The New Fiscal Federalism in Brazil

Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Policy,Resarch,and ExternalAffair| WORKING PAPERS Pubic Economlvic Department CountryEconomics TheWorldBank 1990 December WPS557 The New Fiscal Federalism in Brazil Anwar Shah Fiscal arrangementsin Brazil severely constrain the federal government's ability to fulfill its mandate as a national government. Municipalgovernments,meanwhile,have morerevenues than they need,encouragingfiscalmismanagement.Reform is urgentlyneededto counteractBrazil's fiscalimbalance. wodinpropessand EePo<iligy.R candExtemflAffaiComplexdnbutPREWoddngPaptoaninatethefindingsof to enamungctheexchangeof ideasamongBankstaffand all othas inteeed in developnentissues.lese papes cany thenamesof and conclusio armthe the authors,rflect onlytheir viws. andshouldbe used andcited accordinglyThe findings,interpretations, authonsown.Theyshouldnot be atrbted to theWorldBank,its Boardof Dicos, its managaetnt,or any of its membercountrie.s Policy,Research,and Externa Afbira PubilcEconomlks| WPS557 This paper - a product of the Public Economics Division, Country Economics Departnent - is part of a larger effort in PRE to reform public sector management in developing countries. It is one of a series of discussion papers prepared for the Intergovemmental Fiscal Relations Project of the Public Economics Division. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Ann Bhalla, room N10-059, extension 37699 (117 pages). Brazil is a three-tiered federation of 24 states, two federal territories, a federal district (the capital), and 4,300 municipalities. In 1989 less than half of all govemment spending was controled by the federal govemment. Brazil's new constitution gave autonomous broad powers to states and municipalities on certain tax and spending functions, with municipalities independent of and coequal to states. Shah reviewed and analyzed the intergovenmmentalfiscal relations in Brazil. He found that: * Federal and state governments are involved in purely local functions in an uncoordinated fashion. * Conditional transfers are arbitrary and driven primarily by political considerations. Programs woric at cross-purposes and the subjective nature of these transfers may be sending the wrong signals to lower levels of govemment about laxity in fiscal management. * Revenue-sharing constrains the federal govemment's ability to fulfill its mandate as a national govemment and is conducive to fiscal mismanagement as local govemments are shying away from raising revenues from property taxes and user charges. The municipal govenmments have more money than they need. The state governments also face a financial squeeze but it should be short-lived as they have access to the value-added tax, a dynamic source of revenues The federal govemment's problem is structural. * The administration of sales tax by all three levels creates duplication and confusion. Its revenues fall far short of its spending needs. * In short, existing fiscal arrangements have * Administration of the general value-added tax by the state involves unresolved issues about tax crediting on interstate trade. created a vertical fiscal imbalance. Shah presents policy options to resolve these problems. * The state and municipal revenue-sharing funds do not distribute revenues fairly and equitably. The PRE Working PaperSeries disseminatesthe findingsof work underway in the Bank's Policy, Research,and External AffairsComplex.An objective of theseriesis to get thesefmdingsoutquickly,even if presentationsareless thanfullypolished. The findings, interpretations, and conclusions in these papersdo not necessarilyrepresentofficial Bank policy. Producedby the PRE DisseminationCenter TH NEWFISCALFEDERALISMIN BRAZIL by Anwar Shah * Table of Contents 1.0 PERSPECTIVES ON THE NEW FEDERALISM IN BRAZIL ....... 1.1 Expenditure Assigmr.ent:Theory and Practice in Brazil ..... 1.11 1.12 1.13 1.2 1.3 1.4 D ........... Expenditure Assigmnent Under the 1988Constitution ..... ....... Conceptual Basis of Expenditure Assignment ..... ............. Expenditure Assigmmentin Brazil and Implications for Efficient and Equitable Provisionof Public Services ..... ........ 1 1 3 5 Tax Assignment:Theory and Practice in Brazil. 8 1.21 1.22 1.23 8 9 The Theory of Tax Assignment. Tax Assignment in Brazil. Implications for Tax Policyand Administration.12 Vertical and Horizontal Fiscal Imbalances ........................... 13 1.31 1.32 13 19 Vertical Fiscal Imbalance in Brazil .......................... Horizontal Fiscal Imbalance in Brazil ......... .............. Brazil in Relation to Other Federations: An ImpressionisticView .... 1.41 1.42 1.5 1....... The Practice of Federalism in Selected Countries ..... An ImpressionisticEvaluation .............................. ..... .......... 19 22 24 Conclusions .27 This is one of series of discussionpapers prepared for the IntergovermmentalFiscal Relations project of the Public Economics Division. The project is directed by Anwar Shah. The author is grateful to Antulio Bomfimfor research assistanceand to Dr. Ricardo Luiz Santiago,former Deputy Minister of Planning,Brazil and currently Operations PolicyAdvisor,Inter-American Development Bank, Dr. Consentino Tavares, Director General, Ministry of Finance, Brazil, Jose Pio Martins, Director General, Ministryof Finance, the State of Parana, Diago Loudelle de Mello, IBAM, Rio, Brazil, and Drs. Johannes Linn, Bela Balassa, Peter Knight, Javad Shirazi, Martha de Melo, Gobind Nankani, Demetrios Papageorgiou, Robert Schneider, Ms. Helena Cordeiro and Antonio Estache, World Bank for helpful discussions. ii 2.0 INTERGOVERNMENTAL 2.1 2.2 2.4 2.5 3.0 .. .......... 30 30 2.11 2.12 32 36 State Participation Fund (FPE: Fundo de Participacao dos Estados) . Municipal Participation Fund (FPM) ......................... Specific Purpose Transfers ................... ................... 36 Negotiated Transfers (Conveinios) ........................... 41 Intergovernmental Transfers: A Conceptual Perspective ................ 51 2.31 Grant Types ............. .............................. 2.311 Non-matching Transfers ............................. 2.312 Selective Matching Transfers ......................... 51 51 54 2.32 The Economic Rationale for Intergovernmental Transfers .... ..... 58 Intergovernmental Transfers in Brazil: An Economic Evaluation 64 2.41 Tax Sharing Program: A Closer Examination .................. 2.411 States Participation Fund: A Critical Look ............... 2.412 Municipal Participation Fund: A Review ................. 65 65 66 2.42 Negotiated Transfers: An Examination 68 Reform Options 2.51 2.52 2.53 2.6 ........ Revenue Sharing Arrangements in Brazil ........................... 2.21 2.3 TRANSFERS IN BRAZIL ....................... ....................... ....................... 70 Revenue Sharing ........................................ Negotiated Transfers ..................................... Equalization Program .................................... 71 72 72 State-Municipal Transfers in Brazil ................................ 75 2.61 2.62 2.63 75 80 80 State-Municipal Revenue Sharing ........................... Conditional Transfers .................................... An Evaluation and Policy Recommendations ................... SUMMARY AND POLICY RECOMMENDATIONS .......... ........... 84 3.1 The Assignment Issues ......................................... 84 3.2 Issues in Intergovernmental Transfers .............................. 85 3.3 Implications of Existing Arrangements 88 3.4 Proposals for Reform .......................................... ............................. 89 ... APPENDIX A THE DESIGN OF A FISCAL EQUALIZATION PROGRAM: ISSUES AND OPTIONS APPENDIX B OPERATIONAL MECHANISMS OF CONVENIOS .......................... APPENDIX C THE ADMINISTRATIVE PROCESS OF SUDS ...... REFERENCES .. 92 101 103 ...................... 112 ..................................................... 1.0 PERSPECFIVESON THE NEWFEDERALISMIN BRAZIL Brazil represents a three-tiered federation of 24 states, two federal territories of Roraima and Amapa( to become states on January 1, 1991),a Federal Districtcomprisingthe national capital of Brasilia, and 4300 municipalities. Brazil represents one of the most decentralized federations in comparison to other developing nations. In 1989, less than 50% of consolidated federal-state-local government expenditures were controlled by the federal government. The new BrazilianConstitution promulgatedon October 5, 1988has given autonomous broad powers to states and .ncinalities in certain tax and expenditure functions. Municipalgovernments in other federations are usuallyhandmaiden of state governmentswhereas the Brazilianconstitution recognizestheir independent and coequal status. A review of tax and expenditure assignment is thus critical to an examination of intergovernmentalfiscalrelations in Brazil. In the following,we reviewexpenditure and taxassignment in Brazil;reflect on the conceptualbasis for such assignmentand present a comparative perspectiveon this issuedrawingon the experiences of selected federations. 1.1 1.11 Expenditure Assignment:Theory and Practice in Brzil Expenditure Assignment Under the 1988Constitution. Table 1 provides basic details on expenditure assignment in BraziL The constitution recognizesthree separate levelsof government in the federation and delineates their responsibilities. Defense, foreignrelations, currency, postal'services,planning,nuclear energy,national highwaysand regulation of labour conditions, foreign and ;nterstate commerce, finance, banking and insurance, commercialadvertising,inter-state transport, telecommunicationsand data processing,energy, mining, employmentinsurance,socialsecurity,immigrationand naturalizationand native rightsare the exclusive domain of the federal government. Health, education, culture, protection of environment and natural resources, hydroelectricity,agriculture,food distribution, housing,sanitation, socialwelfare and police are the joint responsibilityof federal and state governments with the federal govermmentsetting standards and state governments havingthe responsibilityfor deliveryof these services. Municipal 1 2 Table 1 EXPENDITUREASSIGNMENTIF BRAZIL Responsible Level of Government ExpenditureCategory Federal only Defence Foreign Affairs InternationalTrade Currency, banking Use of water resources National Highways Planning; regional and natural Postal service Police: federal and frontier a-eas Regulationof labor, inter-statecommerce, telecommunications,inter-statetransport, urban development,energy, mining, employmentinsurance,immigration, citizenshipand native rights Social Security National Statisticalsystem Guidelinesand basis for national education Federal-State(Shared) Health Education Culture Protectionof the environment and the natural resources Agriculture Food distribution Housing Sanitation Social welfare Police Hydroelectricity State only Residual powers i.e. any subject not assigned to federal or municipal levels by the Constitution. Municipal only Public transport (intracity) Pre-schooland elementaryeducation Preventivehealth care Land use Historicaland culturalpreservation Source: NOVA CONSTITUICAOBRASILEIRA,1988. 3 governmentshavebeen giventhe legislativeauthority on subjectsof local interest to supplement federal or state legislation. Furthermore, they have been assigned responsibilitiesin public transport, preschool and elementary education, health care, land use and historicaland cultural preservation. Any subjects not specifiedby the Constitution have been reserved for state government legislation. 1.12 Conceptual Basis of Expenditure Assignment Allocationof expenditureand taxfunctionsto variousmember units isthe most fundamental issuein a federation. Fiscalfederalismliterature argues that expenditure assignmentmust precede tax assignment.This is because tax assignmentwould in general be guidedby expenditure requirements at different levels and these cannot be worked out in advance of expenditure assignment. This literature also provides broad guidance in delineating expenditure responsibilities. For example the so called "decentialization theorem" advanced by Wallace Oates states that "each public service should be providedby the jurisdiction havingcontrol over the minimumgeographic area that would internalize benefits and costs of suc.hprovision"(Oates (1972),p.55 ).1 The abovetheory providesstrong rationale for decentralizeddecit . naking on efficiency, accountability,manageability,and autonomy principles. It is argued that: 1. Local provision allows governments to cater better to the tastes and needs of local residents whereas central provisionoften results in more uniform provision; 2. Decision making is closer to the people for whom the services are intended. This inducesmore responsivenessto local concernsas wellas more fiscalresponsibilityand efficiencyof provisionespeciallywhere financingof servicesare decentralized as well; 3. Eliminates multiple layersof jurisdiction;and 4. Enhances inter-jurisdictionalcompetition and innovations in the provisionof public services. A decentralized system thus ensures consistencyof level and mix of public serviceswith voters' preferences as wellas providesincentivesfor efficientprovisionof such services. The following 4 economicconsiderationsneverthelesswarrantsome degree of centralization or compensatorygrants in the provisionof servicesfor which such factors are relevant. a. Geographicalvariationsof preferences withimperfect mobility:Note that the Tiebout mechanismof votingwith one's feet requires smaliopen area and imperfect mobility. In the presence of imperfect mobility,a decentralized mechanismwould not ensure matchingof publicserviceswith local preferences. b. Spatial Externalities: Spatial externalities arise when benefits and/or costs of public servicesare realized by non-residents. In the case of benefit-spilloutthe jurisdiction providingthe service does not considerthe proportion of benefits of a public service accruingto non-res.dents and thereforewould under-providesuchservice.The reverse result is obtained in the case of cost spillouts i.e. where the public service could be financed byexportingtaxesto other jurisdictions.There are also publicserviceswhose benefits are considered national in scope such as defense and foreign affairs. As a corollary,these serviceswouldbe best providedbythe government at the highestlevel i.e. by the federal government. c. Economiesofscale:Certain servicesrequire serviceareas largerthanalocaljurisdiction for cost effective provision. Examples include transportation services,water and sewerage in a metropolitan area. d. Administrative and compliance costs: It is generally agreed that a centralized administration leads to lower administrativeand compliance costs associated with financingof public services. There are alsocertain policyfunctionswhich canbe carried best by the federal government. For example: (i) StabilizationPolicy.Astabilization policycannotbecarriedouteffectivelybya local jurisdiction.Local pursuit of such a policywouldlead to muchof the gains beinglost to outsidejurisdictions. A monetary policyhas little srope of beingcarried at a local leveL 5 (ii) Redistribution Policies:Effective redistributionis possible only through w.rograms which are national in scope. A local jurisdiction that attempts to carry out redistributivepolicies is likely to driveout the rich. The City of New York faced a fiscalcrisis as a result of redistributiveprograms caoied out in early eighties. The above discussionsuggestsfederal government be givenexclusiveauthority in carrying outstabilization policiesandprovidingpublicserviceswhosebenefitsarenational inscope (seeBoadway 1980,1989).The federal governmentalso has a redistributiverolewhich couldbe exercisedthrough the tax and transfer systemor through joint provisionof publicservicessuch as education and health which primarilyserve as "transfersin kind". The federal government also maybe assigned a role in providing compensatorygrants for spilloutof benefits fromstate level provisionof services. A similarrole for each state is in order for spillout of benefits from local provision of serviceswithin their jurisdictions. All other services would be best provided by the local governments with federal and state governments havingsome role in defining minimumstandards. Table 2 provides a summary view of broad guidance provided by economic theory and discussedabove. As shown by this Table, reasonably clear expenditure assignmentemerges from the basic economic principles. The followingparagraphs provide a commentary on the expenditure assignmentin Brazil based on these guidelines. 1.13 Ex: enditureAssignment in BrazilandImF .rationsForEfficientandEquitableProvision of Publc Services A comparison of Table 2 with Table I indicates that expenditure assignment done by the Constitution correspondsfairlycloselyto the expenditureassignmentsuggestedby economicprinciples. Current practicesin servicedeliverybyvarious levelsof governmentare at substantial variancefrom the Constitutional assignment. In the areas of education, health, urban transportation, recreation and culture, child and old age care and social assistance,all three levelsare involvedin an uncoordinated 2 Current difficultiesin overlapping and fashion leadingto muchconfusion and chaos in servicedelivery. duplicationof functionsin certain areas thus must be attributable to pre-1988Constitution traditions. 6 Table 2 CONCEPTUALBASIS OF EXPENDITUREASSIGNMENT ExpenditureCategory Service Provision Responsibility of Service Defence Foreign Affairs InternationalTrade Environment Currency,banking Interstatecommerce Immigration UnemploymentInsurance Airlines/Railways Industry and Agriculture Education Health Social Welfare Police Highways Natural Resources Notes: F: Federal S: State L: Municipal-Local F F F F F F F F F F,S,L F,S,L F,S,L F,S,L S,L F,S,L F,S,L F F F S,L F F F F F S,L S,L S,L S,L S,L S,L S,L Comments Benefits/costsnational in scope n 7 For examplein the area of education the current Constitution sees Federal governmentsimplysetting the normsand leavingthe provisionto state and municipallevels. Contraryto Constitutionalstipulation discussed above Federal Government nevertheless continues to be involved in direct delivery of education services at secondary school, college and university levels and in an isolated instance in elementaryeducation as well. Furthermore the Constitutiongives localgovernment exclusivemandate in the deliveryof elementary education but the deliveryof secondaryand higher levelsof education has beer,left open. Local provisionof secondaryeducationwouldprovideincentivesfor allocativeefficiency as wellas encourage innovativelearning programs. It is best for the federal government to have a hands off approach in these areas and the State governmentsto assume regulatoryrole by monitoringstudent achievement through standardized tests and by setting minimumservice standards. Undergraduate education because of significant spillo-.ersbeyond local jurisd: tions but presumably insignificant spilloversbeyondstate boundariesis a logicalcandidatefor state level assignment.Universityeducation on the other hand is expected to havesignificantspilloutsbeyond state boundariesand therefore should be a joint responsibilityof Federal and State levels. As envisagedby the Constitution, however, the prc' dsionof thisserviceshouldbe state responsibilitywiththe federalgoverrment assuminga monitoring role and providingcompensatorygrants for benefit spilloutsor to ensure minimumstandards of such service across states. Similarproblemsexistincurrent deliverypracticesforurban transportation. Urban corridors are constructed and maintained by all three levels and sometimes suffer .eglect because of unclear responsibilitiesfor maintenance and upkeep. Intra-urban bus servicesin some states are operated by the State. A decentralized systemof urban transport provisionwouldbe more attuned to mcetinglocal needs and to adjustmentsbased on changinglocal conditions. Thus both federal and state governments have no clear role for intra-urban transportation provisionissues. Their role should simplybe confined to providingtechnical assistance,monitoring and setting service standards. In today's Brazil, federal government is often invo!vedin direct provision of urban transportation network and servicesand this sometimescreates significantdifficultiesforlocal governmentsin their decisionsto upgradelocalservices or modifyexistingnetworks. Some purely local functions such as fairs, public markets, slaughterhouses are also being performed by federal and state governmentsin several municipalities. Electricityand telephones are provided by federal public utilties. Public provision of these essentiallyprivate goods needs to be reviewed. Responsibilityfor water and sewer which ate generally retognized as local services was transferred from the municipal to the state level in 1967in view of the assistance available to state governmentsthrough theNational SanitationPlan (PLANASA).Thischangeoverresulted insubstantial expansion of service to new towns but emphasis on cost r%coveryhas precluded extension of these servicesto slum areas of larger cities and metropolitan areas. In conclusion,constitutional assignmentof expenditure function is broadlyconsistent with economic principlesbut a clear case can be made for bringingthe practice or de facto assignment in conformitywith the decentralization theme enunciated by the new Brazilianconstitution. 1.2 1.21 TAXASSIGNMIENT:THEORYAND PRACTICEIN BRAZIL The Theory of Tax Asslgnent The divisionof revenue sources among federal and lower levels of government is being referred here as the 'tax assignmentproblem'. Once expenditureassignmenthas been agreed upon, tax assignment and design of transfers become critical elements in providing reasonable matching of expenditureneeds with revenue meansforvarious levelsofgovernment. Although taxassignmentcould be undertaken independently of expenditr assignment-a practice which is quite common in developingcountries, yet the tradeoffsbetween the advantagesof a centralized tax administration and decentralized provision of public services become more apparent when tax assignment takes into considerationpre-determined expenditure assignment. In such a situation over-dependence of lower levelsofgovernmentonintergovernmentaltransferswithpotentiallydistortionaryeffectsonexpenditure priorities could be avoided. Furthermme, in those grey areas where theoretical guidance on tax assignment is unclear, expenditure a nt can provide a powerful argument for assigning Ltxing responsibilitytothegovermentwithgreater oedforadditionalrevenues. Musgrave( 1984)usesequity 9 (consistencyof revenue meanswith expenditureneeds) and efficiency( minimizingresourcecost) criteriaandsuggeststhe followingbroad principlesin taxassignment; L Progressiveredistnbutivetaxesshouldbe central; ii Taxessuitableforeconomicstabilization shouldbecentral;lowerleveltaxesshouldbe cyclically stable; iii. Taxbasesdistributedhighlyunequallybetweenjurisdictionsshouldbe centralized; iv. Taxeson mobilefactorsof productionare best administeredat the centre; v. Residencebasedtaxessuchassalesof consumptiongoodsto consumersor excisesare suitedfor states; vi Taxeson completelyimmobilefactorsare bestsuitedfor locallevel; viL Benefittaxesand userchargesmightbe appropriatelyusedat all levels. Basedon theseprinciples,reasonablyclearguidelinesfor assignmentof revenuesourcesto variouslevelsof governmentemerge.Table3 providesa summaryviewofsuchassignment.The Table suggeststhat for certain taxessuch as resourcetaxes or a value-addedtax (VAT) base and rate determinationandcollectionand administrationcouldbe assignedto differentlevelsof government. By followingthis approachboth inter-jurisdictional equity and efficiencyof tax administrationand compliancecouldbe achieved. It shouldbe noted that the theorycontravenesthe advicesometimes offeredbyinternationalagenciesto developingcountriesthat localtaxeson wageand capitalincome shouldbe instituted.Withfactormobility,basesfor suchtaxeswouldbe subjectto erosion. Also,such a regimeencouragestaxcompetitionamongvariousjurisdictions. 1.22 TaxAssignmentInBrazl Table 4 providesan overviewof existingtax assignmentin Brazil. Federalgovernment assumesexclusiveresponsibility for the taxeson income,payroll,wealth(largefortunes),foreigntrade, bankin6financeandinsurance,ruralproperties,hydroelctricityandmineralproducts.It haspartially overlappingresponsibility withstate andlocalgovernmentsfor taxationof industrialproducts.The 10 Table 3 CONCEPTUALBASIS OF TAX ASSIGNMENT Tax Determinationof Base Rate Tax Collection & Administration Customs Income Tax Estates & Gifts Corporate Tax Resource Tax Retail Sales F F F F F S F F F,S F,S F,S F,S S S F F F F S S F VAT Excises Property tax User charges F S S F,S,L F,S S L F,S,L F,SC S L F,S,L Notes: F: Federal S: State/Province L: Municipal/Local SC: The Council of States Comments InternationalTrade Redistributive Redistributive Mobile Factor Unequally distributed Higher compliancecost Harmonized, lower compliancecost. 11 Table 4 TAX ASSIGNWENT IN BRAZIL - 1990 Responsibility Base Rate Adnn. Revenue Source 1993 Disposition of Revenues (S) Federal States Municipalities Federal (F1: Income Tax (IR): Personal Corporate Payroll Tax (CSE) Large Fortunes (Wealth) Tax (IGF) Import Tax (IN) Export Tax (IE) Tax on Financial Operations and Insurance (IDF/IOC) Rural Property Tax (ITR) Tax on Industrial Products (IPI) Hydroelectricity Tax Mineral Products Tax 53.0 24.5(a) 22.5 F F F,S F F F F F F F F F F F F F F F 33.3 100.0 100.0 100.0 66.7 0.0 0.0 0.0 F F F F F F F F F F F F F F F 100.0 50.0 43.0 5.0 5.0 0.0 0.0 32.0(b) 45.0 45.0 S S S S S S S S S S S S 0.0 0.0 0.0 0.0 75.0 100.0 50.0 100.0 25.0 0.0 50.0 0.0 M M M M M m 0.0 0.0 0.0 0.0 100.0 100.0 n M M M 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 0.0 50.0 25.0 50.0 50.0 States (S) General value-added tax (ICMS) Inheritance and gift taxes (CMD) Motor Vehicles Registration Tax (IPVA) Supplementary Capital Gains Tax Municipal±ties (M): Services Tax (ISS) Urban Property Tax (IPTU) Tax on Retail sales of fuels except diesel (IVVCLG) Property transfers (ITBI) Frontage tax (Special assessment levy) Source: 1. 2. 3. 4. Notes: (a) (b) (c) M M r M M NOVA CONSTITUICAO BRAZILEIRA 1988, Sistema Journal Tributacao e do Orcamento", pp.67-78. Lei do Senado No. 165 (11/89) Lei Complemantar No. 104, Camara dos Deputados. Projeto de Lei Complementar No. 104-A, 1989. do Brasil, Article VI, RDa Includes 32 to finance programs to be administered by Development Banks of the NorthEast, North and Center-West regions. Includes (a) plus an additional 102 of IPI as compensation to exporting states for loss of revenues from ICHS on account of exports. The Union must apply at least 182, states 252, and municipalities 252 of all tax revenues and transfers on education. 12 federal government allowsstates to levysupplementaryrates upto 5%on the federal bases for personal and corporate incomes. The mainstay of state governmentsrevenues is the general value-addedtax on goods and services.This taxis administeredby the Councilof states havingfinance ministersof all states including the Federal District as its members. Anychangesin the taxbase or rates must be presented byindividual states for approval by the CounciL The Councilhas resistedchanges in tax rates quite stronglybut has acceded to requests from various states for exemptingcertain commodities/servicesfrom the tax base. Inter-state tax credit issuesfor the value-addedtax continue to take a great deal of the Council'stime. The states have also accessto taxationof inheritance and gifts and motor vehiclesregistrations. States derive 72% of their revenues from these three taxes. Municipalitiesare empoweredto levytaxeson services,urban properties, retailsales of fuels except diesel; property transfers (intervivos)and special assessments(frontage). Municipalitiesraise 18% of revenues from these sources. 1.23 Implications For Tax Policyand Administration A comparison of Table 3 with Table 4 shows that the tax assignment done by the new Constitution in Brazilis broadlyconsistentwith economicprinciplesenunciated above. Some problems remain, nevertheless. These problems are mainlyin the area of sales taxes. Tax bases for the federal manufacturerlevelsalestaxonindustrialproducts and thestategeneralvalue-added tax (ICMS)partially overlapbut are administeredseparatelyby the two levels. Similarlytaxbases for ICMS and local taxon services(ISS) overlap but are administered in an uncoordinated fashion. Brazil is unique in being the only country with a subnational VAT called ICMS. While tax collection of ICMS is in the hands of individualstates, taxrates and base isdetermined bythe Councilof States. This shouldin principleresult in tax harmonization as well as a clearing house for interstate VAT tax credit claims. In practice, the Councilhas been quite receptive to givingindividualstates some flexibilityin definingown tax bases for ICMS. This has the potential of eliminatinguniformityof the ICMS base over time. Also resolution of interstate ICMS tax credit issuescontinue to elude the CounciL 13 It would be desirable to consolidateIPI, ICMS and ISS into a single taxadministeredby the federal government and proceeds shared with state and local governmentsbased on a formula which allocates revenues to the three levels in roughly the same proportion as their current intakes and to individualunits based on originprinciple. Tax base determination for harmonization purposes should be the responsibilityof the federal governmentwhereasstates and municipalities,iftheyso choose,could levysupplemental rates. There is no specificadvantagein federal government administeringthe rural property tax (ITR). 1TRis more appropriate revenue source for state and locallevels. Anyrural property under the jurisdictionof a municipalityshould be subject to taxation by the municipalityconcemed. State level governmentson the other hand should be responsiblefor both the administration and final disposition of revenues on account of this tax in unincorporated areas within state boundaries. 1.3 Vertical and Horizontal Fiscal Imbalances Vertical fiscalimbalance refers to the mismatchbetween revenue means and expenditure needs at various levels and the horizontal imbalance refers to inconsistencybetween revenue raising abilityand fiscalneeds of governmentsat the same level in a federation. These issues are discussedin the followingsections. 1.31 Vertical Fiscal Imbalance in Brazil Some degree of mismatchbetween revenue means and expenditure needs at various levels is common to all federations. Efficiencyin tax administration for certain revenues requires central administration and this in itselfcontnbutes to the vertical imbalance problem. Thus after expenditure and tax assignmenthave been completed, revenue sharing and transfers are frequently used to correct for any imbalances that result from assignment of responsibilities. However, revenue sharing and transfer mechanismsdue to difficultiesin design or due to conflictingclaimsof relativeneeds by various levels of government may not fully resolve this issue. In Brazil, constitutional transfers attempt to 14 address this issue. Tables 5 and 6 quantify revenues at the disposalof various governments before and after revenue sharingimpacts. With the newtaxassignmentand transfers,federal governmentshistorical positionvis-a-visstate and localgovernments has significantlydeteriorated. States now command one of the most dynamicrevenue base (ICMS) and municipalitiesare guaranteed a largeshare of federal and state revenue collections.Whilea precisecalculationof the magnitudeof the squeeze on the bigbrother put by the new fiscal arrangements must await more careful analysis,Table 7 presents some rough estimatesto outline the broad picture of verticalimbalancethat characterizesBrazilof today. According to these calculations, federal and state governments' revenue means significantlyfall short of their expenditure needs. The opposite situation holds for municipalgovernments. The table shows that federal government would be (if it is already not) in dire straits if it continued to follow in future as in past years a broader interpretation of its responsibilities.State level governmentsas a whole face some difficultiesnowbut these maynot persist in the long run in viewof expected growth of ICMS revenues. Municipalgovernmentsin Brazil,on the other hand, shouldbe the envyof allgovernmentsin developing (or even advanced nations) world. So far we have concerned ourselveswith reaching broad judgements on vertical balance in Brazil alone. It would be interesting to reflect howBrazil compares to advanced country federations. Unfortunately, there are no satisfactorymeasures at our disposal to reflect on this question. Three measuresproposed by Hunter (1977)and previouslyusedby Bird (1986)and others attempt to measure the degree of control exercised by the federal government over lower levels of governments. These measures are termed as coefficients of vertical imbalance. The way these measures are structured suggests that a coefficient of zero would indicate absolute federal control over state and local governments and a coefficient of one would indicate that lower levels of governments are absolutely autonomous in their decision making. Note that while a high value on this coefficient is desirable, a value of one has never been a goalin any federation. A value closerto but certainly lessthan one would also be consistentwith the assignmentprinciplesenunciated earlier. Table 8 presents three calculations: one consideringconditionaltransfers and borrowingonly,a second one by incorporatingunconditional transfers and a third one by bringingin shared taxes as well. On the first two coefficients,Brazil does better than selected advanced federations reported there. On the third coefficient, considering 15 Table 5 BRAZIL: TAX REVENUE COLLECTIONSBY LEVEL OF GOVERNMENT Year Union 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987* 1988* 48.1 53.3 19.6 49.5 49.5 49.2 51.0 48.8 50.7 51.3 15.8 51.5 53.7 54.4 56.4 58.3 58.5 59.8 59.0 62.3 60.9 58.2 58.3 58.7 58.2 57.2 57.8 56.9 57.6 53.5 54.2 47.1 Shares (X1 _ States Municipalities 43.3 41.2 49.6 44.5 44.6 44.5 42.5 44.8 42.5 41.4 49.4 44.7 42.7 41.9 40.0 37.8 37.7 36.9 37.0 33.1 34.0 36.1 35.9 36.2 36.7 37.6 37.0 38.6 38.3 42.2 41.6 49.4 * Preliminarydata Source: MINIFAZ/SEF unpublisheddata. 8.0 7.5 6.4 6.0 5.9 6.3 6.5 6.5 6.8 7.3 4.8 3.8 3.7 3.7 3.6 3.8 3.8 3.8 4.1 4.6 5.2 5.7 5.9 5.1 5.1 5.2 5.2 4.5 4.1 4.3 4.2 3.6 All 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 16 Table 6 BRAZIL: FINAL DISPOSITIONOF REVENUESBY LEVEL OF GOVERNMENT Year Union Revenue Shares States Municipalities All 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988* 42.7 45.7 43.3 43.2 41.8 39.6 41.9 39.6 39.0 40.6 36.9 40.6 45.8 45.7 47.7 49.7 49.1 50.2 50.3 51.4 50.2 47.3 47.5 49.3 49.2 48.0 48.4 46.8 44.7 39.5 42.5 33.4 46.0 44.7 47.9 48.2 49.7 48.9 46.0 48.5 48.1 46.3 45.2 42.6 39.8 39.6 38.4 36.5 37.1 36.2 36.0 34.4 34.8 36.7 36.3 35.5 34.8 35.7 35.2 36.5 37.5 40.7 38.6 50.7 11.3 9.5 8.9 8.5 8.6 11.5 12.0 11.9 12.9 13.1 17.9 16.9 14.4 14.7 13.9 13.8 13.7 13.6 13.7 14.3 15.0 16.0 16.2 15.2 15.9 16.3 16.4 16.7 17.8 19.9 18.8 15.9 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1993** 36.5 40.7 22.8 100.0 * Preliminarydata ** Estimate Source: MINIPAZ/SEP Unpublisheddata 17 Table 7 VERTICAL IMBALANCEIN BRAZIL Revenue Share a Expenditure Share b Surplus/ Deficiency c Federal 36.5 43.4 - 6.9 States 40.7 43.0 - 2.3 Municipalities 22.8 13.6 + 9.2 All levels 100.0 100.0 0.0 Notes: a. Final disposition of all revenues based on a fully phased in system of constitutionaltransfers in 1993. b. Actual 1988 expenditures c. Ignores borrowing. 18 Table 8 COEFFICIENTSOF VERTICAL BALANCE FOR SELECTED FEDERATIONS Country Australia Canada West Germany United States Brazil (a) (b) Y1 Y2 0.68 0.79 0.84 0.81 0.93 0.91 0.35 0.71 0.82 0.81 0.89 0.87 Y3 0.55 0.80 0.85 0.85 0.83 0.76 Notes: V1 - 1 - (Sc+B)/E V2 - I - (Su+Sc+B)/E V3 - I - (Su+SC+B+Ts)/E Where Sc au B E Federal conditionaltransfers to states - Federal unconditionaltransfers to states - Net borrowingby states - States expenditures Ts - Shared Taxas - (a) for states only (b) for the consolidatedState-Localpublic sector. Source: (1) Brazil 1988, this report. (2) For other countries, six year averages based on data for 1970s. See J.S.H. Hunter, (1977) Federalism and Fiscal Balance, Canberra: Australian National University, Center for Research on Federal FinancialRelations. 19 federal-state sector only, Brazil is ahead of Australia and Canada but closelybehind USA and West Germany. Note that calculations for selected advanced federaitionsare quite dated and somewhat different valuesare expectedto be obtained based on more recent data especiallyfor Canada in viewof its significantdecentralization experience of the 1980s. The new data is however unlikely to change Brazil's relative position in this matrix. 1.32 HorizontalFiscal Imbalancein Brazil Brazil is a large country with very diverseeconomicopportunities across its vast landscape. Per capita incomein Rondonia, one of the poorest states is only 12% of per capita incomein Sao Paulo, the richest state in the Union (see Table 9). Further, per capita income in the north and northeast regions of the country is less than one-half of the national average. Regional equity issues have, therefore, dominatedthe agenda of all governmentsin its recent history. The federal government uses its tax, transfer and expenditure policiesin an attempt to reduce regional disparities in public services provision. The overall impactof the federal government as shown in Table 10 is stronglyredistributive. Its net impact is negative in the South and Southeast (the have-regions) and positive in the north, northeast and centre-west regions (the have-not regions). Interstate equalization due to lack of any explicitstandard of equalization in current federal puliciesremains an elusivegoal. The net impactof the federal government on the state of Para with 1988per capita income of NCz$266was NCz$0.04as opposed to NCz$52.31in Acre, a state with about the same per capita income. Total per capita expenditures of states also show a great deal of disparityas shown in Table 9. The issuesconcerning equalizationof fiscalcapacitywillbe taken up later under transfers. 1.4 Brazil In Relationto OtherFederations:An ImpressionisticView Earlier sections took issue with the current expenditure and tax assignmentin Brazil and reflected upon their implicationsfor vertical and borizontalfiscalimbalancesin the union. Problems of the sort discussedearlier, however, are not germane to Brazil alone and are a creation of political Table 9 PROFILEOF BRAZILIANSTATES AN ECONOMIC Federal Unit Rondonia Acre Amazon.s Roraima Pars mapa NORTH OWN OWNREVENUES/EXPENDITURES GOP POPULATION POPDENSITYTOTALFEDERALTRANSFERS (per capita)EXPEND 1988 MANUFAC V.A. 1988 current NCz3 REVENUES 1988 (CURRENT NCzS) 1984 1988 1988 1988 per capita per capita thousands (people/sq.k Tax Other Total per capita 1988 (NCzl) (per capita) (NCz3) 0.17 128.74 1019.2 4.28 18 79 96 2.58 49 45 94 122.79 28 21.15% 0.06 268.61 396.5 21 95.24 76 79.09% 1.21 12 10 0.87 502.98 1895.6 0.04 286.01 113.1 0.60 123 123 368.46 307 83.33x 14 31.23 23 74.223.79 9 5 0.19 286.72 4724.4 78 227.34 1ss 88.82x 240.2 1.69 78 0.11 264.14 16 17 33 58.39 28 50.01X 0.34 303.23 8388.9 2.35 Alagoas Sergipe Bahia NORTHEAST 0.05 0.07 0.19 0.17 0.17 0.38 0.29 0.25 0.43 0.27 154.73 136.42 234.74 280.84 203.22 345.S4 271.54 279.12 385.98 276.67 4970.4 26e7.4 8239.4 2238.2 3152.6 7114.6 2342.3 1389 11304.4 41296.3 15.08 10.15 42.33 42.06 58.43 70.43 80.47 62.62 19.94 28.64 9 10 7 12 9 6 10 18 6 8 8 18 12 4 7 30 10 Minas Cerais EspiritoSan Rio de Janel Sao Paulo SOUTHEAST 0.88 0.45 0.78 2.20 1.45 699.29 581.99 989.54 1069.45 913.82 15345.8 2429.4 13558.1 31851.5 82982.8 28.16 53.12 310.54 127.50 88.14 4 5 2 2 3 Parana 0.82 1.14 1.15 0.94 666.12 718.53 890.38 767.30 8732.3 4311.3 8888.5 21932.1 43.81 45.23 31.67 38.12 DistritoFed CENTER WEST 0.16 0.18 0.19 0.14 0.17 588.91 289.73 359.57 838.52 472.81 1714.5 1829.3 4740.4 1743.3 9827.5 BRAZIL 0.88 843.87 Maranhao Piaui Ceara" R GrandeNor Paraiba Pernambuco Santa Catarl R Grande Sul SOUTH M GrossoSul Mato Grosso colas Sourcoe: 144427.6 10 12 8 9 7 8 1S 34.02 38.84 53.28 67.38 47.81 48.76 41.05 74.51 62.35 51.99 28 13 43 26 26 35 20 47 47 38 82.8ox 38.72x 81.23X 38.73x 56.15S 71.40x 48.37% 63.41X 75.30X 88.741 3 9 3 1 2 7 14 5 3 5 89.72 58.97 98.20 135.43 113.33 89 49 es le 96 98.95s 82.881 68.32x 85.41X 84.41X 3 4 3 3 4 5 3 4 7 9 6 7 66.63 90.41 103.05 86.08 47 84 92 73 71.12X 92.37x 89.701 84.52X 4.80 1.81 7.68 300.87 5.22 6 10 6 3 6 12 5 75 20 17 25 11 78 26 109.49 125.48 73.05 188.81 108.84 es 76 42 114 es 61.781 60.23a 57.04X 80.156 59.44U 16.97 6 6 11 88.02 69 78.401 1S 19 22 13 21 18 14 IGE - AnuarloEstatisticodo Brasil 1989 NINIFAZ/STN/SAFEM MINIFAZ/SEF GDP: eatimted based on data extractedfrom 'Stateand Local Finance- Public Policy in Brazil'(June89), by Remy Prud'homae 21 Table10 ONSTATES- 1988 GOVERNMENT OF FEDERAL BRAZIL:NETIMPACT CURRENT NCzSPERCAPITA Fedoral Unit Federal Federal Net Impact Population (thousands) Budgetary qovornent of Federal RevonuosExpendituresGovornmnt 6.64 6.63 80.86 22.65 14.14 16.90 71.90 68.94 23.74 169.96 14.18 91.00 68.26 62.31 -7.10 147.48 0.04 75.11 29.77 12.71 8.49 9.63 8.76 6.67 23.07 8.47 12.34 19.28 13.33 14.56 16.97 17.78 23.12 18.81 38.16 17.64 27.28 12.46 19.20 11.06 11.60 8.16 14.37 12.28 10.08 9.07 14.94 -6.82 6.88 16,845.8 2,429.4 13,56.1 31,651.6 62,982.8 31.31 38.68 118.65 119.90 94.90 12.12 12.68 239.46 9.31 69.66 -19.19 -26.12 120.90 -110.59 -35.24 Parana Santa Catarl R Grande Sul SOUTH 8,732.3 4,311.3 8,888.6 21,982.1 36.33 30.23 43.06 37.46 9.23 11.00 40.62 22.26 -26.09 -19.22 -2.54 -16.20 U Grosso Sul Mato Grosso Goias 1,714.6 1,629.3 4,740.4 10.88 11.68 10.69 13.75 19.02 12.72 2.87 7.34 2.08 CENTERWEST 8,084.2 10.93 14.21 3.28 Rondonia Acre Amazonas Roraima Par& Amapa 1,019.2 39e.6 1,896.6 113.1 4,724.4 240.2 Maranhao Plaul R GrandeNor Paraiba Pernambuco Alagoas S-rgipe Behia NORTHEAST 4,970.4 2,667.4 6,289.4 2,236.2 3,152.6 7,114.6 2,342.3 1,369.0 11,804.4 41,296.8 Mines Gerais EspiritoSon Rio do Janoi Sao Paulo SOUTHEAST NORTH Ceara' 8,388.9 17.05 6.87 Sources: - BalancosCerais de Uniso - 1988 (vol II p 34) - AnuarioEstatisticodo Brasil 1989: Poputation(page75) Notes: - Center West regiondata excludethe FederalDistrict (OF). transfers - FederalExpenditurosincludeall the intergovornmental recordedin the BalancosGerais da Uniao. NET IMPACT=FEDERALEXPENDITURES- FEDERALREVENUES 22 alliances in any federation. The following sections prescnt a review of federal system in selected countries to make an impressionisticevaluation of the Braziliansystem. 1.41 The Practice of Federalism inSelected Countries USA. USA has a 3-tier systemwith states as the weakest linktraditionallyin the system. In federal-localfiscalrelations states are often by-passed.The intrusiverole of the federal government has largelybeen the result of urban and racialproblemsof the 1960sand dominanceof state legislaturesby ruralinterests. Taxand expenditureassignmentin the U.S.is not consistentwith the economicprinciples enunciated earlier. Other than taxes on international trade, exclusivelyreserved for the federal govemment and property taxes for state and local levels, all other tax fields are open to all levels of govemment. Federal, state and local govemments have overlapping and uncoordinated personal and corporate income tax administration. Expenditure assignmentis also not clearlydelineated. Defense, foreign affairsand space administration,foreign and interstate commerce,the postal service, coinage, weightsand measures,patents and copy rights and crimesa-ainst the United States are reserved for the federal govwunment. In housing, education, transportation, and social welfare, all three levels are involvedto varyingdegrees. Federal governmentgets involvedimsuchlocal functionsas fire protection, pothole repair, rat control, urban transit, local librariesand museums,and zoningregulationsas a result of pork-barrelpolitics. Federal govemmentoften exercisesstrong control over local priorities through carrot (specific purpose transfers--in early 1980s there were 492 federal programs) or stick (court ordered racial integration of school pupils and teachers leading to a decay in schoolingin inner cities; highwayspeed limits;withholdingof federal highwayfunds from states not raisingthe drinking age to 21). The hallmarkof the U.S. federal systemis diversity,a "fend for yourself federalism"and a "jungle for tax administration". The efficiencycosts of such a systemare large which only an advanced nation like the U.S. could afford. Major progress to reform this systemwas made during Carter and Reagan years. 23 AUSTRALLA:Australiahas a two-tieredhighlycentralizedsystem. The centre emphasizes uniformityof publicservicesacrossthe nation and uses conditionalgrants to achievethat purpose. Tax administrationand collectionisprimarilycentral (80%of revenues). Localgovernmentsarehandmaiden of states but are given reasonable autonomy in local service delivery. The Commonwealthhas sole responsibilityin defense, trade, immigration,external affairs,socialsecurityand employment. States are responsible for education, health and social services,transport, railways,electricity and water. The federal governmentneverthelessexercisesstrong influencein -heseareas through conditionaltransfers. In tax assignment,customs and excisesare reserved for the Centre and concurrent responsibilitiesare assigned in all other areas. One half of customs proceeds are mandated for states. The Uniform TaxationAct of 1942eliminated any role for states in income taxes and subsequent court rulingsclosed sales and excisetaxation fieldsto states. State-Local governmentsare responsiblefor 50% of the total outlay of the publicsector but raise only 17% of revenues. CANADA: Canada has a two-tiered highlydecentralized system. In 1988, 59% of total expenditureswere undertaken at the state-locallevel. Taxand expenditureassignmentare transparent. Tax assignmentis overlappingbut harmonized. Expenditure assignmentis as follows: Federal: money,banking, trade, airlines, railways,foreign affairs,defense, unemployment insurance. Federal-Provincial: Pensions, immigration,agriculture,industry. Provincial: Education, health, socialwelfare, police, natural resources and highways. WEST GERMANY: The Upper House of the Parliament is called the Council of States (BUNDESRAT). State ministers or their deputies are represented on this council and vote at the direction of their governments. This provides a check to any centralizingtendency in the federation. The expenditure assignmentis as follows: Federal: defense, foreign affairs,immigration,railways,air transport, post office. 24 Concurrent: publicwelfare,regulationofcommerce,industry,banking,insuranceand labour relations, promotion of social responsibility, public roads and shipping. Note that all concurrent responsibilitiesare carried out by states (Laender). States: Education, culture and residual powers. Tax Assignment: Federal government has exclusiveauthority over customs and federal monopolies (alcohol etc.) and priorityover remainingtaxes. Taxes are primarilycollected by the Centre and then shared with the states and local governments based on agreed percentages. MEXICO: Mexico is a highly centralized federation. 80% of public expenditures are controlled by the central government. In addition to the usual functions of a central government (defense,justice, external affairs,commerce and finance), the federal government in Mexicoassumes responsibilitiesfor functionswhich are allocated to other levelsof government in other federations such as health and education. States are responsible for public transport and infrastructure expenditures. Stateshaveno ownsource revenues and solelydepend upon federal transfers (18.1%of federal revenues are transferred and distributed 50% on a per capita basis and the remaining 50% based on historical shares). The designof these transfers are creatingcertain anomaliesin taxadministration. For example, in 1988, several states showed net negative VAT collections. Credit vouchers issued far exceeded collections. 1.42 An Impressionistic Evaluation Table 11 presents a bird's eye-viewof selectedfederal systemsreviewedearlier. The table suggests that Brazil compares quite well to other federations on decentralization indicators. It nevertheless can learn a great deal from other federations in designingtransfers. Of the countries reviewed here, Canada and West Germany offer two alternative neat models of a federation. The former emphasizesdiversityin publicserviceswith minimumstandards achievedby tax harmonization and transfers. The latter emphasizesuniformityin publicservicesachievedthrough rationalexpenditure TABLE II FEDERALSY8TfHS- AN wIPR%SSIOUISTIC" VALUATION Selected Indicators Tax SeDaration Australia Nexico Canada National Unity Strong strong Fairly Weak Strong Fairly State Influence on Federal Falrly strong Tax Overla nDin Unlited States strong Strong Tat Sharing West Germany Brazil Strong Strong veak Strong Strong weak Strong Strong Policy-makers State Government Constitutional Statue Strong Weak Fairly strong de jure; very strong de facto Fairly Actual State Control of Local Goverument Strong Strong Strong Varies from fairly strong to fairly weak Strong Weak Range of Local Goverment Responsibilities L$sited Limited Fairly extensive Extensive Limited Extensive Local Government Influence State Policy-makers on Weak Weak Fairly strong Fairly strong Weak Strong Local Coveroment lnfluenc, StatePolicy on Weak Weak Fairly strong Fairly strong Weak Strong Local Govermnent Inflence Federal Policy on Weak Weak Weak Fairly strong Weak Very strong TWo-tiered; centralized Three-tired; centralized Two-tiered; decentralized Three-tiered unstructured Two-tiered integrated Three-tiered decentralized Important; emphasis on conditional grants Important Important; emphasis on unconditional grants Important; Emphasis on conditional grants Unimportant Emphasis on tax sharing Important Federallunterstate Equalization Performance Very strong; revenue and expenditure disparities reduced substantially. Weak Strong;revenue disparities reducedsubstantiaily. Very weak Strong;revenue Weak and some expenditure disparities reducedsubstantially StateTax Performance Fairlyweak Weak Strong Fairlystrong Fairlystrong Strong Weak Fairlystrong Fairlystrong Weak Weak The Character of Fiscal Federalism Federal-State Transfers Intergovernmental LocalGovernmentFiscalIndependence Fairlystrong Selected Tax SeDaration Australia Mexico Indicators Equalization Formmla Fed. -state Explicit implicit piecemeal Canada a Tax Overlanpint United Fed. -State Equalization Fiscal States Implicit piecemeal and Tax Sharina West Germanv Brazil Explicit complex Implicit piecemeal and State tax base conformity Yes No Yes No Yes No State tax rate uniformity Yes No No No Yes Yes Yes No Yes Yea Yes No No Yes Yea Yes Yea No Single collection and adamin. State-Local revenues match responsibility more or less Source: tax Some data for this table are extracted from a Table prepared "Rating Federal Systems - An Impressionistic Evaluation). by John Shannon, Washington, D.C. ACIR, 1980 entitled and 27 assignmentand tax sharing arrangements. Smaller developing countries like Sri Lanka could benefit from the German modelwhereas largeand diversecountrieslike Brazil,India,Mexicoand Pakistanhave much to learn from the Canadian model. Interestingly,U.S. reforms in recent years have also movedit in the direction of a Canadian style two-tiered system. 1.5 Conclusions Fiscal federalismissues have been more thoughtfully addressed in Brazil than any other developing country. The New Constituf an makes a reasonablyclear assignmentof expenditure and revenue assignment. This assignmentis also broadly consistent with economic principlesdiscussed in this paper. Some problemsnevertheless remain. De facto expenditureresponsibilitiesare not consistent with the constitutional intent. The taxsharing arrangementshave favouredthe municipalsector at the expense of the federal government. The federal government is facing a major squeeze while municipalitiesare reducingown taxefforts due to generous availabilityof funds through tax sharing. As the fiscalpressures on the federal government mount, it is showingwillingnessto discussexpenditure turnbacks on a program by program basis. It is expected that by 1993, much of the kind of federal irvolvement that is criticized here would have vanished due to revenue constraints. The federal government is well advisedto move in this direction sooner. In the area of tax assignment,ICMS (VAT) administration has been handed over to the States Council which contrary to principles enunciated here allows states some flexibilityover its coverage in their jurisdictions. Interstate trade is causinga major havocwith the ICMS administration with energy producing states facing a major crunch on their revenues (ICMS is on the final sales of energyonly). Taxbases for IPI, ICMS and ISS somewhatoverlapbut are administeredby three different levels of government. Further, rural property tax (1TR) which is more suitable for administration by state governmentsis currently a federal responsibility. The followingreform options are suggestedto deal with these issues: 1. Immediate turnback of direct federal involvementin functionsof purely local nature such as primaryand secondaryeducation, urban grading,bridges,zoningetc. Further, 28 administrationof health and educationshould be a state responsibility.Therefore, the roles of federal ministriesof health and education be reduced to setting minimum standards and providingper capita block gr. Asto induce compliance. 2. The three sales taxes,IPI, ICMS and ISS be combinedinto one tax to be administered by the federal government on behalf of state and local governments. Thus proceeds from the taxbe shared bythe three levelsin proportion to their current intake from this source. 3. The administrationof rural property tax be turned over to the state leveL 4. The revenue sharing and transfer programsbe restructured as discussedin the section on intergovernmentaltransfers. 29 NOTES 1. The implementation of above principles require operation of voting with feet mechanism.This brings us to the Tiebout literature which suggests that voting with feet will lead to jurisdiction formationscreating a market analogue to publicserviceprovision.Oates had earlier suggestedthat allocativeefficiencyquestions associatedwith votingwith one's feet could be settled by examining tax and benefit capitalization. The existence of capitalization implies allocative efficiency.This conclusionhas been rejected by Jan Brueckner (1979, 1982)and Ar.war Shah (1983, 1988, 1989, 1990).Brueckner's test is based on the theoretical result that a non-positiverelationship between publicservicesand residentialpropertyvalues is a definiteindicationof over-provisionof localpublic goods beyond optimal levels in the case of typicalmixed communities,i.e. , communitieshaving substantial businessproperty. An alternativetest proposed by Shah uses the criterion that, when the level of local spending is optimal, a balanced budget change in local spending and residential property taxation should leave residential property values unaltered. Thus a positive impact of a balanced budget change would indicate under-provisionand a negative impact over-provisionof publicservices. 2. For specific examples of these see Diogo Lordelle de Mello (1988), "Resources Mobilization Strategies for Urban Development in Brazil",processed. 30 2.0 TRANSFERSIN BRAZIL INTERGOVERNMENTAL The existingstructure of federal-state-localtransfers in Brazilcan be broadlyclassifiedinto two categories,namely:(1) tax transfers or revenue sharing arrangementsmandated by the Brazilian Constitution and (2) Specific purpose transfers includingnegotiated transfers (convenios). In 1989, other than meetingits obligationsforestablishedprogramssuch asthe Unifiedand Decentralized Health Care System (SUDS), the federal government did not disburse any additional funds through the convenios. In a typicalyear, though conveniosaccounted for nearly 10% of federal transfers to states. The followingparagraphs present a brief description of revenue sharing arrangements and other transfers. 2.1 RevenueSharingArrangementsin Brazil Revenue sharing arrangementshave been specifiedin the new BrazilianConstitution. The Constitution provides strictcriteria for the allocationof revenuesto different levelsof government and some guidelines on distribution of these through specialfunds among units at the same level. Specific distribution criteria are specified by parliamentary regulations. There is no stipulation as to the final dispositionof these funds by the receivinggovernment. The Constitution, nevertheless, provides that the Union must applya minimumof 18%, statesand municipalitieseach at least 25% of all tax revenues (including intergovernmental transfers) on education. Sharing of federal revenues from income and industrial product taxes are through participation funds established for this purpose. In 1988, the Federal Government transferred aboutNCz$ 1.5billionto states and municipalitiesthrough the revenue sharing mechanisms(see Table 12). Of this total amount, 52% went to relativelylessprosperous states in the north and the northeast regions of the nation. A description of major programs for revenue sharing is givenin the followingsections. Table 12 BRAZIL: FEDERAL TAX TRANSFERS TO STATESANDMUNICIPALITIES - 1988 (Current NCzS thousands) Federal Unit STATEGOVERNMENT --------------------------Total Per capita X of total MUNICIPALGOVERNMENT -----Total P-r capita X of total TOTAL Total Per capita X of total Rondonia Acre Amazonas Roraima Para Aapa NORTH 18,222 19,390 22,638 13,862 42,730 18,654 135,496 17.88 48.90 11.94 122.56 9.04 77.66 16.15 2.43X 2.68X 3.02X 1.85X 5.70! 2.49! 18.06x 6,431 3,928 10,836 2,394 23,930 2,909 50,428 6.31 9.91 5.72 21.16 5.07 12.11 6.01 0.90X O.55X 1.52X 0.34X 3.35X 0.41X 7.07X 24,662 23,318 33,474 16,266 68,660 21,663 186,924 24.19 58.81 17.86 143.73 14.11 89.77 22.16 1.88x 1.59x 2.29x 1.11X 4.55X 1.47X 12.70X Maranhao Piaui Ceara' R Grand.Nor Parslbe Pernambuco Alagoas Sergipe Bahia NORTHEAST 44,111 26,338 45,134 25,881 28,368 42,048 24,232 24,188 69,702 329,993 8.87 10.26 7.23 11.57 9.00 5.91 10.36 17.67 6.17 7.99 5.88x 3.61X 6.02! 3.45X 3.78! 5.6ox 3.23X 3.22X 9.29x 43.98x 27,961 17,162 33,858 18,651 22,983 36,805 16,090 10,996 65,302 248,698 5.63 6.68 5.43 8.30 7.29 5.03 6.87 8.03 5.78 6.02 3.92X 2.40X 4.765 2.60! 3.22X 5.02% 2.26X 1.54X 9.15X 34.86X 72,072 43,490 78,993 44,432 61,341 77,853 40,322 35,184 135,004 578,691 14.60 18.94 12.86 19.87 16.29 10.94 17.21 26.70 11.94 14.01 4.92x 2.971 S.40X 3.04X 3.51X S.321 2.75X 2.40! 9.22x 39.63x 1 Mines Gerais EspiritoSan Rio de Janei Sao Paulo SOUTHEAST 85,702 11,553 31,070 53,892 162,217 4.28 4.76 2.29 1.70 2.68 8.76x 1.54X 4.14X 7.1ex 21.62X 98,047 12,843 24,619 102,181 237,690 6.39 5.29 1.82 3.23 3.77 13.74X 1.80x 3.45X 14.32X 33.31X 163,749 24,396 65,690 158,072 399,907 10.67 10.04 4.11 4.93 6.36 11.19x 1.67X 3.8OX 10.66x 27.32X Parana Santa Catari R GrandeSul SOUTH 28,686 15,301 25,795 69,782 3.28 3.55 2.90 3.18 3.82x 2.04X 3.44X 9.30x 48,783 28,009 44,524 121,316 5.59 6.50 5.01 5.53 6.84x 3.93x 6.24X 17.00X 77,488 43,309 70,319 191,097 8.87 10.05 7.91 8.71 5.29X 2.96X 4.80% 13.06x M Crosso Sul Mato Grosso Coias (a) CENTER WEST 9,784 15,711 27,306 5.71 9.64 6.78 1.30X 2.09x 3.64X 10,492 12,335 32,507 6.12 7.57 8.86 1.47X 1.73X 4.61X 20,276 28,046 59,813 11.83 17.21 12.62 1.39% 1.92X 4.09x 52,800 6.53 7.04X 55,334 6.84 7.7ex 108,134 13.38 7.39x BRAZIL (*) 750,288 5.26 1OO.OOX 713,465 5.00 1.OOOOX 1,463,753 10.26 1OO.OOX Dist Federal Brazil + DF 4,956 755,244 2.84 6.23 2,287 716,762 1.31 4.96 7,243 1,470,996 4.15 10.19 Sources: Unpublisheddata Ministerioda Fazenda- SecretarisCeral - Secretarisde Economis* Financas(MINIFAZ/SEF): Note: a Data excludethe FederalDistrict(DF) 32 2.11 State ParticipationFund (FPE: Fundode Participacaodos Estados) The federal government deposits21.5%each of incometax (IR) and industrial products tax (IPI) in a special fund for later distribution by the States' Council to individualstates . In determining state shares from this fund, the fund first sets aside 85% of total funds for distribution to states in the north, north-east and centre-west regions of the countryand the remaining 15% for the south and the southeast region. The Act 104-A, 1989 argues that this initial allocation is necessaryto safeguard regional equity objectivesas tax assignmentcarried out by the Constitution appears to favour the rich states more than the poorer ones. The Act has further establisheddeadlines for the federal government for the release of funds for distribution to states. The intent of these deadlines is to limit federal government incentivesto benefit from inflationarygains by withholdingstate funds a bit longer than absolutely necessary. The formula for the distribution of funds among states takes into account population (a proxy for fiscalneed) and inverse of per capita income (fiscalcapacity indicator). The criteria specified for this purposes is expressed mathematicallyin Box 1. A proposal to extend this formulato incorporateland area (fiscalneed measure),interstate trade orientation (spilloverfactor)and ratio of ownrevenues to expenditures (fiscaleffort indicator)is currentlyunder discussionin the Senate (see also Lei #165/89, Senado Federal). This formulayields the participationcoefficientsfor individualstates as given in Table 13. These coefficientshoweverwere found unacceptable by the Councilof Finance Ministersof the States and instead they developed modifiedcoefficientsas givenin column 3 based upon mutual negotiations. These coefficientsare applicabletill 1991. Participation coefficientsfor 1992and later yearswillbe reestablishedbased upon a reviewof the workingof the formula and new data from the 1990census. A similar fund established for distribution of federal transfers to municipalitiesis named Municipal Participation Fund (FPM:Fundo de Participacao dos Municipios). This fund is discussed below: 33 BOX 1 DISTRIBUTIONCRITERIA FOR THE STATES PARTICIPATIONFUND (FPE) (FPE), - 0.85 * G * SN + 0.15 * G * Ss Where G - 0.215 * (IR + IPI) I (POPP)i * (YPCF)1 1 N,SN [ (POPF)i * (YPCF)i] s-i Participationcoefficientfor a state in the northeast,north and SN - and center-westregions. S = Participationcoefficientfor a state in the south and southeast regions. IR IPI POPF - federal tax collectionfrom income taxes federal tax collectionsfrom industrialproducts tax. Population factor. The following table is used for this purpose. Z of nationalpopulation representedby each state upto 2% 2-4x: for the first 2% for each additional0.3% 5-10%: for the first 5% for each additional0.5% above 102 POPF 2.0 2.0 0.3 5.0 0.5 10.0 34 Box 1 (continued) DISTRIBUTIONCRITERIAFOR FPE (Continued) YPCF - state income (per capita) YPCF is determinedaccordingto the following table: Average Per Capita income of all states/ Per capita income of state i YPCF ----------------------------------------------------- Up to .00045 Between .0055 and .0065 .0065 - .0075 .0075 - .0085 .0085 - .0095 .0095 - .0110 .0110- .0130 .0130- .0150 .0150- .0170 .0170- .0190 .0190- .0220 Above .0220 .4 .6 .7 .8 .9 1.0 1.2 1.4 1.6 1.8 2.0 2.5 35 Table 13 FRE: PARTICIPATION COEFFICIENTS FOR BRAZILLANSTATES FOR 1990-1991 Federal Unit Formula Share NORTH Rondonia Acre Amazonas Roraima Para Amapa 2.1164 2.6270 2.1850 1.9090 4.7225 2.6200 2.8150 3.4210 2.7904 2.4807 6.1120 3.4120 NORTHEAST Maranhao Piaui Ceara' R Grande Norte Paraiba Pernambuco Alagoas Sergipe Bahia 5.7415 3.3205 7.0585 3.2045 3.6700 5.3960 3,1900 3.1860 9.4635 7.2181 4.3214 7.3369 4.1779 4.7889 4.3214 4.1601 4.1553 9.3962 SOUTHEAST Minas Gerais Espirito Santo Rio de Janeiro Sao Paulo 7.9545 1.5470 4.2435 3.9460 4.4545 1.5000 1.5277 1.0000 Parana Santa Catarina R Grande Sul 4.2400 1.8800 3.4615 2.8832 1.2798 2.3548 CENTER WEST M Grosso Sul Mato Grosss Goias Distrito Federal 1.4735 2.5530 3.1450 0.7535 1.3320 2.3079 2.8431 0.6902 SOUTH Source: Camara dos Deputados Projeto de Lei CornplementarNo. 104-A,de 1989. Ministerio da Fazenda - Secretaria Geral Secretaria de Economiae Financas(MINIFAZ/SEF) Boletim Informativo No. 166,December 1988 Agreed Share 36 2.12 MunicipalParticipationFund(FPM) FPM wasestablished as 13.5%of federal income and industrial product taxes in 1984 and raised to 20% in October 1988. This share of the specifiedtaxeswillrise by 0.5 percentage point each yearuntil the newsystemis fullyphased in 1993with 22.5%of these taxesearmarked for the fund. 12% of the FPM fundsare allocated to state capitalsand municipalitieswith population greater than 400,000. Of the remaining 88%, 6% is set aside for the Municipal Participation Reserve Fund (RFPM: the Reservado Fundo de Participacao de Municipios).The RFPM isavailableonly to larger municipalities other than state capitalswith 1990population at least 4% of national population. The distribution of funds to all municipalitiesare by a formula which takes into consideration population and per capita income of each municipality. Funds vary directlyby population and inverselyby per capita income. Formulae details are given in Box 2. Table 14 provides details on the level of funding and the participationcoefficientsby population size of the municipalunit. In addition to revenues through the FPM, municipalitiesalso receive 50% of revenues from the rural property tax in proportion to the value of real estate properties located in their jurisdictions;100% of payroll deductions of income taxes of municipal employees;70% of tax on gold by origin; 2.3% of revenuesfrom crudeoil based on the value of production; and 50% of hydroelectricityand mineraltaxes by the sales value of the mineralsby origin. 2.2 SpecificPurposeTransfers Non-Constitutional Transfers: Overall there were 117 umbrella federal-state-municipal transfer programs in 1989of which 19 were open to municipalparticipation. These prog- ns can be broadlyclassifiedinto four categories: (1) The first typeof transfers havebeen instituted to simplycomplywith specificlawsother than the Constitutionalprovisions. Major transfers of this type include transfers to the Federal Capital (38% of total in 1987); transfers related to the creation of new states (21% of total in 1987)and financialcompensation(royalties) paid to states for the 37 BOX 2 FORMULAE FOR THE DISTRIBUTIONOF MUNICIPAL PARTICIPATIONFUND (FPM) State Capitals and Municipalitieswith 1990 PoRulationgreater than 400.000 SCPl4)i 0.12 * G (FPM)8i 0.12 * GFp * xc (YPCF)1 (POPCL) FPM E [(POPCL)i x (YPCF)i] Other Municipalities (POPSM) (FPM)' (FM GFP FP * 0.82 x ____ ZRPOPSM)i + 0.06 * D * (POPML)ix (YPCF) i E (POPML)ix (YPCF)i When G Di - 0.225 (IPI + IR) - 1 If municipalityhas a populationgreater than 156,216 or (POP) /E(POP)> 0.04 (includesnon-capital municipalitieswith populationgreater than 400,000). 0 otherwise Superscriptsc refers to state capital Superscriptom refers to all other municipalities. POPCL - population factor for state capitals and larger municipalitiesdeterminedaccordingto the followingtable: Z of total national populationin this category living in municipality POPCL Up to 1% Between 1 a-id5% 1.0 - for the first 1Z 1.0 - for each additional .5% or fraction,add .5 Above 52 5.0 YPCF - state income per capita factor defined in the FPE section. 38 BOX 2 (Continued) FORMULAE FOR THE DISTRIBUTIONOF F.P.M (Continued) POPSM - population factor for small and medium municipalitieswith population less than 400,000 determined according to the followingtable: Populationof the Municipality POPSM Up to 16,188 - for the first 10,188 - for each additional 3,396 or fraction .6 + .2 Between 16,800 and 50,940 - for the first 16,980 for each additional 6,792 or fraction 1.0 + .2 Between 50,940 and 101,880 - for the first 50,940 for each additional 10,188 or fraction Between 1010,940and 156,216 - for the first 101,880 for each additional 13,584 or fraction Above 156,216 - 2.0 + .2 3.0 + .2 4.0 POPML = population factor for eligible (medium and large municipalities) determined according to the following table: Z of total populationof eligible municipalitiesliving in municipality POPML Up to 2% Between 2 and 5% 2.0 - for the first 2% 2.0 - for each additional .5% or fraction,add Above 5Z .5 5.0 Xt__=_S 8________s __-------------- 39 Table 14 BRAZIL: MUNICIPAL PARTICIPATION7UND COEFFICIENTS (1989) Population (Inhabitants) Coefficient Amount of municipal participationin USS 000's Up to 10,188 10,189 to 13,584 13,585 to 16,980 16,981 to 23,772 23,773 to 30,564 30,565 to 37,356 37,357 to 44,148 44,149 to 50,940 50,941 to 61,128 61,129 to 71,316 71,317 to 81,504 81,505 to 91,692 91,693 to 101,880 101,881 to 115,464 115,465 to 129,048 129,049 to 142,632 142,663 to 156,216 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 698 930 1,163 1,396 1,628 1,861 2,094 2,326 2,559 2,792 3,024 3,257 3,490 3,723 3,955 4,188 4,421 above 156,216 4.0 4,654 Source: Instituto Brasileiro AdministracaoMunicipal, Noticiero, No. 96, September/89. 40 extraction of oil within their jurisdiction. Typicallysuch transfers constitute 60% of annual total non-constitutionaltransfers. (2) The second type of transfers are commonly referred to as convenios or negotiated transfers. These are not regulated by law and are based on negotiations between the federal and other levelsunitsindividually.Support forregionaldevelopment,agriculture, education, health and housingare the priority areas for receipt of funds from convenios. They constituted 40% of the non-constitutionaltransfers in 1987. (3) The third type of transfers are specialinvestmentfunds/projects.These projects maybe undertaken by state and local governments on behalf of the federal government. Financingcomesfromthe General Revenue Fund aswellas the SocialInvestmentsFund (FINSOCIAL:Fundode InvestimentoSocial)and the programsfor NationalIntegration (PIN) and Redistribution of Land to Stimulate the Agrarian Economies of North and Northeast (PROTERRA). The transfers associatedwith these fundsare not recorded in the BalancosGerais da Uniao (BGU) as intergovernmentaltransfers. Nearly70% of these funds are recorded under the umbrellaof planning. There is flexibilityin the use of these funds. The only requirement is that they be used for the development of basic social servicesand infrastructure. (4) The fourth type of transfers are the transfersmade through government agencies. These transfers also do not showup in the BGU. To account for these,it is necessaryto enquire from manydifferent agenciesfor what theyrecord asnon-taxtransfersespeciallytransfers made in the form of convenios.The collectionof data on these transfers is a difficulttask since the governmentagenciesdo not aggregatedata so as to determine whichpart of the transfers went to state and localgovemments. The National Secretaryof Treasury(STN: Secretaria do Tesouro Nacional) reported that in 1986 8.5% of tax revenues went to intergovernmental transfers not related to the sharing of tax revenues. Most of these transfers comefromthe ministriesof Planning,Education,Finance, Urban Development and Health. The northeast regiontraditionallyreceivesnearlyone third of total transfers made through gcvernment agencies. 41 A description of negotiated transfers briefly describedearlier follows: 2.21 NegotiatedTransfers(Convenios) A convenio represents federal transfer of funds to state and local governments for undertaking expenditures on behalf of the federal government in areas of federal government responsibility. These transfers are determined by supplemental lawsor directly negotiated between different levelsof government. These transfers constituted 22% of total federal-state-local transfers and 8% of federal revenues in 1987. Nearly90% of these transfers (excludingthose for enterprises) go to states and the remaining10%to municipalities(see Table 15). Transfers to municipalitiesare usually for urban development and housing prograr is. Occasionallythough large sums of moneywere made available to municipalitiesin seekingtheir politicalsupport e.g. CZ$ 6 billion for support for writinga new constitution in 1987(see Afonso, 1989). In 1988, there were over 3000 convenioswith multitude of objectives. Tables 15 and 16 provide 1988distributionof these transfers bydestination (level of government) and by source (federal ministry). An accountingof these by programsand by objectivesis not possible as it is an activitywhich the federal governmenttraditionallydoes not monitor. A specialsurveydone bythe Ministryof Finance in 1988has analyzedthe data on conveniosby functionalclassificationand by state for 1985-1986.(see Tables 18 and 19 and appendix Tables A.1-A3). This survey established that nearly two-thirds of conditionaltransfers in 1985-86were meant for planning and education. The northeast region received a lion's share of transfers for education (46.4%of total) and the centre-west region received62.6% of total transfers for planning. This surveyfurther confirmedthat most transfers on account of convenios were based on ad hoc decisionsand devoidof any formalcriteria and therefore could not be subjected to any formal analysis. The 1988federal allocationof negotiated transfers asreported in Table 15has an interesting pattemofdistnrbutionamongstates. For example,the State ofMaranhao, then-President Sarney'shome state had a very peculiarparticipationin the overall distributionof negotiated transfers in 1988.Its state government alone receivedhigher fundingthrough negotiated transfers than all the state governments Table 15 - 1988 TRANSFERS BRAZIL: FEDERALNEGOTIATED Convonlos,Agreements,Adjustments,Protocols,etc. (CurrentNCz3 thousands) STATEGOVERNMENT --------------------------X of total Total Per capita Federal Unit MUNICIPALGOVERNMENT --------------------------% of total Total Per capita ENTERPRISES --------------------------X of total Per capita Total TOTAL --------------------------Per capita Total X of total 1.13X 0.07X 4.92% a6s 153 919 467 1,798 3,906 7,697 0.36 0.39 0.48 4.13 0.38 16.26 0.91 0.67% 0.29X 1.72X 0.88% 3.37X 7.33X 14.26X 963 873 6,008 1,569 6,079 3,144 17,602 0.94 2.20 3.17 14.10 1.08 13.09 2.11 0.45% 0.41% 2.81% 0.7sx 2.37X 1.47% 8.26X 8,471 3,793 12,931 2,281 11,776 7,361 46,812 8.31 9.57 6.82 20.17 2.49 30.65 5.58 1.21% 0.54% 1.84% 0.33X 1.68x 1.05X 6.66X 4.56 3.71 2.46 3.91 2.31 1.30 1.29 5.30 0.96 2.28 5.22% 2.19X 3.63X 2.01X 1.68x 2.13% 0.69% 1.87% 2.60X 21.64X 4,605 3,233 1,834 3,013 2,046 2,173 2,835 2,394 6,503 28,636 0.93 .28 0.29 1.35 0.66 0.31 1.21 1.75 0.58 0.69 8.64% 6.07X 3.44X 5.86x 3.84x 4.09% 5.32X 4.49% 12.21% 53.76% 18,641 3,496 10,836 6,032 4,387 6,810 1,990 2,690 9,009 62,791 3.75 1.36 1.74 2.26 1.39 0.96 0.85 1.89 0.80 1.62 8.72% 1.83% 6.07X 2.36% 2.05% 3.18% 0.93% 1.21% 4.21% 29.36% 45,933 16,247 28,017 16,779 13,724 18,236 7,841 12,238 26,386 186,398 9.24 6.33 4.49 7.50 4.36 2.66 3.35 8.94 2.33 4.49 6.65x 2.32% 3.99% 2.39% 1.96x 2.60% 1.12X 1.74% 3.76X 26.43X 11,173 1,506 4,963 243,268 260,899 0.73 0.62 0.37 7.69 4.14 2.67% 0.35% 1.14% 66.02% 60.09% 2,769 633 1,114 2,634 7,140 0.18 0.26 0.08 0.08 0.11 6.18% 1.19X 2.09% 4.94% 13.40X 19,199 1,162 88,209 10,980 99,660 1.26 0.48 6.03 0.36 1.68 8.98% 0.54% 31.89% 5.13X 4e.66% 33,131 3,300 74,286 268,872 367,689 2.16 1.36 5.48 8.12 6.84 4.72% 0.47% 10.69% 36.63% S2.41% Parana Santa Catari R Grande Sul SOUTH 7,346 5,927 13,368 26,631 0.84 1.37 1.60 1.21 1.89% 1.37X 3.08% 6.13X 1,808 2,201 2,271 6,280 0.21 0.51 0.26 0.29 3.39% 4.13% 4.26X 11.79X 8,834 6,984 9,098 21,916 0.78 1.39 1.02 1.00 3.20% 2.80% 4.25% 10.26% 15,988 14,112 24,727 64,827 1.83 3.27 2.78 2.60 2.28X 2.01X 3.63x 7.82% M Grosso Sul Mato Grosso Goias 7,593 7,661 16,202 4.43 4.63 3.42 1.75% 1.74% 3.73X 917 834 1,875 0.53 0.61 0.40 1.72X 1.67% 3.52% 3,064 1,980 6,924 1.78 1.22 1.46 1.43X 0.93% 3.24% 11,6e4 10,385 26,001 6.74 6.36 5.27 1.65% 1.48% 3.66% CENTER WEST 31,346 3.88 7.22X 3,626 0.46 6.81% 11,968 1.48 5.59x 46,930 5.81 6.69x 434,201 3.04 100.00% 63,278 0.37 100.OOX 213,877 1.60 100.00% 701,366 4.92 100.OON 1.86% 0.64% 1.38X Rondonla Acre Amazons Roraiua Para Amapa NORTH 7,253 2,787 6,004 219 4,898 312 21,353 7.02 6.98 8.17 1.94 1.04 1.30 2.66 0.06X Maranhao Piaui Ceara' R Grand. Nor Paraiba Pernambuco Alagoas 22,687 9,518 15,347 8,734 7,292 9,252 3,016 7,262 10,874 93,972 Minas Gerais Espirito San Rio de Janei Sao Paulo SOUTHEAST Sergipe Bahia NORTHEAST BRAZIL Sources: Ministerloda Fazenda - Secretaria do Tesouro Nacional - Secretariade Contabilidade (MINIFAZ/STN/SECON): Unpublished data Table16 - 1988 NEGOTIATED TRANSFERS BRAZIL: FEDERAL Protocols, etc. Agreemnts,Adjustments, Convenios, (CurrentNCzS thousands) INDUSTRYA COMMERCE MINING A ENERGY IRRIGATION EDUCATION MUNICIPAL GOVERNMENTS STATEGOVERNMENTS MINISTRY-FUNCTION X of total Total Per capita 287 0.00 0.06% 19,719 31,166 0.14 0.22 3.81% 6.02% Per capita 46 738 536 21,249 0.00 0.0. 0.00 0.16 Total Per capita 0.08% 1.31% 0.95% 37.e4% 297 1,005 6,017 21,211 0.00 0.01 0.04 0.15 0.12% 0.40% 1.99% 8.42% 42,198 455 112,463 67,883 81 0.30 0.00 0.79 0.48 0.00 18.76% 0.18% 44.64% 28.94% 0.03% 0.46% X of total 162 0.00 1 0.00 AGRICULTURE PLANNINGA ADMINISTRATION TRANSPORTATION INTERIOR LABOR SCIENCE A TECHNOLOGY OTHER 16,416 68,548 860 88,187 68,636 239,585 6,490 0.12 0.48 0.01 0.46 0.48 1.68 0.06 3.17% 13.24% 0.17% 12.78% 13.26% 46.26% 1.25% 1,049 338 1,691 30,761 8 0.01 0.00 0.01 0.22 0.00 1.86% 0.59% 3.00% 64.48% 0.01% 44 0.00 0.08% 1,170 0.01 TOTAL 517,893 3.63 100.00% 56,447 0.40 100.00% 251,942 1.77 COMMUNICATIONS MINISTRY-FUNCTION TRANSFES NEGOTIATED TOTALOF ALL FEDERAL Total INDUSTRYA COMMERCE MINING A ENERGY IRRIGATION EDUCATION COMMUNICATIONS AGRICULTURE PLANNING A ADMINISTRATION TRANSPORTATION INTERIOR LABOR SCIENCE A TECHNOLOGY OTHER TOTAL 0.00% Total ENTERPRISES Per capita X of total 629 1,743 25,272 73,625 0.00 0.01 0.18 0.62 0.08% 0.21% 3.08% 8.91% 163 0.00 0.02% 59,683 69,339 184,821 68,724 239,586 7,704 0.42 0.49 0.81 1.16 0.48 1.68 0.05 7.22% 8.39% 13.92% 19.95% 8.32% 29.00% 0.93% 826,282 5.79 100.00% 115,014 Sources: Unpublished da Ministerio da Fazenda- Secretaria do TesouroNacional- Secretaria de Contabilidade (MINIFAZ/STN/SECON): IBGE) Poputationin 1988: 142,684.3(thousands; Notes: SEPLAN/PR, PRESIDENCIA DA REPUBLICA Planningand Administration includes: MINIFAZ,SADEN/PR, Otherfunctionsare: JUSTICE,NAVY,AERONAUTICS, FOREIGNAFFAIRS X of total 0.06% 100.00% Table 17 BY FUNCTION TRANSFERS BRAZIL: FEDERAL (1968 NCzS Per Capita) states OperatingCapitalSp. Agriculture 0.02 0.22 RegionalDevelopment 4.14 0.43 2.60 0.06 Educationand Culture 0.00 0.76 Energy 0.17 HousingA Urban Development 0.02 IndustryA Co e rce 0.00 0.00 Health and Sanitation 0.a6 0.19 Transportation 0.78 0.01 0.21 0.00 Social Security Other Functions 0.49 0.06 Total 9.66 1.13 Municipalities Inves 0.09 0.09 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.16 0.36 Total 0.32 4.65 2.65 0.75 0.19 0.00 0.87 0.79 0.00 0.93 11.15 States A Municipalities ---------------------------------- Operating Capital Sp. 0.02 0.22 Agriculture 8.84 0.44 RegionalDevelopmnt Educationand Culture 4.17 0.06 0.92 0.00 Energy 0.29 HousingA Urban Development 0.02 Industry A Commerce 0.00 0.00 0.65 0.20 Health and Sanitation 0.01 1.16 Transportation 0.00 0.21 Social Security Other Functions 0.49 0.06 Total 15.97 1.27 Inves 0.10 0.11 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.17 0.40 Total 0.88 8.88 4.22 0.92 0.31 0.00 0.88 1.16 0.00 0.94 17.64 Sources:BalancosGerais de Uniso - 1988 AnuarioEstati;ticodo Brasil- 1989 (Population) OperatingCapitalSp. 0.00 0.00 4.20 0.01 1.57 0.00 0.00 0.17 0.00 0.12 0.00 0.00 0.00 0.01 0.87 0.00 0.00 0.00 0.00 0.00 8631 0.14 Inves 0.01 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.04 Total 0.01 4.23 1.67 0.17 0.12 0.00 0.01 0.37 0.00 0.01 6.49 45 Table 18 EVOLUTION OF FEDERAL CONDITIONAL TRANSFERS TO STATES AND MUNICIPALITIES (1983-87) (percent) Agriculture 983 984 985 986 987 100 92 34 178 189 Regional Development Education & Culture Energy 100 58 40 79 8 100 103 177 199 228 100 136 127 125 61 Houslag & Urban Development 100 62 45 1,418 609 Health E Sanitation 100 97 137 169 204 Transportation 0 0 0 - Other Functions 100 105 154 182 291 Source: Sec. Progr. Financeira/STN e "Balancos Gerais da Uniao", HINIFAZ Deflator: IGP-DI media (FGV) Other Transfers - Intergov.rmental Transfers minus Federal Tax Transfers other Functions - Legislative, Judiciary, Planning and Administration, Defense, Coumerce and Industry. Total 100 88 107 166 161 46 Table 19 DISTRIBUTIONOF TRANSFERSTHROUGH "CONVENIOS" TO STATESAND KUNICIPALITIES 1985-86 (percent) ReRions/ States Planninx Education Health Cash In-kindTotal Cash In-kindTotal Acre 2.6 Amazonas 3.0 2.8 Para Rondonia 3.9 Amapa 0.6 0.5 Roraima Maranhao 7.7 Piaui 4.3 2.5 Ceara R.Gde.Norte 1.9 Paraiba 2.7 Pernambuco 1.8 Alagoas 1.9 Sergipe 2.0 Bahia 3.2 Mato Grosso 4.6 M. Grosso Sul 2.4 Goias 2.1 DistritoPedl. 43.5 Minas Gerais 0.5 EspiritoSanto 3.0 Rio Janeiro 0.2 Sao Paulo 0.7 Parana 0.2 Sta. Catarina 0.9 Rio Gde. Sul 0.2 NORTH NORTHEAST CENTRE-WEST SOUTHEAST SOUTH 13.5 28.2 62.6 4.4 1.3 100.0 Others Total 0.7 1.8 4.1 1.3 0.4 0.3 7.0 4.8 6.6 3.4 4.2 7.6 2.6 2.0 8.1 2.1 2.0 4.4 2.2 9.5 1.9 3.9 3.7 5.5 4.3 5.5 0.7 1.8 4.4 1.0 0.4 0.4 7.1 3.6 8.2 2.8 3.9 7.2 2.9 2.1 9.4 1.8 2.7 3.5 1.4 12.9 2.5 1.0 1.8 7.5 3.6 5.1 0.7 1.8 4.2 1.2 0.4 0.3 7.1 4.4 7.0 3.2 4.1 7.4 2.8 2.1 8.5 2.0 2.3 4.1 1.9 10.8 2.2 2.9 3.1 6.4 4.0 5.3 1.6 2.5 3.0 1.4 0.0 0.5 3.7 6.0 5.1 7.6 4.1 9.0 4.8 3.8 8.8 2.8 2.4 3.5 1.0 8.2 2.3 2.6 7.1 2.8 2.1 3.3 0.4 2.2 4.5 1.0 0.2 0.1 4.5 3.1 7.4 3.8 3.5 9.1 3.6 2.0 10.6 1.1 2.0 4.2 7.5 5.3 1.7 7.2 4.7 4.3 2.4 3.3 0.9 2.4 3.9 1.2 0.2 0.3 4.1 4.6 6.2 5.8 3.9 8.8 4.3 2.9 9.3 1.9 2.2 3.9 4.9 6.9 2.0 4.4 5.9 3.6 2.2 3.3 0.2 1.5 1.7 1.2 0.2 0.1 2.7 2.6 2.3 2.4 4.1 11.5 2.1 2.0 5.9 2.8 2.4 3.9 2.8 7.0 4.3 8.2 12.4 5.4 4.9 5.6 7.8 0.0 0.0 79.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.02 0.0 1.5 0.0 0.0 0.0 0.0 0.0 11.2 0.0 0.0 0.0 0.0 1.0 0.2 4.3 5.4 0.1 1.7 6.8 2.5 4.1 1.9 1.9 3.7 1.2 1.1 7.0 2.5 3.1 5.4 0.6 12.6 1.9 3.5 6.2 11.0 3.7 6.5 2.1 2.0 3.0 10.3 0.4 0.4 5.9 23.7 4.1 2.4 3.0 5.4 2.1 1.6 5.5 2.9 2.2 3.0 16.5 5.5 2.3 3.5 3.1 3.5 2.4 2.9 8.6 46.3 10.7 19.0 15.4 8.8 47.3 9.5 18.2 16.2 8.7 9.0 46.4 52.8 10.3 9.8 18.9 20.2 15.7 8.2 8.5 47.7 14.8 18.9 10.0 8.9 49.8 13.0 19.2 9.2 4.9 35.5 11.8 31.9 15.9 87.3 0.0 1.5 11.2 0.6 12.7 30.2 11.6 24.4 21.2 18.3 33.9 24.6 14.4 8.8 __-------__-------------------- ---------------------------------------------------.---------------- BRAZIL Urban Finance Development 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source:Silvia Paiva, STN/MINIFAZ,Resultsfrom a specialsurvey of governmentagencies. 100.0 47 in the northern region as a whole. Further, it claimed a lion's share (one-fourth) of all the negotiated transfers made to the northeastern state governments in 1988 (see Table 15). Of the regions, the southeast came out a big winner in the negotiated transfers sweepstakes. For example, its state governmentsreceived60% of total negotiatedtransfers in 1988.Note that the sameregion receivedonly 21.48%of total tax transfers. Thus the negotiated transfers substantilalyreverse the equity objectives sought through the tax transfers. A closer look at the shares of individualstates from the negotiated transfers in the southeast presents an interestingpicture. For example,Table 15 showsthat the State of Sao Paulo, the richest state in the region and the nation, received90% of the total transfers made to the southeast region. Here again,as in the caseof Maranhao,it wouldbe interesting to see whether an apparently favourable accessto negotiated transfers availableto Sao Paulo is supported by economic reasons or simplya reflection on its political clout in the federation. From the casualobservationsreported in earlier paragraphs,two important issuesarise:(1) the importance of political factors in the decisionsconcerning negotiated transfers; and (2) the extent to whichthe redistributivecharacterof the taxsharingmechanismsare offset bythe negotiated transfers. The political nature of negotiated transfers is well recognized though not yet fully documented, for reasons descnbed earlier, in Brazil. Equity implicationsof these transfers willbe discussedfurther in a later section on equalization. A brief description of conveniosby major function follows: Health Care: A federal programto f,luancean integrated (unified)and decentralizedsystem of health care named SUDS was initiated in 1987by a PresidentialOrder (Decree Law no. 94657). Its objectives are to coordinate the health policies of the three different levels of government and to establisha unified health policyfor the nation as a whole. The 1988Constitution has reiterated the same objectives (vide article 198). The new Constitution has stressed decentralized deliveryof health care with communitylevel participationemphasizingpreventivecare. In 1987,a large numberof convenioswere negotiated between the Union and States with the followingobjectives: (i) management of INAMPS (NationalInstitute for MedicalAssistanceof the SocialSecurity System)centers; 48 (ii) financingof INAMPS centers; and (iii) servicesto be providedby these centres. Municipalitieshad the option of participatingin the conveniosunder the terms of agreement acceptable to each state. Financing for SUDS primarilycome from the Fundo de Previdencia e AssistenciaSocial (FPAS). FPAS comprisesof socialsecuritycontributions of employersand employeesthrough payroll deductions. The amountsto be transferred isdetermined on a yearlybasisby supplemental agreements (TA:termosaditivos) as established by the convenios. After the signingof TAs , funds are transferred in monthlyinstallmentsto states (see AppendixC for details). The transfer of fundscurrentlydoes not followany formalcriterion. The allocationof funds primarily takes place through the Integrated Programming-BudgetingProject (POI:Projeto de Programacao-orcamentacao Integrado). POI attempts to reconcile needs of various levels of governmentfor health care to the availabilityof funds. Details of the POI processare givenin appendix C. Previous year allocation usuallyserves as a reference point for current allocation. For example,in 1989,the MPAS (the federal Ministryof SocialSecurity)establishedceilingsfor transfers to each state based on a monetary correction of the actual transfers in 1988. Of total spending on health at the state level (from all sources) federal transfers via conveniosamounted to 7% of the total over the period 1981-86.There isa great deal of variationamong states regardingfederal share of health care spending. Alagoasreceived13%of total financingwhereas Ceara received 4.3% of total only. States and municipalities receive 43% of the funds transferred by INAMPS, federal government entities 2.3% and therefore the puL.; sector as a whole receives45.3% of total transfers. Of the remaining funds 11.7% is allocated to private contractors (contratos); 7% to the conveniosnetworkand 26.7%to the participationof the private networkin the HospitalAdmissionsAuthorization Network (AIMs:Autorizacaopara Intemacao Hospitalar). 1988and 1989distributionof fundsby state is presented in Table 20. Regional Development: Not much details are available on programs for regional development. Most of the funds are reported as intergovernmental transfers under the heading 49 Table 20 BUDGETARYDISTRIBUTIONOF SUDS By Federal Unit (in percent) Federal Unit 1988 1989 NORTH Amazonas Para Acre Amapa Rondonia Roraima Tocantins 5.77 1.55 2.23 0.57 0.23 1.14 0.05 - 6.68 1.47 2.72 0.49 0.20 1.00 0.20 0.60 2'..84 1.28 7.66 3.87 2.61 1.82 0.72 3.76 1.14 1.90 28.55 1.45 7.68 4.64 2.97 2.30 0.99 4.60 1.85 2.07 6.77 2.74 0.91 1.03 2.09 6.71 3.00 1.01 1.15 1.55 SOUTHEAST Espirito Santo Minas Gerais Sao Paulo Rio de Janeiro 48.43 1.76 6.60 28.71 11.36 43.42 1.63 7.78 24.44 9.57 SOUTH Parana Rio Grande do Sul Santa Catarina 14.19 6.11 5.39 2.69 14.64 5.88 5.99 2.77 100.00 100.00 NORTHEAST Alagoas Bahia Ceara Maranhao Paraiba Sergipe Pernambuco Piaui Rio Grande do Norte CENTRE-WEST Goias Mato Grosso Mato Grosso do Sul Distrito Federal BRAZIL 50 Investimentos em Regime de Execucao Especial (IREE) in BGU. In 1987,8% of total conditional transfers were for regional development (see Table 18). Education: Typicallyeducation transfers account for nearly one-third of total transfers. Nearly half of these transfers go to the northeast region. Education financing in Brazil comes from earmarked tax revenues (education salarytax:2.5%taxon private sector wagescollected and returned by origin and Finsocial:TheSocial Investment Fund comprisingproceeds from a 0.5% tax on gross receiptsof allbusinesses.All revenuesfrom this are earmarked for socialservices),general revenue fund and project grants through convenios. While not muchdetails are availablefor conveniosfor secondary and post-secondaryeducation,some details are availablefor Convenio Unico - a major instrument for the transfer of federal funding for primary education. This convenio establishes procedures for the preparation, submission and review of states and municipal requests for project financing. Documentation required for this purpose include a work plan organized by major edpicationalgoals, specific objectives under each goal and proposed projects to achieve the specified objectives. This project submissionapproval process is intended to give the federal government strong control over establishmentof expenditure priorities at state and municipallevels. HousingandUrbanDevelopment:During 1984-87,10%offederalnegotiated transferswere made availabieto states (26% of total) and to municipalities(74% of total) to finance various housing and urban development initiatives(see Appendix Table Al). PlanningandAdministration:Conveniosforplanninganddevelopmentaretheresponsibility of SEPLAN. The tunds transferred for planning and administration however need not be spent on planning and administration. The fundsare intended to providefinancialsupport to projects related to the development of basicsocialservicesand for infrastructure. Agriculture: Typically7% of total federal transfers in a given year go towards financing agriculturaldevelopmentprojectsinitiated primarilyby the states and occasionallyby the municipalities. 51 2.3 Intergovernmental Transfers: A Conceptual Perspective * In attempting an economic evaluation of intergovernmental transfers in Brazil, it is instructivefirst to revieweconomic rationale of transfers in a federation. This reviewwill serve as a frameworkfor passingqualitativejudgements on the design of existingtransfers. From the theoryone can predict the general directionof influencea grant mayhave on the recipient's behaviour. Of course, the magnitudeof this influence can oilly be determined by an empirical analysiswhich is beyond the scope of this paper (see Shah, 1980,1989for an econometric modelwhichdistinguishesthe income and price effectsof grants empirically).Some knowledgeof the general direction of grants is useful both in designinggrant programs to meet specific objectives as well as to evaluate the existing structure of transfers. In the followingsections, major types of grant programs are discussedand their rationale is also reviewed. The Brazilian system of transfers is evaluated in terms of its consistencywith stated objectivesand possible reform options are outlined. 2.31 Grant Types 2.311 Non-matching Transfers Non-matching funds may be either selective or general (conditional or non-conditional). Selective non-matchingtransfers offer a given amount of funds without any local matchingrequired provided it is spent for a particular purpose. The effect of this upon a lower level government's (hereafter referred as local government) budget constraint is shownin Figure 1. The post-grant budget line (ACD) isthe pre-grant budget line (AB) shiftedrightby the amount of the transfer (AC). SinceOE (equal to AC) of the assistedactivityis free good from the local perspective,at least OE willbe acquired and perhaps, but not necessarily,mors. Selectivenon-matchinggrants are best suited as a means of subsidizingactivities to which the higher levelgovernment (sayFederal government) assignsa high priority but are givenlow priority by local governments. Such a case would occur if a programgenerated a very high degree of spillovers * lis section is based on McMillan, Shah and Gil!en (1980) 52 Other Public Goods (NCz$) A - - C - 7Med Public Service (NCz$) E GRANT FIGURE 1 : CONDITIONALNON-MATCHING Other Public Goods (NCz$) A B GRANT NON-MATCHING FIGURE 2: UNCONDITIONAL D Assisted (NCz$) Public Service 53 up to some levelof provision (like OE) after whichthe external benefits abruptlyterminate. Although, there is no evidence to suggest that these features characterize state and local provisionof education, health, regional developoment and agriculture in Brazil,yet almost all the funding through convenios is of the selective non-matchingvariety. If the non-matchinggrant is unconditionalor general, then no constraints are put upon how it is to be spent. Thus unlike the previous case no minimumexpenditure in any area is expected. -n this case,the recipient's budget lineis shiftedupwardsand to the right throughoutby the amountof the grant (AC=BD) and the new budget line becomes CD as opposed to AB (see Figure 2). The grant monies can be spent on any combination,publicgoodsor servicesand/or to providetaxrelief to residents. Since general non-matchingassistancedoes not modifyrelative prices and must not be spent for a particular activity,it is the least stimulativeof local expenditures. Typically,local expenditure willonly increase by less than NCz$ 0.50 for each additional NCz$ 1.00of unconditional assistancewith the remaining fundsgoingtowardstaxrelief (i.e.made availableto local residentsto use for private goodsand services). It has been noted that the portion of grants retained for greater local expenditure exceeds local government's own revenue relative to resident's incomes. This is referred to as the "flypapereffect" (Gramlich, 1977); that is grant money tends to stick where it first lands. The implication is that for politicaland bureaucratic reasonsgrants to localgovernmentstend to result in greater localexpenditure than would result if the same transfers were made directlyto local residents. The federal taxtransfersto states and municipalitiesthrough the FPE and FPM areexamples of grantsof thisvariety. There issome evidenceto suggestthat such transfers are inducingmunicipalities to underutilize their own tax bases. Given an amount of available assistance,recipients of grants prefer unconditional nonmatchingtransfers. This is because these grants provide them with the maximumflexbility to pursue their own objectivesas they augment resources without influencingtheir pattern of spending. Hence, the recipient is able to maximizehis ownwelfare. The grantor, however,maybe prepared to sacrifice some of the recipient's satisfactionin order to ensure that the funds are directed towards expenditures on whichhe places a priority. This isparticularlyso when federal objectivesof affordingfiscalassistance to states and municipalitiesis implementedthrough the programso manydifferent departments (e.g. 54 planning, health, education etc.) rather than through the Ministry of Finance which does not have a dominant influence in any specific area. In this case, the different departments do not want their programfunds to be shifted, or even seem to be shifted, by local governmentstowardsexpenditures in other areas. In this situation, conditional (selective)non-matching"block"grants are attractive. Such funding of state-local assistancecan ensure that the funds are spent in a department's area of interest (e.g. health care) yet need not distort local priorities among alternative activitiesor :nduce inefficient allocationsin that expenditure area. This wouldappear to in part be the objectivespursued through the non-matchingSUDS programs, although as presentlystructured, these programs may not be attaining these objectivesin the most cost effective manner. 2.312 SelectiveMatchingTranskrs Selective matching grants are those which require that the funds be spent for a specific purpose (i.e. conditional)and that the recipient matches the fundsto some degree (e.g. 50:50,66.3:33.7 etc.). They are also called cost-sharing programs. Consider a 25 percent subsidy program for transportation. The effect of such a program on localgovernment budget constraint isshown in Figure 3. Without this transportation subsidy,ABindicatesthe combinationsof transportation and other public goodsand servicesa citycan acquire with a budget of OA (equal to OB). Introducing a federal subsidy for transportation amounting to 25 percent of transportation expenditures(or NCz$3.00of local funds for each NCz$ 1.00of grant), the budget line of attainable combinationsbecomesAC. At any level of other goods and services, the communitycan obtain one-third more transportation services than previously. If the communitychose combinationM prior to the grant, it willlikelyselect a combination such as N afterwards. At N, more transportationis acquired. This results from two effects. One due to the subsidythe communityhas more resources (i.e. higher income) and some of those go to acquiring more transportation services.This is the income effect. The income effect is reinforced by the price or substitution effect. Since the subsidy reduces the cost (or price) per unit to the community of transportation services,they acquire more (even if total resources were no greater). Hence, both the 55 Other Public Goods (NCz$) A N~~~ \ 25% subsidy N, Other Public Goods (NCz$) B O C Transportation($ (NCz$) c~~~ AL F Figure 4: CLOSED ENDED MATCHING GRANT B C Education (NCz () 56 price and income effects of the subsidystimulatethe acquisitionof more transportation services. Although the grant is for transportation, more of other public goods and servicesma"'also be acquired, although the relative price of other goodsincrease (ie. they become more expensive)and the price effect works againstthem. If the positiveincomeeffect is sufficientlylarge,it willdominateand a part of the grant's effect willbe greater levelof consumptionof other goodsand services. Most studies tend to find that for grants of this kind, expenditures hi the specified area increase by less than the amount of the grant with remainder shifted towards other public output and tax relief. For example, NCz$ 1.00of grant tends to increase expenditures in the assistedarea by NCz$ 0.80. The other NCz$ 0.20 is diverted to finance other public services(see Shah, 1979). Open ended matching grants are well suited for correcting inefficienciesin public goods provision arising from benefit spillovers. Benefit spillovers , or externalities, occur when services provided and financed by a local govermmentalso benefit membersof other local governmentswithout their contributingto finance their provision. As membersof the providinggovernment perceive all the costs but obtain only a portion of the benefits, theywilltend to under-providethe goods from the social viewpoint. If compensation arranged through negotiation among the affected communities is not feasible, the situation can be corrected by a senior government subsidizingthe provision of the public service. The extent of spillovershould determine the degree of subsidy(matchingratio for grant). While matchinggrants can correct for inefficienciesarising due to spillovers,they do not cope with problemsarisingfrom uneven or inadequate fiscalcapacitiesamongstate-localgovernments. Hence, local governments with ample resources, e.g. Sao Paulo, can afford to meet the matching requirements and so can acquire a substantialamount of assistance. However,other governmentssuch as Para and Ceara with more limitedfiscalcapacities maybe unable to devote as much to match federal fundsand so fail to obtain as muchassistance although their requirements maybe no less and, indeed, their needs mayeven be greater. Other formsof assistanceare needed to equalize fiscalcapacities. Closed ended matchingprovisionsare usuallypreferred by grantors as they then are better able to control their budgets. The effect of closedended matchinggrants on the local budget constraint isshown in Figure 4. ABis the originalbudgetlinewith when no subsidiesare available. However,when assistancefor say education is availableat say 66.7 percent rate up to a limit,the budget line is ACD. 57 Initially,costs are shared on a one-third:two-thirdbasis upto a level of OF at which the subsidylimitof CD'(=CE) is reached. Expenditures for education beyond OF receive no subsidyso the slope of the budget line is 1:1 rather than 1:3 along the subsidizedsegment, AC. Although one would not expect closed ended grants to stimulate expenditures on the subsidizedactivitymore than open-ended grants, empiricalliterature typicallyfinds them to be more stimulating. Gramlich(1977)and Shah (1979,1989)report that the estimatedresponse to an additional $1.00of this kind of grant is typically$1.50. Institutional factorsmayexplainthe rather large response. Gramlichnotes that this type of grant is frequently used to encourage spending in areaswith elastic (i.e. responsive to income and price effects) demands, the grants are large relative to normal spending by recipients in these areas, and/or the granting governments take measures to discourage the reduction of recipients'expenditures on these aided activities. One might reasonably askwhy the extensiveuse of selectiveclosed ended matchinggrants by developedcountries when theyseem ill-designedto solveproblemsand inefficienciesin publicgoods provision.The answerseems to be that correctingfor inefficienciesis not the sole and perhaps not even the primary objective sought with their use. Rather they are employed as a means of assisting local governmentsfinanciallywhile promoting expenditures on activitiesgiven priorityby the grantor. The selective aspectsor conditionson the spending ensures that the funds are directed towardsa desirable activityin the grantor's view. The local matchingor cost sharing component affordsthe grantor a degree of control over and requires a degree of accountabilityby the recipient. Finally,of course, the cost to the granting government is known. The conditional closed end matching grant has certain advantages from the grantor's perspectivebut there are also disadvantages.While they mayresult in a significanttransfer of resources providingfinancialassistance,it maydistort output and cause inefficienciesin doing so. The reason for this is that this aid is often only available for a fewactivitieswhich results in excessivespending in those areaswhileothers are underfinanced. The common argument isthat local priorities are distorted. Also, it is typicalthat capital outlays are subsidizedwhile operating wostsare not. This results in selecting alternativeswhich are too capital intensiveto be least cost. These aspectswillbe discussedlater in the context of the Braziliansystem. A summaryview of the impact of selective transfers is presented in 58 Table 21. This table suggeststhat open ended selectivematching grants are most suitable vehicle in inducingincrease in expenditure on the assisted function by lower level governments. General nonmatchingtransfers, on the other hand, wouldbe preferred if the objectivewassimplyto enb-incewelfare of local residents. 2.32 The Economic Rationale for Intergovernmental Transfers In the preceding sections grants were discussedaccording to the different arrangements under which they are provided, i.e., selective and general non-matching and open and closed ended selective matching. In that discussion, the underlying rationale for grants was introduced but not elaborated upon. In order to re-emphasize this fundamental rationale and its implicationsfor grant structure, the basic reasons for transfers are reiterated and summarized. Gramlich (1977), Boadway (1980,1990),and Shah (1979, 1983,1984)note that there are economicand politicaljustificationsfor grants. Economicjustificationsinclude efficiency(spillovers,common market arguments, differential net fiscalbenefits),equity(fiscalgap, differentialnet fiscalbenefits,redistribution)and to a minordegree stabilization objectives. An interesting aspect of the theory of grants is that efficiencyand equity objectivesare often complementary. Boadway(1980,1990)suggeststhat application of efficiencyand equity principlesresults in four main economicreasons for grants. These are discussedbelow: (i) Inter-jurisdictionalSpillovers:Intergovernmentaltransfers can be used to increase the efficiencywith which public goods and servicesare provided. Their major contribution is to correct for inefficienciesarising in the presence of interjurisdictionalspillovers. Spilloversusuallyoccur because the benefits of a locallyprovidedgood or service itself spills beyond the local jurisdiction to benefit those not contributing to the costs (e.g., benefits from control of air and water poliution, and locallyeducated students who relocate) and because non-residents come to the localityand enjoy the public services provided(e.g. parks, cultural,recreational and transportation facilities,state universities, state welfare and health care system). In the presence of spillovers, state-local governmentsconsiderownbenefits onlyand under-providepublicservices. While other Table21 GRANTS IHPACT-OFcoITmoNAL THE CONCEPTUAL Typeof Grant IncomeEffect Effect U *; _ U Si Function Rankby Objective 8A/BG Price(substitution)TotalEffect _ _ U A S5ncroasein _ _ _ _ _ _ _ _ _ Expenditure _ _ _ _ _ Welfare MATCHING A. SELECTIVE i Open-_nded tt tU >1 1 8 I if U1 2 2 or 3 4 t t t •i 3 2 t t t t t tt t t t 1 1 t t - - ii. Close-ended binding constraint constraint t non-binding 8. NON-MATCHINC SELECTIVE C. ENERALNON-MATCHING Note t UV t t - - - t t - - - subfunction = Assisted A = AssistedFunction U = Unassistedfunctions (services) 0 a Grant Source: AnwarShah (1985) - t t t t t 2 1 (1 3 1 60 approaches such as redrawingofjurisdictionalboundariesand/orseparatejurisdiction for each servicemayalso be used (McMillan,1975),intergovernmentaltransfers are a major and often the most practicalmeans of alleviatingthe inefficienciesarisingfrom spillovers of this kind. Open ended conditionalmatchinggrantswhichmodifyrelativepricesare the kind of transfers that is most appropriate for implementingthese corrections. The extent of cost sharing by the higher level of government should be consistentwith the degree of spiliover. Benefit spilloversis not consideredto be a serious problem in BraziL (ii) FiscalGap: A mismatchbetween revenue means and expenditureneeds at various levels creates a structural imbalance resulting in revenue shortfall usuallyfor a lower level government. This imbalanceischaracterized as a fiscalgap. Four often cited reasons for this imbalance are:(a) inappropriate expenditure and tax assignment;(b)limited and or unproductive tax bases availableto a lower levelsof government so that tax rates would have to be inefficientlyhigh (e.g.in Mexicoand Pakistan);(c):regionaltaxcompetitionie. state and localgovernmentsfearfulof loosingcapital,labour and businessactivityto other jurisdictions do not fullyexploit businesstax potentials and thus provide lower levelsof public services ; and (d): federal government crowds out tax room for state and local public sector. To correct for problems associatedwith (a) and (b) abovejoint occupancy of some tax fieldsor decentralizationof some taxesare advocated. Alternately unconditionalgrants and/or revenue sharingbased on originprincipleare also appropriatt solutionsto dealwith this problem. Higher revenue effort bythe federal governmentand unconditionalgrants are required to dealwith taxcompetition issuesdiscussedin (c) above. Finally,to deal with (d) above some form of tax abatement by the federal government is necessaryto provide more tax room for lower levels in fieldsjointly occupied by the two or three levels of government. (iii) MinimumStandards of Services:A third justification for intergovernmentaJtransfers is based on an efficiency-cum-equitycase for ensuring commonminimumstandards across jurisdictionsina federation. Forcertainservices,expenditureassignmenttostateandlocal government may be based on efficiency of public service provision as well as responsivenessto localneeds and concerns.This mayconflictwiththe nationalequityand efficiencyobjectives(see Boadway,1980,1990). Musgrave (1976) has argued that the redistributive role of the public sector is best performed by the federal government. Mobilityof factorswithin a federation severely limitsthe redistributiverole of local governments. The New York city providesa prime examplein this case. Redistributivepolicies adopted by the city in the 1970screated a major fiscalcrisis and the federal governmenthad to intervene and reverse these policies to restore the financial health of the city. Hence, the case for the assignment of redistributive function to the federal government is quite strong on theoretical grants. Several of the public servicesassigned to state and local governments on efficiencyor accountabilitygroundsare strongly redistributivein character. Social insurance, health, education and welfare are prime examplesof such services. Consider the case of health and education services. These are quasi-privategoods and strictly from a technological efficiencypoint of viewwouldbe best providedby the private sector. Indeed in the U.S. health care is byand large treated as a private good. The WorldBank, in recent years, has marshalleda great deal of resourcesto advocateprivate provisionof health and education servicesin developingcountries based on this view of economicefficiency. Needless to argue that such a viewpointcompletely ignores information asymmetriessuch as moral hazard and adverse selection associatedwith private provision of such services. Fiscal federalismliterature has argued that informationalinefficienciesalone do not provide a convincingcase for the public provision of health care and education. This literature points out that most governmentstreat health care as a fundamentalpublicresponsibility and striveto provide these serviceson a uniformbasis. This is because these servicesare viewed as "redistributionsin-kind". Thus a case for public provision of these services primarilyrests on equity objectives. Consider the case of health services. Incidence of disease is directly correlated with the incidence of poverty and by corollary inversely associated with economicwell-being. Thus public finance and provision of health care 62 enhances the redistnbutive role of the public sector. Similarlypublic provision of education,byimprovingaccessof the poor, servesto further equalityofopportunitygoals. Relative importance of expenditures on health, education and social services further suggeststhat the redistributioneffected by the tax systemor direct cash transfers pale in. comparisonwith the in-kindredistnbution made possible by these public services. In a federal system,lower level provisionof such serviceswhiledesirable from efficiency, preference matching and accountabilityperspectives, create certain difficulties in the fulfillmentof federal equity objectives. Factor mobilityand tax competition influences createstrongincentives forlower levelgovernmentstounder-providesuchservicesaswell as restrict accessof such servicesto those most in need such as the poor or the old in view oftheirgreatersusceptibilityto diseaseand therebyposingpotentiallygreaterrisksforcost curtailment. These perverse incentivescan be alleviatedby conditional selective nonmatching grants from the federal govemment. Such grants would not effect local govemments incentives for cost efficiency while ensuring compliance with federally specified standards for accessand level of services. A second justificationfor commonminimumstandards for publicservicesin a federation is basedon economicefficiencyconsiderations. Commonminimumstandardswouldhelp reduce interregional barriers to factor and goods mobilityand thereby contribute to efficiency gains. Establishment of some minimum standards of social services will encourage labour mobilityand of infrastructure capital willenhance factors and goods mobility. Boadwayhas emphasizedthat harmonization of expenditures would improve gains from interregional trade and help foster a common intemal market. Commonminimumstandardsfor publicservicesacrossdifferent statescan be encouraged through conditional non-matchingor conditional closed ended matchingprograms. A conditional non-matching program is to be preferred due to its non-obtrusive nature. State governmentswould be free to spend grant monies as they chose providedcertain minimumstandardsofservice and accessare met. The higher levelgovernmentwillsimply 63 monitor compliancewith these standards. Note that conditional non-matchinggrants in pursuit of common minimumstandards serve both efficiencyand equity objectives. (iv) Differential Net FiscalBenefits Across States Differential uietfiscalbenefits acrossvarious states arise due to a number of reasons: a. Some states are better endowed in natural resources and therefore have better access to an enlarged revenue base than others; b. Some states or localitieshave relativelyhigher incomes and therefore greater ability to raise revenues from existingbases; and c. Some states or localities have inherited higher cost disabilityfactors such as lacking the mininmumthreshold for scale economies or difficultterrain factors etc. or higher need factorssuch as greater proportion of young,old and the poor. The presence of differential net fiscalbenefits encourages fiscallyinduced migration. Labour and capital may move to areas with positive net fiscalbenefits based on fiscal considerations alone. In the process, some negative externalities imposed on the jurisdictionsthey leave and the jurisdictions they enter are ignored. A fisherman from Para maymigrate to Minas Gerais, although he may not have any gainful employment opportunity there. Overall result of such fiscallyinduced migrationwould be that too many of the factors will move to resource rich areas creating in its wake social and economicproblems and thereby serious inefficienciesand inequities. Inefficiencyarises as factor movement takes place in response to fiscal considerations alone. Inequity is caused by identicalpersons in various states beingtreated differentlybythe publicsector as a whole. National welfare is reduced by the externalities imposed by the fiscally induced migration. Fiscalequalizationgrants to eliminate/reducedifferential net fiscalbenefits acrossstates can enhance both the efficiencyand equity of a federal system. An ideal form of such transfers would be an interstate revenue pool which providesboth negativeand positive equalizationgrantsto memberstatessuch that net transfers equal zero. Thus the program by design willbe self financing. Such a grant systemmust be unconditionaland must not 64 reward strategic behaviour to enhance positivegrant entitlement or minimizenegative transfer by member states. Thus grant design must incorporate factorsover which states have little control. Three grant programsthat have endured and are broadlyconsistent with the aboveguidelinesare the West German, Canadian and the Australiansystemsof equalization transfers. The West German system is a fraternal system of equalization among the German states. The federal government simply acts as an observer and occasionallyasa mediator. The Canadianand Australiansystemare not self financingand instead are federal programs. The Canadian system attempts to augment the fiscal capacityof member provincesup to the nationalaverage capacity. The systemmeasures the fiscalcapacityof a state by the revenue that could be raised in that state if the state goven lent employed all of the standardirevenue sources at the nationwide average intensityof use. The Australiansystemworries aboutexpenditure needs as well(see Shah 1983,1984). Issuesin the designof an equalizationprogramare discussedin an appendix to this chapter. A further but infrequently mentioned objective of these transfers is to advance stabilizationpoliciesof the federal government as discussedbelow: (v) Stabilization Objectives: Intergovernmental transfers can also be turned to assist in achievingeconomic stabilizationobjectives. Grants could increase in periods of slack economicactivityto encourage localexpenditure and diminishduring the upswingof the economiccycle.Capitalgrants wouldbe a suitable instrumentfor this purpose. Care must be exercisedinensuring that fundsare availablefor operatingexpendituresassociatedwith such initiatives. 2.4 Intergovernmental Transfers in Brazil: An EconomicEvaluation The conceptual basisof intergo' r; mental grants enunciated in earlier section providesus with a framework to evaluate the existing s;. ire of intergovernmental transfers in Brazil. The followingsectionspresent observatir ns on the taxsharing and transfer programs in Brazil. 65 2.41 Tax SharingProgram:A CloserExamination As discussedearlier revenue sharing in Braziltakesplace through twomajor funds,one each for states (FPE) and municipalities(FPM) and for certain taxesdistributionof fundsbyoriginaccording to pre-specifiedshares. The followingdiscussiontakes a closer look at the two revenue sharing pools mentioneG 'hove. 2.411 States ParticipationFund:A CriticalLook The taxsharing programfor states (FPE) in place in Brazilismodelledafter the nowdefunct U.S. Revenue SharingProgram of the 1970s.This programrepresents a significantimprovementover the U.S. program but manyshortcomingsremain. The program is transparent and enjoys popular political support. It pays due regard to political autonomy of recipient units and places no special restrictions on the use of funds. It offers predictabilityof grant shares of individualstates and thereby encourages federal units in taking a long term view of their expenditure priorities. The program is redistributive in nature. Grant funds are intended to vary directlywith fiscalneed (population size) and inverselywith fiscalcapacity(per capita income). Incorporation of population and per capita income havebeen done with a finesse. The values of these factors havebeen restricted to a range to avoidundue infuence of abnormal values of one or more factors on a state's share. Furthermore such a range specificationimplicitlyrecognizes a basic minimumgrant level for each state ( per capita NCz$ 1.70in 1988). State Participation Fund (FPE) as currently structured is beset with manydifficulties.It incorporates per capita income as a measure of fiscalcapacity. Per capita incomeis an imperfect guide to the abilityof a state government to raise taxes as it is conceivablethat a significantproportion of income mayaccrue to non-resident ownersof factorsof production. Even if these conceptual problems are ignored, significanterrors in estimatingstate gross domestic product in Brazil are encountered to make the measure unacceptable for an equalization program. Moreover, estimates of state per capita income are quite dated (most recent estimatesare for 1980)and therefore of little relevance for use in 66 a programof fundamental importance in federal-state fiscalrelations such as the FPE. All factors are combined in the formula in a multiplicativemanner and therefore an outlier performance on account of one factor can significantlyinfluence the state entitlement from the FPE. The FPE attempts to achieve a number of diverse and somewhat conflicting objectives such as revenue sharing and equalizationin a singlemeasureand thereforeexpectedlysignificantlyfallsshorton individualobjectives. Consider the equityobjectives. While the formulais indeed redistributivein nature in its overall impact, certain inequitiesin individualstate shares arisefrom its application. For examplethe States oftAcreand Para both had the same 1988per capita income of NCz$266but Para qualified for NCz$9.00in tax transfers asopposed to NCz$49forAcre. Roraima,on the other hand, had onlyslightlyhigherper capita income of NCz$286 but received NCz$123in per capita tax transfers (see Table 9). While federal transfers do promote regional equity, the standard they strive to achieve remains to be specified. Per capitarevenues inclusiveof transfers in the northeast and the north in 1988only reached roughly 2/3rd of the national average. Finally,while there is no visible opposition to the formula for the FPE in politicalor even academic circles in Brazil, States' Council in recent years has rejected its results and instead have decided to allocate the revenue pool on the basis of a political compromisewhich makes a furtherdownwardadjustmentin the participationcoefficients(shares) of richer states. It is remarkable that such a compromisecould be struck. 2.412 MunICipalParticipation Fund: A Review MunicipalParticipationFund(FPM)is auniqueprogramoffederal-municipalunconditional transfers. BrazilianInstitute of MunicipalAdministration(IBAM) has labelled it as "the most generous tax (transfer) system of municipal revenues among all developing count:ies" (IBAM Press Release No.96,September 1989). IBAM notes that federal tax transfer to municipalitiesthrough this fund in Brazilfar exceedsthe total budget of similarsizedmunicipalitiesin other Latin American countries. For a large number of municipalities,FPM represents more than two-thirdof their gross revenues from all sources. 67 The program is transparent and thoughtfullydesignedin certain respects. For example,the formula recognizes population size an important determinant of need and also recognizes certain thresholds in population size. The distributionof fundsis primarilybased on population size and state per capita income. State capitalsand larger municipalitiesbecause of their differential fiscalcapacities and needs are treated separately from smaller and mediumsized municipalities. The FPM formula neverthelesshas significantdrawbacks. State per capita income is used as a proxyfor municipalfiscalcapacityin the distributionformula. This causes a number of difficulties. First, there is no municipalincometaxin Brazil (and none iscalled for due to capital and labour mobility) and therefore per capita income is a poor indicator of a local government's abilityto raise revenues. Second,state per capitaincome is also a poor indexof municipalptr capita income as per capita income variesa great deal withinmost Brazilianstates. In short, the formula failsto reflectupon the differential fiscalcapacityof Brazilianmunicipalitiesin a meaningfulway and as a result does not distribute federal funds among Brazilianmunicipalitiesin a fair and equitable manner. For example,in 1988,per capita transfers to n unicipalitiesin Minas Gerais, a rich state, was NCz$6.39as opposed to NCz$5.07in Para, a have-not state. The so-calledgenerosityof federal funding for municipalfunctionscreates some problems as well. For example, there is some evidence to suggest that generous availabilityof federal fundshas resulted in some municipalitiesunder-utilizingand in some instancesabandoningowntax bases e.g. the property tax. Revenues from the urban property tax (IPTU) have been observed to decline in recent years. Almostall municipalitiesare payinginadequate attention to user charges. Thus the overalleffect of the current arrangementshavebeen to discourageown fiscaleffort by municipalitiesas their revenue gains have outstripped any concomitant increase in their responsibilities. Further, municipal accountabilityhas been significantlyreduced as local politicianshave no great need to convincevoters about the desirabilityof local spending. 68 2.42 Negotiated Transfers: An Examination It was earlier noted that most negotiated transfers in Brazil are of selective non-matching project grants. Such grants are useful when the expenditure priorities of the grantor and the recipient differ substantially. Project review and approval process is an attempt to safeguard the grantor's objectives. Most of the time such an attempt failsas the processcreates incentivesfor recipients to put forwardtheir best and of mutual interest projects which they would havefinanced any way and use the fundsto financeother servicesof interest only to the recipient. The ingenuityof bureaucrats to convert grant monies into fungibleresources can hardlybe understated. The federal government in Brazil, in the past, has attempted to use project review and approval process to influencestate and local priorities to secure federal objectives. SUDS represents an example of such an attempt. In more recent years, these transfers have served not as a means of safeguardingany federal objectivesbut more as a vehicle for pork-barrel politics and therefore, grant programs have multipliedand for manyof these programs,program objectivesare either not specified or specifiedvaguely. Afonso (1988,p.15) notes that examplesabound when fundswere made available prior to the submissionof project proposal. Further, in manyinstances,federal fundsare made available to finance purelylocal functions. In 1988,there were 117federal-state-localmajor programswith over 5000 sub-programs(convenios). Each of the state has several thousandsof its own convenios. Sheer numbers of these programs discourageattempts at an analysis. Further, all three levelsof government do not maintainany central inventoriesof such programs. There are some obviousreasons for this state of affairs. Various governmentscan exercisecomplete discretion over these fundswithout havingany accountability. Enhanced flexibilityis being achieved at the cost of transparency, objectivity and accountability. Of a large numberof conveniosmentioned earlier,the major conveniosfor financinghealth care and education deserve some comments. Integrated and Decentralized Health Care System (SUDS): SUDS represents one of the better thought of convenios. It does not follow any formal criteria for allocation of funds to state and local and private sector entities. Allocationto governmental units is based on a history of past health expenditures and any agreedupon capital projects. Formalprocesseshavebeen establishedfor requests 69 and deliveryof fundsto state and localgovernments.This processseeksto balanceexpendituredemands with availableresources. The existingdesign of SUDS is subject to a number of cufrent and potential difficulties. These center on: the availabilityof funds;degree of federal involvement;differentialtreatment of public and private providersof health care. Availabilityand Predictabilityof funds:SUDSis financedbysocialsecuritycontnbutions and therefore competes with other socialsecurityrelated expenditures for financing. The 1988Constitution has expanded the range of social security benefits available to employees. There has been no correspondingincreases in socialsecurity contributions. Further, new tax sharing arrangements have limitedthe revenues at the disposalof the federalgovernment. Thesefiscalstrains are likelyto constrain the federal funds availablefor the SUDS. Degree of Federal Involvement:The 1988 Constitution has defined health care a shared responsibilitybetween different levelsof government. While the constitutional assignmentis open to conflictinginterpretations by various levels,the intent is fairlyclear. The federal government has the primaryresponsibilityin setting norms and distribution of federal funds; the states are responsiblefor deliveryof health care under the norms specifiedby the federal goverment. Municipalitiesare seen as executingagentson behalf of the state governments. In practice,federal governmentis involvedboth in defininghealth policybut also the deliveryof health care. In 1988,only 41.6% of total expenditures of SUDS were directed bystate and local govemments. Even for the transfers made availableto state and local govemments, for a large proportion of these funds, effectivecontrol remained at the centre in viewof project approvalprocess. This centralizationof authorityis clearlyincompatibleboth with the stated objectivesof SUDS as wellas with the constitutional assignmentof health care responsibilities. Based on the economic principles enunciated earlier also, the federal involvementin health care is justifiedon account of the redistributivenature of this publicservice. Thiswould call for a more passive role by the federal government than is currently exercised by the INAMPS (National Institute for MedicalAssistance and SocialSecurityof the Federal Ministryof SocialSecurity). Differential Treatment of Public and Private Sectors:High rates of inflation and time lags in the deliveryof transfers create certain anomaliesin the treatment of public vs private providers of 70 health care by the INAMPS. Current time lag for the transferof fundsupon approval from the federal treasury to state and local governmentsis about 70 days(see Afonso 1989)whereas the samelag for the private sector is about 45 days. Ihis lag results in a lossof nearly half the value of expected transfers in real terms for the state and local public sector. The loss to the private providers is substantiallyless. Further, in 1989,private providers were entitled to an upwardadjustmentbased on the consumer price index of their entitlements. Similartreatment to the public sector was denied. In conclusion,SUDS represents one of the better conditional transfer programs currently in place yet substantial further improvementsin its design are possible to further its objectives. Education:Asdescribedearlier,throughtheConvenioUnico,federalgovernmentascertains the needs of state and localgovernmentsfor the finance of primaryeducation in Brazil as well as makes in-kindtransfers primarilyin the form of school lunchesand textbooks. Through this convenio, federal government attempts to reshape educational spending priorities of state and local governments. The New Constitution sees mnicly an advisoryfunction for the federal government in the field of primary education. lTus there is no longer anyjustification for this convenio. Further, primaryeducation is primarilya local functionwith very little if any spilloversbeyond the boundaries of area providingthe service. Thus it isbest to givelocal governmentscompleteautonomyin the provisionand financeof such a service. In summary,whilesome of the individualprograms have a number of desirable features, overali the systemof conditionaltransfers in Brazilis in need of majorreform. These options are taken up in the next section. 2.5 ReformOptions This section outlines reform options based on some fundamental flaws of the existing structure identified in earlier sections. Altemative strategies for the reformof the revenue sharing and conditionaltransfer programs are outlined below. 71 2.51 Revenue Sharing Existing tax assignment and revenue sharing system severely constrains the federal government while at the same time it givesmunicipalitiesalmost a free ride. The federal govermment nowcommandsonly 31.5%of final dispisition of revenues as opposed to 45.5% for state governments and 23.5% for municipalities(see Table 7). Policyemphasison trade liberalization,export promotion and international competitivenessfurther limits federal government choices in exploitingits exclusive bases. Tbis suggeststhat the federal governmrentshould withdrawfrom involvementin local functions. Further in subjectsjointly shared with state governmentsit must stay in policydevelopment arena and program development and administration be left to the states themselves. Specific comments on individualprograms follow: States Participation Fund (FPE): It is suggested that the FPE may simplybe used as a mechanismfor distnbution of revenuesbyoriginonlyand equalizationand other objectivesbe addressed through separate and distinct programs designedfor those purposesonly. If this view is accepted, then the [CMScouldbe administeredby the federal gover-ment on behalf of states and revenues returned byorigin. Similarly,revenues currently in the FPE pool could be returned to states in proportion of the revenues raised from their territories. MunicipalParticipationFund (FPM): Direct federal transfersto municipalitiesdo not satisfy any of the economic criteria mentioned earlier and on the other hand provide dis-incentivesfor municipalitiesto exploitownrevenue sources. Thus it issuggestedthat the FPM be disbanded. Instead, states be encouraged to strengthen their unconditionaltransfers programto municipalities. States are in a better position than the federal government to monitor the fiscalpositions of their municipalities and provide appropriate assistance. With the rural property tax (ITR) being returned to states and higher levels of transfers from the federal level, states would be in a better position to provide this assistancethan currently. States transfers to municipallevelscouldbe based on a formula similarto the FPM but which incorporates per capita municipalfiscal capacity as an important factor and further incorporates a basic per capita grant into the formula. 72 2.52 NegotiatedTransfers The system of negotiated transfers as it stands today needs to be completely revamped. These may be replaced by a system of selective (specific purpose) non-matching (block) per capita grants. Such grants will not interfere with the fiscal autonomy of state governments, willencourage innovation by individualgovernments and will help bridge the fiscal gap. Federal objectives will be fulfilledby making these grants conditional on achievingcertain objectivelyverifiable criteria relating to minimumstandards and access. Such grants wouldbe particularlysuitable for health and education financing.Through these grants the federalgovernment couldinducestates to providecertain minimum standards of serviceswhilegivingspecial accessto the needy but poorer segmentsof the populations. Since there is no discernible differences in federal vs state priorities in these areas, project and/or matchinggrantswouldsimplycreateadministrativeinefficiencieswithouthavinganybehaviouralimpacts in the desireddirection. Federal direct participationin municipalprojectsmaybe restrictedto technical and financialassistancefor major projects such as masstransit systemsin large cities. 2.53 Equalization Program Finally, to deal with interstate differentials in fiscal capacity,an over-arching program of equalization transfers be instituted to be administered by the Federal government on behalf of the Councilof States. The programwouldbe financedby applyingnegativeand positiveentitlements based on a formula to the monies availablefor distnbution through the newly constituted FPE Thus final dispositionof revenues to each state will be based on revenues available to that state based on the derivationprincipleplusa positiveor a negativeentitlement based on an equalization adjustment. One possiblestandard of equalization for this purpose could be the per capita national average yield from sourcesto which state and localgovernmentshave direct or indirect access. More specifically,consider the arithmeticmean of allstates as the standardof equalizationand that all state and local taxbases are to included in the calculations. These include: Shared Taxes (partial inclusionbased on percentage share of State-Municipalsector): Income tax 73 Payrolltax Taxon industrialproducts Hydroelectricity tax MineralProductsTax StateTaxes: GeneralValue-addedTax (ICMS) Inheritanceandgifttaxes Motor VehicleRegistrationTax (IPVA) SupplementaryCapitalGainsTax MunicipalTaxes: Servicestax(ISS) Urban PropertyTax(IPTU) FuelsTax(IVVCLG) Propertytransfers(ITBI) FrontageTax Notethat thedataon taxbasesforsharedtaxesandstatetaxesiscurrentlymaintainedbythe federal and state governmentsbut hardlyever analyzedfor effective utilization. The proposed equalizationprogramwillencourageuse of suchdatain taxpolicyanalyses.The dataon munnicipal tax baseswouldinitiallycreate problemsof comparisonand thereforewouldrequireadjustmentfactorsto be appliedbasedon samplevalues- a traditionfollowedin Canada.An attractivestrategyinthisregard maybe to attemptequalizationin the firstfiveyearsof theprogrambasedon statelevelrevenuesources onlyand later phase-inmunicipalrevenuebasesas andwhencomparabledataon municipaltaxbases becomesavailable. Givena decisionon the standardandthetaxbasesto be includedisconsistentwiththe above discussion,an equalizationentitlementfor a state sayx for revenuesourcei couldbe determinedas follows: 74 Ez : (POP)X L (PCTB)na x ti - {(PCTB)x x where x t na } E = Equalizationentitlement of state x from revenue source i. POP = Population PCT'B = Per capita tax base of revenue source i t = National average tax rate of revenue source i. subscript na = national average Subscript x = state x. The equalizationentitlement for a state from a particular revenue source couldbe negative, positiveor zero. These figuresthen wouldhaveto be summedup for all revenue sources consideredfor equalization and the overall sum wouldindicate whether a state would receive a positiveor a negative entitlement from the interstate revenue sharing pool. Note that the application of agreed upon equalizationstandard to existingtax bases wilibe determiningboth the total level of funding available for equalizationpurposes as welias state entitlements. Thus the proposed formula makesequalization explicitand transparent in an objective manner. It will foster a greater sense of participation in the federation by member units especiallyhave-not states. Recapitulating policyreform proposed above have three elements: namelytax sharing arrangements based on origin principle,conditional per capita block federal-state grants with federallyspecifiedminimumstandards as the condition for such grants and an equalization program to raise the fiscalcapacityof have-not states to a certain specifiedstandard. 75 2.6 State-Municipal Transfers In Brazil Thefollowingsectionsreviewtaxsharingandstate-municipal conditionaltransfersin BraziL A broad overviewof these arrangements for the country as a whole is presented while drawingupon specificexamplesfromthe Statesof Para andParana. 2.61 State-Municipal RevenueSharing EarliersectionsofthispaperelaboratedupontheimportanceofFederaltransfersasasource ofmunicipalrevenueanddrewimplications ofthisfor municipalfiscalmanagement.Asecondimportant sourceof municipalrevenuesin Brazilisthe constitutionally mandatedstate-municipal revenuesharing arrangements.Thesetransfersconstituteone-thirdof municipalrevenuesin Brazil(31.6%in 1987,see Table22). It isremarkablethat the municipalities of somestatesin Brazilraiseas littleas 2% of total revenuesfrom own sources(see 1987figuresfor Maranhaoin Table 23). Mechanismsfor these arrangementshavebeen specifiedinthe regulationsdraftedbythe FederalParliament The regulations providespecificsof the formulaaswellas timingfor the releaseof funds. Mostrecent regulationsas givenin Projeto de lei Complementarno.177(1989)are presentedin Box3. This lawspecifiesthat municipalsharesof federalandstate transfersshouldbe immediately depositedin the joint accountof all municipalities.Further,individualmunicipalaccountsshouldbe creditedno later than the second workingdayof eachweek for allrevenuesreceivedin the previousweel. Box3 showsthat a highlytransparentsystemof transfershas been institutedby Federalregulations. Distributionof taxtransfersforthe mostpartfollowstheoriginprinciple.ICMSrevenuesaredistnbuted by a formulawhichmandatesthat at least 75 percentof suchrevenuesto municipalgovernmentsbe allocatedbyvalue-added.SinceICMSisa value-addedtypetax,thisclearlyrecognizesthe originas the guidingprinciple in the distribution of these transfers. Followingthis principle, wide divergence of municipaltransfers in per capita terms by state is obtained as shown in Tables 22 & 23. A smallweight is given in the formula to other factors which the individual states may consider important in the distribution of these monies in their jurisdictions. Table22 RECEIVEDBY BRAZILIANMUNICIPALITIES 1987 - STATETAX TRANSFERS (Values in Current NCzSThousands) Federal Unit - CURRENT STATETAX TRANSFERS value per capita X of total - CAPITAL STATETAX TRANSFERS valu, per capita X of total 0.23X 601 31 785 41 1,792 0.28 0.15 0.o3 0.28 0.17 0.18 0.22 0.5a 0.03o 0.6sx 0.03X 1.49X 0 0 59s 0 0 7 602 Maranhao Piaui Ceara' R GrandeNorte Paraiba Pernambuco Alagoas Sergipe Bahia NORTHEAST 802 414 1,776 547 709 8,176 948 522 5,119 13,660 0.06 0.16 0.29 0.25 0.23 0.45 0.41 0.39 0.47 0.33 0.25X 0.34X 1.47X 0.46X 0.69X 2.63X 0.7sx 0.43x 4.29x 11.24X MinasCerals EspiritoSanto Rio de Janeiro Sao Paulo SOUTHEAST 11,935 2,001 11,988 60,489 76,413 0.79 0.84 0.90 1.63 1.24 9.9ox 1.6ex 9.94X 41.87X 63.38X 37 702 1,129 Parana SantaCatarina Sul R Crande SOUTH 7,749 4,887 9,874 22,480 0.91 1.14 1.13 1.04 6.43X 4.01! 8.19x 18.82X U GrossoSul Mato Crosao Goias 1,882 1,513 2,977 1.12 0.96 0.64 DistritoFederal CENTERWEST 6,372 120,697 277 Rondonia Acre Amazonas Roraima Para Amapa NORTH 67 BRAZIL Source: UINIFAZ/SEF O0.6x STATETAX TRANSFERS, - TOTAL X OF WMNICI. RWSS STATETAX TRANSFERS REVENUES value per capita X of total 277 67 1,196 31 786 48 2,394 0.28 0.15 0.65 0.28 0.17 0.21 0.29 0.23x 0.06x 0.97x 0.0a! 0.64X 0.04! 1.95! 18.26X 6.26X 30.20X 8.49! 13.92X 9.27X 18.28X 0.06 0.17 0.29 0.26 0.23 0.46 0.41 0.39 0.47 0.34 0.28x 0.34X 1.46X 0.46X 0.658 2.60X 0.77X 0.43x 4.26X 11.14X 6.64X 10.13X 17.13X 12.53X 14.25X 26.28x 23.14! 14.5oX 28.64x 19.11x 0.32 27.39X 0.03 0.07 0.32X 27.72x 12 2 20 21 1 16 0 0 s0 122 0.00 0.00 0.00 0.01 0.00 0.00 0.66x 0.09X 0.92X 0.97X 0.05x 0.74x 0.00 0.00 0.00 0.00! 2.30X 5.62X 314 416 1,796 588 710 8,191 948 522 5,219 13,682 376 0.02 0.00 0.02 0.02 17.27X o.69x 1.70X 32.32X 51.98x 12,310 2,016 12,026 51,191 77,542 0.82 0.85 0.91 1.65 1.26 10.03X 1.64X 9.79X 41.70X 63.1ex 33.09x 32.42X 30.08x 37.89! 35.49X 167 23 26 216 0.02 0.01 0.00 0.01 7.69X 1.06x 1.20X 9.94X 7,916 4,860 9,900 22,676 0.93 1.16 1.13 1.05 6.46X 3.9ex 8.06x 18.47X 31.77X 36.30x 41.88X 36.61x 1.56% 1.25X 2.47X 2 101 0 0.00 0.06 0.00 0.09X 4.65X o.ooX 1,884 1,614 2,977 1.13 1.02 0.64 1.53X 1.31X 2.42X 33.93x 24.9!x 26.45X 0.67 6.28x 103 0.01 4.74X 6,475 0.68 5.27X 27.82x 0.85 100.OOX 2,172 0.02 100.0OX 122,769 0.87 100.OOX 81.61x 1S O.Oi 1987 ^ Table 23 OWNREVENUES ANDTOTALEXPENDITURES OF BRAZIUN WIM NCI-ALITIES (Values In Current NCz8Thousands) Federal Unit GROSS REVENUES Voluo per capita X of tOtsl TOTALFED. ANDSTATETRANSFERS value per capita X of total value 034 REVENlIES per capita X of total TOTALEXPENDITLURES Volue per capita X of total Rondonia Acre AMaSonS Roraima Para Amps NORTH 1,704 910 3,960 86 6,689 Si8 18,096 1.74 2.38 2.15 .83 1.23 2.23 1.61 0. 44 0.23x 1.02X 0.09X 1.453 0.13X 8.373 1,847 826 8,348 821 4,40a 428 10,672 1.87 2.14 1.82 2.93 0.96 1.84 1.31 0.83 0.33x 1.323 0.183 1.74X 0.17X 4.21X 357 8s 612 44 1,286 90 2,424 0.as 0.22 0.33 0.40 0.27 0.39 0.80 0.26X o.o0s 0.45X 0.083 0.92x 0.07X 1.79% 1,765 899 4,300 400 5,836 656 18,383 1.80 2.83 2.33 8.65 1.17 2.82 1.64 0.44X 0.22X 1.07X 0.10X 1.8a3 0.163 8.32x Maranhao Piaui Coors, R Grands Nor Paraiba 1.14 1.63 1.71 2.07 1.61 1.74 1.78 2.68 1.99 1.77 1.433 1.061 2.70X 1.17X 1.28x 3.183 Alagoas Sergipe Bahia NORTHEAST 5,565 4,107 10,486 4,533 4,984 12,144 4,089 8,601 22,073 71,582 0.933 5.68x 18.43X 5,429 3,754 7,615 3,528 4,879 9,218 3,278 2,773 14,944 54,916 1.12 1.49 1.24 1.81 1.4). 1.32 1.42 2.06 1.35 1.86 2.143 1.48X 8.013 1.393 1.73 3.643 1.293 1.09X 5.90X 21.68X 136 853 2,871 1,007 606 2,926 811 828 7,129 1e,ee6 0.03 0.14 0.47 0.46 0.20 0.42 0.36 0.62 0.64 0.41 0.103 0.263 2.18% 0.75X 0.46X 2.17X 0.60X 0.61x 6.28x 12.843 5,135 3,940 10,623 4,168 4,795 11,760 3,827 8,484 21,464 69,194 1.06 1.56 1.74 1.90 1.54 1.68 1.66 2.59 1.94 1.71 1.273 0.9ex 2.638 1.033 1.19X 2.913 0.953 0.863 6.32% 17.153 Minas Gerais Espirito San Rio de Janel Sao Paulo SOUTHEAST 87,202 6,219 89,973 135,090 218,484 2.46 2.61 3.01 4.37 3.54 9.583 1.60% 10.29X 34.78x S6.25X 28,299 4,147 16,696 74,207 123,849 1.87 1.74 1.26 2.40 2.00 11.17X 1.64X 6.593 29.29x 48.69X 8,903 2,072 23,277 e0,883 96,136 0.69 0.87 1.75 1.97 1.54 6.659 1.53X 17.24X 45.083 70.45X 38,259 6,865 45,181 148,339 236,444 2.63 2.80 3.41 4.73 3.83 9.48X Z' 1.65X 11.20% 36.27X 58.813 Parana Santa Catari R Grande Sul SOUTH 24,917 18,388 23,637 1,942 2.92 8.16 2.70 2.88 6.42X 3.45X 6.09o 15.95x 17,708 10,911 17,859 46,478 2.08 2.58 2.04 2.18 6.99x 4.813 7.06X 18.353 7,209 2,477 5,778 15,464 0.86 0.58 0.66 0.72 5.34x 1.833 4.283 11.45S 23,662 13,131 23,348 60,141 2.77 3.10 2.67 2.80 5.86x 3.25% 5.79X 14.91X M Grosso Sul Mato Gronso 5,568 6,488 11,257 8.32 4.09 2.48 1.43X 1.67X 2.90% 4,006 4,710 9,205 2.39 2.98 1.98 1.58x 1.8ex 8.63x 1,547 1,758 2,052 0.92 1.11 0.44 1.15X 1.303 1.523 6,165 6,966 11,157 3.68 4.41 2.41 1.533 1.73% 2.77X 28,278 2.43 5.99% 17,921 1.87 7.07X 5,857 0.66 8.97x 24,288 2.54 8.023 2.75 100.00% 258,386 1.79 100.00X 135,048 0.95 100.00% 408,450 2.85 lOO.OO0 Pornmbuco Golas Distrito Fod CENTER WEST BRAZIL Source: 888,a32 MINIFAZ/SEF 1.0SX Table 23 (Continuation) Federal OWN REVENUES/EXPENDITURES Uit Rondonia Acre Amazonas Roralsa Par* Amapa NCRTH 20.23X 9.46X 14.23X 11.00X 23.06X 13.72X 18.11l Maranhao Piaui Ceare' R Grande Norte Paraiba Pernambuco Alagoas 2.66X 8.96X 27.03X 24.17X 12.63X 24.898X 21.19X Sergipe Bahia NORTHEAST 23.77X 83.21X 24.09% MinasCorals Espirito Santo Rio de Janeiro SaoPaulo 23.27% 31.09% 61.62X 41.60X SOUTHEAST 40.24X Parana Santa Catarina R Orand.Sul SOUTH 30.47X 18.86X 24.76X 26.71X M Grosso Sul 25.09% 26.24% 18.39X Mato Crosso Goias DistritoFederal CENTERWEST 22.06X BRAZIL 33.47X Source: UINIFAZ/SEF 00 79 BOX 3 STATE Revenue Source nd Distribution 1 MUNICIPALREVENUE SHARING IN BRAZIL - Distribution Crit.,ria State Value Added Tax Mlv .25 * ICMS (ICAIS) 'V,)* p + ( other factors) (1-p)| VS Sl,ares: where: State 0.75 Ml Mtuni. 0.25 VA = Value Added (average of past two years) Funds allocated to municipality I = value of outflow of goods + value of services rendered within municipality - value of inflow of goods. p = proportion of funds distributed by values added component The following range for p is specified by law (L.C. no. 177) .75 s p s 1. Other factors = Each state is given complete discretion over specific other factors to be included in the formula. x lotor Vehicle Registration Tax (IPVA) Shares: State by origin. Immediate credit of municipality upon coilection. 0.50; Munic. 0.50 c. 10% of Federal IPI (Pass-through revenues) Shares: State: 0.75 Munic. 0.25 - Returned by State Trea-ury - Same as ICMS Note: This is intended to provide financial compensation to states for loss of ICMS revenues on account of exports. 'Source: Projeto de Lel Complementar No. 177 80 The State of Para uses population(7% weight)area (2% weight),and fiscaleffort (9% factors.In addition,it distributes7% of the fundin equalamountsper municipality weight)as specidal (see Box4). The Stateof Paranausesproportionof populationin ruralareas,populationand area as specialneed factors. 2.62 Conditional Tmusfers are not detailsneededfor ananalysisofstateconditionaltransfersto municipalities Varmous avilable. Anecdotalevidencesuggeststhat moststateshave a largenumberof conveniosusuallyin thousandsto provideprojectassistance.Further that mostof these projectgrantsare motivatedby politicalconsiderationsratherthan anysound(or even unsound)economicanalysis. 2.63 AmEvaluationand PolicyRecommeniations The existingrevenuesharingsystemprimarilyservesto return ICMSrevenuesbyorigin. 25%of totalrevenuesare intendedto be consistentwithfiscal edsof individualmunicipalities.It is in this area that formuladesignsof individualstatesneed re-examination.Considerthe stateof Para, 75%of the fundsvarydirectlyby value-added,9% by ratioof municipalrevenuesto state revenues. Thusfor84* of the fundsr -nicipalitieswith aboveaveragefiscalcapacitybut not necessarilyhaving higherfiscaleffortstand to gainmorethan proportionately. sizedeterminedbyrelativepopulation Another9% are distributedbytakinginto accountmunicipality ani area. Finally,the remaining7% are allocatedequallyamongallmunicipalities.Fiscalequalization byvaryinga proportionof fundsinverselywithfiscalcapacity( taxbasesfor municipalsources)is not re oed in this formula. in fact municipaltaxbaseshardlyenter into the formula. Eventhe fiscal onlywithoutanydue effortcompensationispoorlydesignedandservesto benefitlargermunicipalities rgurd fr their fiscaleffort. existingformulacan be improvedby; (1) incorporatingan explicitequalization mhe to componentin the formulabasedon a ratioof per capitaaveragefiscalcapacityof all municipalities 81 BOX 4 MUNICIPALITIESSHARE OF ICMS STATE OF PARA MunicipalitiesParticipationCoefficient:1 ( II V.ADg I V.ADE 75 POPI + 7 POPE E Al )+ E + 9 2 AE E RTI \ _ RTE where: I = Municipality'si participationcoefficient ,.V.AD V,AD = ) V.ADE ,POP ) _- POPE = Ratio of Municipality'sValue Added and State's Value Added. Ratio of Municipality'sand State's Population. E ( RT IT ) = Ratio of Municipality'sand State's Tax Receipts E N = Number of Municipalitiesin the State. A = Area Note: The Municipality's Value Added Is given by the sum of Its Aggregate Value in the two years that preceded the computationof the coefficient: V.AD = I where: V.A + 1,t-1 V.A I,t 82 V.A = Aggregate Value of MunicipalityI at year t. Aggregate Value Formula: VA I+C+P.P + ( ICMRET + EST ) + ( HFG+CFO+COOP ) where: IC = Difference between in-flow and out-flow OI goods (Industry and Commerce). P.P = Primary production. ICMRET= ICM withheld EST = estimates of ICM revenues computed by the regional bureaus of the state secretaryof Finance. CFP = (Company for the Financing of Production). Value of goods purchased. COOP = value of the comme-ce undertakenby cooperatives. 1 Source : Governo do Estado do Para - Secretaria de Estado da Fazenda. ICM dos Municipios - Manual da Cota Parte 1989 83 per capita fiscalcapacityof municipalityi in the formula and distributinga fixedpercentage of fundsby this factor and; (2) the current fiscaleffort indexshould be replaced by the followingmeasure: FiscalEffortComponent= (percapitaown revenues/percapitataxbase)dividedby(average per capita own revenues of all municipalities/average per capita tax base for all municipalities). Shah (1983)specifiesaltemative approaches to incorporatingthese factorsin a formula. The incorporation of fiscalequalizationand fiscaleffort components into the formula for state-municipalrevenue sharing systemwill providespecial compensationfor fiscalneed and fiscaleffort but at the same time willhelp states better monitoring of the fiscalhealth of municipalgovernmentsin their jurisdictions. Conditionaltransfer programsare also in need of restructuring. Only a handfulof programs with explicit objectives say e.g. spillovercompensation or ensuring certain minimumtransportation standards etc. need to be developed and then their design should reflect their objectives. For example, a transportation program which intends to compensate a municipalityfor the use of its roads by nonresidents would be a matchingprogramwith matchingdetermined by the spilloverfactor. A program to upgrade transportation servicesto certain minimumstandards wouldbe a blockgrant program with adherence to the specified standards as a pre-condition for receipt of funds. Followingthis approach, the Brazilianstates couldreduce their conditionalprogramsto a dozen or less programfrom the current count of several thousands. Such a design will enable the higher level government to achieve its objectives in a cost-effectivemanner without distorting local priorities. Given the almost universal criticismof existingconditionaltransfers in Brazil,the reform of these arrangementsshould be a matter of high priority for ail levelsof governments. 84 3.0 SUMMARYAND POLICYRECOMMENDATIONS In this chapter, the major results of previous chapters are brought together to provide an overviewof the existingstate of riscalfederalismin Braziland the directionsfor change advocatedin this paper. For this purpose, first the tax and expenditure assignment issues are reviewed and their implicationsfor horizontal and verticalfiscalbalance are examined. Second, federal transfers to states and municipalitiesare briefly described and analyzed. Third, state transfers to municipalities are examined. Fourth, implicationsof the above analyses as to the potential directions for reform are summarized.Finally,the implicationsof the reformproposals for efficiencyand equityof publicservice provisionand macroeconomicmanagement are addressed. 3.1 The Assignment Issues The 1988Constitution has made a reasonablyclear assignmentof publicservice provision and revenue raisingresponsibilitiesto federal,state and municipallevelsin Brazil. Purely localfunctions such as intracitytransport, zoning,preventivehealth care and elementary education has been assigned to the municipallevel exclusively.The responsibilityfor publicservicesthat are national in scope such as defense and foreign affairsetc. has been entrusted to the federal level and the remainingfunctions are designated as shared responsibilitiesof the federal and state levels with the federal government setting the norms and the state government responsiblefor the deliveryof services. Unfortunately, the de facto assignmentor the practice in Brazil is at substantial variancewith the de jure assignmentand the federal government's direct involvementin purely local functionsis quite pervasive. The constitutional taxassignmenthas created some difficultiesas welL These are primarily in the area of sales taxes. Al three levels have partially overlapping responsibilitiesin tax policy development and administrationfor value-addedtype taxes. The federal government is responsiblefor a manufacturerlevelsales taxcalledbythe name of IndustrialProduct Tax (IPI). The municipalitiescan levy a value-added tax on services (ISS). The states have the mandite of a general value-added tax whosebase encompassesIPI and ISS. This multiplicityin administrationraises the administrationand B5 compliancecostsof these taxes. Further, Brazilis the onlycountryin the worldwith a state level general value-addedtax. This has the potential of evolvinginto a tax with multiple rates on non-uniformbases in the long run. Already inter-state tax crediting issues remain by and large unresolved and are fast becominga source of major concern for the Councilof States. A second source of difficultyconcerns the assignmentof the rural property tax. This tax is more suitable for administrationat the state-local level but is currently being administeredat the federal level. 3.2 Issues in Intergovernmental Transfers The federal transfers to state and municipalitiesare based either on revenue sharing arrangementsor specificpurpose transfers. Revenue sharingmechanismsare welldefind One of the main instruments for federal-state revenue sharing is the State Participation Fund. The Federal Gcvernment transfers a pre-specified share of certain federal taxesto this pool. The Councilof States then determinesstate shares based on a formulathat incorporatespopulation and per capita income as its maincomponents. A proposal currently under discussionwould extend this list of components to include land area, interstate trade auidfical effort factnrs. In recent years formuladetermined shares havebeen found unacceptableto the Counciland therefore, it had to resort to a compromiseallocation based on an arbitrary adjustment to formula shares. The principal merits of the this program are the consistencyof its designwith transparency, predi^tabilityand local autonony objectives. The program further payssome attention to fiscalequalizationobjectives.The programnevertheless has manydesign flawswhich inhibit achievementof its objectives. For example,state per capita income is included as a measure of fiscalcapacity. It is an imperfect guide to the abilityof a state government to raise taxes as it isconceivablethat a significantproportion of income mayaccrue to non-resident ownersof factors of production. Further only a small proportion of to.a state revenues are raised from income taxes alone. This measure is alsosubject to implementationdifficultiesin Brazil. Estimates of state per capita income are subject to significanterrors and are availablewith a long lag. For example,currently only 1980estimatesare available. These difficultiessignificantlydiminishthe usefulnessof per capitaincome for use in a program of fundamental importance in federal-state fiscal relations. The FPE further 86 combinesdiverse and sometimes conflictingobjectivessuch as revenue sharing and riscalequalization at the state level into a singleformula in a multiplicativemanner and therefore significantlyfallsshort on individualobjectives.The programis redistributivein its overall impactbut consistencyof individual state shareswith the formulaobjectivesis not assuredand states withsimilarfiscalcapacityreceivewidely different entitlements. Sincethe formulalacksan expiicitequalizationstandard, it also failsto address regional equity objectives in a satisfactorymanner. These failingsexplains why the Councilof States findsit easier to strike a politicalcompromiserather than accept the formula results. The program to channel federal revenue sharing monies to municipalities is called the MunicipalParticipationFund (FPM). This programconsidersmunicipalpopulation and state per capita income in the determination of shares of individual municipalities. This program has two major drawbacks. First, the formula used for this program fails to incorporate differential fiscalcapacityof the Brazilianmunicipalitiesin a meaningfulwayand therefore formula application does not result in a fair and equitable distributionof fundsamong individualmunicipalities.As there is no local income tax in Brazil(and none iscalled for due to capitaland labour mobility),per capita income is a poor indicator of a local government's abilityto raise revenues. Further, in each state rich municipalitiescoexistwith poor municipalitiesbut state per capita income, by definition,would fail to make a distinctionbetween the twoclasses. Second, this program,is observedto discouragelocalfiscaleffort by meetingnearlytwothird of municipal revenue requirements from the federal revenue sources. Such overwhelming dependence of municipalgovernmentson outside revenuescreates a dichotomybetween spending and revenue raisingdecisionsand contnbutes to reduced financialaccountabilityat the local level. Specific purpose transfers support important policy objectives in a federation. These objectivesinclude:benefit spillovercompensation;bridgingfiscalgap; ensuring minimumstandards of publicservicesacross the nation; fulfilLmentof the redistr.:butivefunction of the federal government; creation of a common internal market; reduction in net fiscal benefits across jurisdictions and achievementof economicstabilizationobjectives. Grant objectiveswould pre-determine grant design. The federal and state governments in Brazil have an incrediblylarge number of specific purpose programs. For manyof these programs,programobjectivesare either not specifiedor specifiedvaguely and in some instancesreflection upon grant objectivesis done after the release of funds. In more recent 87 years, federal specific purpose transfers have served not as a means of safeguardingfederal objectives but increasinglyas a vehicle for pork-barrel politics. There are only a handful of programswith some desirable features. One such programis for unified and decentralizedsystemof health care provision known as SUDS. Federal financing is provided to achievecertain minimumstandards of health care acrossthe nation. The intent of the programis for the federal government to specifypoliciesand state and local governmentsto implement federallymandated programs. In practice, however,the federal government continues to be heavily involved in program administration as weli and therefore the decentralizationobjectiveshave as yet not been fullyachieved. In the comingyears, fiscalpressuresof the new fiscalarrangements on the federal government are also likelyto constrain the federal funding for the SUDS. Finally,the existingprogram gives preferential treatment to private contractors over state and local government agencies. Finally,State-municipaltransfers have two important components. One such component is the constitutionallymandated state-municipal revenue sharing arrangementsor state-municipaltax transfers. The distribi'tionof such transfers for the most part followsthe originprinciple. 75% of state value-addedtax (ICMS) revenuesare distributed in proportion to the value added in each municipality and for the remaining25% states have been givendiscretion in incorporatingfiscalneed factorsas they see fit. Population and area are the two most commonlyused need factors. Somestates have also used fiscaleffort as a special factor. A major criticismof the existingarrangementsis that fiscalequalization by varyinga proportion of funds inverselywith fiscalcapacity(tax bases for municipalsources) is not recognized in the formulae currently in vogue. In fact, municipal tax bases hardly enter into any consideration. Even the fiscaleffort component is usuallypoorly designedand serves to benefit larger municipalitiesonly-iithout any due regard for their fiscaleffort. A second component of state transfers to municipalitiesis specific purpose or negotiated transfers. Most states have a large number of conveniosusuallyin thousands to provideproject assistance. Sheer numbers of these transfers defy any analysis.Anecdotal evidencesuggeststhat politicalconsiderationsdominate in the distnbution of grant funds. 88 3.3 Implicatons of ExistIngArragements This assessmentof existingstructure of fiscalrelations has identifieda numberof significant failings. Most prominent of these are: a. Federal and state governments are involved in some purely local functions in an uncoordinated fashion; b. Sales taxesadministrationbythe three levelsresults in someduplicationand confusion; c. The administration of the general value added tax at the state level creates yet unresolved issuesin tax creditingon interstate trade; d. The FPE and FPM fail to distribute revenues in a fair and an equitable manner; e. Conditional transfers are arbitrary and primarilydriven by politcal considerations. Various programs work at cross purposes and therefore hamper successin achieving major objectives. Highlysubjective nature of these transfers may even be sending wrongsignalsto lower levelsof government regarding fiscalmanagement. For if it is alright for the federal government to follow imprudent policies then lower level governmentsalso would not feel the need for fiscalrestraint. ; and f. The net impactof expenditureand tax assignmentand revenue sharingmeasureshave been to severelyconstrain the federal government's abilityto fulfillits mandate as a national government while generous availabilityof funds to municipalgovernments without any concomitant increase in their responsibilitiescreate 3trong incentivesfor fiscalmismanagement. There is someevidenceto suggestthat munic.palgovernments are,shyingaway(om raisingrevenuesfrom property taxesand user charges. Ihe state governmentsare also fecinga financialsqueeze in the short run but giventheir access to thevalue addedtax-a dynamicsource of revenues,their fiscalproblemsare expected to be short lived. Thus the existing fiscalarrangements have been instrumental in creating a vertical imbalancein the nation. The problem for the federal government is structural in nature. Its revenue means significantlyfail short of its expenditure needs. Fiscalimbalanceforstate governments,on the other hand, is likelyto disappear 89 with increase in revenues from the ICMS and restraints in expenditures. Fiscal imbalance for the municipalgovernments is of the reverse nature. Their current revenue means if fullyexploited substantiallyexceed their expenditure needs. In conclusion,manyaspectsof the existingarrangementscontribute to a fiscalmalaisein the countryand therefore, an urgentneed for reformcan hardlybe over-emphasiz-.d.The followingsections presents some reform options. 3.4 Proposals for Reform The followingreform options are suggested to deal with the shortcomingsof the existing arrangements identified earlier. Towards an Economic Constitution: Earlier analysissuggested that the Constitutional assignment in Brazil is broadly consistent with economic principles. The followingsuggestions, if implemented,willmove Brazil a step closer to havingan economicconstitution. i. Immediate tumback of direct federal involvementin functionsof purely local nature such as primaryand secondaryeducation, urban grading,bridges,zoningetc. Further, administrationof health and education as stipulated in the Constitution should be a state responsibility.Therefore, the roles of federal ministriesof health and education be reduced to setting minimumstandards and providing per capita block grants to induce compliance. ii. Direct federal role in municipal finance and administration need to be severely restrained. iin. he three sales taxes,IPI, ICMS and ISS be combinedinto one taxto be administered by the federal govermmenton behalf of state and local governments. The proceec. from the tax then be shared by the three levels in proportion to their current intake from this source. iv. The administrationof the rural property tax be turned over to the state level. v. A restructuring of revenue sharing and transfer programs as outlined below. s v _ 90 Restructuring Proposals for Federal Transfcrs to States: i. No change in the source of fundsfor the State Participation Fund is proposed. Initial allocationsfrom this Fund wouldbe based simplyon the derivation principleand final allocationsto be determined by equalizationadjustments discussedin the following. ii. It is proposed that the existingsystemof negotiated transfers be completelyrevamped. These transfersmaybe replacedbya systemof specificpurpose blockper capitagrants. Thesegrantswillbeconditionalonmeetingfederallyspecifiedandobjectivelyverifiable criteria relating to minimumstandards and access. iii. It is proposed that an equalization program be instituted to deal with inters'4ae differentialsin fiscalcapacity.This programwouldbe financedby applyingpositiveand nepative entitlements based on a formula to the monies available for distribution through the FPEbased on the originprinciple.The representative taxsystemapproach as outlined earlier maybe used to determine state entitlements. Restructuring Proposals for Transfers to Municipalities i. It is proposed that federal unconditional transfers to municipalitiesbe completely eliminated. To this end, the Municipal ParticipationFund be disbandedand states be encouragedtostrengthen theirunconditionaltransfersprogramstoownmunicipalities. Thus the FPM could be reconstituted at the state level. State transfers to municipal levelson account of a restructured FPM would be based on a formula similar to the FPM but which incorporatesper capita municipalfiscalcapacityas an important factor and further incorporates a basicper capita grant and a fiscaleffort component ii. State specificpurpose transfers be consolidatedinto a handfulof programswith clearly stated objectivesbased on verifiable indicators. 3.5 Implicationsof Reform Proposals The restructuring options discussedearlier willsatisfythe followingobjectives. i Tax and expenditure assignmentwillbe brought in close conformitywith economic principlesand thereby enhancing the efficiencyand equity in publicservice provision. ii. The proposed equalization programwould foster a sense of greater participationby member units in the Union. The have-not states wilHrealize a contnbution by the Union to their uplift and the havestateswillenjoysomesatisfactioninseeingan explicit recognitionof their contributions for the greater well-beingof the nation as a whole. Fiscal management goals will also be advanced by this program as it will provide monitoringof fiscalcapacities of member units. Higher level governmentswould be in a better position to take correctiveactions for fiscaladjustmentin a iimelyfashion. iii. The systemof transferswoud be rationalizedofferingpredictabilitywhilesafeguarding local autonomy objectives. Long term planning and judicious exploitation of own revenue bases would be encouraged. iv. Finally,structural imbalanceswould be eliminated and remaining imbalancescould potentially be corrected byown actions byvarious govermnents. 92 APPENDIXA THE DESIGNOFA FISCALEOUALIZATION PROGRAM: ISSUES ANDOPTIONS An equalizationprogramin additionto safeguardingnationalobjectivesof providingcertain minimumlevels of public servicesacross the nation can foster a greater sense of participation in the federation of member states and therefore is often viewedas a glue that holds a federation together. Economicliterature has long recognizedthat equalizationis justifiedon horizontal equity groundsand in recent years, that under certain conditions it could promote economic efficiency. Recent constitutional changesin Brazilsuggest that equalization is a matter of high priority in the country. A recent World Bank missionhas nevertheless observed that the existingfiscalarrangements fail to deal adequatelywiththis objective. The followingparagraphsdeal withsome fundamental issuesin designing an equalizationprogram. In principle, a properly designed fiscal equalization transfers program would correct distortionscausedbyfiscallyinducedmigration.Sucha programwouldequalizenet fiscalbenefits across states and thereby promote economic efficiency. To measure net fiscal benefits reasonable approximation of costs and benefits of public servicesprovision in various states is essential. ITis requiresdevelopingmeasuresof differentialrevenue raisingabilitiesand the costsof provisionof public servicesof the Brazilianstates. Equalizationof net fiscalbenefits couldthen be attempted by adopting a standard of equalization and establishing the means of financingthese transfers. These and related issuesof unconditionality;tax effort; stabilizationeffects; and employmentof strategy are discussedin the followingsubsections. Measurement of Fiscal Capacitv The estimationof fiscalcapacity,i.e.,the abilityofgovernmentalunitsto raiserevenues from their own sources, is difficultboth conceptually and empirically. The alternative measures of fiscal capacityare unlikelyto show approximatelythe same results. Of a large variety of such measures that are availablethe two most prominent measuresare discussedbelow. 93 Macro Indicators Various income or output measuresserve as indicatorsof abilityto bear tax burdens by the residentsof a state. Among the better knownmeasures are: (i) Personal Income: The personal income of a state is the sum of all incomes receivedby the residentsof a state. It is not a satisfactorymeasureof overall fiscal capacity as it is a measure of abilityto bear tax burdens but a highly imperfect and partial measure of abilityto impose them. (ii) Personal Disposable Income: Personal disposable income is defined as personal income less direct taxes. This concept of income shares the weaknessof personal income as a measure of fiscalcapacity. (iii) State Gross Domestic Product: It represents the total value of goods and servicesproduced withina state. It also is an imperfectguideto the abilityof a state government to raise taxes as it is conceivable that a significant proportion of income may accrue to non-resident owners of factors of production. Evenif these conceptualproblemsare ignored,significanterrors in esti.natingprovincialgross domestic productin Brazilare encountered to make the measure unacceptable for an equalizationprogram. We already noted that the equalizationof net fiscalbenefits acrossstates is required from the standpointof economicefficiency.The estimationof these net benefits is best done by a comparativeanalysisof taxingand spendingbehaviorof state and localgovernmentalunits. Various income concepts do not relate to the taxing practicesof the states but merely indicatewhat they potentially have available for taxation. Fortunately, a representative taxsystem approach is well suited for such a task. This approach is discussedbelow. The Representative Tax System (RTS) This systemmeasuresthe fscal capacityof a state by the revenue that couldbe raised in that state if the state governmentemployedall of the standard sourcesat the nationwideaverage intensity of use. 94 To estimate equalizationentitlementsbasedon a representative taxsystemapproach, informationon both the taxbases and taxrates for each state is required. For most revenue sources this informationis usuallyreadilyavailable. Then a decision has to be made as to the standard of equalizationi.e. whether fiscalcapacityof the have-notstates shouldbe brought up to the median, arithmaticmean or some other norm based on all states data. As an example,consider arithmatic mean of all states as a standard. Then equalization entitlement for a state say (x) for resource source i could be determined as follows: EL (POP). - where {(PCTB), x t.) {(PCTB)5 x t,,. E' = equalizationentitlement of state from revenue source'. POP = Population PCTB' = Per capita tax base of revenue source i. e = national average tax rate of revenue source i. subscript na = national average subscript x = State x The equalizationentitlement for a state froma particularrevenuesource couldbe negative,positive or zero. These figureswould then have to be summedup for all revenue sources considered for equalization and the overall sum would indicate whether a state would receive a positive or a negativeentitlement from the interstate revenue sharing pooL It should be noted that the data on tax bases and tax collections required for the implementationof a RIS are alreadybeingpublishedon a regularbasisby variousdepartments and agencies of the federal Government. Thus the RTS does not impose any new data requirements and could be implementedusing the existingdata. Measurement of Expenditure Needs Economictheory suggeststhat an ideal equalization transfers program should also 95 take into considerationthe expenditure side of the provincial-localbudgetaryoperations. Many economists have argued for taking expenditure needs and differential unit costs of provision of public services into account. Several countries follow this approach. The followingparagraphs examinethis issue. We alreadynoted that the casefor equalizationrestson differentialnet fiscalbenefits acrossstates. These differentials could arise either due to differences in revenue raisingcapacity and/or due to differences in the cost of provisionof publicservices. Consider two states with the samerevenue raisingcapacitywherethe residentshave identicaltastes for provincialpublicservices but the cost of providingthem differsdue to supplyfactors. For example,differences in the degree of urbanization,population density and age distribution amongstates will have significanteffects upon the relativecostsof publicservices.The degree of urbanizationcaneffect the costsof salaries and wages,land, and construction, as well as particular servicessuch as pollution control, public transit, police and fire protection and the provisionof utilities. Population densitywilleffect the costs of providingpublic utilities and willalso effect the costs of highways. Age distribution will influencethe need for school rooms, hospitals,recreational facilitiesetc. These differential costs are likely to cause substantial variations across the two jurisdictionsin the level and mixofpublicgoodsprovided,resultingin differentialnet fiscalbenefits. A strong casefor equalizationcan. therefore, be establishedon both efficiencyand equitygrounds to compensate for cost differentialsthat giverise to differential net fiscalbenefits. The fical federalismliterature, ingeneral,treatsdifferentialcostsassynonymouswith differential needs but it must be noted that some cost differencesmayarisedue to deliberate policy decisions of the provincial governments and thus do not constitute need. Compensation for unavoidablecost variations resulting from differences in the costs of inputs and from dissimilar input-output relationships which might arise because of distance from sources of supply and geographic features can be justified on equity grounds. Equalizationgrants should offset such inherent disabilitiesbut shoulddisregardcost differencesdue to differencesin efficiencywithwhich resources are used. These questions do not pose any special difficultiesfor a regression based approach to the measurement of expenditure needs. 96 Expenditure need is more difficult to define and derive than a measure of fiscal apacity.Thedifficultiesinvolvedinmeasuringexpenditureneedaresubstantiallyhigherthan those encountered in usinga representative taxsystemto measure fiscalcapacity. They include defining an equalizationstandard,determiningdifferentialcostsdue to differinginput-output relationships, nature of service areas, composition of population and isolating need/cost differentials due to differentialtastes or policydecisionsas distinct from inherent cost disabilities. A further concern would be the susceptibilityof the grant shares based on need factors to strategic behavior on the part of the recipient states. The experience of Australia, West Germany,Switzerlandwith federal unconditionaltransfers, of the U.S.with highwaygrants and of the Canadian states with provincialmunicipal transfers, indicates that these concerns can be addressed and expenditure need incorporated in formula grants in a manner acceptable to both the donor(s) and the recipients. Some empirical questions are resolved easily. For example, to avoid problems associated with subjective standards such as 'minimum service levels' or 'reasonable levels of services',expenditure need could be defined as "the cost of supplyingaverage performance levels for the exsting mix of provincial- local programs". Relative expenditure needs could then be determined empiricallyeither using direct imputation methodsor by adopting a simpler approach usinga representative expenditure system. The latter approach is preferred for its objectivityand ease of computation. Furthermore, it enables the analyst to derive expenditure need measures based on actual observed behavior of the provincial-localgovernments under study rather than basingit on ad hoc value judgements. The relativeweights to be assigned to various need factors in the representative expendituresystemcouldbe determined byeconometric analysis.This method requiresspecificationofdeterminants foreachservice categorytobe analyzed.Thesedeterminants would include relevant fiscalcapacityand publicservicesneed variables. The estInating equation so specified would then yield quantitative estimates as to the independent influence of each specified factor in determining the spending level for that category of public service. This informationcould also be analyzedfurther to determine as to what each state would have actually spent if it had average fiscalcapacity and average tastes but actual need factors. Morespecifically,the formulaforequalizationentitlementon accountofexpenditure 97 classificationi for state x could be stated as follows: EEIX = (POP), {(PCSE)I - (PCSE)',,J) where EEi, = Equalizataion entitlement on account of expenditure classificationi for state x POP. = (PCSE)i, = Population of state x Per Capita standardizedexpenditure by state x on expenditure classification(i). This is the estimatedexpenditurewhicha state would havespent if it had national average fiscalcapacitybut its actual need factors. (PCSE)',,. = National average per capita standardized expenditure on expenditureclassificationi. This isthe estimatedexpenditure for all states based on national averagevaluesof fiscalcapacityand need factors. Equalization entitlement on account of a particular expenditure classificationcould be positive, negativeor zero. These entitlements would have to be summedup for all expenditure categories considered for equalization. Overallentitlement of astatebased on a comprehensivesystemofequalizationwould be determined by summation of its separate entitlements from the Representative Tax System (RTS) and the Representative Expenditure System (RES). Only the states with positive entitlementswouldbe eligibleto receivetransfersin equivalentor somefraction of the total amount (the fraction to be determined by the centre depending upon the availabilityof funds) from the centre. A phased approach to a comprehensiveequalization maybe well advised. Initially, a representative tax systemcouldbe implementedfor a fiveyear period. Then depending upon this experience a reSresentative expendituresystemcould be brought in to complementthe RTS in the next fiveyears. A joint Federal - States FiscalArrangementsSub-Committeemaybe instituted to monitor the working of the systemclosely. 98 The Equalization Standard Equalization of net fiscal benefits requires that we adopt an explicit standard of equalization. The specifiedstandard wouldbe the level to whicheach state wouldbe entitled to be raisedto enable it to providepublicsector net benefits per household comparablewith other states. Simplicitydictates choosing either the arithmatic mean or the median of the govemmental units involved as the standard. Arithmatic mean provides a good representation of the data in the absence of extreme values. In the event that the sample values have wide range, the median or the arithmaticmean after eliminationof extreme valueswould providea better representation of the sample. Mean is to be preferred over the median, however, for ease of computation. Costs and Financing An ideal fisca; equalization program would be self-financing. The member governmentsare assessedboth pvsitive and negativeentitlementswhich sum to zero. The federal government merely acts as conduit for such a policy. If such an interstatal revenue sharing pool in practicewouldcreate, administrativedifficulties,then the equalizationprogramcould be financed out of general federal revenueswhich are in part derived from the equalization receivingstates. Other Considerations So far the discussionhas focussedon the basicelementsof an equalization program. Several related aspects of equalizationtransfers are considered in the followingsubsections. Unconditionaliq There is a general consensusin the academicliterature that an equalization system shouldenable stau. governmentsto providea standard bundle of publicservicesif the government imposesa standard level of taxes on the bases at its disposaL The state governments(or certainly their citizens)should, however,be permitted to substitute lower rates of taxation for higher level of services and vice versa. As such the equalization payments should be in the nature of unconditional grants having only income effects. Service areas in which there appears a good 99 reason to actuallyset minimumnational standards are better handled by conditional grants and shared cost programs. It should be noted that by raisinga state's fiscalcapacity,the unconditional equalization grants enable the poor states to more easilyparticipate in the shared cost programs. Tax Effort Incorporating taxeffort into the formula for determiningequalizationwouldinNolve makingthe equalizationentitlement a functionof the ratio of actualtax collectionsin a state to its own tax base. Potentially non-recipient states maywishto see such a factor incorporated into the program to prevent states with a positive fiscaldeficiencyin an area from collectingequalization paymentseven if they maynot levya tax in the area. Potentiallyrecipientstates maywishto see tax effort incorporated becausewithout it, extra taxeffort on their part willbe relativelyunproductive compared to a wealthy state. Problemsexistwith incorporatingtax effort into the program. First, the inclusionof tax effort willcause the programto depart from its unconditionalnature. A state should be free to substitute grant funds for revenue from own sources. Similarly,if a state raises taxes in order to provide a bundle of servicesthat is higher than the standard it should not receive equalizationfor doing so,e.g. other states should not haveto pay most of the cost if a state decidesto paint its roads. Incorporatingtaxeffort wouldtie the federalgovemmenttoexpenditure philosophiesof thevarious states. A problem also arises in that some states do not have tax bases in all areas. Incorporating tax effort may also encourage the employment of strategy by a state. Another major problem associatedwith the inclusionof tax effort in the formulais that in viewof the differential abilities of the states to export taxes,the measurement of tax effort wouldbe crude. Inclusionof taxeffort in the formula could also result in increase in taxes on the poor states. In view of the above considerations, it appears that a program of equalization payments would not be improved by includingtax effort. StabilizationEffects If the equalization payments in Brazil were to be based upon relative measures of 100 fiscalcapacity,they willbe expected to have a stabilizingeffect upon state revenues. The level of paymentswillmove in a directionopposite to that in whichthe states own revenue raisingcapacity moves. Maximumstabilizationof state-local revenueswilloccur when the payments are based on all revenue sources, a national average standard of equalization is used, cyclicalfluctuations in provincialeconomies are small and the time lag in calculatingthe grants is relativelyshor. When any large component of the total base is quite volatile, such as natural resource revenues, the destabilizingefiects can be quite large and some sort of averaging formula would have to be employedto ease the difficultiesassociatedwith provincialbudgeting in the face of uncertainty. Strategy Strategy refers basicallyto the actionsthat provincialgovernmentscan take to affect the level of payments they receive. A scheme that enables a state to employ strategy must be considered undesirablebecause in general such extra paymentsreceivedmaynot have any relation with actual disparities. For example,a programemployingtax effort could enable states to raise their payments by imposingheavytaxes in areas in which they have a below average base. This problem, however,is muchless serious in practice than it might appear as the room for additional taxation from sources in which the potentially "have-not"states are not well-endowedwould be extremelylimite.d. ConcludingRemarks Economictheory providesa strongrationale forfiscaldecentralizationas it promotes efficient provision of public services by promoting a better match of these serviceswith citizens preferences; by minimizing the cost of political decision making and by encouraging political accountabilityand by addressingregional and local concerns. In Brazil such a program could be helpful addressingregional equity and stabilizationobjectivesof the federal government. 101 APPENDIXB OPERATIONALMECHANISMSOF CONVENIOS The provisions of convenios vary by functional objectives. There are some general procedures,however,that shouldbe observedwhen forminga convenio(Decreto-lei #200 of Feb/2/67, 2300of Nov/11/86,93872of Dec/23/86.) A convenio is legallydefined (200/67)as a way to decentralize the activitiesof the federal governmentby delegation of powers and funds for the conduction of federal projects (projects under the responsibilityof the Union) to the local authoritieswhere the project is to take place. The source of the funds can vary in manyways. Usually the funds come as supplemental credits or as recursos vinculados (earmarked resources). In the first case, the federal government uses the excess in tax collections to supplement the funds allocated to conveniosby more than the other funds. As far as earmarked sources are concemed, these would be the FINSOCIAL, PIN/PROTERRA, education salary,and lotteries, as well as the share of the special fund (FE) for whichthere is no fixeddistribution criteria. Stages forthe implementationof a convenio:proposal, authorization, penho", delivery of funds,monitoring and control. For each petition filed by a state/local government, there is a separate proces-. In the petition there must be a justificationand, in mostcases,a detailed plan for the use of the funds(schedule of tasks and expenses). The pctition isexaminedbya technicalcommittee. This analysischecksfor the compatibility of the petition with the activitiesof the ministry. Once the technicalcommitteegivesa favorableappreciation,the petition isforwardedto the general secretary of the ministryso the value of the convenio can be approved and so that it can be verified whether funds are available. A convenio contract has the followinggeneral characteristics:definitionof the authorities involved,the object of the convenio, the duties of each part, a schedule for the deliveryof the funds, deadlines and conditionsfor alteration. The conveniois usuallysigned in a publicceremony in the state 102 capital. Convenioslargerthan CzS2 millionmustbe publishedat the DiarioOfficial.(The official dailypublicationof the federalgovernment. For each conveniothere existsa checkingaccountat Banco de Brasilwhere to federal governmentdepositsthe fundsaccordingto the predeterminedschedule.There are then hundredsof suchaccounts.Thebeneficiary(stateor localgovernment)shouldreportto the federalgovernmenton the progressand utilizationof thefunds.However,thisnot alwaysoccurssincesomelocalgovernments lackthe personneland the expertiseto submitthe detailedreportsaskedby the federalgovernment. As far asthe employmentof the fundsisconcerned,however,the beneficiaryhas to at leastpresentthe statementsof the abovementionedaccountas a wayto documentthe proper use of the funds. The misuseof suchfundsconstitutesan actsubjectto criminalcharges.Usuallyit isonlyat the end,afterthe lastwithdrawalismadethat the federalgovernmentasksfor receiptson the expensesmade, In manycasesthebeneficiaryhasalsoto commitsomeof theirownrevenuesto the convenio, but there are no formalcriteriafor suchmatchingof funds. A completeand detailedaccountof the transfersreceivedthroughthe convenioisnotalwayssubmittedto theappropriatestate/locallegislature as so it isnot alwaysthe casethat onecan findthemin the officialaccountsof statesand municipalities intergovernmenttransfers. It isalsoverycommonto havechangesin the originaltermsof the convenioor to readjust predeterminedvalues If thelikelihoodof thesechangeswasnotforeseenwhentheconveniowassigned than a newconveniohas to be made. In manycases,valuesneed to be readjusteddue to the highrate of inflationand to the time it takes for the actualdeliveryof funds. To avoidthis problemsome local/stategovernmentsoverstatethe valuesof theirconvenios. Source:Afonso(1989). 103 APPENDLC C THE ADMINISTRATIVEPROCESS OF SUDS 1. Mechanismsof Transfers of Funds The funds transferred fromthe INAMPS (National Institute for Medical Assistanceof the SocialSecurity System)to the states come from the Fundo de PrevidenciaAssistenciaSocial (FPAS). This fund isconstitutedby the SocialSecuritycontnbutions of employersand employees(deducted from pay roll). The amount to be transferred is determined on a yearly basis by the "termos aditivost (additional agreements) as establishedby the convenios. Deliveryof Funds from INAMPS to States INAMPS = once the Additional Agreement (TA) is signed, it makes the transfer of funds in monthly installmentsaccording to the schedule of withdrawals (CD-see below) and to the availabilityof funds in its budget. BANCO DO BRASIL = funds are deposited in earmarked accounts. States uses part of the funds and transfers the rest to the municipalities. States = willreceivethe 4th, 7th, 10th and installmentsconditionedon the approval of the renderingof account (PC) coJncerningthe lastquarter. Consolidates the PCs of the municipalities. INAMPS/Office = has 15 daysto evaluate the PC. DOF/INAMPS = approvesthe TA elapses,they have 30 days to present a consolidatedPC for the entire year. The signingof a newTA isconditionedon the approval of the PC. 2 Matchbngof Funds As it was noted before, PO is the processby whichresources are allocated to the providers of health-care. There is some matchingof funds here since aUllevelsof government participate in the processwith some of their own revenues. There seems to exist no clear and predetermined matching condition, however. What followsis a description of the way the processworks: 104 CIMS (municip) - prepares the plan of application of funds (PA) and the schedule of withdrawal of funds (CD) for the coming year. PA and CD should discriminate between the allocation of go revenues and transferred revenues. CMS (municip) = evaluates and approve the PA/CD. Sends them back to CIMS. CIMS = has up to September 30 to send the PA/CD to the state (CIS). CIS = consolidatesthe state PA/CD wit1inhemanydifferentmunicipalones. The consolidated report will indicate the real growth in the participation of states in the financingof SUDS and the identificationof funds according to their source: municipalitvsae -NAMPE,and other. CES = evaluatesand approves the PA/CD. Sends them back to CIS. CIS = has up to November30 to send them to the Regional Officeof INAMPS. INAMPS = analysesthe PA/CD. Its main role is to provideorientation to CIMS and CISin the preparation of the PA/CD. Sendsthe PA/CD to DOF/INAMPS. DOF/INAMPS = analysesand consolidatesthe PC/CDs from allstates. It mayask the states to change them. Prepares the POI which is the unified health budget for the nation for the comingyear.Prepares the "termosAditivos"to be signed everyyearbyStates andMunicipalities.Submitseverythingto the approval of CIPLAN. Hasup to December 10to send approvalnotice to officesand states. CIPLAN = Approvesthe HTermosAditivos." INAMPS Offices = States and municipalitiessign the "TermosAditivosw(TA). The TA must specifythe amounts of co-participationof the parties involved. States/Mun. Source:Afonso (1989). managementof the budget through SES and SMS, respectively. Table A. TRANSFERSBY FUNCTIONANDBY LEVELOF COVERZIENT FEDERALCONDITIONAL (1987) (in percent) Functional functiona and Planning Administration Agriculture Regional Developvent Energy and Minerals Houain% and Urban Development Health And Sanitation Labor Total Source: - 'Balancos Gerais Distributions Total States Hun. 71.2 10.5 11.8 0.0 4.6 1.9 0.0 72.3 1.4 7.9 1.0 16.8 0.5 0.0 71.7 6.6 10.1 0.4 9.9 1.3 0.0 100.0 100.0 100.0 between Distributiors levels of govaroment States hun. 56.4 43.6 9.4 90.6 33.9 66.1 0.0 100.0 73.8 26.2 17.7 82-3 23.1 76.9 56.7 43.3 Total made transfers capital Intergivarnment account through the special investments (as a 2 of ...) Other InterOther interSpecial gav, transfers gaw, capital lnvestments transfers 100.0 100.0 100.0 100.0 100.0 100.0 100.0 69.9 40.5 53.1 0.4 30.1 34.1 19.6 1,220.3 1,364.0 439.3 107.4 17.5 184.1 165.9 87.8 405.7 13.9 107.4 5.8 72.7 100.0 37.4 302.6 70.7 da Uniao - 1987r, HINIFAZ and to Municipalities made ln the Special Investments Acccunt, to States --d DP, Include Capital Trausfers, BGU the in 4130 Account Execution: in Special Investments Account (4320) Capital Ttansfers: Intergoverment CurrentTransfers:Account (3220) Intergovernment Other Transfersa (3220+ 4320) minus FederalTax Transfers 4230.47 CapitalTransfersmade to iratesIn the SpecialInvestments:Account lutergovernment 4730.48. Account Investment Special the in Mun. to Transfers Intergoverimant 106 Table A.2 8RAZILs FEDERAL CONDITIONAL TRANSFERS TO STATZS AND MUNICIPALITIES BY FUNCTION (1983-1987) (percent) %griculture Regional Development Education & Culture Energy Houslng & Urban Development ------------------------------.------------------------------ 983 984 985 986 987 4.5 4.7 1.4 4.8 5.3 Health & Sanitation Transportation Other Functions Total 24.9 29.7 35.6 27.3 45.0 100.0 100.0 100.0 100.0 100.0 __--------------__---------------------- 34.4 22.8 12.8 16.5 1.8 15.8 18.4 26.0 19.0 22.4 6.0 9.2 7.1 4.5 2.3 1.7 1.2 0.7 14.7 6.5 12.8 14.0 16.4 13.1 16.2 0.0 0.0 0.0 0.1 0.6 ource: Sac. Progr. Financeira/STN e "Balancos Gerais da Uniaom, MINIFAZ eflatort IGP-DI medio (FGV) cher Transfera - Intergovernmental Transfers minus Federal Tax Transfers ther Functions - Legislative, Judiciary, Planning and Administration, Defense, Commerce and Industry. 107 Table A.3 DiSTRIBUTION OF FEDERAL TRANSFERS THROUGHTEZ COIVENIOS TO STATES AND HUNICIPALITIES BY THE GRANTINOFEDERAL KINISTRY 1985-86 (percent) Regions/ States Planning Acre Amazonas Para Rondonia Amapa Roraima Maranhao Piaui Ceara R.Gde.Norte Paraiba Pernambuco Al.agoas Sergipe Bahia Mato Grosso M4. Grosso Sul Goias Distrito Fedl. Hinas Gerais Espirito Santo Rio Janeiro Sao Paulo Parana Sta. Catarina Rio Gde. Sul 43.5 52.4 32.2 13.2 53.2 44.8 45.3 41.4 21.4 27.9 31.5 12.5 31.0 39.0 20.5 54.7 38.4 24.8 92.1 2.8 42.2 1.8 7.9 1.5 13.9 2.7 Education Cash In-kind Total 7.4 18.7 28.5 2.6 20.0 18.6 24.9 27.1 33.9 29.4 29.3 31.6 26.4 23.3 30.7 14.7 20.4 30.9 2.9 37.4 20.2 24.9 26.4 34.6 37.6 40.7 4.0 11.0 17.2 1.2 17.4 10.0 14.7 12.2 23.2 14.4 16.0 16.2 16.5 13.9 20.3 7.6 14.9 13.9 1.0 27.5 13.0 3.5 7.0 25.4 18.3 20.6 Health Cash In-kLad 2.7 4.4 3.6 0.5 0.0 4.9 2.2 6.0 4.5 11.5 5.0 6.9 8.2 7.7 5.7 3.4 4.1 4.2 0.2 5.5 3.7 2.7 9.4 2.8 3.2 4.2 11.5 29.7 45.6 3.8 37.4 27.9 39.6 39.3 57.0 43.8 45.3 47.8 42.9 37.2 50.9 22.4 35.3 44.8 3.9 64.9 33.2 28.4 33.3 60.0 56.0 61.3 Urban Finance Others Total Total Develoument 0.8 4.4 6.1 0.4 3.5 0.9 2.9 3.1 6.9 6.1 4.6 6.2 .7 4.1 6.8 1.5 3.6 6.0 2.3 3.8 2.8 6.8 5.4 4.8 4.0 4.3 3.4 8.8 9.7 0.9 3.5 5.8 5.2 9.1 11.4 17.6 9.6 13.1 14.9 11.9 12.5 4.9 7.7 10.2 2.5 9.3 6.5 9.5 14.8 7.6 7.1 8.5 0.7 8.5 5.1 0.6 4.7 2.5 4.3 6.9 4.8 6.8 10.4 22.5 8.2 9.2 9.5 8.7 11.3 11.4 1.3 11.5 13.4 23.1 33.1 14.4 14.7 16.2 38.4 0.0 0.0 78.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.1 0.0 0.0 0.0 6.0 0.0 33.2 0.0 0.0 0.0 0.0 2.5 0.6 7.4 2.7 1.2 18.9 5.7 3.3 5.3 4.0 3.3 4.2 3.0 2.6 6.6 4.2 7.2 8.8 0.27 11.4 4.7 4.1 10.8 16.4 8.8 11.3 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 _---------------------------__--____---_-----__---_-_----------------- -------------------------------- NORTH NORTHEAST CENTRE-WEST SOUTHEAST SOUTH 25.7 29.1 74.6 9.9 5.2 10.0 28.7 9.2 29.4 37.4 5.8 16.6 4.7 15.0 21.9 15.8 45.3 13.9 44.4 59.3 1.8 5.7 1.4 5.3 3.4 2.0 5.1 2.8 4.6 4.4 3.7 10.8 4.2 9.9 7.8 2.6 10.2 4.4 19.4 15.12 48.6 0.0 0.6 8.0 0.0 3.6 4.6 2.3 8.4 12.5 100.0 100.0 100.0 100.0 100.0 BRAZIL 34.9 21.2 12.0 33.1 3.6 3.8 7.4 9.3 10.2 5.1 100.0 Source: Silvia Paiva, STN/IfNIFAZ, Results from a special survey of government agencies. 108 Table A4 OVERALLIMPACTOF THE NEW FEDERALISM ON STATE ANDMUNICIPALFINANCES: States of Para and Parana Fiscal Impact on Available Revenues - 1993 Para State 17.70% Municipalities 31.20% Parana State 17.35% Municipalities 35.65% Sources:- Secretaria de Estado da Fazenda do Parana - Coordenacao de Assuntos Econ6micos.- Barrato, G. (1989) Impactoda Reforma Tributaria na Receita Disponiveldo Parana. (manuscript) - CondurCiJr., R. P. (1988). 0 Ganhos Tributarios do Estado e MunicipiosParaenses.Para DesenvolvimentoNo. 24 () Average Impact on all Brazilian municipalitiesas reported by Condur(u(1988) Table A.S RECEIVEDBY BRAZILIAN MUNICIPALITIES 1987 - FEDERALTAX TRANSFERS (Values in CurrentNCz3 Thousands) Federal Unit (FPM) TRANSFERS CURRENT X of total value per capita (FPM) CAPITAL TRANSFERS value per capita X of total 4.80% 0.44% 7.89% 671 522 763 237 2,756 331 5,270 0.68 1.35 0.41 2.16 0.60 1.42 0.65 0.75% 0.68% 0.84% 0.26% 3.07% 0.37% 5.86% 0.23 0.36 0.09 0.24 0.26 0.13 0.21 0.06 0.23 0.20 5.47% 4.37% 2.77% 2.65% 4.04% 4.45% 2.40X 0.43% 12.57% 39.16% 3,354 2,260 4,455 2,547 3,187 4,568 2,038 1,435 7,951 31,793 0.89 0.90 0.73 1.18 1.03 0.65 0.89 1.07 0.72 0.78 3.73% 2.61% 4.96% 2.83% 3.S5% 5.08% 2.27% 1.80% 8.85% 35.38% 4,840 868 520 1,377 7,202 0.31 0.28 0.04 0.04 0.12 22.91% 3.28% 2.57% 8.80% 35.56X 12,567 1,677 2,798 12,863 29,905 0.83 0.70 0.21 0.42 0.48 13.98% 1.87% 3.11% 14.31% 33.28% 1,071 485 734 2,270 0.13 0.11 0.08 0.11 5.29% 2.30% 3.62% 11.21% 6,466 3,726 5,783 15,975 0.76 0.88 O.88 0.74 7.20% 4.15% 6.44% 17.78% 0.09 0.28 0.18 0.76% 2.0e% 3.57% 1,300 1,372 4,251 0.78 0.87 0.92 1.45% 1.53% 4.73% 0.14 6.40% 6,923 0.72 7.70% 100.00% 89,886 608 413 533 234 1,784 241 3,713 0.62 1.07 0.29 2.14 0.39 1.04 0.48 0.73% 0.59% 0.77% 0.34% 2.58% 0.36% 5.33% 163 109 220 3 972 90 1,567 0.17 0.28 0.12 0.03 0.21 0.39 0.19 0.80% 0.54% 1.09% Sergipe Bahia NORTHEAST 2,248 1,374 3,893 2,011 2,369 3,864 1,551 1,348 5,405 23,861 0.48 0.65 0.64 0.92 0.76 0.52 0.67 1.00 0.49 0.59 3.23% 1.97% 5.59% 2.89% 3.40% 5.26% 2.23% 1.94% 7.76% 34.28% 1,108 888 562 538 818 902 487 87 2,546 7,932 Minas Cerais EspiritoSanto Rio de Janeiro Sao Paulo SOUTHEAST 7,927 1,012 2,278 11,486 22,703 0.52 0.42 0.17 0.37 0.37 11.39% 1.45% 3.27% 16.5ox 32.62% Parana Santa Catarina R Grand.Sul SOUTH 5,395 3,281 5,049 13,705 0.63 0.77 0.68 0.64 7.76% 4.68% 7.25% 19.69% 1,148 954 3,527 0.68 0.60 0.78 1.85% 1.37% 5.07% 5,627 0.59 8.08% 1,298 100.00% 20,257 Rondonia Acre Amazonas Roraima Para Amapa NORTH Maranhao Piai Ceara' R GrandeNorte Paraiba Pernambuco Alagoas M GrossoSul Mato Crosso Goias DistritoFederal CENTERWEST 89,609 BRAZIL Source: MINIFAZ/SEF 0.49 (FPM) TAX TRANSFERS TOTALFEDERAL value per capita X of total 154 418 724 0.14 0.01% 0.64 100.00% 1987 - Table A.6 NON-TAXTRANSFERSRECEIVEDBY BRAZILIANMUNICIPALITIES (Values in Current NCz3 Thousands) Federal Unit CURRENTST. A FEDNON-TAXTRANSFERS CAP. ST. A FED NON-TAXTRANSFERS TOTAL ST. A FEDNON-TAXTRANSFERS value per capita % of total value per capita X of total value per capita X of total 339 42 127 20 411 26 965 0.36 0.11 0.07 0.18 0.09 0.11 0.12 1.56% 0.19% 0.68% 0.09% 1.87X 0.12X 4.40% 60 204 1,272 33 451 23 2,043 0.06 0.63 0.69 0.30 0.10 0.10 0.26 0.32X 1.09X 6.77% 0.18X 2.40% 0.12X 10.88x 399 248 1,399 53 862 49 3,008 0.41 0.64 0.76 0.48 0.19 0.21 0.37 0.98% 0.e0% 3.44% 0.13% 2.12% 0.12X 7.39X 382 221 803 278 216 790 198 263 712 3,883 0.08 0.09 0.13 0.13 0.07 0.11 0.09 0.20 0.06 0.10 1.74X 1.01% 3.8ex 1.27X 0.28 0.34 0.09 0.06 0.09 0.10 0.04 0.41 0.10 0.14 7.34X 4.68X 2.99X 0.71X 1.42X 3.67% 0.51% 2.96% 6.86% 29.71X 1,761 1,078 1,364 411 482 1,481 294 818 1,774 9,441 0.38 0.43 0.22 0.19 0.16 0.21 0.13 0.61 0.16 0.23 4.33% 2.65X 3.35X 3.8oX 0.90% 1.20% 3.25% 17.62X 1,379 867 681 133 286 671 98 653 1,062 5,678 1.18% 3.69x 0.72X 2.00% 4.36X 23.20% 1,328 219 EspiritoSan Rio do Janei 1,631 7,720 Sao Paulo 10,898 SOUTHEAST 0.09 0.09 0.12 0.26 0.18 6.06% 1.00% 7.44X 36.21% 49.71% 2,094 236 242 2,433 5,004 0.14 0.10 0.02 0.08 0.08 11.15X 1.26X 1.29% 12.9ex 28.86X 3,422 464 1,873 10,163 15,902 0.23 0.19 0.14 0.33 0.26 8.41X 1.12X 4.60X 24.96x 39.07% Parana Santa Catari R GrandeSul SOUTH 2,068 1,100 1,497 4,863 0.24 0.28 0.17 0.22 9.38% 1,270 3.62% 16.9OX 3,328 2,326 2,176 7,827 0.39 0.65 0.26 0.36 8.17X 1,226 679 3,174 0.15 0.29 0.08 0.16 8.78X 6.02% 6.83X 19.23X M Grosso Sul 388 473 705 0.22 0.30 0.16 1.68X 2.18X 3.22X 464 1,261 1,272 0.27 0.79 0.27 2.42% 6.88x 6.77X 822 1,724 1,977 0.49 1.09 0.43 2.02X 4.24% 4.88X 1,648 0.18 7.06X 2,977 0.31 15.88X 4,623 0.47 11.11X 21,926 0.15 100.00X 18,776 0.13 100.00% 40,701 0.29 100.00% Rondonia Acre Amazonas Roraima Para Amapa NORTH Maranhao Piaui Ceara' R Grande Nor Paraiba Pernambuco Alagoas Sergipe Bahia NORTHEAST UinasCerais Mato Grosso Goias DistritoFed CENTER WEST BRAZIL Source: MINIFAZ/SEF 0.99X 21.22X e.62% 1.01% 6.71X S.36X Table A.7 TRANSFERSRECEIVEDBY BRAZILIANMUNICIPALITIES 1987 - SUMMARYOF INTERGOVERNMENTAL (Valuesin CurrentNCzr Thousands) Federal Unit TOTAL FED AND ST. TAX TRANSFERS value per capita X of total OTHER ST. A FED TRANSFERS value per capita X of total TOTAL FED. AND STATE TRANSFERS value per capita % of total 948 679 1,949 288 3,641 379 7,884 0.97 1.60 1.08 2.45 0.77 1.63 0.94 0.4S% 0.27% 0.92% 0.13% 1.67% 0.18K 3.80% 399 246 1,399 63 862 49 3,008 0.41 0.64 0.76 0.48 0.19 0.21 0.37 0.98X 0.80X 3.44% 0.13X 2.12X 0.12X 7.39X 1,347 825 3,348 321 4,403 428 10,672 1.37 2.14 1.82 2.93 0.96 1.84 1.31 0.53X 0.33% 1.32% 0.13% 1.74% 0.17% 4.21% Maranhao Piaui Ceara' R Grand. Norte Paraiba Pernambuco Alagoas Sorgipe Bahia NORTHEAST 3,688 2,676 8,251 3,115 3,897 7,757 2,984 1,957 13,170 46,476 0.76 1.06 1.02 1.42 1.26 1.11 1.30 1.46 1.19 1.12 1.73% 1.26% 2.94% 1.46% 1.83X 3.86% 1.40% 0.92X 6.19% 21.39% 1,761 1,078 1,364 411 482 1,481 294 816 1,774 9,441 0.38 0.43 0.22 0.19 0.16 0.21 0.13 0.61 0.16 0.23 4.33X 2.66X 3.36X 1.01% 1.18X 3.S9X 0.72% 2.00% 4.38% 23.20X 5,429 3,754 7,615 3,626 4,379 9,218 3,278 2,773 14,944 64,916 1.12 1.49 1.24 1.61 1.41 1.32 1.42 2.06 1.36 1.36 2.14% 1.48X 3.01X 1.39K 1.73X 3.84X 1.29% 1.09% 6.90X 21.68X Minas Coraia EspiritoSanto Rio do Janeiro Sao Paulo SOUTHEAST 24,877 3,693 14,823 64,064 107,447 1.65 1.66 1.12 2.07 1.74 11.70% 1.74K 6.97% 30.12K 60.63% 3,422 464 1,873 10,153 16,902 0.23 0.19 0.14 0.33 0.26 8.41% 1.12% 4.60% 24.95X 39.07K 28,299 4,147 16,696 74,207 123,349 1.87 1.74 1.26 2.40 2.00 11.17X 1.64K 6.69% 29.29% 48.69X Parana Santa Catarina R GrandeSul SOUTH 14,382 8,586 16,683 38,661 1.69 2.03 1.79 1.80 8.76K 4.04X 7.38X 18.18% 3,328 2,326 2,176 7,827 0.39 0.66 0.26 0.38 8.17K 6.71% 6.36% 19.23X 17,708 10,911 17,869 46,478 2.08 2.68 2.04 2.18 6.99x 4.31X 7.06X 18.35X J,184 2,986 7,228 1.90 1.89 1.56 1.60% 1.40% 3.40% 822 1,724 1,977 0.49 1.09 0.43 2.02% 4.24% 4.88K 4,006 4,710 9,206 2.39 2.98 1.98 1.68X 1.86X 3.63% 13,398 1.40 6.30% 4,523 0.47 11.11% 17,921 1.87 7.07K 212,636 1.60 100.00% 40,701 0.29 100.00% 263,336 1.79 100.00X Rondonia Acre Amazonas Roraima Para Aapa NORTH M CrossoSul Mato Grosso Goiar DintritoFederal CENTERWEST BRAZIL Source: MINIFAZ/SEF 112 REFERENCES Afonso, J.R.R. (1989, May), "Despesas Federais com Transferencias Inter-governmentais: Uma Revisao de Concentos, Estatisticas e Diagnostico." INPES - Relatorio Interno #12. 102.103. Afonso, J.R.R. (1989, October). "MecanismosInstitucionais das Relacoes Intergovernamentais no Brasil." mimeo. 21-A - 21-H. Bird, Richard (1986). Federal Finances in Comparative Perspective. Canadian Tax Foundation, Toronto. Boadway,Robin (1980). Intergovernmental FiscalTransfers in Canada. Canadian Tax Foundation, Toronto, Canada. Boadway, Robin (1990). The Reform of Federal-State Transfers in Mexico. Processed. Bomfim,Antulio and AnwarShah (1990). "MacroeconomicManagementand the Divisionof Powers in Brazil: Perspectives for the Nineties". Intergovernmental Fiscal Relations in Developing Countries Project Discussion Paper No. 4, Public Economics Division, The World Bank, September. Camara dos Deputados (1989, November). "Projeto de Lei Complementar No. 104-A,de 1989." Brasilia, DF: Centro Grafico do Senado Federal. Camara dos Deputados (1989, November). "Projeto de Lei Complementar No. 177-A,de 1989." Brasilia, DF: Centro Grafico do Senado Federal. Castro, Firmo de (1989). "Justificacaoao Projeto de Lei Complementar No. 104-A/89da Camara dos Deputados." Brasilia, DF: Centro Grafico do Senado Federal. Gramlich, E. (1977). Intergovernmental Grants: A Review of the Empirical Liteirature. In W.E. Pates (ed.) The Political Economy of Fiscal Federalism. Lexington, D.C. Heath, Mass. USA. pp. 219-239. Hunter, J.S.H. (1977). Federalism and Fiscal Balance. ANU, Canberra, Australia. Mahar, Dennis and Dillinger,W.R. (1983). Financing State and Local Govermnents in Brazil. The World Bank Staff Working Paper No. 612. McLure, C.E. (Ed.) (1983). Tax Assignmentin Federal Countries. ANU Press, Canberra, Australia. McMillan, M., Anwar Shah and David Gillen (1980). The Impact of Provincial-Municipal Transportation Subsidies. A report prepared for the Alberta Transportation Minister. August 1980. Oates, Wallace (1972). Fiscal Federalism. Harcourt Brace Jovanovich, New York, USA. 113 Shah, Anwar (1979). Provincial Transportation Grants to Alberta Cities: Structure, Evaluation and a Proposal for An alternate Design. Alberta Transportation, Edmonton, Canada. An abridged version published in Roy C. Brown (editor), Ouantity and Ouality in Economic Research, Volume 1 (1985), New York and London: UniversityPress of America, pp. 59-108. Shah, Anwar (1981). The reform of Provincial-MunicipalTransfers in Alberta. A report prepared for tne Minister of Municipal Affairs, Alberta. Shah, Anwar (1983). Municipal Fiscal Needs and Provincial-M;inicipalUnconditional Transfers in Canada. DiscussionPaper, Finance Canada, Ottawa, March 1983. Shah, Anwar (1983). Efficiency, Equity and Fiscal Equalization Grants: Issues and Alternatives. Finance Canada DiscussionPaper, October 1983. Shah, Anwar (1983). The New Federalism in Australia. Finance Canada DiscussionPaper, February 1983. Shah, Anwar (1984). Alternative Approaches to The Measurement of Expenditure Needs of Canadian Provinces. Finance Canada DiscussionPaper, March 1984. Shah, Anwar (1984). An Empirical Analysisof Public Transit Subsidiesin Canada. Finance Canada DiscussionPaper, June 1984. Also published in Roy Brown (ed.). Quantity and Quality in Economic Analysis. Vol. 11(1986), UniversityPress of America. Shah, Anwar (1984). A Preliminary Analysis of Expenditure Need Compensation in Fiscal Equalization. Finance Canada DiscussionPaper, November 1984. Shah, Anwar (1988). An Overviewof Fiscal Decentralization Issues in Sri Lanka. The World Bank, processed. Shah, Anwar (1988). Capitalization and The Theory of Local Public Finance: An Interpretive Essay. Journal of Economic Surveys,Vol. 2, No. 3, pp.209-243. Shah, Anwar (1989). A Linear Expenditure System Estimation of Local Fiscal Response to ProvincialTransportation Grants. Kentucky,Journal of Economicsand Business,pp.150-168. Shah, Anwar (1989). A Capitalization Approach to Fiscal Incidence at the Local Level. Land Economics, Vol. 65, No. 4, November 1989,pp. 359-375. Shah, Anwar (1990). Intergovernmental Fiscal Relations in Sri Lanka. Paper presented at the 46th Congress of the International Institute of Public Finance, Brussels, August 27-30, 1990. Senado federal (1989,July). "Projeto de Lei do Senado No. 165,de 1989- Complementar." Brasilia, DF: Centro Grafico do Senado Federal. Senado Federal (1989, March). "Projeto de Lei do Senado No. 45, de 1989." Brasilia, DF: Centro Grafico do Senado Federal. 114 Senado Federal (1989, September). "RedacaoFinal do Projeto de Lei do Senado, No. 45, de 1989." mimeo. Tribunal de Contas da Uniao (1988). "Relatorio e Parecer sobre as Contas do Governo da Republica." Brasilia, DF: Centro Grafico do Senado Federal. 69-70. The World Bank (1986). Brazil: Finance of Primary Education. A World Bank Country Study. PREWorkingPaperSeries Contact A nh eac DAfor SilviaSagari and WPS540 VentureCapitalOperations Rolein LDCMarkets GabrielaGuidotti TheirPotential 1990 November WPS541 PricingAveragePriceOptionsforthe StijnClaessens SwedervanWijnbergen 1990MexicanandVenezuelan Clauses Recapture S. King-Watson 31047 Choe WPS542 TheMetalsPriceBoomof 1987-89: Boum-Jong and TheRoleof SupplyDisruptions StockChanges 1990 November S. Lipscomb 33718 DonaldB.Keesing Gone Assistance WPS543 Development ServicesHave AndrewSinger Wrong:WhySupport Failedto ExpandExports 1990 November S. Fallon 37947 CanExpand DonaldB. Keesing WPS544 HowSupportServices AndrewSinger Exports:New Manufactured Methodsof Assistance 1990 November S. Fallon 37947 What WPS545 HealthandDevelopment: Contribute? CanResearch NancyBirdsall 1990 November L. Mitchell 38589 Grov h to Export-Led WPS546 TheTransition in SouthKorea,1954-66 Haggard Stephan Kim Byung-Kook Moon Chung-in 1990 November E. Khine 39361 1990 November M.Hileman 31284 Matter?An AndreaBoltho WPS547 DoesHighTechnology Applicatbnto UnitedStatesRegionalRobertKing Growth in Developing WPS548 DepositInsurance Countries SamuelH.Talley IgnacioMas 1990 November M.Pomeroy 37666 in a Substitution WPS549 Intertemporal Evidence Framework: Monetary fromChileandMexico PatricioArrau 1990 December S. King-Watson 31047 to RelativePrice JohnL Newman WPS550 Firms'Responses VictorLavy in C6ted'lvoire:The Changes for ExportSubsidies RaoulSabmon Implications PhilippedeVreyer andDevaluations 1990 December A. Murphy 33750 Experience GaryBanks Antidumping WPS551 Australia's 1990 December N. Artis 37947 NancyGillespie WPS552 SelectedWorldBankPoverty of Approaches, Studies:A Summary andFindings Coverage, 1990 December M.Abiera 31262 WPS553 Money,Inflation,andDeficitin Egypt MarcebGiugale HinhT.Dinh 1990 December V. Israel 36097 PRE WorkingPaperSeries Contact DAblhor D for papr December1990 M. Schreier 36432 of PriceReform SimonCommander WPS555 The Macroeconomics FabrizioCoricelli in SocialistCountries:A Dynamic Framework December1990 0. del Cid 39050 WPS556 TaxingChoicesin DeficitReduction John Baffes AnwarShah December1990 A. Bhalla 37699 WPS557 The NewFiscaiFederalismin Brazil AnwarShah December1990 A. Bhalla 37699 WP5558 AlternativeInstrumentsfor Smoothingthe Consumptionof PrimaryCommodityExporters KennethM. Kletzer DavidM. Newbkry BrianD. Wright December1990 J. Carroll 33715 WPS554 Korea'sLabor MarketsUnder StructuralAdjustment DipakMazumdar