Report No. 11094-CHA China Budgetary Policy and Intergovernmental Fiscal Relations Volume I Main Report July 28, 1993 Country Operations Division China and Mongolia Department East Asia and Pacific Regional Office FOR OFFICIAL USE ONLY MICRPOGRAPHICS Report No: 11094 CHA Type: SEC V~~~ Of -the Worl &ank This document has a restricted distributi~on and nivay be used by recipients only in the performnance of their official duties. Its contentts may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS Currency Unit: Yuan (Y) $1.00 Y5.46 Y 1.00 = $0.18 FISCAL YEAR January 1 - December 31 WEIGHTS AND MEASURES Metric system ACRONYMS AND ABBREVIATIONS AAPRC - Audt Admistrton of the People's MIS - Management Infonnation System Republic of China MOP - Ministry of Fiuance ABC - Agricultural Bank of China MOFERT - Ministry of Foreign Bconomic Relations ADB - Asian Development Bank and Trade BOC - Bank of China MR - Ministry of Railways BOF - Bureau of Finance MYE - Multiyear Expenditures CASS - Chinese Academics of Natural and Social NCGA - National Council on Government Sences Accounting Cc - Capital Construction NFEBR - Nonfiscal Extrabudgetary Revenue :CP - China's Communist Party OECD - Organi -iion of Econonmc Cooperation CDs - Certilicates of Deposit and Dv -lopment CEM - Country Economic Memorandum NPC - National People's Congress CESH - Culture, Education, Scie=e and Health PBC - Peopl's Bank of China CIECC - China International Engineering PCBC - People's Constucdon Bank of China Consuling Company PEMS - Public E1xpnditur Manaaent System CPA - Certified Public Accountant PIC - Provincial Investment Company CPE - Centrally Planned Economy PPC - Provincial Planning Commission CSY - China Statistical Yearbook PRC - People's Republic of China 1BR - Extrabudgetary Revenue PSBR - Public Setor Borrowing Requireme FEBR - Fiscal Extrabudgetary Revenue R&D - Resarch and Development PYP - Five-Year Plan RCCs - Rural Credit Coopeatives GAAP - Generally Accepted Accounting Prctc RIPS - Research bIsite of Fiscal Science GAO - General Accounting Office RMC - Resident Mission, China GFS - Govenment Fmance Statistics SAMB - State Asset Management Bureau GNP - Gross National Product SB - Specialized Bank GDP - Gross Domestic Product SEC - State Economic Commnision GVIAO - Gross Value of Industrial and Agricultur SIC - State Investmeat Corporation Output SNA - System of National Accounts IBRD - nternational Bank for Reconstruction and SOE - Stat-4Owned Enterprise Development SPC - State Planning Commission ICBC - Industrial and Commercial Bank of China SSB - State Statisdcal Bureau ICOR - Incremental Capital-Output Ratio STIN - State Fixed Investment IDA - International Development Association SWF - Social Welfare Function IOT - Investment-Orientation Tax TB - Tresury Bond IM - Inner Mongolia TVE - Town and Vilage Enterprise IMP - International Monetary Fund UB - Univend Bank (L/D) - LoansDeposit UCCs - Urban Credit Cooperatives ZBB - Zero-Base Budgeting FOR OMcCIAL USE ONLV CONTENTS Phfwe ..... . . . . . . . . . . .. . . . . . . . . . . v Excutve Summary and ..R.onun.datlon. . ........ vii I Budgeta Polcy and Intergov nmental Fiscal Relatoun: An Introduction I . . .. . .. .. . I A. Chia'sBudget and theChangingRoleofGovnment I B. TrendsinExpenditures andRevenues .. 3 C. Issues... 9 D. Outline of the Rep ............................... 14 2 AFameworkfor Budgetary Reonn .............. ........... 15 A. Introduction . ......... I5 B. Ihe ScopeoftheBudget .............................. 17 Ismes . .... 17 ReoomnlnezKlationh ...................... . .24 C. TheB udgetasaPolicy Tool ............................ 28 Issues ............ ......... 28 Reconmmnendations ................ ......... 31 D. Thelnlr Role onitRheEudgRee t he e... 40 Issues ........ . . . ... . . 40 Reconnendat io.s.... 42 E. Concluding Remaks .................. 43 3 Eyxpdltur AsdgmnentIn aDecentralized Eonomy ............ .. 46 A. ntroduction ...................................... 46 D. An bternational Perspective on Expenditure Assignment Over Levels of Government ..................................... 46 Theoretical Considerations . ......................... 46 Some Internaiond Evidence ...... .................. 47 C. Expenditure Assignment in China . ........................ 50 The Istibuiona Framework ...... .............. . 50 Low Autnomy in China .. ......................... 54 Trends in Central and LocA Expenditures and Revenues ....5... S D. FincingExpenditures ........ ....................... 68 hemationd Experiee ....... .......................... 68 Central-LocalRevenueShang I China ............. .... 72 bItrgovermentalFinancial Flows . . ................... 78 This document ha a restrictd distribution and may be used by recipients only in the performance of their offlcial duties Its contents may not othervise be discksed without World Bank authofiation. -il- 3 E -nditure Assdgmnat Ina Debo4ralued Ewnomy (cont!d) E. Expedir and nteproilnla Equity .. ................... 83 Cros-Prvi,iicWa Expenditure Patterns .. . ... .. . ... .. . .. .. 83 Culture,Education,SSdleneandTechnology, eadlh ......... 84 Pdice Subsidies .......... ....... 8S Does the FlscaSystem Equall? .......... ....... 86 F. Fiscal ncentivesandDysfunctlonalOuomes ................. 91 G. R......e.d.ton......... .. .. .....* . 95 ibllWraphy .. .... ........ ........ 99 1.1 Csfcannd DefintionofBudgetary Acwuts ... ... 108 1.2 Domnestic GovenmentDebt ...... . . .... . 113 1.3 Govnnmenes W'ageBill.. .. . .... .... 117 2.1 Thenstiudtonalando IganIzatlonal iaFramework ................ .... 120 2.2 RegulatlosonStatelBudget Mngement . ... .. 133 2.3 Financi Accountig and Reportg .. . ... 147 2.4 FinancingInvesment . ... 154 2.5 ForeIg Capita and Investmnt .. .... ......... . .. . 180 2.6 RailwayTransportInvemenmt aPming andtheBtdget . .187 .3.1 Expenditureand R eeAsslgnment in aM MLedE aoomy ............ 193 3.2 Aocation of Bank Credit Acas Provinces ..... .. .... ... 204 3.3 Paying for Price Subsie .......... ........... .. ... 209 3.4 Emprical Analyds of Denbaliaion . ................. 213 3.5 Pwivdal Finance in Ci bib ase o sses... ................ 217 asusulcalAnnas ..... .. ........... .. . 224 TAZ IN TErXT 1.1 Stae Budgetary E..x .......................... 6 1.2 Explic*ltFof State Budget Defiflct ....................... 13 3.1 Conceua Basis of Expedit Assignment . ................. ... 48 3.2 Expendr A nient in Selected Conis ............................... . 51 3.3 Provinci Shae In Local Expmfirs ad Reveues for Seected Prvinces . 57 3.4 Breakdown of Expendir on Culte, Educadon, Science, and H h ..... 60 3IS Adus iEnlsr...veExpendlt.r.s ............4444**4* **....4.4.4.4. 65 3.6 Nonx ieveorSelddOECDCoundtes,C1988 .....untrles,19........... 72 3.7 Chin: Central-Local RevnueShaig Contrt, 19880 ............ 79 3.8 EsdImeCed tra-Loca Trandss, 1990 ........ .... .. . . . . ..... . 81 3.9 ProvionI CESH ad CC Expendiu ...... ................... 85 Dow RN Tr 2.1 1990SEtrbugetar Funds . .............. 19 2.2 New Prna of the1 StatBudget . . . . . . ... .. .. . . . . . .. . . .. . 23 23 AtiU"e Budget Cledar. Tietable (o Compiig the Ste Budget ..... 30 2.4 JapmneBuodoPtoces .....*...............*. ............ 34 2X FomultIn of th Budget l Nehelands ..................... 36 2.6 Effimnymnd Ef ia Revienw: TheNethedands ..... 38 3.1 ntralzngEducatn In Chin . .......................... . 61 3.2 DeHalth Care In China .................................... . 63 3.3 Tax SSwing andFaEqu uao an Gtm .....ionlan.............. 70 3.4 Provinc Tax Eflb In China ................................. 73 3X RevemxShwingFormulas: SomStaegs ......77............... X !iGuI E N 1TxT 1.1 ShftsinBudgetaryExpendfltures ...... .................... .. 5 1.2 On d ise: R B ......RealBudgetarylevenuesaiExpenltu. ....... 9 1.3 ShiftsnP.enpios: Budgetary ExendinusAdjused .............. . 11 1.4 Cut of Control: Unplanned Exedtures ........................ 12 2.1 Stop In lpan'sBudgetPeo.es . . 33 3.1 Govrnumt StucW In Chin ....................... ..... 52 3.2 ntral Local Expnis: Shafg a SbinkInPg ................ 56 3.3 Keepg Cohntol: Centra and Locd Expendins and Earm d G .... A8 3.4 Cor Local Expeadibs: Second Courms ..................... 59 33 ranesa FinanlaiFlows ..... ....... ................ so 3.6 DwIdig Shaing: Central and Locd Revens and Net TaxShaig S....... 82 3.7 PeCapita xpendt andlvR es nC insProvenls,1990 ........ 87 3.8 Through the Budget, 1990 ....... .......... ... 88 3.9 eThunFiscalAlow: Fisca andResoure Transfer,1989 .......... 90 'tN * ri1  *1 I iIjP ill *1' Ir1di U 1i 0 a Iii IiiIJIji. jf;m; ii U *A*I IIIJI4IIIJJIIUII I I1jji Ijij IA' j j illhlill I - vli - EXECUIIVE SummARY AND RECOMMENDATIONS IntroductIon L. Ecoomic rermn sInce 1978 signficandy danged the role and proffle of the governmet budgeL Th buge has evoved ftom an all-encompassing acounting for the Plan toward an accoun of genal government. Decentrizaton of decsion power Sratly reduced and altered the scope of the budget no e teeneura d tizo a pwved ctucial in te sucoess of CWs econmic ref'm d*cressed the budgeay presence of StateOwned Enmtrpie (SOEs). Decenization of govermet f ions e d the budgetary presen of local government, and the devolution of decision power Inrased th Importan of locad goveme In the budgetary process. These significat aDges in te govnme wer accompid by only margnd changes in the budget syotem and in Intergovenmel fiscal iStuon. i. ia's budgt system seved the oury wel under cenr plann but k incres y alls sbort of the demands which Chia's emerging socaist mart economy put on t, ad budgety praices I b the economic decentralizaon in the refrm period. The currt budget system does not adequately refect the preset and fiur fisca resoibilii of Chna's govem The budget is ot well oquipped for ita growing funcion as a onoic policy tool, and Ib still overly geared toward the taditiondal role as a resour mobilizer. tovnmental fiscal relatio have not fuly adapteDo the nw IndeenDence gained by loc nmen, leadg to a proliferation of extra- and ofibudgetary prc tbat uderm i budget controi and m o mic sablity. The dwindling fisa reces In the daing system led to an erosion of the fiscal equaizaton between provic. In. he main challen for the Chinese autries b to mabe the budget a mor effecte policy tool, both In its role os an indirect lever In manging m , a a bewer-deid Inrmen for meeting nw soci objectives. This report nalyzes the urent problems wih tie budget sysem and recmmens how t ca be transformed to meet the dhallenge. The rort ihat many of the recomdations can oily be Implementd as the overl reform progm unlds and the future role of governm Is clarified, but some chges can m Imprvwe the effecnss of the budget. iv. The report focuses on two crucial featu of the fiscal n (a) the stuctur and itudonal aspects of the budget system; and (b) China's intergoverunental financs with patar rferenc to expendue assimen by levels of government Chinas fisca systeq needs refm in other areas as wel, and earlier World Bank stUi hav - viii - addrese theso in deil4/ Fiscal policy In a maket based economy aims to (a) promote stability, (b) enable officient allocation of scar=c resources, and (c) guaanty equitable outcomes of the economic process. The incaed openmess of Chia's economy will make fiscal policy die domiiant macroeconomic policy tool, but fiscal policy will also play the dominant role in allocaion and income redistribution, as the Importance of the Plan furher diminishes. Reforms to equip goverinment with the appropriate tools for fiscal policy in the market based economy are therefore a matter of urgency. v. Neco.sary reforms outside the scope of tis report Include (a) transforming China's revenue system into a tme tax system, (b) broadening China's tax base, and (c) modemizing tax collection. China's govermnent has recognized the need for reforms, but the Initiatives ta have not been without problems. The experiments of separating taxes and profits goes in the direction of a trme tax system, but Implementation is held up by reluctance to phase out the tau deductibilit of debts, an issue that could be solved by grantig these debts. Ithe ongoing extension of VAT and Business Tax to a wider category of goods and services is diveifyin the tax system, but the complicated rate structure remains a barrier to an e4uitable and efficient system. The reformed tax system wfll only be effectively enforced with a solid collecdon eanism and with the increased income flows across provinces, and the loosening of ties between SOEs and government, a nadonal tax system becomes neesary. Budgetary Trends Over the Reform Period VI. Chia's budget reflects the profound cunges in society tat have taken place since 1978. Following the overall retrenchment of the goverment from the economy, budgetary eedimes declined by over 10 percentage points of GNP over 1978-91. To a considerable degree, this is a reslt of conscious and desirable policy reform, as SOEs were allowed to keep more resources, and at the same time saw their excess profits decdine as competition increased. However, including thefiscal part of extrabudgetay funds (that is, funds that are essentaly of the same natre as budget resources), government stfill commands more than a quarter of GNP, a share comparable with that of other Asian developing economies. Priorities shifted away from resorce mobilization for production enterprises toward more traditional government tasks. The greatest cuts have been in capital expenditures, as Stat-Owned Enterpris (SOEs) were excluded from the state budget in the course of rerorm, as they now rened profits for reinvestment. Further sharp cuts took place in reported national defense spending, whereas socia expenditures, adminitation, and price subsidies all grew moderately, and debt service grew rapidly as a share of GNP. The trend toward soci expenditures is likely to continue as enterprise reforms will move so expeaditurs of enterprises to the budget, th enlargig governmet's share in GNP. Overall, the decline In goverment's revenue share of GNP outaced expendiure decline, leading to persistent budget deficit over the reform period, reaching 2.5 percent in 1992 (IMF definition). Vi. abudgetary revenue gained strongly in Importance, rising from a third of budgetay finds in 1978 to over 90 percent in 1990. However, in 1990 only one sixth of total extrabudgetay fiuds, or 2.6 percent of GNP, are Jiscal extrabudgetary fimds, related to government fiuctions. The bulk of extrabudgetary flunds consist of reined eanings and 1/ Word Bank, VW=ina Revenue Mobiliaian and Tax Policy," Washngton, D.C., 1990; and Cbihn Refoging nteroU FlFiscaldReatons,' WasgionD.C., NovemberZ7, 1991. -ix- deprecito funds of SOBs, which have been growing fast In the course of increased indepndence of SOEs. viii. Responsibility for admisteing public expenditure shifted towards lower levels of govment over the reform period, ncem t with the genera decentralization. Cental governmet's share of tota govement expenditure is now below 40 percent. Although this may appear low in inationa coWmparison, it Is by no means exceptionally low for a country of China's size and development stage. Also, China's central government share is suppressed by the largely oftbudget finance of Social Security and Defence, usually central government functos. The downward shft of expenditures over the reform period can be attributed largely to: (a) the new emphasis on functions traditionally administered at local levels, such as social expenditure; (b) wage Increases and increases in admnisave expenses, which fell more heavily on local goverments; and (c) sbarp rises in locally administered, but centrally set price subies. Local governments' own sources of revenue did not keep pace with vising pendires, and they became increasingly dependent on earmarked grans from central govermnt, and strongly expanded their reliance on extrabudgetary funds to perform goverInment fucions. Tax sharing through the tax contcs decreased rapidly in importance, as central govemment gradually augmented its fixed revenue base through the reassignment of individual taxes or a chao in enterprise ownership. Such sharing now accounts for about 20 percent of local revenues and expenditure, and about a quarter of the center's. ix. Budgetary finace of Chinashigh investment share in GNP decreased btaily over the 1980s, a trend even more pronounced for investments in the sta sector. Budget finane was replaced by: (a) own funds, that rose with the growing independence of enterprs; (b) firein f bnd-increasingly in the form of direct foreign investment-as China opened up; (c) domestic loans; and, more receny, (d) new mechanisms of fnance such as stocks and bonds. The naue of budget finance of enterprs investments changed from direct capial grants to an annual lump sum allocation for the capital construction fund, managed by SPC. The decline b budgetary finnce of Imvestment was partclarly pronounced in local budgets, and consequently the recourse to altenative fiance sources was stronger at that level. x. In smmary, the expenditure composition of the budget has substamdiaUy changed over the reform period. In recent years, more of the budget has gone to the provision of public goods (social sves, infrstrture), while the share of enterprise finance has gone down, and the locust of experes has shifted to local government. The functional specialization between government and enterprises, as reflected in these numbers, has been the major dring force behind China's economic success. Moreover, the decentralization of ownership to local governmen has strongly enhanced competition among SOEs, and substnily reduced x-efficiencies. However, the fundamental changes in Chinas society have not been witho problems for the budgetary system. Mai Issues and Problems Xi. The Budget System. While many changes have occurred in the scope, coverage and size of govaement expendhre in recent years, and many fiscal reforms have been instied, budgetary processes and practces in China have not kept up with the accelerated pace of reform In the 1980s. The scope of the budget In China is lagging behind the changing role of govemnment oW vermUent activities ae not unified In the budge:, wbereas nongoven actvities sti remaIn on the budget. Entrpses still have a strong presence in the fiscal -'- accounts, although financ indepndec b the stated policy. The prolifeaton of ofibudget and euxabudget funds blur,ed piority* seing and has weakened the budgetary contol medis essenal to a well-functioning fiscal system xli. Acal expen conistey supersede budget esm, by as much as 18 prcent in the mid-1980s, but In 1992 stll by over 8 peren of budgeted expendites. Multyear axeditu plang is hardly developed, and the curre method for budget plaing invtes incrementalism. Individual economic uits hawe resisted periodic program nd proJect aessmen and i-crementalism in the budget process has confbutd to Wee cOntrOl and Inreased presure on the budget. xiii. ne Chie governmen has moved away from detaied planning activities of oWer sectors in the economy to guiding these decisions. ITis should til accompanying budget rfdorMs, especly to ehance the budget as a provider of signals to economic agents, and as an instmen of momic poy, but this has yet to be seen. In pardcul, govements capacity to def a system of ienives through the budget is inibPted by the reticted publicity of the detied budget ad the late approva of the budget wbich impedes economic agets to take acount of govenmen plans in a timely way. Macoecontrol is weakene by the wide divergence between eed and actual experes and revenues as noted above. xIv. Recourse to bgaufnig and noncompliance with fiscal regulations have contbud to fiscal undiscipline across institons. Rules and codes regarding entprse tax liability are evolving, but entprises at tax liability Ughtly in an environment where the finnc bureaus play the double role of tax collector and owner, and in a sitWion where taxes are frequently contracted ther than aessed. Budgetay finance of SOE investmer_ is still virly fe, as repayment raio of the PCBC are extremely low, or as repaymens are canceled out against tax obligation. This bas negative oecoomic and macroeconomic onn, d work to the detimet of bankig sectr stability. The Chinese authorities have in ed a number of chages in the budget system to deal withis situation, but thes masures fdl short of preparingthe budget for its new role. Many of these problems are not unusual r an ecnomy in transition. However, futu budgetay refoxm needs to address them if macrconom difficuties w to be avoided In the fuue, and a number of acdons set out below can be tae Immediately to impove the funciing of the budget. xv. The recodations of ths report are therefore desigped to address these issues by suggestng ways in which the scope and coverage of the budget can be redesigned in order to make it more closely match the new role of governmen as it is emering in the course of reform. Similarly, such refrms of the budget are necessary if the government is to be equpped with a budget that will ass it to manage maecnomic policy, and pemit itot use the budget to guide the decisions of other economic agents. xi. lntu o n Jiance. China has not yet fundamently ched the sstem of intergovenmena ce despite the increased role and responsibility of local govnment, and this has given rie to a number of problems. Furthermore, the devolution of power to an already deconcntrated goverment has given local government de facto command over functons that determine macroeconomic stability. The lack of reform has driven central and local governMent to take actions that furher undennined the fiscl system. An incresing mismatch between local rvenue and expenditu assignment has occurred over the reform period, as local -xi - gove ts took on ic resed responsiblities for, for eample, price subsidies, but wihout new sources of reve. Un aty incased In venm relations, due to cdang levd of ownership of entprs, pansion of central fixed taxes, and ceta 'borrowing from localities. Repeaed reoratig undemied the tax cotact which, in principle, wee well- designed to generae revenues; kt reduced revenue collections well below potential; and made fiscal planning at the local level more difficult As the resources In the tevenue-sharn system stagnathed centra pvemnt allowed the reditbutive element of the fiscal system to weaken, at tm when divergifg economic condion-a posle reat to reforms-would require moro equalization among regions. Fil reditbution now takes ple mail) through ad hoc, earmarked gran allocated on a one-by-one basis, often poorly targeted and too small to meet basic needs in ttie poorer regions. Moreover, the arbitry nature of centra grant allocations led to extensive negotiatons and rent-seeking by localities, tying up valuable administrative resources. xvii. Besides ineased rent-seking, local governments responded to the fiscal squeee in other ways that undermine the desired fiscal system and jeopardid macroeconomic stability. Ihe local reactions ilnclude- (a) exanding the local tax base at the cost of cent governmet; @b) tuing budgeay int getay fumds and tapping enteprises' xgetar finds for govnment puwposes; (c) pushing expendit" ofibudget, sometmes financed by indirect local borrowing that puts addional demands on local banis, and strais macroeconomic stanility; and (d) run arrears in payments or issue IOUs, a practice tiat fueisnftion. Loal gove have reduced effdve tax rates on entetpris proft below the stauory rae via entprise profit co ts. Loca govrnments offer tax concessions to enterprises ta affect the size of the taxable income to the detent of tax buoyancy, for insta by manipulating the rules for pretax repayment of investa loans. More , the resourcetrapped local governments depend on local entrpris for their revenues, and are tus tempted into inefficient regioa competiton and loca protctionism, made possible by voids in competition reguation. h Chinese autorities have repeatedly changed central-local fiscal contract, but without fundamentally altering the icentives the iegove fiscal relations ihat give way to the observed dysficional ouomes. mviii. Future refms in this area wil tiherefore need to focus on clarification of expenditure assignment, on formulation of crieria for fis d buon, on the menu rt of flsS need and on an apropriate system for assigning revenues to each level of govermen The rort contends that the prese system is creatg distotions and inefficiencies by its lack of clarity, and that, in patcular, the fist two issues need to be resolved before a detaied design of tax- sbaring can take place on a rational basis. A LngRun Budgetary Fmework xix. Within the cone of overl fisl reforms, China's budgetary process and itergover a relations noed fameal change to support fiscal policy that protes efficicy, equity and maeconomic stability. Ihe inc g mportance of fiscal policy in macroeconomic management urges for speedig up fiscal rdorms. Fisa roms can only be brought to success if al parties gan, which i more likely the more comprehensive the reoms ar. Budgetary refom and chang ingv rlations should threfore go hand in hand with a broadening of the tax base, the creation of a local tax base, a rationalized ta collection system, and a reblis of central ovenMets macroeconomic control. The evidence from the past with fiscal reforms suggests ta the experme , regioaly limited - xii - pieceme reorm method, sucefl In many areas, may not work with fiscal refotms. A graual natiowde Implemtion of an established reform plan encompassing the varius aspects of fiscal refom promises to be the better method. xx. he Budget and the Scope of Governmet. Budget reforms must change the scope, conten and objectives of the budget and budgetary policy. In the long run, the budget should reflect the role Chiln's authorities see for govement in society in a unified and clear way. Nongovernment activities should be removed from the goverut budget and fiscal accounts, includingentprisextrabudgetary fund. Budgetaryinvestmentexpendituresshouldconcenate on govermnent functions. Fiscal watraudgetary funds should become an intgral part of the budget. Goverment actities now perfLrmed by enteprises, such as social security, sould be brought on the budget. Quasi-fiscal activiti3s of the banking system, such as subsidies to loss- making SOEs and for agriculral Inputs should be made transparent en the budget The establishment of separate policy bank to manage policy loans seems superfluous, if not damaging, and designatiot of an exosting institudon as the main policy bank seems preferable. Unifying government activities on the budget is a precondition for the budgetary process to become a meaningful mechanism for allocatng scarce resources over government objectives. A unified budget will reflect the macroeconomic significance of government action, a precondition for rational macreonomic policy. xxi. Govens share of the economy should be an outcome of the change in budgetary scope instead of being a goal in itself: there is no objectively opftim share of goverment in an economy. More Important is the ability to fiance the desired share of government in a non- Inflationary way: tax reforms and tax adminisaion reforms should in the long run achieve that goal, and eliminte rcourse to the balking system. xxii. China's authorities desire to keep a substantial state ownership of enterprises. A Nonfinaial Public Sector (NFPS) Account consolidating govermen and State Owned Enterprses would be a useful supplement to the government budget. The NFPS is based on the crterium of government control, instead of the criterium of govermTent functions. As long as China's government exerts considerable control over SOEs, and SOEs remain instruenbts for implementing govenmet policies, public sector accounts would yield a better picture of the overall govermsne. establishment. The public sector borrowing requirement (PSBR) would be more relevant than th8 government deficit in detrmining the overall macroeconomic stance of goveivmt. FinancW ianshtiutions owned and controlled by government should not be consolidated in the public sector accounts, becase this would eliminate information on which part of the PSBR Is fianced by banks. Government ownership does not pjclude a beneficial separati of functions within governmet. Separating the ownership function from the tax collector function Is indbspensable for the implementation of rule based policies, for the imposition of hard budget constrains, and therefore for improving SOE performance. xxii. TMe Budget as a Poley Tool. The budget's increased role as a macroeconomic policy tool could be emphasized by integrating the budget with macroeconomic developments. Th requires an explicit recognition of the macrnomic assumtions behind the budget in the form of a backgunddocument on the economic outlook to be published with the budget. xxiv. Chins wudget classificatiec and presentation needs adjustment such that the budget wi be usf lor multiple users and uses, and adoption of the Iational standard of budget categorizaon and presaions strongly reommended. The prentaton of the budgae oud - xiii - allow economic analysis of goverment plas by means of an c onomic presentation, in additon to a functional and administrative classification that assigns spending authority to specific ministries, unit and levels of goverment. xxv. MOF could improve the quality of expenditure decision makng. 'rhe icementa bas mmber method should be abandoned, and more output and program-oriented budgeting techniques shoud be adopted. Introducing multiyear budgets and commitment budgets, together with crossreference between Plan and budget would improve expenditure planning, and recognize the multiyear nature of many government programs and policies. Capital budgeting capabilities of MOF need further strengthening. Ihe fiscal reporting system must be geared more toward policy analysis, and the AAPRC mandate should be expanded to include output and effectiveness audits. Institutional mecism , such as the expenditure review committes many countries have, should be established to raise efficiency of government expenditures. Such epeaditure zeviews would be Instrumenl in effectively increasing the tight fiscal envelope, and therefore contibute to overall fiscal reform. xvi. The legal basis for budgeting and expenditure control needs strengthening. The draft Budget Law should explicitiy assign responsibilities to units and ti ons concerning tigb and obligations on expenditure budgeting, approval, authorization, disbursement and audits. Procedures for redressing expenditure overnms should be made explicit, and sanctions in case of abuse should be codified. The budget law should be supplemented by a detailed budget manual, explaining in detail the rights and responsibilities of those involved in the budget process. xxvii. Intergovemntal F1nnce. China's authorities should assign functions clearly over levels of governent, with efficiency of service delivery and fisca equity as guiding principles. For erample, education and health care should be primarily a local responsibility with cenr government only settng general guidelines, whereas other functions, especially those of interl market regulation and competition policy should be assigned to central government. xxvii. Control over government functions crucial for macroeconomic stability should be eCen haized, including control over aggregate government borrowing, and overall tax burden. The increasing intergovermmentl flov-. of goods, production factors and income gives ctl govenment a comparative advantage in tax collection, and a Natonal Revenue Service (NRS) is highly desirable. Such a recentralization of control over =dmbnirra*on of revenue sources (boowing, taxes) may well coincide with a decentrization of the revenue acca from these sources. Mhe expenditure assignment should be the basis of a ew rmeme a ignment, and the exp ture necessary to perform the assigned fmctions efficienty should be covered by the revenue assignment-provided the overall fiscal envelope is suffrciently large due to China's tat reforms. The efficiency of government service delivery depends to a considerable exten on the autonomy a level of government has, on the financial Icentives it has for cost savings, and therefore on the memns with which the expenditures are financed. xxix. China's authorities should assign each level of government a substanti own tax base, with which it finances functions f which it has full responsibilities. China's authorities hould assign local governments a true local tax base-including significant control over tax rates-and exploit the opportunity for more accountability and the incentive for better and cheaper public services such a tax base gives. Own tax bases can futhermore serve to absorb vaiations in expenditure needs without a tiresome renegotiation of intergovernmental fiscal relations. Finally, - xhv - extension of the local tax base contributes to brk the present deadlock in the system of intergho menW fiscal relations. To some exent, fiscal etudgetay funds are a de fact local tax base, but a legal basis for this should be provided, and as argued before, these fumds should be brought on the budget. For local govment, prope taxes would be an excellent own tax source, which is thus far hardly exploited in China. However, the subordination principle in income taxation distorts incentives for local government, and should be abandoned in further reforms of the tax system. xxx. Tax Sharing and Equaslzatdon. A full separation of tax bases in China is probably not feasible, nor desirable in the light of efficiency of collection, assuming that the NRS will be established. The fiscal gap between expenditure needs and revenues from the own tax base can be filled by tax sharing, general grants, specific grants, or a combination thereof. If tax sharing is chosen, a necessary condition for tax buoyancy is that the collecting agent has a large stake in the taxes it collects, i.e., its share of the tax should form a substantial part of total revenues for that agent. To accommodate for the increasing divergence in economic permmance among regions, and the consequential growing divergence in fiscal capacity, a fiscal system with equalizing properties is necessary. Horizontal equalization of fiscal capacity between localities can be achieved though tax sharing arrangements or through equalization grants. Whatver the mechanism chosen, China's authorities should desip horizontal equalization mechnLim that are resilient to inflation, and can accommodate substantial shifts in relative welfare positions of localities, which are likely to occur as reforms deepen. xxxi. Adminisatively, it is more complicated to incorporate equalzation cerations in tax sharing arrangements, and usually a separate equalization fund is set up, or centra government allocates funds directly too poorer provinces. Ihe latter solution would, however, require a larger share of tax revenues going to central govermnent, which seems difficult in the present Chinese situation. For China, an interprovincial equalization fimd could be the appropriate way to strengthen horizontal equalization. This closed-end fund could be filed by a certain percentage of taxes shared between central and local government: this eannaking will have no effect on tax effort in the cas.; of a NRS. The fund could for instance be adrnnsered by a body at arm's length of central govemment, with both local and central representatives. Disbursements from the equalization fund could be based on broad indicators of exendie need, such as popuation, area and urbanization, and on fiscal capacity of the locaity stemming from its own tax base. A variety of formula-based grants schemes or tax sharing arrangements can be found around the world, but China's authorities should make the primarily polta decisions on what factors should be given weight in the formula, and to what extent fiscal capacity should be equalized acs localities. xxxii. ITe elements of intergovernment fiscal relations as set out here need to be captured in a Local Government Finance law. The law would work out the constitutional division of labor between levels of government, and specify the financing mechanisms, including the above menioned tax sharing and equalization arrangements. The law would in part serve as a guantee for local govenment that the NRS will not be abused to reverse the devolution of the reform period, and would establish the increased local autonomy by means of codifying local taxation rights. xv - Imediate Actions xxxiii. China's authorities can take a mnmber of immediate actions that clafy the scope of the budget, improve its function as a policy tool, and improve intergovernmental fiscal relations. xxidv. To restore budgetary unity, the various levels of govermentfiscal extrabudgetary funds should be reported as part of the budget, without necessarily changing the control over these fiuds. Enhrpise extrabudgetary funds should be removed from fiscal accounts, as they no longer fall widtn the scope of govenunent. xxxv. The budget should clearly differentiate between government and non government where already possible. Among ohers, this requires that the budget clearly distinguishes size and nature of government budgetary support to enterprises. To prepare a further delineation of government and nongovernment, MOF should prepare an inventory of those units and agencies that operate In the budgetary and pscal extrabudgetary sphere. Such a listing and the change therein as reforms proceed should be reported along with the budget. xxxvi. Tho Chinese authorities should consider a wider publication of a more detailed budget, togehr with a detailed descrption of the categories. This would enhance their capacity to guide decisions of other economic Ments. MOP should initiate a recoding of budgetary classifications, which would allow multii le users access to budgetary data for mltiple uses. This would require taining of all levels of government involved in budgeting. xxvii. MOF should facilkate a wider flow of information between levels of government, between ministries and even between its own departments. MOF should strengthen the capacity to analyze budgetary implications of economic data, both macroeconomic and microeconomic. Data and analytic capcity combined would enhance the accuracy of the budget forecasts, ngthen the programmatic effect of budgets, and would make the macroecnomic effects of budgetary decisions more transparent MOF, SSB and SPC should establish a working group on data reporting, which should work on consistency in reporting on budgetary data. xxxviii. MOF should adopt stricter measures to contain the impact of expenditure overuns. Inclusion of contingencies In the budget is a first step, but MOF should issue more detailed and stricter regulations concerning budget readjustment, including detailed rules on which instituton Is responsible for absorbing the expenditure overrun and sanctions for the case of noncowizpllance. Enforcement of this new reguation requires more regular monitorng of budget implementation by MOP, especially at local level. Announced and unanounced inspections may be ntrmental in enforcement. MOF's and AAPRC's audit capacity therefore urgently need strengthening. xxxix. The budget should clarify the size and nature of budgetary support for enteprises. A clearer di on between loans, grants and subsidies should be made. Policy loans hough the banking system should be registered as a contingent budgetary obligation. MOF shoud issue clear petformance and eligibility criteria for the various means of bu%igetary support. xi. The present tax sharing experiments do not match the fundamental revision of intergovernmentlfiscal relations as sketched above. Notably, the expeiments largely maintain the fical cotacting system, which lies at the heart of the present problems. Moreover, the - xvi - experimens have not been implemented as designed by MOF, and therefore seem more and more irrelevant for the evolving situation in China. China's authorities should therefore consider substantily modifying, or even giving up these experiments altogether. xli. An intergovernmental committee to tackle the various issues in intergovernmental fiscal relations should be established. The committee, manned by both local and central representatives should solve the technical issues in implementing intergovernmental fiscal reform. The commuittee should make proposals for the criteria governing the equalization fund. They should be in charge of drafting the suggested Local Government Finance Law. The committee may evolve Into the administrative body that manages the interprovincial equalization fund. xdii. The erosion of equalization through the fiscal system should be stopped now. Government should consider to increase the lump sum transfers to poor provinces right away, without waiting for the fundamental revision of the system. Also, cofinancing requirements for earmarked grants could be differentiated among provinces, to account for the different fiscal capacity. Both measures could be financed with cutbacks in price and SOE subsidies. The equalization payments could be gradually enlarged as the new interprovincial equalization fund will be established. This fund should be built up gradually, starting with a low percentage of al shared taxes. This percentage would then gradually be increased until the structural level is reached. In this way, local governments could easily adjust to the small changes, and probably can pay their *contribution" to the fund out of overall revenue increase. Unks with Other Reforms xliii. The above recommendations could be implemented independent of other reforms, their impact would be enhanced by reforms in other sectors. ITe main linkages are spelled out below. xliv. Financial Reforms. The speed at which delineation of government f*om the rest of the secoors can proceed and thus the budget unification can be completed depends on the speed of reform in other sectors. In particular, financial sector reform and enterprise are essental for progress in the budgetary sphere. The further development of the emerging stock markets will yield better altrnatives to the financing of enterprise investment than the banidng system. Developing a long-term bonds market would equally improve enterprise funding, would reduce the need for bank finance of budget deficits, and would enable goverment to better spread debt repayments. A well functioning financial market could relieve the budget from the present task to mobilize resources for national investment programs. Finally, a reformed banking system will help imposing hard budget constraints on state owned enterprises, a necessary condition for entprise reform. Xlv. Enterprise Reform. Enterprise reform will substantially affet the budget, as it wfll reduce subsidies, and may increase revenues. Until SOEs operate on a for-profit basis, and are no longer used for implementing govenment policies, such as minimizing open unemployment in hef course of price reforms, the finamcial links between SOEs and government will go beyond those in other mixed econi nies. Ihese links, as long as they last, should appear on the budget, however, and not be hidden in the financial system. x1vi. Regulatory Framework. Cental government should fill the reguatory void that gives leeway to regiona protectionism. Government should prepare and isue nadonal laws - xvii - regarding competition, company law and contract law to fill the instiutional void, which gives localities protonist leeway. Central govemment should strengthen its capacity to monitor local govenment's implementation of national tax laws and regulation. MOF should start the preparatons of a naton tax system. The system would curtail illegal and inefficient tax exemptions by local government, and would facilitate the introduction of a new tax and tax saing system I. BUDGETARY POLICY AND INTERGOVERNMENAL FISCAL RELATIONS: AN iNTODUCIION A. CIINA's BuDGE A TE CANGING RoL OF GOVERNMENm 1.1 Govment budgets reflect the outcom of the political procs in which chois ars made among competing demands for limited resources, and the size of govenmet actvie In the econmy decided upo. Budget can be a powerl tool for allocating nadonal resours, set natonl priorides, determining the relative size of the prvate and pubic so. and expressing the naure of intergoermental rtions. To the tthat budget weig, articulate, and litmize the demands from all intested parties, trslag them Io a collective interest,' the budget could be seen as an expression of the country's 'reveaed 1.2 Toward a Mixed Ecnomy. In broad terms, the goal of economic refrm in China 2/ has been to isttute an economic system in which: (a) te government would continue to play a central role in econmic m _ageme in terms of fomdatng and implementing laws and regulations; formulating and applying economic policies, and establishing and readjust the management system. Insruments should focus more on m omic management and rule-based approaches with emphasis on market-determined pices in products and production factors; (b) the govement would participate in providig of social services and wdefre but hrough mecm and instons hat would citate efficiet risk pooling and service delivery, with more equitable burden sharing between individuals and the govement sector; (c) the govemment would continue to decentralize its participative role and enable the enterise sector to make fts own decisions on invesmet, products, and quantities to be produced and sold, responding to market forces. Martet forces 1I As discussed in sm of the pubic finance _tnstr dealing with h thoy of th public int_t It is ls elegant, but not less sietific, to tbk as a saiting point for etn_ of social acto tde oved objectives of soiety insted of the derived ones.' P.O. Sier, Public Expenditure Budgeting, in A.S. Blinder (ed.), 1ke Economics tf Publkc FThace, Washing, Brolpings, 1974. Z o Pw a eredv ew of the refor decade,'" se 0w Ceauy Ecum M. 1adur-plnbeg and Rq1n in dse ISM, Wodd Bank, Wasln D.C., 1992 - 2 - would Increasingly determine the allocation of resources, and the govenmenut would rely more on indirect policy instrumeuts;2/ (d) the government would decentralize operations within the government sector and enable locad governments at different levels to perform specified government functions, encouraging competition and performance comparisons in the provlsion of public goods; (e) planning, focusing on channeling of public investments to key sectors, and market principles would coexist in a complementary fashion. 1.3 Progress along these liaes are transfoming China into a "mixed-economy,*4I where public policies would be mainly conducted according to rules and indirect fiscal and monetary instruments common to industrial and developing countries. 1.4 Flscal Poficy and Fiscal Reform In China. Fiscal policy in a market based economy aims to: (a) promote macroeconomic stability; (b) enable efficient allocation of scarce rurcs ; and (c) guaranty equitable outcomes of the economic process. The increased openness of China's economy will make fiscal policy the dominant macroeconomic policy tool, but fiscal policy will also play the dominant role in allocation and income redistribution, as central economic planning firther recedes. Reforms to equip government with the appropriate tools for fiscal policy in the market based economy are therefore a mater of urgency. 1.5 This report discusses two crucial elements in the fiscal infrastructure: budgetary policy and intergovenmental relations, but China's fiscal system needs reforms in other areas as well, and earlier World Bank Studies have addressed these in detail.jI Necessary reforms iMnlude: (a) transforming China's revenue system into a true tax system; (b) broadeing China's tax base; and (c) modernizing tax collection needs to be modenized. The experiments of separating taxes and profits goes in the direction of a true tax system, but implementation is held up by reluctamce to phae out the tax deductibility of debts, and issue that could be solved by grandherg these debts. The extension of VAT and Business Tax to a wider categoy of goods and services is diversifying the tax system, but the complicated rate strucure remains a barrier to an equitable and efficient system. The reformed tax system will only be effectively enlforced with a solid collection mechanism, and with the rased income flows across provinces, and the loosenig of ties between SOEs and government, a national tax system becomes necessary. Proper arrangements of intergovernmental finance as discussed in this report will make it femsble as well. 1.6 China's budget system mirrors the profound changes that have taken place in society and the organization of the state since the late 1970s. Gradually the state budget has evolved from an all-encompassing account of the public sector into a reflection of the accounts , In the Chinese lHteature, this is usually referred to as 'the state regulating the market and the markt guiding the Lterrse z .4/ Sine th 14th patty congres, oe could ue the term 'socialist markt economy.' Wordd Bank, -China Rene Mobiltio and Tax Policy,- Washington, D.C., 1990; and China Refomingn vementFiscal Relatons,' Washington D.C., November27, 1991. -3 - of the general government (central plus local governments). This redefinition of the scope of the fiucdons and economic activities of the government has been carried out through a process of devolution of ncions to financial and nfiancial public entrse and local governments. The budget lso reflects the changing gvvenme's economic role in carying out dvelopment objectives, while reducing its direct role in the production of goods and services. 1.7 In addition to the change In the coverage of governmental activities, the evolution of the budget also reveals the extent to which expenditure priorities have been altered during the years of reform, as a response to changing policies, for example, away from military expendiures and toward soci sectors like education and health. Ihe budget also summaizes the finanal implications of the compromises made during the.reform to achieve somedmes contradictory objectives: reforming relative prices in favor of the gicturlS sector, whie protecting urban standards of living; and giving a larger role to markets and competition, while at the same time shielding state-owned enterprises (SOEs) from those same forces. 1.8 OutUine of the Chapter. This chapter prese a brief overview of the main trends In budgetay revenues and expditures and in the fiscal deficit and its financing since the begining of the reforms. Then, it discusses the change in governmental expenditure priorities in a few main categories whose importance has increased or declined during the last yeas. Three annexes to this chapter include a detailed definition of the main expenditre items in the Chinese classification of the budgetary accounts (Annex 1.1), and a detailed discussion of the fstest growing budgetary expenditures (debt service, Annex 1.2 and the Government's wage bill, Annex 1.3). B. TRED IN EXPENDIUS AND REVENuES 1.9 Trends in aggregate budgetary revenues and expendiwres and in the fiscal deficit and Its financing in 1978-91 are reviewed below. 1.10 Budgetar ExpMenditures. Following the overall retrenchment of the government from the economy, budgetary expenditures declined from 31 percent of GNP in 1978 to 20.4 percent in 1991 51 Crable A-1.2). The greatest cuts have been in capital expenditures, from 12.6 percent of GNP in 1978 to 3.9 percent in 1991, as SOEs were excluded from the state budget in the course of reform. Further sharp cuts took place over those same years in national defense spending, from 4.7 percent of GNP to 1.8 percent. Against the general trend, nondefense current expenditres rse slightly, from 13.7 percent of GNP to 14.7 percent; social expenditures, adminisation, and price subsidies all grew moderately in terms of GNP, while debt service grew very rapidly (see Annex 1.2). 1.11 Budgetary Revenues. Over the reform period, China's budgetary revenues have declined significantly, from 31.3 percent of GNP in 1978 to 16.8 percent in 1991.71 While the decline in revenue in terms of GNP was intended, nonetheless, it became a major concern of the Chinese government, a main driving force in the decline in expenditures, and the cause of recrt changes in the fiscal system and in central-local fiscal relations. However, with the gi Chieso definiion Due to lack of dled atistics, it was not always posible to us dat compatible with tdo in onal pracdc 11 Chinm definition, excluding debt issued -4- benefit of bldsigh, the decline In revenues Is not srprisng and can largdy be explained as a direct comeence of the govemens economic reform. The main fctor behind the decln in revenues has ben the reduced budgeay conribudons from SOEs-die tu paymentB ad profit mi -which fell fom 20.6 perct of GNP to Im ta S pacent dudrn dte peiod, whs entere loss-retred as negav revenue-grew strongly. Eneris inme rached 55 pent of total revenues through the mid-1970s before tax reorm reduced th compont to virtay zero in th mid-1980s.SI 1.12 The reduction in SOE contribuons was caused by a combinaon of refom InDiati tha transfrred resources from the govement to the enterrse sector and policies ng ety, reslting in increased competon.2/ Ihe profit-retenion scees enterprise rtenton of dercation fmds, and the deduction of pretax amortization before tax payments ga the SOEs more auonomy, but at the same time, reduced fscal control ovr resources. More pronounced was the erosion of the govenmens fisal capacity by maket forcs: Q) relaive price adjsment shfted proft from the havily txed industri secto to less axed secto, notaby agrice and rura eeaprises;1Q/ (increasing compedtion in th hxdustra setr, especially from town and village enterprises (Vs) and collective anterprises but also from abroad, depleted SOE monopoly rents and reduced govrnent revemes. During this period, taxes on namaonal trade and personal incomes did not increase nearly eno to offse the dedline In enterprise income taxation. 1.13 ExpedItre Shares.Uj In recent years more of the budget has gone to prvid *purer public goods and services (Inftructure and social services) and less iusrial nvesmn and production fincing, as the general govenme's scope and coverg ws nduced. With the notable exception of debt service, some of the most rapidly rsing expdo co ries have increaingly fallen on esource-saed local budgets, forcing them to look to other rabe sources of income to finance their curren and capital e (Cbt 3). The main changs in the rative r of the gods and seices financed by do budget between 1978 and 1991 were (Figure 1.1): (a) Capital Construction. A large portion of fixed capital formation could be decenrized from the budget to nonfmcial public entrprises. As a result of flI C. Wong, Fiscul Reform and LocAd l I M Probmatic Seuencing of Refm in Pout-Mao China, Modern Cina, VoL 8, No. 2, April 992, pp. 197-226. 2 aBrq Naughton, 'hmplcatio of the State Monopoly over hdustry and its Relaxation, Modem Cina, VoL 18, No. I (Januy 1992), pp. 1441. IW See Wodd Bapk, *aWda Rewaw MobIladon and Tax Poky,- Wahingt D.C. 1990; Chmne P.W. Woog, Ce*tl-Local Reations i an Era of Fisca Decline: the Patadox of Fiscal liCtn,QasaQi.iy (Decenbe 1991). ly, lb Chiwne classfication of buddgeay expenditures has not Impt pace with the ching natume oovege, and incesg complexity of publc activities. Te tradinad incldOs 13 andheads of acount Crab 1.1), combiing fimetional and ecomic rhie, nd wltdM distinguW etwee curen and captl exnditres. t a mo N& _ level (Annem 1.1), Cbines budSay e du lasifie ito 29 main item, with th _a woeomp mooticned above, which lmitsdte inbr otonal contet and uW of the budgetr incno fo antyeaG puspooe .5- Pl%e : Sin IN 3VJM AItY v_it owwfo cum CESH -wedlem 47%% 2U I _ Mln _ w _ ~~~~~~10% udo Olh 8% OHM 4% 31% 11~17 Nob ad USINiGUpI or1vckmaTa M svao h1clessng dieHIaton betw th eonomic fimcto of the public secor.-I A pon o whih wa filanced by qtrabudgay fbads (aptes 2 and Amex 2.4-2.6). he share of capitl cons don jJ In tot budgetary oulays fel by half, fom over 40 percent of toWal qeadu in tie lat 1970s (almost 13 peet of GNP) to about 20 peret Om ha 4 percen of ahits G(NP) i 1991. A u*rher dedine s budgetd for 1992. The cag In the share of capia construcdon e spenurs in the stats budgot hides the drastc declw hi loca pgovemes shwe of thu exedtu in 1978 the centrallocal slit w about 56:44, In 1991 it was 74:26, nd it Is xpected to decin to 79:21 hi th 192 budget While theo s of capil mcdon wa cut In haf durn the refom yea, t o W be . an expenditure responsibt - concentaed at the cent. IV a smal tobmo mu.t nd loPlo g d ), nofnacl and fiac)d W,1 So ud 1am,inAm 1.1. XtI W4SP '.dNLPD IW ma U'SNP6 (g. U 91-1W J 9.ei mp :~ :_om V'tCOD" 9-tS@t 01*01' GMS"T *'SU't T'Ottl 9'G 6" M 8-90 $Otf 6 9`11t- tS t-Ut M o- 69- NIt "It 9rt "9It "It "It -tie tVg 093- 50'S V- -t o't- o-S* t d . t t9-t 6e-t atl' WIC nts' WS SS- 0*0I 00eo o0X0 00e00 000 00e0 oeoo 0 Colo I1"mv .We tO't "'I iO'i t0't Wi' U-1 WI "'I Lt'Ul 9-t "'I 01 't@ st*i iT re-t dMIM "Q,dft eAIwaeeTuwY Solst U"I "If OSc "ICs ist Wi Gi Ost "It "It "It os', ut' OSS gilt ml O0I5 061O 95 961 9 ' s5t " 'S 54*9 fl* SOIL of,t so's 5'9 IS,# WIT 09 -n u e'Kg ss'w 09e #Wt 95'? £9sS 90.9t S-tt 90 *u sru g't8 eu 9 *S ts'Ut ie*s ws 9 in Ilu . tS.Q O-tt u*-tt iO* s@'s 6t-0 gSri oeo i'W O*-O 4C*I- 90'S' 0l't t*' O is U-O - o e "ITT SI* 9U'55lt go'," 69 "t 4tOII 0 1 - - - - ' ' ' - -- O@t 6V'9 OS'S Ls's W0' Ct9 n sit till r " - 'tW I'S 9's Sol; CO, e-o* p ooo dwP 9tO6 tol' 594 "If Colo gilt toll 00'S "It "I Ii £'9 go's t t s''t M.se,z.1e;" mU.inzA. ' oi'2 0"lo -8 9O'S Is'0 3 9- T5905 ti-tt 91't 9'65t SO'S? W6it 6 c-n06' t n-n"" T t1 W9It "LIt "Ist W9 t SS-t tS-t K-t tt 59't "It gC-t £0. P9It all tilt P5 T PW W - 0os0 st'O eros It-Ot ire os'e ire is's OO'S t'SO-OI S i- *i 0-9 94*5"* 69'-n "l9oSt Uiti StP n I O t-91 0rnIOUI ''tit W9t O - i L "II "t51 90t tro01 ntot ot pm .'.w 'wsu.-4 '0 *=mm tt-9 ttf9 59*9 "t9 tSo' titS SIS it'S tt? i t5i' "l9 f*99 K's9 SO' Celt6- 900535 fVft tt'? 95*T "It "It' SCIT ii1 0S-1 to l il t 00'S S1 O t S9 -t r99 t9i .e%eem VW umIdemta '£eswu "It 00-1 "t1 6Ot 5*i W'l nfS s9i "ISt 99'I "It lo' 06-t filil Solt he.dbzd t * L'i t-55' 9It Celt "It ot 9 0V9 5999 "sI tilt so's so's 09-9 so's it9s U-P31 PM h "lo 9'S9 U0 6S0o isro 49 0 tO 9i-0 S9-0 lol0 "It 90'S SOlt olt Col9 ot;b t t0ff- ti-C t t5't WsUS3 OSIg -9* trot 59'i5 59955 t594t 00.93 09,6 "6.9 Kn net o"re ansm3am" - muqg _------- ---- s19mon_ Tvw* S .zq I, 9 C 9-6C 9'e5 9 Ol 09t i.0 0@ tIO- It-- VI0I I'-- i5. *It, 5@ 9s *i 'tt t-t5 i*ii*Ct1C t-t f- - - - ' ' ''- - '~eq m* S'n 9rig 01 Vi VI' 0- S OlS 0*' 4 S'9 V'S C-9 # *-h - P x 3.09 O-'t S'CC flog e*if 6t slt $I? I1 ? t UVOlII r 8 ti i9 is ItS =W" '39FW0 05SmZ*O Sl5 @'OS 0'6E 5'i3 *-n O-tt iP et-61 t-8t ,it 9nLI I,i taes cn ot9 *MM wPM"a o9 8 -S S- S 0.s V-t I's 's tsC s's t -' V9 *It TIC cs - 1 m M i - 9-0O -ss 9 s eitg i-' 9-tc *lot 'tt *,it t-tt SlO t*6 I'i 9-9 P O-U 6e69 T9 ' sSS 9-" Slot 0-8C IIsI 1S9 ,'9 sGe I *'it s'ai rsst qt 9 -ut.P pm I'st U0tt 'IIusga'e 0-SX $la 'SIR 6tst e's *sC t'n r'os re s's 0'. *'i s's 06 ' l i 9'S VS9 *9 S' 6' Ils L'St 6S I's t I- t *St VIZ 5ST M' *,I q p W * -Vt l' 9-5 5C5 s's s-C 't s Olt 9 SS it *-E s's VS 0S .d let flat 'Sn 9'Stt t'St MTS, O'S ClOt CIt 6-' 69 S-9 Olt ri 59 uess.Pmwzs p M at""M Pesw. TT Olt t*5 t'- 0-1 tlI 0-1 9 5tOt C-t C 9 r't Is S'S 1'9 TOS&* t-p9 t'U K-i 9-Ui 9'S9 C-C9 t-g9 :it9 S*I e' 6-n s's eos slet- SIss Slit US-UICZM Textfd ~~W P1W PWt V~~~~ Pizw GPW P1w VW P1w P1 VW PMW P1 IMT w m P ees tees 066t 006t 5S4 iet i 961 65i 5941 5a 5 tt t 1S115 0et "ST us i (Mond pu ugq A) As _ZDf ZiVIS :1' SjqqL -7- (b) Tecal Upgrading and Transformatlon.Wl IhW are of these Investm s bhave made up a stabl- portion (5.5-7 percem) of total budgetay expendires over the first ten years of reform. Since the 1988 austerity program, these expenditures have slightly declined in relative Imporce, a utend that I expected to continue as SOEs finance an increaingly large propordon of their investments with depreciation funds, retained earnings, and commercil borrowing (Annex 2.4). The share of capital constuction and technical renovations taken together provides a better picture of budgetary investments. lbis share fell from 47 percent of total budgetary outays in 1978 to 24 percent in 1991. (c) Culture, Education, Sdiece and Health (CESH). Social expenditures are by now the largest category P' budgetary oudays, almost doubling their sbare In budgetay ecpenditres between 1978 and 1991, from 10 percent In 1978 to about one fifth in 1991. Tbis rapid growth can be largely explained by the rise in educational etenditures, up from about 6 percent of total expeditu In the late 1970s to almost 11 percent in 1991. The new, higher plateau was achieved early in the refom period, where the most drastic changes in the compositon of xpendiur wok place. Since then, the share of education has remaied constan, at about 58 percent of social expenditures. Health expenditures have lso increased their share in total expenditures (Box 3.3). For both categories of expu , aminisive costs (especially labor), absorb as much as 80-90 percent of tota oudays. The rapid rise in real wages over the last years might have eroded the supply of these services in real terms. Contrary to the cbange in the center-local split for capital construction, the split for CESH expenditures has remained practically unchanged at a ratio of 11:89, with most social expenditures in the local budgets (Table A-3.8). (d) Governmt Admiistrtion. The share of administrative expenditures has grown .yotandy since 1978, from 4.4 percent of total budgetary outlays in 1978 to about 10 percent in 1991, reflecting a rising number of goverment employees and rapidly rising nominal wages (Annex 1.3). (e) Repayment of Loa. MOF started to place government bonds in China's dometic market only in 1981. At first these placements were carried out through mandatory allocations, but more recently via voluntay placements increasingly among households. The late start and the modest size of the early Issues W led Initally to a dow Increase of the government's domestic debt. Ihe outtanding volume of government paper in December 1985 amouted to less than Y 24 bilion, 2.8 percent of China's GDP. Since 1987, however, the U/ TemoT ogidal upgradn and tra fomion is arficially dinguised fiom capital construction by the sin of the investmes nd share of civil works in ivestme proects as well as by thir orgoizanal affilon, so Annex 1.1. j/I The bonds issued from 1981 to 1984 averaged about Y 4.4 billion a year and were issued with a maturt of ten years, redeemable in equd portions stating in 1986. ITh larger 1985 bond ism (about Y 6 billion) caried a shorte, fiveo-year maturity. - 8- volume of treasy scuries issued swelled; diversificao (and cot) of the u min n used have increased; and mauiies have shortened. Although the ovel "vlsible W govenment dect to be finced (by domesc and foeig sure) Wli has averaged 2.3 percent of GNP annuy sine 1987. the size of the governs domestic debt In December 1991 was mor tha five imes Its 1985 level, Y 122 billion lW or 6.5 percet of GNP. The oera sde of the governments domestic debt remaXis small by te inal standards, not only vis-&-vls China's GNP, but also considring Its high svings Me2l and pgid economic growth. Yet the debt stmctu has ceated persitet finacing and servicing difficulties for the authorities. Ihe rapid rise in the public debt after 1987 (Annex 1.2) in a stMil embryonic domestic capital market 2W and, more important, the ntimwous shortening of new treusy bond mamities IV has resuted in a bunching of amortization and Interest paymentr (payable upon maturity) which ended In a severe 'liquidity crunch3 in 1990. As a result of these changes, the share of debt repaymet in toal budgetary outlays has risn from 2 percent in the mid- l980sto 6.5 percent in 1991 and an estimated 10.5 percent in 1992-all of ithas falen in the lap of the central goverment. Management and finncing of other budgetary categoelas has Increased, forcing the central govement to renegotate tax contat and other predatory pracdces, with the negative inctives mentoned above (Chapter 3). (f) Price Subsidies. After 1986, when price subsidies were reflected as an e Item, thei sae in total budgetary outlays grew to a peak of about 12.3 percet in 1989. As price reforms were adopted, igbsidles' share stard to decline, to an expected star of 8.2 perceat in 1V %2. In terms of te budgetay outlet," about 85 percent of all price subsidies are reflected in local budgets, though rgely financed out of earmd cena grans. AI To contai eo sa of the measured deficit, many budgetay obligations ve been shiftd to th banindg systm fli Betweefn twbrds and thvlre t of the govnmt's overall deficit i th lat fiv yeas (1987-91) has been financd domleaticlly. II/ udes the volume of govern bonds and isses classified vnder Financial ,intttions," with the exception of th "Bak' Finac Bonds" (Annex Table 1.2). II/ In 1987-91, 's ross G ational Sriing averaged 39 peret of GNP. 2W Itay, withnoscondarymakes,MO Pissusaountedforpractdicaly all -narket num.ats (not ndig CDs). Sic Decmber 1991, goewment paper has aouted fo about 70 pnt of al outsnding debt (ain excluding CDs). See Annex Ta 1.2. IV Pro te year in dt eay 1980s, to five yeats in 1985, and recnly to tre ya -9- C. 181 1.14 Msung the Rete Decline In the Size of the Governmet. he decin of govn share In GNP has been a major concen of Chlnvse authoris. Howeve, despie tho undeniable decin In the share of govermt In GNP, both governmt xpendirs and revenus 2 have epeienced signfican grcwth In redl emms. Ral gprew p by an annual 4.8 percent over the whole refom perod (1978-91), and by 7.3 pereen betwn 1981 and 1991 (Table A-1.3 and FIgir 1.2). Rwal rvenues grew less, but stI increased by an average of 3.2 pecet per year between 1978 and 1991, and by 6.3 peren In 1981-91. Thus, governmet'sshdinking sha of GNP does not Imply a loss of control over re resouces, but merdy a laggn bdeid the tota economy's excepdonal rea growth. FIgwu 1.2: ONTz 1E RISE Real Budgetay Revenues an Expenditues (1987 prices) Y blllon 250- 250 --------------- --------------- -- -- 200 --------------------- a 3--------_~--------------------- 10t 197879 80 81 82 83 84 85 86 n7 88 89 90 91 Year -Expenditures 4- Revenues 5M.,, MOP Dab'; CM"a CE 1.15 Recatg the Budget. To mak a maningfa ae f the rcee chp in he gvenmens dsr, certain adjustments need to be made, as Chie fisa st8a do not foflow interatonal practime.2 Mie main ones are: (a) Subsidies to losa-making emnte s and, befor 1986, prie subidis for daily livg necessities, ae teted as negative reveumes rather thane elpel W Chino dndfniIiaa coueled for dsbts. fl Foneadatdbo modifd hrenssoo, ssWoddB Bank, 'XCwRaiweMcEWvaa Tax PbUq,6 Washnon D.C. 1990, p. 9-10. - 10- (b) Debt isued s treated as revenue rather than as a fincing iem 'below the line; (c) Debt (incipal) repayments o treated as expenditures, rather tan as a negative financing item. Adjusted for these accountng differences over the reform perd, both expediures and revenues sti fbllow the deciig trend in terms of GNP, but at a higher level. Expenditures decline from 34.1 percent of GNP in 1978 to 22.5 percent in 1991; and revenues fal from 34.4 percent to 19.8 percent over the same period (Table A-1.4). (d) Chnese fiscal statistics differ from the International Monetary Fund (M) sandard classification 2W/ in the treatment of extrabudgetay revenues. About 80 percent of total extrabudgetary funds consist of enterprise depreciation funds and retained eaning, which should not be included at all as part of the government's fiscal accounts.2W The other categories of extrabudgetary revenues of local finance bureaus and anmasdsrtve and nonprofit units hould be considered fiscal revenue. We label these fiudsflscal exrabudgetwayflds, and consider them as part of general government revenue.4i Ihey should therefore be brought into the budget (Chapter 2). On the expenditure side, no statistics on extabudgetary expenditure per institution are available. However, as a general rule, entities that collect the revenues are also entitled to spend them.2V/ Fiscal extrabudgetary expenditures are therefore assmed to be equal to the fiscal extrabudgetary revenues.211 1.16 Ad4ustd fIguru. The budgetauy expenditures adjusted for the GPS fbomat, including fiscal eatrabudgetry expenditures, show the same declining trend in terms of GNP as the Chinese figures discussed above. However, the governments much higher share over the wholereform period fell fom 36.8 percent in 1978 to 26 percent in 1990 (Figure 1.3 and Table A-1.4). The revenue decline is smilar as before, but the levels differ from the Chinese figures. 7A1 IMP, " Manuo of Gownment Finance &athtkw,' Washingt D.C., 1986. IV One madification applie here: the Chinesw goverment exercises osiderble control over the SOB x ty funds, inditecdy throu regulation but also directly, becas MOF and the finance buteaus still exrcs ubdsantia influence over day-to-day management (the Stat Council's 'COadar on Sv=8gth dw Managenwm offa budgetwy Fundt' (Aprl13, 1986) speaks of paned with policy guidancse). To dte extent that this influence is used to purse _overnt goals intead of enteprse goals, extabudgetaby venues could be considered quasi-axatiaL W See IAP, A MJWi of Gowrwnemt F (na Statsda, p.8, whee 'Nonprofit nsitutid may finncd and effectively controle by government fall under the geneal govmenL During th m;ission it was suggested tat some of the nonprofit units wee financially self-sufficient, but for lack of deied data, all of them are tatted as genea governnmn. ;f Se the St Council's 'Cbwk4g on Strengtheing te Managemn of Ewurabdgetary Fundr,' (Apnil 13, 1986). 21 Ts amsmptin neglects (i) a differn in timing of revues and expenditwe and (ii) the use of fisbal y funds for rd*g budgetary deficits (Box 3.2). - 11- 1gure 13: SHIFNG PERCEIONS Budgetary Expenditures Adjusted (ercent of GNP) Pefrent 40 35 ------------------------------------ 30 -__. 25 ----- 20 -…---------------------------- ---- 197879 80 81 82 83 84 85 86 87 88 89 90 91 year Chinese Definftlon + Excl. Exrabudgetary * Ind. Exrabudgetay Soua MO dat for adjustments am toxL 1.17 The adjusted figures shed a different light on govrmet_s declig shar (a) a 26 percent share of GNP is comparable with the share of other Asian developing counti which averaged 27.0 percent in 1988;/ (b) the decline was to a considerable etot th consequence of the policy of increased autonomy for SOEs, which accounts for almost half te fiscal decline.&/ If the decline in defene expendures is included, about 80 perce of dte fal can be accounted for. 1.18 Apart from the etrabudgetary expendiures, a number of offbudget activiie can be considered as government, the most importat being: quasi-fisca actvies of the baking stem; social expendiures of SOEs; and offbudget defence expenditures. Quaui-fiscal actvie of the baking system could add up to as much as 6.8 percent of GNP in 1991.2/ Social spending by SOEs could amount to as much as 34 percent of GNP, and an equal amount could be spend on defence outside the budget. Since detailed data are lacking, a comprehensive view of expenditures on government functions canmot be established, which Is a concen in iself. VI Seo KX-youg Chu and Richsrd Hemming, ads., Publlc Egmad" 1wkd k (Wahigtoa D.C. 1991), IMP p.23. 30/I Ihs xuabudgetax funds of SOEs rses from 7 peroet of GNP in 1978 to almost 12 pera hun 1990. jJ, Wang, Chitie at al, Ecnomc R(oiw and Fical Mnaganwn In Cwa, Asn dw eop Bank, Fabuazy 1993 (draft). -12- 1.19 To aluo the govenments dae in GNP, Chis autes should first mak eh fihde polical decisions about ts role In society, then judge what reources are required to perform dt role in a cost-effective way. 'Me budgetay process could be an effedve tool for both step, but at preset in China, It Is not. 1.20 Th Rel Issues. The oncn over the flt rato' Is acually a cocrn over the Inabgity to find a sustainable, noninflationay source for government expenditu, and as such no at the heart of this report. However, the appare inability of the ceta govermn to cotin spending wii the limit of substatl ra revenue growth and the restg dicit is equally worisome. A good illustration of this lack of control is the excs of actual expitu over the budget expenditures (Figure 1.4) , which, after a brief decline from the 1985 peak have again been rising for the last four years. Appareny, budgetary proceres and practices cannot acieve the statod objecves of exp_ent control and defici redon. Furemore, the budget sem to be faiing as an strumen for rational priort seting. For example, ependitres on price subsidies and SOE subsidies grew during the second half of the 1980s, pushing other more presing exedh off the budget The lack of control and the lack of priority setting wihin the budget point to the need for budgetary reform (Chaper 2). Fire 1.4: Our OF COaNTOL Unplanned Exediures (1983-91) Y bIlion peent 30 30 25 ---------- ------------------------------- 25 20 ---------- --------------------------- 20 15 - ---------------------- 15 10 --- ---------- 1° 0 .5 ..5 1983 1984 1985 1986 1987 1988 1989 1990 1991 (eft Madse) Years (g hand soul4) U Unplanned Expend. 0 % of Budgeted Expend. No: UnpandW * ual _. budgetd expesdlur Sou= Vawu BudtW Spchs an MOF. 1.21 al Ddefidt. The State budget was in continous deficit dating the 1980s. Since 1987, the anual defcit has amounted to an aveg 2.3 pcent of Chis GNP. Ihe defick in 1986-88 was largely atributable to lagn stoa revenues after the adopton of provincial and SOE tax contracts. From 1989 through 1991, the fincid posidon of the - 13- govrnmemt underwent a marked detoration, widh an oveshootng of the planned dec from 1.5 percent and 1.6 percent of GNP in 1990 and 1991 to aaul deficits of 2.1 percent and 2.6 prcn, respectively.3V Since 1988, fiscal deficits have mainly reflected the se's difficulty in conrlling expeditu. After the rectificaton progsam was adopted in 1988, the eosion of the etate's budgetary revenues was temporarily reversed,32/ but other expendiures became more difficult to control.al Table 1.2: EXuT FNANG OF STATz UDGE: DEFICIT (Y billion and percent) 1987 1988 1989 1990 1991 Deficit (Y) 24.7 33.4 37.4 36.8 50.1 Total 100.0 100.0 100.0 100.0 100.0 Domestic 766.0 69.8 64.4 ZIA P30 /a 47.0 52.3 -28.7 73.4 55.2 Other / 53.0 47.7 128.7 26.6 44.8 J-gsr 27. 3. 30.2 36.6 23.8 Gross 156.0 120.9 127.4 135.9 146.2 hmortization -56.0 -20.9 -27.4 -35.9 -46.2 I Gross creditto MOP las MOP's depits at PBC. b Largely domestic bond issues (inudidg Special Construction Funds). Source: MOP and IMP. 1.22 Jbandng the Defidt. Since 1987, about a third of the stats average annual deficit (MP defnion) has been financed by foxlp credit (Table 1.2). Amortzato paymes more than doubled between 1988 and 1991, and gross borowing requireme have also Increased rapidly (hough from a sma base). The complemen of the stat's finacing ir emaib' 70 percet) has come from domesc sources. Among the domestic sources, tie relative sbhes of the People's Bak of China PC) and the domestc bond market have shown IV lhe swm. occured in 1991, whem pneiminary dat point to a deficit of 3.1 pucent; of GNP, 0.9 per high than budgted. .Ui Toa revms as a perentap of GNP icesed by 0.5 points betwen 1988 and 1989 but again fel, to about 20 peren of GNP in 1990 and an estimaed 19 peceM in 1991. 311 Durng dt firs ya of dt rificei prgram, subsidies (daily living acosstis and operting loom of SOEs) ms by 0.7 points of GNP, reach 6 pece of GNP in 1990 and abmising over a quarter of tot budgary exediue (26.7 peme). Ihe need to simulate dt coomy and finance ncasing invenories d pte of ta imncee (dh buinss t rate in 1991), th nrlxom of administbe conrob ovr caiad contucto and unexpeded exadtr- s (flood) nma the fiscl pewfomc wom tan expected. - 14- substna Istability over the last five yeas, although on averago PBC has provided about 40 percent &! of the domestc fnds, while 60 percent were raised In the domesdc capital market. D. OUm OF TE REFoRT 1.23 This report emphasizes the structra and instional aspects of the budgety system; it suggests options for reform and for allocating expenditure responsibilities by levds of government; it documents and analyzes the consequences of local government underfinding; and it examines the major changes in the investment finance system and its relaionship with the plan and the budget. 1.24 Chapter 2 deals with the changing nature of the govenment budget system In China. It reviews the extent of modernization needed in budget institutions and proceses to support economic reform and ways of improving their effectiveness as management tDols. T review is meant to provide a framework for changing the budget system The Institional famework and the associated budget processes are the focus of the chapter. A brief discussion of the evolution and current status of the budgetary system and review of recent nitiadves to modernize it is followed by suggestions for future directions of improvements. 1.25 Chapter 3 deals with the assignment of expenditr responsibilies over levels of government In China, and its financing. China's expenditure and revenue igmnent is put into intationa perspective The dhapter reviews the major trends in centrallocal shares of goverment expenditures and revenues over the refbrm period, and the patter of _ and taxes across provinces. SpecW attention is given to the redstbutive properties of Cias system of intergovernmental transfers. The recommenions address the correcdon of the dysfunctional outcomes that emerged over the last decade. 1.26 The annexes to this report provide details on a nmiber of ibstional feat and issues of China's fiscal system that fall outside the scope of the main report, but that may serve the reader who is unamiliar with China. &/ Excluding 1989, PBC's dwae is closer to SS percen - 1S- 1@. A FRAMEWORK FOR BUDGETARY REFORM A. IRODUCRON 2.1 Ihis chapter examines the cLanging nature of the government budget system in China. It reviews how far budget institutions and processes need to be modernized to support economic reform and how to make them more effective policy tools. Ihis review provides a framework for changing the budget system and focusses on the institutions and the associated budget process. The chapter discusses the current status of the budgetary system, recent Initatives to modernize it discussed and offers suggestions for fiuther improvements../ 2.2 An assessment of China's budgetary system, and rcm ons for Improvements require a well-stated vision of the emerging needs of China's economy. Chapter 1 argues that China is heading toward a mixed economy; this chapter evaluates current budgetary practices against those practices required for this type of economy. Howeer, the chapter recognizes the features incompatible with the role of government in mixed economy may remain on China's budget for some time to come. The chapter provides-where appropriate-recommendations on immediate steps toward the future role of the budget, while accomodating for these special features. 2.3 Government budgets have three main functions in a mixed economy: (a) a tool of accountability at govermment agency level; (b) a tool of management, emphasizing efficiency in achieving certain goals; and (c) an isuent of economic policy, contribng to growth, macroeconomic stability, and equlty.2I To perform these functions, budgets should ideally he a number of featr: 1J Some limitations of tbis report should be noted at the onset. First, becaue my budgetary details are kept confidential in Chia, not all the infomation tat is usually pubJic in other counties is available for analysis. Second, many is tat ase in the acal worldng of a system might be misod whe work is consrined to a more gneral and static examinaton of the budgeay process. This chapter therefoe deas mainly with the structural, institutional, leg and systemic aspects of Xt budgetprocss, rter than opeional podues and related isos. Tbud, abundt defition discrepancies i t oncepts used im China make rncng budgeay operations into t lycmab fonts most difficlt Fourth, for lak of di_aggegtd information, three ars elevat to the budget procss could not be covered in detail: accounting, cah _ aend fiscl rpotg. The annes to this chapter summaize th mis simod of key budgeary regulains (Ann 2.1); descibe the fuietions and organiztion of core agies and de (Ane 2.2); outine the characteristics of the t infomation sysem (M and fiscal rportg system (Annex 2.3) under the aegis of the Ministry of Pinance (MOP); and descibe the main channel and mecaims of ivestment finc (Annexes 2.4-2.6). 2,1 A. Prmhand, Gommu, Budgetg wd pardhw Control, IMP, Washigton D.C., 1990. a Om 'F - it ~ ~lloooP VTA or A9~ali - 17- B. Tun SCOPz OF 1K BuDGE: (a) Unity 2.4 A budget's scope depends on the scope of govenment and the purposes of the budget. As gov er s scope changed in the 1980s, a retrenchment of the budget's covage and a corsponding expansion of extrabudgetary tasatons were normal. Not surprisingly, extrabudgetary funds inased enormously during the reform decade. At end-1990, etabudgetary funds reaced Y 270 billion compared with total budgetary revenues of Y 330 billion, over 80 percent of all cetal and local revenues. However, as analyzed below, alumst 80 percen of ths exabudgta revenues (Y 207 billion) were enterprise-owned funds, while the balance (Y 63 billion) were fiscal etrabudgetary funds controlled by public instiuo and financial departmes. Financial deparments contributed to undermining budget unity by oigiaing many ovedrapping budget and nonbudget tansacons and by allowing governnt agencs to shid opeations on and off-budget. Use of extrabudgety fds for capi cosrion also =conuted to the oveiiieating of the economy and ofte created budeay liabiities for fure maintnance and operational expeditures. 2.5 Borderlne Betwee the General Government and Other Sectors. Government units are broadly divided into thre groups: (a) full budgetary units whose revenues and are included in the budget (e.g., line ministries); (b) d e budgetY units that recive budgetay support and whose revenue and expendtre gaps are filed with budgettransfers (e.g., goveanmt print shop sellingsubsidized publications); and (c) self-rdiant units that do not need ay budgetary support and are permitted to retain their rpluses of receipts over epeditures (e.g., selfsfficient utility companies). Some of these unit are empowered to leY fees and surcharges.3I In addition, entities simiar to enpes fn th they are subject to taxes, receive subsidies, and must remit their proft to the budget) should be treated off-budget, with only the net flows shown in the budget.4/ However, this dif rtI of the agencies' role vis-&-vis the budget has not been completed in Chna and confision aod ovelapping peist 2.6 Ezrabudgetr Funds. The present defition of eudgetary funds is all- inclusive, encompasing transactions that should not be channeled through the State budget.5I The principle of unity demands that all operations that are In the nature Qf goverua transacdons should be fully accuned for in the budget. We define those extabudget funds that are generated in such operations as fc exrudgetary funds. Correspondingly, operao that are not in the nature of the govarme transactions should not be included in the budget, , While thei power to imis fees in theor is cshn by the Price Bureau, in pratie dthey exeis discrtn in taising mvenues 11 Copating surpluss of SOEs dwdd be shown ndw nonta govemt rfeveo III PI I and Prpe Iome), while SOE loss should be clasdsfied under current expendite as subsidies to n c public enteprim (d tem from cqp Alhough th shoulbe_s capt). Sed b P (198nb. 51 Albhough they tboud be capWrd by a nonfmmda public sector blldget- - 18- except for thir net effect, such as profit remittances from and subsidies to SOEs. Along these lines and consiste with the authorities' intntion of fostering fiurther economic refotm and differentiation nside the public sector, SOE finds should not be included in, or denominated, as exudget funds. This would be In line with the "Regulations for Transforming the Operating Mehaisms of the SOEs.' Fiscal extrabudgetary funds should be brought back into the budget and reclassified (ideally) according to internationally accepted practices. 2.7 Extrabudgetay funds expanded very rapidly as a result of the reforms initiated in the 1980s (Table A-2.1). Since not all the fincial needs could be covered while continuing to manage the accounts of many units and SOEs within the scope of the State budget, a more flexible and decnrized management of 'social funds" was found necessary. The explicit objective of budget transactions was to allow increased flexibility. They were also supposed to change the struce of incentives to help tevitalize SOEs, speed up growth, and renew the delopment of horizontal ties athied by prolonged overcentralization, and, in gneral, improve incentives for government units. 2.5 The decentaizing reforms are often seen by the Center as having gone too far and many advocate for a stricter control (that is, recentralization) of "social funds."fi The arument needs to be examined by disaggregating extrabudgetary funds into their various components. Extrabudgetary funds fall under three broad institutional categories: (a) extIabudgeay funds of local finance bureaus, including, on the revenue side, surcharges on taxes set by local governents (for example, agricultural surcharges); (b) extrabudgetsuy funds of administraive agencies and insiuons, including highway maintenance and other cost- recovery fees, market and other fees, collected by government units; and (c) extrabudgetary funds of SOEs, including earmarked funds for the Tehical Transformation and Major Maintenance Funds (depreciation fund), retained profits, and short-term loans for circulation purposes (that is, working capital). Foreign investment and intemational loans are sometmes icluded in this category. In terms of composition, about 80 percent of these funds are owned by the entrise sector. On the expenditure side, extrabudgetary funds are used to finance fixed-asset ivsment, major maitenance, bonuses and welfare payments, administrative expendiu, expe es in the soci sectors, transfers and taxes paid to the central government, increases in working capital and other earmarked expenditures (Box 2.1). 2.9 At present, exabudgetary funds are not part of the budget, but they are subject to regulation in terms of those who are permitted to use them, the scope, the nature, and the purpose for which these funds can be used.2/ The original intention was to increase SOEs' and government units' flexibility in budgetary operations and provide incentives consistent with the policies of unified leadership and decentralized management. However, since 1986, extrabudgetary funds have been increasingly regulated, subject to annual budgets and presentation of final accounts. There is an awareness in China that 'if we continue to limit our thinking to funds within the budget, the reform of the fiscal system would be limited in scope" fi MTh main reasos behnd th call for tgbter contrl over extabudgetay fnds were: to stop illega fons of txaton, conon of budetary fimds into exabudgetay funds, _exsse capital constucon outside of the SPC apprval pmcess and plan, and diversion of earnad extabudgetay funds. See State Council Circular on Strngtheing the Management of Exurabudgetary Funds,' April 13, 1986. 2/ See Stat Council, 'Circuar on Stengthening the Mangement of Extrbudgay Funds, Beijing, Apr 13, 1986. - 19- Box 2.1: 1990 ExrRAsuD6ErARY FuS (Y billion ) Revenues: Classified according to agent Expenditures: Classifiedaccordingtouse (i) Local fiace dept. 6 Fixed asset Investment 93 (ii) Administrative Agencies : 57 of which: (iii) SOEs, of which: : 207 Capital construction 27 Tehnical transformation : 66 Technical transformation : 66 Maintenance : 37 Major Repairs : 30 Retained profits : 70 Welfare : 24 Special purpose fund : 30 Bonuses : 11 Other : 4 City maintenance and new product trial 5 Road use fees : 14 Administrative agencies 18 Culture, education & health : 17 Changes in woring capital : S Transfers to budget 29 Other : 24 (IV) :2 : and, that therefore," [we must] direct our eyes and thoughts to the entire sum of socal funds and deliberate the reform of the fiscal system comprehensively.]1 Ihe argument for restoring budgetay unity should not be confised with a call for recetralaion. It Is a cal for reclassifying fiscal operations and agencies to bring into the budget "conventional sources of government funds and expenditures, which level of government should control these funds is discussed in Chapter 3. Commercial operations and remaining financial and nonfcial entprises and institutions should be taken out of the budget so as to demarcate the boundary between the general government sector and other sectors.2/ 2.10 The Credit Plan. PBC fomudlates the annual credit plan in consultation with MOF and SPC, taking into account the needs for invesm in fixed capital and working capital, as well as PBC's direct fimancing of the goverment's deficit.jQ/ In recent years, the credit plan has been more involved in microeconomic, sectoral, and regional financing (Annex 3.2), by means of five channels of 'policy lending": (a) allotting credit to supplement budgetary resources for investment In key sectors and enterprises not covered by the capital construction budget; (b) making up for shortls in budgetary allocations to individual projects when act expenditures exceed budget appropriations (Annex 2.4); (c) providing support to loss-makig .1 Xiang Huaichag, op. cit., p. 38. 21 For definitions and covenons, se I, A Manual on Gownmem FPnance Stadusde, Wadhngt D.C., 1986. JQI Wodd Baok, nancial Sector PiUda mand 1wimnaDeWopemw, Wasnton, D.C., 1990. .20 - SOBS; (d) fnace agricultural inputs; and (e) cleing npriso ears. The crdit plan als involves allocation of credit at prerential Iterest rates to some regions and sector while It is not clear how interest subidies are met by the budget. PBC allocates crdit to ecWalized banks which make project funds available, particularly to PCBC for capital constructieon lending. Banks must also finance out of their own fund "key construction projects"; ter apprais, tecical ftion projects; and exercise choice In fimding smaller project. Thus, the credit plan operates as an etended arm of the State budget making it diffictlt to assess the sectora and regional allocation of funds 1I1 or the efficiency of their use, whie conceaUng the true size of the public-sector fiscal deficit. 2.11 Debt. Mobilization of credit by government units is directly linked with the scope of the budget. The financial requirement of the general governmet should receive special attetion for policy-making; different forms of classification are essential for different analytical purposes (for example, by type of debt holder and by type of debt instrument). In China, the lack of clarity regarding the budget's coverage leads to ambiguities about the nature of the obligaions and gutee that emerge (for example, sovereign versus nonsovere foreign debts). Ihe government's financial requiemen need to be coordnated with those of nonfinancial enterprises for conducdng short-term monetary policy. However, a comprehensive picture of the governmen's oveall demand on aregae savings very difficult to obtain due to: te dispersion of the approval process for local currency borrowing; the unconvenonal forms of borrowing somems used at the local level; the forced nature of some debts nrnse "tiangula debts')lZ; the institutional weaknesses and lack of coordination of centa agencis (MOF and PBC); and the speed with which securities markets develop and the embryonic state of the regulory frmework. In addition, while government routes budgetary loans through specialized institutions (PCBC), the debt servicing from these institutions to the govment is not cleady reflected in the budget The lim information of the budgetary authorites about fiscal extrabudgetary funds hinders assessing, the financial needs of goverment units and their repayment capacity. Ways of improving the efficiency of SOE use of credit is yet another aspect requing the attention of the Chinese authorites. The existng system does not provide a separate recording and analysis of SOE use of borrowed fumds. J, As part of the preparation of a consolidated budget for the nonfinanci public sector, this issue should receive atenton. ,/ The govement explcity recogn that the ftate budgt and bank cedts and ls an two cnnels through which tho Ssta mobilz and istnibus fiuds,' i -Budget Plang, Implemtation and Contol at Vaius Levels of Govenmnt," MOP, Backround Matrial One, Aprl 1992, p. 2. See discussion of the provicid allocation of the cedit pl in Anex 3.2. W Debt that aem often duitangl with govefmet diretd credit (often from PBC and in excess of the credit t ). Thse credits miugt becom contnget hablities of government with finaa setor when such rescue opeaions ae used to addre solency and vo merly tmpory lquidity probms IV Th low en on governmnt asses was mentiod as a major sour of concem for th budgety authores. The fir step in adressing this i is to establi a regular flow of _iform and financ ndican nonfnncial pubhc etrses either te Stae Asset Bura or to MOP directly, in parallel from the rpotg u ; o linoe ministries resonsible for SOEs. -21- (b) Clity 2.12 aarly poss the most serios problems In the areas of. (a) cover oA/ (b) clsfication and definitio, nd (c) prentaon of the budgeL 2.13 Casf aon d Defdions. Prior to the 992 budget, revenes were classified by source: (a) taxes (the main source); (b) receipts from enteprdses; (c) receips fo maor projects; (d) miscllaneous recipts; (e) subsidies to cover SOE losses (treated as a negaive revenue); and (t) domestic and international borrowing (treated as a budgetary eveue above the line, not as financing). Expenditurs were not classified by the naue of the transaction (economic clsification), but rather by the purpose of the financing prvided (functional cmassification). In neither case, was a systematic distinction made beten curnt and capital expendiures, and pooling major expenditure items into homogeneous and consitt groupng was problematc(Chaptser 1 and Annex 1.1). No cros-assificadon ofen was p- 2.14 1 !PS entatn of the Budget. Deficiencies in the traditional budget pe led the authorities to begin significant reforms in 1992: * lhe 1992 Budget. lhe regulations for the management of the State budget stpulae tha, stting with the 1992 Budget, its presentation would chan "from a sine budget system to a dual budget system." classifying expeni according to their "nature (that is, 'productive' or "nonproductive") into the regar budget and the constuction budget. The new presentaton applies to the State budget only; below the central government, adoptiag the new systm s inialy optional. The Aorities see three main advantages in the new, experinme system: (a) it establishes a clearer disincton between sources and uses of budgetary funds in each of the two budgets and shows any savings Inthe regl budget; (b) it allows stricter cotols to be adopted over the reglar budget and use of any surlus to fund capital constuction; and (c) it wfll facile "perfoce analysis and supervion of budgetary revees ad expendite to strengthen fIaci control."-IV Newspaper unts of the new system Antiond four expected improvements: Q) the pomotonof capkal constuction, while protcting people's consmption; CH) more transparency in government accounts; (iii) ineased restrictions on _ ; and (iv) incentive to reform SOEs and the investm system. * /AsseIng Chinasl Budgey Refm. Ihe 1992 budget is not a strctura refom but a new presentatio of exstg accounts, and it may not bring about the hoped-for advantges. Moreover the 1992 budget stil difer signficaty from interona practice: (a) (planned) losses from noproductive and productive SOEs from current and constuction revenues, repctdvey, are nette out; (b) "oproductive constucion" is classified as a cuet expendire; (c) amorton and interest paym are classified together as n WAI The man uus deolig with cover wo leady dscussed above, under th beadng Unity. W Wan Bini Repr on dw hy8 1 of rhe Sw Budga for 1991 aad dweDA*S w Baefor 1992,5t Sso Of teo 7tNPC, Bejing, Marh 21, 1992, p. 9. -22- expenditure in the contuction account, while national convention considers Iter paymen a currn ependiture and amortzaon as (negative) fnancing;If/ and (d) revenue under the constuction account includes the curent account surplus and tax-like revenues that are eamarked for capital construction. The deficit on the constucdon account has been identified, and recourse to domestc debt borrowing, externa borrowing and domestic bank borrowing has been indicated. The transition to the new presentation might prove difficult, becas the new format is not mandatory at provincial level and below. The mission was informed that some provlnces (Jiangsu) bave adopted the new presentaion at every goverment level, while others (inner Mongolia) will continue with the old system. Additional effort wUi be required from MOF to consolidate the State budget. 2.15 Taxes and enterprise Income are intermingled in the current account. Enterprise Income is based on a negotiation with each enterprise. Expenditure classification still combines administrative, economic, and functional orientations. Before the dual-budget system, ent expenditures were broadly ditingished from capital ewenditures. Most capital expenditures for fixed investment were shown as one item (capita construction); other expenditures with a substaial capital component were shown under technical transfomaion. In practice, some mainnance expenditures appear to be included under the capital constuction budget, and some minor capital expenditures show up under other heads of account. An attept was made to Identify the predominat nature of expendure (current or capital, developmental or nondevelopmental, cenal or local, or both) under each of the detailed 29 heads of expenditures. Thus, under many expenditure heads, only current expenditures were shown, but capital expenditures were lumped together mostly under two heads. Under the new system of State budget presentation, expenditures under the current account include expenditures on nonproductive capital construction (Box 2.2). 2.16 In the 1992 Budget, under the costuction accoun expenditures, all the productive capital construction is lumped as a single item, but it is subdivided into expenditures financed by domestic funds and those finacd by foreign borroving. Tbis is essently presentational. Other expenditures for fixed capital formation, technical transformation, and city maintenane are also shown as a lump-sum figures. Paradoxically, provision for working capital is included as part of the construction account. Geological prospecting is brought under the onstruction account, but it is not clea whether as an investment in developmental activity involving current expenditures or as a%^quisition of capital assets. Repayment of debt and payment of interest are again lumped uwfer the construction account. Thus, expenditures under the new State Budget presentation, viz., bween current and construction account appear to be based on a distinction between development and nondevelopmental expenditures. By Implication, expenditures on education and health are treated as nondevelopmental. The adoption of the IW standard economic and functdonal claifications used by most countries would still allow China J6 inerest is defined as paynt for the use of boroed money (curret expenditure). It should be distigsd from repyment of borrod money which is classfied as (negative) financing. So 1MP, A Maa on G e,w, astFiaw Sta, 1986, p. 179 and p. 175, Table C. -23- Bon 2.2: NEW PRDVIaTIor OF IM319 hrAll Btmos? LA (Mlhosts of Yua* lad.32.2 Tax 304.2 P #adnet tax 67.2 Valus.eddsd tax 47.6 DaalBussa tax 61.6 sotat uCOAPala lome tax 55.1 Sta esipiasd3u _men tax Si Coaue;tvoe out.lna. lcon tax 10.6 T*iff 20.0 Agricultur tax 73 5.7 Sb*A0jFusld 123 0dw 10.9 2, TdlD _ e 257.6 Jnadzmatloa day ovul and social Welra 102.0 Al, 1b^ad cesisemsy 123 77.0 couvenaldmom wochl tobf 67 3 sumsamdevftdioue '~~~~~~~~~~~~77.2 De_ene 37.0 AdMnilnlaetive expenditure 36.7 paO. uabads 33.6 Odom 26.7 Caitl and lca andy find 2.7 &Win ~~~~~~~~MA !MIM fiaw carted =Gouut 44.6 Special coosatauiouo revenue 47A Key ocoasamcdoa Saud for .mwV and tasnepoatafic 16.2 Cn y_abdatenance "ad acttx 10X9 txondlrctiofcaptal a m 3.1 Occupatio ta b of culivatd lud 1.7 2.8 2. Total _pmdlbaIe 1S4.4 Co _aatnacion finaaed by domedc hand 378 Co=Uwdm Onsund by fordp bomwing ~~~~20.6 T ad p rodudw JP innavadon 18.0 _wma"We 1.0 Gorpr_ 4.3 13.7 4.4 9.3 Cit mat 13.2 Developuteaa Sand for Ieee developed areas ~~1.8 1.0 43.0 34.7 8.4 A S~~~~~~~~~~~~~~~2 Doasgc debt li_sg 38.0 Exnal g 20.6 dom sfic 20.8 L inew _nfib a sym wa Irdc 1994. houses Miulatry of Finance ?1on)_ -24 - to distiguish, for example, between developmental and Onondevelopmental expenditures.1/ Recmedatin 2.17 'he immediate actions would consist of: (a) establishbig clear criteia for determining the boundwy between the general government and other sectors, including diffrentating the government units into t three categories mentioned above; (b) removing the tasactions in the nature of govemmental activities from extrabudgetary funds and bringing them into the budget while ceasing direct government intervention in maMging extrabudgetary funds owned by SOEs;UI (c) reviewing the extent to which eamaring of fisca extrabudgetary funds is to continue; (d) introducing clear accounting and classification systems to sepaate policy loans from budgetary operations, and budgetary loans from caital grants, and developing a measure of the deficit of the nonfinancial public sector and its financing. 2.18 Budget Coverage. The management of the budget-coverage problem requires action in several stages: (a) First, government functions must be classified by major programs/activities, starting with the broad objectives and moving progressively toward the narrower ones. lbis should correct the current mixing of economic clasifcations (for exaample, capita construction) with functional classifications (for example, education). This vital first step entails more than simply reclassifying edudng functions. It requires decisions on which functions government should carry out and where to place those functions-on or off budget. The coences of these decisions must be considered on relations between agencies performing a particular function and the budget).12/ The fuAture system should clearly define budget coverage, showing the tanactions between govermental units and the budget, indicating the support to be given to semigovrment_ units and government transactions with enteprises as owner in a format distict from those as a sovereign.2Q/ (b) Second, present efforts to refine an economic classification of government expenditures should continue (see below), in particular, the distinction between J1/ The exclusion of the fist three item (that is, general public sevices, de and, public order and safety) fiom total ovenmet expenditures i tl IMP fu nctional daificaon (op. cit., Table B/C, p. 193), would pvide an pproximation to 'productive expenses. W / An arms-length relationship of the vous levels of government with their entpiss se antWi foDting this pcess, for examp, by adoptig forms of indirect cnol trough holing co_mi a nd a mor active role of the State Asset Maagemen Bura. 12/ Three types of quetons need to be answered: (a) who should be the poducer? (b) who should pay? and (c) when does the governmet have a compative advata as an economuc organiztion? The answes to these questons should indica the nsue and best form of goenment intervention (for example, govnment financing or government prvision). See L. Stiglitz, op. cit. 2Q/ A list of institions attahbed to th budge hould clarify its coveg. -25- current and capital expenditures. The payoff of having the fiuctional and economic dcsifications of government expenditures (cross-classification) for economic analysis Is very high and should be one of the objectives of budgetary reform. (c) Third, a Nonfinancial Public Sector (N}PS) Account consolidating govement and State-Owned Enterprises would be a useful supplement to the government budget The NFPS Is based on the criterium of government control, intead of the critedum of goverment functions. As long as China's govemment exerts consierable contol over SOEs, and SOEs remain instruments for implementing government policies, public sector accounts would yield a better pictue of the overal governent establishment. Ihe public sector borrowing requirement (PSRB) would be more relevant than the government deficit in detming the overall macroeconomic stance of government. Financial instittions owned and controlled by government should not be consolidated in the public sector accounts, becuse this would elininate informaion on which part of the PSBR is finmced by banks. 2.19 FIscal Extrabudgetary Operations. There will be resistance to restorig the unity of the budget, from spending depments and units. which have substantil (do facto) freedom under the existing system. As a fir step, it may be necessary to either: (a) consider stringent measures such as a state regulation cowled with administrtive mechanisms and time limits to bring fiscal extrabudgetary funds into the State budget; or, preferably but not exclusively, (b) create inceives, like manining earmarking of previous fiscal eotrabudgetary funds to reassure agencies, units and local amnstrations that they wil not lose resources or control over them as it may happen if these funds are grouped with other general budgetary funds. Tne Chinese authorities have accepted the general principle of bringing surcharges ito the budget and have already implemented it for education. This will have to be extended, and all surcharges brought into respective local government's budgets. 2.20 Regarding user fees (e.g., road use), the govment might appropriately consider more offwctive statuoy mechanisms-which need not be under the control of the Price Bureau-whereby the amount of fees chargeable, the purpose and the use are defined and modified from time to time in consultation between MOP and the respective government units. The budgetary support to some of these units would become autonomous and would be determined by objective criteria based on reco ions by possibly new regulatory bodies.ZI/ Identification of cost and profit centers in each unit would be an essential step in bringing about their clear dffrdation (nto full budgetary units, Intermediate, and self-reliant units). MOF would have to play a crucial role in idenfying such centers. Management of retained profits would also be automatically linked to unit differentiation, determination of the accounting practices, and relationships with the budget. Earmarked revenues should be constaly reviewed to keep them within reasonable levels-consistent with the flexibility the government needs in allocating resources. This will be especially importan as government's total revenues shrink in relation to its increasing responsibilities as a result of the reforms. 2.21 Debt and CreIt. The management of debt inred by SOEs is highly decenalized and should not become a direct liability of government at some future date. PBC, W1/ Most couztries hv MT Boardto Reguatoy Commisso for is prpo. -26 - as ChiW bank supevisor, should have enough autonomy to assess the quality of the banks loan portfolios, ordering provisions to cover credit risks fully and writing off nonrecoverable credits. Bortowing by govenment units should be separately Identifed and formally approved by MOF and local finance bureaus; the budget should explicitly recognize these liabiities. Further, since the credit plan allocates debt for various activities and debt financing is often used simply because budgetary appropriations leave cost escalations unfunded, a process of formal goverment should be adopted. Hence, an overall and continuous review of public- sector debt, would be an essenti Immediate step, pending differentiation of governmtal units. Banks and bank supervisors should increasingly see SOEs as fully accountable for their profits and losses, and the commercial risks Involved in lending to them should be assessed with no recourse to the budget in case of defult. L other cases (lending to nonprofit units fully dependent on the budget and mandtory policy loans), should be registered with and cleared in advance by MOP. In the latter case, the explicit guarantee of the government (as approved by the relevat conact authority) would back these lending operations and this admiistrative decision should be clearly reflected as an effective or contingent liability in the budget. For bond issues, a similar process of differentiation and disclosure needs to be developed. 2.22 PoUcy Lending and Poliy Bank. China's authorities have recognized the necessity of relieving the specialized banks of quasi-fiscal "policy lending.' China is considering settig up policy banks" for taking over the policy lending tasks.22/ However, creating specil tutions for policy loans may only slow down the commercialization of the specialized banks, and designation of an exiting instttion as a policy bank would seem preferable. urthermore, part of the policy lending-notably the subsidies to loss-making SOEs, and the subsidies for agricultural procurement-are a fiscal task, and should be on the government budget if Chia' authorites choose to contnue them. Only the development lending for long- gesating projects, such as inatructure investment, could possibly be handled by a policy bank, but such a task seems to ft in well with the tasks of the existing PCBC. 2.23 ClasIfication. To align budget classifications with the increasing market orientation of the economy, to make it useful for economic analysis and policy-making, this report recommends the following: (a) FWEh agency (units, departments, local governments) should modify its accouning system to show every revenue and expenditure under separate and redefined current and capital categories. A simple reordering of the old classification is not sufficient; new definitions need to be inoduced. (b) On the current receipt side, taxes can be differentiated into three categories: taxes on income and profits; taxes on international trade; and taxes on goods and services. All taxed income (to the extent it is recurrent) should be brought under the current account, whether earmarked for capital construction or not. (c) Subsidies to cover enuprise losses should not be netted out and should be shown as part of the government's current expenditures. Any capital transfers to SOEs should be classified as a capital exedit. This presentation is Z2/ Indeed, with the decision to set up an export-import bank, China has created such a policy bank. -27- essential to maintain the gross nature of taxes and the true nature of the goverment support to SOEs (acquisition of a capital asset). (d) The State budget should show the repayment of loans to the government as a separate item. (e) On the expenditure side, the capital budget should differtiate between acquisition of capital asset, capital tranfers, loans, and the acquisitIon of shares. Acquisition of capital assets can be divided Into land, equipment, etc. Every agency must break down curmt and capital expenditures In ths formaL In other words, the capital construction and technical tnoron budget should be disaggregated by finction and agency. Ihe proposed pr tion would enable the Introduction of cross-classifications (combin ftional and economic classifications) which are so Important for analysi. (f) For analydcal purposes, the economic classification should separate current and capital expenditues. Currt expenditues shoild be shown under tee main categories: (i) wages and salaries and purchases of other goods and services; (Ii) interest payments; and (iii) subsidies and current sfe (to SOEs, to other levels of government, and 'other'). Each agency wiUl have to identify its main programs and separate expenditures into current and capital by purpose. Each agency should have objectives, programs, and cost centers. Simultaneously, a code should be assigned to each account (with clear definitions and instuedons) to guarantee consistency, facilitate automation, and serve multiple uses, including distusingbetween SPC-approved expenditures, developmental and nondevelopment expenditures. (g) As regards debt, the budget should classify interest payments under cumrent expenditures; amordzation under fiacing. (h) The Chinese authorities should enlarge the role and responsibilities of MOF and the finance bureaus to include reponsibility for implemeing the proposed improvements in the budget system, and these entities should revise their organization. MOF should resolve possible conflicts of interest resdting from MOF's triple role as owner/financial manager. tax collector, and super (Annex 2.1).2WI (i) MOP should establish separate unit to work on classification of budgetary transactions. Mhis unit wi have to organize its work into working grop at grassroots (units in central departm and at local level). O) As budgetary accounts are being redlassified govemment unis should be identifed as govenmental, semigovermmental, and commercial. Aftr agreement on the broad direction of the reforms, an organizational study would be desirable of the core agencies partcipating in ihe budgetary process. 211 In partiular the of. Industy and nComm_m Trd. sd BuMar, and A4d_m -28 - (k) Finaly, government stadstics cannot ignore tfie needs of other users and other data systems. 'he Mission observed Utile coordinaton between MOF with SSB and SPC, which has serious Implications in tem of Inconsitent information, noncomarable classifications, and great difficuldes In producing the new System of Natioal Accounts. An rinsttonu task force should examine these issues and offer solutions. C. Tim BuDGrr AS A Poucy TooL hsues (a) Complsice 2.24 Expendture overruns are the main cause of missed fiscal targets (Chapter 1). The persistent fiscd deficits which are higher ta budgeted increase ma onomic instability, as they are financed by Increased recourse to PBC or Bank fianced, or by running arrears in the case of local govement. Furhermore, expenditure overruns undermine the efficiency of expeitures, as spending agents have no incenives to save, since the overruns will be financed anyway. Although cost overruns occur reguarly, capital investment budgets do not seem to conan contgencies, and the overall reseves of one to four percent (state budget regulations) are not nearly enough to cover overruns such as the 8 percent of 1992. Aldiough the budget has effecively isolated itself fom the consenes of overspending of SPC directed projects, by means of a lump sum allocation to SPC, the macroeconomic consequences of overspending on such projec still remain. 2.25 Exenditure overn are caused by a combination of projection errors, wilful underestimaton, and lack of incenive to stay witiin the spending limits. Ihe lack of linkage between economic planning and the budget, and the absence of asrfence documents of multiyear planning and the budget seems to add to projection errors and overspending. Wilfid underestation seems to be prome by the "base number method* budgetig principle, through which expenditures within the base mnmber are approved more easily than those superseding the base umber, regardless of their merits. Furthermore, MOP does not seem to have the insttionl capacity to addrs the costs ove s. he regulaons on sta budget maagement specify legal sanctions agaist overspendig, but the terms overspending and mspendmg seem to be reserved for the budgetary contibtion while overall cost overruns may go unsanctionA, as long as they are financed by other means. Finally, although ther are mid-year budget reviews within the adminis on, overspending does not have to take the political hurdle of a supplementary budget, and the overspeig wfi become generaly klown only at the time of next year's budget. 2.26 On the revenue side, the exting fiscal system is chracterid by contcting with local governments and entpises, tying to provide incentives for collecftng and paying taxes.ZAI Tax cotactOg with tpri es makes compliance nonparent and increases tax administaton and control costs. Field inspections and postng of tax personnel are practced to increase the cost of noncompliance. However, this nonmlebased approach to tax 24/ See BRD, Oa Rombsg Inrowrw,eata fTel Rko, November 27, 1991. -29 - adminitratIon cannot Solve the principal agent problem2X which leaves ample scope for noncompliance. In addition, a negoted tax syst exacerbtes the inrinsic conflict of ineres among the ttiple roles played by govement agencies as owners, tax collectors, alnd supevisors. This exarcise of power creates plenty of incentives and oppormnities for tax avoidance (and corruion). Finally, the tax conac Inhibit tax buoyancy in times of accelerated growt, when predetermined contracted taxes form an even smaller share of GNP. Tbis contrasts sharply with the tax system in other counties, where progressiveness in the tax system forms an automatic stabilizer of economic cycles. (b) Reglaity 2.27 The Budget Cycle. The budget cycle in China currently consists of: (a) analysis of the current year's budget implementation and projection of the next year's revemnes and expenditures. lhis exercise is carried out in the third quarter of the current year; (b) determining next year's 'control quotas' by MOF and SPC, using current year preliminary figures ('base method"), adJusted according to the (annual and FYP) plan's targets and taking into account the proposals of local governments ('factor method"). After their approval by the State Council, quotas are communicated to the ministries and provincial governments, forming the basis for compiling unit and local budgets 2W (reported to MOF in the unified budget forms deibed in Annex 2.1); (c) compiling the draft State budget by MOP, followed by the Stat Council's aproa; (d) disussion and approval of the drft State budget by NPC's Fnmanci and Economic Committee and submission of a resolution to the NPC; (e) implementation of the aual State budget; and (t) examination and approval of the f}nal account by the State Counci and the People's Congresses (Annex 2.1). 2.28 The Fisal Year. The fiscal year coincides with the calendar year (January 1 to December 31). The budget becomes effective as late as April, which reduces the actua span of the fiscal year to nine months. Thus, the budget is executed in the first quarter on a 'tentative basis.' The delay is attributed primarily to legislative lags, but there are also difficulties in lig next yeas projected budget to the short-term oudook for the economy so that MOF can establish control quotas early enough to finish consultations with the centra departments and local governme and draft the budget before year-end (see following section). Sinc control quotas' for the next year are largely an extrWolation of the outcome of the cur year, MOF has Indicated that it would be very difficult to anticipate the timetable for compiling the draft budget. The abbreviated chronology of the budgetary cycle is shown in Box 2.3. 2I Arising wh th principal aget (that is, center) is unable to peform a certain task drectly (that collecting tan or exrcise owneship rights) and must cnd / incentives to motivate a thid party (that is, pvinces or eteps m gers) to perform the task in the picipal's interest (that is, mxim tax collecis and enterpisoe profits). 2 Control quotas wae determined by *w caa for povic with conracts 'lu local rves and expeditues d p-ccg t' rvn g (dhat i fixed percentage and added r e sharing conact), as a forin of setting minimum tax collonDs and maximm e _m&tures, settn a tax colecto targets, and playig a mwciio« fuam4n For locaitie with contc no linking reveues and expenditures and con_tr g at difernt level (tht is, sha v at goted es and revig t fers fom t ceoter), MOP sets oldy the cntrl quota for e, lettig locaities deterine their expenditures according to *'heir financia capabilities.' - 30- BOx 2.3: INDICATm BUDGET CALNAPt: TMETA FOR COMfG TwE STATs BuDGET LL Juno - July W issues contol figures over major indices of tho next yeas economc and social development. July MOF issuos tho next year's budgetay rovenue mn expenddit contol quotas. August - Ptovinces, autonomous regions, and municipalities provide suggestd Sepmr figus to SPC and MOP, based on the contol quots. Around October SPC and MOF hold joint or sepazate national plannin and financa cofence to detei next year's budget targets tougb conutatao and dehlbato After October Fint Iteration: budt are comiled fm the bottom up. Second Itraion: budgets are couied fom h center dow to he lowr levels. Bdem the end After collecting genal budget plans, MOFPoompilesth e draft Stat budget of the year plan to be submitted to the Stat Council for rview and approval. Arud March MOP, on behalf of tie State Council, submits a report on last yes's budget etecution and next year's draft ste budget to to aud NPC sesn for apval After approval, the budget plan acqures legad oc and beoomes applicable to provincos, iepartm_ b, and uuLt. LI he calendar show is indicative. FPom year to year, the actul pwcedure may vwy. (c) Opmroio Adequacy 2.29 Under the present budget system previous year's expenditures play a large a role in the deminaton of next yea's expendiures, and less consideration is given to program and macrecnomic effects of these expendit. MOF seems to lack the informtion and the caaity necesary for assessing the uselness of expenditures budgeted by line ministries and unis. he AAPRC ocuses its audits on compliance, and the economic efficiency of public expenditus is of low prioriy. 2.30 E planning is lttle hrnked to economic developments and in price deveopments. Actual tax revenues seem to be closely linked to the Plan growth figures, which determine thu tax contracts, irrespective of the actual economic developments. -31 - Reomndation (a) Complance 2.31 Three areas of compliace were flagged for ation in the 1992 budget speech: the role of finance departments; the need to control and increase the efficiency of public expenditures and resource use; and ensuring compliance with laws and regulations. To deal effectively with these concerns, the following actions arecommended: * Compliance and Information. Compliance is a key area of the budget system. Compliance is possible only when clear-cut, equitable, and realistc rules formaly provide for flexibility and exemptions. Compliance, therefore, is not related simply to administrative discipline but to the pull of incentives faced by goverment employees and agenci is which may or may not reinforce the norms embodied in rules and regulationw. At the same time control procedures and the content and tmeliness of MOF s infomation flows and anadytical capabilities need strengtening. Field i pections and siting government functionaries should be complemented by better stascal methods of reporting and supervision, off-site surveillance and targeted sample inspections. * Autonomy and Control. Incentives must be provided for compliance by spending agencies. A number of countries (for example, Australia) have found successful approaches, where MOF acts as an investment banker in negotiating with spending agencies and making financing decisions. Ihis approach has the advantage of putting the 'burden of proof' on the shoulders of the spending ministries and agencies, forcing them to present objective and well-documented cost/benefit analyses for review and approval by MOF. At the same time, built- in inentives should encourage agencies in raising the efficiency of expenditures. MOF provides incentives for agencies to do their own evaluations and allows them enough flexibility to reallocate funds among programs within their budget ceilings. In this process, forward expenditure estimates become a valid benchmark for cost effectiveness of reallocating funds to existing or new programs. * Transparenc. Ihe fute system should introduce a rulebased, transparent system backed by detailed instrucdons, audits, and appeal procedures, as well as penal clauses and policydriven incentives. The immediate actions consist of identifying major areas of noncompliance in tems of specific departments and development of a rule-based system. Systematic review of fiscal laws and regulations to identify the reawns for noncompliance should be adopted so that more rule-based systems can be introduced, especialy in budgetary relationships with enterprises and local governments. In the medium term, revenue-sharing arrangements have to replace the contract system with local governments; and a rule-based intface has to replace contcing with enterprises. For this to be possible, changes in rules, procedures and budget process are necessary, a dauting task in an economy so large and complex as China's. -32 - * Budget Manual. The role of fiance departmen can be more lay defied and steghu ed by publishing an Itrucdon book or manual to explain how the Stalt budget reuatons work. * Expnadlture Review Committee. Initutional mechanisms (ike some countes' exenditure review commitees) should b established to deelop metods of rising the efficiency of expenditure and, more broadly, the public adminisdan Such bodies duld as rowmmend budgetay refoms and expenditrcontol mechim. Ihe budget, in the last aysib, b itsef an ieve mechanism for triggering nspo by public ag . Pardlr at on should be given to the rdative inceties being given for or aganst complac. lhe NPC role shoud be enaced for studying, approving, and wmending e State budget.27J Ihe Fia and Economic Committee, with its 28 knowledgeable members in fIani matters, may be in the best position to examine the budget in detail. Strengthening this cmits powers for meaingi examination and discussion of the draft budget is highly (b) Regularty 2.32 The Peoples' Conesse usualy do not change draft budgets, but it camot be presumed (nor is it desirable) hat approvals will continue to be profrma.2/ The *alt'rnves are either to amend the dates of the budget cycle, including the NPC schedule, or to change the fiscal year. Whle thaere is no nally ecommded budget yea, and prces vary, it is reomded that China's future system ensure completion of the budget- proval process by the time the fisal year begins. A chronology of the budget-approval proess and an illusratio of the steps in Japa's budget proces are shown in Box 2.4 and Figure 2.1, where the budget cycle starts a year before the parlimentary approval of the budget, whie in the Netherlands (Box 2.5), the cycle also s early in the year, with draft budgels ubmitted to MOF before April 15 of yea (t-). T Chinese authories should mmdiaty beg studies considerig the two options-aging the fisal year or amen th timiog of te preset approv proe-so t the process is completod by the time budget bes to operate. To meet at deadline, MOF needs to initiate the budget compiladon v4uch earlier tha June-July of the previous year as at present. fl Mm fin appoad by NPC involves cosuideaon of the budget by a body h a v mng of 3,000 ddelgt Te Standing Committee las about 150 nembe. 2Vj II 1991, to NPC oeq n exoinry Y 1 blioes reducti in the fis deficit of tho _odraft b -33- Ilgr. 2.1: STEWS IN JA?M'S BUDGE= PROCESS muAletzi.e 3ntqo l Ageuelee 1*aoace (331) Cablat 11t Pepaatloa sPteprtos GulAsli for the letbiate budget ZRequeste (Cablt WlerstandUM (Jul 27, 1990) 0~f =d~t (Ieari. and RequeetRo"s (by Aug 31) (Sol 1tIJe) Cabinet decinion on I the general pryciple Ptreparation of of bugt ccaapilti |ItO! ~ ~ o budget druftt . g ~~~~~(Mmc 22) - . Presentatis o to otitiction of the inute to the I gotiati t am (Dec 24) I ILi te1 X~~~~~~~~~~~~~~~~~~~~~~(p 11)rl 1 I I I I IIDoe, 2-211 t- I I I - - ' I lparati of 1 | w - ,1 1 lisovezot draft| & Cb1mat deeXlotl I I I } I | (Dec ~~~~~~~~~~291 ff { 11 ~~~COlpSLAtlt Of 4 I 1 ||i~~~buget docut8ts l l l l ~~~~~~~~~(J8an 0,1991) 1l l l l l ~~~~~~~~~~~~~~~~~~(Apr 11) Now AJlhougb tdo Con_aw C_mwitle of Both Hoses was held a*er the House of Couv s recin oan Apg 11, do Houe ou no }ach a *oosas. Thm th FY1991 budget beme effec_de u apoed by do Ka. of Repmesetatlew on March 14 becu of its supedor powVr of deCision oo budgets Sourwo 2w qou D g is laW, Japan's MOP. Tokyo, 1991. -34 - Box 2.4: JAPANESE BUDGET P1OCESS I/ L Cadar o Nain Pon* of Decsion-Makg and Actvitie The fisnd year (PT begins on April (year t) and ends on the following March 31 (yewr t11+1). Months before 8atat of FY(t Man events and acvites (Year t-1): 12- (April/May) Minis (and agecies) sart prepafion of eports an et yeas 10-09 (June/uly) Cabine discu_ses and agrees policy guidelines on budet request, prepad by MOP. 08 (Aupst) Ministries submit revenues, expenditure estmtes and budgt requess to MOP. Ministies submit request of Fiscal Investment and Loan Proam ( ) fimds. 07-04 (Septerbel MOP holds budget hearings with Minities and agencies about Doeember) budge reuet (including PILP finds). MOP negotiaons with Ministries take place hierarchically, from lower official lovels to Ministerid level (if nece). 04 (December) Cabinet decides on budget estimates (including FILP progm), t g aso into accout the macoeconomic results for th nw fisca year. (Year t): 03 (ate January) Cabint submits budget documents to the Diet. MOF delives two budget speeches to the two Houses. 010 (Marcl/April) Diet debates and approves the budget. 0 (I Apri) Stat of fiscal year (t) 2. Tb e Anna Budget Cyle Each ministry and agency bens estimatng its revenue and expenditue for the foUowng fial yea budget aoud Aprl or May 2 each year. They m PrePare rePor On expt revenue, equ exditurs and other necessary matters und their sdwtian and send them to MOF no later tha August 31. .1/ See OECD, Dhe Coxol and Maagew of Gownmew Evxdkre (Pais: OECD, 1987) and Budget Breau, MOP, De Japwe Budgd in Bilft 1991, Tokyo, 1991. ...Coninued -35 - ox 2.4 (cone.d) Ta June or July, the cabinet discusse and decides on gudelan for the MOF budget ruquest. lhe guidlines contain intuctions to be followed by acub nstry and ageny i nakng its cwlulatiolns. Acording to the cabinet agreemt on the guideine for the estimate budget request for th next fiscd year, the. guidele include a pere adjusme for cure and capi exnditure, m_ued against the nomia expeaditue level under the prvious budget Excluding national debt service cost and local allocadon tax grants, the guidies apply only to about 33 pert of total general account expenditures, also tag ito account the obligaoy character of sevealW prognms. a prparng its budget, each minstry decides on requests after earing the opio of to Policy Affairs Research Coumcil, consisting of membs of Diet of the rulig paty in lat AugusL The MOP d budgt for tdq_ budget incoludes unallotted fin a rwoures for the 'reviv netaos'- gthatnewnegotitio td place biger in ie y. Ce guidelis for budget esoimat requests do not conta unallotted financia resour for the Wvival aegotiaios). After the emination prwess which includes hearinp and reviews by MOF frm Septembe to eardy Decembe, MOP prepares the drft budget and sbmits it to miies and ageocies in late December. After the nodficadon of the MOP draft budget to ministries and agence, th 'revival negotiations' betwee MOP and each ministry and ageny tda place. For neary a week, fina decisiomakig place tho negoas at tr offici levels up to minit level. ITh MOF drft budget includes unallotted financiad resoures tt a distributed to spcific programs in th course of the reviva negotat I total amount of expenditue in th draft does not change after the revival negotiations. When negotions do not reach a slmentw, are sthen n by the the meaor exectves of the rumg LDP; e g l, cam of th Policy Affas Rearch Coumcil, and th chma of the Executive Counil. Tlh final decision is made by e pdme minister, if the second level of negotiadons fails. At the end of December, the Economic Planning Agey pruaes th Outlock and Baic Policy on the Natidon Economy" for the following fiscad year through coultaion with relvt minis including MOP. Thle cabnt must appro outook At narly te Sam tim, 'Genl Piciples of Budgt Compilaieon ar decided by th cabinet as well. Pinally, tho government approves fth budge for suibmson to e Dietl Usualy the de_l n proces of both houses ends in ealy Aprl. (c) Opemional Adequac 2.33 Various measures can then be taken to improve the budget formulation d implementation systems and improve the operadonal adequacy. Among these later measurs, the following should be conidered to improve operational adequacy: 2.34 Reduce Inranm . The "base method' should be deemphized and the "fiactr method' given a heavier weight and new meaning in forecasting x years revenue and -36 - Box 2I: FIbMATION OF T0WE BUDGET TMe NE LANDS Teclnal Asectf M1a Estmate Mutyear eods (MYD) (fo conal puposes ase reisted for bt pit and nontax mvus of ceta government. Ther s noi to a mfdium-tamB sev as a meas of conol fr MOP, rater than as proclaid resourc for the g tea Scabiet ( ~ ~ ~~~ . . - a 3.0 . a JIM a a . . - . _a 1.1 J7a~ - . . . - . . _ 0.5 Sama (j. _*i _ . . * . a 1.2 -lX a , , , *a. _ _ 1.1 ft1S ( _elu 1959) - . . . . . 0 .5 7_n Euielt . a . . . . _ a 1.4 - , , a , _ , _ 6.1 ibet . . . . a . 9. |sad~~ ~ ~ ~~ - . . - _ * . 15.3 Usabssh - - . - - - - a 7.4 Tom= ~ ~ ~ ~ aa a-6.4 a1.4 so e e D a MA n aaae a a thu c , _ as Nanig, N , et. & lio payresithi 4. peba of Wuas rolun_ce. Sia: ail n Wafib 1992 . so - Ugure 3.: IMuRGovhEM NTAL FINANCIAL FLOWS centralrI P, Budget U L2 * C. Cs Local 4 BudgetC Central-local transfers C1 - quotas subsidies under fiscal contracts C2 - earmrked $rants C3 - compensatory payments Local-central transfers LI - remittances under fiscal contracts L2 - special taxes L3 - compensatory payments 3.64 In additon to contractd reeue sfes, th ctal govenmet sends earmasrkd grants to provinces for specific uses (C2, Figure 3.5). ITese include appwropiatio fo capital consactionprojects, price subsidiesf u ain consumption, social reief fnds, nd spal subsidies for heath and education of the poor, minority, and border provices. lhe thid mpoet of cen trasfes to provinces comprses 'final account subsdies," which ar en P191t1ofy, payments for the loss of revenues from profitable local entepriss tha have bee transferred to cent ownership (C3). Centalization and decentraliation of aeter is proposed by the rsponible line minist, and approved by the industy deartm of MOF. Alouh o finl accounts ubsidies shold guaran budget neuay of ownership _ m m n Rogow"; toed gov_ments complain ta al governmenwtamks the yawted brad, " to enterpises that have big growth potenti. Final accoua subidies ae tmoray: whnve revuesing contracts ae d , they ar fcted Into the bas nmbers, so that the fina ows Are either added to Cl, or subtracted from Ll.5/ 3.65 On the side of provincial tansfers to the centra government specal taxes accrue to h cen goverent. Introduced snce 1988, these taxes include two on cgattos and one n liqor (L2).$1 The third cowmponet compimry payments to the ceral Bovrnmen x enterpise ta have been transferred downwad to local onol (3). Like the MI f t1V fows ny as be negative, if th tus nd en ws losuw_mking. Mw Revnue from tm txe do not ene te _ndo contrau dwing but ashard sqmate. - 81- downward paymn f recentlized etp , ihese upwad paymens ar to be teplacd by adjustmts In Ihe revenuesharing conac upon their revision. 3.66 A fia tafer f*m local to central gOVeme is central govennts "borrowing from the provices. In 1982, centra govemnt 'borrowed' more ta Y 8 billion from provinces, but the Incurd debt was written off upon recontractin in 1983. Although cet govrmet has repotedly promised tih It would no longer borrow frm prOvin, ShanghaI, for eample, regItered for both 1989 and 1990 a 'loan' of Y 0.4 billion to cntral sovrmen./ 3.67 The aistence of these multiple flows upward and downward significuay altrs our of the budgetary process, since resources avalable for expenditure at eac level of government are net of all these positive and negative flows. At the ceral level, then resources ar: the sum of centr fixed revenues, provcial remittances, special tax , and loca p y paymentg for enteprise trasfers; mlaus the sum of quota subsdies to deficit p .vinc, eamared gas to provinces, and compentory paymen for oeneise transe. Only two of these components (Cl and Li) are governed by th revenue-sharn contract. 3.68 Table 3.8 presents the breakdown of central-local tasfers for 1990. Tough the fiscal refr of the 1980s, revemearing tansfers have gready diminished In nce. Figure 3.6 illustrate the sharp decine in net transfers through the revelue-sharing system, which fell from 9.3 percent of GNP in 1981 to 1.9 percent in 1990. Thble38: EgMAwTD CEmRAL-IACAL TRANSnMS, 1990 (Y billion) source of trnfer Asount Quota subsidies under fiscal contract. (C1) 8.1 Earmarked grants (C2) circa 30.0 Compensatory payments (C3) 8.5 Total 46A. Source: Derved from MOF data; see also Figure 3.5, this 3.69 By 1990, quota subsidies to poor provinces accounted for only 0.43 percen of GNP, a mer 17.3 ercent of the total transfers from the centra goverm t to local 51 ZJaimia Li, 'Provincialqxm to Fisca Doe talmiow- Mke Cas of Anhui,ns" and Shanha, pae penac at dh conf_une 'Tunaforming Concepts io Wodrkn Redaies, Theo azs Uni y of HaSg Kg, September 9-11, 1991. -82 - Ilgure 36: DwINDLuIG SHAIUNG Central and Local Revenues and Net Tax Sharing Perct of rvenu 108% le gm ~~ ~~~~~~_- ....... ...... ----.......... 25% _==- = ~~~~~~~~~~~~~~~~~~~~~~~~.-................. ,...,., -_ 1981 1982 1983 1984 1985 1986 1987 1988 1989 1"0 YeAr CeWtm fied re. Not shard rnues Local rOnu aftr S,_ Ti _o A-.., gove-nments. In contrast, earmarked grnts have grown steadily, and in 1990 they comprised nearly two thirds of total central-local transfers (Section 3.C). Similarly, while litle detailed inrmation has been provided on the size of compensatory payments, they also appear to be significant-almost 20 percent of total central-local transfers In 1990..U/ For its own expentures, the cent government eliminated its dependence on remances dtrough the revenue-saig system. A net receiver of 5.3 percent of GNP in revenue from local govemments In 1983, the cetal govenment has become a net provider of 0.8 percent of GNP to local governments In 1990. In contrast, local governments' own revenue is falling increasingly short of expenditures, covering only 77.5 percent of its budgetary expenditures, down ftom 87.6 percent in 1983 (Table A-1.2). 3.70 By increasing the central fixed revenue base, the central government reduced its dependece on shared revenue over the 1980s. In contrast, local governments' own revenues fell progressively short of expenditures, thereby Increasing their dependence on earmarked grants fl( This estma is reached by subtacting the amount of earmarked grants given to the mission from total eamaked grant and fnal account subsidies derived in Table A-1.1. - 83 - from the central government, and increasing their incendves to develop edgetary opeadons. E. ExPENums AND INTROViNcLa EQu4TY Cross-Provindal Expenditure Patterns 3.71 General Trends. Actual expenditures equal the resources avaiable for local use after all transfers. Table A-4.2 shows provincial per capita budgetary expenditures from 1978 to 1990. Two trends stand out: (a) expenditures grew rapidly in all provinces throughout the period, and (b) there was a relative convergence of per capita expenditures across provinces, although the differences In the level of per capita expenditures still remain large. 'he range in per capita expenditures had narrowed between 1978 and 1990, and the top-to-bottom exendtre ratio was reduced from 6.90 to 5.95. The coefficient of varian in per capita provinc expenditures also decreased secularly, from 0.70 in 1978 to 0.55 In '990. Table A-4.2 shows that rankings of most provinces changed between 1978 and 1988, especially in the middle rank. However, the same five provinces appeared at the top in both 1978 and 1988 (Beijing, Shanghai, Tianjin, Qinghai, and Ningxia). At the bottom, there is a good deal of inertia: the lowest five provinces in 1990 came from among the lowest 1978-tier (Anhui, Henan, Sichuan, Hunan, and Guizbou). 3.72 Some Recasted Figures. Correcting the budgetary expenditure data for SOE subsidies (that is, countlng them as expenditures), a somewhat different view emerges: (a) dispersion, as measured by the coefficient of variation, is larger L 1987 and 1990; and (b) the degree of dispersion increases between 1987 and 1990, and the coefficient of variation ris from 0.62 to 0.68 CFable A4.2). Including fiscal extrabudgetr e for 1987 521 reduces the dispersion in expenditures per capita, suggestig tat provinces wih smaller budgetary resources had relatively more extrabudgetary expenditures. 3.73 Determinants of Expendithrs. Variation in expendiare level across Cbinese provinces can, to some extent, be explained by demand and cost factors (Annex 3.4A). Regression analysis reveals that provincial budgetary expenditures per capita are strongly positively correlated with provincial income per capita. 'his could be due to the high income elasticity of demand for public goods, but also due to the budget constraint of the provincW government, as tax revenues are equally strongly correlated with Income per capita. Highly urbanized provinces have higher per capita budgetary expenditures, as urban areas need more infastructure. The higher the provincia population, the lower the expenditures per capita, a fitding that may indicate economies of scale in the provision of provincid public services. A higher share of dependents (students and pensioners) also increases budgetary expenditures per capita. The higher the share of SOEs in the gross value of industrial and agriculural output (GVIAO), the higher the expenditures per capita tend to be. Given the average practice across the Chnese provinces, and controlling for income, population and urbanization, Tibet, Beijing, Qinghai and Tianjin have unusually higher per capita expenditures. To explore the differences in expenditures across provinces further, ihe major expendit categories will be analyzed below. i2/ 1987 is th only year for which pvicia fisa l c;:ul~ ~ ~ ~ ~~~~I ws a iIi4i a IV I. i Ii 's~~~~~~4 toC v~* i jlj~~~~_l -8S - poor provinces try hardor on socal expenditurs (devoting a larger share of ter budge) but are faincrasinly bed the richer. Iable 3.9 PsoVNQAL CES AND CC 1978 1990 cnn Mc =_8CC CRI Per capta I a 2 x S i Per capita Ii S IS S (Y) (Y) ToD live .l,jlt 28.7 11.9 53.4 65.3 147.9 23.4 17.3 40.7 Sh}ghai 27.8 11.7 40.6 52.7 128.4 22.7 10.7 33.4 Qiuai 24.9 13.4 34.4 47.8 106.9 23.5 18.5 42.0 Tianjin 24.3 12.1 41.4 53.5 80.4 21.0 8.5 39.5 ingusa 22.2 13.7 39.6 53.3 79.3 25.4 9.8 35.2 Bott Sichuan 7.8 21.3 31.2 52.5 27.5 25.9 5.9 31.8 Huamm 8.2 20.9 29.0 49.9 29.4 27.1 5.6 32.7 Anhui 8.3 21.5 23.1 44.6 33.7 25.5 7.0 32.5 Ruen 8.4 17.7 31.2 48.9 34.3 23.1 9.3 32.4 uduzhou 8.5 18.5 27.7 46.2 35.5 26.6 6.6 33.2 Jiangsu 8.5 17.5 - - - - - - Slhsdong 8.6 19.3 - - - - - - National average 10.4 17.0 33.9 50.9 46.3 26.3 8.9 35.2 Note: CESH = Cultur, Education, Scdence and Healt; CC = Capital Conston B Budgetay Exp_diur. 3.77 The budgeay des of CESH mad capital spending are inversely related. Table 3.8 sows hat, compad to nation average levels, the top provinces in CESH expendiu devoted less of thek budge to CESH ad mome to cpital consbuction, while the botom provinces spe ey more on CESH and ths had les for other categories, cluding capita consuct This fact has twr mportn imications: (a) lower lvels of speng on capital constucdon may doom the poor provinces to a lower growth path; (b) dnc econok growth I necey to gnert revnue growth, pooe prvinces wil coni to lag behn in fisca capct, starving their socia and econmic progrms; (c) becse CESH spending repren investme in human cpital, the poor provinces could be locked ino a vicious circlo of povrty. Ptk Subd 3.78 Tablo A-20 shows the shae of toa sbsidies in ta e diues by provi for 1987. That ye, the nional avea for h shae of provinacil na to e sbidies w 26.1 pet. Aco prvine, h share raed from a low (of 15.5 pece in -86- Nigla to 36.8 pecn In Beijing. Because most subsldies go to ura grain conumption, which are relatively fixed per capita, povinc with a high pcentagp of urban population would be expected to spend more on subsidies than lesn urban provinces. The presence of two of the the municipalities at the top of the raking, Bsefing and Tianjin, confirms this epectation. In ad n, the three highly urbaized northeastern provinces of Liaoning, lDin, and Heilongjlang, also devoted more ta 30 pen of their expeaditures to subsidies.0/ ITe rapid growth of subsidy costs tbrough the 1980 Imposed a growing burden on local budgets, but appwed to have had no clear regional income consequences. 3.79 The growth of interregional disparities within specific expenditure component during the same period highlights the central problem in expenditure assignments, wher too much is devoted to nondiscrelonvary expenditures at the local level. Becase poor provinces have too litde left to finance both development and social services, they are forced into a choice that perpeut their low-income status. Paradoxically, the bulk of mandated, wondiscretlionary ependitres is absorbed by price subsidies that are devoted to protecting the basic needs of the urban populace. By targeing the better-off urban populace, subsidies exacerbate urban-rural inequities and are regressive In nature. Does the Fiscal System Equalize? 3.80 The lack of personal Income toatlon and a nationwide socal security sysem, and the restiction on people's moving to places where they can improve their income or receive more and better public services, puts strong demands on China's fiscal system for both Income --~ ~. .. . .... ... - . - ... , .. .4 .$ _V 87 - Ugure 3.7: PER CAPITA EXEN uiEs AND REVENUS IN CHINA'S PROvINCES, 1990 (Yuan) 0 a C to IC xi I . o~~~~~~~~~~I - z l__|_ ___ ___ _. _ .I! ! ! ! a | 4 _ l m :S E__ F Q '5 Qlhd (which has a large Ta population) and te minotity and botde am of bnet Mongolia, Nhin1a, and Xnjlng. 3.83 ExcludI Shangha and Tlbet, the rlaton btwe ad asfs per capIh and hme per cpita is weak Figure 3.8 shows te relation betwe per capita income and per capha transfer to (+) or received from (-) the center in 1990. he ton between Income and trasfs b positve, but the slope of the curve Is small and dispersion around the trend Is larg. The wido dispersion gives the redistribution arbitrary features. To Illustate: the pwvinces bunched around a per capita national Income of Y 1,000 receive widely different amounts of transfers. Henan (income Y 872) receives only Y 6.9, whereas Gansu (income Y 919) receives Y 52.0. Flgure 3.8: TMOUGnTM BN UDGET, 10 (Ym) Per aple truiur Ur.qb Us bu d 100 50------------4.--------- 0 4100- - --- ------------- 4150 -------- -------------- -200 --------i--------------- o 500 1000 150 2000 2500 3000 3500 Per capita Incme N*W bOdIo T"h an Sha. Samr= TUs A4L1 and A-il. 3.84 Over the 1980s, the surpus provinces sharply reduced thirk transfers to the center. Shaaghai-,frm Y1,128per capta in1983 to Y613.5In 1990,TlanjIn, from Y231.1 In 1983 to Y52.9 in 1990, and Beijing, fiomY 216.6 to Y68.97 over the sam perlod. In real terms, the declne was steeper: for example, Shanghai tranferre in 1990 only about a dthid Of the amount transferred in 1983.(& Over the 1980s, however, the amount of central fixe revene Inrased strongy (Secton 3.3), to becme the main source of the trasfers. The W In 1987 prices Shangha transferred Y 1,413 per capita in 1983 and Y 480 in 1990. -89- prwv overal aro "in d* towatd e ceer, wi& Y 14.31 per capita in 1990, agaist a surplus of Y 22.15 per capita in 1983.(/ 3.85 ba d Transfers, More Emaig. Tansfers from the central government to lcal govnments inrased sharply (15 percent rea growth in 1985-90), but the composition shifted from quota subsidies to earmarked gts (Saction 3.D). In 1987 prices, quota subsidies declined from Y 9.4 billion in 1985 to Y 6.1 billion in 1991 (Table A-1.3); In 1985, the quota bsidies stUI ondtted 44.5 percent of all centra-local asfrs, but in 1990 only 14.6 percat (Table A-1.1). For prvinces for which earmarked grants cod be derived, tL same patten held: quota subsidies declined and markod ra increased (Table A- 15).fI However, the most str&lng fe Irin the data on emarked grants was its dbtion across 1wovinces. Relatively rich provinces libe Shanghai ad Guangdong rceived subsanti amouns of earmarked grants, Y 171 and Y 57.5, respctvelv, per capita in 1990, whfle poor provinces like Guizhou and Guangxi received Y 23.9 and Y 28.3 per capita, respectively. 3.86 l he distution of eamarkd grants a be partly explained by the impact of centrl policies across provinces: the inefficient and regsive system of prico- subsidies and the subsidiion of enepris losses is cofinanced by the center through emaked grant. Old industrial centes, where SOEs dominate, are the main beneficiaries. A second cause may lie m the high cofinancg requirement atached to cetral grants for CES}I projects. Due to the poor provinces' weak own-revee base, they camot always comply with the cofinaig requiemet and might therere forgo cen eamked gas. Differentiation of coficing requiremens by province seems to be the exception rater tan the rule, and only i limited msount of central budgetary fluids is eamarked for projects in poor rural and minority mueas. / Thus, although total central-local amsfers increased, the growth came from hie earmarked grants, benefitg primarily the richer provinces, and increasing the urbnrural divide. 3.7 p Equity. Whie China's mass povey has been dramatcaly educed by the strong ecomic growth of the reform periotd, and urbn povery has virualy disappeared, 100 million poor remain, concenaed In remot and mouno counes, in the Northeast, the North, and the Southeast province.ZDI For those -pockets of poverty' with a weak fiscal capacity, ovinc fis transfrs are crucial. Although few dat on subprovmcial finance are published, evidence from the education and the health sectors (Boxes 3.1 and 3.2) suggest may of the social seri are not provided or are of substndard quality 671 Ovral transf per cita we found by divdng th central-loca t duoug tota populaton. / Eamare grants we derved from total flows fom or to tme center, corctd for the flows fom th fiscal conact. Beca th con0dtc flows are not known for each provinc, a complee distribution of ard grans camot be derved. §2/ The igurs for 1991 for the pecia finds am Fund for Educato of Poor and M iity Aea, Y 22 umiion; cerl contuion to te Develmnt Fd, Y 800 millon (tho fimd iv aedit to poor areas); th 'hee Wat Fund,' Y 200 million; Basic Nods Fund, Y 24.8 million 2W See Wadd Baok, -Mina Sraes fir Reducng Povt in th 1990s.' Repot No. 10409- CHA, MW 1992 'I5 U l@III I0~a0 hA~~~~~~~C a % tz jJ IX itt j~~~~~~'ja - - - - -- - -- - -- - Wo * |X 'liii IIii w 0 Iii} [1 31iX I II II -91 - resowce outflow smaller. Sbaanxl, a poor province recving a reaively low fiscal trasf receives a much larg nofscal tafer ./ 3.90 A significant part of the nonbudgetary redistribudon takes place though the banking system, which has become more and more a substitute for the budget over the rfom peiod (Chapter 2). Although more complete data ar lacking, Annex 3.2 provides some prediminay ev by looking at the loan/deposit rado in the baking systm per pvnce. Ihe ratlos teveal ta both the poore provinces and prince with a strog SOE dustril base receive credit aocatows well above their deposits, yielding above-average loan/depit rat. Thus, thei reour envelope s extended, over by PBC's cedit allocaton system. 3.91 The price system Is a second Important fct. Most state-et pices are blued against provinc with naurl resou and, in genal, beneWl to dustrial provi . As a reslt of price reforms, howsev, this chanel i gradually losing It Importance as a tool fi rebution. In conast, a third administtve chanmel has gained Imporn over the refom period, "policy subsidies," for exmple, permision from the centr to form *sci oomic mon or receve some other preferential tratmet. Plans to extend such police to th Inner provins have been proposed. In the meDani, th conceatio of hies zones in tho coasta regions has been tinsumeta in their strn growth pebrmace. 3.92 In summay, tie equalizin poperides of the fiscal system declined over the 1980s, and trasfers from the central goverment to the provinces are only weay relat to Pe capita income. Redstributive quo gra have given way to eu armd ssidies, notargetd toward poor provinces. Off-budget flows play an mporta role in h rdiit of resources among provinces and between them and the center. 3.93 The Chines authorities point out tha, during the 1980s, fiscal capacity in eah prvice has iased strongly and that a convergence has taken placo. Whil ft Is true the increase in the poore provic fiscal capacity has bsee achieved not by fiscal ri but by a strong icase in their oveal tax rate, while the ta rate of richer provine has declined Cable A-6.3). Wheter convergence in fisca capaci and hxpdb as en place is, to a certain extent, irrelen The present systm of quotas and emared grats fais to accut proprly for relative cages among income of provinces, canot accozm-odate for infladon, and sems to result in an arbitay distribution of fiscal capt among provinces. ThIS sysem will thus ustain or accentuatvincW icme diffens. F. FICAL AcmmvB DyUNCONAL OUTCOMES 3.94 Strain on the fcal system comes as no sprise, given t depth dspeed of reform in China and is to some extent tolerable, given the phenomea owth tat th enow,ra decentrizaidon has caused. However, the rductance to dmp the integovernmeal fiscal system in the light of de facto devolution of power has led to a number of avoidable problems, and to actions by centra and local government ht further W This nysses only as n ia As China retches fiom the cenba pl systena sour flows can no lon 0 be as intendod o af cer polc. o fow of _m es to Shagi f*om outid h fi t may bM d proviocs uea in attting forig investMen -92 - undermined the SYSem and macoeconomic stablity. lhb most impotnt ones are lised beow. 3.95 Mia of Eapuidita' and Revenue Assinments. Revenes and ependitures ar easy mismatcbed at subutonal levels of govement. Chins intevrr_ fiscal redao have ben largely revenue-driven. Little atenion has been given so far to the ependtr side, especly to mcing expendi with revenues. At the county and township level in pcula, venue sem indequat for the ependitues assigned to those levels, leaving ent govenment funcdons neglected. 'he mismatch has been accntuated by the cents policy-induced chane In expendiure responsibilities and his somedmes undermined ceta govenmet policies, for example, In educotion and inastuctre. CESH In poor regions has been indequat and is failing faither behind richer regions. Ihe long-term implicatiom for Interregional income dibuto are negative. 3.96 UncetatY. 'Me ceta governmen's acton throughot the proc often helped to undermine efficiency ad equit of the fiscal system and created a set of adver incentm lead to dysfnctional outcomes. As revenues fell in GNP terms, and as the budget deficit grew, the centra gvemnt rcted by expading its claims on reenues by Increasing the cenr fixed taxes, protetng its own finces at the expense of local governm finances-borrowing" from locaities. 3.97 LocA goverm considered predatory repeat aempts to renegoti nveneharing contcs, to "borrow- local surpluses, to reclim oweship of enterprises in high-ow indutrie, and to radjust die subordinaon of ckies with big revemn sutpluses. Repeated unilatea rcntrctng by the central govement has blurred and undermined contrt incentives, reduced revenue collection well below its potental, and made fiscal planning at local level much mor difficult nder the ad-hoc chwacter of the ermarkod grant (see below). 3.98 Reduced Reditributins. As fiscal resources in the revenue-haring system stated, the centra govemnmet aowed the redistdbutive demevt to weakel By fixing the quota subsidies to poor regons in nominal tms since 1988- 5hngn the policy to increase them by 10 perc a year-it allowed thekr real value to erode with inflation.ZZI Instead, fiscal d on now takes place maiy through ad hoc, earmed grns, itroduced piecemeal as neded and ofe poorly targeted and too small to meet basic needs in the poorer regions. Moreover, the arbiWay nature of central grt allocations has gien way to extensive eOtadons and rent-eeking behvor by the localities, tyi up valuable administrtve resources. 3.99 Extra- and Off-udgetary Activities. Besides increased rent-seekng, local governments have responded 'to the fiscal squeexe in other ways hat ru counter to reform objecdves and that unmne the fiscal systm itself. These local "reactions Inlude inefficient, nnransarent,and iom lle egal attempts to: (a) expand the local tax base; (b) tap e sp' etrabug funds; (c) push expendiure off budget, sommes fianacd by 72 Under t 1985-87 oobtrwt, quuts subidies were scheduled to inas by 10 pecen a year, to provid, for rl growth afte cmatce.. Due to he, ecers own wrning fiscd difficultie, t polcy was abandoned in 1988. -93 - ldlrwt local bonowbg, or paid for by dhe oetprise and private secors; and (d) Impose public sevfice f and levies-excessive n some cwases, inequitable I others. 3.100 A vaiant of the above streagy is of pushing governmental expenditur -off budget. IIJ Xuzhou City in Jiangsu, most sff salaries In township-level finace depatmns ae paid by towship eonomic commissions from off-bget industrial funds.7j/ Earlier mission dicusn of educational ficing snowed tha provinces are increasingly seelkig nbudgetary funds to supplement budgetay apoprions, Including uning factories to ear Inomes for schools, collectog 'social contributions,' etc. In poor provinces, especily in the rural sector, many costs a passed along to parents, by charging tuition ad fees and obliging the parents to help sbsidiz the m i teachers' wages. Although cost recovery can enhance efficiency, the level of paet's cotribtion can fom a seiu access barrier. Cont os fom frms consftw a form of taxatisn not based in any law. From a policy formulation perspeive, the exta- sad off-budgetary atvities blur the view on what govenment aculy does, and how it is fhanced. 3.101 Iefftdaiet legional Competition. Evoking the localities' Interestin epding the local ta bse was a goal of decentaization of fiscal responsibility. However, as the budgety resourc were no longe srMicient to achieve the goal (as seen in the virta off of cital construcion from the local budgets), the Intruments used for this purpose became inpropria. Distortd price and tax structures channeled resources to the 'wrong' secos and policyguided bank loans more and nmore replaced budgetary finance (Annex 4). Tbh diversion of resources led to excess capacity in procesing industries, whee revenue potential is high epecially whOn loca monopolies can be established. Becausm the price and tax stres ar uniform ntionwide, the regional composition of industry has been duplicative, ne reurc are channeled to the same high-taxlhigh-reven Idustries eveywhee. To mimize revee grwth, pvices have even subsidizd high-cs producs in high-ax Industries like boveag and tobacco.7: The conegence of Industri sucturme acros regions exceat he tendency toward p m and hinders the development of a national keL ho ot the reform decade, local revenue dependence on the financial health of local entepises has induced local officias to protect their resources and markts. A rest has been _ outb of 'wool was,, 'silk cocoon wr,' and other bariets to rource os between regio, icuding limis on 'capital outfow' through the interbank market. 3.102 Pscal Predato . Under severe fisa pressure, many local govermn hve snumbed to tappig etprise fiuds by impog a variety of legal and illegal lees clu educsional surcges, utiity surcharges, 'envionmenta fees,* and the like). They have also shifted costs to enterprises for responsibilities ranging from education,75 road-building, and ZI To hold down admds_wtive costs oan C budget, only 160 of the 660 p_Ison int de ciWs fil ierwhy (with offices in 180 townships, 6 conties and 1 district) bave bee put on te payrol for admi s_wdve perond (x_ygAg hia). 21 In 1986, Gn paid out sbd alingY 29.85miiontosupportpoductiof skaypoint, t producs in he and cigartte induies (Guag4 Niaj, 1987, p. 5S4). Zw( oPr exapl, TVYs pnd abot Y 20 bloon educatio d wela outsldd e bud which is m dw 40 perca of bdrty expeditue on ctegories. Se Speh by Vie. Premier7b Tiesynn (Jun. 199). -94 - environntal protection to hosting banquets for govemmental functions. Lacking a strong lega fameork to protect ntises from fiscal predation, such practices have prolifrated despite myriad rqulations promulgated to prohibit them.261 Governmeat predation over enterprise funds violates the spirit of enterprise reform, which aims to I crease enteprise wtonomy and to isul them from govemment interference. Predation blurs the Incentives fkr entpses to respond to market forces and therefore impair allocative efficiency. 3.103 Tax Expenditres. Another set of problems are tax expenditures. AItough the prvinces have no formal powers for seting tax rates, local governments can set the effective tax rates on enterprise profits via enterprise profit contracts.17 They can aso offer tax expenditres to enterprises and affect the size of the taxable income, by stipulaing portons to be exempt from taxation, by setting the rules for pretax repayment of Investment loans, etc. In exteme caes, they can exempt enterprises from indirect taxes, by waiving part of sales taxes for e rises in financial difficulty. As long as the revenue-retention rate is less than 1, loca govments have an incentive to pursue evasive or avoidance tactics to keep resources at home instead of sharing them with higher levels of government, thereby trning budgetary funds into extrabudgetary fiuds. Funds 'stored' in this way in the enteprises can then be tapped for communt projects.21/ 3.104 Allowing lower level governments to make tax expditues under revenue-sharing arrangements creates a set of highly regressive inentives. Makig revenue-rch provinces more generous in offering tax, expenditu than poor ones has adverse Implications for regional income growth. Rich provinces can 'share' the costs of tax expenditures upward to the centra government; poor provinces have to bear the full cost themselves. Moreover, bse of the int pressures on their budgets and the restriction on borrowing, poor provinces canmot forgo present revenues for fimre revenues. The increasing use of tu ependit ultimately results in lower effective tax rat in rich areas than in poor ones, and In lower overall taxation than desired. 3.105 New Directions? In partial recognition of the problems in the present fiscal system, the Chinese govermment has initiated experimenta reforms, to be implemented in the cities of Dalian, Wuhan, Shenyang, Qingdao, Chongqing, TaIJin, and the provinces of Llong, Xinr iang, and Zhejiang. Labeled 'tax-sharing system," the erimena system reins much of the present system: revenues are divided into central fixed, local fixed and dshred revenues, and the local retentbon rate will still be driven by base number expenditures. Ihe main difference is the assignmew of different taxes to different levels, with the most Imptant change the local SOE income tax, which now becomes 'local fixed,' thereby inring the incentive for local protectionism, although reducing the incentive for-ilegal-tax exemWtion. 7& Donald C. Cbae, The Leal Bacoud to the Behior of Sumed Eteqrses,' Decmber 1989, processed. 27a By sefing a lmp-sum payme ta iclue iome ta, ahdjusment tax, and profit rmitne, tax conts essentially undid all the efforts of separating taxes and income in the ta reform of 1983-84. 7j1 Xu Riqig, 'Som Thoughts an Local Govenmet Fis Contracting,' Caima JiagiW inance ad Trade Econodics), 1988(11), pp. 53-54. -95 - 3.106 Ihe stated aim of the reforms Is to increase central govenment tax buoyancy. At the same time, however, the expenditure assignments remain basically the same, and the shortage of funds at the local level is potentially aggrvated. The experiment does not anticirate a true local tax base, with tax base and rae demined at local level. Tax revenue will therefore only by accident follow expenditue levels, ad regulat adjustment of either the local fixed taxes, or the shaing rate and accompanying local lobby activides is to be expected. The base maber expenditures are determined by past expenditures, retaing the incrementalism of the old system, and forgoing the oppornity to put expenditures on a more radonal basis. Moreover, no fiudamental reconsideration of expenditure assignment over levels of government has taken place. The experiment therefore seems insufficient to deal with the more fundamental problems in China's fiscal system. G. RECOMMEATONS 3.107 Ihe dysfunctional outcomes of the fiscal system that have occurred over the reform period blur the fudamental gains of decentralization. Instead of recentralizing-an option that may no longer be open, given the rising power of the provinces-the Chinese government should realign the instional stucture to give localities t'*.e right incentives for providing public services in an efflcient and equitable way withoutjeopardizing macroeconomic stability. 3.108 China should embark on a more fundamental reform of intergovernmental fiscal relations than the incrementalism and expeiments thus far pursued. Fiscal reforms can only be brought to success if all parties gain, which is more likely the more comprehensive the reforms are. Budgetary reforms and changing intergovernmental relations should therefore go hand in hand with a broadening of the tax base, the creation of a local tax base, a rationalized tax collection system, and a reestablishment of central government's macroeconomic control. The evidence from the past with fiscal reforms suggests that the experimental, regionally limited piecemeal reform method, successful in many areas, may not work with fiscal reforms. A gradual nationwide implemention of an established reform plan encompassing the various aspects of fiscal reform promises to be the better method. 3.109 China's authorities should make clear assignment of functions among levels of government, with efficiency of service delivery and equity among localities as guiding principles and avoiding the exisng duplication of functions at various levels. In assigning functions, Chinese authorities should account for the de facto devolution of functions that has already taken place over the reform period. 3.110 Some govemment fictions, especially those determining China's internal market, should be recentalized. Regional development should not lead to inefficient duplication of indust investments. The centra government could, for example, pursue an industria policy through the investment orientation tax, interest rate subsidies, or other measures. Local govmen should be allowed to purue regional development goals, but impediment to goods and fictor movement should be strongly opposed by the central authorities. In this contet, it should be noted at the policy to restict the movement of labor enhances local protectionism. To counter local proonism, central govemment should fill the instutional void with realm of ompetition, re ion, company law, and contract Uw in order to restore the unified marke. -96 - 3.111 tanlo of MaoenoIc FUonso,. Control over govrnmen funtons cucial fbr macnomic stabilty shoud be recentralized, ncluding control over g govennt borowing, and overall tax burdean Io resmin Itergovemental flows of goods, production fators and income gives centra govemment a comparative advange In tax collecton, and a Nationa Ievene Service ) is highy desirable. Such a reentrizaton of conrol over adminluon of revenue sources (borrowing, taxes) may well cncde with a decentralization of the revenue accrul from these sources. The expenditure asson shoul be the basis of a new revenue asigment, and the expenditures neossay to perom the assigned functions efficienty should be covered by the revenue assignment-provided the overall fiscal envelope is sufficiendy large due to China's tax reforms. The efficiency of govemmen service delivery depends to a considerable exet on the autnomy a level of gent has, on the firnnaAl incentives it has for cost savins, and thefore on the meam the expenditr are financed. 3.112 Own Tax Base. China's authorities should assign each level of governent a 1h snd own tax base, with which it fl7ances funcions for which it has full responsibilides. China's auftories should assign local goveraments a true local tax base, and exploit the opportunityfor more accountability and the incentive for bear and cheaper public services such tax base gives. Own tax bases.canfu ermore serveto absobvariations in pendieneeds witou a dresome renegotation of iteoveromental fiscd rdations. Finally, extension of the local tax base contbute to break the present deadlock in the system of intergoven fical ltio. For local genment, proper taxes would be an excellet own tax source, which i ths far hardly exploited in China. -However, the bordiationpinciple in income taxaion ditor incentives for local governmert, and should be abandone in further reforms of the tax syste 3.113 Tax Saring and Eqtion. A full separadon of tax bases in China is probaly not feasible, nor desirable with respea to efficiency of collection, assuming hat the NRS wiR be established. Ihe fiscal gap between expenditre needs and revemnes from the own tax bae can be filed by tax sharing, genera grants, specific grants, or a combinaion thereof. To accommodate for the increasing divergence in economic pefonnance among regions, and the consquential growing divergence in fisca capcity necesita a fisca system with equalizing propties. Horizontal equization of fiscal capacity between localities can be achieved though tax sharing arrangements or through equalization grants. Whaever the meni dshosen Cina's auithorides should design horizontat equalization mecisms that are reslient to Iatin, and can accommode substania shifts in relative welfre positions of localities, which are likely to occur as reforms deopen. 3.114 Adm iniratvely, it is more complicated to incporate equalization p ~ II I Itin in tax sharing arrngemet, and nsualy a separate equalization fund is set up, or central governm allocates funds directiy too poorer provinces. The latter soludon would, hower, tequire a larger share of tax revenues going to central govent, which seems difficult in the prest Chinese suadon. For China, an rindal equalization, fxd could be the appropriate way to strengthen hoontal quization. i closed-end fund could be fled by a cerin pcenta of taxes shared been tral and local pvernmet this earmarg will have no effet om tax effort In the case of a NRS. The fund could for hntance be aiLitred by a body at arm's length of cenal government, with both local and central rreqsntves. Dis se from the equalization fund could be bad on broad indicators of expenditue need, such as population, area and urbanzaion, and on fiscal capacity of the -97 - locality stemmn from Its own tax base. A variety of formula-based grants whemes or ta saing arrangements can be found around the world, but China's authorities should make thte prdmarlypo l decisios on what fauos should be given weiot in the formula, and to what etn fal capact, shud be equalized across localities. 3.11S Lal Govenment ilname Low. The element of intergovernmental fcl reaions as set out here need to be captred in a Iocal Government Finance law. TIhe law would wotk out the constitutonal division of labor between levels of government, and specify the fiacing mchanisms, including the above mentioned tax sharing and equaizati arrangements. The law would In part serve as a guaraee for local govermnt that the NRS wil not be abused to reverse the devolution of the reform period, and woud establish the hweased loc autonomy by means of codifying local taxation rights. 3.116 Inenental Comittee. An Intergovernmental commite to tacle the various Issues in int-oenent fiscal relations should be established. The committee, manned by both local and cental representives should solve the tecbnical issues in InpIeRentg ineg vemental fiscal reform. The committee should make pposas for the ctea govening the eqaation find. They should be in age of dafting the suggested Local Government Financ Law. .The committee may evolve into the administave body that manes the totrprovinca equalizadon fund. 3.117 Exprm . The present tax sharing experiments do not match the fundamenl reisio of iner nent fical relations as sketched above. Moreover, the experimet have not been implemented as designed by MOF, and therefore seem more and ore Irrelevan for te evolving situation in China China's authorities sould therefore consider gvn up these altogether. 3.118 Short-Term Quota Tranfer Increase. In the short n, further erosion of equaiztion funds should be prevented. As a start, certain tax revenues could be partally aigned for equalization, equal to the quota transfers now given. The taxes feeding the equaization fund should be income elastic, so that revenue keeps up with overall economic development Resources saved from phasing out SOE and price subsidies could be used to supplement the quo transfers. Itpovincal equity could be further srengtned by differ tng th cofi ement for earmaked grants from the ceal governmnt 3.119 Moitori Capacity. Concuent with the Improvements I ther l, ceta gawarment should strethen Its capacity to monitor local Implemention of national law. Rgular nounced and unannounced ispections of local govrnment policy would iase compliance with them law. Further study on tax effort and representative tax systems could help the centra government Identify provies that are giving excessive tax concessions. The costs of industrl policies shoud be made transparent by declarng in the budget revenues forego and subsidies made In exercising regional development funcdons. 3.120 Earmarked Grants. The fiscal system should give localities more ientves to purse natinal obJectives, eqeally in the social sector. This requires a larger financial c n1hm ft rom centra govenment in the form of earmrked grants. The fuins for such progm can be released by phasing out the Iefficient ear gs for SOE losses and pdc subsies. he eamared grn should be tied to objectve crtia (possibly formula) and shuld come from dosd-d fuds so as to retain conol over cent governme -98 - expenditures. Appropriations for earmarked graSs should be program-based, not entiement based 72! and, once decided, should remain stable for some time to Improve the local environmet for budgetaey planning. Cental earmarked grants should be reflected in local budgets. 3.121 FIscal E budget Funds. Fiscal extrabudgetary funds should become part of local budgets (Chapter 2). To presrve local autonomy over the funds and enhance the cost- recovery incentives for local public services, fiscal extrabudgetary expenditures and revenues should remain separate from the fiscal contract, and be integrated in the local tax base in the future. In the longer run, self-sufficient nonprofit units producing marketable goods could become Ineendent, and consequently be removed from the government' accounts. Fees and levies of adtinistrative units should in principle be treated as general fiscal revenue, which should not necesay remain within the unit that generates them. Some earmarking might, however, stll be instrumeal in increasing revenue-raising incentives. Enteprise extrabudgetary fuds should no longer fall under government control or appear on local fiscal statements. 3.122 Cost Reovery. Without a nationwide system of income redistribution and considering the wide divergence in fiscal capacity from region to region, the cost-recovery principle should be applied with umost care to social services, like health and education, as fees could become barriers to access. Central government should set guidelines for local user fees. Increasing health insurance coverage and introducing education vouchers could be effective in enhancing efficiency of supply, while assuring access to social services. Until such instruments are operative, however, increased budgetary financing of social services seems mandatory. 3.123 Centra involvement in local functions in a country as large as China can only be limited and should be well specified in terms of goals and instrumens. The general principle of unified leadership and decentralized managementa is a strong one, and should not be eroded by a complex system of central-local or central-subprovincial earmarked grants. Such a system would undemine local responsibility for local public finances. 3.124 Subprovindal Information. The central government appears to have litle knowledge on subprovlnca finance which hinders the formulation of nationwide policies. If the ceral government wants to play a substantial role iL policies on the subprovincial level and better target its national policies, it should build up the necessary information bases to address the relevant policy issues. As a first step, a unified system of fiscal reporting among provinces is imperative. Widely published and dealed provincial fiscal accounts would also allow provinces to learn fom each others' eperiences, thereby enhancing imitation of best practice. Combined wih other data, provcial stastics could become the basis of output and efficiency indicators.1/ Subprovincial equity should be further studied to determine whether equalization fumds reach the levels of government that most need the funds. 711 That is, be tied to specific prgam for pecifically set out goals, not as general suppt for local govenmet budgots, which dsod be covered by eiter ow revenues or genal grants. QI A mor acotive exchag of governt peisomel between provinces may enhmce th dinvinination of best pmactie, iras g of efcienicy of govnment overall. -99 - BIBLIOGRAPHY AAPRC. Regulations on Audit of the PRC, Beijing, November 30, 1988. I Interim Regudladons on Penaltesfor Violadons offFYnndanlLaws andDecrees, Beijing, June 17, 1987. S Stte Audit In Ohna, Beijing, May 1989. A_zeman, Joshua, and Peter Isard. "Externalities, Incentives, and Failure to Achieve National Object in Decnized Economies." Mimeo. March 1991. Aronson, J. Richard, and B. Schwartz. Management Polices In Locl Govenent Fnne. Washington, D.C.: International City Management Assodation for Institte for Trining in Municipal Admini, 1981. Babi, Roy. "The Design of Intergovernmental Transfers in Industrialized Countries." Public Budgeting and Finance. vol. 6, Winter 1986. and Christine Wallich. intergovermental Fiscal Relations in China.' World Bank. Country Economics Department Working Paper No. 863, Febnay 1992. Blejer, Mario I., and Adrienne Cheasty. "The Measurement of Fiscal Deficits: Analytical and MethodologicalIssues." Jounal of Economic LIterature, vol. 24 (December 1991), pp. 1644-78. Chen, Chien-ping. "hItroduction of Civil Service System Planned." 'Civil Service System To Be Introduced in China Soon." Hong Kong, Wen Wei Po (n Chinese), December 11, 1991, p. 1. Chen Rulong. Coeporary adnese Fnance (Dangdal Zhongguo Caizheng. 2 vols. Beljing: Chinese Academy of Social Sciences Press, 1988. Chikn, A. 'Chinas Latest Economic Statistics." Confidenti Monthly Report Pts. 1 & 2. China's Statstics information Consltancy Service Center, Hong Kong, November 1991. - Microfonmdation of Macroeconomic Inventory Behavior in Hungary." In A. Chikan and M.C. Lovel (eds.), lhe Economa of Inventory Management. rdam: Eler Science Publishe, 1988. Chines Communit Party Centra Committee, Oina's Economic Stcvure Rrm, Bejing, 1984. Cinese Economi Studes, FaIl 1990 and Winter 199011991. Chu, Ke-young, and Richard Hemming (eds.). Publc enditure HaNdbook, A ude to Publc Po1Wy Issues In Dewloping Counies. Washington D.C.: Inteationa Money Fund, 1991. - 100- Coombs, Hugh M., and D.E. Ieni. Public Sector Fia Managegmnt. London: Chap_an Hail, '991. Csdis-Nagy, Blla. 'Unsolved lsues hi the Principle of Inventory Management In a Sociist Planned Economy with Reermee to Hungary.* In A. Chikan and M.C. Lovell (eds.), lhe Economics ofhwentoy Managem. Amsterdam: EBsevier Science Publishers, 1988. Delfs, Robert. China: Power to the Provinces." Far Eastemn Econondc Reviw, vol. 151, no. 14 (Apri 4, 1991), pp. 21-30. Dimitrov, P. 'Macroeconomic Aspect of lIventory Management in Bulgaria.' In A. ChHn and PLC. Lovil (eds.), 7he Economics of Invenory Managemt. Amsterdam: Eiever Science PublIshers, 1988. Dittus, Peter. 'The Stae Investment System of China: Development Issues ad Optio.' Backgound Pape No. 4, World Bank, Structural Change Project, Industry and Energy Opeati Division, China Deatment, Washington, D.C., November 29, 1989. Ohall, Moheb A. 'ne Meament and Economic Effects of Excess Invetory Holdings in Socist Eonomies." In A. Chikn and M.C. Lovell (eds.), The Econmi f Innty Mag . Amsterdam: Elsevier Science Publishers, 1988. Gong Zbong. Analysis on the Repayment of Budget Investment Loans to Indusal Entepises During the Seventh Five-Year Plan Period (English aslation), 1992. Govenmnt of Japan, Minist of Finance, Budget Bureau. lhe Japese Budget In Br, Tokyo, 1991. Ministry of Finance, insttute of Fiscal and Monetary Policy. Fanial Stat of Jap 1991, Tokyo, 1991. Governent of the Peoples Republic of China. The Drecory of anes Gornme Organs: 1989-90, Beijing, 1991. G,aagx Yeaook [Gungli NIaniani, 1986. Guo, Dalmo. "A Review of the Current Status and a Forward Look o the Puture of Finacl Dist inRelaonship Between the Central Government and Localities.' FbnnilResearch, July 1991. Guo Shuqing et al. 'Study Report on Infrastructure Constmction in Guangdong Province." Par presented at the Inatnal Workshop on Fiancg Ilnvestment In InfRastructue and Basic Industties, Habou, China, March 24-28, 1992. Guo Shuqing. "The Finmcing of China's Development of Ifastructureand Basic Induties: Curet Situation, Problems and Solutions.' Paper presented at Internationa Workshop on Finncing invmen in Inastnrucbre and Basic Industries, Hangzhou, China, Mach 24-28, 1992. Harld, Peter. 'ivestm System Reform in China: The Role of the State Investment Corporations." Wodd Bank Dscuon Paper No. 5, September 26, 1989. - 101 - Heady, Chdstopher, and Pradeep KI a. Taratn in Decn SocWist Economies: The Case of ChLna.* Wodd Bak, Asia Country Departmnt 1, Working Pa No. 120, Juy 19M. Hu HR.. *Estimatlng Ret in Our Prices in 198." BRUao, No. 5, 1989, p. 11. Rua, Rarld, and Anthony Pellechio. Oa Managing Extenal Debt. World Bank Draft Bacgound Pper, Apil 1990. Rubel Pvicina Finance DeatmnL "Budget Procedures and Government Expendittes. Badgound Paper, Wuha, April 13, 1992. Hughe, Gordon, and Stephen Smith. 'Economic Aspects of Decentralized Govenment Stuce, Funcdons and Finance. Eowmi Poliy, October 1991, pp. 426-59. Hunyadi, C. lnventory Rglaton In Hungary.- In A. Chikan and M.C. Lovell (eds.), Ihe Eaonoa of inwnor Mnagena. Amsedam, Elsevier Science Publishers, 1988. HsIo, Kadtine and H.Y. Huang. "The Government Budget and Fiscal Policy in Mainland Cina. Mdalad China Economic Seies No. 5, Chung-Hua Insdtution for Eoonomic R _earch, Taipd, Taiwan, Province of China, December 1987. Inunational Monetary Fund. A Manua on Gowrnmwna Fnuce Statlta. Washinton D.C.: IMP, 1986. Jeffson, Gary H., and Wenyl Xu. 'The Impact of Reform on Socialist Enterrises in Transition: Stture, Conduct, and Perfomance in Chinese Industry." -Jourwl Qf Cmoara E&onomic, vol. 15, 1991, pp. 45-64. JUn S7aengq1ng, Jilin Provincial Affairs, 1987. Ji Megf and Li Xaobo (Ministry of Fmance Budget Management Office). "Tlb Key Pindciples in Stat Budget MangemenL" CaM Gongzlwo SWahimunan, 1984. Kobyasbi, Mlnru (Advisor, ndustr Bank of Japan). "Problems of Fund Prouemet to Strenghn Supply Capaciy of Infrastructue in China." Paper presnted at the taonal Worsop on Fincing Ivetment In Infastuc and Basic Idustes, Hogzhu, China, March 24-28, 1992. Lacy, Robet M. -Managi Public Expendie. An Evolving Word Bank Prpetve. - World Bank Discusion Paper No. 56, Wasington D.C., 1989. Lampton David M. 'The mpleenti Problem in Post-Mao China." In David M. Lamptn (ed.), POlic l a In Post-Mao Chna. Univ. of Califoia Pres, 1987, pp. 3-24. Lardy, Nicholas R. "I China Diffet'? .he Fate of Is Economic Reform." In D. Chirot (ed.), lbc Crtd1 of L _iim Seattle: Univ. of Washington Press, 1991. -. Economic Gh and Driution In Ohna. Univ. of Cambrdge Press, 1978. -102 - Lee, Hong Yung. 'China's Now Bureaucracy?' Paper presented at the conference State and SocietY in China: the Consequences of Reform, 1978-1990,' Caremont McKenna College, Febnuay 16-17, 1990. Levin, Jonathan. 'Measurig the Role of Subuadonal Governments, Washington, D.C., JMF, Fiscal Affairs Depament, January 1991. Li, Renhu, and Jiang Jun. Chen Junsheng on Reforming Admnbtrative Organs.' Beijing Xlnhua Domestic Service in Chinese aMr, December 17, 1991. Lin, C.Z. "Open-Ended Economic Reform In China.- In V. Nee and D. Stark, Remaking the Economic Insil ions of SocloL Ohina and Easern Europe. San Francisco, Calif.: Stanford Univ. Press, 1989. Liu Hulyong, Ma Wanfu, and Hu Guangming. bThe PCBC's Leading Position and Role in Investment Field.' Paper presented at Intenational Workshop on Financing Investment in Infastructure and Basic Industries, Hangzhou, China, March 24-28, 1992. Lyons, Thomas P. Interprovincial Disparities in China: Output and Consumption, 1952-1987.' Economc Development and Cltural Change, vol. 39, no. 3 (April 1991). Ma, Jun, and Yong Li. 'Chinas Regional Economic Policy: Real Effects and Alternatives.' Preiminary draft. McCleary, Wliam. "Earmarking Govemet Revenues: Does It Work?" World Bank, County Economics Department Policy, Planing and Research Working Paper No. 322, December 1989. McLure, Charles (ed.). Tax Assigpnent In Federal Cowarles. Canberra Australia University Press, 1983. Ministry of Finance. 'Budget Planning, Implementation and Control at Various Levels of Government.' Background Paper No. 1, Beijing, April 1992. Coordination Relationship Between Budget Preparation and the Development, Credit and Investment Plans.' Background Paper No. 2, Beijing, April 1992. Genera Planming Office. Zhongguo Clzheng Tongdi, 1950-1988 (Chinese Financial Statistics). Beijing: Chinese Finance and Economics Press, 1989. Naughton, Barry. 'Why Has Economic Reform Led to Inflation?" Cinese Economc Reforms, 197.989 vol. 81, no. 2 (May 1991), pp. 207-11. The Decline of Central Control over Investment h Post-Mao China." In David M. Lampton (ed.), Policy hIpleation in Post-Mao Odna. Univ. of California Press, 1987, pp. 51-79. 'Center-Local Intergovernmental Relations In China.' World Bank, Discussion Paper, First Draft, December 1991. - 103 - -. mpcations of die Stae Monooy over Idusy and Us Reaxain." Mdm , vol. 18, no. 1 (nuary 19M), pp. 14-41. -. 'Inflatin Chins: P8at , Causes nd Cures." Study Par, Join Economic Commit of the Cones of te Und Sttes, Washing, D.C.: 1991. -. 'MCoe Obstacles to Refom in Chins: The Role of Fina and Montay Poiy." Paper prented at the 12th Sino-Amedcan Conference, Taipei, Taiwan, P.C., June 1990. Osebeg, Michel, ad James Tong. 'The Evolution of Centra-Provlncial Fisca Reladons i ChIn , 1971-1984: The Form System.' ahina Quarterly, March 1991, pp. 1-31. nizationfEconomic Cooperaton and Development. 7he ControlandManagement Gowrmentm Expeniture. Paris: OECD, 1987. Owe, J. and G. Panela (eds.). Loca Gobwnwmnt: An Internatonal Penpecdw. Amsterdam: North NIoilnd, 1991. Paul, Samuel. Strening Public Service Accountability: A Concepual Framework.' Wodd Bak Disceusi PaP No. 136, 1991. Prmchad, A. Gt hwnen Budgeting and Exwditure Control. Washigdton, D.C.: Intenaionl Monetay Fun 1983. , nad L. Gamfalvi. 'Government Budget and Accountng SYMS.' In V. Tamer (Od.), MIcal Polies in Ecownom lnT)anslWon, Washington, D.C.: Intenational Monetary Fund, I992. Ernes PEre, and Mai L. Seguiti. 'China: te Exsting Govment Budgt Systm' Interntional Monetary Fund, Fiscal Affairs Department, Washingn, D.C., lanuay 31, 1990. =. '"Chin: Govement Budget System: Issues and Proposals.' Ienatonal Monetary Fun, Fiscal Affairs Dpartmt, Washington, D.C., Jamuy 31, 1990. Prime, Pedelope B. -Regional and Development Implicadons of China's Tax Reform: Iepronci Comparison of Revenue Collection and Distribution." Mimeo, Center for Inernational Researh, Bureau of the Census, U.S. Departmen of Commerce, October 1991. 'Ptoblems and Optm In China's Public Finance." Washington D.C.: Bureau of the Census, Mimeo, June 1991. Prudhomme, Rdmy. 'Urban Finance in China.' Papr prepared for a UNDPIBRD-sponsored seminar in Sha Fderay 1991. Qiao Rou_hag. Prie Sbsdies. Chinese Price Bureau Press, Beijing, 1990. Rabin, Jack (ed.). Puic Budgetng and Fanal Management, vol. 3, no. 1 (1991). Rajaram, Anad. 'Tbree Cites Study: Issues in Industrial Organiation.' World Bank, Daft Back nd Paper, 1992. - 104- Rd, A. Ramdhlag. Ana Plans ad Budgein." b Y. Venopu Roddy nd K Kasairm (ed.), MUlUedd Plannng In IndIe Procs aed Perpu . Now DdK 1988. Reserh Group on Financing BasiC Juduts I Financing of ias DvdlofpIntof and Basic ndustries: lhe oCr Situaio, Prble and Solutions. Pa Preset at the -, -d -- Workho on Fin n Invemen In I and Basi Idusti, Ranih. China, Marh 24-28, I992. "Reuati on Sta Budget Mananz" Biing Xinhua Domestc Seovin Chines, WT November 2, 1991. Shah, Awar. wl1N Now Fia Fedalism in Brzb.* Wodd Bank Couwty Ecoomic D1arn Workiag Pap No. SS7, Wasligton D.C., 1990. Prspectives on the Desig of bntegvnmo Fisca Relations." Wold Bak Country Economics Depatmn Woig Paper No. 726, Washingto., D.C., July 1991. Shue, Vivien. *Beyond the Budget, Fian Oranation and Reor ins Chne County. Modem Chba, vol. 10, no. 2 (Apr1 1984), pp. 14746. Skhuan Jl Mai jan (Schn Eomic Yeabook). Bijig: State Statistical Bur Pulishing House, 1986. Silvman, Jery M. olnatin Menr R of Public Sector Dd_ I Red lMalgI-n mdw Draft 10/3191, Wodd Bankr ares, Ste . Road Fundn Study. Daft Repo Wod Bank Jun 1991. "Stat Coune. Restict Loa n . Hox KonS Mafg o ( in Chinesei), Decmber 13, 1991, p. 76. Sarrett, David A. Fowdado if Pbc Econi, Cambrdg Ecoomic Uandbodk Unid Kindom: Univ. of Camhbidg Pres, 1988. Stat Statistical Bue. Chiadatic Yearboo 1991. Beijig, 1991. Sdgt, Joseph E., at al. The Ecwndc Rok f se State. In nold HA I (ed.), UKL Ba Blackwell Ltd., 1989. Su, Ning. "Spport the BasicIdus Constructim by UtiliUing Household Saings Deposits." Paper presnd at the Intankd Workshop on Fiacing n Infrtmt Iu and Baic Industresnobou China, Mauch 24-28, 1992. SunarastJ, V.; Luo Ci; Bill Nloto, et al. Chin bIse In Deveoping Govenment Socides Mr Domesmtc Public Debt Mangemeont nd Mety Piy Teiques." P 1, Main Repor Wa_Sion, D.C.: h Monetay Fund, Centr Bakig Depr_ent, August 1991. - IS- Sundar", V.; Mc QNy, nd Cathrn McAu , et a. "Cina Jsues In Delopn Govemet SeItls Morket Domeic Public Debt and Moneay Poicy Teclquus." Pt. 2, Backgwd ormation and Techiab Appece. Wasigton, D.C.: nteatonal Monetry Fund, Cetra Baking Depatmen, August 1991. 1at, XInyl. Financial Dficles fo Stae Entepises." In An LuminS (9e.), S&aesaftsjfr Owvrcming Fiaa DLJIOksut (in Chine), hU#nl JLnSl Ord, no. 8 (Aug s, 1991), pp.8-12. Tsuo, Kai Yu8L "ChinsRegionall uqualt, 1952-1985." JoawxQf Cop,utIivEcoa,vol. 15 (1991), pp. 1-21. Wang Bngqian 'Report on the Ilem of the 1989 State Budget and the 1990 Daft Budget.' Seventh Natial PoWs Congroe, 3d Plenay Sess., Mach 21, 1990,' Renml Riao, Apri 8, 1990. 'Repo on theo pleumdon of the 1989 Stae Budget," Sevent Natond Pooples Conges, August 31, 1989, CaWleag, vol. 10 (1989), pp.33-36. W 'itain Budgesty Baance to Sev Economic Dvelopment and Systm Reform.' Cd*ng YauuIb, vol. 9(198), pp. 1-. Reprned in CaJhmtg An , vol. 12 (1986). *Report on the of th State Budget for 1991 and on the Daft Sae Budget for 19.' Delieed at th Seventh Nil People's Cogrs, S Sess., BeIing, Mrch 21, 1992. Wang, Mong. Tough on the Rdbrm of FdaxustrialInvstment in China (Ouline).' Paper presented at the Inentioa Workhop on Financing Investme In T-s trU iu and Basic dustries, 4gdh, CMina, March 24-28, 1992. Wang Song Nian and Tag YunweL Bgetary ACOunt In China.' In A. Premiand (ad), rwnewt Fnanda Managewat, Issues and Cowany Sudes. Wuhington, D.C.: Intenatonal Montay Fund, 1990. Wasylenko, M. 'Fiscal Dece&talization and Economic Delomen.' Pulc Budgetig and Fac. vol. 7, 1987. Wildavsky, Asron. nhJ Polii Oxse dgewy hPocess. 4th ed., chae 4-8. BostonTrorto: Lite, Brown, 1984. Wolson, Dirk J. Publeke Seaor e Ecwm=dhe Orde. Wolte Nordhoff, Gronnen, 1988. Wong, Cbdstie P.W. From &maldc Aeforjbn Ise People's Rpucf Ciana. Report to the Asian Devlopment Bank, Manila, Jamay 199. - iscal Rerm and Local Idustrializatlo Th Prblmatic SNupwing of Rdom In Po-Mao China.' Muran Chna, vol. 18, no. 2 (Aprl 1992), pp. 197-226. - . 'Central-Loca Reatons In an Er of Fical Decliae: t Paadox of Fis ca Post-Mo China.' Oa Qurf (Deceme 1991), ch. 1, a. 15. -106- Wodd Bank. "China: Revneu Mobibation d Ta Policy." Wodd Bank Country Study, Wasigto D.C., 1990. - Chin Finai Sector Polides and iotutonal Development." World Bank Country Study, 1990. -. "The Reftm of Publi Sector Mangem Lessons fm Experience.' Country EconJcs Department Policy and Researh Seies No. 18, September 1991. . 'China: Reforming Intgv ernental Fical Relations.' Repot No. 10050CHA (Gree Cover), November 27, 1991. -. 'Bank Work on Mltuy Expenditre. December 9,1991. 'Ste Enterpries In China: Dealing with Loss-Makers.' VaDw on, Socia Ecnomic Reform Unit (Country Economics Departmen), vol. 2, no. 11 (December 1991). 'China: Education Development in Poor Provinces Project.' Staff Appraisal Repo, Report No. 10010-CHA, Fderuary 11, 1992. 'Chia Stateie for Reducing Poverty In the 1990s.' Report No. 10409-CHA, May 1992. - >'Long-Term Isues and Optons in Health Transition.' Report No. 7965-CHA, June 25, 1990. -. 'ManagingDvlopmen TheGovernanceDimenslion.' ADscussionPaper,August29,1991. Wu, "ang,La, Lou JI-Wel. 'China: Intergove Fiscal Relaos and Macroeconomic Management.' Pap pre_ nted at Senior Policy Sem o terg Fiscal Relations And Macr mic _agem Economic Development Instlttt, 1be World Bank, New D , dIa, Febr 2628, 1991. Wu Xiaotng. ssues on Senhg the Management of Policy Lending." Paper preented at InteratioalI Workshop on Finanng Investment in Infistructure and Basic Industries, Hanlzhou, China an March 24-28, 1992. Xiang, Huacheng. 'A Reiew of the R6fotm of the Fiscal Syste and Overall Pr,jections for the Futr. ainnse Ecwono Studies, vol. 24, no. 1 (Fall 1990), pp. 15-39. Jkmg CWAeng MWI= (Xaglanig Finance Yearbook), 1986. Xu Riqing. 'Some Thoughts on Local Goverment Fiscal Contacting." Calmao Jingli (Finance and Trade Economica), 1988, pp.53-54. Yu, Guan 'On the Advantage and Shortcomings of Financial Contracting System and Proposals for Improvement.' Jlagu Economic Research Cent. Zbho Bokun and Z&ang Hongbing. 'Prom Insvtging the Siton in Ai and Jiangsu Provinces to Disimssing Some Issues in Buiding Xlang(zhen) Fisc Organs." Caizheng, no. 2 (1985). . 107- Zhimin, LU. 'Prvnial Responss to Fiscal Decentralization: lhe Case of Ad, Guangdong, and Shanghd.' Paper prnted at the Confeenc on Reform ad Policy Iplemenion China: Transformin Conces Ino Woking Realities. lh Cine Univesity of Hog Kn, Septmber 9-11, 1991. Zhou Harong. 'Fiacing of China's Bc Industies: Th Penaicon of Chanmels. Pe pesened at Intenaiona Workshop on Financing Investm In Inratructr and Basic Industres HanghOU, China, Mar 24-28, 1992. Zhu, Fulin. 'Control and Managemnt of Government Expendltue: 'Me Case of CbInaw Economic and Social Commission for Asia and the Paciflc, Unied Nations, New DeI Augu 1991. Zhu Yaolng. 'Funds Raised for Highway Constucdon: Curet Sita, Issues and Measures.' Paper presented at the hoItmaional Workshop on Financig tmet in Inftruce and Basic Indutries, Hangzhou, China, March 24-28, 1992. Zhu Jialang. 'Diver Types and Fiacing of Basic Industies.' Paper prese at the Itrnaiona Workshop on Finncing bI _nent in InfrIstructur and Basic Industi, Hangzhou, Chi, Mach 24.28, 1992. Zou Jiahuam 'Report on the Implemenaion of the 1991 Plan for Ntonal Eoomic and Soil Developmet and the Drat 1992 Plan.' Delivered at the Sevth Naonal Peoples Party Congress, 5th Sess., Beijing, March 21, 1992.