoyant income tax, a point discussed in the following chapter. All data are from table 3, p. 73 and table 134, p..208, in OECD: Revenue Statistics, 1965-1995. Paris: 1996. Note also that Japan's 1994 tax levels are depressed due to recession and income-tax cuts. The 1990 figure, by contrast, is 31.3 percent. 17 As for fiscal institutions in other centralized states. France's Ministry of the Interior is often believed to wield a tight rein over local government. But there was considerable fiscal decentralization as a result of the Socialist Party commitment to refashion France in the early 1980s. Moreover, in 1994 local revenues in France were only 18 percent of total government revenues, and so offer a much smaller terrain for fiscal competition. 2 4 The U.K.'s case is even more remarkable in this regard, as the figure shows how miniscule is the weight of both local revenues and expenditures in the composition of the fiscal state. There is an exceedingly narrow base for inter-ministerial fiscal competition, both because of the small scale of local government and because the British Ministry of Finance holds most of the jurisdictional cards in the sphere of intergovernmental taxation. Moreover, confrontation in the U.K. between local and national government is generally played out between the Conservative and Labour parties, based on their contradictory philosophies concerning the proper role of the state as well as their desire to control local party organizations. 2 5 Jacques Le Cacheux and Laure Tourjansky, "The French Decentralization Ten Years On: Local Government Finances," in Local Government Studies, 1992, Volume 18, no. 4, and Peter Morris, "Local Government in a Centralized State," French Politics Today. Machester University Press : 1994. Morris notes that French local authorities still are highly dependent on the centre for revenues, but they "now have complete autonomy over a number of taxes, including those on property and bus inesses. As of 1987, departments and regions gained the freedom to set the rate of local taxes"; ibid., p. 100. Gerard Stoker, "The Battle Over Local Spending," The Politics of Local Government, 2nd Edition. London: MacMillan, 1991 and John Gibson, "British Local Government Finance under the Conservat ives," Local Government Studies, Vol. 18, no. 4, 1992. 18 Indeed, intergovernmental fiscal relations in the United Kingdom has a history of "particularly ruthless" strategies to marginalize local government, with Thatcherism in the 1980s being only the most recent example. For example, a century before the ascent of Thatcher serious differences in rural and urban fiscal capacities emerged, but as with "every Prime Minister since his time, Gladstone refused to give localities a more flexible tax even though a local income tax was recommended by his able financial advisor, Goschen." By contrast, Japan's wartime tax reform saw the Home Ministry (MOHA's predecessor) snatch a share of the income tax for subnational government and its own institutional ends. 2 6 Japan is also notable in that most developed countries in the postwar era decentralized their wartime fiscal systems. They did not completely dismantle centralized structures of income taxation, because of the pressures for funding the expansion of welfarism and the enlarged role of the national state in economic management. But they largely relaxed controls on their respective local tax regimes and made various reforms to institutionalize intergovernmental consultation and negotiation. Japan went rather in the opposite direction, in reaction to Occupation-era reforms and as part of the emphasis on husbanding resources for reconstruction and rapid economic 2 6 On the U.K., see Douglas E. Ashford, The Emergence of the Welfare States. Oxford: Basi l Blackwell, 1988, pp. 121-32. Concerning the Home Ministry and the income tax, see Jinno Naohiko and Andrew DeWit, "Institutions, Ideology, and the Intergovernmental Allocation of Taxes," University of Tokyo Faculty of Economics Discussion Paper Series 96-F-18, September 1996. 19 growth. Moreover, other states, including the U.K., have overtime made great changes to their intergovernmental regimes, but Japan's postwar system has been unusually stable, with only incremental changes occurring in the wake of the rollback of the Occupation's reforms. 2 7 Table 1.1 gives evidence of another unusual feature of the Japanese state: its heavy orientation towards public works spending, in particular at the subnational level. The first column in the table shows the percentage of Gross Domestic Product consumed by general government sector fixed capital formation (which is largely public works). Japan's 6.6 percent of G D P is nearly twice that of France and Sweden, which spend 3.4 and 3.3 percent, respectively, of G D P on capital formation, a level that is itself noticeably greater than the rough average of about 2.0 percent for the rest of the sample. The second column shows that gross capital formation constitutes almost a third of Japan's local authorities' expenditures. This means that subnational government in Japan is to a large extent a conduit for public works spending, more so proportionately than in any of the other countries sampled. Akizuki Kengo, "Institutionalizing the Local System, The Ministry of Local Affairs and Intergovernmental Relations in Japan , " The Japanese Civil Service and Economic Development. Ed. by Hyung-ki Kim, et al. Oxford: Oxford University Press, 1995, p. 338. 20 Table 1.1: Government Fixed Capital Formation, 1994 General Government Fixed CapitalFormation/GDP (%) Subnational Fixed Capital Formation/ Subnational Expenditures(%) Aust ra l ia 2 . 0 9 .7 C a n a d a 2 . 3 6 . 9 W e s t G e r m a n y 2 .0 10.1 U . S . 1.6 9.1 J a p a n 6 .6 3 1 . 8 F r a n c e 3 .4 2 3 . 0 Italy 2 . 3 1 0 . 7 S w e d e n 3 . 3 7 .0 U . K . 2 . 0 8.1 Source: OECD: National Accounts, Detailed Tables, Volume 2, 1982-1994. Paris: 1996. Japan's immense public works spending is a core characteristic of its fiscal state. Figure 1.2 displays various countries' government-sector spending as a percentage of gross domestic product, 2 8 and Japan stands out as unusual in all categories save its annual outlay to cover interest on the public debt. The U.S figures are for 1993 and the Japanese data for the 1990s are elevated, due to anti-recession measures, from the 1980s rough average of a little over 5 percent of G D P . Other countries' data vary as well, but within narrower limits. 21 Figure 1.2: Government Sector Spending as a Percentage of GDP. 1994 Note: US data is 1993, "Austral"=Australia, "Total Government" is total national and subnational spending as a percentage of G D P . Source:OECD: National Accounts, Detailed Tables, Volume 2, 1982-1994. Paris: 1996. The figure shows that the Japanese government's final consumption expenditures (these are mostly for civil servants' salaries) are the lowest of the nine countries, and Japan's "social transfers" (government spending on social security and social assistance) are the second lowest, next to Australia. Japan 22 is therefore clearly not an interventionist welfare state of the European type, where high public sector employment, service provision, and transfers are reflected in elevated levels of spending on final consumption and social transfers. And even though the Anglo-American states are generally smaller than those in Europe, they cannot match Japan for stinginess when it comes to the number of civil servants and the limited scale of redistribution through social transfers. Nor do any of the other countries, whether Anglo-American or Continental, come close to Japan's outlays for fixed capital, especially as a proportion of total government sector spending. Rapid ageing of the Japanese population, working in tandem with the country's generous social security programmes, does raise the possibility of a heavy expansion in income transfers. If the transfer programmes remain essentially as they are, or are enhanced, then Japan will be on a trajectory towards the European-style welfare state. 2 9 But for the present the Japanese fiscal state's statistics rather confirm several critics' claim that it is a "construction state" (doken kokka), one that redistributes an inordinate amount of national wealth through regional spending on infrastructure. These critics, and others who analyze the political economy of state spending in Japan, highlight the extensive porkbarrel role of public works. 3 0 This is one of the arguments in John C Campbell, How Policies Change. Princeton: Princeton University Press, 1992. For example, Gavan McCormack's chapter on "The Construction State,"The Emptiness of Japanese Affluence, London: M.E. Sharpe, 1996. 23 Such claims are correct and illuminating, but from a more historical and macrosociological point of view another interesting fact is that the central state has long sought sociopolitical stability in the regions. As one expert on Japanese public finance notes, suggesting a large measure of path dependence, "aid for rural districts was a policy that originated with the bureaucracy." Funding thus flows mainly to key rural interest groups, such as builders and farmers, not only because of postwar politicization but also because the Japanese state has a tendency to innovate in political institutions and public works designed to bolster stable social structures in the local areas. 3 1 The point here is neither to set up an argument of Japanese uniqueness nor to suggest that bureaucrats unilaterally run the Japanese fiscal state. Rather, my aim is to highlight differences relevant to a fiscal sociology account. All states are concerned with social stability, but state managers -- including top bureaucrats and ruling-party politicians -- promote it in very different contexts and with disparate levels of zeal . This is also true of economic growth: all states desire it, but some, such as Japan, have pursued it as the overriding objective of public policy. The counterpart of Japan's relentless quest for 3 1 The quote is from Noguchi Yukio, in "Time to Grow Beyond the 1940 System," Tokyo Business Today, September 1995, p. 9. For details on the Japanese welfare state and the state's emphasis on maintaining stability, see Jinno Naohiko, "Nihongata Zei, Zaisei Shisutemu" [The Japanese Fiscal System], in Gendai Nihon Keizai Shisutemu no Genryuu, op. cit., and "Nihongata Fukushi Kokka Zaisei no Tokushi tsu" [Characteristics of the Japanese Welfare State], Fukushi Kokka Zaisei no Kokusai 24 greater economic efficiency in export-oriented industries, the key to the economic miracle, has thus been a concern for political effectiveness in expenditures. Unlike the semi-authoritarian measures favoured in other high-growth regimes throughout Asia, such as stridently anti-liberal Malaysia and Singapore, the postwar Japanese state has to a large extent bought off potentially threatening opposition with the pacifier of public spending. 3 2 Much of this spending has gone to public works, whose legacy includes woefully inefficient construction firms, 3 3 entrenched and expanding networks of political corruption, 3 4 and a monumental load of debt. The system is undergirded by the fiscal instruments of the 1940 regime, which give incentives for MOHA and MOF to oppose fiscal decentralization. But in addition, the massive flow of pork reinforces these incentives, as MOHA seeks to maintain order in local finances and MOF strives to preserve its large role in national-level fiscal policymaking. Hikaku [Comparative Welfare-State Finance]. Ed. by Hayashi Takehisa and Kato Eiichi. Tokyo: University of Tokyo Press, 1992. On As ian regimes, see Political Oppostion in Industrializing Asia. Ed . by Garry Rodan. London: Routledge, 1996. Also, Kent Calder argues that the ruling LDP emphasized distributory politics in order to quell opposition. Though he sees the bureaucracy as technocratic agents of rationalism and austerity, this would seem specifically the orientation of MOF. MOHA, for example, does not oppose f iscal expansion and public works so long as excess ive costs are not placed on the local authorities. See Calder, Crisis and Compensation: Public Policy and Political Stability in Japan. Princeton: Princeton University Press, 1988. O E C D figures show 10.6 percent of Japan's working population in construction in 1994, whereas the percentage in Canada was 4.8, the US 4.3 (1993), Australia 5.4, France 6.4, Germany 7.0, Italy 6.8, Sweden 4.7, and the UK 3.8. The data are from OECD: National Accounts, Detailed Tables, Volume 2, 1982-1994. Paris: 1996. Note also that according to a 1993 study by Japan's Ministry of Construction, the cost of building a home -- exclusive of the price of the plot of land -- in Japan averaged twice that in the United States. See "Comparing Home Construction Costs , " Japan Link, Vol . 2 Issue 14, July 22, 1994. Japan's top domestic news story for 1996 was the arrest of former vice minister of health and welfare, Nobuharu Okamitsu, who apparently took 60 million yen in bribes in exchange for providing government subsidies to construct old-age homes. Underlying the incident is the increasing importance of this area as yet another outlet for public works spending. See "Tokuyou, kensetsu tanka o kensa e" [Towards unit-cost construction for special-care homes], Mainichi Shinbun, January 7', 1997, p. 1. 2 5 Japan's Taxes and Fiscal Transfers One of the main means through which central agencies, and especially the Ministry of Construction, target public works at the local level is via specific subsidies. These subsidies are among the principal elements in the continuing debate over decentralization, and as I noted earlier provide over 40 percent of the intergovernmental fiscal transfer. Specific subsidies are much loathed in the press and by progressives because they are, as classic political pork, often wasteful means by which Diet politicians and their bureaucratic allies can impose developmentalist and other agendas on local communities. They are also carrots that subnational governments will go to great lengths to secure. Municipal governments at times get so caught up in the feeding frenzy on budgetary allocations that they spend more in lobbying the centre than the subsidy they are after is actually worth. Moreover, many of the specific subsidies are linked to the delegated responsibilities (inin jimu) through which central agencies are able to make local authorities, and especially the local executive, act as agents for the implementation of centrally determined tasks. Critiques of this delegation of tasks centre on the problems engendered because local executives are thus removed from oversight by their legislatures and local resources and government employees are effectively put at the disposal of central agencies. 26 Nor are these marginal problems, as "the Local Autonomy Act specifies, in no less than 500 separate statutes, the tasks that the state requires prefectural governors and mayors to perform." 3 5 Targeted subsidies are thus an important part of the 1940 fiscal regime because of their role in centralizing the fiscal regime, stabilizing the local areas, and mesmerizing politicians. But the present work will not discuss the subsidies in more detail than is necessary to highlight their role in the fiscal regime and especially the influence they have, in the aggregate, on MOF and MOHA's fiscal competition. One reason I take this approach is that researching the institutional channels through which the distribution of Japan's 2200 different types of specific subsidies is determined would expand this dissertation to an unmanageable size because many line ministries, groups of Diet politicians, local governments, and private interests are involved. 3 6 The second reason is that the subsidies supplied 15.3 percent of local revenues in 1994, down from 22.8 percent in 1979. Though statistics for 1995 and 1996 will likely show an increase due to anti-recession measures and the Liberal Democratic Party's concern not to fall again from the grace of government, the long-range prospect is for further attrition in specific subsidies. 35 The Government and Politics of Japan. Ed. by Hitoshi Abe, Muneyuki Shindo, and Sadafumi Kawato. [Trans, by James White]. Tokyo: University of Tokyo Press, 1994, p. 64. 3 6 Many works analyze the politics of subsidies in Japan. See , for example, Hirose Michisada, Hojokin to Seikentou [Subsidies and the Governing Party]. Tokyo: Asahi Shinbunsha, 1981. The number of subsidies was taken from "Hojokin no seiri, kanbandaore" [A big show of cutting subsidies], Nihon Keizai Shinbun, August 25, 1996, p. 1. 27 This decline means a corresponding reduction in explicitly "political" intervention in local affairs and, as I argue in chapter 5, a likely politicization of the Allocation Tax. 3 7 As a result, I instead devote my attention to the politics of relevant national and local taxes, and the increasing level of general subsidies (ippan hojokin) delivered through the Local Allocation Tax (Chihou Koufuzei). My case studies centre on the clearest examples of interministerial fiscal competition and opposition to decentralization. Though I will discuss these taxes and transfers in detail below, it is important at this point to show that they are not marginal and unrepresentative aspects of the overall tax state. The Local Allocation Tax referred to earlier transferred 16.8 trillion yen to the local authorities in 1994, or well over half of Japan's unusually high level of intergovernmental fiscal transfers. Moreover, the locally collected taxes that are also featured in my case studies provided, in 1994, 35.5 percent of total prefectural taxes and 53.5 per cent of total municipal taxes. And the inter-ministerial struggle over control of a share of the Consumption Tax involved a tax that supplied 7 trillion yen, or over 8 percent of total national tax revenues in the same year, a level that is expected to increase greatly in the coming The 1994 figure is an increase over the 14.4 percent recoded the previous year, and reflects public works spending to stimulate the economy. See figures for 1993 and earlier in Table 12.2 "Local Finance Revenue Structure," in Kengo Akizuki , "Institutionalizing the Local System," op. cit. Note also that according to an anonymous participant on the above-noted Committee for Promoting Decentralization, the specif ic subsidies were among the first items for f iscal reform. Progress is stymied on this front largely because MOF feels too burdened by the scandals of 1995 and 1996 to take on other ministries and their allied zoku giin in a contest concerning further cuts in specific subsid ies. 28 years. To be sure, fiscal competition and conflict between levels of government are at the very least universal tendencies, because state agencies all desire access to finite fiscal resources. But the important point is that the institutional context in which these tendencies play themselves out varies in politically significant ways. Japan, as we have seen, is unusual because it combines an enormous scale of fiscal transfers and divides central-ministry jurisdiction over vast local and national tax fields. This particular combination of features is not evident elsewhere, and in conjunction with the fiscal history discussed below, has resulted in bureaucratic manoeuvres to protect tax turf or gain control over sections of the contending agency's tax base. This phenomenon of bureaucratic tax competition, or trench warfare on the tax fields, helps maintain Japan's centralized system, as I demonstrate through my case studies. In order to explore these issues further, some definitions and a review of the most relevant literature are in order. Hence, in the next chapter I turn to consider the concept of decentralization, in particular fiscal decentralization, and then examine how it has been treated in political science works on Japan's intergovernmental relations. Chapter three discusses in detail the 3 8 From Table 50 "Chihouzei Shunyuu no Kousei no Ruinenhikaku" [Changes in the Structure of Local Revenues], in ZaiseiKin'yu ToukeiNenpo [Ministry of Rnance Statist ics Monthly], April, 1996, pp. 114-15. The data for the Consumption,Tax can be found on page 131 oWECD Revenue Statistics, 1965-1995, op. cit. 29 prewar fiscal order, the wartime reforms, and the postwar revisions. Chapter four sketches the structure of interests within the contemporary fiscal regime, while chapters five, six, seven, and eight analyze how the politics of specific transfers and taxes support the thesis I am advancing. My final chapter, the ninth, summarizes the arguments made and draws conclusions concerning Japan's intergovernmental tax politics and the role of the tax structure in fiscal politics. I would also add that the Appendix reviews the relevant theoretical approaches to fiscal politics, and shows why I have opted for macro and meso levels of analysis in examining Japan's intergovernmental fiscal regime. 30 Chapter 2 Institutions and Fiscal Decentralization Local public entities shall have the right to manage their property, affairs, and administration, and to enact their own regulations within the law. Article 94, The Constitution of Japan Decentralization is generally defined as the transfer of administrative responsibilities and the finances to perform them from national to subnational levels of government. Such a shift of resources and authority is inherently political, because it concerns the question of which level of the state should wield the most effective jurisdiction in specific areas of policymaking and over particular fiscal resources. Moreover, transfers of revenues and responsibilities do not necessarily go hand in hand. Central states often delegate responsibilities while retaining control over the associated revenues, a characteristic of Japan's system of vertical administration. Hence an apparent case of decentralization might not be substantive if the subnational governments do not have the means to make reasonably autonomous 31 decisions in the areas placed under their administrative purview.1 For this reason I argue that fiscal decentralization is the most important and politically contentious dimension of decentralization. I would emphasize, however, that this dissertation neither advocates nor opposes fiscal decentralization for Japan. Liberal political theory and neo-classical economics make, of course, strong normative arguments for fiscal decentralization as a means, respectively, to educate the citizen about democracy and promote the efficient allocation of resources when citizens have the opportunity to choose among a range of options. 2 But these are mere assertions derived from theory and may not hold true, depending on the context, in actual practice. The kinds of interests whose representation might be enhanced through decentralization must be considered before the change can be deemed to bolster the democratic character of the state. A notable case in this respect is the United States. America's variant of fiscal federalism abhors centralized administration, leaving the system very much in the hands of the bond markets and highly organized groups of electors who mobilize quite readily against property taxes, the fiscal base of Anglo-American local government. Thus even wealthy American communities can be rendered incapable of taxing themselves at what non-ideologues would 1 B.C. Smith gives a comprehensive treatment of these issues in the introductory and finance-related chapters of his Decentralization: The Territorial Dimension of the State. London: George, Allen and Unwin, 1985. 32 consider appropriate levels. The outcome is perfectly "rational" in the institutional environment of American local politics, because narrow fiscal self-interest is provided with a comparatively vast scope to express itself.3 Transfers are another problem area in America. Robert Kuttner notes this concerning the wide-eyed neoliberalism that forged the 1996 welfare bill: "bipartisan blather about state flexibility and experimentation is designed to obfuscate an inconvenient key fact: Washington assumed responsibility for public assistance sixty years ago because no other level of government had either the fiscal resources or political will to consistently do so." Federal programmes such as Aid to Families with Dependent Children have always been suspect in the American political milieu, for a host of reasons that certainly includes a strong, institutionally grounded faith in the virtues of fiscal decentralization. Without some degree of federal standards and sufficient resources, a wholesale devolution of responsibilities opens the door to a recrudescence of the human tragedies that called forth the federal presence in the first place. As Kuttner reminds us, "States' rights, once a code word for discrimination and backward social policy, are in vogue again. But Justice Louis Brandeis, author of the oft-invoked phrase that states could be 'laboratories of democracy,' warned simultaneously against a 'race to the 2 See B.C. Smith's chapter on "Decentralization in Theory," ibid. 3 A related d iscuss ion of this point is in Sven Steinmo, "Why is the Government So Small in America?" Governance. Vol . 8, No. 3, July 1995. 33 bottom.' In Canada too, recent proposals for fiscal and administrative decentralization appear quite unattractive from the standpoint of equity in society. Particularly in Ontario, under a conservative regime, plans to shift provincial responsibilities down to the municipal level seem motivated more by a simple zeal for cost-cutting rather than a larger vision of promoting local democracy. 5 On the other hand, neither is decentralization a sure route to securing the goals of fiscal conservatives, as is evident in the lack of budgetary discipline among certain Canadian provinces. In constituencies that often seem institutionally prone to turn to deficit spending to support distributive politics, devotees of self-reliance are unlikely to feel "empowered" by decentralization. 6 Hence I restrict my institutionalist arguments to explaining in part why fiscal decentralization has not occurred to a significant extent in Japan. I find some progressive arguments for fiscal decentralization in Japan to be persuasive, 4 Robert Kuttner, "You Say You Want a Devolution?" The Washington Post, September 24, 1995 (http://epn.org/kuttner/ bk950924.html). 5 On Ontario, see Martin Mittelstaedt, "Municipalities to pay more," Globe and Mail, January 13, 1997, p. 1 and Michael Valpy, "The harvest of a patronizing, patriarchal government," Globe and Mail, October 22, 1996, np. 6 The Canadian federal level of government has been more fiscally responsible than its counterparts among the provinces in recent years, increasing its deficit to GDPra t io f rom 3.2 percent in 1988 to 4.6 percent in 1993. But the provinces, led by Ontario and Quebec, went from a f iscal surplus of 0.3 per cent to a deficit of 2.3 percent in the same period. The data are from the table on p. 94, Kokusai Hikaku 7bu/ce/[lnternational Comparative Statist ics]. Tokyo: Nihon GinkoToukeikyoku, 1995. Also, my thanks to Alan Siaroff for noting that "what is really striking about Canada is the interprovincial variations in deficits." 34 but to adopt such a normative approach would clearly require a different institutional and comparative analysis than is the object of this dissertation. Potential Technical Constraints on Decentralization On the other hand, a related question most certainly does have to be addressed. This is the potential claim that intergovernmental fiscal regimes are dictated by virtually iron laws in economics, and that technical constraints rather than political hurdles bar further progress on fiscal decentralization. If this were true, then the intergovernmental assignment of the tax base would be an administrative matter, largely beyond the realm of politics and the methods of political science. There are cases in which this caveat is highly relevant. Potential efficiency losses via local taxation at times preclude, or at least make very unwise, some forms of fiscal decentralization. For example, it would be wasteful for Australia to return to its highly decentralized regime of state and federal income taxes, which was in place until the federal state unilaterally took control of the income-tax field in 1942. Australian taxpayers in 1935 could, under certain circumstances, find themselves subject to as many as 14 different income taxes. This variety of levies brought with it an administrative Byzantium whose rebates and amended assessments could take up to three or four years before 3 5 they ran their frustrating course to a conclusion. 7 But such cases are rarities, and stand as cautionary tales in the folklore of fiscal studies. Modern tax regimes are instead marked by a strong emphasis on administrative and economic efficiency, with governments at all levels generally cooperating to maintain the legitimacy of the tax state.8 Such cooperation allows for a remarkable variation in the assignment of tax fields between the levels of government in developed states. Sweden, for example, permits local governments to collect the lion's share of the personal income tax, a practice particularly uncommon among centralized industrial democracies. Other taxes along the income distribution stream -- i.e., taxes on production, distribution, and expenditure -- can be found assigned to the various levels of government with apparently no hard and fast restrictions. There are, however, some fairly regular features of the management and assignment of tax bases. One is that agreements can generally be made to rationalize the overall tax base, as was noted above. This tendency applies especially where there are few subnational governments or they have been 7 On this see , Julie P. Smith, Taxing Popularity: The Story of Taxation in Australia. Canberra: Federalism Research Centre, 1993, pp. 49. 8 For example, see the sect ions on various means of cooperation between Japan 's central state and the local authorities in "Zeimu Gyousei Une iJou noKyouryoku Kankei" [Cooperation in Tax Administration] JichitaiZaiseiKiso no Kakuritsu [Enhancing the Fiscal Base of the Local Authorities]. Mutai Shunsuke, et al. Tokyo: Gyousei , 1995. Note that this kind of cooperation is also a common feature of tax relations among national entities, and is typified by tax treaties that allow credits for taxes paid on foreign-source income; see Richard A. Musgrave and Peggy B. Musgrave in "Tax Equity with Multiple Jurisdict ions," Taxation in a Sub-National Jurisdiction. Edited by Allan M. Maslove. Toronto: University of Toronto Press, 1993, pp. 3-43, esp. pp. 19-20. 36 relatively constrained by a central authority, leaving an institutional legacy of more or less cooperative interaction.9 This latter point suggests that decentralization in Japan would not descend into a US-style fiscal free-for-all. Second , the more visible types of taxes, especially property taxes, tend to be concentrated at the local level. The reasons for this distribution include history and social structure as well as the fact that the taxes are less buoyant and attract more political hostility. For example, "Eighteenth-century Americans preferred direct taxation - primarily the taxation of property -- in part because they wanted taxation to hurt and thus retard the growth of government." Though mainstream public finance suggests the property tax is ideal for collecting revenues for local services, the decision to follow this logic is a political choice. The Swedish state has given ample confirmation of this point through its shift over the past few decades from heavy reliance on property taxes to more buoyant income taxes. 1 0 9 This is not always the case , as the Canadian federal and provincial governments respectively impose a VAT (the Goods and Services Tax) and provincial sales taxes. Attempts to harmonize these taxes have been hampered by several provinces' unwillingness to give up some measure of their f iscal autonomy; see Ikegami Taihiko, "Kanada ni okeru Renpou, Shu aida Zeigenhaibun Ronsou" [The Intergovernmental Debate over Revenue Distribution in Canada], Chuo Daigaku Keizaigakuronshuu, Vol. 36, No. 4, October, 1995. Moreover, in the US, it is seemingly impossible to get agreement on tax harmonization among the states, whose tax mixes are greatly divergent. In the absence of constitutional authority to secure any measure of uniformity, America largely relies on the marketplace to regulate tax rates and tax mixes among the fifty states; see Taryn A. Rounds, "Tax Harmonization and Tax Competition: Contrasting Views and Policy Isues in Three Federations," Discussion Paper Number 15, Australian National University Federalism Research Centre, March 1992. 1 0 The quote is from Elliott Brownlee, "Tax regimes, national cr isis, and state-building," Funding the Modern American State, 1941-1995. Ed. by W. Elliott Brownlee. op. cit., p. 41. A lso B. Guy Peters points out that the heavy reliance on the property tax in the Anglo-American democracies has at least as much to do with the historical rate of home ownership as it does the more theoretical rationale of assess ing for locally provided serv ices; see The Politics of Taxation: A Comparative Perspective. Oxford: Basi l Blackwell, 1991, pp. 39-42. On public finance arguments and the property tax, see John F. Graham, 37 In other words, the relative assignment of direct and indirect taxes 1 1 between governments often depends on bargaining power and what taxpaying publics will put up with rather than the theoretical constraints emphasized in most public finance arguments. 1 2 These brief notes are necessary, because there is a mountain of orthodoxy that would obscure the politics of inter-governmental tax assignments. For example, mainstream public finance approaches, including those used in Japan, rely heavily on the Tiebout model for drawing conclusions regarding the appropriate tax mix. Stated sufficiently simply for the purposes of this dissertation, Tiebout is said to operate when citizens of the local state can "vote with their feet" and thus leave tax jurisdictions they do not approve of. "The Place of the Property Tax in the Fiscal System," Retrospectives on Public Finance. Ed . by Lorraine Eden. Duke University Press : 1991. On mainstream f iscal theory's reliance on the Anglo-American tradition and differences from that of continental Europe, note Richard Musgrave, "The Role of the State in Fiscal Theory," International Public Finance, 3, 1996, pp. 247-258. 1 1 While there is not complete agreement on the division of all taxes under the rubric of "direct" and "indirect," the standard definition is that direct taxes are those "in which the rates of tax can be adjusted to reflect the circumstances of the taxpayer, such as his net income or total consumption, asse ts or net worth, and special needs related to age, dependants, or disability." Hence this definition includes income, wealth, and inheritance taxes, whereas indirect taxes are "levied at rates which are independant of the taxpayer's attributes." These taxes are most often imposed on products or acts, and include value-added taxes, customs duties, and admissions taxes. On this definition, see Jonathan R. Kesselman, "Role of the Tax Mix in Tax Reform," Changing the Tax Mix. Edited by John G. Head. Sydney: Australian Tax Research Foundation, 1986, pp. 49-94, esp. pp. 51-2. 1 2 B. Guy Peters, The Politics of Taxation, op. cit., pp. 47-9. The politics of debating tax assignment is also seen in Peter Groenewegen, "Taxation Issues," Governing in the 1990s: An Agenda for the Decade. Ed. by Ian Marsh. Melbourne: Longman Professional, 1993, pp. 170-98. Also, as Russel l Matthews notes, the Australian states had the constitutional right in the mid-1980s to impose income-tax surcharges, but they refrained from doing so because "in the absence of a partial Commonwealth withdrawal from the field, tax rates had become so high as to make surcharges politically unattractive"; see "Fiscal Federalism in Australia: Past and Future Tense," Reprint Series Number 74, The Australian National University Centre for Research on Federal Financial Relations, 1986, p. 14. 38 The model thus advises the taxation of things at the local level and the taxation of persons at the national level; in other words, indirect taxes are recommended for subnational governments and direct taxes for the centre in order to prevent the option for exit from posing drastic limits on tax choices or producing excessive tax competition among local authorities. 1 3 However, there are serious problems with Tiebout and similar public finance analyses of intergovernmental fiscal relations. In particular, the models are hampered by their positivist methodology, which is far too abstract to deal adequately with the rich institutional history of a given intergovernmental fiscal regime. 1 4 Though taxes per se do seem truly as certain a phenomenon as death, the precise mix of taxes cannot be infallibly predicted or prescribed by neoclassical models. This fact is evident in the above-noted variability of intergovernmental tax regimes. Another constraint on local taxes is said to lie in the functions of the various levels of the state. The theory of public finance which underpins most thinking on the subject sees the role of local taxation as a means for extracting returns for the benefit derived from local services. This is the so-called "benefit 1 3 On Tiebout and its reading in Japanese public f inance, see Kato Hiroshi and Yokoyama Akira, Zeisei to Zeisei[The Tax System and Tax Politics]. Tokyo: Yomiuri Shimbunsha, 1994, pp. 230-36. 1 4 P. Bernard Spawn, "On the Assignment of Taxes in Federal Polit ies," Taxation and Fiscal Federalism. Ed. by Geoffrey Brennan, Bhajan S. Grewal, and Peter Groenewegen. Sydney: Australian National University P ress , 1988. Note that the criticisms are in the context of a d iscussion of f iscal federalism, but the critique of the formalism of Tiebout and other theories holds as well for the assignment of taxes in a centralized fiscal state. 39 principle," whose roots lie in the restricted activities and revenue base of the nightwatchman state. Benefit taxation contrasts with the principle of ability to pay, wherein taxes are levied in association with the state's redistributive function. Ideally then, taxes levied at the local level should not be very progressive, if at all, and should aim solely at recouping the cost of local services based on a rough evaluation of the benefit taxpayers accrue from them. 1 5 In practice, this kind of evaluation is very difficult to make, as taxpayers may live and work in different tax jurisdictions. Nor can tax for services -- as opposed to direct fees -- be levied with surgical precision on those who consume them, even if the taxpayers remain within the same jurisdiction. This is because the "tax price" paid is an inefficient mechanism for distributing costs. Such problems thus open up a wide scope for essentially political judgements in determining the intergovernmental tax mix. The empirical record also shows, as we see in Table 2.1, that the intergovernmental allocation of major taxes in developed countries' tax regimes varies greatly. The Swedish example sees the heavy imposition of income taxes, albeit at flat rates, at the local level versus a rather light burden 1 5 A short description of the principle can be found in Richard Musgrave and Peggy Musgrave, "Tax Equity with Multiple Jurisdict ions," Taxation in a Sub-National Jurisdiction. Ed. by Allan M. Maslove. Toronto: Fair Tax Commission, 1993, pp. 5-7. A crit icism of the principle of benefit taxation as a tax ideology can be found in Naohiko Jinno and Andrew DeWit, "Institutions, Ideology, and the Intergovernmental Allocation of Taxes, " op. cit. 40 at the national level. Canada, Germany, and Japan, as is also clear from the table, stand roughly between the example of Sweden and its diametrical opposites, Australia, France, Italy, the U.K. and the U.S., where income taxes play a minor role at the local level. Property taxes as well show considerable variation according to the particular country, with Australia, Canada , Japan, and the U.S. showing rather high levels of local funding from this source, whereas the Swedes levy no local property taxes since their reforms. Consumption taxes, meanwhile,. are prominent local revenue sources in Germany and the U.S., but are levied almost wholly by the central states in Australia, France, Italy, Japan, Sweden and the U.K. Table 2.1: Tax Structures at National and Local Levels, 1994 (Units: Percent of Gross Domestic Product) Country Income Taxes Property Taxes Consumption Taxes National Local National Local National Local Federal Austral ia 16.2 0.0 0.0 2.8 6.1 1.0 Canada 9.8 6.2 0.0 4.0 4.3 4.4 Germany 4.6 6.9 0.0 1.1 7.8 2.8 U.S. 10.6 2.3 0.2 3.0 1.1 3.3 Unitary Japan 6.8 3.6 1 .1 2.1 3.1 0.7 France 7.2 0.6 0.8 1.6 10.9 0.2 Italy 14.0 0.5 1.4 0.8 10.5 0.1 Sweden 4.5 17.2 1.6 0.0 12.8 0.1 U.K. 12.1 0.0 2.4 1.3 11.1 0.0 Source: OECD: Revenue Statistics, 1965-1995. Paris: 1996. The table also indicates that the most productive taxes tend to be at the national level. This concentration of fiscal capacity is due to the redistributive 41 and economic management functions performed by the relatively centralized postwar fiscal regime. But as spending responsibilities grow and diversify at subnational levels of government, there are increasing calls for the transfer of tax room from the national authority.16 The point, to reiterate, is that there are few demonstrably iron-clad rationales for depriving subnational levels of government of access to specific tax bases, and more generally, of enhanced fiscal autonomy. By extension, limited scope at best exists for denying that the assignment of taxes among the various levels of government is not strongly determined through politics. The argument that intergovernmental tax divisions are matters of politics, not principle, is strengthened by the fact of the globalization of the market economy. Many analysts have come to insist that the attendant massive movements of funds around the globe and around the clock, has seriously undermined orthodox assumptions about the appropriate distribution of powers in the fiscally centralized postwar industrial state. They conclude that as the national state loses its capacity to act as an interface between the global and domestic economies, the rationale for investing it with fiscal and administrative pre-eminence over the local authorities also erodes. 1 7 1 6 For example, C Ruggeri et al., "Vertical Fiscal Imbalance and the Reallocation of Tax Fields in Canada," Canadian Public Policy, Volume 19, No. 2, 1993, pp. 195-215. 1 7 On this development, refer to Jinno Naohiko, "Reinforcing the Fiscal Resources of the Local Authorities: The Japanese Experience," Tokyo University, Faculty of Economics Discussion Paper Ser ies, 95-F-20, August, 1995. 42 Degrees of Decentralization I have argued in the above that political variables help shape intergovernmental taxation, but I have not offered a consistent explanation for variations in the kinds of institutions involved or the intensity of the associated politics. The tax regime concept cannot explain such differences, for example, due to its macro-level focus. This lack of a rigorous theoretical approach here would seem problematic, so far as there is a premium on comparative work with a consistent and restricted set of variables. 1 8 But one thing that cannot be ignored is that analysts disagree on the very interpretation of fiscal and administrative centralization (or decentralization) in a particular country. The difficulty stems in large measure from the significant institutional diversity in countries' intergovernmental fiscal relations. This diversity not only includes -- as we have seen -- the character of the tax mix, but also historical influences on taxation, 1 9 the rules and institutions that shape intergovernmental tax relations, and a host of other related features. 1 8 Michael Loriaux notes this problem in his review article, "Comparative Political Economy as Comparative History," Comparative Politics, Vol. 21, No. 3, April 1989. Loriaux cites Alisdair Maclntyre in support of his contention that the search for laws of political behaviour should be abandoned in favour of a system-specif ic approach, using context and history to compare "the efforts of some element - a political elite, government, or political party -- to assure its political survival." The concept of an organizational mission would appear to fit nicely into this style of comparative work. 1 9 A c lass ic example here is the French tax regime, whose light burden of direct taxation arguably dates from the Revolution. The French nobility was legally entitled not to pay taxes, and the outrage this engendered appears to have influenced the subsequent character of the tax mix. Certainly this argument about path dependence also relies on the role of small business and other anti-income tax interests in modern French politics, but there is an interesting continuity. Unfortunately, there does not appear to be - at least in English or Japanese - a work that traces this legacy. 43 The division between unitary and federal states might seem the the most obvious place to start in constructing a system of classification, but even it does not demarcate more and less centralized regimes. Rather, grounds exist for arguing that there are too few differences to warrant discussing centralized and federal states as truly distinct entities. 2 0 Among the federal states, for example, Australia's strong vertical imbalance in the division of tax fields (see Table 2.1) gives the federal authority, or Commonwealth, a great deal of influence over the policies implemented by the states. Canada, by contrast, has greatly decentralized its access to fiscal resources over the postwar period, driven in large part by demands for greater autonomy from wealthy provinces (especially Alberta, British Columbia, and Ontario) and by nationalist aspirations in the province of Quebec. Furthermore, the local authorities in centralized Britain became even more clearly dominated by the centre during the 1980s, when free-market Thatcherism's ironically aggressive agenda of centralization constrained Labour-dominated local councils and their already limited fiscal autonomy. The Thatcherites so disliked local councils that "in 1991 the cabinet considered abolishing local government altogether."2 1 To restrict the local authorities' capacity to spend, the Tories enacted controls on local taxation 2 0 See Dale A. Crane, "Does the Federal-Unitary Dichotomy Make any Difference?" Subnational Politics in the 1980s. Ed. by Louis Picard and Raphael Zariski. New York: Praeger, 1987. 2 1 See "Never say die," The Economist, May 4, 1996, p. 64. The article further comments on the strong role of party politics in Britain's centre-local relations: "Labour...is all for local democracy - provided that means Labour counci ls, controlled by moderates who can be relied upon to do the national leadership's bidding." 44 and then imposed the regressive and ill-fated poll tax. One effect of Conservative efforts to suffocate local autonomy saw local revenues in Britain plummet from 11 percent of total revenues in 1975 to 4.1 percent in 1993. Devolution of authority overtaxes in the U.K. remains strongly opposed by the Conservatives and, reflecting a clear institutional interest in fiscal centralization, the British Treasury. 2 2 By contrast, many other unitary states have been devolving responsibilities --and the resources to meet them - onto their local governments. The Thai Government is one example, with its success in passing and implementing a raft of decentralization measures in 1994. Important prerequisites in this case were decentralization's salience on the agenda of party politics and elected politicians' ability to overcome the Ministry of the Interior's efforts to subvert the process. This case reflects yet again the significance of institutions: Thai local governments are dominated "by the central government through an administrative network of provincial governors, district officers and other officials belonging to or appointed by the Ministry of Interior (sic) in Bangkok. The power of the ministry to oversee the whole of the country outside the Bangkok Metropolitan Area has remained relatively unchanged and unchallenged for decades." In order to get around the obstacle of central References include "Britain's Constitution," The Economist. November 11, 1995; Gerard Stoker's chapter on "The Battle Over Local Spending," The Politics of Local Government, 2nd Edition. London: MacMil lan, 1991; John Gibson, "British Local Government Finance under the Conservat ives," Local' Government Studies, Vol . 18, no. 4, 1992, and O E C D , Revenue Statistics of OECD Member Countries 1965-1994, 1995. 45 agency obstructionism, the ruling party made certain to include a number of independent academics on the Local Revenue Improvement Committee that studied and made recommendations on the issue of decentralization. 2 3 Another notable case is China, where one might think the unabashedly anti-democratic regime at the centre would be firmly in charge of the fiscal state. However, China's 1980 reform of the intergovernmental fiscal system replaced the highly centralized fiscal structure that had characterized the era of the centrally planned economy. The reform implemented a "fiscal contract" system in which a complex system of negotiations determines the level of fiscal transfer and whether it will be from the centre to the local authority or, in the event of a local surplus, vice-versa. Local governments thus have powerful incentives and opportunities to cheat on their tax obligations. Enormous problems have thus ensued for the Chinese central state, which finds itself in a fiscal crisis due to a shrinking revenue base coupled with the need to prop up large and inefficient corporations that perform a wide range of welfare state functions for millions of citizens. Abandoning these firms to the marketplace The quote is from Medhi Kronkaew, "The Political Economy of Decentralization in Thailand," Southeast Asian Affairs, 1995, p. 345; see also the same author's "Fiscal Reform in Thailand," a paper presented at the June 24-25, 1995 conference on Fiscal Regimes in the Asia-Paci f ic Region, Australian National University, Canberra; note also the report on this conference and its d iscussions on local f iscal autonomy by Andrew DeWit, "Ajia Taiheiyou Chiiki no Zei, Zaisei Seido ni Kansuru Kyanbera Kaigi Houkoku," in Chihouzei, October 1995. It is interesting that the Thai Ministry of the Interior was comparatively flexible concerning proposals for f iscal decentralization, but resolutely opposed to reductions in its powers of administration and appointment. This could be because f iscal controls are less certain and blunter instruments than are direct orders transmitted through appointees in local administrations. If so, then this suggests that Japan's Interior Ministry bureaucrats found a fall-back position through tax controls in the wake of the Occupation's dissolution of the ministry and its extensive powers. 46 would create massive and politically unstable levels of unemployment, but convincing the rapidly growing local areas to go along with a scheme to redistribute revenues is proving surprisingly difficult for the ruling elite in Bei j ing. 2 4 In short, simply dichotomizing unitary and federal states obscures a great diversity in institutions and the degree of fiscal decentralization. Furthermore, the categories are not stable. A voluminous literature on federalism examines changes the concept has undergone through the wartime and postwar expansion of the federal state as well as more recent pressures for restraint and decentralization. There is also a large literature dealing with similar processes, as well as countertrends, in unitary states. 2 5 The research indicates that a given national community will show, unless there is profound institutional ossification, some variance in its degree of fiscal decentralization as governments respond to the pressures of interest groups and socio-economic change. It is, however, well beyond the scope of this thesis to review this literature in any detail. The main points to draw from such works and the preceding brief 2 4 J inno Naohiko, "Shijo keizaika to soze i seido: chuugoku no zeisei to seifukan zaisei kankei" [Marketization and the tax system: China's intergovernmental f iscal relations], Kounan Keizaigaku Ronshuu, Vol . 34, No. 4, March 1994 and Wu, Jinglian. "China's Fiscal Reform: Centre-Locality Relations," Tax Policy and Planning in Developing Countries. Ed. by Amaresh Bagchi and Nicholas Stern. London: Oxford University Press, 1994. 2 5 See , for example, Percy Allen, "Subcentral Government and Centre-Local Relations," State and Society in Western Europe. Cambridge: Polity, 1995, pp. 412-470 and Douglas E. Ashford, British Dogmatism 47 discussion of recent changes in Thailand and elsewhere is that there is much institutional variability among states regarding their intergovernmental relations. Drastically polarized and constitutionally based definitions of federalism and centralized states describe, as Allen puts it, only "the visible face of power" and hence a more finely honed perspective on institutions is required. Moreover, these institutions themselves generally change over time in ways that indicate neither a linear path towards increased centralization nor to greater decentralization. 2 6 Decentralization as a Comparative Variable Fiscal decentralization is clearly an inconstant category between types of states and even within a particular political community -- though as we saw above, the Japanese case is notable for its postwar stability. As a result of the general tendency towards change, there is no robust comparative theory to explain the degree of fiscal decentralization across countries, and constructing one seems a daunting challenge. After reviewing several analyses that focus on geography and other non-institutional variables, Gibson and Batley are led to conclude that "there is a large variation between countries unexplained by and French Pragmatism: Centre-Local Policymaking in the Welfare State. London: George, Allen and Unwin, 1982. The quote is from Percy Allen, op. cit., p. 416. Arnold Heidenheimer et al. note that increased centralization has been argued by many analysts to be an inevitable trend, based on exapanding welfarism and other extrapolations; see Comparative Public Policy, 2nd Edition. London: Macmillan, 1983, pp. 274-310. Michael Pei l lonalso d iscusses these issues in, "Welfare and State Centralization," West European Politics, Vol . 16, no. 2, April 1993. 48 such variables." They argue somewhat vaguely that "the wider political background" is necessary to explain the variation. 2 7 This dissertation examines closely the "wider political background" in the Japanese case. The goal here is not to construct a rigorous comparative theory, because the institutional context can affect the interpretation of even seemingly unambiguous indices of measurement. To take what should be a rather straightforward case, Dutch municipal and regional governments collect only about 10 percent of what they spend and hence would seem enmmeshed in a centralized system. However, some analysts suggest the Dutch system of intergovernmental relations confers abundant local autonomy. Their arguments centre on the representative function played by the Vereniging van Nederlands Gemeenten (VNG), the Dutch association of local governments. Through this institution, local governments negotiate their levels of aid from the central state, which has "little direct leverage on policies in direct localities." Hence, what seems on a statistical basis to be a highly centralized intergovernmental regime of fiscal relations turns out instead to be much less so . 2 8 Or is it? Jan Simonis, an expert on Dutch intergovernmental affairs, interprets the same institution in a completely different way. Simonis argues that the 2 7 John Gibson and Richard Batley "Introduction: Trends in Financing European Local Governments," Local Government Studies, Vol. 18, no. 4, 1992. 49 Dutch national government has managed to take over the lion's share of decisionmaking on local policies via a steady expansion of "co-governance." This phenomenon sees the national authority able to use its control of finances to impose its preferences on the recipient local authorities. Moreover, the administrative style is supported by established institutional practices in the Netherlands, which tend to discourage the expansion of differences among the local authorities through "increased revenue from local taxes." The author notes that many central government departments, concerned about increased taxation of their respective client interests, opposed higher local taxes. In fact, not all the local authorities were fully in support of the proposal: "On the one hand they are keen to obtain greater autonomy; on the other they are wary of the risks and consequences this kind of independence would create for them." 2 9 This account, and others noted above and in the pages that follow, show that intergovernmental fiscal regimes are institutionally complex and cannot meaningfully be summarized by a restricted set of variables. I deliberately belabour this point because the desire to find simple explanations for, or measures of, complex phenomena is both the genius of good social science and the source of appalling reductionism by even brilliant social scientists. 2 8 Arnold Heidenheimer et al., Comparative Public Policy, 2nd Edition. London: Macmillan, 1983, pp. 283-84. 2 9 The Ministry of Economic Affairs in particular "feared a considerable increase of rates on bus inesses and land"; see Jan B. D. Simonis,"Decentralization in the Netherlands: an Analysis of Administrative Differentiation," Local Government Studies. Volume 21, Number 1, spring 1995, p. 39. 50 Arend Lijphardt, for example, constructs a comparative table that uses only the relative proportion of revenues collected locally to measure the level of decentralization across countries. Lijphardt concludes that because Japanese local authorities collect over 30 percent of their own revenues, they are highly autonomous actors. I will show below that Lijphardt's approach is only the most glaring case that neglects the institutional context of MOHA's tax controls. 3 0 Abandoning the quest for a readily quantifiable parsimony might seem to leave my account adrift, with no reasonably clear idea of the variables that are important in assessing centralization and the associated fiscal politics. But fiscal sociology wields a razor worthy of William of Ockham in its focus on control over finances. In this respect, the most fractious intergovernmental fiscal politics centre on subnational governments' autonomous access to fiscal resources. Indeed, the fiscal history of many, if not most, states is notable for the centre's efforts to restrict the local authorities' range of independent action in the sphere of revenues and expenditures. What has varied is the institutional context in which these efforts occurred, permitting greater or lesser degrees of central control. 3 1 3 0 Note the d iscuss ion of "Decentralization" in Arend Liphadt, Democracies: Patterns of Majoritarian and Consensus Government in Twenty-one Countries. New Haven, Conn.: Yale University P ress , 1984. Indeed, even where multiple variables are used to a s s e s s relative levels of centralization, the results tend to be quite ambiguous; see Michael Peillon, "Welfare and State Centralization," West European Politics., Vol. 16, No. 2, April 1993. 3 1 Detailed accounts of the France and the UK can be found in Douglas Ashford, British Dogmatism and French Pragmatism: Centre-Local Policymaking in the Welfare State. London: Allen and Unwin, 1982, and Douglas Ashford, The Emergence of the Welfare States. Oxford: Basi l Blackwell, 1988. Ashford 51 Hence, it does to a large extent matter whether subnational governments collect their own revenues and set their own tax rates. Peters, for example, notes in his impressive comparative work on tax politics that "local governments with more independent revenue sources can, everything else being equal, do more of what they choose rather than what the central government wants them to do." 3 2 The caveat, however, is that the larger institutional environment has to be assessed in tandem with the simple statistics on tax collection, as in Japan MOHA occupies the ambiguous role of both dominating the local tax base and defending it against the depredations of the centre, represented by MOF. These points have not adequately been understood in analyses of Japan because many scholars, like Liphardt, rely on misleading statistical data without researching the institutional relationship behind the figures. Others have been keen to overturn the old orthodoxy by showing that specific subsidies do not afford close, top-down control, but they too have neglected to analyze the institutions and incentives of intergovernmental tax politics. By contrast, progressive Japanese public finance scholars often do better comparative work on intergovernmental relations than their counterparts in d iscusses the institutional background to the apparent paradox of stronger centralized control in the UK than in France. B. Guy Peters, The Politics of Taxation: A Comparative Perspective. Cambridge: Basi l Blackwell, 1991, pp. 46-9. 52 political science, who tend to ignore or dismiss their work. Public Finance and Local Autonomy in Japan In Japanese, and especially among progressive public finance scholars, there is an extensive literature that calls for fiscal decentralization. This work is part of a critical tradition that stretches back at least to the Taisho (1912-1925) democracy period, when pressures for fiscal decentralization were frequently asserted. 3 4 Nor is it surprising that Japan should have an indigenous tradition of insisting on the need for local autonomy, as the country was quite decentralized prior to the 1868 Meiji Restoration. 3 5 The Meiji centralization of the state and the rapid modernization of feudal structures greatly altered, but did not eliminate, the sense of localism, which conservatives of the day often tried in fact to bolster. And in contemporary Japan, the local community and its needs are an increasingly potent source of decentralist pressures. Hence, Japanese debates on decentralization are based on a good deal more than the Occupation's quintessential^ American concern for local autonomy as the This claim is based on the present author's experience from discussions and from checking who cites whom. Pre-war editions of journals such as Toshi Mondai carry numerous articles showing how salient these pressures were, especially during discussions of major tax reforms in the 1920s and 1930s. Kurt Steiner, Local Government in Japan. Stanford: Stanford University Press , 1965, pp. 1-18, argues that one cannot persuasively argue there was local self-government in pre-Meiji Japan because there was no strong conception of rights. Hence largely autonomous, de jure as well as de facto, urban centres such as developed in the West were absent. However, his claim that the relations between the feudal state and local areas were "diffuse" and characterized by a "vacuum of government" rather than the "concept of the right to local self-government" does not logically preclude the existence of a large sphere of relatively autonomous local affairs, especially activities governed by custom rather than explicit law. 5 3 school of democracy. In any case, the power of decentralization as a rhetorical device is undeniable. Certainly there are dissenting voices, from individuals who point to fiscal discipline and other presumed benefits of centralization, but they are clearly the minority. The rise of decentralization as a somewhat ambiguous new orthodoxy is reflected in other central state ministries' concern about MOHA's ability to use the concept as a slogan, in a campaign in which the prize is more likely to be a recirculation of power among the central ministries than its actual devolution to the local authorities.3 7 But Japanese public finance scholars dwell less on the political-institutional roots of fiscal centralization than they do on its tax-transfer mechanisms. Few of their accounts examine the central state as the disaggregated entity it actually is, nor are they particularly concerned with the policymaking milieu that reproduces the centralized fiscal state. They focus instead on the means of fiscal control, and to this end many of them provide excellent accounts of the workings of the Japanese intergovernmental fiscal system. Many also examine the system in a comparative light, and this dissertation has benefited from having them to draw on. 3 6 Carol Gluck 's work on political debates in the late Meiji era notes, for example, that "local autonomy" (chihoujichi) was long a focus of competing ideological agendas; see Japan's Modern Myths: Ideology in the Late Meiji Period. Princeton: Princeton University Press, 1985, pp. 189-205. 3 7 0 n this, see Kato Yoshiaki , "Chihou bunken chakuchi e no michi: Kengen ijou gutaisaku naku" [The road to decentralization: no concrete policies for transferring authority], Nihon Keizai Shinbun, October 22, 1994, p. 1. 54 The public finance work is thus not to be faulted for remaining largely within the confines of its own methodological concerns. What is needed, though, is to take the insights from such studies and invest them with the greater political understanding that can be achieved through an historical institutionalist analysis. Political Science and Local Autonomy in Japan Political science accounts of Japan's intergovernmental relations have come largely to emphasize the autonomy of local governments, asserting that fiscal constraints are much exaggerated. 3 8 Many of these studies are strongly informed by the 1960s trend towards viewing local government as an important arena of politics rather than, as was previously the case, merely an agent of central administration. This signal change in perceptions engendered major works that depicted a large scope for local activism and effectiveness Much of this work followed the growth of ci t izens' movements and the election of local progressive politicians in the 1960s, which appeared to herald a new era of intense local act ivism in Japanese democracy; see, for example, Ronald Aqua, "Politics and Performance in Japanese Municipalities" (Unpublished PhD. dissertation, Cornell University, 1979) and Kurt Steiner et al. Political Opposition and Local Politics in Japan (Princeton: Princeton University Press, 1980). Later contributions include Michio Muramatsu, "Centre-Local Political Relations in Japan: A Lateral Competition Model," The Journal of Japanese Studies, 1986. Stephen Reed, Japanese Prefectures and Policymaking. Pittsburgh: University of Pittsburgh Press , 1986, and Richard Samuels, The Politics of Regional Policy in Japan: Localities Incorporated? Princeton: Princeton University Press , 1983. They do not reject the fact of f iscal centralization, but they do argue that there is more local autonomy than was previously thought. A more recent analysis concedes that local f iscal autonomy made little headway in the 1980s, but it employs a public choice framework of analysis that fails to disaggregate Japan 's central government into separate bureaucratic institutions. See Hanai Hiyohito, Intergovernmental Grants and Distributional Issues: The Case of Japan. (Australia-Japan Research Centre unpublished PhD diss., 1993). 55 even in such highly centralized political communities as France. 3 9 While not wholly incorrect, claims that Japan is already quite decentralized have been too enthusiastic, and hence have generally overshot the mark. The pluralist approach favoured by the new orthodoxy causes it to neglect the larger institutional environment, or fiscal regime. Moreover, in asserting the importance of local activism against the centre, it has not paid adequate attention to the divergent and at times manipulable interests among the local authorities nor the emphasis in tax control on promoting uniformity as equity. This last item is especially important, I argue, because the crisis-ridden history of Japan's intergovernmental relations has left deep imprints on the shape of the fiscal regime and the purpose of its associated institutions. The earlier work also centres for the most part on the fiscally expansive years before the oil shocks of the 1970s. The 1973 shock heralded the secular decline in the Japanese economy's rate of growth; and the shock in 1979, the rise of a protracted austerity campaign under the slogan of "administrative reform." The related fiscal weakness of the local authorities, from the mid-1970s on, did much to temper their late 1960s uprush of assertiveness. This was especially notable in the case of Tokyo Governor Minobe. The experience of Japan's largest subnational government surely has significance for the debate on the degree of decentralization and the relevance of the centre's 3 9 A notable example is Douglas Ashford, British Dogmatism and French Pragmatism: Centre-Local 56 fiscal controls. Hence this aspect needs to be incorporated into discussions of local autonomy in Japan. Financial Dependence and Local Autonomy Ronald Aqua confronts head-on the question: "Does financial dependence limit local autonomy?" In his view, the fact that Japan's local authorities do not collect the bulk of their revenues is not in itself an impediment to local autonomy. This is because the local authorities do the majority of programme spending and receive a large transfer of general grants from the centre. Moreover, he emphasizes cases where the local authorities have mounted legal challenges to specific tax provisions and where they have ignored MOHA directives, such as on the need to reduce local salaries due to Japan's mounting fiscal problems in the wake of the 1973 oil shock. But Aqua overlooks the fact that locally collected revenues are not based on autonomous taxes. My discussion of MOHA's tax controls in chapter 3 of this dissertation shows that local taxes are assessed standard and restricted rates by MOHA. There is some scope for choice at the local level, to be sure, but MOHA is able to enforce the rates through the operation of the transfers it manages, including the Local Allocation Tax and the Consumption Transfer Tax. Most local authorities that would seek to exceed the rates that MOHA Policymaking in the Welfare State. London: Allen and Unwin, 1982. 57 advises for a particular tax, can be penalized through the transfer system because the amount of transfer is calculated on the basis of local tax effort. MOHA's calculations assume, unless the ministry has given explicit approval to an exception, that a given local authority is imposing a specific tax at the standard rate. If such is not the case, the local body will forfeit some portion of the transfers it would otherwise be entitled to. Moreover, Aqua comments approvingly on the shift to general grants (mainly via the Allocation Tax) and away from the specific subsidies favoured by line ministries such as Construction, Education, and Health and Welfare. He sees this trend as indicative of a reduction in fiscal centralization, as local governments may use the general revenues "to meet any legal financial commitments." 4 0 But higher levels of general transfers do not necessarily expand local fiscal autonomy. The context in which the level of grant increases has to be examined before such a verdict can be rendered, yet this was not done in Aqua's work nor in subsequent studies that cited his argument as evidence of fiscal decentralization. Many public finance analysts in Japan point out, by contrast, that the role of the Local Allocation Tax has shifted from one of simple fiscal equalization to one of paying down accumulations of debt. This new role "Polit ics and Performance in Japanese Municipalities" (Unpublished Doctoral Dissertation, Cornell University, 1979), p. 22. 58 restricts the tax's utility to the local authorities.4 1 In addition, my case study of the politics of the Local Allocation Tax shows that the transfer and its institutional nexus are far more complex than Aqua and other political scientists tend to assume. Debates in the comparative literature also recognize the need to inquire into the institutional circumstances of the particular country's intergovernmental relations before concluding whether increased general transfers actually mean greater local fiscal autonomy. One authoritative source notes, for example, that unrestricted grants do not necessarily mean that the recipient local governments are free to spend the money as they see fit. The reason is that the centre may impose "specified services at specified levels." The same work also argues that: a gradual shift toward unrestricted grants usually means that communities are increasingly reliant on transfers to finance their traditional core services. Once this occurs, local governments become far more dependent on national transfers, not simply to support peripheral projects, but also to support the very heart of the local operating budget. In this sense the shift to unrestricted grants increases local dependency. 4 2 A further problem with Aqua's account is his suggestion that local debt is free from the intervention of the central state. That is, Aqua bolsters his claim of a 4 1 Iwamoto Kazuak i , "Chihou Kouf uzei Seido no Henyou" [The Change in the Local Allocation Tax System], in his Nihon Chihou Zaisei Ron [Theory of Japan's Local Finance]. Tokyo: Doubunkan, 1986. 59 relatively high degree of fiscal autonomy in Japan by, among other things, lumping local bonds in with local fees and charges and indicating that these are autonomous local revenues. 4 3 But in fact the local authorities are compelled to secure the permission of MOHA before they float local bonds. This constraint can be quite a severe one at times, as former Tokyo Governor Minobe discovered in the mid-1970s. 4 4 Lessons From Minobe's "Fiscal War" Political scientists have largely failed to examine the implications for local autonomy of former Tokyo Governor Minobe's defeat in his "fiscal war" on the centre -- and especially MOHA -- in the wake of the first oil shock (i.e., from 1973 to 1977). 4 5 This omission is quite puzzling. One would think the new wave of literature stressing the growing scope for local policymaking --especially under the progressive leadership of Kyoto and Tokyo a generation ago -- would discuss the role MOHA's fiscal controls played in derailing the Minobe regime's plans for expanding Tokyo's expenditures and tax base. Kurt See Arnold Heidenheimer et al., Comparative Public Policy, 2nd Edition. London: Macmillan, 1983, p. 308. Aqua, op. cit, p. 20. Minobe Ryoukichi, along with Kyoto Governor Ninagawa Torazou, was a major symbol of progressive control of local government in the 1960s and 1970s. See Ellis Krauss, "Opposition in Power: The Development and Maintenance of Leftist Government in Kyoto Prefecture," and Kurt Steiner, ""Progressive Local Administrations: Local Public Policy and Local-National Relations," Political Opposition and Local Politics in Japan. Ed. by Kurt Steiner, Ellis Krauss, and Scott Ranagan. Princeton: Princeton University Press, 1980. See "Progressive Local Administrations: Local Public Policy and Local-National Relations, Political Opposition and Local Politics in Japan, op. cit, pp. 334-35. 60 Steiner, for example, appears satisfied in depicting Minobe's problems as the consequence of recession and reduced revenues. But this is only half the story: through its debt controls MOHA denied Minobe access to short-term substitute funding, and hence narrowed his policymaking options to deciding where to impose cuts in planned programmes. Steven Reed also makes note of the case. But he does so only in passing and as an example of a renegade among the local authorities, most of whom "accept [MOHA's] definition of sound financial management." 4 6 The problem here is, of course, that the comment fails to refute the argument that MOHA's ability to rein in Tokyo's metropolitan government showed that fiscal controls do constrain local policymaking. A recent and well-documented work by Jinno Naohiko finds much more significance in Minobe's defeat in the fiscal war. Jinno notes that Minobe's strategy after 1973 was to fight for control over revenues at the local level and thereby expand authority over expenditures as well. This plan required that the central authorities back down in the face of his challenge, but the effort ended in humiliation for Minobe rather than the acquiescence of MOHA. The fiscal bind for Tokyo's government in the early 1970s was that it ran out of ways to shift funds around in order to cover personnel expenses. These costs 4 6 Steven Reed, Japanese Prefectures and Policymaking, op. cit, pp. 31. 61 had rapidly increased as a result of expanded outlays on the metropolitan welfare state. Tokyo's 20 percent annual increases in spending were, prior to 1973, paid for through equivalent increases in revenues. But the economic effects of the oil shock slashed corporate tax receipts, upon which the city is highly dependent. Moreover, as an anti-inflation measure fees for public services were not to be raised. Though cuts were made in capital investment and the funds transferred to cover salaries, the limit had been reached with that strategy. Hence the Minobe Administration turned to MOHA for permission to float bonds; but MOHA declared that it would no longer extend such permission for revenue replacement. This move, and the refusal to yield to Tokyo's pressure for more tax room, shoved Minobe onto the path of fiscal reconstruction. In other words, the spur to austerity was the centre's control of the fiscal system through MOHA's administrative authority.4 7 With a more flexible tax base and autonomous access to the bond markets, it is quite possible that Minobe could have found the fiscal resources to ride out the economic downturn. Minobe's welfarist spending priorities had a significant base of popular support, but one inadequately organized for the task of bringing overwhelming pressure on the postwar fiscal regime. Many health and welfare measure adopted by the central government in the 1970s were in fact originated late in the previous decade by progressive local 4 7 J inno Naohiko, "Toshi Keiei no Hatan kara Saiken e" [From Failure to Reconstruction in Urban Management] Toshi o Keieisuru [Urban Management]. Ed. by Jinno Naohiko. Tokyo: Toshi Shuppan, 1995, pp. 67-138. 62 leaders, especially the Minobe administration. So with greater fiscal autonomy, local authorities might well have had the fiscal stamina to maintain policy leadership and push harder against the boundaries of conservative policymaking. But the centralized fiscal regime and its tax and debt controls held against Minobe's challenge to the constraints on the local tax base. This challenge can, in retrospect, be seen as effectively centring on whether progressives or conservatives would lead the expansion of Japan's welfare state. Certainly even had Minobe triumphed at this stage, he may in the end have failed to maintain Tokyo taxpayers' support with a programme of providing enhanced services for increased local levies. But on the other hand, the fiscal scale of Tokyo in the Japanese system is such that if the Minobe administration had triumphed against MOHA and the conservative hierarchy, he would have profoundly shifted the base of Japan's intergovernmental fiscal regime. This in turn might have helped erode in part the Japanese public's pronounced and institutionalized aversion to even moderate levels of taxation. The rout of Minobe in the fiscal war therefore opened the door to the "administrative reform" fiscal conservatism of the 1980s, which implemented cutbacks to an already small fiscal state. Refuting the "Vertical Control Model" Other political science work, as I mentioned above, tends to follow Aqua's lead 63 in downplaying the significance of financial controls from the central state. Muramatsu Michio, for example, rejects the "vertical control model," which he characterizes as one that centres on financial and administrative dominance of the local authorities by the central bureaucracy. The advocates of the model, Muramatsu notes, point to the early postwar proliferation of delegated functions that ministries had written into the Local Autonomy Law in order to maintain control over the authorities. The functions allow central ministries to compel local governments to undertake specific duties under the former's tutelage, and as noted in the introduction generally involve the extension of financing through specific subsidies. 4 8 Muramatsu critiques the vertical control model by pointing to instances where central direction has been deflected or modified by the local authorities. Local governments, depending on the issue at hand, can play central bureaucracies off one against the other in Japan's fragmented system of centralized government. 4 9 Hence one finds examples where local authorities have exceeded central ministries' recommended guidelines concerning day-care centres and medical care. Muramatsu thus correctly argues that the vertical control thesis has been content to assume that formal lines of authority are actual lines of authority, unmediated by the politics of local policymaking and 4 8 The best description in English of how the sys tem works is in Richard Samuels, The Politics of Regional Policy in Japan: Localities Incorporated? op. cit., esp. pp. 37-82. 4 9 The phenomenon is called "vertical administration," or tatewarigyousei'm Japanese . It refers to the fact that ministries and agencies of the central government have formal administrative authority over their functionally similar counterpart local agencies; e.g., the Ministry of Education can instruct local branches on issues pertaining to education, school facilities, and related matters. 64 uncomplicated by the intervention of other actors, such as politicians from the majority party. In place of vertical control, Muramatsu advances a "lateral political competition model." In this model the local authorities compete through policy innovation and political entrepreneurialism so as to gain funding (via specific subsidies, in particular) for favoured programmes and projects. Muramatsu argues that "three-fourths of total government expenditures are made by local governments. Sensitive to people's needs and demands, local governments have created a series of new programs." On the other hand, Muramatsu also notes that this "pluralization" of local policymaking came up against fiscal controls in the 1980s: "the pluralization which resulted in enormous expenditures causing central deficits has invited central government intervention once again, while its real impact remains to be seen." Fiscal deficits thus brought, in reaction, a strategy of recentralization in order to trim the cost of local government. Local expenditures, as a proportion of total government spending, declined from a peak of 76.9 percent in 1980 to 65.1 percent in 1989. But that only returned them to about the level 65 they had been in 1970, when they were 66.3 percent. 5 0 If the policy can be viewed as a recentralization of control over expenditures, then the fact that the outcome was to return to the conditions of 1970 and earlier suggests there was considerable relevance in the vertical-control model's public finance arguments. In fact, Japan's 1940 fiscal regime has always been marked by a great degree of deconcentration in expenditures versus concentration in revenues. Tracking statistical movements in the relative scale of local expenditures does not really "prove" much, except that the share of local government in overall spending has declined since the peak in 1980. Whether that means there was increased fiscal centralization is unclear. The type of expenditures and the influences on them need a deeper analysis before it can be concluded that a sea change occurred. What is clear from the 1970s, though, is that the central state used tax and debt controls to rein in the policymaking autonomy of the most powerful of the local authorities, Tokyo. Here, then, is another case where MOHA's fiscal controls and its two-faced role vis-a-vis local governments are not adequately incorporated into the analyses. While it is correct to question the assumptions of the vertical control model, it is also important to recall that its adherents did not construct it out of thin air. Certainly those who depicted specific grants as unambiguous means 5 0 The data are in Jinshichiro Yonehara, "Relations Between the National and Local Governments," 66 of central control were exaggerating the effectiveness of formal controls and underestimating political variables. But it is equally unwise to downplay the significance of MOHA's fiscal controls, for as Steven Reed concedes, "MOHA's administrative guidance concerning finances cannot be ignored." 5 1 But Reed, like Muramatsu, tends to sidestep the issue of MOHA's controls in his effort to depict a larger sphere for local policymaking autonomy than that advanced by the earlier orthodoxy. Reed notes that the 1970s saw a surprising spurt of policymaking at the local level in Japan. Though he also points out how dependent the activism was on ample local finances, he fails to note the effectiveness of MOHA's debt controls in curtailing local innovation (p. 58-9). The specific mechanisms of control in the Japanese or any other case is important. That is, it makes a significant institutional difference if the limits on local fiscal autonomy are, as in the United States, imposed through the bond markets and groups of electors or, as in Japan, a bureaucratic agency and the conservative political hierarchy in the central state. Institutionalizing the Local System A more recent contribution from Akizuki Kengo, part of a larger study of the role of Japan's bureaucracy in economic development, adopts a clearly meso-level institutionalist perspective on the intergovernmental system and the role Japan's Public Sector, op. c/t., p. 168, Table 10.1. 67 of MOHA. Akizuki depicts MOHA as the pivotal agency in intergovernmental relations and stresses that it and the intergovernmental fiscal system present Japan's local authorities with a "financial safeguard" and provide "a set of stable rules for actors involved in intergovernmental processes, which enhance their willingness to participate." But he also points out that "a well-institutionalized system takes some goals (such as financial independence) and functions (such as rule-making) away from local government." 5 2 Akizuki offers an interesting institutional sketch of MOHA. He points out that the ministry is very small and hence well able to inculcate its members with agency goals. Moreover, the unusually extensive practice of sending member bureaucrats for service in the local areas creates a strong network of contacts between MOHA and its clients, the local governments. Thus MOHA intercedes between the local authorities and the other central-state agencies both in order to protect local interests as well as its own organizational turf. In effect, Akizuki takes a step back from the revisionist perspective that has come to shape most political science commentary on Japan's intergovernmental relations. Instead of downplaying the significance of fiscal controls, he concedes that the controls tend to hem in subnational governments even though they do enjoy more policymaking autonomy than 5 1 Cited in Samuels, The Politics of Regional Policy in Japan: Localities Incorporated?, op. cit, p. 46. 5 2 "Institutionalizing the Local System: The Ministry of Local Affairs and Intergovernmental Relations in Japan , " The Japanese Civil Service and Economic Development op. cit., p. 337. 68 the old model of vertical control suggested. Yet Akizuki neglects the historical dimension, a problem that is most notable when he sets out to explain MOHA's near-fetish for fiscal stability among local governments. Akizuki suggests the source of this policy orientation can be found in MOHA's structural position in the central state: [MOHA] is situated in central bureaucratic politics. It must negotiate with MOF over financial matters and must compete with the line ministries for policy initiatives and jurisdiction. These central rivals will jump on [MOHA] as soon as local governments show any signs of disarray. In order to cope with the pressures from these ministries, [MOHA] must keep local governments in order. [MOHA] is driven to tighten control of local governments by its institutional position at the centre, not necessarily by its perception of local governments. Ironically, [MOHA], which is perhaps the most liberal, understanding, and supportive agency for local governments, is also the most strictly institutionalized for central control over them. 5 3 But there is in fact little irony in the matter, as MOHA and its predecessors were set up precisely to exercise central control over local taxes and transfers. MOHA's powers stem from fiscal history and deliberate choices by postwar bureaucratics to avoid the chaos of decentralization. In other words, it is necessary to examine history in order to understand institutional structure and how the latter helps shape choices in the present. MOHA and the tax structure it oversees (along with MOF) were quite consciously set in place to perform 3lbid., p. 365. 69 the role that Akizuki indicates derives from the current institutional order. Moreover, one of MOHA's major incentives to defend this fiscal system is the competition from MOF. The next chapter turns to examine in detail the fiscal history and politics that gave rise to Japan's 1940 system. 70 Chapter 3 Institutions and Japan's 1940 Fiscal System This tax system was given a revolutionary reform and in design is splendid. But in operation, it has revealed numerous major faults, especially the worsening inequities in local tax burdens...this tax system in fact has no substance (kakaretarumochi), and in my opinion is just a living corpse {ikeru shikabane). Miyoshi Shigeo Home Ministry Bureaucrat (cited in 1931) Here we have with the clarity of a textbook example what we mean by the crisis of a fiscal system: obvious, ineluctable, continuous failure due to unalterable social change. Joseph Schumpeter, "The Crisis of the Tax State" The Prewar Tax System Japan's prewar tax system, at the national level, was largely composed of the Business Tax (Eigyouzei), the Household Tax (Kaokuzei), the Income Tax (Shotokuzei), the land tax, and the Liquor Tax (Shuzei). The local authorities relied heavily on a myriad of local taxes and contributions as well as surtaxes imposed on the national taxes, at flat rates set according to central-state approval. Transfers from the central government, as was the case in most countries, were negligible in the prewar era. The period's fiscal system has in consequence been characterized as one that was "decentralized and 71 deconcentrated" (bunkenteki bunsan shisutemu), because central-state authority over the local tax regime emphasized tax denial rather than outright control (hence, "decentralization") and the public services that did exist were mostly provided at the local level (hence, "deconcentration"). Through tax denial the central authorities sought to keep their local counterparts from expanding and encroaching on tax fields that the central state was using to fund national defence and industrial development. 1 In other words, the, prewar central state's approach to local finances was generally one of benign neglect, so long as local spending and fiscal problems were kept within reasonable bounds. From the fiscal sociology perspective, subnational government was virtually a separate sphere of fiscal activity, one whose scale was restricted by the emphasis on benefit taxation and extremely limited redistribution. As was the case in many other countries, however, the growth of local needs and differences in local fiscal capacity began to pose a fundamental crisis for the prewar system. The crisis in Japan was essentially the same as elsewhere, though of course viewed from the meso-level of analysis it manifested itself and was resolved in a very different institutional context. 1 On "deconcentration and decentralization" as well as "tax denial" and "tax control" in the Japanese intergovernmental f iscal system, refer to Jinno Naohiko, "Nihongata Zei, Zaisei Shisutemu," op. cit. Note also that prewar public services were quite rudimentary and local taxation often involved customary exact ions, for local projects and facilities such as shrines, imposed roughly according to ability to pay. 72 Japanese central state agencies with a direct interest in local fiscal affairs, the Home Ministry and MOF, had thus rather an intermittent interest in local finance until the 1920s. This does not mean they completely ignored this area of administrative activity, because even in the prewar years there were grounds for fiscal competition between the two. In fact, the Home Ministry's founding in January of 1874 was soon accompanied by a struggle with MOF over local tax reform. MOF had been solely in control of local affairs prior to the institution of the Home Ministry,2 and was working on a tax reform at the time. Tax reform had become an issue because of unfairness in the tax system, but local bureaucrats used the opportunity to press for the separation of local from national taxes. Their goal was to secure a measure of autonomous local taxation. The effectiveness of this lobbying for local fiscal autonomy greatly increased with the institution of the Home Ministry, and in September of 1876 MOF was compelled to draw up a law that recognized taxation at the prefectural level as distinct from that of the national level. 3 In the wake of that reform, though, the focus of bureaucratic activism turned to the business of building the state. Local public finance, and especially transfers from the central state, was too small-scale to excite much central bureaucratic attention and internecine conflict. 2 In fact, the Home Ministry was in many respects made up of functions stripped from MOF. On this, note Sydney DeVere Brown, "Okubo Toshimichi and the First Home Ministry Bureaucracy 1873-1878," Modern Japanese Leadership: Transition and Change. Ed . by Bernard S. Silberman and H.D. Harootunian. Tucson: The University of Arizona Press, 1966. 3 In Shinpo Tetsuo, "Meiji Reinendai no Chihou Gyousei : Naimusho to Ookurashou no Soukoku" [Local Administration in the Early Meiji Era: The Conflict Between the Ministries of Finance and Home Affairs], Hougaku Seijigaku Ronkyuu, Winter, 1994. 73 Moreover, the Home Ministry's powers over subnational levels of government were pervasive and direct. The prefectural level was the fulcrum of the overall system, as the Home Ministry appointed the prefect and he oversaw administration at the lower levels. Certainly the ministry had to contend with other central agencies' efforts to use local government as a tool for implementing national policies. But the Home Ministry contained within itself an enormous range of functions; so many in fact that in postwar Japan they were parcelled out to the new ministries of Construction, Home Affairs, Labour, Health and Welfare, and Justice. Hence the Home Ministry fairly enveloped local government, and could impose its ends directly via ordinances and the power of appointment. MOHA's contemporary leverage through control of much of the local fiscal system is a poor substitute for its predecessor's far larger sphere of administrative preeminence. 4 The intergovernmental tax regime came to the forefront of fiscal issues in the wake of the first world war, though, because Japanese capitalism's rapid but uneven development exacerbated discrepancies in fiscal capacity. The search for answers to the problem led to various studies, especially through the Special Fiscal Economy Advisory Commission (Rinji Zaisei Keizai Chousakai), a shingikai set up in 1919 in order to investigate Japan's political economy for areas in need of reform. Of particular concern was stagnation and decline in " S e e , e.g., Kurt Steiner, Local Government in Japan. Stanford: 1965. 74 the rural economy as well as disturbances, such as the 1919 Rice Riots, caused by the failure of urban incomes to keep pace with prices. Alleviating rural distress made the adoption of a redistributive system and comprehensive income taxation appear necessary to social-policy bureaucrats, especially those in MOF and the Home Ministry.5 The major tax issue, and the one most germane for this dissertation, was the heavy incidence of the land tax and other taxes, especially indirect taxes, on rural interests.6 Fiscal imbalance became increasingly serious in the wake of the first world war, as the locus of economic activity shifted from farming to the highly productive new industries in urban areas. The decline of Japan's rural economy relative to the modernized, industrial sector therefore impelled large-scale reform of the tax system. The 1920s saw the bureaucracy mount an effort towards comprehensive tax reform; but strong opposition from business interests stymied this effort to adjust the structure of public finance to 5 Sheldon Garon, The State and Labor in Modern Japan, Berkeley: University of California P ress , 1987, pp. 76-98, d iscusses the specialization of functions in the Home Ministry from the early twentieth century and the emergence of special ist "social bureaucrats" in the area of labour legislation. The phenomenon was also evident in fiscal affairs. 6 Harry Oshima notes that the burden of the land tax played a major role in the sad plight of agriculturalists in late Meiji Japan. The problems the farmers suffered are evident in the frequency of uprisings and other disturbances, along with the rapid increase of tenanted landholdings. He also observes that "[t]he fact that the percentage share of the consolidated tax receipts contributed by the land tax fell sharply from about 60 percent in 1880 to 15 percent in 1910 did not necessari ly lessen the tax burden. In fact it might have increased the burden since a substantial portion of the indirect taxes fell on the peasants as purchasers of the taxed products." See "Meiji Fiscal Policy and Economic Progress," 77ie Sfafe and Economic Enterprise in Japan. Ed. by William Lockwood. Princeton: Princeton University Press , 1965, pp. 353-390. 75 correspond with an industrializing economy. 7 To a large extent, then, the pressure for tax reforms can be seen as intrinsic to the prewar fiscal system. According to comparative research done by H. Hinrichs, general fiscal evolution shows a movement from land taxes and poll taxes to "customs duties and modern forms of direct taxation such as income tax and corporation tax." Seen in this light, we can read fiscal choices in Japan in this era as stemming from the following: the land tax was relatively inelastic, and thus "could not meet the increasing demands of the treasury. Indirect taxes (e.g., on alcoholic beverages, sugar) were raised to supplement [it], but a more income-elastic tax was required -- income tax. It was introduced in 1887 and raised considerably in the 1920s and again in 1940." As we have seen, however, there was a good deal of politics within a process that public finance sees as a secular movement. 8 0 Fiscal stress worsened with the Showa Panic of 1927. Economic development was heavily subsidized via the agricultural surplus, but development drew labour away from the rural areas and into the cities. This contradictory state of affairs naturally eroded the social stability and fiscal capacity of the 7 A good d iscuss ion of f iscal politics in the 1920s is found in Ikegami Takehiko, "Daiichiji Taisen ato Nihon ni okeru Zeisei Seiri Mondai no Hassei " [Japan's Problem of Tax Reform after the First World War], Tohoku Daigaku Kenkyu Nenpou .Volume 48 Number 6, November 1986. S e e also Fujita Yasukazu , "Takahashi Zaisei Shita ni Okeru Chihou Zaisei no Saihensei (1 & 2)" [Reform of Local Finance under the Takahashi Fiscal System], Keizai Ronsou, Volume 143, Number 2-3, 6, February-March, June 1989. 8 The work is cited in Ryoushin Minami. The Economic Development of Japan: A Quantitative Study, 2nd Edition. New York: St. Martin's Press, 1994, pp. 259-60. 76 countryside, and further exacerbated discrepancies among local tax rates. In this milieu, the Showa Panic arrived like a typhoon to wreak even more havoc, and compelled the Home Ministry to undertake intensive research on a programme for fiscal redistribution. In August of 1932, The Home Ministry in fact wrote up a bill, the "Chihou Zaisei Chousei Seido Youkouan" [Outline for a System of Local Fiscal Adjustment], which sought to redistribute according to population and the fiscal resources of local authorities. MOF's Tax Bureau, too, was driven to examine the means available for a system of fiscal redistribution, as organized rural protest manifested itself in politics. However, differences emerged in the two bureaucracies' plans for funding the structure of redistribution. The Home Ministry programme for fiscal redistribution was aimed solely at funding reductions in local tax burdens, through increases in national income, capital-interest, and inheritance taxes as well as the institution of a luxury tax. By contrast, the MOF proposal was for the transfer of the local Household Tax to the national level, after which the funds would be used for fiscal equalization. In other words, MOF was seeking to centralize taxation -- a programme that would have eroded local fiscal autonomy. Nor were these the only proposals on the agenda, as in the wake of the Showa Panic, the Seiyukai 's Tanaka Cabinet came to power and instituted a Tax Advisory Commission (Zeisei Chousakai). The Commission reported in 77 December of 1928, and advocated that the land tax be shifted to the local authorities. This recommendation constituted a reversal from the previous reliance on surtaxes, the hallmark of the Kenseikai governments' approach to local fiscal reform. Reflecting local pressures for devolution of major national taxes, 9 the proposal was drawn up in a bill that included a transfer of the Business Tax to the local authorities, and was then presented to the Imperial Diet in 1929. However, the bill came to naught in the legislative proceedings because of opposition from the aristocratic clique. Afterwards, simply delegating taxes to the local level was not to re-emerge as a serious policy option. The highest political hurdle confronting the reallocation of taxes was the concern that such would undermine the fiscal resources of the central state. Opposition in the Imperial Diet focused on the possible need for greater central revenues in order to fund industrial development and equip the country's military. In addition, however, some actors recognized that the uneven growth of the local economies made such a tax shift unwise. The reason was that the fiscal imbalance among the local authorities would have worsened in the absence of a redistributive mechanism. 1 0 That is, the wealthier local entities would have benefited from the expanded powers of taxation because of their stronger fiscal capacity; but the poorer areas would 9 "The History of Local Government," The Government and Politics of Japan. Ed . by Hitoshi Abe, Muneyuki Shindo, and Sadafumi Kawato. [Trans, by James White]. Tokyo: University of Tokyo Press, 1994, p. 58. 1 0 Suzuk i Takeo, Zaiseishi[Fiscal History]. Tokyo: Toyo Keizai Shinposha, 1962, pp. 146-52. 78 not have been able to exploit the expanded tax room due to their weak local economies and consequently restricted fiscal capacity. Gradually, the policy options narrowed in the direction of fiscal redistribution based on centralized taxation, and especially the lucrative income tax. Table 3.1: Ratio of Local to National Direct Taxes. Selected Prefectures. 1934 (Units: Thousand Yen) Prefecture Direct National Tax (A) Local Tax (B) —67S Tokyo finite 70,642 6.844 Osaka 45,676 42, 981 0.941 Hyogo 20,607 35,720 1.733 Aichi 14,415 27,111 1.881 Aomori 1,538 7,044 4.567 Kagoshima 2, 531 11,830 4.674 Tokushima 1,255 7,083 5.643 Okinawa 237 2,139 9.014 Source: MaruyamaTakamitsu, Nihon Chihou Zeiseishi [The History of Japan 's Local Tax System], Tokyo: Gyousei , 1985, pp. 306. The scale of the fiscal problem in the prewar period is indicated in table 3.1. This table presents the ratios of the amount of local taxes to national direct taxes 1 1 for selected prefectures and major cities. As the data show, the spread of relative tax burdens was very great in prewar Japan, due to uneven economic development and the attendant wide variations in fiscal capacity. 1 1 Note that the direct taxes consist of the Income, Business Profit, Land, Business Transaction, Mining Business, Mining Asset, Capital Interest, and Inheritance Taxes. 79 The crisis of the prewar system is also reflected in the rates for the surtax on the land tax and the rates for the Household Tax. Table 3.2 shows clearly that in 1935 the urbanized and richer prefectures were able to keep their tax rates far lower than those found in their weaker rural counterparts. Hence the plans for fiscal redistribution. Table 3.2: Tax Rates in Selected Prefectures, 1935 Prefecture Rate of Surtax on the Laid Tax Household Tax Rate Toky- 0355 0~65 Aomori 1.467 4.15 Iwate 1.490 2.70 Okinawa 1.695 9.75 N a t i o n a i Average 1.325 4l>3 Source: table in Henry Ludmer, "Japan's Tax Structure 1," The Bulletin For International Fiscal Documentation, Volume 1, 1946-47, p. 313. Note: "Nat"=National. The Home Ministry had a particularly strong interest in these fiscal problems, and not only because of its extensive networks of personnel in local governments. An even more powerful motivating force was the ministry's de facto loss of authority to dictate local tax rates. That the Home Ministry was not able to enforce its jurisdiction over local tax rates (ie, the surtaxes on national taxes) is an important and, for some perhaps, a surprising point. The ministry is the archetype of Japan's prewar bureaucratic authoritarianism, and dominated local governments through the appointment of key personnel and a vast array of directives. But the prewar 80 depression caused such a divergence in local fiscal capacities that the ministry could not credibly demand that tax rates be balanced among the local authorities. Uniform tax rates on grossly unequal tax bases produce commensurately disproportionate streams of revenue per capita. This simple fact made it necessary to increase tax rates in fiscally weaker areas, with the ministry's resigned acknowledgement of the unavoidable. Restoration of effective control over, and balance among, local taxes required redistribution to bolster the fiscal position of weaker administrative entities. But a sufficiently redistributive scheme would require increased taxes to fund it; and increasing taxes, under the prewar fiscal system, meant essentially to impose an even greater burden on the rural interests who were already being heavily taxed relative to urban interests. Shifting more of the tax burden, via income taxes, onto urban business interests had not been accomplished in the 1920s, due to the political coalitions of the Taisho Democracy era (1912-1925). The need for income-tax reform thus presented itself again as a prerequisite to the modernization of the prewar fiscal regime. 1 2 The early 1930s saw an additional powerful actor arrive on the stage of Japanese politics: the military. Pressures mounted for increased spending on armaments and other expenses for more aggressive colonial policies and militarism. This was especially the case in the wake of the assassination of the 1 2 The late 1920s and early 1930s saw heavy but futile pressure from the central state for the local authorities to cut their expenditures, which proved too difficult due to fixed patterns of spending. The only alternative for the locals was to increase the local tax burden. See Suzuk i Takeo, Zaiseishi [Fiscal History]. Tokyo: Toyo Keizai Shinposha, 1962, pp. 155-56. 81 Finance Minister Takahashi in February of 1936, after which the state budget rose dramatically, as is evident in Table 3.3. Table 3.3: National Government Expenditures and National Income in Japan. 1936-1945 (Units: Million Yen) Fiscal Year National Government Expenditures Estimated National Income National Expenditures, % of National Income 1936-37 2,282 16,320 14 1937-38 4, 661 20, 476 23 1938-39 7, 705 23,200 33 1939-40 8,418 28,300 30 1940-41 10,534 32,000 33 1941-42 18,728 36,800 51 1942-43 24, 695 45,000 55 1943-44 36,036 50,000 72 1944-45 49,652 60, 000 83 Source: table in Henry Ludmer, "Japan's Tax Structure 1," The Bulletin For International Fiscal Documentation, Volume 1, 1946-47, p. 313. Tax policy also shows the mark of the new and militarist regime, though in a rather contrary form than cartoons of the absolutist Japanese state suggest. There were efforts to expand revenues to fund the increasing expenditures on armaments. However, the tax policy bureaucrats continued to encounter effective opposition from business interests ~ to the extent that a thoroughgoing overhaul of the tax system, proposed in the 1936 "Baba Tax Reform," was blocked. 82 The Baba reform, which took its name from the minister of Finance, included what were -- even by the standards of the time -- rather drastic remedies for the local finances. In the mid-1930s, subnational governments were receiving 53.1 percent of their revenues from taxes and contributions over which they exercized great discretion, while 35.5 percent came from surtaxes, and 11.4 percent from specific subsidies that had in the interwar period been instituted to cover the cost of primary education and other responsibilities. Indeed, there was an effort to use these transfers to reinforce control over local rates of surtax, but the scale of transfer was insufficient to secure this end. The Baba programme thus sought to restructure the local fiscal regime so that independent local taxes would provide no more than 17.2 percent of local revenues, with 32.4 percent coming from surtaxes, 39.0 percent from general subsidies and 11.4 from specific subsidies. In the words of Kikuchi Shinzo, a contemporary observer, the Baba reform was "an effort to suffocate local autonomy and impose, in place of a deliberately withered local administration, a uniform, centralized bureaucratism." 1 3 But the effective opposition to the Baba measures was not focused on the plans for local finances, and instead centred on the proposal to introduce a comprehensive form of the income tax. This plan was, like previous initiatives, stymied because it would have led to much higher taxes being placed on urban business interests. In the wake of the failure of the Baba Reforms, the 1 3 The details and commentary are in MurayamaTakamitsu, Nihon ChihouZaiseishi [A History of Japan 's 83 North China War Emegency Taxes and others were levied from 1937 to 1939, but their actual fiscal significance was not very great; certainly they produced nowhere near the volume of revenues needed to match Japan's escalating expenditures. The low productivity of the tax system plus the continued instability in the rural areas hence forced a major tax reform in 1940, one that drew heavily on the Baba programme. 1 4 The 1940 Tax Reform The argument that there were at least two major policy objectives in the 1940 reform is still somewhat controversial in Japanese public finance circles. Mainstream opinion -- to the extent that it pays attention to the war years --views the reform as motivated solely by the need for revenues for warfighting, an empirically untenable argument that perhaps reflects an understandable revulsion for the wartime fascist state rather than an objective appraisal of fiscal history. 1 5 Recent empirical research shows the moralistic position to be clearly misinformed, as there is indisputable evidence that a powerful system Local Finance], Tokyo: Gyousei , 1985, pp. 314-25. 1 4 Henry Ludmer, "Japan's Tax Structure," Bulletin For International Fiscal Documentation, Vol. 1, 1946-47, p. 295; and Henry Shavel l , "Postwar Taxation in Japan, " Journal of Political Economy, Volume 56, 1948 (February-December), pp. 124-25. 1 5 A c lass ic statement of the former position is Hayashi Takehisa, Nihon ni okeru Sozei Kokka no Seiritsu [Constructing the Tax State in Japan], Tokyo: Tokyo University Press , 1965, pp. 294-300. Contrary views concerning the wartime f iscal record can be found in Ak i raAmakawa "The Making of the Postwar Local Government Sys tem," in Democratizing Japan: The Allied Occupation. Honolulu: University of Hawaii P ress , 1987, pp. 256-7; Jinno Naohiko, "Nihongata Zei, Zaisei Shisutemu," op. cit.; and Jinno Naohiko, "Gendai Nihon Zeisei no Keisei Katei" (1 & 2) [The Process of Constructing the Contemporary Japanese Tax System], Keizaigaku Zasshi, Vol . 88, No. 2-3 and 5-6, 1991. I would add that I use the term "fascist" in spite of Peter Duus and others' criticisms of applying it to wartime Japan . 84 of intergovernmental redistribution was constructed, one aimed both at maintaining social stability in the rural areas as well as reinforcing them for their role in the war effort. Table 3.4: Changes in the Structure of Local Finance (Units: Percentage of Revenues) 'Year Local Taxes (A) LTT(G) Fees/Charges (D) Subsidies (E) Transfers from Centre (B+C+E) 1931 6.26M 6 . 1 2 4 6.686 6.686 1933 6.182 6.688 6.161 6.161 1935 0. 231 0.114 0.093 0.093 19371 6.237 6.131 6.699 0. 099 1938 6.238 0.129 6.16s 6.108 1939 0.230 0.123 0. 149 .0. 149 1946 6.266 6.692 6.119 0. 120 6. 2l2 1941 6.268 6.166 6.113 6.131 6. 251 1942 0. 194 0.126 0.084 o. 146 0.273 1943 6.164 6 . 111 6.694 6.188 6 .29§ 1944 6.204 0.162 6.116 6.197 6.358 1945 0.184 0.158 0.097 0. 261 0.419 1946 0. 118 0. 082 0. 058 0.436 0.568 1947 6.262 6.154 0.623 6.323 0. 5l7 1948 0.280 0. 180 0.024 0.320 0.500 1949 0.343 0.163 0.030 0.244 0.407 1950 6.346 0.199 6.628 6.209 6.463 1955 0.328 0.019 0. 138 0.036 6.254 6.41o 1960 0.355 0. 017 0. 149 0.032 0.227 0.393 1964 6.345 6.611 6.164 6.0271 6.228 6.463 Note: LTT=Local Transfer Tax, LAT=Local Allocation Tax . Source: Table in Noguchi Yukio, 1940-Nen Taisei [The 1940 System], op.cit, pp. 124. 85 This new redistributive role of the central state is evident in table 3.4, which shows the prewar, wartime, and postwar levels of redistribution to the local authorities. The table depicts very well the fact that the level of transfers from the centre underwent a sea change during the Pacific War. From 1940, the Local Allocation Tax (LAT) 1 6 and subsidies from the national treasury provide an increasingly significant portion of local revenues. These transfers were based on the centre's assumption of control over local resources, which gave it potent fiscal tools that will be discussed in detail below. Japanese politico-economic studies are constantly under siege from culturalist claims and other superficial assertions of the country's uniqueness. 1 7 Hence, it is important to repeat that wartime centralization of the tax system -- and especially the income tax -- was the norm among belligerent nations. All countries faced the need to raise funds efficiently, with as little harm to price levels as possible. The war years saw, in consequence, the institution of income-tax collection systems that took regular deductions from the working class, which previously had been exempt from such taxation. Japan's 1940 1 6 In fact, during the war it was -- as we have seen -- referred to as the Distribution Tax, but as the principle of redistribution remains the same, it is more convenient to assign the most recent name to the tax. 1 7 For example, Mochida Nobuki, a public finance economist at the University of Tokyo, writes that "Japan's sys tem of equitable tax allocations, which has no equivalent in the West, drastically reduces residents'Tieboutian voting with their feet between regions. At the same time, Japanese society also places great value on long-term relationships and ex post facto equity, and this shared idea of fairness prevents the American idea that bankruptcies are an integral part of decentralization from gaining ground in Japan"; see "Balancing Equity and Decentralization," Social Science Japan, November 1995. In fact, not only is f iscal equalization common elsewhere, but Sven Steinmo's work handily shows that 86 institution of regular income tax deductions, wherein the employer acts as the collection agent for the state, preceded that of its adversaries by a few years and was in fact a policy "re-import" from the Japanese colony of Manchukuo (Manchuria). 1 8 But in overall fiscal terms, Japan did not deviate much from the rough norm of tax policymaking among the warring nations. 1 9 Moreover, it was common elsewhere for fiscal centralization to meet other objectives in addition to the goal of raising revenues for warfighting. In Canada , for example, the fiscal pressures of the war provided an opportunity to overcome the opposition of the larger provinces and implement a landmark set of recommendations to centralize taxation and create a system for redistributing resources to the poorer provinces. Some of these areas had been so ravaged by the Great Depression and the protracted drought of the 1930s that large numbers of their young male residents were unfit for wartime service due to chronic malnourishment. Against a national average of 48 percent, the prairie provinces of Saskatchewan and Alberta saw respective declines in provincial per capita income between 1928-29 and 1933 of 72 American public policy choices are due more to institutions rather than vague and generally conflicting culturalist variables (especially "values"). Jinno Naohiko, 1993, op. cit., p. 234 notes that at-source collection of the income tax was instituted in Manchukuo in 1911. The UK and the UK, on the other hand, adopted the same measure in 1943 and 1944, respectively. The Nazi leadership in Germany also centralized the tax system, extended the redistributive sys tem set up to cope with local f iscal cr ises during the Weimar Republic, and shifted the burden of income taxation downward onto middle and lower-income earners. A useful review of the tax changes in Germany, including comparisons with those implemented by the wartime Allied powers, is presented in Mabel Newcomer, "War and Postwar Developments in the German Tax System," National Tax Journal, Volume 1 Number 1, March 1948, pp. 1 -11. 87 percent and 61 percent. Japan's highly centralized "1940 fiscal system" saw local governments lose, via redistribution, the capacity to determine rates of local surtaxes. As seen in tables 3.1 and 3.2, the Home Ministry had previously been forced to allow enormous deviations in local tax rates because it lacked the fiscal tools to enforce its own area of jurisdiction. But the 1940 tax reform modernized the income tax and thus shifted the burden of taxation away from the fiscally strapped countryside. Moreover, the more equitable tax structure underpinned a new system for redistributing taxes, one that gave the Home Ministry the means to enforce standard rates, as it could withhold transfers in the event of deviations for which it had not given explicit approval. As one analyst notes: The Home Ministry initiated the local tax reform of 1940, which established a Local Distribution Tax system to ensure adequate local revenues and adjust financial inequalities among the local governments...This system increased the local tax revenues as a whole, but it also made the local governments increasingly dependent on these tax allotments. 2 1 The new local tax regime had two major facets. One was the centralization of local land, Business, and House taxes through collection by the national authority; these revenues were subsequently returned to the prefectures from which they had been taken. The measure was hence called the system of 2 0 See "Book 2: Recommendations," in Report of the Royal Commission on Dominion-Provincial Relations. Ottawa: Queen's Printer, 1954. Data cited are on page 150, Table 50, of Book 1 of the above Report. 88 "Refund Taxes" (Kanpuzei) and helped bring local taxes, and decisions concerning them, firmly within the grasp of the central state. The second element of the new intergovernmental tax regime was the system of Distributive Taxes (Haifuzei). This structure saw set percentages of selected national taxes transferred to the Distribution Tax account and thence redistributed to local authorities on the basis of calculated revenue capacity and overall fiscal need. The former assessment was made by totalling revenues from the respective locality's surtaxes on the land tax, House Tax, and Business Tax, with the distribution of central funds being done in inverse proportion to this sum. The latter item, fiscal need, was based on the local authority's population, with disbursements being made in direct proportion to this figure. The tax base funding the structure included 16.55 percent of the classified income tax, the aggregate income tax, and the corporate income tax; and added to this, 50 percent of the Amusement Tax and the Admission Tax. 2 2 Thus the hallmarks of the new intergovernmental tax structure were centralized collection and redistribution. The system emphasized extensive intervention towards promoting uniformity, and both the fiscal means and the end they served marked a sharp contrast to the previous separation of local 2 1 Amakawa Akira, "The Making of the Postwar Local Government System," Democratizing Japan: The Allied Occupation. Ed by Robert E. Ward and Sakamoto Yoshikazu. Honolulu: University of Hawaii Press, 1987, pp. 256-7. 2 2 An excellent source on the tax policies undertaken in this period, one used for much of the above d iscuss ion of the Local Apportionment Tax, is Saburo Shiomi, Japan's Finance and Taxation, 1940-1956. New York: Columbia, 1957. Note especially "The Tax Reform of 1940," pp. 33-52. 89 and national taxation and the large gaps in tax rates among the local authorities. In one analysts words, "the 1940 reform was in formal terms the final reform of the old local tax system, but in truth it marks the point of departure to the contemporary local tax structure."23 Enhanced Tax Control One of the most important points to emphasize here concerns the framework for decisionmaking on two types of local taxation. The prewar regime of tax control was rather loose, or "deconcentrated." On the one hand this meant that the Home Ministry lacked the fiscal means to enforce its directives on rates of surtax. But it was also the case that much local taxation was highly informal, consisting often of contributions for local services such as shrines and their maintenance. These contributions were not mandated by law, but rather by community norms, and were generally levied on the basis of externally assessed ability to pay. The centralization of the tax state during the war put de facto decisionmaking authority over official local taxation -- i.e., surtaxes on national taxes -- into the hands of the Home Ministry's Local Affairs Bureau. This change thus brought nominal and effective control over local taxation into the same hands, for as noted earlier de jure authority had always resided in the Local Affairs Bureau. 2 3 Murayama Takamitsu, Nihon Chihou Zaiseishi [A History of Japan's Local Finance], Tokyo: Gyouse i , 90 But the transformation of the rural socio-economic landscape also greatly eroded the place of the customary local levies, and therefore fiscal redistribution and the modern tax state had to replace more informal mechanisms of funding local services. This process reinforced tax centralization and deepened the local authorities' dependence on the central state. Both of these changes contributed to the emergence of the contemporary "centralized-deconcentrated system" (shuukenteki, bunsan shisutemu), which couples centralized fiscal controls with the decentralized delivery of services. 2 4 The Postwar Reforms After Japan's surrender in 1945, a series of fundamental tax reforms was implemented under the American Occupation. The reforms generally centred on raising revenue, both to introduce a greater degree of fiscal equity via income taxation and to control inflation through the absorption of mass spending power. In addition, there was the pressing need to finance the postwar state and the burden of the Occupation, as well as the perceived need to expropriate large fortunes. These policies, implemented in the midst of a defeated economy that had lost nearly a quarter of its physical stock, took a heavy toll and helped produce a robust anti-tax sentiment. Indeed, popular opposition to tax assessments at times erupted into "mass demonstrations, 1985, p. 350. 91 attacks on tax offices, burning of tax records, and beating of tax officials." 2 5 This chaotic environment was exacerbated by great fiscal instability among the local authorities and consequently a heavy dependence on central-state transfers. This is evident in Table 3.4, which shows that subsidies skyrocketed to comprise nearly half of local revenues in 1946. Fiscal Decentralization Under SCAP In this very early postwar period, S C A P ' s desire to foster "the spirit of local autonomy among the people" was the motive force behind Japanese tax policymaking, and the initial response to local fiscal crisis was to decentralize the tax system. Revisions were made to the Local Tax Law and Local Apportionment Tax Law to allow higher levels of local surtax on the Land, House, and Business taxes. But these were relatively minor changes. S C A P ' s decentralist reforms included far more aggressive measures, such as extending an Inhabitants Tax (Juuminzei) to the prefectures and allowing both levels of subnational government to levy extra-legal independent taxes in order to secure a larger measure of flexibility in local taxation. 2 6 Jinno Naohiko, "Nihongata Zei, Zaisei Shisutemu," op. cit., pp. 224. Martin Bronfenbrenner and Kichiro Kogiku, "The Aftermath of the Shoup Reforms" (Parts 1 and 2], op. cit. Shiomi Seisaburo, Japan's Finance and Taxation, 1940-56, op. cit., p. 67. Note that "extra-legal" (houteigai) refers to taxes not written into the local tax law. A cornerstone of MOHA's contemporary fiscal control is its ability to permit these taxes, of which there are very few. 92 Later, in the spring of 1947, revisions to local taxation included a wholesale downward reallocation of decisionmaking authority and the right to autonomous occupancy of important tax fields. The new constitution, promulgated in the same year, guaranteed local autonomy; and as one means to this end, major elements of the Local Apportionment Tax regime were placed in the hands of the prefectural governments, with municipalities permitted to levy surtaxes on the new prefectural taxes. Standard rates still applied to the independent prefectural taxes, but to avoid rigidity in the local tax base there was again a generous recognition of excess taxation. In addition, several narrow-based consumption taxes were transferred to the local level as well, including the Admissions Tax. The next major development came in July of 1948. Shiomi Seisaburo expresses the character of the changes and the rationales for them quite succinctly: As local autonomy progressed, the necessity for local independence was stressed and the Local Finance Law was enacted to implement the Local Autonomy Law. Accordingly, this latest revision was so framed as to make local taxes independent of national taxes and to minimize supervision of the central government over local finance. The Ministry of Home Affairs [i.e., Naimushou], which had been handling the local administration for many years, was dissolved at the end of 1947 and the Local Finance Commission was created to take over these duties. 2 7 "Ibid. p. 75. 93 Shiomi goes on to note that S C A P was hardly through with its reforms. The same package of measures included the elimination of central controls over several local taxes as well as the institution and revision of extra-legal independent taxes. In Shiomi's words, "[t]hese were entirely novel ideas and particularly noteworthy as a turning point in the history of the local taxation system in view of the fact that local finance had been at the mercy of the central government ever since the institution of the taxation system in Japan far back in the Meiji Era (1868-1968). 2 8 Of course, the diametrical opposite of the rapidly decentralizing postwar tax system was the 1940 system. Prior to the war, the national and local tax systems were, as Shiomi himself notes, quite separate entities, and the absence of a redistributive system posed a strong hindrance to effective centralized tax controls. However, Shiomi's comments suggest how profound a departure from past Japanese practice the reforms were, because they sought to enshrine local fiscal autonomy in the law. Shiomi 's ready appreciation of the extent to which the reforms posed a challenge to central authority hints at how strongly opposed many former Home Ministry bureaucrats were to such an explicit and codified erosion of central authority. But these bureaucrats could only, for the duration of the Occupation, attempt to subvert decentralization through linguistic obfuscation and legislative ™lbid. p. 76. 94 obstruction behind the scenes. Indeed, in the middle of the apparent dominance of policymaking by the Occupation authorities, the bureaucrats were often busily crafting means to maintain or enhance their powers vis-a-vis one another and in relation to civil society. Kurt Steiner and others have, for example, handily documented the efforts made by the Home Ministry bureaucrats to manage the process of decentralization the Occupation was committed to imposing on Japan. Of particular concern to the Home Ministry leadership was the Occupation's insistence on democratizing local administration through direct election of the prefectural governor. The prefecture was the linchpin of the Home Ministry's administrative rule over the local authorities, a hierarchy of centralized control that was borrowed from the Napoleonic French model. In both regimes, the prefecture -- in France, the prefect-- was a middle level of government set up specifically to oversee local affairs and was headed by an appointee from the Interior Ministry. Losing the power to select this official was thus perceived in the Home Ministry as a serious detraction from the bureaucracy's authority, and the reform was therefore obstructed as diligently as was possible under the Occupation. Nevertheless, the Home Ministry was stripped of the power to appoint prefectural leaders, even as its control of the local tax base was being undermined. And then it lost its integrity as a ministry per se when the 95 Occupation dismantled it in 1947 as part of its reorganization of local government and the Local Autonomy Law. From the remains of the Home Ministry's dismemberment, the Occupation spun off an array of agencies concerned with construction, labour, elections, and local finance. Subsequent Occupation reforms sought to bolster the autonomous role of the latter of these new agencies. But having lost direct authority over local government personnel, the former Home Ministry bureaucrats fell back and regrouped at the line of control over local finance, there to make their stand against fiscal decentralization. Indeed, though Occupation-inspired purges hit the Home Ministry particularly hard, 2 9 its displaced bureaucrats were prominent in the subsequent rollback of the Occupation's reforms to the taxes, laws, and institutions that comprise Japan's intergovernmental fiscal relations. The Dodge Line Shock and the Shoup Report In fact, the Occupation's shifting priorities appear to have encouraged the reassertion of centralized fiscal controls. Most notably, the 1949 imposition of the Dodge Line included a massive cut in local taxation, which caused havoc in local finances. Joseph Dodge was an American banker brought to Japan by S C A P and given wide-ranging powers to impose drastic solutions to the country's prolonged bout of inflation. As part of his programme to crush 2 9 Kent Calder notes that 1300 officials from the Home Ministry were purged, "representing 70 percent of those purged in the entire civilian sector of the government"; see Crisis and Compensation: Public Policy and Political Stability in Japan. Princeton: Princeton University Press, 1988, pp. 150-1. 96 inflation and "over-balance" the Japanese central government budget, the proportion of the national personal and corporate income taxes transferred to subnational governments went from the 1948 setting of 33.14 percent to 16.29 percent. The redistributive tax system was thus abruptly axed in half to accomodate Dodge's demand for drastically reduced expenditures. Moreover, Dodge's nostrum of economic dogma, through starving Japan's corporate sector of funds, nearly achieved the economic collapse that General Curtis LeMay's wartime bombing had not. However, the bureaucracy wisely found the means to supply alternative sources of finance and thus check the ominous spread of bankruptcies and unemployment. But for the local authorities, especially the weaker ones, the Dodge Line's insensate fiscal austerity hit like a shock-wave. 3 0 Several months prior to the appointment of Dodge, the public's anti-tax sentiments and the confusion in intergovernmental affairs saw Japanese authorities and elements of the Occupation's Internal Revenue Division request the appointment of a blue-ribbon Tax Mission. The mission came to be headed by Carl S . Shoup, a highly respected public finance scholar, who hand-picked a small team of America's best and brightest in the field. The Note Steiner, Local Government in Japan, op. cit., p. 106. 97 Japanese were hoping for recommendations of recentralized taxation, 3 1 but what they got instead was effectively a plan to federalize the fiscal state, through a strict division of tax fields and functions among its component levels of government. Shoup's report can, as Shiomi suggests from an historical perspective, be seen as a strong distillation of Anglo-American tax ideologies applied liberally to Japan. Up to that point, Japan had been governed on the tax front largely according to indigenously selected Continental European examples. The YnsXShoup Report was delivered in 1949, with a less significant second installment coming in 1950. The analysis of intergovernmental fiscal relations in Japan dwelt much on the need for local fiscal autonomy, defined largely as a clear separation of tax bases, tax collection authority, and programmatic responsibilities among the three main levels of government. The philosophical core of this approach was the characteristically American conviction that democratic behaviour was fostered by strong local governments and the ability of electors to perceive which level of administration bore responsibility for specific taxes and programmes. In order to bolster local fiscal autonomy, Shoup also recommended that the Distribution Tax (known as the Haifuzei from 1948) be replaced by a system of equalization grants (Heikou Koufukin), whose methods of calculation and transfer to recipient governments were to 3 1 Martin Bronfenbrenner, "Dr. Shoup Revisits Japan: The Shoup Tax Mission Forty Years Later,"777e Japan Foundation Newsletter. Vol . 16, No. 4, March 1989. Bronfenbrenner was a Tax Economist in S C A P ' s Internal Revenue Division in 1949-50. 98 be transparent and based solely on local fiscal requirements. A further recommendation in the Shoup Report was that a Local Finance Commission be set in place and take an independent role in coordinating fiscal relations between the central state and the local authorities. Such a commission was, in essence, to be a greatly altered and far more autonomous version of the agency that had already been created through the dissolution of the previous Home Ministry. The membership of the new Local Finance Commission, in an effort to ensure strong local representation, was to be drawn heavily (3 out of a proposed 5 members) from among the six national associations of governors and other local administrators (Jichi Roku Dantai). Moreover, the powers exercised by the Commission were to include the allocation of shares of maximum allowable local debt, determination of the basis for distributing equalization grants, and temporary suspension of restrictions on the maximum rates of local taxes when it deemed appropriate. 3 2 While not America's radically decentralized federalism, where only the financial markets and local electors hold the local authorities accountable, Shoup 's intergovernmental tax regime sought to amalgamate the previous Occupation reforms and create a coherent system of decentralized authority over taxation and spending. It was thus an enormous red flag waved right in the faces of many of the Japanese fiscal authorities. See the sect ion on "Intergovernmental Fiscal Relations," The Report on Japanese Taxation by the Shoup Mission, Vol. I.Tokyo: General Headquarters for the Supreme Commander for the Allied Powers, 1949, pp. 21-31. 99 The Deconstruction of Shoup This last round of the Occupation's tax reforms is of great significance, but not because -- as many still hold -- the current Japanese tax system is based on Shoup . 3 3 On the contrary, the reforms are important because the tax and institutional changes based on Shoup and other S C A P measures were so quickly and consistently deconstructed in favour of the wartime fiscal system. With the advantage of hindsight and the literature on transwar phenomena, we can see that the Shoup episode and its aftermath do not bespeak the failure to recommend "realistic" tax policies that fit with Japan's fiscal traditions. In fact, many of Shoup's reforms to modernize Japanese tax administration were retained. Revisions to the Shoup system centred on fostering capital accumulation and gutting local fiscal autonomy, core elements of postwar Japan's centralized developmental state. These particular foci of rollback give evidence of wartime learning, expressed in the revanche of the 1940 system and its institutions.3 4 Hiromitsu Ishi adopts this position concerning the relationship between the Shoup Report and the current tax system, in "The Impact of the Shoup Mission,"The Relevance of Public Finance for Policymaking. Ed. by Hans M. van der Kar and Barbara L. Wolfe. Detroit: Wayne State University Press, 1987. One of the best accounts of the political background to the Shoup Report and its deconstruction is Martin Bronfenbrenner and Kichiro Kogiku, "The Aftermath of the Shoup Reforms" (Parts 1 and 2), National Tax Journal, Vol . 10, Nos. 3 and 4, September and December, 1957. Their work suggests that Shoup was in large measure a failure, due to a lack of realism in some c a s e s as well as an unfavourable political context. See also Kozo Yamamura, Economic Policy in Postwar Japan: Growth Versus Economic Democracy. Berkeley: University of California Press, 1967. 100 Fiscal Control for Equity as Uniformity To repeat an earlier point, the existence of a centralized fiscal regime is not, in itself, remarkable. The postwar expansion of the welfare state in the advanced industrial countries was often accompanied by the central or federal level's dominance of major tax fields, and especially the income tax. What stands out in the Japanese case is the return to much of the pattern of 1940 in the wake of large-scale fiscal decentralization under Shoup and earlier Occupation reforms. Understanding the process of recentralization involves grasping the range of motives of the actors involved. It is important, of course, to point to the role of naked self-interest; but not to the extent that illustrating the politics of recentralization becomes merely a caricature, wherein sworn enemies of fiscal democracy callously put S C A P ' s system to the sword. Emphasizing malign intentions focuses the analysis too much on control for its own sake and thus highlights only one aspect of a larger story. For as I have alluded to earlier, a defining experience of MOF and Home Ministry bureaucrats was the prewar fiscal crises of the local authorities. The political salience of dire local circumstances is evident in repeated prewar efforts to secure tax reform and the institution of a redistributive system. The Home Ministry, institutionally exposed to the crises through direct contact with subnational governments, and having de facto lost its control over local taxes, was certainly far more 101 cognizant of and concerned about the problems than MOF (a fact, we shall see, that is even more apparent in the postwar era). But the scale of the problem was such that MOF as well was drawn into the search for solutions. The postwar re-emergence of fiscal stress at the local level thus helps account for the return of the 1940 fiscal system. Japan was thoroughly vanquished in the war and gripped by chaos on multiple fronts. Economic output in the wake of the surrender plummeted, and the central state tried desperately to pump liqidity into the economy via subsidies, thus igniting a furious inflation. Table 3.5, it will be recalled, shows that subsidies also flowed copiously to the local authorities in the years 1946 to 1948, when aggregate transfers constituted half or more of the local authorities' revenues. Then the sudden imposition of fiscal austerity via the Dodge Line saw local governments thrown into very dire straits indeed. With massive wartime destruction, raging inflation, and severe housing and employment shortages, the fiscal problems of local governments were one emergent catastrophe among many that demanded solution by the state and the Occupation. The Shoup reforms offered one answer, but one inconsistent with past practice and with no popular social base. The reforms were also a deliberate repudiation of the fiscal controls that MOF and Home Ministry bureaucrats had crafted to deal with similar difficulties in the pre-Pacific War period. The return of an old problem, especially in such an environment of great uncertainty, simply begs to be resolved by indigenous methods that are not only believed to have worked in the past, but also 102 enhance the authority of the actors in a position to advise or engineer their adoption. This is not to say, however, that fiscal recentralization was wholeheartedly supported by all the bureaucrats within the Home Ministry and its successors. Many of the bureaucrats within the old ministry and the local finance agencies that were created with its dissolution were, on the contrary, favourably disposed to decentralizing the fiscal state. There is, as was discussed earlier, an indigenous Japanese history of effective local autonomy from the central feudal state as well as movements against the centralization of the later Meiji state. Local fiscal autonomy was for some a progressive move in reaction to a national state whose policies had for long years emphasized militarization and the regimentation of daily life. However, conditions in the early postwar era were not particularly amenable to this goal of decentralization. The centralists held the upper hand, in large measure because of the extreme fiscal crisis that Dodge exacerbated and that continued into the late 1950s. Debt climbed among the local authorities, with the worst year being 1954's 64.8 billion yen of local debt versus the budgeted forecast of 14.0 billion yen. This brought about, on November 25 of 1955, the opening of an Emergency Diet on countermeasures for the local fiscal crisis --103 an "unprecedented measure" (zenko mizou no sochi).35 Under such conditions, the centralists were able -- in drafting reforms to the local finance laws and deliberations of relevant shingikai committees -- to insist that the choice was the risky path of bolstering local fiscal autonomy or the old policy of centralization in order to reinforce local finances. The records of the shingikai and other agencies clearly show the concern of the bureaucrats and their consistent post-Shoup emphasis on fiscal recentralization. The Local Finance Commission Report of 1951, for example, calls attention to the local fiscal crisis and the problem of taxation beyond the legally fixed rates (which was allowed through S C A P and Shoup reforms). By 1953, both the Local Autonomy Agency and MOF, via the GTAC, were expressing serious concerns about increasing divergence in taxation among the local authorities. And in 1954, the Local Autonomy Agency report argued explicitly that, in order to avoid fiscal imbalance, local autonomy had to be curtailed through the recentralization of taxes and other measures. The report portrayed two choices: Shoup's independent taxes (dokuritsuzeishugi) or the tradition of surtaxes (fukazeishugi), and strongly suggested that reversion to the latter was most appropriate for a "small country like Japan" (semai Nihon no kuni).36 3 5 Note p. 47 in Ogita Yasu , "Chihouseido Chousakai no Ayumi" [The Course of the Local Sys tem Advisory Commission], Chihou Jichi no 30-Nen [50 Years of Local Autonomy], (ed. by Nihon Gyouse i Gakkai) , Tokyo: Gyousei , 1979. 3 6 These reports are "Chihouzaisei no Jokyouhokoku" by Chihouzaisei linkai, 1951; "Zeiseichousakai Toushin" by Zeiseichousakai, November, 1953, pp. 4-6; "Chihouzaisei no Jokyouhokoku" by J ich ichou, March, 1953, pp. 46-52; and "Chihouzaisei no Jokyouhokoku" by J ichichou, March 1954, pp. 27-33. A 104 Pressures for recentralizing local finance did not only emanate from fiscal bureaucrats. One of the major revisions of Shoup was the rejection of the innovative Value Added Tax, which Shoup had envisioned as an independent tax base for municipalities. The measure failed to make it through the implementation stage to a large extent because of opposition from local businesses and many residents. These interest groups were adamantly against what they perceived as increased taxation and hence opposed the new local property taxes as well. Local politicians and administrators, too, generally felt uncertain that their governments possessed the fiscal and managerial capacity to perform the new tasks they had been assigned. This opposition of local interests and governments helped reinforce the position of the central bureaucrats who stood against Shoup's programme for local fiscal autonomy. 3 7 But interest groups mobilize on specific tax policy issues, the ones that affect them most, and hence cannot be the authors of an entire fiscal regime. useful d iscussion of the f iscal history can be found in Honma Yoshihito's interview with Kobayashi Yosoj i , a former bureaucrat from both MOHA and the Home Ministry, in Shougen, Chihou Jichi: Namushou Kaitai-Chihou Bunkenron [An Oral History of Local Autonomy: The Breakup of the Home Ministry and the Debate on Local Decentralization]. Tokyo: Gyousei , 1993, pp. 38-45. Martin Bronfenbrenner, "The Japanese Value-Added Sales Tax," National Tax Journal, Volume 3, 1950, pp 298-313, d i scusses popular resistance to the VAT. One example of bureaucratic efforts to stymie Shoup's plan for a VAT is seen in MOF Tax Bureau Chief Hiraita Keiichirou's call for further study - the administrative k iss of death in Japan -- of the tax because of large-scale opposition and the question of whetherthe prefectural officials had sufficient skill to manage the tax's collection. Indeed, Hiraita goes onto reject the whole premise of Shoup, that there should be separate taxes assigned to the central, prefectural, and municipal levels of government, on the grounds of Japan's being a "poor and small country" (binbou de semaikuni); see "Zeisei Kaikaku no Mondaiten" [Problems in Tax Reform], Kinyuu Zaisei Jijou, Volume 1, Number 1, 1952, pp. 42-47, especially pp. 45-6. 1 0 5 Certainly their piecemeal efforts in postwar Japan did not add up to an overall programme of reform of the intergovernmental tax system. Hence, in terms of agency, the large-scale restructuring of this system in postwar Japan was not due to pressures from private lobbies, but rather the contest between the decentralist path opened by Shoup and the Occupation and the mainstream, centralist tendency of the postwar bureaucracy. The Redistributive System One of the most salient areas of the wartime system's revanche was seen in redistribution. Shoup's system of equalization grants, supervised by an autonomous Commission, was to a great extent implemented in June of 1950. But as part of a general recentralization of the state, in the "reverse course," the Local Autonomy Agency (Chihou Jichichou) was reorganized in August 1952, and renamed simply the Autonomy Agency (Jichichou). The new agency also absorbed the Local Finance Commission (Chihou Zaisei linkai), which caused some administrators in the latter agency much consternation concerning the fate of the recently implemented reforms in local finance. Worry centred on the issue of whether the proposed reorganization would undermine the autonomous management of the redistributive system. 3 8 The chair of the Local Finance Commission went so far as to publicly express his concern that the administrative changes would detract significantly from local autonomy. See Maruyama Takamitsu, Nihon Chihou Zeiseishi[A History of Japan's Local Tax System], Tokyo: Gyousei , 1985, p. 452. 106 However, the position taken by the new Autonomy Agency was to avoid as much as possible any outside interference in its jurisdiction. Hence, during preparation of the Diet bill for amalgamating the Local Autonomy Agency and the Local Finance Commission, the new Agency's leadership strongly opposed proposals to have the Local Finance Advisory Council (Chihou Zaisei Chousakai), one of the Agency's affiliated organizations, become anything more than simply an institution for submitting questions to (i.e., shimon kikan).39 In addition, changes were made to the redistributive system, effectively pulling it back in the direction of the wartime fiscal state. Shoup had sought to implement a system that eliminated the centre's control through specific subsid ies 4 0 and the absence of local involvement in determining the levels and means of redistribution. But the pressure of central-state actors saw the equalization grant system eliminated in May of 1954, and replaced by the Local Allocation Tax. The amount of funds to be redistributed was the focus of much friction between 3 9 On this, see "Chihou Zaisei Shingikai no Seikaku ni Iken Chousei: Chizaiin, Haishigo no Kouf ukin Santei , Haifu ni Kuryo" [Ideas Concerning the Character of the Local Finance Advisory Council], Kinyuu Zaisei Jijou, Volume 3, Number 16, April 14, 1952, pp. 11-12. 4 0 Shoup sought to clarify national-level and local-level functions as well as those which are shared. The Shoup Report then recommended that the costs for such programmes be covered by the relevant level of government, in order to avoid the recurrent problem of centralized control through the manipulation of subsidies and the imposition of excess ive costs on the local authorities. But the sys tem was not put into place, and instead the current order allows the national ministries to impose burdens on the local authorities using thelatters' resources. See , e.g., Kurt Steiner, Local Government in Japan. Stanford: 1965, pp. 283-91. 107 MOF and the Local Autonomy Agency in the annual budgetary process. MOF, especially the Budget Bureau, constantly sought to cut the level of transfers to the local authorities whereas the Local Autonomy Agency pressed to have them increased. The scope for conflict was narrowed a good deal when the two compromised, in 1954, on including fixed percentages of the national corporate and personal income taxes and the liquor tax in the new Allocation Tax. The latter tax was then effectively transferred to the Autonomy Agency and distributed to the local authorities according to a complex formula. This arrangement continues, with the current Ministry of Home Affairs having been set up in 1960 to replace the Autonomy Agency. 4 1 At first glance, the above redistributive structure appears quite depoliticized. There are, as noted, regular and legally sanctioned procedures for determining the approximate volume of Allocation Tax revenues to be distributed as well as a formula governing that distribution. However, as I examine in more detail in chapter 5, the local authorities are not represented in an autonomous and disinterested commission.-- Shoup's proposal -- when decisions are made on such matters as the percentage of national taxes to be included in the Allocation Tax, the subsequent redistribution of revenues, and allowable levels of local debt. Moreover, 6 percent of the total funds for the 4 1 The institutional changes leading from the December 1947 breakup of the Home Ministry to the July 1960 creation of MOHA are d iscussed in chapter four, "Chihoujichi ni okeru Gyouse i Kikou no Hensen" [Changes in the Administration of Local Autonomy], Chihoujichi [Local Autonomy]. Ed. by Takatsuji Masami, Tokyo: n.p., 1983. See also Kurt Steiner's more explicitly political account of the gradual revanche of the Home Ministry's Local Affairs Bureau in his Local Government in Japan. Stanford: Stanford University Press, 1965, pp. 302-9. 108 Allocation Tax are deposited into the Special Allocation Fund, which allows for explicit ministerial discretion in transfers. 4 2 The point here is that the wartime system's centralized fiscal control has in large measure been reintroduced. Shoup was not revised willy-nilly, as one would expect from short-term political calculations, but rather with a fairly clear aim of returning to what had been in place prior to the Occupation. Controlled local tax rates and direct management of local debt have been reestablished, in place of institutions with mandatory direct local input into decisions concerning these matters. The strong resemblance between the fiscal chaos of the prewar and the early postwar periods apparently made the 1940 fiscal system seem, for the actors most familiar with Japan's fiscal order, the only realistic choice. MOHA's Tax Controls MOHA's controls over local taxes are comparatively strict, a fact that is clear to most observers. Steven Reed, in his probing effort to bring a more comparative perspective to the politics of Japan's intergovernmental relations, is forced to concede that "[w]hile regulation of local taxes is common in unitary 4 2 Kurt Steiner, Local Government in Japan, ibid. pp. 287-8, points out that this measure dates from the Shoup reform package of 1950, when a 10 percent special grant was created at the behest of Japanese officials, who "argued with great tenacity" for this division of special and general grants. The concess ion was finally made by S C A P , but only for the years from 1950 to 1952. However, the special grant was afterwards made a permanent measure. Purnendra Jain's interviews at MOHA indicate that local 109 states, the Japanese system seems particularly restrictive."43 Indeed, MOHA's formal powers over local taxation were a focus of criticism in the Second Ad Hoc Administrative Reform Commission's call for greater fiscal flexibility at the local level. But the powers remain in place, meaning local taxes are set by recommendation of the MOHA Local Tax Bureau. The recommendations are then ratified in the Diet, and there is no record of them having been changed. 4 4 While this hardly means that MOHA steadfastly ignores opinion from politicians, other central agencies, and the local authorities themselves, it does indicate that MOHA is in a position to aggregate such information and then define the point of compromise among potentially conflicting aims. MOHA's fiscal controls are of several types, depending on the particular tax. One type is the Standard Rate (Hyoujun Zeiritsu), which applies to personal and corporate income taxes and property taxes at both the prefectural and municipal levels. MOHA sets a standard rate for these taxes. When a local authority makes a convincing case for the need to exceed this rate temporarily, MOHA can recognize the appeal, allow the extra tax burden, and use the standard rate in calculating the local authority's share of the Allocation Tax. The latter measure both prevents an inadvertent tax penalty being visited on the local authority and obviously serves as a means of encouraging general competition for revenues from the fund grants MOHA a gatekeeper role. See his Local Politics and Policymaking in Japan. New Delhi: Commonwealth Publishers, 1989, pp. 156-7. mJapanese Prefectures and Policymaking, op. cit, p. 28. This information is according to a communication from the MOHA Minister's office, in November of 1995. 110 compliance with the standard rates. Another control is the Restricted Rate (Seigen Zeiritsu), which applies to most of the aforementioned taxes and represents an upper limit that may not be exceeded. For example, the standard rates for the prefectural and municipal taxes on corporate income {Houjin Juuminzei) are 5 percent and 12.3 percent respectively, and the restricted rates are 6 percent and 14.7 percent. Of less importance, in terms of the number of taxes affected, is the Fixed Tax Rate (Ittei Zeiritsu). As the name implies, it is a single rate that has to be adhered to. This rate applies to the prefectural and municipal taxes on tobacco, for example. 4 5 MOHA controls debt via an ostensibly temporary addendum to the Local Autonomy Law, Article 250, which states that the permission of the MOHA minister or the prefectural governor is required "before floating a loan or changing the manner of floatation, interest rate, or the manner of liquidation of the existing loan (sic)."4 6 The control over local debt is an important constraint on the local authorities, especially when troughs in the business cycle cause their revenues to decline. Current rates for local taxes are presented in tabular form along with detailed information on local f inance law, on pp. 37-49 in "Chihouzei Kankei Shiryou Handobukku, Heisei 8-nen, 10-gatsu" [Handbook of Materials on Local Taxation, October 1996], op. cit. Local Autonomy Law. Tokyo: Jichi Sogo Centre, 1993, p. 82. I l l Table 3.5: Change in Tokyo Metropolitan Government Revenues, 1989-1993 (Unit: 100 Billion Yen) Item/Year 1989 1990 1991 1992 1993 Tax Revenues 477 47.2 48.5 43.8 40.6 % of total 76.4 71.7 712 62.4 56.8 Debt 185 1.84 2.99 8.66 12.1 % of total 3.0 2.8 4.4 12.3 16.9 Total Revenues 62.3 65.7 68.1 70.1 71.4 Source: Table 4 in Kanai Kiyohito, "Keizai Shakai Kankyou no Henka to Toushi keiei" [Urban Management and Changes in the Economic Environment], Toshi o Keieisuru [Urban Management]. Ed. by Jinno Naohiko. Tokyo: Toshi Shuppankai, 1995, pp. 152-53. This weakness is especially pronounced in the case of the wealthier prefectures and the metropolises such as Tokyo and Osaka, as is evident in the Table 3.5, which shows Tokyo's declining revenues and rising debt from 1989 and the onset of recession, to 1993. Japan's metropolises collect well over half of their own revenues, in contrast to most other subnational governments. But they are especially dependent on direct income-tax revenues and generally receive no Allocation Tax to substitute for lost revenues in recessions. In Japan, corporate income tax revenues readily contract ineconomic downturns, 4 7 forcing the worst affected local authorities to float bonds to cover the shortfall. By law, MOHA has to give its approval before 4 7 T h e average contraction of revenues, due to economic cyc les , among the prefectures is 30 percent. See , Igarashi Fumihiko, Ookurashou Kaitairon[Tbe Argument for Breaking up the Ministry of Finance], Tokyo: Toyo Keizai, 1995, pp. 178. 112 these bond flotations can occur. The 1940 tax reform sought to secure stability and a rough measure of equity in fiscal outcomes among the local authorities, and hence transfers from the central state doubled, as was seen in Table 3.4. But this figure was only twice an already low base of transfers in a relatively small economy, meaning that the aggregate level of expenditures was not great. By contrast, the rapid growth of the Japanese economy in the 1950s and 1960s generated an enormous volume of tax resources for collection and redistribution. Thus, compared to the prewar and wartime years, MOF and MOHA now have powerful motivations for maintaining their trenches on Japan's tax fields. The next chapter examines in more detail the constellation of interests, including MOF and MOHA, that is concerned with Japan's intergovernmental fiscal relations. Information on the above tax and debt controls can be found in Mutai Shunsuke, et al., Jichitai Zaisei Kiban no Kakuritsu, op. cit., especially on pp. 34-44. The book is produced by MOHA staff and presents an interesting defence of f iscal control from the standpoint of preserving equity among the local authorities and intergenerational tax burdens. 113 Chapter 4 Interests in Japan's Contemporary Fiscal Politics The power structure in Japan has four parts. There's the central government, the Home Affairs Ministry -- which represents the regions -- the prefectures, and the municipalities. And each part has its own philosophical bent. Thus when I hear people advocating decentralization, I'm quick to ask where they want power to devolve. Kato Hiroshi Chairman, Government Tax Advisory Council MOHA's Organizational Mission This study of the reconstitution and reproduction of the 1940 fiscal system thus centres on MOHA's "organizational mission." 1 One of the smallest agencies in Japan's central state, MOHA is a distillation of the dismantled Home Ministry's local finance functions, as the ministry at present contains a Minister's Secretariat and three bureaux: Local Administration, Local Finance, and Local Tax. 2 Other areas under the old Home Ministry's vast purview, such as public 1 John C. Campbell refers to the Japanese Ministry of Labour's "organizational mission," or "continuity of bureaucratic motives, style, and actual policy." The ministry's organizational mission is rooted in its autonomy and strong motives for policy sponsorship, even though it is not one of the powerful economic ministries, such as MITI and MOF, that are generally cited in studies of bureaucratic power in Japan . See Campbell, How Policies Change: The Japanese Government and the Aging Society. Princeton: Princeton University Press , 1992, 277-81 and fn. 53. My thanks to John Campbell for much-appreciated hints that MOHA's role might profitably be viewed as an organizational mission. 2 According to Kengo Akizuki , MOHA's personnel number 569 bureaucrats in a ministry that "is recognized as one of the most important bureaucracies." MOHA also includes the Fire Defence Agency, but that is an external bureau whose business is not central to MOHA's work. See "Institutionalizing the Local Sys tem, " op. cit.p. 342. 114 works and labour relations, were transferred to newly created central agencies when the Home Ministry per se was dissolved in 1947. Shorn of intraministerial cross-pressures from spending bureaux, MOHA found the administrative space to develop a clearly defined agenda that centres on equity and uniformity in local finance. In other words, MOHA's organizational mission is in part possible because its structure does not incorporate the contradictory interest, characteristic of the postwar era's highly politicized spending ministries,3 of offloading a large share of the burden for public works onto the local authorities. This downward shift of burdens is a major problem for MOHA. Particularly troublesome are specific subsidies {tokutei hojokin), which line ministries direct to their counterpart agencies at the local level. The subsidies often leave much of the cost of a given project to be borne by the recipient local authority, and hence ostensibly local funds are effectively used as the respective central agency desires. The structure within which such phenomena occur is referred to as "vertical administration" (tatewari gyousei), because intergovernmental relations are fragmented into discrete channels of authority between the 3 A s Kent Calder notes, the postwar break-up of the Home Ministry weakened component functions such as land-use planning. This weakness, in the new constitutional regime of electoral politics, led to a high degree of politicization as postwar politicians sought to secure their electoral bases through distributional politics. See Crisis and Compensation. Princeton: Princeton University P ress , 1988, especial ly the sect ions on "Regional Policy" and "Land Use Pol icy." Note that Calder's work (e.g., pp. 289-98) does not investigate the details of the reassertion of f iscal controls through MOHA and its early postwar predecessors, and hence he overlooks the role of bureaucrats in the redistributive sys tem that funds the "crisis and compensation" dynamic from the centre to the periphery. 115 individual ministry and its counterpart bureaux in the local administrations. 4 Vertical administration, like competition with MOF overtax fields, is one aspect of MOHA's often contentious relationship with other ministries in the central state.5 Both of these phenomena condition MOHA's role as an intermediary between agencies of the central government and the local authorities, whose interests MOHA represents. MOHA thus opposes MOF and other central agencies' interventions in local administrative and fiscal affairs. To accomplish its objective, indeed duty, of protecting the fiscal health of the local authorities, MOHA requires considerable authority. The reassertion of the wartime system secured this authority to a fair degree. Therefore, seen through the lens of MOHA's organizational mission the maintenance of the 1940 system's fiscal controls is not for MOHA's power, per se, but rather to restrict one potential source of renewed fiscal chaos. 6 In many developed countries, subnational governments are directly 4 See Richard Samuels, The Politics of Regional Policy in Japan: Localities Incorporated. Princeton: Princeton University Press , 1983, pp. 42-4; and Kurt Steiner, Local Government in Japan, op. cit., pp. 232-62. 5 On this point, see Tahara Soichiro, "Jichishou" [The Ministry of Home Affairs], Nihon no Kanryou 1980 [Japanese Bureaucrats, 1980]. Tokyo: Bungeishunju, 1979, pp. 48-9. 6 Expressions of the need to maintain equity among the local authorities are a constant thread in MOHA statements and publications. See , for example, Mutai Shunsuke et al., JichitaiZaiseiKiso no Kakuritsu, [Enhancing the Fiscal Base of the Local Authorities]. Tokyo: Gyousei , 1995., pp. 2-6. 116 represented at the national level. 7 Japanese local authorities do not enjoy the same institutionalized access, and require a representative in the central state. This is because the scale of Japan's intergovernmental fiscal transfers and their focus on public works spending create enormous incentives for intervention in subnational affairs. Moreover, the country's political institutions are highly fragmented. Not only does vertical administration produce strong divisions at the level of intergovernmental relations, but Japan's central political leadership is comparatively weak. This is seen especially in the passivity of the Prime Minister's Office and the role of the electoral system in encouraging relative indiscipline within the ruling party.8 These features make it difficult to achieve the level of supra-ministerial coordination and party discipline that is elsewhere characteristic of parliamentary systems of government. 9 Japan's fragmented political institutions have given Diet politicians comparatively much room to engage in distributive politics. This is manifested 7 A d iscuss ion of various countries' institutions can be found in Percy Allen, "Subcentral Government and Centre-Local Relations," State and Society in Western Europe. Cambridge: Polity, 1995. Other accounts are J . J . Hesse, "The Federal Republic of Germany: From Cooperative Federalism to Joint Pol icymaking," and Y. Meny, "France: the Construction and Reconstruction of the Centre 1945-1986"; both of these articles are in West European Politics, Vol. 10, No. 4, October 1997. 8 A good account is in Robert C. Angel, "Japan's Premiership and Cabinet: A Study in Democratic Institutions and Processes , " a paper prepared for the Washington and Southeast Region Japan Seminar, February 25, 1995. Note also that the 1994 introduction of single-member and proportional representation districts appears to be having "the perverse effect of exacerbating intra-party rather than inter-party policy differences"; see Gerald Curtis, "Reform requires deregulation of political activit ies," Asahi Evening News, May 26, 1996, p. 3. 9 Leonard Schoppa's work on education policymaking in Japan suggests this fragmentation and lack of coordinating mechanisms quite strongly; see "Zoku Power and LDP Power: A Case Study of the Zoku Role in Education Pol icy," Journal of Japanese Studies 17, 1991, pp. 79-106. 117 especially through so-called "policy tribes" (zoku giin) of legislators, which wield considerable influence in many distributive and symbolic areas of policymaking. 1 0 Not only are these politicians in search of votes for reelection, but when their politicking focuses on public works, they have the added incentive of bribes and kickbacks that can be used to finance their campaigns (and presumably other, more frivolous purposes as well). 1 1 The various zoku and their often allied spending ministries tend to come into conflict with MOHA's efforts to maintain order in local fiscal affairs. Hence MOHA cannot entirely manage the fiscal environment of the local authorities, notably because of countervailing power wielded by shifting coalitions of politicians, bureaucratics, and private interests. Indeed, in many instances of local policymaking, MOHA's authority is limited by the developmentalist agenda that still heavily defines the Japanese state. 1 2 1 0 O n zoku giin and their relationship to policymaking, see Inoguchi Takashi , "Shin, Zokugiin Taibouron" [Aims for the New Policy Tribes], in ChuoKoron, March 1994, pp. 104-12. Also, the role of t hezoku giin and the P A R C in tax policymaking is d iscussed in Takashi Inoguchi and Tomoaki Iwai, "Zeisei Zoku no Seij iryokugaku" [Study of the Political Power of the Tax Policy Tribes], ChuoKoron, March 1987, pp. 96-106. 1 1 A detailed d iscussion of such "structural corruption" can be found in Gavan McCormack's chapter on "The Construction State," The Emptiness of Japanese Affluence, op. cit; see especially, pp. 32-44. 1 2 For example, a recent study of urban redevelopment in the Japanese city of Omuta shows that MITI and the Construction Ministry were both decis ive actors in this case of local policymaking, along with the Japan Development Bank. MOHA's role certainly restricted the scope within which redevelopment could be undertaken, indeed envisioned, but the ministry in turn made concess ions on financing through debt allocation and tax changes. These comments are based on communications during December 1995 with Ted Gilman concerning his Ph.D. dissertation, "No Miracles Here: Urban Redevelopment Efforts in Japan and the United States," University of Michigan, 1995. For an interesting account of developmentalism in Japan, see Gavan McCormack's chapter on "The Construction State,"77je Emptiness of Japanese Affluence, op. cit, 25-77. 118 MOHA is therefore on the one hand a hegemon at the local level, willing to pursue its own interests; but on the other hand, MOHA is a crucial buffer between the local authorities and the politicized designs of the spending ministries as well as MOF's fiscal austerity concerning local finance. Without MOHA, the local authorities' capacity and resources to place their own stamp on policymaking would almost certainly decline. The weak representation of local governments in the central state in fact led them to support the administrative changes that saw MOHA become a full-fledged ministry in I960. 1 3 Yet the bulk of the intergovernmental fiscal regime has been outside the rather narrow sphere of interest and influence of zoku politicians. This ostensibly depoliticized fiscal terrain is the site of MOHA's struggles against MOF's persistent efforts to appropriate sections of the local tax base as well as offload fiscal burdens through cuts in transfers from the budget of the central state. The two ministries see local finance from virtually opposite points of view, and thus devise their policies accordingly. 1 4 The fiscal conflicts between the ministries will be examined in much greater detail in the ensuing chapters, so 1 3 See Takatsuji Masami, Chihoujichi [Local Autonomy], Tokyo: Gendai Gyouse i , 1983, pp. 70-1. In describing MOHA as a "hegemon," then, I am emphasizing the aspect of consent at least as much as that of coercion. MOHA at times is forced to resort to the latter, as in the c a s e of Tokyo Governor Minobe; but the ministry also has to respond to local needs and demands in order to remain an effective representative. An unresponsive MOHA would likely see local authorities turn more determinedly to other means, such as Diet politicians, to lobby the central state. 1 4 A sampling of opinion among bureaucrats on the issue of local revenues showed 100 percent of the MOHA respondents argued the revenues were insufficient, whereas 37 and 46 percent, respectively, of M O F bureaucrats thought the revenue levels were either too great or satisfactory. See Table 6.3, 119 at this point I will only note that MOHA's main hurdles vis-a-vis MOF is that it does not have independent authority over much of the general government tax base, especially income taxes, nor can it decide on its own the mix that will fund the Local Allocation Tax. In addition, what shapes Japan's intergovernmental distribution of taxes is power, taxpayer tolerance and convention, rather than constitutional rules. 1 5 MOF and MOHA's shared jurisdiction on major tax fields has thus engendered intermittent turf wars and collaborative truces between them, with both the conflicts and the truces generally impeding efforts toward enhanced local fiscal autonomy. Recent years have seen conflict between the two ministries escalate. The reasons for the upsurge are several, and will be dealt with in detail in the individual case studies. However, noteworthy is the reemergence of fiscal stress due to slowing economic growth and globalization of the economy coupled with increasing demands on the state. Another related source is the long-range programme to "rebalance" the tax system in order to promote economic vitality and perceived fairness in distributing the tax burden. This is evident in Table 4.1, which shows the overall changes in the revenue structure since 1955. "Chihou no J ichizaigen" [Local Revenue Autonomy] in Michio Muramatsu, Chihou Jichi [Local Autonomy]. Tokyo: Tokyo Daigaku Shuppankai, 1988, p. 199. 1 5 Note that some federal states have more or less exolicit rules for allocating intergovernmental tax bases . The interpretation of the rules often varies overt ime, as it certainly has in Canada, but there is at least a legal basis for one level of government to file suit in protecting its tax base. 120 As the table shows, the weight of taxation incident on the income field peaked in the mid-1980s, and has thereafter declined relative to taxes on consumption and assets (e.g., property taxes). The government policy to shift the burden of taxes over time away from income derives from a belief that the tax system overall is too dependent on the income base. Table 4.1 hows there has been some success with this strategy, though of course the 1990s recession had a dampening influence on income-tax revenues as wel l . 1 6 Table 4.1: Changes in the Japanese Tax Mix, 1955-1992 (Units: Percent of revenues) Tax Field 1955 1965 1975 1986 1992 Income 48.4 53.1 54.9 59.1 54.3 Consumption 39.0 33.6 24.4 19.0 20.9 Assets 12.6 13.4 20.7 21.9 24.8 Source: Chihouzaisei no Hatten to Aratana Tenkai Finance]. Ed. by YuasaTosh io . Tokyo: Gyousei , 199 The Development of and Future Directions in Local 5, pp. 317. There is, of course, a good deal of politics behind the choice to rebalance the tax fields. The shift away from income taxation is strongly linked to a problem explored in one of the case studies below: perceptions of unfairness in the personal income tax have convinced ranking fiscal bureaucrats and ruling party politicians that permanent cuts are necessary. But the point to be made here concerns the effect that the changing tax base has on intergovernmental 1 6 One ranking M O F interviewee in fact suggested that the ministry's long-term aim is to raise for the Consumption Tax to near-European levels while reducing dependence on income taxes. 121 tax politics. For as the balance of taxation changes, MOF and MOHA are encouraged to compete to expand their tax room in new fields as well as protect their absolute shares in areas slated for cuts or gradual attrition. Cooperation is also sometimes evident. A recent example of cooperation between the two ministries, in a reform that cost a degree of local fiscal autonomy, is seen in MOHA's acquiescence to the elimination of the Electricity and Gas taxes (denkizei and gasuzei) when the Consumption Tax (shouhizei) was adopted. Competition is, by contrast, visible in the Land Value Tax (chikazei) episode, the struggle over the Local Consumption Tax (chihoushouhizei), and the current sparring concerning corporate income taxes and cuts in the Local Allocation Tax {chiho koufuzei). Politicians and Tax Politics Returning briefly to the wartime fiscal centralization and the roots of MOHA's organizational mission, we have seen that the former is not unique. Nor is the simultaneous construction of a system of intergovernmental redistribution. But other redistributive fiscal systems, such as those established in Australia and Canada , either included or soon came to include explicit mechanisms for intergovernmental political and bureaucratic negotiation. These mechanisms 122 were, in other words, aimed at alleviating centralized fiscal controls. 1 7 By contrast, Japanese central-state bureaucrats devised their reforms in reaction to the legacy of American military occupation, which had sought to tear apart the intergovernmental fiscal system that the bureaucrats had long laboured to build. Once the Occupation had ended - indeed, to some extent even before -- Japan's MOF and the predecessors of MOHA worked to reconstruct the 1940 system, thereby giving themselves great latitude to impose fiscal discipline and maintain a virtually uniform regime of local finances. In other words, bureaucratic preferences generally set the agenda for the postwar reconstitution of the 1940 intergovernmental fiscal system. Political scientists will naturally ask where the Diet politicians were during this concerted bureaucratic agenda-setting on intergovernmental tax relations. The answer to this question fits into the argument made earlier; that is, the national-level politicians' attentions were devoted mainly to the distributive end of intergovernmental fiscal relations. Driven by the electorally based need to show results to supporters, 1 8 such politicians for the most part sought 1 7 In Austral ia, for instance, the Commonwealth Grants Commission and its f iscal equalization formula are sites of contest and negotiation among the federal and state governments; see Christine Fletcher, "Competition Between Regional Governments and the Federal Culture of Fiscal Equalization, Australian Journal of Political Science, Vol. 28, 1993, pp. 58-9. 1 8 Note that postwar Japan's Single Non Transferable Vote electoral sys tem tended to place a premium on politicians' being seen to "deliver the goods" to the local consti tuency. Securing a shift of f iscal resources to a lower level of government would thus not appear to be in the interests of a Diet politician, unless voters were organized to reward such policies. On the importance of local sev ice for Dietmembers, see Samuel Kernel, "The Primacy of Politics in Economic Pol icy," Parallel Politics: 123 influence at the expenditure rather that taxation end of the fiscal process. Japanese politicians' role in tax politics per se centres on highly visible and politicized forms of taxation, especially in the realm of intergovernmental fiscal relations. This was also the case in the decades between the two world wars, when attempts by the bureaucracy to rationalize intergovernmental fiscal relations and the overall structure of taxation foundered on the political power of urban capitalist interests. Earlier, through the late 1800s era of state-building, politicians' attentions had focused on the tand tax (chiso) and whether it should be raised to fund armaments production and other strategies aimed at bolstering Japan's military-industrial base. The political parties' positions on such industrial development differed only marginally, and they gradually abandoned their pressure for reductions in the land tax; then, at the turn of the century, they ceased their opposition to increases in the tax. Party tax platforms in this period were largely patterned on the fiscal structure and the incidence of the tax burden. For even as the land tax burden increased in absolute terms, it declined as a percentage of total revenues and thus elicited a commensurately shrinking degree of interest. The land tax supplied over forty million yen, or 60.6 percent of total national tax revenues of just over sixty-eight million yen in 1890. By 1920 the land tax yielded seventy-four million yen, but that was only 7.9 percent of total revenues of 938 million Economic Policymaking in Japan and the United States. Ed. by Samuel Kernell. Washington, D.C.: The 124 yen. Most of the burden of revenues was shifted instead to indirect taxes, which amounted to 543 million yen in 1920, of which the liquor tax (Shuzei) provided the largest share at 164 million yen, or 17.5 percent of total national revenues. The period's burning tax issue had as a result become one of lifting the tax burden somewhat off the lower classes, especially rural inhabitants, and transferring it to wealthy urban residents and capitalists. But the parties were by then effective spokesmen for the modern sector's opposition to bureaucratically led efforts to make the obsolete tax system more productive and fair. 1 9 In the postwar era, the focus of party politics turned in particular to the personal income tax. Distress concerning Japan's great wave of anti-tax sentiment in the early postwar period was transferred from the front-line offices of the tax-collection branch to MOF's career bureaucrats in Tokyo. These officials, in turn, inculcated in the conservative parties of the 1950s a strong appreciation for the political utility of personal income tax cuts. This helped institutionalize anti-income tax sentiments as a strong influence on Japanese party politics and tax policymaking. 2 0 Brookings Institution, 1991, pp. 342-4. 1 9 Much of the above information comes from Banno Junji's excellent interweaving of f iscal issues and party-bureaucratic politics in the Meiji era, in The Establishment of the Japanese Constitutional System. London: Routledge, 1991, especially pp. 170-202. Land-tax data can be found in table 7.1 of this dissertation, and the liquor tax information is found in Table 12-8-a, "National Taxes by Principal Items (1st term - 1926)," in Historical Statistics of Japan, Volume 3. Tokyo: Japan Statist ical Associat ion, 1987, p. 9. 268-69. 2 0 On this period, see Ohtake Hideo, "Hatoyama, Kishi Jidai no Chisai Seif uron" [The Small-State Debate of the Hatoyama and Kishi Era], in Seijigaku Nenpo 1991. Note that other factors provide the background to anti-income tax politics in Japan. One of the major items of comparative signif icance is the fact that 125 Tax politics in recent years continue to exhibit this focus on more visible taxes, as reductions in income tax are always attractive options for party political manouevring. The long years of the Heisei Recession in the early 1990s saw the former Socialist Party (now the Social Democratic Party of Japan) press for income-tax cuts, which became a dominant theme in the public debate on tax policymaking. Competition over tax cuts indeed became so intense that the LDP hierarchy felt threatened by the prospect of rank- and-file defections. 2 1 Also, the October 20, 1996 general election centred very much on the tax platform of the New Frontier Party, led by Japan's foremost Machiavell ian Ozawa Ichirou. The party sought to exploit public opposition to the ruling coalition's commitment to raise the Consumption Tax rate to 5 percent, and so offered a "freeze" (touketsu) of the 3 percent rate until 2001 plus a continuation of income-tax cuts aimed at stimulating the sluggish economy. Though Ozawa failed to win the election with this basket of populist tax measures, that would appear due more to voter distrust of him and his Soka Gakkai religious supporters in general. Certainly polling date show the tax cuts themselves there was almost no support from labour and the Left for income taxation, in sharp contrast to the case in North America. In Japan, the Socialist Party was resolutely opposed, for example, to the Shoup plans to centre the tax sys tem on the personal income tax and the ideology of "ability to pay," as Japanese social ists saw income taxation -- in Marxian terms -- as a tax on labour rather than surplus value. By contrast, the US and Canada have f iscal histories wherein the role of labour and progressive political forces were crucial for the institution of the income tax and the maintenance of its political legitimacy until recent years. These issues are discussed in greater detail in Chapter 6. 2 1 See , for example, "Shushou: shintai, taisei ni shitagau, 'shotokuzeigenzei no j isshi hitsuyou'" [The P M says to followthe general trend, "implementing an income tax cut is necessary"] , Nihon Keizai Shinbun, July 11, 1993, p. 1. 126 garnered strong voter support. 2 Intergovernmental Tax Politics Intergovernmental tax politics are a less attractive issue area, however. Diet politicians tend to exert little deliberate influence at this level, compared to politicians in many other centralized and federal states, 2 3 save when personal income tax cuts they have engineered restrict local finances. This latter effect is indirect -- not the result of a policy concerning local resources themselves --and stems from the fact that local income taxes are calculated in terms of their national-level counterparts. The political dimensions of the intergovernmental tax structure per se are outside the calculus of the majority of Diet politicians, who are far more interested in the specific subsidies of the line ministries. 2 4 There are, as we will see, occasions when national politicians express interest in intergovernmental fiscal relations, but these are generally responses to the agenda set by the competition or cooperation of the two strongest ministries, MOF and MOHA, in this policy area. Polls in the October 1 Asahi and the October 5 Yomiuri newspapers in 1996 both showed that prospective voters ranked the tax issues highest or among the highest priorities when deciding how to vote. Examples d iscussed earlier showed, for instance, that British, Dutch, French, and Thai politicians are more involved in intergovernmental fiscal reforms than are their Japanese counterparts. Kengo Akizuki makes the very interesting point that MOHA's formula for redistributing the proceeds of the Allocation Tax have removed those funds from the sphere of explicit political contention, and that the alternative of specif ic subsidies helps perpetuate this condition; see "Institutionalizing the Local Sys tem, " op. cit, 359-60. 127 To be sure, MOHA has a sizable cadre of supporters in the Diet whose votes can be mobilized on issues related to intergovernmental f inance. 2 5 Yet this is according to the ministry's agenda rather than the politicians' independent activism. There is at least one significant case where a politician, as policy entrepreneur, took a strong role in intergovernmental fiscal politics. But the politician, former Prime Minister Kakuei Tanaka, predictably used the rhetoric of decentralization to advance a clearly self-interested agenda. The incident occurred in the early 1950s, when Tanaka was a rising Diet politician. Tanaka became involved in the issue of how to manage the revenues from the Gasoline Tax (Kihatsuyuuzei), which had been implemented in 1937 and was producing a increasingly heavy yield as the economy recovered from its wartime damage and early postwar turmoil. For example, the Gasol ine Tax drew 20.5 billion yen, or just under 5 percent of total national indirect tax revenues of 435 billion yen in 1953. By 1960, this percentage share of total national indirect taxation had increased to 12.5 percent; and by 1970, it had climbed to 18.9 percent. 2 6 Tanaka could not have known the scale of road-building that was to occur in Japan as the 1950s wore on; but, being from Niigata with its long and 2 5 Igarashi Fumihiko, Ookurashou Kaitairon, op. cit, pp. 187-88. 128 snowbound winters, he was certainly aware of the importance of roads to economic activity. Hence Tanaka stressed the importance of road-building to the recovery and future growth of the Japanese economy. However, for him and his Diet supporters, a major problem was the centralization of decisionmaking on road-building and its finances within, respectively, the ministries of Transport and Finance. This administrative structure was a legacy of the prewar order, wherein decisions on road construction and maintenance were removed from the pork-barrel concerns of party politics and based instead on tranport's contribution to Japan's military power. The need to decentralize highway planning and contruction decisions, by moving them into the Construction Ministry, quickly became a pet issue for Tanaka. In fact he repeatedly invoked the idea that the postwar era was a break from the past and the associated dominance of the excessively centralized and bureaucratic state. Moreover, he insisted that funding for road-building be earmarked in order to get around potential obstacles posed by MOF's control of the budget. Though MOF strongly opposed the loss of its discretion through the institution of an earmarked tax, Tanaka's coalition of private-sector and political allies carried the day. The result saw a quarter of the revenues from the Gasol ine 2 6 These figures are calculated from information in Table 12-8-b, "National Taxes by Principal Items (1927-1984)," in Historical Statistics of Japan, Volume 3. Tokyo: Japan Statistical Associat ion, 1987, p. 270-71 . 129 Tax deposited directly into a special account for the highly politicized Construction Ministry. 2 7 In short, this postwar example of politically driven "decentralization" was more a deflection of revenues to one of the most politicized arms of the central state rather than an actual devolution of fiscal resources to the local authorities. The above outcome indicates that Diet politicians, notably the more powerful ones, certainly can overcome bureaucratic opposition to their tax initiatives. But it also suggests there is a relatively limited sphere in which such politicians are motivated to act. Local Politicians and Fiscal Decentralization On the other hand, local government politicians have an obvious and direct interest in intergovernmental fiscal relations. This is because the local authorities suffer most from the vagaries of national policymaking, the uncertainty that stems from a high dependence on direct-tax revenues, and especially MOF's frequent efforts to restrict local budgets. Local politicians are organized into the Six National Associations of Local 2 7 On Tanaka's campaign and the rhetoric he employed, see Hayasaka Shigezou, "Douro no Seibi" [Road Maintenance], Chapter 4 in Seijika Tanaka Kakuei [The Politician Kakuei Tanaka]. Tokyo: Tokyo Kouronsha, 1987, pp. 54-90, and Nishioka Toshirou, '"Douro San H o u ' no Seitei Katei" [The Process of Setting Up the Three Road Laws],//? Tokyo Toritsu Daigaku HougakukaiZasshi, Volume 30, Number 1, 1989, pp. 355-405. 130 Governments (Chihou Roku Dantai, hereafter CRD) , 2 8 which were seen by the U.S. Occupation as a means for encouraging local autonomy. The former Home Ministry bureaucrats at the time proved themselves to be major impediments to the Occupation's decentralist goals. In order to bypass the Home Ministry, the heads of the CRD were, as Richard Samuels puts it, used "quite aggressively." Samuels also reveals that the C R D were given standing on: important national administration committees such as the Local Systems Investigation Council and the Town and Village Amalgamation Promotion Committee. In addition, each of the six C R D chairmen sat as members of the Local Administration Committee (Chihou Gyousei linkai), which was created in the wake of the dismantling of the Home Ministry. The C R D emerged in the immediate postwar period from its semi-governmental torpor at S C A P ' s direction, and its leaders became some of the most influential executives of MacArthur's local policy. All six were formally reinaugurated after the war as part of the S C A P effort to nurture interest groups and to introduce local autonomy to Japan . 2 9 But Samuels goes on to note that the CRD have been the targets of extensive criticism for their complacency in the face of the post-Occupation recentralization of state functions. Japanese and foreign observers, he writes, have often seen the C R D ' s as a transmission belt for central directives and the These six organizations comprise the National Associat ion of Prefectural Assembly Chairmen (Zenkoku Todoufuken Gikai Gichoukaf), the National Associat ion of City Assembly Chiefs (Zenkoku Shigikai Gichoukai), the National Associat ion of Town and Village Assembly Chairmen (Zenkoku Chousongikai Gichoukaf), the National Associat ion of Governors (Zenkoku Chijikai), the National Associat ion of Mayors (Zenkoku Shichoukai), the National Associat ion of Towns and Villages (Zenkoku Chouson Choukai); see Jichi Yougoshuu [Japanese-Engl ish Glossary on Local Government]. Tokyo: J ich i Sougou Sentaa, 1994, p. 130. The Politics of Regional Policy in Japan: Localities Incorporated?, op. cit. pp. 23-4. 131 implementation of national policymaking. Samuels seeks to leaven this characterization somewhat by depicting the CRD as an important collective actor in lobbying for funds from the centre's budgetary process. Most prominent for our purposes here is the fact, which Samuels concedes, that there is often considerable division among the constituent organizations of the C R D on tax policies. In particular, the views of the associations representing municipalities and smaller administrative units often diverge from those of the representatives of prefectures concerning the distribution of tax revenues and the advisability of fiscal decentralization. 3 0 My case studies show there are not always large splits over fiscal issues, but there is a strong tendency toward this and it has clear structural roots. It is, in fact, a potential that MOF has in the past tried to exploit in its contests with MOHA. One of the most limiting features for the C R D ' s effectiveness as an agent for fiscal decentralization is this division of interest. The division arises because large and wealthy prefectures and metropolises, with Tokyo on top, receive no Allocation Tax whereas the rest of the local authorities generally do. Tokyo collects upwards of 70 percent of its own revenues, and effectively pays large amounts, especially through its residents' personal and corporate income tax 3 0 Recent commentary on this phenomenon recognizes the dependence of the poorer local authorities on the Allocation Tax and their consequent unwillingness to work with larger governments for f iscal autonomy. Local Allocation Tax revenues create, for governments with restricted f iscal capacity, a disincentive to increasing their own tax effort. The position of such weaker local authorities in supporting the tax makes, of course, perfect f iscal sense; but it also helps MOHA legitimate its role in 132 payments, into the coffers of fiscally weaker local authorities. Of the latter, those who cannot manage to raise more than 30 percent of their total revenues from the local tax base include 66 percent of prefectures, 31 percent of municipalities, and 83.6 percent of towns and villages. MOHA itself often calls attention to these fiscal divisions in its efforts to set the agenda of policymaking on the issue of decentralizing control over tax revenues. 3 1 Moreover, as is evident in my case studies, the C R D generally rely on MOHA to plead their cause so far as tax policymaking is concerned -- in the main fora at the centre of the state. In particular, the local interest is represented by MOHA in the Government Tax Advisory Counci l , or GTAC {Seifu Zeisei Chousakai), wherein virtually all policy proposals relating to the intergovernmental tax system are argued before representatives of major interest groups and among expert members of specialized subcommittees. 3 2 The GTAC is one of a large number of institutions known generically as deliberation councils (shingikai). Many are fairly inconsequential in policymaking, but some -- such as the GTAC -- are important sites where the maintenance of tax controls. See "'Amae' sasaeru koufuzei seido" [The Local Allocation Tax structure supports dependence], in Nihon Keizai Shinbun, March 4, 1996, p. 30. 3 1 A typical example of this approach is seen in the fact that the data cited in the above come from a MOHA publication that argues the need for f iscal decentralization while also assert ing the need for uniformity. The only way to reconcile those two contradictory ends is to define decentralization as the control of tax rates and redistribution through MOHA. See, Mutai Shunsuke et a/., JichitaiZaisei Kiso no Kakuritsu, op. cit., pp. 2-6. 3 2 The Local Sys tem Advisory Commission {Chihou Seido Chousakai) was an important site of policymaking influence in the period when Shoup was being revised; but afterwards it dwindled into relative insignificance in tax policymaking, to the point where one rarely sees mention of it in the p ress . 133 representatives of major interests groups are brought together to debate appropriate turns in public policy. Though MOHA bureaucrats have some jurisdiction in the selection of members and related functions, MOF personnel play the central role. It is MOF Tax Bureau staff who select the bulk of the representatives for the GTAC, determine the information they will discuss, and even write up their final report. In spite of this obvious scope for manipulation, the product of the discussions is not regarded as a sham. A more subtle process is at work, one in which bureaucrats learn about positions and perceptions among major interest groups in their areas of authority, while the represented interests are brought into policymaking and thus encouraged to compromise with one another. Such features give a fairly high degree of credibility to the institution, as an adjunct to more effective policy design and implementation. MOF is thus usually able to manipulate the debate along desired lines without grossly eroding the GTAC report's legitimacy. A strong report can give a considerable degree of momentum to a policy initiative.3 3 It has of course become a commonplace to note that the GTAC's influence relative to that of the LDP 's Tax Advisory Counci l , or LDPTAC {Jimintou Zeisei 'General information on deliberation councils is plentiful. See , for example, Ehud Harari's "Mass Media Participation in Japanese Public Policy Processes : The Case of Shingikai" (Paper presented to the ISS Contemporary Japan Group, University of Tokyo, March 1995) and Frank Schwartz, "Of Fairy Cloaks and Familiar Talks: The Politics of Consultation," in Political Dynamics in Contemporary Japan. Ed. by Gary All inson and Yasunori Sone. Ithaca: Cornell University Press , 1983. On the importance of the GTAC, one that "stands like a mountain among the rest," see Norio Kakizaki , "Shingikai: Kanryou e no Houshin no Kisek i , " [Deliberation Councils: Places of Service to the Bureaucracy] in Ekonomisuto, July 31, 1979. 134 Chousakai), has declined since about the late 1960s. 3 4 But the GTAC remains the pre-eminent site for agenda-setting on intergovernmental tax issues, if not most tax policymaking. Other Interests Other major interests involved in intergovernmental tax politics include big business lobbies such as Keidanren, though their salience and the character of their participation vary depending on the specific issue. For example, business organizations were opposed to the Gas and Electricity taxes that were eliminated in the rationalization of the indirect tax regime as part of the 1989 institution of the Consumption Tax (Shouhizei). Business groups were also against the prefectural-level collection of the Local Consumption Tax, sought by MOHA from the fall of 1993. Hence, on tax matters, the corporate sector is generally a fiscally conservative force, focused on seeking tax reductions that apply to business as a class (e.g., cuts in corporate income taxes) or to sectoral interests. This does not mean business opposes fiscal decentralization. Indeed, when the cream of corporate Japan assessed intergovernmental fiscal relations during the famous Second Ad Hoc Commission on Administrative Reform in the early 1980s, they called for a relaxation of MOHA's controls over local tax 3 4 The representative study of this phenomenon is Kishiro Yasuyuki , Jimintou Zeiseichousakai [The 135 rates and the institution of new local taxes. The purpose was to introduce a stronger form of self-government among local residents, with an apparent public choice assumption being that such would result in fiscal conservatism. 3 5 The corporate consensus would appear to be that decentralization can be a good thing, so long as it means smaller government or at least no increase in tax burdens. In other words, when it comes to taxes the business community acts largely according to the general fiscal principles sketched in the introduction to this dissertation: it does its best to reduce its tax burden. On the other hand, there is a pronounced split when it comes to the expenditure end of fiscal politics. Businesses that do not rely heavily on public spending, especially Japan's internationally competitive firms, encourage reform of the public sector. But the recipients of largesse are not at all keen on spending cuts, at least in the programme areas from which they benefit. This is true, for example, of construction firms, which employ 10.6 percent of the Japanese workforce --about twice the average of other industrialized states 3 6 -- and greatly depend on Japan's high level of public works spending, as was displayed in table 1.1. Liberal Democratic Party Tax Advisory Council]. Tokyo: Toyo Keizai Shinposha, 1985, pp. 43-52. The report is cited and described briefly in "Sentaku to Futan: Kazei J ishuuken no Kakudairon" [Choice and Burden: The Argument Concerning Expanding Autonomous Tax Authority],in Mutai Shunsuke, ef al., Jichitai ZaiseiKiso no Kakuritsu, op. cit, pp. 41 -4. O E C D figures show 10.6 percent of Japan's working population in construction in 1994, whereas the percentage in Canada was 4.8, the US 4.3 (1993), Australia 5.4, France 6.4, Germany 7.0, Italy 6.8, Sweden 4.7, and the UK 3.8. The data are from OECD: National Accounts, Detailed Tables, Volume 2, 1982-1994. Paris: 1996. 136 Another interest whose role varies according to the particular tax and its calculation of political advantage is the Local Employees' Union, Jichirou. The union is the most powerful interest group in the Socialist Party and a resolute opponent of MOHA as an institution. However, as a union it is generally keen on seeking to maintain or expand its membership -- through local tax collection, for example. Hence, at times Jichirou assesses its interests as coincident with those of MOHA. As was described at length in the previous chapter, this has frequently been the case under the coalition governments of the 1990s. Other Central Agencies The tax aspect of intergovernmental relations is of little interest to most other central-state ministries, because they have no jurisdiction in the area. Of course, the ministries of Agriculture, Forests and Fisheries, International Trade and Industry, Transport, and Health often lobby for client groups' interests with respect to specific taxes. However, there is little evidence that such ministries are concerned with which level of the state levies a particular tax. 3 7 By contrast, the ministries are very keen on expanding the role of specific An example of the lobbying is seen in the various ministries' "wish list" for tax reductions for the 1997 f iscal year; see "Futan keigen youbo, aitsugu" [Demands for lower burdens come one after another], in Nihon Keizai Shinbun, August 31, 1996, p. 5. 137 subsidies, often working in tandem with LDP and politicians and such client interests as construction firms. From MOHA's point of view, then, other ministries can, by transferring fiscal burdens onto local governments, bedevil efforts to maintain local fiscal stability. This often antagonistic role, to MOHA's , of the spending ministries is particularly evident in the delegated functions (inin jimu) that are a commonplace in Japan's centre-local administrative relations. According to a recent account, "[t]he number of agency-delegated functions has increased significantly. At present, more than half of the functions performed by prefectural governments are agency-delegated." 3 8 Delegated functions stem in part from the local government codes of Japan's Meiji era (1868-1912), when duties that the local authorities were expected to carry out were assigned by the central government. Under S C A P 3 9 the codes were revised in order to break from the old regime's dominance of local governments. However, the Local Autonomy Law instituted in the early postwar reforms was itself considerably amended during the Occupation's "reverse course" and afterwards, thus expanding the scope for intervention in local affairs. Pressure for amendments to the law arose, not surprisingly, in 3 8 See Jinshichiro Yonehara, "Relations Between the National and Local Governments," Japan's Public Sector. How the Government is Financed. Ed . by Tokue Shibata. Tokyo: University of Tokyo Press , 1993, p. 170. 3 9 This is the acronym for Supreme Commander Allied Pacif ic, or the commander of the forces --overwhelmingly American -- that occupied Japan after the surrender and until April 28 of 1952. The acronym actually designates Generals MacArthur and later Ridgway, but in practice it refers to the occupation as a whole. See Kurt Steiner, Local Government in Japan. Stanford: Stanford University Press, 1965, p. 64 fn. 138 large part from central government agencies' efforts to recover control over their respective corners of local affairs. Delegated functions allow central ministries to foist administrative responsibilities and fiscal burdens onto local authorities. But these functions are not, strictly speaking, an element of the intergovernmental tax system. They are, moreover, declining in their direct fiscal significance because of the gradual reduction of specific subsidies. 4 0 The above shows that the policymaking regime for intergovernmental taxation is largely dominated by MOF and MOHA. Other agencies and organizations are important, and often require some degree of attention in the effort to explain policy outcomes. But these other interests lack the authority, cohesion, and clarity of purpose that MOF and MOHA possess when it comes to intergovernmental tax politics.The following case studies therefore centre on MOF's interest in cutting transfers from the centre without ceding tax room to the local authorities or MOHA; and MOHA's interest in wresting tax room from MOF without ceding its own controls to the local authorities. On such "agency" and "entity" delegated functions and the problems they entail, see Richard Samuels, The Politics of Regional Policy in Japan: Localities Incorporated. Princeton: Princeton University Press , 1983, pp. 42-4; and Kurt Steiner, Local Government in Japan, op. cit., pp. 232-62. See chapter 1 for a discussion of the gradual withdrawal of specific subsid ies. 139 Chapter 5 Controlling Redistribution: The Local Allocation Tax The rules for distributing the Allocation Tax are a black box -- only the MOHA people know what goes on inside. MOF Tax Bureau Interviewee Simply substituting specific subsidies with the Allocation Tax raises questions about the current structure that is strengthening MOHA's intervention in local affairs. Mayor of Takeo City, Chiba Prefecture Nihon Keizai Shinbun, January 6,1997, p. 28. Recentralizing Redistribution The current Local Allocation Tax (LAT) and its predecessors were described in some detail in chapter 3. The object in that exposition was to depict the central fiscal devices of the prewar, wartime, and postwar periods; in other words, to relate the institution, reform, and re-institution of the 1940 system. This section will retrace some of that ground, but along rather a different path. The focus here will be on the interministerial politics of the LAT as well as the tax's role in inhibiting local fiscal autonomy and impeding reforms in that direction. In particular, I stress that determining the rules for allocation is not the apolitical process many political scientists believe it to be. 140 One of the first items to clarify concerning the LAT is that it is not actually a tax. It is, rather, a transfer mechanism whose funding is derived from fixed percentages of selected national taxes. 1 The designation of the LAT as a "tax" stems from the 1940 reform, which the previous section noted saw the institution of the Local Apportionment Tax (Chihou Bunyozei) on the basis of set percentages of income and other taxes. The funds were accumulated in a special account for the transfer, and the accumulated revenues were then distributed equally to the prefectural and municipal levels, according to several measures aimed at calculating relative fiscal need. The significance of the transfer increased, as we saw in Table 3.4, because as a proportion of the local authorities' aggregate revenues it rises from 9.2 percent in 1940 to 15.8 percent in 1945. Japan's 1945 surrender brought chaos to the militarized economy and administration. Tax revenues in 1946 plunged to 7.2 percent of national income from the previous year's 11.2 percent. Moreover, the source of revenues shifted from the wartime average of 58.6 percent direct and 33.7 percent indirect taxes to 43.6 percent and 46.9 percent respectively in 1946. This decline in revenues from the major direct taxes reflected not only Japan's woeful economic circumstances, but also a host of administrative problems that allowed income taxation to be evaded or paid in the next or even the 'Spec i f i ca l ly , the Local Allocation Tax in 1995 drew its revenues from a 32 percent share of the national personal and corporate income taxes as well as the liquor tax; a 25 percent share of the national tobacco tax; and a 24 percent share of the national consumption tax. 141 second fiscal year in greatly depreciated currency. 2 Of course, in this period S C A P ' s ideology of local fiscal autonomy was the strongest influence on Japanese tax policymaking. This ideology went in tandem with, it is important to note, the Occupation's efforts to decentralize policing, education, and other centralized state functions. In consequence, the responses to the early postwar local fiscal crisis show a decentralist trend in both finances and administrative burdens. Unprecedented fiscal reforms began with the earliest tax revisions in the wake of the surrender and continued through the legislation based on the Shoup Report. The Shoup reform sought, as part of fiscal decentralization, to put the Distribution Tax on firmer and more favourable ground for the local authorities. Hence, Shoup recommended the rules governing the tax's distribution be modernized through an increase in the range of local circumstances taken into account in calculating financial need. Shoup also recognized that the central state's sole jurisdiction in deciding the level of and rules for fiscal redistribution greatly circumscribed local autonomy. The Shoup Report therefore sought to revise the structure of the Local Autonomy Agency, the descendant of the 2 On these figures and the administrative problems, see Henry Shavel l , "Postwar Taxation in Japan , " in the Journal of Political Economy, Volume 56, 1948, pp. 126-30. A curious aspect of Japan 's f iscal history, when seen against the rather authoritarian prewar state and its heavy expenditures on national defence, is that tax enforcement was remarkably lax. Taxpayers incurred no penalty for not reporting particulars about their income, though they were legally obliged to do so. Shavell notes that there was not s o much as one case of conviction or punishment for tax evasion up to 1948, and that Japanese tax officials regarded criminal punishment for tax evasion as "contrary to the Japanese philosophy of 142 Home Ministry's Local Affairs Bureau, so as to secure direct representation from the local authorities. In this respect, Shoup was evidently taking a lesson from S C A P ' s inclusion of the six C R D chairmen on the Local Finance Commission, which was one means through which S C A P sought to bypass the uncooperative former Home Ministry bureaucracy in the early postwar period. The new institution recommended by Shoup was to have been called the Local Finance Commission and, ideally, would mediate between the local and central authorities on the Allocation Tax, the apportionment of debt, and other fiscal issues. But this plan did not come to fruition due to the reconstitution of the Home Ministry's Local Affairs Bureau through its absorption of the Commission. Conflicts with MOF Recentralization of the fiscal regime in the 1950s virtually eliminated institutionalized local input in the management of the system, but did not depoliticize the process of decionmaking on redistribution. Maintenance of the old role of the Distribution Tax entailed the redistribution of centrally controlled tax revenues; as in the 1940 tax reform, two central agencies (MOF and the Local Autonomy Agency [Jichichou]), were in charge of general subsidies. But in the postwar order, these institutions lacked the exigency of stabilizing the local areas in the midst of a global war. just ice" (p. 130). It is small wonder, then, that the Occupation's later efforts to double revenues through 143 The initial post-Occupation structure for determining the amount of the LAT involved annual negotiations between MOF and the Local Autonomy Agency, whose fiscal interests were much at odds. MOF approached budgeting issues in the wake of the Dodge Line with an austere cast of mind and a commitment to keeping the burden of the state below twenty percent of gross national product. This policy orientation led MOF, during annual budgeting sessions, to overestimate the level of anticipated local revenues. On the other hand, the Local Autonomy Agency challenged MOF's figures with its own underestimates, as it sought to enlarge its jurisdiction and secure funds to cope with local fiscal crises. The absence of a fixed framework for deciding the level of redistribution left the matter to interministerial political maneouvring. A former official from the Local Autonomy Agency describes the negotiations between MOF and the Agency as having been particularly tough because of this problem. 3 One solution to the difficulty was offered through the Local System Advisory Commission's (Chihou Seido Chousakai; herafter, L S A C ) report of October 1953, which recommended that the structure of redistribution be revised. This report, produced very much as a reflection of the Local Autonomy Agency's aggressive tax collection caused riots and other disturbances. 3 Note the interview with Furuya Keigi, in Shougen, Chihoujichi: Naimushou Kaitai, Chihou Bunkenron [Oral History of Local Autonomy], by Honma Yoshito. Tokyo: Gyousei , 1993, pp. 340-77, especial ly pp. 353-55. 144 aims, 4 advocated a major reversal of the Shoup reforms. Against the backdrop of continuing fiscal crisis at the local level and the concern about politicization of the intergovernmental tax-transfer system, the report called for stable and centrally directed redistribution. Specifically, the report recommended the recentralization of several important taxes, including the Admissions Tax (Nyuujouzei) and the Amusement Tax (Yukyouinshokuzei), as well as the transfer of a share of the Fixed Assets Tax base [Koteishisanzei) from the municipal to the prefectural level. Other proposals for centralization included terminating Shoup's local Value Added Tax (Fukakachizei) and reinforcing the Enterprise Tax (Jigyouzei) instead; adjusting tax controls; and introducing a system of debt management that monitored the tax effort of indebted local authorities and placed them under the authority of the Local Autonomy Agency. On August 7, 1953, the cabinet created the GTAC and thus formalized MOF's earlier ad hoc tax system of deliberation councils. The GTAC also went on to study the problem of fiscal transfers, as its mandate permitted this sharing of administrative purview with the pre-existing L S A C . The creation of the GTAC at this point, and its grant of administrative purview over local governments' tax matters, clearly displeased the head of the Local Autonomy Agency's Finance Department.5 This reaction is indicative of the friction between the two 4 See supranote 33, chapter 3. 5 The Finance Department head indicated that he felt the new deliberation group was redundant (dasokuteki), but added that he welcomed as many views as possible on the question of local taxation. 145 central agencies. However, the annual fights over the LAT needed resolution in order to avoid politicization of the system. The GTAC's November 11, 1953 report on the local tax problem was thus virtually identical, on the major points, with that of the L S A C . The MOF-dominated GTAC report differed, for the most part, only in advocating a more restricted scale for local finance and in making minor adjustments to some intergovernmental tax arrangements. 6 Both reports were taken up by the political process in February, 1954, and resulted in a series of legislative changes that 1) implemented the compromise on the LAT; 2) centralized the taxes identified in the reports and set up the redistributive structures called for; 3) installed the system of debt control; 4) and enacted several of the other measures desired by the two deliberation councils. The outcome most germane to this section was the 1954 institution of the LAT, with the rates of its component personal income, corporate income, and liquor taxes set at twenty percent. These rates were raised through the high-growth period with little friction between the two ministries, until revenue constraints See Takamitsu Maruyama, Nihon Chihou Zeiseishi [A History of Local Taxation in Japan]. Tokyo: Gyousei , 1985, pp. 462-3. 6 See "Zeiseichousakai toushin, Shouwa 28-Nen, 11-Gatsu" [Tax Advisory Commission Report, November, 1953], in Zeiseichousakai Toushinshuu: Shouwa 28-Nen irai, Shouwa 37-Nen 9-Gatsu [Tax Advisory Commission Reports, 1953-September 1962]. Tokyo: MOF, MOHA Tax Bureaux. 146 reemerged in the late 1960s. An additional reason for the relative quiescence was that both MOF and MOHA benefited from cooperating on increases in the rate of contributions to the LAT during this period. The ministries enhanced their jurisdiction through the negotiations concerning the tax and, by extension, the superintendance of local fiscal affairs. However, it also meant that their cooperation became more crucial to the fiscal process. 7 By 1965, the two ministries had found enough common ground, in a regime of rapid economic growth, to increase the rates of personal income, corporate income, and liquor taxes in the LAT to thirty-two percent. The rates have remained unchanged since then, however, because the era of relatively easy financing came to an end and the MOF-MOHA conflicts resumed. Allocating the Rules of Distribution The rules governing the distribution of the LAT come under considerable suspicion from many local governments, public finance specialists, and not surprisingly, bureaucrats within MOF. In fact, the 1954 compromise implemented an extremely detailed method of computation for determining if a local authority is entitled to LAT revenues, and if so how much. The computation involves adding up the relevant local government's tax effort, or Basic Financial Revenue (Kijun Zaisei Shuunyuu), exclusive of non-statutory 7 See Takamoto Shousaku, "Sengo Taisei no Keisei" [The Construction of the Postwar System], in Nihon no Chihou Seifu [Local Government in Japan], Ed. by Satou Seizaburou, Oomori Wataru. Tokyo: Tokyo University Press, 1986, pp. 47-110; p. 96 147 taxes and other specified items. Also calculated is the local authority's Basic Financial Need (Kijun Zaisei Juyou). This figure expresses the local authorities' demand for expenditures according to a complex formula that includes the particular jurisdiction's total area, population, number of students and classrooms, length of roads and rivers, degree of coldness and amount of snowfall, and other variables. These variables are meant to ensure that all local authorities are treated in a manner that takes adequate note of possible special circumstances, such as a large number of retirees in the community. Akizuki argues that the calculations are removed from the arena of zoku politics and other distorting influences. In other words, the distribution of LAT revenues, once they are calculated, is not politicized by MOHA or other agencies, and this helps explain why Diet policians do not try to intervene in the process. 8 There is, however, considerable room for discretionary control over LAT funds, because MOHA's Finance Bureau makes adjustments to the formula for redistribution. This may seem a trivial matter; and, as was noted earlier, virtually all political science accounts of Japan's general transfer regime regard the LAT as an unproblematic shift of funds from the centre to the fiscally weaker local authorities. But in fact, redistribution is not that simple if one looks at the phenomenon from a comparative standpoint. "Institutionalizing the Local System: The Ministry of Local Affairs and Intergovernmental Relations in Japan , " in The Japanese Civil Service and Economic Development, op. cit. 148 The issue here is not merely, as noted in chapter 1, that the LAT revenues are often effectively earmarked for paying down debt. Nor is it that 6 percent of the LAT is distributed according to discretionary criteria. A further problem is the question of who chooses the elements that compose the formula for redistribution. Though the process may be properly bureaucratic in the wake of the selection, the items deemed appropriate to include "are largely the result of political decisions." The truth of this assertion is seen, for instance, in the enormous variation in the criteria for redistribution throughout Europe and even among the German states. Redistribution to local authorities is predicated on conceptions of spending needs, and "it must be recognized that overall assessments of spending needs inevitably involve many value judgements as to the relative importance of different local services and the 149 weight to be given to different indicators of need." 9 Because of the inherently political nature of these choices, the selection of redistributive criteria is generally debated in open fora that include not only bureaucrats but local and national politicians as well. In Australia, for example, the high degree of centralization in the tax system (i.e., the federal government monopolizes the income tax field) has brought about compensating institutions for intergovernmental bargaining and ratification of the criteria for distributing grants. In Canada, in fact, successive rounds of "executive federalism" have seen provincial leaders dismantle piece by piece the federal government's powers to attach conditions to its fiscal transfers. Similar mechanisms for intergovernmental negotiation are evident in most other federal and unitary states, especially those where the scale of transferred resources approaches that seen in Japan. But Japan does not have institutionalized consultation. There are ad hoc consultations with prefectural governments, but these are not routinized and do not confer explicit bargaining rights on local authorities. Certainly it is more efficient to have MOHA act in this area on behalf of Japan's 47 prefectures and other subnational governments, due to the difficulties of achieving consensus among such a large number of actors. Yet the choice of institutions to cope with this collective action problem is not so constrained that centralization in 9 Philip Blair, "Financing Equalization Between Local and Regional Authorities in European Countries," 150 MOHA's Finance Bureau is the only viable alternative. Indeed, the degree of centralization is so great that only ten members of the Finance Bureau actually do the work of adjusting the rules. This may be an efficient and unproblematic solution, if viewed in rational choice terms as an agency relationship in which politicians ultimately control the delegated responsibilities. But one can also argue that the trade-off of centralization in order to enhance efficiency gives MOHA an indisputably large scope in which to impress its own institutional agenda on the rules of distribution. These points were confirmed by a recent nationwide survey of Japanese mayors, wherein 67 percent of the respondents complained that MOHA's method of calculating the LAT was opaque, 44 percent argued that the distribution does not reflect the wishes of the local authorities, and a further 22 percent felt it was improper (okashii) that MOHA changed the rules for distribution every year. As the mayor of Sagamihara City in Kanagawa Prefecture argued, "authority for determining the distribution of the LAT should be transferred from MOHA to the C R D and other local organizations." 1 0 Further evidence that the LAT affords discretionary powers to MOHA is seen in the pressures from other spending ministries and their allied zoku politicians Local Government Studies, 1992. 1 0 The comprehensive survey was undertaken in order to highlight problems in the current trend towards shifting specif ic subsidies into the general subsidies account (i.e., the Allocation Tax). See "Kuni tanomi no kouzu kawarazu" (The structure of dependence on the central government remains unchanged), Nihon Keizai Shinbun, January 6, 1997, pp. 28-9. 151 to have the LAT rules altered to favour particular distributive ends. This intra-state lobbying includes, for instance, the Construction Ministry and affiliated politicians' efforts to have more fiscal room granted for the support of road-building and road maintenance. Such line ministries also gave tacit support to the Local Autonomy Agency's early postwar pressures for higher rates of the LAT's component taxes. The Education Ministry was supportive because the LAT funds, which are general revenues once disbursed from MOHA's Finance Bureau, could be used as funds for compulsory education expenses. In other words, the role of the Local Autonomy Agency itself expanded because it could decide whether assigned responsibilities for the local authorities should be paid for via subsidies or be included in the calculations of a specific local authority's basic fiscal demand. If the latter, the cost of the delegated work would become an element in the determination of the amount of LAT to be disbursed to the particular local authority.11 This utility of the LAT to MOHA is often unappreciated in studies that view it as merely a general fiscal transfer. 1 2 The bargaining that takes place with other central agencies concerning the rules of distribution is neither institutionalized nor in any manner of speaking open. It is in thus understandable that many academics and public officials 1 1 See Takamoto Shousaku, "Sengo Taisei no Keisei" [The Construction of the Postwar System], in Nihon no Chihou Seifu [Local Government in Japan]. Ed. by Satou Seizaburou, Oomori Wataru. Tokyo: Tokyo University Press , 1986, pp. 47-110; p. 94. On the computation of the Allocation Tax itself, see the detailed description in Yonehara Junishichiro, "The Financial Crisis of Local Government in Japan , " Occas iona l Paper No. 9 of the Centre for Research on Federal Rnancial Relations, The Australian National University, 1979. , 2 Kengo Akizuki makes this argument, for example, one which is common to conventional and revisionist accounts of MOHA's tax controls; see "Institutionalizing the Local System," op. cit, pp. 360-61. 152 regard this backroom dealing and MOHA's dominance of the rules over distributing the LAT as a "black box." 1 3 Therefore, rather than a simple fiscal transfer, the LAT must be viewed as part of a complex of institutions that restrict broader input into decisionmaking. The LAT is both a "no strings attached" transfer and an important element of MOHA's structure of influence and fiscal control. Below I sketch the means through which the latter phenomenon is said to operate. The Politics of Allocation Tax Increases The 1954 compromise between MOF and MOHA on the LAT turf depended on larger fiscal trends, as I alluded to above. Even during the rapid average annual growth of the Japanese economy, there were troughs in the wake of the biggest crests of the economic boom. These slowdowns in growth showed up quickly in reduced natural revenue increases from the highly elastic income tax, and brought pressure for fiscal readjustment from MOF's Budget and Tax Bureaux. Morever, the scale of the state expanded through public works projects and the mounting, albeit modest, outlays on welfare programmes and services. 1 3 The phrase was used in interviews at M O P s Tax Bureau, and it is not uncommon in newspaper and academic sources . 153 Conflict between MOF and MOHA over redistribution reemerged in the mid-1960s. From MOF's perpective, a major problem was excessive local spending, especially on salaries, which was gradually pushing the central state toward the much-loathed step of deficit spending. The dislike MOF had of this circumstance can be gauged from the fact that the ministry had scrupulously avoided deficits after the imposition of the Dodge Line in 1949-50. As a result, in 1965 MOF sought to offload more of the burden of local projects onto local debt even as it conceded to the final increase (from 29.5 to 32 percent) in the rates of component taxes for the LAT. This policy line was enunciated in the November 1 mid-term report of MOF's Fiscal System Advisory Council (Zaisei Seido Shingikai). MOHA's Local System Advisory Council (Chihou Seidi Chousakai) answered the challenge in a March 8,1966 report on local taxation and transfers. MOHA clearly prefers general transfers to local debt or specific subsidies. Debt was a major aspect of fiscal crisis in the prewar and early postwar eras, and hence is anathema to MOHA's organizational mission of stability and uniformity. The latter source of funds, specific subsidies, is also disliked because it involves generally having to share administrative turf with the spending ministry in charge of the subsidy. 1 4 The MOHA advisory council, not surprisingly, came 1 4 Ak izuki Kengo, "Institutionalizing the Local System,"op. cit., pp. 354-55, confirms that there is considerable friction between MOHA and line ministries on the issue of specif ic subsidies. A lso, note that both M O F s General Affairs sect ion in charge of the budget for MOHA, and MOHA's Fiscal Affairs Bureau share interests here. The MOF agency would be in charge of a larger budget were all subsidies made general and thus transferred via the Allocation Tax. The agencies' d iscussions of this are well-154 down clearly in favour of general transfers versus the position put forward by MOF. These reports were official expressions of the new fiscal reality in intergovernmental affairs. Long in advance of Diet politicians' turn to fiscal austerity in the 1980s, MOF had announced its concern about spending pressures and its resolve to do something about them. By contrast, MOHA signalled its readiness to stand in opposition to MOF's plan to offload fiscal burdens onto local governments. The conflict between the two ministries escalated the next year, when MOF was compelled to float bonds to cover a revenue shortfall. Being driven toward deficit spending -- albeit the ostensibly productive outlay of construction bonds -- was bad enough. But in addition, MOF was losing its capacity to make adjustments in the budget because expenditures were effectively becoming earmarked. In November of 1967, MOF came out strongly against "fiscal rigidification" (zaisei kouchokuka) and pointed to four major sources of this problem. These sources included such politicized transfers of tax dollars as the rice-price guarantee, the cost-inefficiency of the railways, and the mounting deficits of the health insurance system. The fourth contributing factor, though, was the LAT, which MOF declared that it wanted to cut from the 32 percent level (of component taxes) that had been set only two years before. known in Japanese public finance circles, and could -- as spending ministries fear -- some day become 155 MOF and MOHA's squabbles over funding the LAT continue, their intensity varying inversely with the rate of increase in tax collection. During the steep decline in revenue collection in the early 1990s, for example, MOF's Budget Bureau again pressed hard for outright cuts in the component taxes. The argument was that local budgets were robust relative to that of the central state. But these recent efforts to cut the component tax rates have so far been strongly and successfully resisted by MOHA. MOF-MOHA Compromise on the Special Account Possibly contributing to the stability on the rates of the LAT is the MOF-MOHA compromise on using it as a tool of financial administration. This unusual compromise on the transfer was reached in 1987, with the institution of the Specia l Account for Loans from the LAT (Koufuzei Tokubetsukaikei Kariirekin). This and subsequent measures allow for temporary reductions to be made in LAT transfers from MOF in the event of serious fiscal stress. The unpaid balance of the LAT is treated as a loan from the special account to the central state, and is to be repaid in the next year or afterwards. This system might, at first glance, appear to be a harmless bureaucratic expedient. However, it has come under heavy criticism in public finance circles because the movement of the funds in and out of the account is unclear and not under the authority of the institutional roots of a major policy change. 156 local governments. Moreover, shifting funds around across years -- rather than within the same fiscal year -- is an irregular procedure and beckons further study. 1 5 MOF's reasons for entering into the compromise instead of simply pushing for cuts in the LAT can only be speculated at. According to Japanese public finance scholars, MOF wanted to keep politicians from interfering in this aspect of the fiscal system and perhaps turning it into a massive and expanding porkbarrel. A full-scale bureaucratic fight on this issue would almost certainly invite political attention, as alterations to the basic framework of the LAT require the approval of the Diet. On the other hand, MOHA's reasons for participating in the scheme seem obvious. Cuts in the transfer would reduce MOHA's means with which to fulfill its organizational mission of smoothing out fiscal gaps among the local authorities. This feature, in tandem with ministerial discretion concerning the rules for allocation, is likely incentive enough for compromise. Yet here again, the outcome of the compromise is a structure outside of the direct influence of the local authorities. 1 5 A short description of the problem is in "Zaisei kenzenka ni 'gyaku kouka"' [An outcome contrary to the pursuit of f iscal soundness] , number 8 in the Nihon Keizai Shinbun series on decentralization, May 1, 1996, p. 25. My thanks to John C. Campbell, an expert on Japanese budget politics, forcomfirming that the transfer of funds across the years is highly unorthodox . 157 General Subsidies and Fiscal Dependence A related reason for MOHA to strike such bargains with MOF is that, in spite of the complaints from many mayors and other local officials, the LAT does serves MOHA well in maintaining a large bloc of satisfied local governments. This phenomenon has long been evident in Japanese intergovernmental fiscal politics, and is likely a property of any system of redistribution between --to use the Canadian parlance -- "have" and "have not" subnational entities. Moreover, as is perhaps also the case in all such systems, the entities that pay more than they receive in transfers generally support fiscal decentralization while the net beneficiaries are considerably less enthusiastic about such policies. Hence the unwillingness of many smaller prefectures and municipalities to follow the lead of Tokyo and other large constituencies, which favour a more liberal fiscal regime. A large number of these less fiscally robust local governments have become dependent on LAT transfers. Indeed, some administrations opt to take the easiest path by reducing their tax effort to the standard rates, thus gaining their specified share of LAT revenues and avoiding the political pain of taxing up to the maximum permitted by MOHA. There are, in this case as well, at least two ways of looking at the phenomenon. One perspective, implicit in the revisionist literature, suggests that a large bloc of satisfied local governments means the LAT is fairly distributed among local governments. Were the system of general transfers grossly inadequate, 158 capricious, or coercive, there would be less willingness to rely on it. Hence, if one views the LAT in isolation from its institutional environment, it is easy to argue that it satisfies local fiscal needs at a marginal perceived cost in local autonomy. On the other hand, a more critical point of view would note that MOHA certainly has to ground its fiscal hegemony on a large measure of consent. Japan is a centralized state, but it is also indisputably democratic, with other opportunities for local governments to air grievances if they choose to bypass MOHA. If my argument concerning MOHA's organizational mission of perpetuating fiscal control is true, it follows that it is in MOHA's interest to create a bloc of dependent and supportive, or at least compliant, local governments. A corollary of this condition, however, is much dissatisfaction among the larger administrations, which pay into the LAT base but receive little or nothing in return. This phenomenon is, as I noted earlier, common for wealthier constituencies in systems that redistribute revenues among subnational governments. Rich provinces in highly decentralized Canada have long felt, for example, that they are unfairly burdened in order to pay for fiscal federalism. 1 6 In Japan, Tokyo contains 9.4 percent of the population and 1 6 Some of the politics of Canada's intergovernmental f inance are reviewed in Ikegami Taihiko, "Kanada ni okeru Renpou, Shu aida Zeigenhaibun Ronsou" [The Intergovernmental Debate over Revenue Distribution in Canada], in Chuo Daigaku Keizaigakuronshuu, Vol. 36, No. 4, October, 1995. 159 represents 18 percent of gross national product, but supplies fully 29 percent of personal income tax, 37 percent of corporate tax, and 35 percent of consumption tax revenues. 1 7 These revenues are the main elements of the LAT, from which Tokyo receives no transfers. Since some smaller local governments are not taxing to the extent that they could be, revenues gleaned from Tokyo and other areas are effectively being misallocated. Such monies, the argument goes, could be spent on the notoriously poor infrastructure of Japan's larger metropolises. 1 8 That is, many local authorities do not challenge MOHA's dominance and scope for discretion concerning the LAT (including the 6 percent portion), because the only perceived alternative is a fiscal autonomy that would compel them to place heavier burdens on local taxpayers or rationalize their expenditures. The political implications of arguing that the local authorities should be "weaned" from subsidies, both specific and general, will seem 1 7 The figures are from, "Zaiseichousei ni 'kyoudouzei ' kousou" [Fiscal adjustment and cooperative taxation], item number 9 in the series "Chihoubunken no riron" [The debate on decentalization] in Nihon Keizai Shinbun, May 2, 1996, p. 29. 1 8 Note that the calculation of basic financial revenue for prefectures and municipalities is based on 80 and 75 percent, respectively, of total revenues. This is to provide incentives to enlarge the local tax base, by avoiding the situation wherein "any increase in local tax revenue would reduce the local allocation tax by the same amount"; see Junschichiro Yonehara, "Relations Between National and Local Governments," Japan's Public Sector: How the Government is Financed. Ed. by Tokue Shibata. Tokyo: University of Tokyo Press , 1993, p. 174. Estimating the efficiency loss through transfers managed by MOHA is beyond the scope of the present work. Some comparative light can be shed on the matter through the example of Australia, however. Australia's f iscal equalization mechanisms are "often described as the most comprehensive and explicit in the world," include far more variables than Japan 's , and are also ratified through intergovernmental bargaining at the Premiers' Conference. Even so , the net welfare loss of the distortions on allocation has been estimated at "around twenty per cent of the aggregate cost equalization transfers." See Peter Groenewegen, "The political economy of federalism s ince 1970," State, Economy and Public Policy in Australia. Ed. by Stephen Bell and Brian Head. Oxford: Oxford University Press, 1994. 160 familiar to anyone who has followed the trajectory of fiscal decentralization debates elsewhere. In many other countries with large-scale redistribution, those arguing decentralization are often hoping for, so to speak, a leaner and meaner local state. The not-so-hidden agenda of devolutionist Republicans in the United States, for example, has with good reason been characterized as a mean-spirited (and likely ineffective19) fiscal "stimulus" to the poor. Moreover, most work on fiscal decentralization is aimed at the question of whether devolution can shrink the state, with rather a technocratic disregard for what the expenditure cuts might cost in human terms. 2 0 But in Japan support for decentralization has long been the preserve of the Left, and only in recent years has it gained a strong following among the conservative establishment. Rationalizing expenditures at the local level can be viewed as a progressive policy option in Japan because of the scale of waste and political corruption associated with the public works that constitute a large share of local spending. Attempting to secure fiscal discipline at the local level, via the reduction of transfers and a shift towards greater reliance on local resources (with a concommitant transfer of tax controls), is therefore not a neo-liberal strategy to remove the so-called "undeserving poor" from the welfare rolls. Instead, the aim is to enhance reliance on local fiscal resources 1 9 According to the Dave Lesher of the Los Angeles Times, California Governor Pete Wilson's "highly touted attempt in 1992 to shake people off welfare with carrot-and-stick financial incentives has failed, a two-year study has concluded"; see "State's Welfare Program Fails, Study S a y s , " January 5, 1997, p. 1. 2 0 Note, for example, Jaber Ehdaie, "Fiscal Decentralization and the Growth of Government Spending," The World Bank Policy Research Department, Public Economics Division, June 1996. 161 and thus reduce wasteful public works through breaking the existing pattern of corrupt fiscal politics. 2 1 Proposals for Reform The Japanese public debate recognizes some of the problems posed by excessively centralized control of the LAT. In fact, the final report of the Decentralization subcommittee (Chihoubunken bukai) of the Administrative Reform Promotion Group (Seifu no Gyousei Kaikaku Suishin Honbu), submitted on October 4, 1994, advocated major changes to the LAT law. The subcommittee reasoned that the constituent taxes should be earmarked for direct deposit into the LAT fund rather than flow, as they do at present, into general revenues and then be transferred to the special account by MOF. This measure would not in itself be a major decentralist reform; rather, it is seen as the first step in that direction. 2 2 Moreover, the current debate on fiscal decentralization centres on, as was noted in the introduction to this dissertation, the reduction of specific subsidies and their replacement by general subsidies. The process is referred to, in direct translation, as "general finance-ization" {ippan zaigenka), and is 2 1 An example of the progressive argument is seen in Jinno Naohiko, "Zeigen utsuseba, zaigenryoku no jichitai kakusa heru" [Transfering taxes can reduce local f iscal gaps], Asahi Shinbun, December 13, 1996, p. 4. 2 2 Refer to "Chihoubunken no saishuu houkokuan katamaru" [The final report on decentralization is determined], in Nihon Keizai Shinbun, November 1, 1994, p. 2. 162 hindered by the uncertain political situation in Japanese party and bureaucratic circles. Yet were the policy line to meet with unanticipated success, it does not follow that fiscal decentralization would be the result. For one thing, the 1980s record of cuts to specific subsidies not being compensated by commensurate increases in general subsidies gives pause for doubt. Experience in other countries, too, suggests that general subsidies are more visible targets for restraint, as "national governments can place a limit on the total allocation to local governments by deciding first what the total funding pool will be and then distributing the pool according to an entitlement formula." 2 3 But even if fiscal restraint did not complicate the issue, the absence of formal avenues for local input into the system would not change. Prospects For Politicization Other actual and potential problems with the tax came to light in the protracted negotiations over the 1997 budget. One of the major sticking points in the negotiations between the ruling coalition and the bureaucracy concerned funding for the construction of new high-speed shinkansen ("bullet train") lines. A proposal for five new lines has long been left languishing on the drawing boards for lack of finances, especially because the projects are truly massive porkbarrel public works. On this, see Heidenheimer, etai., pp. 308. 163 The funding hurdle confronting negotiators seemed insurmountable, for two major reasons. The first is Japan's debt-strewn fiscal landscape of the 1990s. The legacy of pump-priming via public works and income-tax cuts is that Japan has, in the mid-1990s, become the industrialized world's most debt-ridden nation. The combined debt of the government sector in Japan in 1996 stood at 440 trillion yen, which has created a consensus among the public finance community that great prudence in spending is imperative. Thus large-scale projects such as the shinkansen lines are viewed as not only wasteful but a dangerous anachronism, the "dregs" (zanshi) of the high-growth era's porkbarrel politics. 2 4 The more immediate barrier arose because Japan Rail was unwilling to take on 50 percent of the construction costs, which is the funding arrangement in place for previous shinkansen projects. But attention eventually turned to whether LAT revenues could be used to cover part of the costs. MOHA has long been in principle against this use of general subsidies, and had thought the matter a dead issue after it came up on the agenda in the summer of 1996. Moreover, when it reemerged later in the year, MOHA requested a postponement of any decision to August of 1997. MOF also opposed a deal on funding, citing the country's massive debt load. However, a deal was concluded just before the end of the year, one that will see MOHA transfer LAT The comments are from Ishi Hiromitsu, a major figure in Japan's public f inance community; see the interview with Ishi in "Zaisei saiken honenuki no shouchou" [A symbol of the mutilation of f iscal reconsrtuction], Yomiuri Shinbun, December 26, 1996, p. 7. 164 monies -- equivalent to about 30 percent of construction costs of 173 billion yen -- to the local authorities where shinkansen construction is to occur. 2 5 In fact, MOHA is ambivalent on the issue. It stands to gain much prestige through its gatekeeper role with the LAT, because of the bullet-train funding sheme and other such "economic development" (keizai kaihatu) projects. These latter envision Japan as a series of separate regions of technopolises and other public-works bonanzas. Big schemes involving so much concrete and steel cost a lot of money, especially in Japan, but the targeted subsidies that are largely the purview of the Construction Ministry are on a slow decline. Hence politicians' recent interest in the LAT. But MOHA's opportunity also confronts the ministry with the risk of creeping politicization of the LAT, through the activities of the zoku giin. Moreover, as MOHA's influence increases there are almost certainly to be challenges from other ministries, which already resent its authority to vet proposals where the use specific subsidies entails a corresponding financial contribution for the local authorites. 2 6 In other words, the LAT, is certainly not a simple transfer of fiscal resources, and is apparently about to become even less so. Lacking in open and "Hashiru seiji, kasumu zaisei saiken" [Fast politics leaves f iscal reconstruction in the dust], Nihon Keizai Shinbun, December 18, 1996, p. 5. These points were confirmed via a January 7, 1997, telephone interview with a public f inance scholar who works closely with MOHA and therefore must remain anonymous here. On other ministries' worries that MOHA authority will increase through decentralization, see Yoshiaki Kato, "Chihou Bunken Chakuchi e no Michi: Kengen Ijou Gutaisaku naku"[The Road to Decentralization: No Concrete Policies for Transferring Authority], in Nihon Keizai Shimbun, October 22, 1994, p. 1. 165 institutionalized intergovernmental consultation, the transfer system is part of the regime of fiscal centralization. Moreover, because it is already a point of friction between MOF and MOHA, the likelihood of the latter's voluntarily relinquishing its discretionary authority seems minimal. This is likely to be even more true if the recent trend continues and more fiscal resources are shifted to the general versus specific subsidy account, thus inviting further attention from zoku giinin search of finances for vote-oriented expenditures. 166 Chapter 6 The Income Tax Field and Corporate Taxes Dr. Carl Shoup...once said that governments are prone to create a corporate tax system that is based not so much on economic theory or economic rationale but on one simple criterion: That tax is easy to collect and can be collected in large amounts. Nakamura Yoshio, Managing Director, Keidanren Japan's corporate income tax structure bears strong evidence of bureaucratic intervention and the revanche of the 1940 fiscal system. Of special note here is the division of the tax field between MOF and MOHA. This structure gives them strong incentives -- in the context of the party competition for personal income tax cuts -- to oppose fiscal decentralization as well as corporate pressures for tax cuts. In consequence, Japan is notable for high corporate tax rates and overall burdens relative to other advanced industrialized countries. 1940 and the Modern Income Tax The 1940 reform set the stage by modernizing the income tax through clearly distinguishing the personal and corporate income tax fields. As was noted earlier, this reform realized in large measure the prewar bureaucracy's aim of shifting the tax burden onto Japan's modern sector of salaried urban workers and capitalists. 167 Table 6.1: National Taxes By Principal Items, 1940-1945. (Units: Thousands of Yen) Year Total Taxes Total Direct PIT CIT Total Indirect 1940 4,140, 843 2, 616,123 1,488,679 182,873 1,524,720 1941 4, 817,942 3,048, 246 1,401,363 534,907 1,769,696 1942 7, 350, 520 4, 611,232 2, 236,191 775,946 2, 739,288 1943 9,731,154 5,442, 242 2, 604, 097 993, 616 4,288, 912 1944 12, 715,146 8,375,566 2, 040, 581 1,326, 514 4, 339, 581 1945 11,541,379 7, 334, 136 3, 820, 426 1,180,835 4, 207, 242 Note: Year= Fiscal Year; PIT=Personal Income Tax; CIT=Corporate Income Tax. Source: revised datafrom Table 12-8-B "NationalTaxes by Principal Items, 1927-1984," in Nihon Chouki Toukei Souran, Dai 3 Ki [Historical Statistics of Japan, Vol. 3]. Tokyo: Japan Statistical Associat ion, 1987,270-71. The 1940 reform was not, however, the ideal of comprehensive income taxation sought in the 1920s. Rather, the bureaucrats learned from earlier policy failures and thus carefully tailored the wartime tax reform so as to gain as much acquiescence as possible while still securing the essential goal of shifting the burden. The outcome centred the tax system on schedular types of income taxes, a regime the Japanese bureaucrats modelled on contemporary British and French examples. Put simply, the schedular taxes did not treat all forms of income equally as is the case with a comprehensive tax system (a characteristically North American tax style). The new system-instead taxed different forms of income, such as income from interest and dividends, at different rates, though once aggregate income exceeded 5, 000 yen (in 1940), 168 a general surtax was applied at progressive rates to the excess. 1 Income taxation thus became the base of the Japanese state's revenues in the 1940 reform. The national-level Corporate Income Tax (hereafter, CIT), in particular, rapidly emerged as a salient source of revenues for warfighting and redistribution, as Table 6.1 indicates. Table 6.1 shows that the CIT's 641 percent increase from 1940 to 1945 greatly outpaced the other sources of revenue listed, including the rates of increase of direct, indirect, and total taxes. This situation is reversed in the chaos of the early postwar period, as Table 6.2 notes. In this period, corporate profits were extremely low and often hidden from zealous tax assessors whose Occupation overseers sought to maximize revenues. Consequently, revenues from corporate income taxation were quite low relative to other taxes, and especially the Personal Income Tax (hereafter, PIT), through the remainder of the 1940s. However, shortly before the 1950-51 passage of the Shoup tax reforms, the revenues from the CIT recovered and then began to match those of the PIT. After the implementation of Shoup consolidated the Occupation's PIT reforms of 1947, the CIT resumed its movement relative to the PIT and in fact came to 1 Good sources are Henry Ludmer, "Japan's Tax Structure 1," in The Bulletin For International Fiscal Documentation, Volume 1, 1946-47, and Henry Shavell , "Postwar Taxation in Japan, " in The Journal of O E C D Political Economy, Volume 56, February, 1948, pp. 124-37, especially, p. 125. 169 exceed the latter as a source of revenues in 1960. Table 6.2: National Taxes By Principal Items. Selected Years 1946-90. (Units: Millions of Yen) Year Total Taxes Total Direct err— Total Indirect " W 1 37,438 21,332 12, 241 1,316 T6TT06 1947 189,600 99,409 79,273 7,290 90,191 1948 447,745 222,669 190,832 28, W " " 225, 976 1949 636,405 344, 351 278, 754 61,264 292,055 1950 570,849 313, 620 220,134 83,790 257,230 1951 723,144 424, 995 225, 672 183, 881 298,149 T955~" 936,892 481,120 278, 675 1960 1,801,464 978,336 390,606 573,353 823, 128 1965 3,279, 653 1,941,565 970, 359 927,120 1,338,088 1970 7, 775, 386 "5,134,413 2,428,170 2, 567,168 2, 640, 973 1975 14, 506, 758 10, 058, 267 5,482, 283 4,127, 908 4, 448,491 1980 28,373, 115 20,162,840 10,799,624 8, 922, 656 8,210,275 1985 39,150,200 28, 517, 000 15,435, 000 12, 207, 000 12, 207, 000 1990 62, 779, 800 46, 297, 100 25, 995, 500 18, 383, 600 16, 482,700 Note: Year= Fiscal Year; PIT=Personal Income Tax; CIT=Corporate Income Tax. Sources: Table 12-8-B "National Taxes by Principal Items, 1927-1984," in Nihon Chouki Toukei Souran, Dai 3 Ki [Historical Statistics of Japan, Vol. 3]. Tokyo: Japan Statistical Association, 1987, 270-71, "Appendix V: Corporation Tax Burden" and "Appendix 7: Revenue," in An Outline of Japanese Taxes, 1993. Tokyo: Tax Bureau, Ministry of Finance, 1993, pp. 343 and 346-7, respectively. The very strong position of the CIT relative to the PIT in early postwar Japan is not, in fact, anomalous. The Canadian case, for example, shows a generally equally heavy reliance on both corporate and personal income taxes at the federal level from the end of the war through to 1953. From 1953 on, though, the PIT's place in the Canadian federal revenue structure increased from 4.6 170 percent of G N P to 6.1 percent in 1989, whereas the tax take from the CIT more than halved, in dropping from 4.7 percent of GNP to 2.0 percent in 1989. 2 Table 6.3: Taxes on Corporate Income as Percentage of G D P . 1 9 6 5 1 9 7 0 1 9 7 5 1 9 8 0 1 9 8 5 1 9 9 0 1 9 9 4 Aust ra l ia 3 .8 4.1 3 .4 3 . 5 2 .8 4 . 4 4 . 2 C a n a d a 3 .9 3 . 5 4 . 4 3 .7 2 .7 2 . 6 2 . 5 F r a n c e 1.8 2 .2 1.9 2.1 2 . 0 2 . 3 1.6 G e r m a n y 2 . 5 1.9 1.6 2.1 2 . 3 1.8 1.1 J a p a n 4.1 5 .2 4 . 3 5 .5 5 .8 6 .8 4.1 S w e d e n 2.1 1.8 1.9 1.2 1.7 1.7 2 . 8 U . K . 2 . 2 3 . 3 2 . 4 2 .9 4 . 7 3 .9 2 . 7 U . S . 4 . 0 3 .6 3 .0 2 .9 2 .0 2.1 2 . 5 Source: The source for the table is Table 12 in Revenue Statistics of OECD Member Countries, 1965-95. Paris: 1996. The corporate tax revenue figures are from table 52 of the same publication. What is unusual about the Japanese case, therefore, is not the mere fact of the CIT's current salience in the country's revenue structure. Rather, Table 6.3 shows that Japan stands out because it increased its heavy reliance on corporate taxation at a time when most other developed countries (save the 2 See Table C-3 in W. Irwin Gillespie, Tax, Borrow, and Spend: Financing Federal Spending in Canada, 1867-1990. Ottawa: Carleton University Press, 1991, pp. 288-299. 171 notable example of the U.K. after its 1984 tax reform)3 were reducing theirs. Indeed, even though Japan's corporate tax revenues dropped from 29.3 trillion yen in 1990 to 19.8 trillion yen in 1994 (due to the effects of the Heisei recession), the table shows that as a percentage of G D P Japan's burden was second only to that of Australia. Public finance work often argues that the salience of the CIT in Japan's revenue structure reflects the high rate of corporate income in the national income accounts. Yet between 1950 to 1960, America, Canada, New Zealand, and the U.K. were all taxing corporations -- as a percentage of G D P -- at higher levels than Japan, whereas France, Sweden, and West Germany were imposing lower burdens. The Anglo-American countries' high levels of corporate income taxation was the legacy of political mobilization for fiscal democracy coupled with the ratcheting effect of the second world war. By contrast, the Europeans and Scandinavians taxed corporate income rather lightly, providing very broad-based tax breaks for investment that allowed the state to shape economic development. Hence the standard argument that Japan has a high level of corporate-derived revenues in its mix because of the relative importance of corporate activities in 3 Sven Steinmo suggests , in his account of Britain's 1984 corporate tax reform, that the majority of Tory politicians, including the Chancellor of the Exchequer, expected the burden on capital would be rationalized and reduced. The latter was not realized in large part because the reform was drawn up under conditions of budget secrecy, with virtually no consultation outside Treasury circles. In other words, the government "miscalculated" and then held fast to its reform. See Steinmo, Taxation and Democracy, op. cit, pp. 174-7. 172 the national economy appears rather problematic if offered as the sole explanation for the tax burden. 4 Much of the variation among countries' corporate tax burdens in the early postwar period appears instead to depend on the scale of tax expenditures and the degree of targeting in them. Also, the public finance argument neglects Japan's postwar fiscal politics, when virtually annual PIT cuts created pressure to turn to the CIT. Japan's Postwar Income Tax Cuts Until the Tanaka administration made an explicit commitment to welfarism in 1971, postwar Japanese governments followed a policy of keeping the overall tax burden as a proportion of G N P to below twenty per cent of national income. Achieving this aim in a rapidly growing economy, as Japan's was for close to two decades, required large and virtually annual income-tax cuts, due to the sharp progressivity of the tax system. 5 In the absence of such reductions, the volume of revenues would quickly have become a flood as taxpayers' accelerating incomes moved them up through the progressive rate structure of 4 For data on corporate tax burdens in the 1950-1960 period, see Kojima Ryutaro, "Sengo Nihon no Zeisei to Shihon Chikuseki" [Capital Accumulation and Postwar Japan's Tax System], in Zaisei Seisaku [Rsca l Pol icy], ed. by Fujita Sei . Tokyo: Nihon Keizai Shimbunsha, 1973, Table 5.1, "Ippan Seifu no Keijou Shuunyuu Kousei no Kokusai Hikaku (1950-1960 Nen)," p. 89. The standard argument about Japan 's comparatively high level of corporate activities in the economy can be found in Aoki Tarao, "The National Taxation System," in Japan's Public Sector: Howthe Government is Financed. Ed. by Shibata Tokue. Tokyo: University of Tokyo Press, 1993, 114-16. 5 Joseph Pechman and Keimei Kaizuka note that annual natural revenue increases via the high elasticity of the PIT in Japan were used to fund more generous exemptions for individual taxpayers; in "Taxation," in Asia's New Giant: Howthe Japanese Economy Works. Ed. by Hugh Patrick and Henry Rosovsky . Washington: The Brookings Institution, 1976, p. 346. See also Ishi Hiromitsu, The Japanese Tax System, 2nd Edition. Oxford: Clarendon Press, 1993, pp. 44-55. 173 the PIT. But the straight technical explanation for virtually annual tax cuts belies the tax politics of the early postwar era. The goal of the Occupation, prior to the arrival of Shoup and his tax reform mission, was to maximize revenues in order to pay for the costs of reconstruction and the burden of the occupying forces. The emphasis on the American side was to limit as much as possible the burden to be borne by American taxpayers, and this led to comparatively heavy exactions on the Japanese public. This problem was compounded by Japan's poorly developed tax administration, which afforded few impartial means of assessment to the tax collectors. Hence, the collectors, with little training but lots of pressure to fulfill individual tax collection quotas, at times levied assessments that exceeded taxpayers' net worth. The result of this was public furor, to the extent that tax collectors were at times physically assaulted. Front-line personnel in the revenue-collection offices, under siege by enraged taxpayers, thus became a major source of institutional pressure on MOF for leniency concerning the PIT. This pressure from the front-office personnel was further passed along as MOF inculcated the conservative parties with the political utility of cuts in the PIT rates and increases in exemptions. Indeed, the early postwar political environment was such that the lesson was obvious to conservative politicians save those who sought a large state, remilitarization, or tax cuts to be centred on their small-business 174 supporters. Moreover, the Socialists too emphasized the need for PIT reductions. This position reflected their reliance on company unions and the general lack of an association between income taxation and democracy in postwar Japan. In addition, the Left worried -- not without reason --that a highly elastic tax base would encourage Japan's unrepentant conservatives to rearm. Party politics in postwar Japan therefore centred quite early on income tax cuts. Of course, few people anywhere actually enjoy paying the income tax, so the mere fact of opposition to it in Japan does not explain outcomes in policymaking. What is important, in a comparative sense, is that no major interest in Japan stood in defence of the PIT. This outcome was not evident, by comparison, in the Anglo-American democracies, as Canadian and American social democrats looked to income taxation to play a central role in redistributing the burden of the state. Japan's strongly negative politicization of the PIT hence confounded the U.S. Occupation's efforts to impose the Anglo-American version of fiscal democracy on an institutionally quite dissimilar polity.6 In addition, differences in corporate and personal income tax burdens stem at On tax protests, see Ohtake Hideo, "Hatoyama, Kishi Jidai no Chisai Seifuron" [The Small-State Debate of the Hatoyama and Kishi Era], op. cit.; also Ronald King, "From Redistributive to Hegemonic Logic: The Transformation of American Tax Politics, 1894-1963," in Politics and Society, Vol . 12 , No. 1, 1983. 175 least in part from differences in political institutions. Sven Steinmo's work, reveals that the US relies heavily on progressive forms of income taxation which over the postwar period were hollowed out by a myriad of highly specific loopholes. Steinmo explains this fact through reference to America's highly porous and fragmented set of political institutions: these create a context in which it is both possible and rational to pursue and extend very selective corporate income tax expenditures. 7 Ronald King also remarks that the early postwar equation of income taxation with democracy, an ideology given political force by the Democrats, labour, and other progressive interests, saw business and the Republicans eventually respond with an insistence on incentives. King notes, for example, the Ways and Means Chairman's hyperbolic comment in 1947 upon introducing a 20 percent across-the-board stimulative income tax cut, that "we Republicans have been warning that the short-haired women and long-haired men of alien minds in the administrative branch of the Government were trying to wreck the American way of life and install a hybrid oligarchy at Washington through confiscatory taxation." But this aggressive and overtly ideological tax agenda was not appropriate to the new postwar politics of the income tax. Efforts to make wholesale cuts to the tax gave way to less politically charged, and hence more effective, means of securing selective tax reductions for the Republicans' client interests. The 1954 Revenue Act then codified the selective incentives approach, such that "official rate schedules would be kept sharply progressive as a matter of 7 Sven Steinmo, Taxation and Democracy. Cambridge: Cambridge University Press, 1993. 176 conscious policy yet would be combined with a number of categorical exemptions and dispensations." 8 Postwar Japan's Special Tax Measures Japan, too, made considerable use of incentives in the field of corporate income taxation, but with the important difference that the exemptions and deductions were constructed in the classic across-the-board bureaucratic pattern. The process of tax policymaking in fact strongly resembled Germany's. The end of direct Allied administration saw Germany's MOF regain control overtax policymaking and then use the Allies' confiscatory levels of taxation to craft broad-based incentives for business. Corporate income-tax policy gave business investors the choice of either using their capital in ways the state preferred, or paying high rates of tax. Because bureaucrats generally enjoy a greater degree of autonomy than politicians from the lobbying of specific interests and firms, in both Germany and Japan they were generally able to craft more broadly based tax measures than was characteristic of, say, Anglo-America. 9 Further evidence of this point is provided by Daniel Okimoto's comparison of industry-oriented tax policymaking in the US and Japan. Okimoto notes that 8 Ronald King,"From Redistributive to Hegemonic Logic: The Transformation of American Tax Poli t ics, 1894-1963," in Politics and Society, Vol. 12 , No. 1, 1983. 177 studies show a greater variation of tax expenditures in the US as well as most European countries. The reason is the political might of "old-line industries," whose lobbying wins them tax breaks that effectively put the tax burden "onto the shoulders of the most efficient and promising sectors, a perverse reward for efficiency and growth." By contrast, Japan's relatively weak legislature and powerful bureaucracy shape a tax policymaking process that limits the potential for economically irrational business taxation. 1 0 MOF's Initiation of Special Measures According to Chalmers Johnson, postwar Japan's tax expenditure process began when Minister of Finance Ikeda Hayato persuaded the MOF bureaucracy to accept preferences as a replacement for the direct subsidies that the Japanese Government had been dispensing to business since the end of the war. 1 1 In other words, Johnson (who is generally labelled a statist) suggests MOF was reluctant to use tax expenditures, and that they instead emerged as a result of pressure from politicians. But another, more persuasive account is found in interviews with former Japanese bureaucrats who worked in the MOF Tax Bureau at the time. The interviews reveal that in 1951 Ikeda was obsessed with increasing the CIT rate from 35 percent to 42 percent in 9 See , for example, the d iscussion of Germany in Carolyn Webber and Aaron Wildavsky, A History of Taxation in the Western World. New York: Simon and Schuster, 1986. 1 0 Daniel Okimoto, Between MITI and the Market. Stanford: Stanford University Press , 1989, pp. 86-90 includes comparative notes on the incidence of corporate tax expenditures in the US and Japan. The direct quotes are from ibid. p. 87. 178 order to tax some of the profits from the Korean War boom and thus overcome Japan's chronic postwar revenue shortages. These revenue shortages were causing serious problems during the annual budget-setting process, especially because a lack of funds imposes difficult trade-offs on politicians' natural inclination to disburse benefits to client groups. 1 2 MOF Tax Bureau chief Hirata sought to counter Ikeda's tax hike with the introduction of special tax measures aimed at keeping the corporate burden from increasing. This sequence of events "cleared the road" for tax expenditures, especially the reintroduction of special tax measures enacted during the war but eliminated by Shoup and previous US-inspired reforms. Ikeda found himself unable to oppose the tax expenditures, which of course reduced corporate tax revenues, because the influential business community strongly supported them. 1 3 11MITI and the Japanese Miracle. Stanford: Stanford University Press, 1982, pp. 232-34. 1 2 This fact is evident, for instance, in the struggles between LDP politicians and the M O F Tax Bureau group in GTAC. From the mid-1950s and well into the 1960s, there were frequent disputes between these groups over how much of the natural revenue increase from the Personal Income Tax to devote to increased expenditures (the politicians' choice) and how much to use for increasing tax exemptions and other forms of tax reduction (the bureaucrats' preference). On some occas ions, the squabbling became public, as in 1957 due to the enormous surge in tax revenues; see Takeda Takao and Hayashi Takehisa, "Zaisei Se isaku no Keisei Katei" [The Evolution of Fiscal Policy],in Gendai Nihon no Zaisei Kinyuu: Showa 30-Nendai [Finance in Modern Japan: 1955-1965]. Tokyo: Tokyo Daigaku Shuppankai, 1978, pp. 244-46. 1 3 Ito Mitsuharu, Sengo Sangyoushi e no Shougen: Sangyou Seisaku [Oral History of Postwar Industry: Industrial Policymaking]. Tokyo: Mainichi Shimbunsha, 1977, pp. 19-26. Confirmation of the interview remarks is seen in the article by Minomatsu Izumi, the then head of the MOFTax Sys tem Department of the Tax Bureau, in "Houjinzei Shikou Kisoku no Kaisei" [Revisions to the Regulations for the Corporate Income Tax], in Kinyuu Zaisei Jijou, Volume 3, No. 6 February 4, 1952. 179 Indeed, the precedent set, tax expenditures were enacted in a virtual flood. Major corporate lobby groups, beginning with the well-known Keidanren, pressed the LDP , the MOF Tax Bureau (especially via the GTAC), and MITI for tax expenditures. 1 4 The process of allocating tax breaks involved annual negotiations between MOF and MITI, in which a global limit was placed on the yen amounts of measures to be awarded for a given year. Then MITI's various internal divisions and bureaus would "vie with one another to win special tax provisions for industries under their jurisdication." S ince the demand for special tax measures exceeded supply, MITI was placed in an enviably influential position of distributing benefits to the corporate sphere. 1 5 The institutionalization of this process made MITI into a stalwart champion of tax expenditures, as they helped the ministry channel resources into areas that it deemed to deserve them. MITI's institutional interest in fiscal politics thus 1 4 A cursory look at the business journals of the period shows a seemingly interminable argument for tax expenditures. In the journal Keizai Rengou, for example, one representative article focuses on the Korean War Corporate Income Tax increase and asks for tax relief through special measures and, if possible, reduced tax rates; see Keizai Dantai Rengoukai, "Toumen no Zeisei Kihon Se isaku ni Kansuru Iken" [Opinions Concerning a Temporary Basic Tax Policy], December, 1952, pp. 24-17. More generally, statements in journals and the press by business leaders, bureaucrats, and others about the need to reformthe tax sys tem so as to encourage capital accumulation are striking in their uniformity. Indeed, though the Shoup reforms were guided in part by a desire to bolster incentives for business, much of the sustained attack on Shoup centred on its alleged neglect of the needs of capital. On these issues, see "Genzei oyobi ze isei kaisei no mondaiten: Shaupu zeisei takei no houkai ka" [Issues in Tax Reduction and Reform: The End of the Shoup System?], in Kinnyu Zaisei Jijou, Number 34, Volume 2, August 20, 1951, p. 7; and on the criticisms of Shoup, see Martin Bronfenbrenner and Kichiro Kogiku, "The Aftermath of the Shoup Reforms" (Parts 1 and 2), op. cit. 1 5 Okimoto, op. cit., pp. 88-89, argues that the work of making trade-offs among the requests from industries was performed in MITI. Specif ically, "The Business Behavior Division {Kigyou Koudouka) of the Industrial Policy Bureau (Sangyou Seisakukyoku) decides which industries deserve to receive tax exemptions, and in what amounts, after consulting with all the sector-specif ic divisions and bureaus and huddling with the Accounting Division (Kaikeika) and the General Coordinating Division (Soumuka) of the Minister's Secretariat {Daijin Kanbou). 180 came to centre on advocating increased tax expenditures for its corporate clients. Indeed, in a shocking display of economic heresy for the contemporary neoliberal mind, the ministry even opposed overall cuts in corporate tax rates as an inappropriate means to achieve the goal of capital accumulation. 1 6 By contrast, MOF's Tax Bureau and its associated academics soon became perturbed by the erosion of the CIT base through the 1950s. In the late 1950s, the academics, former bureaucrats, and labour-affiliated members on the GTAC began publicly to voice their disapproval of special measures. 1 7 Some years after the emergence of GTAC's concern over tax expenditures, large corporations gradually began to lose interest in special measures and instead came to prefer outright tax cuts. The latter trend is reflected in Table 6.4, whose columns on corporate income tax "Rates" and "S.T.M." (Special Tax Measures) display the revenue changes due to adjustments made in CIT rates and associated special tax measures from 1950 to 1974. 1 6 A typical example is seen in MITI's arguments before the G T A C in August of 1956; "Tsusanshou, Zeisei Kaisei Iken o Torimatome" [MITI's Opinions on Tax Revision], in Kin'yuuZaisei Jijou, Volume 7, No. 34, August 27, 1956, pp. 10. Note that state act ivism via the deployment of tax expenditures was hardly unique to postwar Japan, and was one of the main characteristics of advanced industrial states that "supply side" and other neoliberal political movements have sought -- with much s u c c e s s -- to extirpate; see the comments on this in Steinmo, Taxation and Democracy, op. cit. 1 7 Takeda Takao and Hayashi Takehisa, "Zaisei Se isaku no Keisei Katei" [The Evolution of Fiscal Policy], op. cit., pp. 256 ff. and Joseph Pechman and Keimei Kaizuka, "Taxation," in Asia's New Giant: How the Japanese Economy Works. Ed. by Hugh Patrick and Henry Rosovsky. Washington: The Brookings Institution, 1976, p. 329-30. 181 Table 6.4: Estimated Income Tax Changes in Japan, by Type. 1950-74 (Billions of Yen) Corpora te Income T a x Persona l Income T a x Fiscal Year Change Rates S.T.M. Change Exemptions Rates S .T.M. 1950-53 -25 31 -56 -386 -272 -86 -28 1954 -3 0 -3 -31 -29 0 -2 1355 -12 -14 2 -53 -23 -13 -18 1956 14 0 14 -23 -23 6 6 1957 22 -2 -24 -110 -40 -85 15 1958 -22 -20 -2 -6 6 6 -6 1959 -4 0 -4 -23 -28 -12 17 1960 0 0 0 0 0 0 0 1961 -40 0 -46 -56 -38 -23 5 1962 -1 0 -1 -50 -25 -23 -2 1963 13 0 13 -67 -32 0 -35 1964 -59 -5 -54 -75 -66 6 -8 1965 -57 -28 -28 -65 -92 6 26 1966 -99 -50 -49 -158 -101 -53 -4 1967 -30 6 -30 -93 -142 11 38 1966 0 6 6 -125 -135 -11 6 1969 2 0 2 -183 -142 -41 0 1970 75 97 -22 -289 -173 - 131 15 1971 12 6 12 -415 -286 -167 -22 1972 31 0 31 -32 0 0 -32 1973 27 0 27 -375 -335 6 -40 1974 352 424 72 -1,783 -1,467 -266 -56 Note: S.T.M.= "Special Tax Measures." Source: Joseph Pechman and Keimei Kaizuka, "Taxation," in Asia's NewGiant: Howthe Japanese Economy Works. Ed. by Hugh Patrick and Henry Rosovsky. Washington: The Brookings Institution, 1976, p. 329-30. 182 Table 6.4 also shows that rate changes to the CIT per se were generally infrequent during the early postwar period and the era of exceptional economic growth. Special tax measures were, by contrast, more frequently applied. However, the late 1960s saw the emergence of a full-scale rollback of tax expenditures for large corporations -- which had drawn the lion's share of the benefits up to then -- and a shift towards introducing special tax measures for small and medium-sized companies. This shift towards tax expenditures for small and medium-sized companies is seen in public finance circles as in part a consequence of the growing power of L D P zoku politicians and the LDPTAC in tax policymaking as well as large Japanese firms' declining interest in tax breaks. The latter benefited less from broad-based special measures as the economy slowed, and they also began increasingly to chafe at MITI's administrative guidance. The late 1960s thus marks the close of the era of large-scale and relatively uniform tax exemptions, the hallmark of bureaucratic tax incentives. Tax expenditures for the small and medium-sized sectors of Japanese business have since been implemented in fairly large numbers, but their combined effect is far less fiscally significant than was the case in the heyday of tax expenditures for big business. In part, this is because the tax breaks themselves are less comprehensive, but it also reflects the fact that well over half of small and medium-sized businesses are technically unprofitable and 183 thus exempt from income taxation. Figures for 1994 -- admittedly the peak year of the Heisei recession -- show that 62.7 percent of Japan's close to 2.37 million firms were in the red, with the average since 1970 being 51.8 percent. 1 8 The available tax expenditures are of little or no use to firms without taxable income, and their large numbers greatly limit the loopholes' capacity to erode national tax revenues. Thus the aggregate revenue loss from all special tax measures in 1993 was only 520 billion yen, or about 3.3 per cent of total corporate income tax revenues for the same year. This amount contrasts strongly with the peak loss of 28.6 per cent of corporate tax revenues recorded in 1955. 1 9 However, this does not mean that large corporate lobbies in Japan simply acquiesced to the closing of special tax measures in the 1960s. In fact, they and their sponsor MITI fought very hard to prevent such closures even when, as was the case with export promotion tax incentives, they violated Japan's international commitment according to trade agreements. One of the major sources of opposition was the concern that eliminating tax loopholes meant, in the absense of rate reductions, an effective tax increase. Hence, the lobbying ' 8 The data can be found on p. 81 of "Chihouzei Kankei Shiryou Handobukku, Heisei 8-nen, 10-gatsu" [Handbook of Materials on Local Taxation, October 1996], Tokyo: Chihou Zaimukyoukai, 1996. 1 9 Table 4, "Revenue Effects of Special Taxation Measures for Corporations," in An Outline of Japanese Taxes, 1993. Tokyo: Tax Bureau, Ministry of Rnance, 1993, p. 78. The table's descriptions of the tax expenditures reveal that many of them specifically designate small and medium-sized corporations. Figures for revenue losses in the early postwar period can be found in Ryutaro Kojima, "Sengo Nihon no Zeisei to Shihon Chikuseki , " in Zaisei Seisaku. Ed. by Sei Fujita. Tokyo: Nihon Keizai Shimbunsha, 184 was often focused on trading off the loss of tax preferences for cuts in the overall corporate rate. 2 0 To sum up at this point, then, tax politics in postwar Japan saw the general rejection of Shoup's insistence on economic neutrality with respect to the structure of income taxation. Indeed, the reforms had hardly been implemented before the rollback began, driven by bureaucratic concern that the level of corporate income taxation not hamper the still-tentative economic recovery during the boom of the Korean War. Where Finance Minister Ikeda sought to increase the CIT rate and thus solve in part the crisis of financing the national budget, the MOF Tax Bureau leadership countered his move with a series of special tax measures. The special measures rapidly mounted up, though, as a host of corporate lobbies were successful in demanding equal treatment. Moreover, the process of allocation was managed through MITI, which imposed an industrial policy rationale and broad sectoral equity on tax expenditures. As a result, the erosion of the CIT base proceeded apace through the application of special tax measures reminiscent of those instituted during Japan's crisis of 1973, Table 5.6, "Sozei Tokubetsu Sochi ni yoru Zeishuu no Genshuu Suiteigaku (Showa 25-40-Nendo)." 2 0 Takeda Takao and Hayashi Takehisa, "Zaisei Se isaku no Keisei Katei," in Gendai Nihon no Zaisei Kinyuu: Showa 30-Nendai. Tokyo: Tokyo Daigaku Shuppankai, 1978, pp. 256 ff. Note that the logic of the trade-off became very common in the tax reforms implemented among developed countries in the 1980s. "Base-broadening," as the elimination of tax expenditures came to be called, and rate reduction became very popular, even visionary, in the wake of the 1986 U.S. tax reform; see some of the c a s e s in 185 production in the midst of the Pacific War. 2 1 However, the erosion of the CIT's tax base could not continue indefinitely. One of the limits was the loss of a convincing policy rationale as the award of tax expenditures became an output of the machinery of interest-group politics. As Okimoto's comparison of tax expenditures in Japan and the US suggests, Japan's process of allocation remained strongly bureaucratic, meaning tax expenditures applied to whole sectors of industry rather than the much more selective outcomes characteristic of the U S . But Japan's broad-based loopholes also led to a stronger erosion of the tax base that, taken in tandem with the decline of persuasive industrial-policy goals, contributed greatly to growing opposition within the MOF Tax Bureau and the GTAC. A second limit was imposed by the fact that the cuts in individual income tax burdens were even more deep than those made to the levy on corporate income. Table 6.4 displays this comparatively, and leaves no doubt that the largest reductions were visited on the PIT. The virtually annual trimming of the natural revenue increase from the PIT, even though supported within MOF as a restriction on the growth of the state, imposed in itself an effective limit on how much revenue could be allowed to leak out of the sphere of corporate World Tax Reform: Case Studies of Developed and Developing Countries. Ed. by Michael Boskin and Charles E. McLure Jr. San Francisco: International Centre for Economic Growth, 1990. 2 1 On this, see Jinno Naohiko, "Sozei Se isaku to Keizai Tousei (2): Taiheiyou Sensouki no Soze i Se isaku" [Tax Policy and Economic Control: Tax Policy in the Pacif ic War], in Keizaigaku Zasshi, Volume 84, No. 2, July, 1983. 186 income taxation. The Tax Bureau's stance was the dominant influence on corporate tax policymaking in that period, especially as manifested in the authoritative reports of the MOF-dominated GTAC. Thus, the stability in the CIT rates relative to those of the PIT, as we can see from table 6.4, throughout the postwar period reflects in large part the increasing hesitancy in the Tax Bureau. Moreover, table 6.4 shows that the PIT cuts became increasingly pronounced in the 1970s, whereas the CIT became the target of incremental upward adjustments. This upward movement in the CIT -- both in terms of the statutory tax rates and the effective burden -- peaked in 1984, and was unambiguously a consequence of deep cuts in the regime of personal income taxation. 2 2 The Role of MOHA and Local Taxation Alongside MOF, an important source of opposition to cuts in the CIT was MOHA and its predecessors. The postwar record shows MOHA resisted efforts by the LDP, MITI, and business lobbies to get the corporate tax burden reduced. 2 3 The reason for this opposition.is that corporate taxes are important to the fiscal health of the local authorities. As was pointed out in the earlier section, a major function of the 1940 tax reform was to provide the revenues 2 2 Miyashita Tadayasu and Miyawaki Atsushi, Zaisei Shisutemu Kaikaku [Reform of the Fiscal System], Tokyo: Nihon Keizai Shinbunsha, 1995, p. 93. 187 for redistribution and help cope with the recurrent fiscal crises of the local authorities. Indeed, MOHA's fiscal interest in this regard likely exceeds that of its predecessor Home Ministry. The prewar Home Ministry was effectively unable to control the revenue decisions of the local authorities once they got into fiscal crisis, because the ministry lacked a redistributive structure through which to overcome the problem of widely different tax capacities among its subordinate administrative units. But the Home Ministry had far more potent powers than MOHA through its ability to appoint prefectural governors and use them as instruments of national policy. Postwar reforms and the breakup of the Home Ministry eradicated much of the legal and administrative framework for intervention in local affairs, leaving the intergovernmental fiscal structure as the strongest alternative. 2 4 Hence the fact that MOHA is a channel for redistributing 32 percent of PIT revenues to the local authorities gives it a powerful interest in maintaining that tax base. 2 5 But MOHA's interest in the CIT base has even deeper material roots than this, due to the revisions of the Shoup reforms and especially the passive legislative euthanasia of the VAT. Shoup's plan to implement a VAT at 2 3 Ohtake Hideo, "Hatoyama, Kishi Jidai no Chiisai Seifu Ron" [The Small State Debate of the Hatoyama and Kishi Years], op. cit. 2 4 Jinno Naohiko, "Nihongata Zei, Zaisei Shisutemu," op. cit, pp. 240-41. 2 5 This redistribution is v ia the Local Allocation Tax, as was described in earlier chapters of this dissertation. 188 the prefectural level in order to provide a steady source of independent revenues for the local authorities was the focus of strong opposition until the tax was repealed in 1954. The elimination of the VAT forced the fiscal authorities to turn back to the Enterprise Tax -- a tax on profits -- and institute a prefectural inhabitants' tax in order to maintain prefectural revenues. As a result, there are three taxes on corporate activities atthe local level in Japan: the already noted Enterprise Tax, which in 1994 supplied 30.9 percent of prefectural tax revenues; and the prefectural and municipal local local inhabitants' taxes, which accounted, respectively, for about 5.5 percent of prefectural revenues and 11.7 percent of municipal-level revenues. 2 6 In consequence, Japan's regime of corporate income taxation has an unusually heavy weight of business taxation at the local level. Table 6.5 shows that a much higher percentage of corporate income, measured in terms of G D P , is taxed at the local level in Japan than is the case in a representative sample of other advanced industrial countries. This alone would be sufficient to make MOHA wary of cuts to corporate taxes: the municipal corporation tax is levied as a percentage of the national tax, and therefore the revenues derived from it decline with reductions in the latter's rate. 2 7 The f iscal history is d iscussed in Kurt Steiner, Local Government in Japan, op. cit., pp. 264-66. The 1994 tax data are from p. 8 of "Chihouzei Kankei Shiryou Handobukku, Heisei 8-nen, 10-gatsu" [Handbook of Materials on Local Taxation, October 1996], Tokyo: Chihou Zaimukyoukai, 1996. According to the M O F Tax Bureau's An Outline of Japanese Taxes, op. cit., pp. 244-45, the municipal level's standard tax rate for corporations is 12.3 per cent of the national corporation tax, and -- as is specified by MOHA's tax controls ~ "should not exceed 14.7 per cent in any case . " 189 Table 6.5: Corporate Income Taxation at National and Local Levels for Selected Countries as Percentage of GDP, 1991. Country Type and Name Corporate Income Taxation National Local Federal States Germany 0.7 1.0 USA 1.8 0.4 Unitary States France 2.0 0.0 Japan 3.7 2.4 Sweden 1.6 0.0 United Kingdom 3.2 0.0 Source: revised version of the table in Mutai Shunsuke, "Nozomashii Chihouzei Taikei," mChihou Zaisei no Hatten to Aratana Tenkai [Past and Future Developments in Local Finance]. Ed. by Yuasa Toshio. Tokyo: Gyousei, 1995, pp. 322-3. On top of that consideration, though, is the 32 percent of the national-level CIT earmarked for the Local Allocation Tax. Tax expenditure-induced erosion of the CIT base costs MOHA funds that it transfers to the fiscally weaker local authorities, because the earmarking of the CIT in the Local Allocation Tax is a fixed proportion of the overall revenues, not an absolute sum of revenues. Hence, one sees cooperation between MOF and MOHA towards eliminating 190 special tax measures that undermine the base of the CIT. However, though MOF-MOHA cooperate in certain areas of potential mutual benefit, their split jurisdiction on the corporate tax field also gives rise to conflict. MOF's turf is the CIT, whereas MOHA's is the Enterprise Tax and the prefectural and municipal corporate inhabitants' taxes. The conflict between MOF and MOHA centres on the maintenance of the respective tax bases under their authority, and is implicated in the current anomaly of Japan's comparatively high corporate taxation in spite of having had decades of conservative government. As Table 6.6 reveals, Japan's corporate tax rates exceed those in the U S and the UK, though not the German rates. The argument that intergovernmental fiscal politics exerts a strong influence on Japan's effective corporate income tax rates is, in this example, given support from a comparative perspective. German local governments' heavy reliance on corporate income taxation is a legacy of the Prussian state's enactment of the income tax as a national tax and the consequent transfer of property and business taxes -- taxes on profit --to the local authorities. The business tax exposes the local authorities in Germany, as is the case in Japan, to the fiscal uncertainty that derives from the fluctuations of the business cycle. However, local governments in Germany 2 8 See , for example, "Sozei tokubetsu sochi no seiri, gourika, taishou no shiborikomi de icchi: zeifu zeichou ga shuuchuu tougi" [Discussions in G T A C centre on rationalizing special tax measures], in Mainichi Shinbun, December 3, 1994, p. 11. The article d iscusses ajoint submission by M O F a n d MOHA for eliminating CIT tax expenditures that the ministries believe have lost their public policy rationale. 191 strongly defend the tax because of its importance to local finance and the fact that they are in autonomous control of rate setting. 2 9 Table 6.6: Effective Corporate Tax Rates Among 4 Countries. Selected Years. Japan US UK Germany 1980 49.47 50.86 52.00 56.52 1985 52.92 45.00 45.00 56.52 1990 51.04 35.00 35.00 56.52 1993 49.98 33.00 33.00 52.87 1995 49.98 33.00 33.00 48.26 Source: Table 5.1 in Mutai Shunsuke, etai.,JichitaiZaiseiKiban no Kakuritsu, op. cit, pp. 212. Increasingly significant for Japan is that its level of corporate taxation is even more out of line with its actual and emerging economic competitors in East and Southeast Asia. For example, the weight of taxes on business in Hong Kong is a full 21 points less than that in Japan, a fact that contributes to growing Japanese concerns about the potential for "hollowing out" of the domestic industrial base. Indeed, though the focus of attention remains on tax differences with North America and Europe, attention is turning in the 1990s to the even greater gaps with Asia, a trend that is probably linked to Japan's shift towards higher levels of trade and economic linkages with its Asian George Milbradt, "Business Taxes and Local Autonomy in Germany," in Local Business Taxes in Britain andGermany. Ed. by R . J . Bennett and H. Zimmerman. London: Anglo-German Foundation, 1986, pp. 95-103. 192 neighbours. Comparisons of Current Business Tax Regimes It is somewhat misleading to express tax burdens in simple lump-sum statistics, however, as this approach tends to give little idea of the actual tax circumstances that specific industries face in various constituencies. Fortunately, a study 3 1 is available that compares the tax positions of selected Japanese companies were they operating in a variety of foreign tax jurisdictions. This kind of detailed empirical study provides a much better indication of the actual comparative tax position of Japanese firms versus their counterparts in trading partners and in regional sites that are often indicated as areas where hollowed out employment and production might migrate. The data reflect what a Japanese manager in a given corporate enterprise must take into account when considering a move offshore, and thus indicate whether competitors overseas have real cost advantages through the tax "Houjinzeiritsu sage kentou: sangyou 'kuudouka' soshi nerau" [Studying the reduction of the corporate tax rate: goal is to avoid hollowing out of industry], Nihon Keizai Shinbun, May 3, 1995, p. 3. See also "Zeisei Kaikaku no Ji tsugen o" [Towards Realizing Tax Reform], Tokyo: Keizai Kouhou Sentaa, July, 1996, p. 6. The study was done by XheZaidan Houjin Kigyou Katsuryoku Kenkyuujo, of Tokyo, in July of 1995 and on the basis of published materials from the corporations chosen. Profits and expenses of selected Japanese corporations were calculated with the actual tax measures of the foreign countries researched. In other words, the study investigated the tax obligations that -- to use one of the target corporations -- Matsushita Electronics would have in Indonesia, Thailand, the United States, and other tax jurisdictions. Note that the years used are 1991 for Japan; 1994 for the U.S. ; 1988 for France, Germany, and the U.K.; and 1992 for the Asian countries. Hence Japan's tax competit iveness vis-a-vis the European countries is, if anything, exaggerated due to the fact that -- as is evident from the table on corporate tax rates -- Japan's rates have remained relatively stable in the 1990s while those of other countries have been reduced. A mimeographed copy of the study is in the present author's possess ion . 193 system. Table 6.6 presents selected data pertaining to Toyota Corporation, a well-known Japanese car-maker (the assumed profit on sales for each country is 1, 150, 002 million yen). Table 6.7: Toyota Corporation's Tax Position. Various Countries. (Units: Million Yen) Profits and Taxes/ Country Social Security National CIT T o c a l c T T " Property and Con. Taxes Total Taxes Levied Total Social Security, Tax Burden Japan 33, 351 55,726 27,6^4 13,289 96,471 129,822 France 90, 619 62,168 0 16,864 78,972 169,591 Germany 69,785 49,052 21,170 19,285 89,507 159,292 U.K. 27,643 94,108 0 3,546 97,648 125,292 U.S. 44,341 79,345 18, 215 16,563 114,123 158,464 China 18,548 82,996 14 0 83, 010 101,558 Indonesia 18,548 39,159 6 2o3 39,362 61,156 Malaysia 18,548 98,728 27 6 98,755 117,363 S. Korea 18,548 89,456 0 1,612 97,777 116,325 Singapore 18,548 82,466 6 159 82,625 101, 173 Taiwan 18,548 52,428 0 1,323 53,751 72,299 Thailand 18, 548 78, 516 0 8,297 86, 814 105, 361 Note: "CIT" = Corporate Income Tax; "Con Tax" = Consumption Tax. The table shows that the combined tax burden for Toyota is not particularly high in Japan relative to hypothetical competitors, particularly if one includes the often overlooked burden of social security. Including the latter indicates that European firms labour under particularly high costs. 194 Table 6.8: The Tax Position for Matsushita Electronics, Various Countries. (Units: Million Yen) Profits and Taxes/ Country Social Security National CIT TSSTNT' Property and Con. Taxes I otal Taxes Levied Total Social Security, Tax Burden Japan 39,178 34,698 17,952 12,348 63, 998 103, -176 i-ranee 106, 451 16, 591 0 10,928 21,519 127, 970 Germany 81,977 26, 317 14,898 14,896 59,111 141,088 U.K. 32,473 44, 178 0 3,507 47,685 80,158 U.S. 52,088 26,411 7, 815 16,416 53,636 165,725 China 21, 788 36,267 6 0 36,452 61,246 Indonesia 21,788 39,159 0 203 39,362 61,150 Malaysia 21,788 43,437 6 0 43,727 65,515 S. Korea 21,783 46,539 3,494 1,306 51,389 73,177 Singapore 21,788 40, 072 0 158 40,230 62, 018 Taiwan 21,788 16,436 6 237 16,723 41,511 Thailand 21,788 37,937 6 8,437 46,424 63,212 Note: "CIT" = Corporate Income Tax; "Con Tax" = Consumption Tax. The electronics industry is represented in Table 6.8 by Japan's Matsushita Electronics (the assumed profit on sales for each country is 1,143, 077 million yen). Note in the table -- based on the same pattern as the one above -- that the tax burden on Matsushita in Japan compares unfavourably with that in trading partners and other countries in the region. Once the social security burden is taken into account European firms are again put at a clear disadvantage, but Japanese electronics firms' competitors tend to be Asian companies or fellow nationals who have set up offshore production sites in the Asian region for import into Japan. This suggests that the generally lower tax 195 burden in the region is a major problem for Japan, and that concerns about tax-driven hollowing out of this industry are not wild exaggerations or merely corporate lobbies' self-serving propaganda. Table 6.9: The Tax Position For Representative Steel Companies, Various Countries. (Units: Million Yen) Profits and Taxes/ Country Social Security National CIT "Local ClY Property Taxes Total Taxes Levied Total Social Security, Tax Burden Japan 23,437 28, 514 15, &41 29,335' yn, 796 95,227 France 63,682 16,849 0 21,070 37, 919 101,600 Germany 49,641 17, 126 8,466 15,361 38,955 87,993 U.K. 19,426 44,599 6 3,662 52,682 72,168 U.S. 31,160 27,902 9,896 37, 816 75, 614 106, 775 China 13,654 48,625 6 6 49,662 62,656 Indonesia 13,654 58,629 6 261 58,896 51,924 Malaysia 13,034 38,442 371 0 38, 813 51,847 S. Korea 13,654 39,308 2,948 1, 186 45,456 56,476 Singapore 13,034 56,525 6 365 56,888 49,922 Taiwan 13,034 17,454 0 648 18,101 31, 136 Thailand 13,034 41,095 0 20,034 61,129 74,163 Note: "Corp. Inc. Tax" = Corporate Income Tax. Table 6.9 is based on the average financial data for two Japanese steel-makers (the assumed profit on sales for each country is 747, 593 million yen). It shows that the burden of corporate income and other taxes is generally quite heavy in Japan compared to that in other countries. Moreover, this case also 196 plainly shows the tax advantage possessed by Japan's closest competitors in South Korea and elsewhere in the region. Therefore, these data would appear to lend further strong support to arguments that Japanese producers' comparatively high tax burdens afford potentially strong incentives to relocate to less costly offshore sites. Corporate Taxation and the Fiscal Crisis An important backdrop to the above tables and discussion is the fiscal crisis of the early 1970s. In this period, revenue declines from earlier income-tax cuts, plus the post-oil shock slowdown in the Japanese economy worked in tandem with rapid increases in welfare state spending. One consequence of the ensuing haemorrhage of deficit spending was that greater pressure came to be placed on drawing revenues from the CIT. Japan's slide into fiscal crisis in the 1970s was surprisingly steep, with bond issues climbing from 12.4 percent of total expenditures in 1971 to 34.7 percent in 1979. 3 2 In 1979 Prime Minister Ohira broached MOF's initial plan for introducing a VAT-type tax to stop the snowballing deficit, but the scheme rankled the LDP 's core supporters and cost the party dearly in the ensuing general election. With the VAT thus rendered taboo, the focus of policymaking turned to proposals for administrative reform. At the same time, however, MOF continued to hold the 3 2 From Table 11 -2 in Public Finance and the Tax System in Japan. Tokyo: FAIR, 1992. 197 line on tax expenditures as well as introduce effective tax increases on corporations. Examples of the latter were a 2 percent rate increase in 1981 and a 1.3 percent corporate surtax, introduced ostensibly for two years in 1984 in order to cover the cost of reductions in the income tax; the surtax boosted the national-level income tax rate for large companies from 42 percent to 43.3 percent. Though Keidanren and other industrial lobbies mobilized in opposition to these measures, their efforts were strikingly unsuccessful. 3 3 The business community was therefore bearing much of the cost of fiscal reconstruction. Or to put it a way that highlights the comparatively surprising angle more clearly, corporate Japan was picking up much of the tab for the expansion of the welfare state whereas the general trend elsewhere - and certainly in Anglo-America -- was to reduce the burden on corporate income. Table 6.3 showed, for example, that between 1965 to 1990 Japan increased its taxation of corporate income (as a percentage of GDP) from 4.1 percent to 6.8 percent, while the O E C D America burden dropped from 3.9 percent to 2.3 percent. Admittedly, the O E C D Europe burden rose over those years from 1.8 percent to 2.6 percent of G D P ; but the overall level was still relatively low and, more important, Europe's economic heartland, Germany, decreased its corporate tax burden from 2.5 percent to 1.8 percent in the same period. 3 3 On closing loopholes, see Wada Yahigashi, chapter 8 "Sozei Tokubetsu Sochi no 'Seiri, Gourika'" [Reforming Special Tax Measures], in Sozei Seisaku no Shintenkai: Zaisei Kaikaku to Zeisei Keikaku [Changes in Tax Policymaking: Rsca l Reform], Tokyo: Bunshindo, 1986. M O F s "war" with Keidanren over the Corporate Income Tax burden was a staple in the press through the mid-1980s; a representative example is "Ishida Keidanren Zeisei l inchou, Ookurashou to no Zeironsou de Kyuusen 198 By contrast, Japan's taxation of personal income (as a percent of GDP) remained well under the levels in most other O E C D states in spite of fairly rapid increases. Thus, while Japan increased its taxation of personal income between 1965 and 1990 from 4.0 percent to 8.4 percent, the O E C D America levels went from 6.8 percent to 12.5 percent and the O E C D overall saw an increase from 7.1 percent to 11.4 percent. Japan's tax burdens on corporate income and the fact that they increased over the years is unusual. It may be that the country had little alternative, as over the same years the total revenue take as a percent of G D P (excluding social security) outpaced the O E C D average by increasing from 14.3 percent to 22.2 percent versus the latter's more leisurely upward movement of 21.5 percent to 28 percent. 3 4 Revenues had to come from somewhere, in an era marked by continuing party competition over cuts; in the PIT and the stymied plan to introduce a broad-based consumption tax. Corporate Taxes and Intergovernmental Fiscal Politics Indeed, as noted earlier, MOF had already been effectively increasing corporate taxes by closing loopholes its personnel had helped initiate in the Yaburu Hihan no Isshi: "Toshi Genzen Zehi Hitsuyou" [Keidanren Tax Committee Head, Ishida, Breaks Truce in Tax War with MOF], Nihon Keizai Shimbun, October 12, 1984, p. 27. 199 1950s. For intergovernmental fiscal politics, the upshot of MOF's move against special tax measures is that MOF and MOHA developed generally common interests concerning the CIT field. The shared jurisdiction remains, as we will see below, a potential source of friction that has of late ignited a full-blown case of interministerial tax competition. But there is no evidence that MOHA has sought, unlike MITI, to battle with MOF for tax expenditures and other erosions to the CIT base. On the contrary, MOHA opposed cuts to the tax regime; and quite rationally, as the 1960s saw the protracted economic miracle get solidly under way and, to a large extent, lift local governments out of their long postwar fiscal crisis. MOHA certainly allows temporary adjustments to local corporate tax rates when specific local authorities are in financial difficulty. But shorn of public works and the other line-ministry activities that were part of the old Home Ministry's vast purview, MOHA has no good reasons to pursue tax expenditures. 3 5 That is to say, MOF and MOHA shared a positive-sum game on the CIT field until recent years. Increases in the CIT, whether through higher rates or base-broadening (i.e., closing loopholes), benefited both bureaucracies in their concern to maintain revenues. The game depended on room for high rates, The above data are from tables 4, 10, and 12 in Revenue Statistics of OECD Member Countries, 1965-95. Paris: O E C D , 1996. In fact, MOHA seeks the eimination of exemptions that in the past were introduced in the Entrprise Tax base. These exemption, or tokureisochi, include tax reductions for broadcasters, publishers, and other mass media outlets as well as educational film producers; see, "Rainendo zeisei kaikaku, yuuisochi o minaoshi" [A review of preferences for next year's tax reform], in Sankei Shinbun, December3, 1993, p. 11. 200 which was ample in the era of high growth and a largely closed domestic market. Japanese companies were able to pass the tax burden onto consumers. Moreover, the corporate tax regime provides a further example of MOHA's generally negative stance towards substantive local fiscal autonomy. This is because the ministry, and especially its Finance Bureau, pressures the prefectures not to raise the Enterprise Tax beyond the standard rate. MOHA does this in order to prevent increased local corporate tax revenues from causing automatic reductions in the amount of Local Allocation Tax funds from MOF. The MOF-MOHA Stalemate over Corporate Taxes However, cooperation between MOF and MOHA concerning the corporate tax field has in the 1990s turned into a zero-sum contest. Corporate tax cuts are virtually inevitable because of the above-noted concerns over strong tax competition from major industrialized and neighbouring Asian countries, competition which carries the threat of hollowing out of employment and industry. As was also pointed out earlier, these concerns have led Japanese tax policymakers to shift more towards the taxation of assets and consumption as opposed to income. MOF and MOHA are responding to this new context in a way that is consistent with the central thesis of this dissertation: both are hunkered down in the 201 trenches, employing tactics aimed at ensuring the burden of cuts is borne by the other agency's business tax regime. The ministries have, to this end, conducted extensive lobbying campaigns through various institutions, such as the GTAC, in order to impress their own preferred tax reform scenarios on the agenda. For example, MOF and MOHA have sought information on foreign tax strategies that would help legitimate their preferred approaches to tax reform in Japan. Hence, MOF's scenario relies on Germany's recent plans to cut local business taxes in order to revitalize the country's underperforming economy. 3 6 MOHA's strategy includes refuting MOF's arguments through advancing a counter-example, one that suits MOHA's organizational goals. As part of the process of preparing this strategy, in the summer of 1995 MOHA sent a team of personnel to study Michigan state's innovative Single Business Tax. 3 7 These and other examples were debated extensively before a special study group {kentoukai) of the GTAC, one composed of representatives of MOF and MOHA as well as other experts, during the fall of 1995. The object of the study group was to judge the quality of MOF and MOHA's respective arguments concerning the proper balance of national versus local taxation on the corporate and land tax 3 6 This information is according to interviews with G T A C members at the University of Tokyo during 1995. See also "Houjinzeiritsu sagekentou: sangyou 'kuudouka'soshi nerau" [Studying the reduction of the corporate tax rate: goal is to avoid hollowing out of industry], Nihon Keizai Shinbun, May 3, 1995, p. 3. On the German tax-reform plans, see "Sangyou kyouka e houjinzei keigen: doku no zeisei minaoshi honkakuka" [Corporate tax cuts for industrial revitalization: Germany's tax reform debate goes into high gear], Nihon Keizai Shinbun, February 16, 1996, p. 9 and "Germany: Is the model broken?, The Economist, May 4, 1996, pp. 17-9. 202 fields. The gravity of the situation can be appreciated by the fact that "this style of examination is an unprecedented measure for the GTAC" [seifu zeichou dewa zenrei no nai shingi sutairu to nam].38 The study group was unable to come to any definitive conclusions as a result of the hearings, and thus the issue was put off for further consideration -- to a special GTAC Subcommittee on Corporate Taxation headed by Professor Ishi Hiromitsu of Tokyo's Hitotsubashi University. The stalemate is politically noteworthy because, as we recall, there is a relatively high tax burden on Japanese firms and their political pressure on the issue is so intense. The data presented earlier in this case study certainly show that Japan's effective corporate tax rates and burdens greatly exceed Japanese companies' competitors in East Asia. And pressure for a lower overall tax burden from Japanese business has been evident since the early 1980s when MOF effectively shifted much of the cost of fiscal reconstruction to business, apparently as one way to convince the latter to support the institution of a VAT. But it would appear to be the case that MOF and MOHA's divergent fiscal interests are playing a large role in thwarting corporate pressure for an effective tax cut. 3 7 The Single Business Tax and the German local business tax are d iscussed in the MOHA publication, Jichitai Zaisei Kiso no Kakuritsu, op. cit, pp. 260-75. 3 8 See "Seifu Zeichou, Ookurashou, J ichishou to Chokusetsu Taiketsu e -- 'Tochi, ' 'Houjin' no Kentoukai de" [MOF-MOHA contest in the G T A C study group on land and corporations], Mainichi Shinbun, September 12, 1995, p. 9. 203 In other words, though the GTAC and its subcommittees generally serve to legitimate MOF's goals, the process in this case bogged down over the two ministries' split jurisdiction on the corporate tax field and their representation of larger fiscal interests. MOF stands for the maintenance of responsible budgeting in a period of repeated fiscal stimulus and demands for reduced income taxes. MOF is thus loath to concede to cuts in its array of revenue sources, especially the lucrative CIT. And MOHA, for its part, is charged with the fiscal stability of local governments, who rely heavily on the taxation of corporate income as well. This Gordian Knot of fiscal interests would not lend itself to a simple and decisive GTAC recommendation for reduced corporate income tax rates, in particular because of MOHA's obstructionism on the issue of revising the Enterprise Tax. MOHA's stalling tactics meant the Enterprise Tax could not be reviewed by the GTAC until 1998 at the earliest, and thus left MOF's corporate tax as the centre of attention for tax cuts. 3 9 Shortly after the October 20, 1996 general election, Nakamura Yoshio, the managing director of the main business lobby, Keidanren, became so frustrated with the process in the GTAC that he spoke out publicly on the issue. According to Nakamura, a major problem with the Japanese tax system generally, and corporate taxation in particular, is that the "authorities in charge of collecting taxes usually have the biggest say" in reform. He pointed out that a major avenue for the bureaucracy's influence is the advisory councils, and 3 9 " B a p p o n kaikaku sakiokuri tsuyomaru" [Strong possibility full-scale reform will be put off], Nihon Keizai 204 called on elected politicians to take a more decisive role in policymaking. Not surprisingly, his first task for them was that they cut back corporate tax burdens to levels prevailing in the United States. 4 0 Some ranking members of the LDP responded favourably to the call for reduced corporate taxes. Yamasaki Taku, the head of the L D P Policy Affairs Research Counci l , announced himself in support of a 5 percent cut in CIT rates, at a projected cost of 2 trillion yen. This policy statement appeared to throw a wrench into MOF's compromise strategy -- the CIT being left in the spotlight through MOHA's intransigence -- of insisting on revenue neutrality via closing loopholes in exchange for any such cuts. But MOF quickly shifted its arguments, and started pointing to the potential political costs of lowering corporate taxes when individuals' burdens were expected to increase by about 7 trillion yen in 1997, due to slated increases in the Consumption Tax and the end of the 3-year special income-tax cut enacted in 1994. 4 1 Moreover, MOF's position appeared to receive considerable backing from the LDP 's Tax Advisory Counci l (LDPTAC) head, Hayashi Yoshiro (himself a former MOF minister). Hayashi let it be known that he regarded securing tax cuts a difficult matter, because Japan's fiscal problems required there be compensating base-broadening (i.e., revenue neutrality) but closing loopholes would attract Shinbun, October 17, 1996, p. 3. 4 0 Nakamura Yoshio, "Give companies a break when it comes to taxes," Japan Times, November4, 1996, p. 8. 4 1 "5 % sage nara zaigen ha 2 chou en" [A 5 % cut would cost 2 trillion yen], Nihon Keizai Shinbun, November 9, 1996, p. 3. 205 strong political opposition from corporate lobbies and their their allied zoku giin.42 Shortly afterward, on October 26, the GTAC Subcommittee on Corporate Taxation turned in its report. Reflecting the stalemate in the GTAC between MOF and MOHA, the report essentially rejected business arguments for a tax cut by insisting on the need for revenue neutrality and highlighting the existence of 38 types of special measures that favour large firms. 4 3 This report was followed on the 18th of December by the release of both the GTAC and LDPTAC outlines for tax reform in 1997, neither of which called for a tax cut per se. But in line with the LDPTAC 's greater sensitivity than the GTAC to political interests, especially the possibility of a corporate backlash, the L D P T A C report noted that it was only putting off any business tax reform for an additional year. It would appear there was no other route to take, as in a December 3 statement to the Diet Prime Minister Hashimoto Ryutaro had already committed the governnment to revenue neutrality.4 4 Hence, Japan's heavy weight of corporate income taxation is due not only to "Houjinzei genzei nado shouten" [Focus on corporate income tax reduction and other items], Nihon Keizai Shinbun, November 16, 1996, p. 2. On the tax zokugiin, see "Zeisei kaikaku e ugoku zokugiin" [Policy tribes moving in on tax reform], Yomiuri Shinbun, November 13, 1996, p. 3. "Houjin kazei kaikaku houkoku: daikigyou no keneki ni mesu" [Report on reform of corporate taxation: a mess of vested interests for large firms], in Mainichi Shinbun, October 27, 1996, p. 5. On the report, see "Houjinzei, tetsukazu" [Corporate tax left untouched], Mainichi Shinbun, December 19, 1996, p. 9. For Hashimoto's statement, note "Yugamu futan: houjinzei, zeiritsu to hikiatekin" [Warped burdens: corporate tax rates and special allowances], Nihon Keizai Shinbun, December 5, 1996, p. 5. 206 economic peculiarities and the electoral logic that places a premium on personal income tax cuts, making corporate income taxation the ready candidate to provide substitute revenues. An additional strong influence is that both the local and central governments impose taxes on this field; and because of this split jurisdiction within the context of the 1940 fiscal system, neither MOF nor MOHA is ready to yield ground. 207 Chapter 7 The Property Tax Field The Land Value Tax was adopted when MOHA showed reluctance to reform the Fixed Assets Tax...But concerned at the centre's encroachment on land-holding taxes, MOHA hurriedly raised the Fixed Assets Tax assessment...and sought incremental tax increases in spite of the introduction of a burden-adjustment scheme and the decline in land prices. Because of this, anger with the Fixed Assets Tax has exploded. Kakinuma Shigeyoshi Shukan Toyo Keizai October 5, 1996 Land taxation in Japan has a long and intensely politicized history, as preceding discussions have indicated. This chapter explores the fiscal history in depth, and offers an explanation for why MOF and MOHA have in recent years escalated their conflict on this tax field. The particular significance of this argument is that bureaucratic fiscal competition over land taxation spurred what threatened to become a tax revolt and has strong potential to engender further fractious fiscal politics. Prewar Politics and the Land Tax The background to the current politics is found in the infamous land tax, the centrepiece of the early Meiji era (1868-1912) fiscal system. The tax grew increasingly inequitable over the years, bearing down far heavier on rural and 208 agricultural interests than on the growing ranks of modern urban and capitalist Japan. Taxation of land was also a perennial focus of party politics,1 especially as the guiding ethic of the prewar era, "rich country, strong army" (fukoku kyouhei),2 and its polar opposite emphasis on reducing the burden on the public (minryoku kyuuyou), came into conflict. In 1873, the shift from paying the land tax in kind (i.e., rice) to having to pay it with coin in fact brought forth peasant riots throughout Japan. Moreover, to a large extent, the effort to avoid being taxed to pay for the military reinforcement of the turn-of-the-century state saw Japanese business enter into politics as a major actor.3 Table 7.1 shows that the land tax remained an important, but dwindling, source of the central state's revenues right up to the eve of the Pacific War. This decline is in relative rather than absolute terms, and one to be expected due to the inelasticity of the land tax base and the need to find other revenue sources for the expansion of the state.4 However, the severe revenue shortages that emerged as the war went on saw the state turn again to the tax 1 This contrasts with the c a s e in the United States, where the tariff was the centrepiece of the national tax regime, and the focus of party politics, until the first world war. On this, see John Mark Hansen, Taxation and the Political Economy of the Tariff," International Organization, Vol . 44, No. 4, Autumn 1990. Of course, the "unequal treaties" imposed on Japan by the Imperial Powers in the late 19th century restricted her from using tariffs, which left the land tax as the mainstay of the state. 2 Literally "rich country, strong army." On the politics of the land tax in this era, see Banno Junji, The Establishment of the Japanese Constitutional System. London: Routledge, 1992; especial ly pp. 170-99. 3 Kurt Steiner, Local Government in Japan, op. cit., p. 26. On the rise of corporate lobbies in this era, see Lonny E. Carlile, "Dealing with Diversity, Maintaining Autonomy and Seeking Integration: Zaikai amid Taisho Demokurashii, 1900-1930. February 1996, mimeograph. 4 This trend of relative decline in the weight of land taxation is a common phenomenon in f iscal development; see Ryoushin Minami, The Economic Development of Japan: A Quantitative Study, 2nd Edition. New York: St. Martin's Press, 1994, pp. 259-60 209 as a supplentary source of revenues. Table 7.1: Land Tax Revenues. Selected Years 1880-1947. (Units: until 1945, thousands of yen; after 1945, millions of yen) Fiscal Year Total Taxes Land Tax 1886 55,262 42,346 1890 68,007 40,084 1900 153,459 46, 718 1910 406,595 76,292 1920 938,057 73,945 1930 1,103,085 68,035 1939 2, 927, 975 48,685 1940 4,140, 843 3,921 1941 4, 817, 942 25 1942 7, 350, 520 13 1943 9,731,154 6 1944 12, 715,146 37,828 1945 11,541,379 27,651 1946 37,438 . 38 1947 189,600 0 Source: Tables 12-8-aand b, "National Taxes by Principal Items (1927-1984)," Historical Statistics of Japan, Volume 3. Tokyo: Japan Statistical Associat ion, 1987, pp. 268-71. By contrast, the land tax became more significant to the local authorities. This was because they -- through the agency of the Home Ministry - imposed surtaxes on the tax at the prefectural and municipal levels. In 1939, for example, prefectural and municipal (cities, towns, and villages) surtaxes on the land tax brought in, respectively, 48.3 million yen and 44.0 million yen. 210 This compares with the same year's total of 48, 685 million yen for the national land tax.5 Fiscal Democracy and the FAT The Occupation's 1945 assumption of virtually dictatorial control over the Japanese tax system brought wholesale changes to the structure and role of the land tax. Rather than leave the tax at the national level and merely permit local authorities to levy surtaxes on it, S C A P handed it over to the prefectures. The municipalities were then permitted to continue levying a surtax on this new prefectural tax. The tax itself was not very productive, due to the fact that its base was rental values and these were set by administrative fiat. Hence the tax rate was frequently increased during the early postwar period in an effort to have it keep pace with the rampant inflation. By 1949-50, for example, the tax rate had reached 250 percent of assessed value. This unstable situation continued until the enactment of the Shoup Tax Commission's recommendation that the tax base be widened to include capital values of land (rather than rental values) as well as depreciable business assets, and that the tax be renamed the "Fixed Assets Tax" (hereafter, FAT). Moreover, reflecting a precept of American fiscal democracy, Shoup proposed that the tax be shifted in its entirety to the municipal level. The rationale underlying this recommendation was that of providing municipalities with a stable and 5 The figures are taken from Tables 463, 463, and 467 of the Japan Statistical Year-Book, 1949. Tokyo: 211 independent tax base, one whose rates local electors could determine through their representatives. The American model to which Shoup turned derives from America's high rate of home ownership, which made the property tax appear perfectly suited to the task of funding the local community's historically limited supply of services. Since homeowners pay the tax, the argument goes, they can be counted on in elections to express their fiscal preferences concerning the package of services they desire, thus providing strong supervision to local government. The tax has therefore long been seen in public finance studies, which are dominated by American academics, as the ideal local tax base. This tax ideology also finds concrete expression in Anglo-American local governments' unusually high reliance on property taxation. 6 The FAT's wholesale transfer to the municipal level was thus accomplished through enactment of the Shoup reforms. It remained in place during the revisions of the early 1950s, though an amendment was made in 1954, partly out of concern that the munipalities would not be able to administer the tax on their own. This amendment transferred, to the prefectural level of local government, the FAT for assets whose value exceeded a specified amount. As Statistic Bureau of the Prime Minister's Office, pp. 856-7, 860-1, and 864-5. 6 This is shown below in Rgure 7.3. Also, B.Guy Peters notes that "the United States uses the property and wealth tax to a greater extent than any other country"; see "Determinants of Choice in Tax Pol icy," The Determinants of Public Policy. Ed. by Thomas Dye and Virginia Gray. Lexington, Mass . : Lexington Books, 1980, pp. 207. See also John F. Graham, "The Place of the Property Tax in the R s c a l Sys tem," Retrospectives on Public Finance. Ed. by Lorraine Eden. Duke University Press, 1991. 212 Steiner puts it, "[t]he partial transfer of this tax from the municipalities to the prefectures was a significant reversal" of Shoup's policy to emphasize the strengthening of the municipal tax base. 7 Though Steiner's point about the tax base is correct, for this dissertation the 1940 system received its strongest reaffirmation when control overtax rates was firmly placed in the jurisdiction of MOHA's early postwar predecessors rather than the local communities themselves. Table 7.2: Land Tax Revenues by Level of Government. 1930-1944 (Units: Thousand Yen) Year National Land Tax Prefectural Surtax Municipal Surtax Towns, Villages' Surtax 1930 68,035 70,904 7,110 34,605 1939 46,685 48,288 25,117 18,919 1946 3, 921 32,088 26,289 45,145 1941 25 31,619 28,060 47,127 1942 13 31,809 29,164 50,457 1943 6 29, 618 26,543 49,118 1944 37,828 40,123 22, 596 60, 861 Source: data from Tables 463, 463, and 467 of the Japan Statistical Year-Book, 1949. Tokyo: Statistical Bureau of the Prime Minister's Office, pp. 856-7, 860-1, and 864-5. The 1940 tax reform itself had begun to shift the land tax to the subnational level, and especially to the cities, towns, and villages. This is evident in Table 7.2, which shows that the land tax's decline as a national revenue source saw the opposite movement at the subnational level. The municipal and lower 7 See Steiner, Local Government in Japan, op. cit, pp. 274-5. 213 levels of government began to draw the lion's share of the revenues from this tax base for the first time in 1940. The difference with the Shoup system is that the 1940 reform did not make the local land tax a truly independent tax, but rather transferred it nominally to the local authorities while keeping collection and rate-setting functions firmly under centralized control. The reason for this was that the land tax in 1940 Japan was viewed as the quintessential national tax (cf. the Anglo-American experience). Even the Japanese Imperial Household paid the tax, so its transfer to the