Report No. 30035-VN Vietnam Managing Public Expenditure for Poverty Reduction and Growth Public Expenditure Review and Integrated Fiduciary Assessment (In Two Volumes) Volume I: Cross Sectoral Issues April 28, 2005 Joint Document of the Socialist Republic of Vietnam and the World Bank Prepared with the Support of the Like Minded Donor Group Abbreviations ADB Asian Development Bank MOLISA Ministry of Labor, Invalids and Social AFD Agence Francaise de Developpement Affairs AFTA Asean Free Trade Area MPI Ministry of Planningand Investment BEC BidEvaluation Committees MPDF Mekong Project Development Facility CFAAs Country Financial Accountability MPI Ministry of Planning and Investment Assessment MOT Ministry of Transport CIDA Canadian International Development MTEF Medium-TermExpenditure Framework Agency MTFF Medium-Term fiscal Framework CIEM Central Institute of Economic NGO Non-Governmental Organization Management NMDC Netherlands Ministry Development CMS Cooperative Medical System Corporation CPARS Country Procurement Assessment NORAD NorwegianRoyal Agency Reports NPL Non-PerformingLoans CPRGS ComprehensivePoverty Reduction and NTP National Target Programs Growth Strategy ODA Official Development Assistant DAF Development Assistance Fund OECD Organization for Economic DANIDA DanishInternational Development Cooperation and Development Agency O&M Operation and Maintenance DFID Department of International PAR Public Administration Reform Development PER-FA Public Expenditure Review and DMFAS DebtManagement and Financial Integrated Fiduciary Assessment Analysis System PFMRP Public Finance Management Reform DPP Department for Public Procurement Project EC European Commission PIP Public Investment Program ECORY Research and Consulting P M U Project Management Unit EVN Electricity of Vietnam PRS Poverty Reduction Support Credit FDI Foreign Direct Investment SAV State Audit of Vietnam FDIC FederalDepositInsurance Corporation SBD Standard BiddingDocument GDP Gross Domestic Product SDC Swiss Development Cooperation GFS Govemment Financial Statistics SDU Service Delivery Units GFSM Government Finance Statistics Manual SEDP Socio-Economic Development Plan GI0 Government Inspection Office SIDA Swedish International Development GS0 General Statistics Office Cooperation Agency HCMC Ho Chi Minh City S I 0 State Inspection Office HCFPs Health Care Funds for the Poor SOCBs State-ownedCommercialBanks HEPR-JC Hunger Eradication, Poverty Reduction SOEs State Owned Enterprises and Job Creation TABMIS Treasury and Budget Management HIV/AIDS Human Immunodeficiency Information System VirusIAcquired Immunodeficiency UCOA United Charts of Accountants Syndrome UNCTAD United Nations Commission on Trade IMF International Monetary Fund and Development IPSAS International Public Sector Accounting UNDP United Nations Development Program Standards USBTA United States Bilateral Trade JBIC Japan Bank for International Agreement Cooperation VAT Valued Added Tax KFW Kreditanstalt fur Wiederaufbau VDGs Vietnam Development Goals L A N S Local-AreaNetworks VDR Vietnam Development Report LMDG Like Minded Group of Donors VEA Vietnam Economics Association MARD Ministry of Agriculture and Rural VHLSS Vietnam HouseholdLiving Standards Development Survey MIS Management Information System VND Vietnam Dong MOET Ministry of Education and Training vss Vietnam Social Security MOF Ministry of Finance WANS Wide-area Networks M O H A Ministry of Home Affairs M O H Ministry of Health ACKNOWLEDGMENTS This Public Expenditure Review and IntegratedFiduciary Assessment (PER-FA) was commissioned by the Prime Minister of Vietnam, H.E.Mr.Phan Van Khai and was prepared jointly by the Government of Vietnam and the World Bank with the support of the Like MindedGroup of Donors (LMDG). On the Government side, overall guidance was provided by Mr. Tran Van Ta (Standing Vice Minister of Finance), Mme L e Thi Bang Tam (Vice Minister of Finance) and Mr Truong Van Doan (Vice Minister of Planning and Investment). On the World Bank side, overall guidance was provided by Mr. Homi Kharas (East Asia Chief Economist) and Mr. Klaus Rohland (Vietnam Country Director). Additional guidance was provided by Ms. Marisa Fernandez-Palacios (East Asia Operational Services Director), Ms. Barbara Nunberg (East Asia Public Sector Manager), Ms. Helen Sutch (Acting East Asia Public Sector Manager), Mr.Martin Rama (Vietnam Lead Economist) and Mr.Rakesh Nangia (Vietnam Portfolio Manager). Valuable additional guidance was provided by Dr. Graham Scott (Chief Technical Adviser, Vietnam Ministry of Finance and former Secretary of the New Zealand Treasury). The task managers and principal editors of the report were Mr Do Hoang Anh Tuan (Director of the State Budget, Ministry of Finance), Mr.PhamDinh Cuong (Deputy Director, State Budget Department, Ministry of Finance), Ms. Tran Kim Wen (Manager, State Budget Department, Ministry of Finance), Mr. Edward Mountfield (Senior Economist, World Bank) and Mr.Nguyen Van Minh (Senior Economist, World Bank). Key contributors to specific chapters are listed below: 0 FiscalTrends and Sustainability Mr.Do Viet Duc -Manager of State Budget Department, Ministry ofFinance(MOF) Mr.Vivek Suri-Senior Economist, the World Bank Mr.NguyenMinhTan-Instituteof Finance Research, Vietnam Mr.Albert de Groot-ECORYS ResearchandConsulting, the Netherlands Trends inCompositionof PublicExpenditure Mr.DaoXuan Tue-Managerof State BudgetDepartment, MOF Mr.Le Quang Thuan-Instituteof Finance Research, Vietnam Mr.Albert de Groot-ECORYS Research andConsulting, the Netherlands Institutionsfor PublicExpenditureManagement Mr.PhamDinhCuong-DeputyDirector, State BudgetDepartment, (MOF) Ms.RosaAlonso ITerme -Senior Public Sector Specialist, the WorldBank Mr.EdMountfield-Senior Economist, the World Bank Mr.NguyenVan Minh-Senior Economist, the WorldBank Institutionsfor FinancialAccountabilityand Transparency Mr.BehdadNowroozi-Senior FinancialManagement Specialist, the WorldBank Mr.Tran QuangThong -Financial ManagementSpecialist, the WorldBank Mr.Le Viet Hung-FinancialManagementSpecialist, theWorld Bank 1 Mr.DangQuang Huy-Expert, EnterpriseFinanceDepartment, MOF 0 Decentralization to Sub-national Government Professor Jorge Martinez - Georgia State University (USA) Ms.Tran Thi LeTrinh - Deputy Director, HanoiFinancialDepartment, Vietnam Mr.NguyenDoanToan-Manager of BudgetDivision, HanoiFinancialDepartment Mr.BuiDuongNghieu-Instituteof FinanceResearch,Vietnam 0 Delegationto SpendingUnits Professor Christine Wong - University of Washington, USA Ms.DoThiThuy Hang- Deputy Director of Public ExpenditureManagement, MOF Ms.NguyenThi MinhChau- Expert of Public ExpenditureManagement, MOF Mr.SorenDavidsen-Governance Specialist, the WorldBank Ms.HoangThi Thuy Nguyet -Institute of FinanceResearch, Vietnam 0 Public Investment Management Dr.Cao Viet Sinh,Director, Ministry of Planning andInvestment(MPI) Mr.LeDinhQuy - Deputy Director of Investment Department, MOF Ms.DoThiBich-Expert, Investment Department, MOF Mr.StephenLister -consultant, MokoroConsulting, UK Mr.NguyenQuang Thai -consultant, VietnamEconomicsAssociation (VEA) 0 PublicProcurement Management Mr.NguyenViet Hung,Directorof Public ProcurementDepartment, MPI Mr.NguyenXuan Dao, Vice DirectorPublic Procurement Department, MPI Mr.BuiHa- Director of Finance andMonetaryDept, MPI Ms.HoangMinhThoa - DeputyDirector of Finance andMonetary Department, MPI Ms.Vu Quynh Le-Expert of Public Procurement Department, MPI Ms.Nguyen Thi PhuHa - Expert of Finance andMonetaryDepartment, MPI Ms.IrinaLuca-Senior Procurement Specialist, the World Bank Mr.JoelTurkewitz -Senior Procurement Specialist, the World Bank Mr.Giovanni Casartelli -LeadProcurement Specialist, the World Bank Mr.Ahsan Ali -Senior Procurement Specialist, the WorldBank Mr.NguyenChienThang -Senior Procurement Specialist, the WorldBank Ms.PhanThi CUC-Consultant, Academy of Finance, Vietnam Mr.DinhViet Ninh-Consultant, Electricity of Vietnam (EVN) 0 Education Sector Expenditure Mr.Nguyen Van Ngu- Deputy Director of Finance andPlanningDepartment, MOET Mr.BuiHongQuang- Senior Expert ofFinance andPlanningDept, MOET Ms.DangThi ThanhHuyen-Manager, EducationManagement Department, MOET Mr.PeterBrooke-Bannock Consulting, UK Mr.SamuelLiebennan -LeadEconomist, the WorldBank Ms.MaiThi Thanh -Senior EducationOperation Officer, the World Bank Transport Sector Expenditure Mr.HaKhac Hao-Deputy Director of PlanningDepartment, MOT 11 Ms. Nguyen Thanh Hang - Senior Expert, Planning & Investment Department, M O T Mr.Doan CanhHoang-Expert, FinancialDepartment, MOT Mr.Simon Ellis-Senior Transport Specialist, the WorldBank Ms.LeThuHang-Institute of Training andProfessional Enhancement, Vietnam Mr.PhamNgoc Thuy -HanoiPolytechnic University, Vietnam Mr.LeThan -consultant, Vietnam RoadSafety Project Mr.Jacques Yenny, independent consultant, USA 0 Health Sector Expenditure Mr.NguyenNamLien-Vice Director, Finance & PlanningDepartment, MOH Mr.LeVan Quan -Expert,PlanningandFinance, MOH Mr.Samuel Lieberman-LeadEconomist, the WorldBank Mr.James Knowles -Independent Consultant, USA Ms.NguyenThi HongHa-Mediconsult, Vietnam Ms.NguyenKhanhPhuong-Instituteof HealthStrategy andPolicy Agriculture and Rural Development Sector Expenditure Mr.NguyenVan Than-Acting Director of Finance Department, MARD Ms.PhanNgoc Thuy -Vice Director ofFinanceDepartment, MARD Mr.PhamTrungKien- Assistant andSenior Expert of Finance Department, MARD Mr.NguyenThe Dzung-Agriculture & RuralDevelopment Specialist, WorldBank Mr.StephenMink-LeadEconomist, WorldBank Mr.William Cuddihy -Independent Consultant, Australia Ms.PhamThi LanHuong-Central Institute of Economic Management, Vietnam National Target ProgramExpenditure Mr.BuiHa- Director of Finance andMonetary Dept, MPI Ms. Hoang MinhThoa - Deputy Director of Finance and Monetary Department, MPI Ms.NguyenThi PhuHa - Expert ofFinance andMonetary Department, MPI Mr.Mark Minford-Bannock Consulting, UK Mr.Ngo Van Minh- DeputyDirector of Finance andPlanningDepartment, MOLISA Mr.TaVan Thieu - Manager, Finance andPlanningDepartment, MOLISA Mi.NguyenHuuTu-consultant, Instituteof FinanceResearch, Vietnam Mr.BuiQuoc Bao-consultant, Institute for Market andPriceResearch, Vietnam Peer reviewers were Mr. Bill Dorotinsky (Lead Public Sector Specialist, World Bank), Mr. David Shand (Financial Management Adviser, World Bank) and Ms. Jane Rintoul (Senior Governance Adviser, UK DFID). Other valuable contributions were provided by Ms. Keiko Sato, Ms. Carrie Turk, Mr. Rob Swinkels, Mr. Daniel Musson, Ms.Nga Nguyet Nguyen and Mr.Robin Mearns, all of the World Bank. Administrative support was provided by Ms. Gloria Elmore and M s . PhungThi Tuyet (World Bank). The LMDG comprises the Canadian International Development Agency (CIDA), the Danish International Development Agency (DANIDA), the Netherlands Ministry for Development Cooperation (NMDC), the Norwegian Agency for Development Cooperation (NORAD), the Swedish International Development Agency (SIDA), the Swiss Agency for Development and Cooperation (SDC) and the UK Department for International Development (DFID). Valuable contributions were also provided by other ... 111 official development assistance agencies including the Asian Development Bank (ADB), European Commission (EC), International Monetary Fund (IMF), Japan Bank for International Cooperation (JBIC) and UnitedNations Development Program (UNDP). iv Table of Contents ACKNOWLEDGMENTS ................................................................................................... i EXECUTIVE SUMMARY ............................................................................................... i x 1. 1 Background..................................................................................................................... BACKGROUND, APPROACH AND SCOPE.......................................................... 1 Approach......................................................................................................................... 1 Scope............................................................................................................................... 4 Recommendations on FuturePublic Expenditure Analysis............................................ 4 2. 13 Introductionand Overview ........................................................................................... FISCAL TRENDS AND SUSTAINABILITY......................................................... 13 Macroeconomic Performance....................................................................................... 13 Overview of Fiscal Performance .................................................................................. 15 RevenueTrends ............................................................................................................ 15 Overall expenditure trends............................................................................................ 17 18 Future evolutionof government debt............................................................................ Budget deficit, off-budget items and public debt.......................................................... 24 Strengthening Management of Public Debt andFiscal Risks....................................... 25 Fiscal Transparency ...................................................................................................... 26 3. Recommendations ......................................................................................................... 28 29 Introductionand Overview ........................................................................................... TRENDS INCOMPOSITION OF PUBLIC EXPENDITURE................................ 29 Economic Composition of Expenditure........................................................................ 29 Functional Composition of Expenditure....................................................................... 33 How Pro-Poor was the Composition of Public Spending?........................................... 35 Recommendations......................................................................................................... 39 4. INSTITUTIONS FOR PUBLIC EXPENDITURE MANAGEMENT..................... 47 Introductionand Overview ........................................................................................... 47 Legal and InstitutionalFoundations for Budgeting ...................................................... 47 Strengthening State Budget and InvestmentPlanning.................................................. 51 Strengthening Budget Execution And ManagementInformation ................................ 58 TRANSPARENCY........................................................................................................... 5. INSTITUTIONS FOR FINANCIALACCOUNTABILITY AND 63 Introduction and Overview ........................................................................................... 63 Accounting and Reporting Processes inthe Public Sector........................................... 63 Oversight Functions inthe Public Sector...................................................................... 66 Financial Reporting and Auditing in SOEs .................................................................. 70 Public Oversight............................................................................................................ 72 6. Action to Prevent Corruption........................................................................................ 73 DECENTRALIZATION TO SUBNATIONAL GOVERNMENT.......................... 75 Introductionand Overview ........................................................................................... Structure of Government And Assignment of Expenditure Responsibilities ...............75 75 RevenueAssignments................................................................................................... 81 87 Budgetary Autonomy .................................................................................................... The System of Transfers ............................................................................................... 90 Transparency and Financial Accountability At Subnational Level .............................. 93 Subnational Government Borrowing............................................................................ 94 V 7. 97 Introductionand Overview ........................................................................................... DELEGATION TO SPENDING UNITS ................................................................. 97 Overview of Delegation Measures inVietnam............................................................. 97 Implementation of Delegationto Date........................................................................ 101 Initial Impacts and Risksinthe Implementation of Decision 192.............................. 103 8.Initial Impacts and Risks inthe Implementation of Decree 10................................... 107 PUBLIC INVESTMENT MANAGEMENT .......................................................... 115 Introductionand Overview ......................................................................................... 115 Reviewo f InvestmentPlanninginVietnam ............................................................... 115 Key Issues................................................................................................................... 128 Summary of Recommendations.................................................................................. 135 9. PUBLIC PROCUREMENT MANAGEMENT.,.................................................... 137 Introductionand Overview ......................................................................................... 137 Overview o f ProcurementReforms ............................................................................ 138 Legal and regulatory framework For Procurement..................................................... 140 Transparency and FairnessinProcurement................................................................ 141 Procurement Methods and BiddingDocuments ......................................................... 144 Evaluation and Award................................................................................................. 146 Contract Implementation ............................................................................................ 149 InstitutionalFramework.............................................................................................. 150 The Consulting sector ................................................................................................. 154 Procurement environment ........................................................................................... 155 Tackling Corruption inProcurement.......................................................................... 156 Summary o f Key Recommendations.......................................................................... 157 APPENDICES Appendix 1.1ProgressinImplementation of Recommendations ..................................... 6 Appendix 3.1 Composition of State Budget Expenditure. Final Accounts. 1997.............41 Appendix 3.2 Composition of StateBudget Expenditure.Final Accounts. 1998 .............42 Appendix 3.3 Composition o f State Budget Expenditure. FinalAccounts. 1999.............43 Appendix 3.4 Composition of State Budget Expenditure. Final Accounts. 2000 .............44 Appendix 3.5 Composition of State Budget Expenditure. FinalAccounts. 2001 .............45 Appendix 3.6 Composition of State Budget Expenditure. Final Accounts. 2002 .............46 BOXES Box 2.1 Key Contents of Decision 192/2004/QD-TTG on Fiscal Transparency..............27 Box 3.1 Returnsto NewInvestmentversus Operation and Maintenance Spending: The Case o f Large-Scale Irrigation............................................................................... 31 Box 4.1 Objectives of a Medium-termExpenditure Framework ...................................... 53 Box 4.2 Preliminary Findings from the Education Sector MTEFpilot............................. 53 Box 4.3 Piloting of the CitizenReport Card Survey inFour Cities .................................. 57 Box 4.4 Key Characteristics of TABMIS .......................................................................... 64 Box 7.1 Australia's Running Cost Systemfor Administration Units................................ Box 5.1 Progress inImprovingFinancial Accountability and Transparency inVietnam.61 98 vi Box 7.2 Introduction of Cost-Recovery Pricingin Saint Paul's Hospital. Hanoi............109 Box 7.3 Doctors Invest inMedical EquipmentinBinDuong......................................... 110 Box 7.4 Two-tiered Education inHanoi.......................................................................... 110 Box 7.5 The Unravellingof Public Provision of Health Care inChina .......................... 112 Box 7.6 Decision 139/2000 on HealthCare Funds for the Poor ..................................... 114 Box 8.1 Relevant Provisions of Resolution N O 3 (Decentralization) ............................... 131 Box 9.1 Costs Associated with Registry Management:An Example from the World Bank ..................................................................................................................... 143 Box 9.2 Vietnam: Good Practices inProcurement.......................................................... 146 Box 9.3 Examples of specifications to Buy High Quality Computerswithout Indicating Trade Name or Country of Origin ....................................................................... 148 Box 9.4 PossibleFramework for MonitoringProgress inProcurement.......................... 151 Box 9.5 The Complaint System inthe Republic of Slovenia.......................................... 152 Box 9.6 Transitional Arrangements for Procurement...................................................... 155 FIGURES Figure 2.1 Trends inGovernment Revenue and Grants .................................................... 16 Figure 2.2 Share of DifferentTaxes inTotal Tax Revenue............................................... 17 Figure 2.3 Trends inGovernment Expenditures................................................................ 18 Figure 2.4 Budget Deficit .................................................................................................. 18 Figure 2.5 Government Debt from Budget and Off-Budget Items.................................... 19 Figure 2.6 On-Budget Debt and External Public Debt ...................................................... 19 Figure 2.7 Implicit Interest Cost of Debt........................................................................... 20 Figure 2.8 Interest Payments as Share of Current Expenditure......................................... 20 Figure 2.9 Commercial Bank Credit to SOEs.................................................................... 22 Figure 2.10 DAF On-Lending........................................................................................... -23 Figure 2.11A Base Case Scenario for Public Debt........................................................... 24 Figure 2.12 Alternative Debt Scenarios............................................................................. 24 Figure 3.1 Economic Classification of Public Expenditure, as share of Total State Budget Expenditure, 1997-2002............................................................................ 31 Figure 3.2 RecurrentExpenditure by Sector, 1997-2002.................................................. 32 Figure 3.3 Functional Composition of Budget Expenditure, 1997-2002........................... 34 Figure 3.4 Trends inKey Sectors, 1997-2002................................................................... 34 Figure 3.5 Budget Transfers and Poverty across Provinces in2002 ................................. 36 Figure 3.6 State Investmentand Poverty Rates across Provinces ..................................... 38 Figure 6.1 Central Government Share inBudget Expenditure, 1997-2002....................... 79 Figure 7.1 The Growing Role of Fees inHospitalFinance............................................. 108 Figure 8.2 State Investment 1995-2003 under Central and LocalManagement .............117 Figure 8.1 Composition of State Investment 1995-2003................................................. 117 TABLES Table 2.1 Overview of Fiscal Trends (1997-2003)............................................................ 15 Table 3.1 Targeting Effectiveness of HEPR in2002......................................................... 39 vii Table 6.1 Assignment o f Expenditure Responsibilities inVietnam .................................. 78 Table 6.2 Revenue Sources for Provincial Governments for 2002 ................................... 83 Table 6.3 Outcomes of State Budget Revenue and Expenditure Decentralizationin Vietnam.................................................................................................................. 85 Table 7.1 Implementation of Administrative UnitDelegation at Central Level .............101 Table 7.2 Implementation of Decision 192at Provincial Level...................................... 102 Table 7.3 Central MinistriedAgencies Implementing Decree 10 (2003) ........................ 103 Table 7.4 Provinces Implementing Decree 10 (December 2003) .................................... 104 108 Table 8.1 Composition of Total Investment2000-2003.................................................. Table 7.5 Sources of Revenue inSt.Paul's Hospital, Hanoi........................................... 116 Table 8.2 Overview ofthe PIP 2001-2005 List of Key Projects ..................................... 123 Table 8.3 PIP 2001-2005 Sector Breakdown of PlannedInvestment.............................. 125 Table 8.4 Public InvestmentCapital 2001-2005 by Regions .......................................... 125 Table 8.5 PIP 2001-2005 PlannedFinancing of Public Investment................................ 126 Table 8.6 Three-year Status of PIP 2001-2005 Implementation ..................................... 127 ... V l l l EXECUTIVESUMMARY 1. Public expenditure i s one of Government's most important tools in the struggle for poverty reduction and growth. This Public Expenditure Review and Integrated Fiduciary Assessment (PER-FA) was commissioned by the Prime Minister of Vietnam inDecember 2003 and conductedjointly by the Government of Vietnam, the World Bank and the Like Minded Group of Donors (LMDG) during 2004 and early 2005. Its aim i s to review and assess the contribution that public expenditure has made to poverty reduction and growth in Viktnam in recent years, and to identify priorities and actions for strengthening that contribution over the coming years through better resource allocation and better public expenditure management. 2. Fiscal sustainability. Strong fiscal management i s a pre-requisite for poverty reduction and growth. The Government of Vietnam has been acknowledged for its fiscal prudence, with relatively small budget deficits, a relatively small stock of debt (both domestic and foreign) and a sustainable share of public spending in GDP. Fiscal trends have been positive, both in terms of revenue collection and expenditure outturns, resulting in a sustainable fiscal balance. Nevertheless, a number of threats to fiscal sustainability exist and require attention over the coming years. The Government should restrict further off-budget bond issuance, channeling all Government borrowing through the budget. It should take urgent action to resolve the current expenditure arrears- largely associated with public investment in the transport and agriculture sectors-and prevent their further build up. It should strengthen arrangements for the monitoring and management of fiscal risk, starting with better recording of domestic debt and credit through the Development Assistance Fund (DAF). It should rapidly implement the plan to prepare a realistic and sustainable Medium-Term Fiscal Framework (MTFF) as part of every budget cycle. 3. Composition of public expenditure. Between 1998 and 2003, total government spending rose at the remarkable average annual rate of some 16 percent in nominal terms. This rapid growth presented the Government with an extraordinary opportunity to restructure the composition of its expenditures and better align them with its goals of poverty reduction and growth. There have been notable achievements. The shares in total capital spending and total aggregate expenditure of Education and Training and of Science and Technology have increased significantly over the review period. Successful efforts have been made to make the transfer of resources between provinces more pro- poor, resulting in an inter-provincial transfer and revenue sharing formula which indeed benefits poorer provinces. However, non-wage operations and maintenance expenditure fell to just 17 percent of total expenditures by 2002 while the share of capital expenditure (including major repairs) increased to over 40 percent. It i s necessary now to take urgent action to address an imminent maintenance crisis and strike a better balance between ix x VietnamPublic Expenditure Review and Integrated Fiduciary Assessment capital and recurrent spending. Recurrent expenditure should take an increased share of the budget, but with strengthened and integrated planning of public pay and employment. 4. Institutions for public expenditure management. Over the past decade, Vietnam has steadily strengthened its institutions for both budget planning and budget execution at every level of Government. In particular, the 2002 State Budget Law heralds several major reforms: clarifying powers and responsibilities, strengthening the management of decentralization, promoting administrative reforms and strengthening transparency and accountability in public finance. However, further reforms are required. Efforts are required to strengthen joint working between the Finance and Planning functions at every level. The Ministry of Planning and Investment (MPI) and the Ministry of Finance (MOF) should collaborate actively with sector ministries and provinces in the preparation of Medium-Term Expenditure Frameworks (MTEFs), building on the experience and success of sectoral pilots. M O F should play a more active role in the development of the new Socio-Economic Development Plan (SEDP). Implementation of the pilot MTEFs in four sectors and four provinces should be accelerated, building on the success in the education sector. Links need to be strengthened between performance indicators and budgetary decisions, with strengthened monitoring of service delivery using "citizen report cards" and other techniques. Budget execution and management information will be strengthened further by the successful implementation of the Government's new Treasury and Budget Management Information System (TABMIS). However, TABMIS should be implemented so as to ensure that MPI, sector ministries and provinces are given direct access to Treasury data on execution of sector spending at every level of government. 5. Financial accountability and transparency. Experiences around the world suggest that public expenditure i s unlikely to result in cost-effective service delivery without financial transparency and accountability. Vietnam has taken significant steps in improving institutions, laws and regulations and practices to ensure greater financial accountability and transparency in public expenditure. Oversight by the National Assembly and the provincial People's Councils over public finances has been increased. Public access to financial information continues to improve. Nevertheless, challenges remain. The Government should ensure independence for the State Audit of Vietnam (SAV) by converting it into an independent organization established by the National Assembly. Audit reports should be made available to the public. The Government should further rationalize the roles and responsibilities of the audit and inspection functions, Appropriate monitoring measures should be implemented to ensure that fiscal transparency and reporting regulations are properly implemented at all levels of government and by all spending units. The Government should complete its corruption diagnostic as soon as possible and ensure appropriate follow-up action. 6. Decentralization to sub-national Govemment. Over the past decade, Vietnam has embarked on an extensive decentralization program. The share of sub-national governments in total expenditure has increased dramatically, from 26 percent in 1992 to 48 percent in 2002. The 2002 State Budget Law represented in many ways a marked improvement over the 1996 Law in terms of the management of decentralization, further clarifying the roles and responsibilities of central and sub-national tiers of government, and giving new and real powers to the Provincial People's Councils. Nevertheless, the Executive Summary xi new powers at sub-national level create an urgent need to enhance administrative capacity and to strengthen transparency and accountability institutions. Furthermore, it will be important to monitor and review the new fiscal decentralization arrangements in order to identify areas requiring adjustment and further strengthening. The Government should monitor and evaluate the impacts on public service delivery of granting provinces discretion over the public finances of districts and communes; and consider in the longer term the case for specifying revenue and expenditure assignments for districts and communes. The Government should consider the introduction of a full formula-driven system of equalization grants; and the strengthening of the conditional grants system. Current borrowing rules for subnational government should be strengthened, with all bond issuances by provinces includedinthe State Budget. 7. Delegation to spending units. Increased delegation to spending units i s one of the key developments in public expenditure management in Vietnam in recent years. Such delegation i s proceeding on parallel but separate tracks for Administrative Units (under Decision 192/2001) and for Service Delivery Units (under Decree 10/2002). These reforms are rational responses to citizens' demand for more and better public services as their living standards rise; and there i s evidence to suggest that they may have promoted cost reduction and service innovation in certain spending units.However, such delegation needs to be managed carefully if it i s to result in better service delivery and if it i s to support rather than jeopardize equity and poverty reduction. The Government should maintain the pilot status of Decision 192 and use the "citizen report card" approach to help monitor and evaluate the impact on citizen satisfaction of block granting in administrative units. It should consider more tightly constraining the flexibility that managers have over salaries and remunerations. With Decree 10, the Government should develop further mechanisms to protect the access of the poor and near-poor to essential social services; and should conduct continuous review and assessment of the impact of Decree 10, again using the "citizen report card" approach. State Budget resources for Decree 10 agencies should increasingly be tied to the "purchase" of specific public goods and specific services for the poor-with mechanisms developed for monitoring and enforcing delivery against such contracts. 8. Public investment management. Public investment management i s a critical part of public expenditure management in every country. In Vietnam, it i s particularly critical to the extent that State Budget-financed investment has grown rapidly and now constitutes about 40 percent of the total State Budget. The capacity of the MPIto manage public investment has been strengthened over the past decade with the development of five year Public Investment Programs (PIPS),the third of which i s now underpreparation, In practice, however, the PIP i s only one element in the investment allocation system in Vietnam. Although often described as a compendium of projects, i t i s not a comprehensive screening and approval process even for large projects. Provinces are performing an increasingly important role in public expenditures. The Government should undertake a strategic review of the allocation of responsibilities for public investment between tiers of government, as part of the implementation of the new Government Resolution 08/CP. MPI should further strengthen its capacity to guide, train, support and supervise bodies with decentralized investment authority. An investment planning manual and standard project registers should be developed and rolled out to every level of government. Analysis of recurrent costs should be built in to investment xii Vietnam Public Expenditure Review and Integrated Fiduciary Assessment selection, and the MTEF approach should be used as a way of balancing and ensuring consistency between recurrent and capital costs. 9. Public procurement management. Sound public procurement i s a central pillar of good public expenditure management in a market-based economy. In Vietnam, the importance of procurement in public spending i s rapidly rising. The value of procurement financed through the State Budget more than doubled between 1999 and 2003 from US$2.0 billion to US$4.9 billion. A significant portion of public spending occurs through the public procurement process and even modest efficiency improvements in the workings of the system could potentially yield substantial savings-perhaps of the order of 1-2 percent of GDP. Over the past few years, progress has included the strengthening of the Department of Public Procurement (DPP) in MPI; and the important steps taken towards increasing transparency and standardization of public procurement, including the mandating of the creation of a Public Procurement Bulletin and of Standard Bidding Documents (SBDs) to guide procurement action. Nevertheless, important challenges remain. The Government should establish a mechanism for receiving and responding to procurement complaints. It should finalize the Procurement Ordinance/Law and submit it to the National Assembly. Procurement monitoring mechanisms should be established. An ongoing capacity building effort i s required, including capacity not only in management of initial procurement but also contract supervision and enforcement. 10. Education sector expenditure. Vietnam's rapid progress with poverty reduction and economic growth was underpinned by solid progress in promoting literacy and broader human development, both before and during the reform period. Duringthe PER- IFA review period, substantial progress has been made in increasing financing for the sector. Participation rates have increased and physical infrastructure has improved. Gaps in those participating and benefiting from education have narrowed. Nevertheless, important challenges are faced in improving the quality, equity and effectiveness of the service. Past and planned increases in education sector expenditure, combined with demographic changes in the school age population, have provided an opportunity for increased focus on service quality, equity and effectiveness. Rather than further improving Pupil-Teacher Ratios, the Government should use additional resources to prioritize the introduction of full day working more widely, reducing user charging and improving teacher qualifications and classroom infrastructure, particularly in the poorest localities. Mechanisms should be developed to ensure that at least a minimum level of budget funding reaches the frontline of service delivery for each student. The Government should consider linking improvements in teachers' pay to changes in practice and performance; and should give priority to the development of tests of attainment at key stages. 11. Transport sector expenditure. Transport infrastructure makes an important contribution to both poverty reduction and growth in Vietnam. Investment in rural roads has led to all but 220 communes being connected to all weather roads. Research has shown that these investments have had a significant impact on reducing rural poverty, and particularly in the poorest areas. The quality and extent of transport infrastructure have improved, the size of the vehicle fleets has increased and the efficiency and cost with which transport services are delivered has improved. Improved efficiency and lower cost provision of logistics services has been a key factor in facilitating Vietnam's export- Executive Summary xiii led growth. Nevertheless, serious concerns remain regarding inadequate maintenance expenditure in the sector. The Government should shift funds from new investment expenditure to operations and maintenance in the roads sector; and should take urgent action to strengthen commitment controls, address expenditure arrears, and ensure that investment in new roads does not reach unsustainable levels. It i s also clear that the transport sector needs to attract significant private sector financing to support public- financed expenditures. The Government should accelerate implementation of policies and measuresto attract private sector investment in transport. 12. Health sector expenditure. Vietnam's poverty reduction and growth has gone hand-in-hand with significant progress in health. Health indicators have continued to improve nationally since the early 1990s, while once high levels of fertility appear to have been brought under control. At the same time, some problems have persisted. There have been limited gains for some population segments, leading to growing inequalities. New health challenges have also appeared, including HIV/AIDS and more recently SARS and avian flu. The Government has taken steps to address some of these remaining challenges. Most dramatically, it has developed a national strategy for assisting the poor with healthcare costs (Decision 139), resulting in an increased level of health insurance coverage. The Government should encourage provinces, in implementing Decision 139, to purchase health insurance cards for the poor rather than reimburse healthcare providers directly. It should carefully monitor and evaluate implementation of Decision 139, including the methods used to identify the poor. It should proceed cautiously with implementation of Decree 10 in hospitals, with careful monitoring to ensure that public funds are not diverted to support the expansion of services for the rich. Strong and effective accreditation mechanisms should be developed; and all accredited hospitals, public or private, should be eligible to provide services to the insured, with public funding gradually shifting to the demand side for curative health care. 13. Agriculture and rural development sector expenditure. Agricultural transformation has been a vital component in Vietnam's pro-poor growth. The use of market mechanisms to provide incentives to farmers, beginning with the doi moi reforms of the 1980s, resulted in substantial agricultural sector growth. However, this growth was supported by public provision of basic infrastructure, by substantial progress in human development and by publicly financed research and extension services. In coming years, the efficient, effective and equitable provision and management of rural public services and infrastructure will play an even more important role in sustaining Vietnam's agricultural growth and rural development. To this end, the Government should re- prioritize and strengthen management of public expenditure within the agriculture and rural development sector. In particular, it should improve the balance between recurrent and capital costs, particularly in irrigation, with more attention given to catching up with deferred maintenance and system completion. It should increase the budget share of agricultural research and extension services; and reduce the burden of agricultural SOEs on public expenditure. As with transport, action i s needed to strengthen commitment controls and address the expenditure arrears that have built up in the sector. 14. National Target Program expenditure. In the early 1990s, Vietnam established a number of National Target Programs (NTPs) to support local governments and sector ministries to achieve national targets such as poverty reduction and economic and social xiv Vietnam Public Exuenditure Review and Intenrated Fiduciarv Assessment development. Since 2001, these programs, including the Hunger Eradication and Poverty Reduction and Job Creation Program (HEPR-JC), Program 135 (P135) and programs in the areas o f Health, Water, Education and Culture, have grown into an important instrument for economic development and poverty reduction. Spending on NTPs has grown rapidly and has contributed in important ways to the achievement of key development and poverty alleviation goals. The Government should continue to use NTPs as a complement to mainstream funding and-given Vietnam's rapid fiscal decentralization-should consider growing them as a share of public spending to enhance central Government's ability to ensure delivery of national targets. However, Government should rationalize NTPs and strengthen their coordination both with each other with main programs. It should strengthen mechanisms to ensure that NTPs deliver against national targets, including establishing clearer annual performance agreements based around specific and measurable outputs and outcomes. Fiduciary controls for NTPs,though not as weak as i s sometimes claimed, should be strengthened. There should be increased local consultation and participation in NTP management and planning. 15. Using public expenditure analysis. Periodic assessment of recent public expenditure and its management i s a crucial part of a sound planning and budgeting process. Such assessments are part of the planning and budgeting cycle in most developed countries. As the first chapter of the report describes, this PER-FA i s the result of a year-long partnership and dialog, not only between Government and donors but also between the different ministries within Government. For the first time, the Government has taken a leading role in such a public expenditure analysis exercise. The Government should incorporate the analysis and recommendations of the PER-FA (summarized in Appendix 1.2) into the SEDP and forthcoming budgets and should monitor implementation closely. I t should take steps to further enhance the capacity and status of public expenditure analysis, not only in M O F and MPI but across Government. It is recommended that the Prime Minister should commission a further PER-FA, to be completed within the next three years, again led by Government and produced in partnership with donors. In addition, annual updates should be conducted on a smaller scale. 16. The following table summarizes the main recommendations of the PER-IFA. n x .3 P 3 ._E m Yv) 2 um M t;l ei m Y mm I- I- 6 e, 81 0 $ 5 s N N EN WI I c'1 I c m or= 13 10 I In % 9 8 0 N 0 s 0 13 0 i n m 0 0 N N d .Y N 1 00 i? 0 r- 0 r- w o 0 8I 0 0 0N 0 O 0 N I m z c;' $s 8s CO 0 c;' c;"m 0 vl m m 8 0 0 0 0 N 0 0 0 0 N 0 N N N s8 N N N e 3 2 0 -2 2 m cd 0 C .e Y cd 13 C in a Y cd in 8 .e Y 2C u e, 0 e c cd G a c m .-M C C g e *a 0 g b -0 -Ya 0 e, em5 L .-a Y -$ 0 a 8 0 Y L Y 3 20 .3 E Y 5 e, o x B 3.- zM Y C 0 .3 .-m Y N 5i?? 0 Y 6 -0 E8 -0 .-G ei 9 a, zE Y .-.- 3K Y 2 .-M 5> K Y 2 e, C OD U a a 8x D U 9 > 2 LE: L C 3 a, 5 a, 42E c 0 s.-8 Y x o m $ 2 m a E E V 5 a, l-49 * d C e, 2b -.-mZ 0 bo .-> E -0 2 28 e C 83 Y De, m e, x m Y 32 -5 rr 0 YE .-C08a 8 Y 4 5e, 9 .- 0 Y k id 3 M w$2 6 E E E E E S . 3 E hl N N hl c.l N 3 In I sI mI sI m I sI O 0 0 O m O N 0 N 0r-4 0 hl 0 N .-E ."A M Y c 5 s .e Y C & Y2 $a C YE 'U 0 9 c .-G Y0 .3 0 3N 3a c .3 su .3 cd Y b .-N s 4 Y f 0 K '5 0 Y .-2 u m a$0 I, .- % 5e, m % 2 3 d ca -9 ce, ba a m f .-e" Y E Ed s M 9 .-x Y 2 3 u K 5e, M b e, C LCI r4 e, 5 8 M f Y -B$ a 25 G 0 1 .3 i v1 .3 111 2 3 s%I z2e Y 1 0 Y C cd Y M 5 U g ?! E2 C E 23m a, f Q C h; 4- I cd b 2 U rc -2 Y m 0 na, -M2 -E8 M Y c i cd C ui .3 YC .3 m x Y C 2 8c%I .3 Y 3 cd 0 C G- Y 5 M i a, K Y111 0 C 0 5 a, aCC 5 s Y a, cc 0 % .-x U cd :: 0 0 0 c 0 E cd C -0 .-C 0 U a, M C 0 C 3 5 -.- Y m 0 m U m B a C 8 0e z 9 2 .-M > 5 .3 Y 8E Y .3 e a, E b Q 5 111 S aa, C m .-i .3 E Ym .-8C P 4 Y D U m 0 -.- C m E 2 .3 C m 0 3E Y e cd x Em 2a, I- g cd UY 3 i 5 b cd cd M .-111 C 5 c, 2." -.- -a 0 c, .3 D a, z Y Y cd i;i w" 6> W 2 (c; i G w ; b 0 \1 Ec.l 0 I N 9 I 10I 10 s s s 10 I s 3 0 0 0 0 3 0 N 0 0 0 0 0 0 0 0 N N N N N c1 32u N N Y 1 8 s: x 5 2 .-d 0 se, 2 Y .-C ei 0 s Em 8x Y v) v, 5 gE a: .3 8P a a x 3 e, 0 e, U 9 Pe, e, s8 0 9 Y 5 .3 Y0k us .-E 2 Y .eC v1 U it2 3x .-0 92 Y m c Y Y C 0 9 $8 B .-Em e, 9 Q 5 Y +. m v1 8 0 .- Y c iv1 U 0e, M 6 Y Y Y- 0 5:e, .eC -6@ Y Y YW, 6 8> E 8E Y g M> e) > .-C Ue, .- e) 23 e0 5 - 3 E 0 Y C .-x Lz -22 Y .I s82 '1 6 4 ae, e, .? 3 6' %t i 5 c v) d w 0 m ETA ET E E L3 0 0 r- N N 0 N 0 CI m I In I 0 2 0 2I mI 0 0 Z C I 0 0 CI 8 0 N N 0 E E 8 P N sI NA0 A0 N 0 N 0 N 0 N 5 e, x Y 9 3 v1 B28 EE 8 rc a2 0 .-5% ac C 5 3 E: E * ."C0 Y 8 * 3 2Cm v1 0 -M U ga .-elC2 ze,m 2a P .-0 M Y 2m .-a Y 8 8cdm c E 8 k b m m C C 0 -.-u e, e, ' 3 D 2 a a su U E 0 '2I a--2 0 m u ij v1 e, wMu r 2 ; m 5 0 Y 9 a m 0 --Em m z .3 ze,m &I rr m -m 5 0 v1 a8 2 .-C a3 .-c3 c r E e, c x 5 a, .-F Y c 2\d .- v: 3 0: I E92s 0 0 m h l 4& E C .3 z ab C 5 -9 .3 c 'S W Q 9 -2 sb0 m YC e, c E %e, e, 2 5 .-3 2 5 PI E D 0 2 .-C0aC C .3 C Y 0 .3 .-ma Y 3 .-0 Y% E 5 e, zam Do zE C 0 .3 Y W -m Y 6:9 v1 3 C E -8 e, N 3 .3 m .- 2 Y -2g cd CC cj \d 1. BACKGROUND,APPROACHAND SCOPE BACKGROUND 1.1 In December 2003, the Prime Minister of Vietnam issued Official Letter No. 5932NPCP-QHQT instructing the Ministry of Finance, the Ministry of Planning and Investment and other concerned agencies to conduct a Public Expenditure Review and IntegratedFiduciary Assessment (PER-FA). 1.2 In a market economy, public expenditure is one of Government's most important instruments for supporting poverty reduction and growth. Periodic assessment of recent public expenditure and its management i s a crucial part of a sound planning and budgeting process. Such assessments are part of the planning and budgeting cycle in most developed countries. Traditionally in the developing countries, such assessments have been conducted by donors: for example, as part of World Bank-led Public Expenditure Reviews (PERs), Country Financial Accountability Assessments (CFAAs) and Country Procurement Assessment Reports (CPARs). Increasingly, however, reform-minded developing countries are taking ownership and leading such exercises themselves, as an integrated part of their own planning and budgeting cycle. 1.3 InVietnam, a series of public expenditure assessments has been conducted. These include the UNDP-World Bank PER of 1996; the IMF-World Bank study of 1999, "Towards Fiscal Transparency"; the joint Government-donor PER of 2000; the World Bank CFAA of 2001; and the World Bank CPAR of 2002. These assessments have made an important contribution to Vietnam's reforms, influencing the formulation of the State Budget and the Public Investment Program, helping to shape sector strategies and donor programs, and making an important contribution to the Comprehensive Poverty Reduction and Growth Strategy (CPRGS) and the Public Administration Reform (PAR) initiative. Progress in implementation of the 2000 PER i s summarized in Appendix 1.1 and discussed further in the following chapters. 1.4 The PER-FA launched by the Prime Minister would involve updating the previous PER, CFAA and CPAR, and for the first time conducting these exercises in a fully integrated manner. Like the 2000 PER, the work would be conducted jointly with donors and published as a joint Government-donor report. For the first time in Vietnam, however, Government officials as well as international experts would participate substantively in the analysis and would draft major written contributions. APPROACH Objectives 1.5 The objectives agreed for the 2004 PER-FA were as follows: 1 2 Vietnam Public Expenditure Reiielr,und Integrured Ficliiciai?* Assessment 0 to undertake a review of public expenditure and its management in Vietnam, including strengths and weaknesses; 0 to align development planning and budget planning across Government, as well as the country assistance strategies of financiers and donors, with the Government's Socio-Economic Development Plan and Comprehensive Poverty Reduction and Growth Strategy; e to provide analytical inputs to the 2005 State Budget, the 5 year Socio-Economic Development Plan for 2006-10, the Government's pilot Medium-Term Fiscal Framework and the Government's pilot Medium-Term Expenditure Frameworks at sectoral and provincial levels; 0 to support the Government's ongoing program to strengthen budget planning, execution, reporting and accountability; 0 to strengthen capacity to conduct public expenditure analysis in the Ministry of Finance, the Ministry of Planning and Investment, in sector ministries and provinces and in the Vietnamese research community; and 0 to provide an assessment of fiduciary risks for the Government, its financiers and donors, and its citizens. Participation 1.6 Public expenditure management i s a "whole of government" process. The PER- F A included participation from across Government. Central ministries leading study teams and contributing major working papers to underpin each section of the PER-FA included the Ministry of Finance (MOF), the Ministry of Planning and Investment (MPI), the Ministry of Agriculture and Rural Development (MARD), the Ministry of Education and Training (MOET), the Ministry of Health (MOH), the Ministry of Labor, Invalids and Social Affairs (MOLISA) and the Ministry of Transport (MOT). These ministries were also involved in editing the final report. A number of provinces, districts, communes and frontline spending units also made written submissions to the PER-FA. 1.7 The Government was determined to play a more active role in conducting this PER-FA. Nevertheless, the Government was keen that this PER-FA should involve strong participation from donors; and that, like the 2000 PER, it should be a joint Government-donor assessment, with the final report endorsed by all parties. Donors helped conduct the PER-FA, contributing working papers to underpin each study and assisting Government in preparing the final report. Donors providing major financial and technical contributions included the World Bank and the Like Minded Group of Donors (LMDG), which comprises donor agencies of the Governments of Canada (CIDA), Denmark (DANIDA), the Netherlands (NMDC), Norway (NORAD), Sweden (SIDA), Switzerland (SDC) and the United Kingdom (DFID).The PER-FA also benefited from valuable interactions with representatives of the European Commission (EC), the International Monetary Fund(IMF), the Japan Bank for International Cooperation (JBIC) and the United Nations Development Program (UNDP). 1.8 The participatory approach extended beyond the Government policy makers and the donor community. A large number of Vietnamese and international academics, Background Approach and Scope 3 researchers and consultants supported the Government and donor teams, including both private individuals and representatives from Bannock Consulting (UK), the Central Institute of Economic Management (CIEM, Vietnam), ECORYS Research and Consulting (the Netherlands), the Education Management Research Institute (Vietnam), Electricity of Vietnam (EVN, Vietnam), Georgia State University (USA), Hanoi Polytechnic University (Vietnam), the Institute for Market and Price Research (Vietnam), the Institute of Finance Research (Vietnam), the Institute of Health Strategy and Policy (Vietnam), the Institute of Training and Professional Development (Vietnam), Mediconsult (Vietnam), Mokoro Ltd (UK),the University of Washington (USA) and the Vietnam Economics Association (VEA, Vietnam), Process 1.9 Consultations on a concept note for the 2004 PER-IFA were conducted between Government, the World Bank and the LMDG in October 2003. Discussions were conducted between donors and officials in all participating ministries. Government officials, World Bank and LMDG staff also participated in a seminar jointly organized and sponsored by the Government of Vietnam, the Government of Korea, the World Bank and the LMDG in Halong Bay on October 9 and 10, 2003, which served a brainstorming and launch workshop for the PER-IFA. 1.10 The PER-FA was built up from a series of cross-sectoral and sector-specific studies. A working paper for each study was produced by an inter-ministerial Government team, working with Vietnamese experts. Cross-sectoral study teams were typically led by the Ministry of Finance or the Ministry of Planning and Investment, whereas sector-specific study teams were led by the relevant sector ministry, with participation from the Ministry of Finance, the Ministry of Planning and Investment and, where appropriate, other relevant ministries and provinces. 1.11 For each cross-sectoral and sector-specific study, a working paper was also prepared by one or more international experts, including staff of the World Bank, consultants and academics. Typically, the Government's paper was prepared first and the international expert then drew upon and validated this material. In some cases, the Government's paper formed a later commentary on the international expert's input. In all cases, there was extensive interaction between Government teams, international experts and donors. The international experts' fieldwork was mostly conducted during two missions: one in April and one inJuly 2004. 1.12 The final PER-FA report itself was drafted jointly by representatives of the Ministry of Finance and the World Bank. There was widespread consultation across Government and with international and Vietnamese development partners during a drafting mission in October 2004 and in the months following. In April 2005, the Prime Minister approved this draft of the PER-FA for publication, following consultation with ministers. A workshop to disseminate the published PER-FA i s scheduled to be held in May 2005 in Danang. 4 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment SCOPE Cross-sectoral studies 1.13 The PER-FA involved eight cross-sectoral studies. These studies are presented in Volume 1and include: 0 Fiscal Trends and Sustainability (Chapter 2); 0 Trends in Composition of Public Expenditure (Chapter 3); 0 Institutions for Public Expenditure Management (Chapter 4); 0 Institutions for Financial Accountability and Transparency (Chapter 5); 0 Decentralization to Subnational Government (Chapter 6); 0 Delegation to Spending Units (Chapter 7); 0 Public Investment Management (Chapter 8); and 0 Public Procurement Management (Chapter 9). Sectoral scope 1.14 The PER-FA involved five sector-specific studies. These sectoral studies are presented in Volume 2 and include: 0 Education Sector Expenditure (Chapter 10); 0 Transport Sector Expenditure (Chapter 11); 0 Health Sector Expenditure (Chapter 12); 0 Agriculture and Rural Development Sector Expenditure (Chapter 13); and 0 National Target Program Expenditure (Chapter 14). RECOMMENDATIONS ONFUTURE PUBLICEXPENDITURE ANALYSIS 1.15 Key recommendations from each study are presented in the relevant chapter and summarized in the table following the Executive Summary. 1.16 Regarding future public expenditure analysis, the PER-FA makes the following recommendations : Reflect the analysis and recommendations of the PER-FA in the forthcoming Socio-Economic Development Plan and forthcoming budgets and monitor implementation closely. 0 Further enhance the capacity and status of public expenditure analysis teams within the Ministry of Finance, Ministry of Planning and Investment, sector ministries and provinces. Lead a further PER-FA within the next three years, again working in partnership with the World Bank and other donors. In addition, annual update of the PER-FA Background Approach and Scope 5 should be conducted by Government, though with much more limited scope - covering perhaps one or two topics each year. E 3 d sG3 0 e, 5E E 3 za, Y VI a 0 i VI e, Q M 3 I) a, 9E y1 a ei n C Y rA aj S a 0 n e, E & B E. Y e, Y 2 20 .3 Y 'd S h 0 .-S0m Y I m > 48 e, Y 3 n 2. FISCALTRENDSAND SUSTAINABILITY INTRODUCTION AND OVERVIEW 2.1 Fiscal sustainability i s a pre-requisite for poverty reduction and growth. Public expenditure i s one of various types of government intervention designed to compensate for the failure of markets and to secure distributional equity. The level and composition of public expenditure has macroeconomic as well as microeconomic implications. In particular, unless a sustainable medium-term fiscal strategy i s developed and implemented, i t i s difficult to ensure adequate resources for socio-economic development and instability may cause difficulties for budget execution. Both of these eventualities are likely to be at the expense of growth and poverty reduction. 2.2 Vietnam has received wide acknowledgement for its fiscal prudence, with relatively small budget deficits and a relatively small stock of debt (both domestic and foreign). During the review period, Vietnam's fiscal stance has remained extremely prudent, with the budget deficit averaging less than 2 percent of GDP over the period 1999-2003 and the stock of debt remaining under 40 percent of GDP. The country has received increasingly solid grades from international organizations, foreign governments and sovereign risk rating agencies. Despite these impressive achievements, however, a number of risks to fiscal sustainability have emerged and will require ongoing attention over the coming years. 2.3 This chapter begins with a review of macroeconomic performance and aggregate revenue and expenditure trends in Vietnam over the past 5-6 years. The effect of these revenue and expenditure trends for the government's budget deficit and debt are discussed. The impact of certain off-budget items such as the fiscal costs of recapitalizing State Owned Commercial Banks (SOCBs) and the operations of the Development Assistance Fund (DAF) for government's indebtedness are also highlighted. Following discussions of public debt management and fiscal transparency, the chapter concludes with key recommendations. MACROECONOMIC PERFORMANCE 2.4 Real GDP growth has been high both by regional and international standards. It averaged 6.4 percent over the period 1997-2000 and 7.1 percent over the period 2001- 2003. Over the review period, GDP growth has been steadily rising: from 6.9 percent in 2001 to 7.1 percent in 2002 and 7.3 percent in 2003. In 2004, Vietnam recorded a GDP growth rate of 7.7 percent despite the challenges posted by the surge in world energy prices and the avian influenza. 2.5 The performance has also been remarkable in terms of poverty reduction. As demonstrated in the 2004 Vietnam Development Report, based on the 2002 Vietnam Household Living Standards Survey (VHLSS), 29 percent of the population had 13 14 Vietiinin Pirblic .upenditlire Revieit. and Integrated Fidircian*Assessment expenditures below the poverty line by 2002, compared to 37 percent in 1998 and 58 percent in 1993. Thus Vietnam lifted around 20 million people out of poverty in less than a decade. 2.6 Robust export performance has been a major driver of growth. During 1997-2003 exports grew at an average annual rate of 14 percent in value terms and now amount to half of GDP, with the share of manufactured products representing around 46 percent of the total. Exports grew by 21 percent in 2003, despite a lackluster global economy. The garments sector recorded the highest growth rate (45 percent) and equaled crude oil as the most important export for Vietnam in value terms. Rapid export growth has been aided by production expansion and by bilateral and regional trade agreements that came into effect during this period. Participation in the ASEAN Free Trade Area (AFTA) and the USBTA signed in 2001 were key milestones inthis process. 2.7 At the same time, strong growth has been accompanied by a rapid rise inimports. Import growth in 2003 stood at 28 percent compared with an average of 14 percent over 1997-2003. The main import categories witnessing increases in recent years are machinery and equipment, petroleum products, and other production inputs needed for the fast expanding export industries. This rise in imports therefore reflects a sustained expansion in production capacity. 2.8 With imports outpacing exports in the last two years, the trade balance has widened from a surplus of 0.8 percent of GDP in 2000 to a deficit of around 7 percent in 2003. However, strong inflows of remittances helpedkeep the current account deficit at 4.7 percent of GDP in 2003. The State Bank of Vietnam (SBV) reported record remittances into Vietnam of around US$2.6 billion in 2003 and over US$3 billion in 2004. If transfers operating through unofficial channels are taken into account, total remittances from abroad are probably in the range of US$3 to 4 billion a year. This i s the equivalent of one fifth of export earnings. FDI inflows are estimated to have reached about US$1.5 billion in 2003. ODA disbursements reached US$1.1 billion. This helped foreign reserves to rise from US$3.7 billion in 2002 to about US$5.6 billion in 2003. This i s the equivalent of 10 weeks of imports of goods and non-factor services. 2.9 The investment to GDP ratio stood at 35.1 percent in 2003 and 35.8 percent in 2004, compared with 28.3 percent of GDP in 1997. The share of the public sector in GDP has averaged 56 percent, while those of the domestic and foreign invested private sector has averaged 24 and 20 respectively. High levels of investment, especially in the public sector, have had an important impact on infrastructure development, economic transformation and implementation of national poverty reduction programs. Concerns persist about the quality of investments, in the sense that more growth and poverty reduction could have been attained from the same level of investment. 2.10 Vietnam has maintained prudent macroeconomic policies during the period. After two years of mild deflation, prices rose b y 3-4 percent in 2002-2003. Duringthe last year, inflation has accelerated to an annualized rate of 9.5 percent. These include increases in food prices due to the avian influenza outbreak and higher steel and oil prices. But there are also deeper economic forces behind the acceleration of inflation. Labor costs and prices in Vietnam remain very low by international standards, and they are bound to increase in a context of rapid economic growth. Fiscal Trends and Sustainability 15 2.11 Rapid growth has been accompanied by an ongoing monetization of the economy. The ratio of broad money to GDP rose from around 28 percent in 1998 to about 68 percent in 2003. The ratio of credit to GDP stood at nearly 50 percent at end-2003. The share of credit going to the non-state sector continued to increase and i s now close to two thirds of all credit extended. OVERVIEW OFFISCAL PERFORMANCE 2.12 During the period under review, strong economic growth was accompanied by strong growth in government revenues, as Table 2.1 illustrates. Even with a quite modest budget deficit, this has facilitated what as a share of GDP i s a quite gradual and sustainable increase in total expenditure but what in absolute terms i s a very rapid rise. This i s discussed further in sections below. Table 2.1: Overview of Fiscal Trends (1997-2003) (percentage of GDP) 1997 1998 1999 2000 2001 2002 2003 Total Revenueand Grants 20.8 20.2 19.6 20.5 21.4 22.7 23.1 Tax revenue 15.8 15.4 15.1 14.8 15.7 16.8 16.7 Non-tax revenue 4.2 4.2 3.9 5.3 5.4 5.5 6.1 Grants 0.8 0.6 0.6 0.5 0.4 0.4 0.3 Total Expenditure 22.6 20.3 21.2 22.6 24.2 24.1 25.1 Current Expenditure 16.3 14.7 13.8 15.9 15.9 15.7 16.8 CapitalExpenditure 6.2 5.7 7.4 6.7 8.3 8.4 8.3 Deficit -1.7 -0.1 -1.6 -2.0 -2.8 -1.4 -2.0 Primary balance -1.1 0.4 -1.0 -1.2 -1.8 -0.4 -0.9 On-lending (ODA only) 2.2 1.5 1.8 2.2 0.8 0.9 1.2 REVENUETRENDS 2.13 During the period under review, strong economic growth was accompanied by strong growth in government revenues. In nominal terms, government revenue and grants increased by an average of over 14 percent per year between 1998 and 2003. As a percentage of GDP revenue and grants rose from around 20 percent of GDP in 1998 to over 23 percent in 2003. The increase in revenues from consumption of petroleum products and production of crude oil from 4.1 percent of GDP in 1998 to 6.5 percent in 2003 explains the largest part of the revenue increase. The share of such revenues in total revenue and grants rose from 20 percent in 1998 to 22.4 percent in 2003. As a consequence, the public budget has become more vulnerable to oil price shocks. Furthermore, although the oil production outlook for the years ahead appears to be quite robust, crude oil exports cannot be relied upon as a source of revenues in the very long term. 2.14 To strengthen revenues as well as to encourage domestic and foreign investment, the State has lightened the tax burden through reduction of tax rates in many areas. Enterprise income tax rate has been reduced from 32 percent to 28 percent; the number of 16 Vietnam Public ExDenditureReview and Integrated Fiduciary Assessment VAT tax rates has been cut from 4 to 3, including the abolishing of the highest rate of 20 percent. The National Assembly has decided to eliminate overseas profit remittance tax and surtax on enterprise income. Most notably, in 2002, agricultural landuse tax has been reduced by 50 percent; and since 2003 nearly all farmers are exempted from such tax: now approximately 75 percent of Vietnam's population pay no direct taxes. Figure2.1: TrendsinGovernmentRevenueand Grants (% GDP) 1-1997 1998 - 1999 2000 2001 2002 2003 Tax revenue Non-tax revenue -Total revenue & grants ~ Import duties (tariffs and some specific consumption taxes), plus export taxes. Excludes VAT on imports. Fiscal Trendsand Sustainability 17 applied, will tend to rise. It may be noted that in the last two years imports have risen by nearly 25 percent per annum. Second, the VAT and specific consumption taxes2 on imports will continue to be applied. Third, increasing international integration i s likely to boost economic growth which will lead to rising revenue from other sources such as personal and corporate income taxes. International evidence suggests that the share of tariff revenues tends to decline in importance as GDP increases while the shares of other taxes rise. Figure 2.2: Share of Different Taxes in Total Tax Revenue 1998 2003 Other Corporate Other .*D, income tax Trade Taxes 28% 9% income tax Special 21% 3% VAT consumption 32% taX 10% 2.18 Ensuring a strong revenue response from these other sources will most likely entail reforms to improve the structure of such taxation. At the same time, ensuring that trade liberalization has its intended effect on GDP growth will entail significant "behind- the-border" reforms in areas such as provision of infrastructure services, financial sector, legal reforms, and reforms designed to improve the environment for private sector development. These reforms will be essential for enhancing the competitiveness of Vietnam's exports. 2.19 Finally, there i s a wide gap between the revenues that should be collected on the basis of statutory duty rates and what i s actually collected. This i s due both to abuse of exemptions as well as weaknesses in tax administration. As exemptions are reduced and customs and tax administrations are strengthened this gap will tend to decrease, thereby providing another offset to declining revenues. Clearly, these reforms will assume greater importance as closer integration occurs with the world economy. The upshot of the above discussion i s that trade liberalization will need to be accompanied by an appropriate set of complementary reforms to ensure that budget revenues are kept strong. OVERALL EXPENDITURE TRENDS 2.20 The year 1999 marked a turningpoint in the size of overall government spending. After steadily declining as a ratio of GDP in the years preceding, government spending (excluding on-lending, carry-over, and expenditure from retainedrevenue) has recovered during the period 1999-2003. From a low of 20.5 percent of GDP in 1999, it reached These would needto beWTO-consistent once Vietnam becomes a member. 18 Vietnam Public Exvenditure Review and IntePrated Fiduciarv Assessmen2 ~ ~~ 22.6 percent of GDP in 2000, 24.2 percent in 2001 and 24.1 percent in 2002 (Figure 2.3). B y 2003, this ratio was boosted to over 25 per cent. Figure2.3: Trends inGovernmentExpenditure a (% GDP) 30 20 25 15 10 5 0 1997 1998 1999 2000 2001 2002 2003 2.21 Between 1998 and 2003, total government spending rose at the remarkable average annual rate of 16 percent in nominal terms, with capital expenditure growing at some 20 percent. As a share of GDP, current expenditure rose from 14.7 percent in 1998 to 16.8 percent in 2003, while capital expenditure rose from 5.7 percent to 8.3 percent over the same period. Capital expenditure has averaged 34 percent of total expenditure over the last three years. 2.22 A more detailed discussion of expenditure trends, disaggregated by functional and economic classification, i s presented inthe next chapter. BUDGETDEFICIT, OFF-BUDGETITEMSAND PUBLICDEBT 2.23 The net result of these revenue and expenditure trends has been a budget Figure 2.4: Budgetdeficit deficit that has remained manageable, ranging between 0.1 percent and 2.8 (% GDP) percent over the period 1997-2003 (Figure 2.4). The National Assembly has been prudent in stipulating that the budget deficit plus amortization should not exceed 5 percent of GDP. The revenues collected by the government fully financed recurrent expenditure and partly Y I financed capital expenditure. Thus, in 1997 1998 1999 2000 2001 2002 2003 accordance with the so-called "Golden Rule", borrowing was undertaken to finance only capital expenditure and borrowing was always less than capital expenditure. 2.24 B y the end of 2003, the government debt that i s attributable to budgetary expenditures including ODA on-lending stands at around 33 percent of GDP. However, to get a more complete picture of the wider public debt position it i s important to add items that are currently not reflected in the budget. Two such items are bonds issued to \ Fiscal Trends and Sustainability 19 finance certain infrastructure and education projects and the recapitalization costs of SOCBs. Their inclusion in public debt numbers i s likely to raise its level by about 3 percentage points (Figure 2.5). This i s by no means an alarming level, but it would be prudent to be watchful of these trend^.^ Figure2.5: Public DebtfromBudgetand Off-budgetItems (as percentage of GDP) 40.0 30.0 --tDebtfrombudgetoperations 25.0 -.c-plus SOCB recapitalization -+- plusoff-budgetbonds ~ 20.0 2000 2001 2002 2003 2004 2.25 The next section takes a closer look at certain aspects of the debt emanating from on-budget items. Subsequent sections examine the risks arising from off-budget bonds, SOCBs and the operations of DAF, and the factors that might mitigate these risks. Government debt from on-budget operations 2.26 On-budget capital expenditure i s financed by ODA sources or b y issuing domestic government bonds. Inthe last two years Figure2.6: On-budgetDebt andExternalPublicDebt there has been - (as percentage of GDP) increased reliance I on domestic bonds I I1 40.0 I I for financing capital - 35.0 expenditures. This 30.0 - trend i s reflected in 25.0 - Figure 2.6, which 20.0 - 1 shows a divergence ---cOn-budget Debt 15.0 - i -m- External Public Debt in the trends for total and external public debt. 2.27 One consequence of greater reliance on domestic finance i s that the interest cost of debt has been rising. This i s because virtually all foreign public debt i s on concessional lending terms, while domestic debt carries higher interest rates. Figure 2.7 depicts the rise It may also be argued that the domestic resources mobilized by the DAF be added to the government's public debt. MOF officials disagree. Their explanation i s that DAF i s more like a bank and so its liabilities should be seen as contingent liabilities rather than as direct public debt. 20 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment in the implicit interest cost of total debt. At the same time the share of interest payments intotal current spending has also been rising (Figure 2.8). Figure 2.7: Implicit Interest Cost of Debt Figure 2.8: Interest Payments as Share of (%) Current Expenditure (%I :::II__i1:: ~~ 1.o 1.o 1998 1999 2000 2001 2002 2003 1 1997 1998 1999 2000 2001 2002 2003 2.28 Given Vietnam's current level of development, it i s expected that the government will continue to run budget deficits in the foreseeable future. If the government runs deficits of around 2-3 percent of GDP, and the economy continues to grow around 7-8 percent, then the debt-GDP ratio attributable to on-budget borrowing will remain manageable. Government bonds to finance infrastructure and education projects 2.29 The Government has been issuing bonds with a 5-10 year maturity to finance various expenditures kept off-budget. These bonds are part of the Government's strategy to raise VND63 trillion (US$4 billion) by 2010, to finance infrastructure and education projects including the north-south Ho Chi Minh Highway, roads in mountainous areas and in the Mekong River Delta, and irrigation projects in the disaster-prone central provinces. These bonds should be added to the above debt numbers to get the complete picture of public indebtedness. Even though the bonds are off-budget their issuance has been subjected to the scrutiny of the National Assembly. The interest payment on these bonds will be on-budget. In 2004, the Government issued about VND 5 trillion of bonds for infrastructure development and VND 2.5 trillion for the education sector, representing about 0.7 percent of GDP. Bonds of 2 years and 5 years maturity have carried coupons of around 8 percent. On maturity, the 5-year bonds will be rolled over for another period of 5 years. The main players on the purchasing side of government bonds have been State- owned Commercial Banks (SOCBs) and insurance companies. These interest rates have been marginally higher than the deposit rates offered by the banks for similar maturities. Off-budget expenditure arrears 2.30 Another significant concern regarding fiscal sustainability i s the large build-up of off-budget expenditure arrears for infrastructure projects, particularly in the transport 4Totalinterest payments in current period divided by previous period debt. Fiscal Trendsand Sustainabilitv 21 sector. As Chapter 11 discusses in more detail, major discrepancies exist between the Ministry of Finance data on expenditure in the transport sector (captured on a cash basis by MOF's State Treasury Department) and the data recorded by the Ministry of Transport (which i s on a quasi-accrual or commitment basis). M O F data shows transport spending reaching some 3.5 percent of GDP by 2002 whereas MOT data suggests transport spending accounted for perhaps as much as 5 percent of GDP-some 1.5 percentage points of GDP's difference. 2.31 The difference between MOF's cash expenditure figures and MOT'S expenditure accrualkommitment figures reflects the annual build up of arrears in the transport sector. M O T data suggests that these appear to be in the order of VND6.5 trillion, of which about VND3 trillion i s at central level and VND3.5 level i s at provincial level. 2.32 This situation has drawn attention to the need for much stronger commitment controls at every level of government. MOT, provinces and other parts of Government should not be allowed to make expenditure commitments that cannot be financed through the State Budget. It also draws attention to the need to reform transportkonstruction SOEs (an issue discussed further in Chapter 11). The need to clear these arrears represents a major burden on the budget and a threat to fiscal sustainability which, though currently manageable, i s a cause for serious concern requiring urgent action. Public debt from recapitalizationand DAF 2.33 The four large SOCBs have been implementingrestructuringplans for the last 2-3 years. Non-performing loans worth VND23 trillion as of end-2000 were identified for resolution. B y end-2003 around 70 percent of these had been resolved. In conjunction with these restructuring plans SOCBs have received a combined capital injection of about VND10.9 trillion. This recapitalization has largely been in the form of 20-year government bonds carrying a coupon of 3.3 percent. A full picture of the government's debt would include such bonds, both those issued previously as well as those that may be issued in the future. The bonds issued previously may be interpreted as a realization of the government's contingent liabilities. The current capital adequacy ratio of the banks at under 4 percent i s considerably below generally accepted standards. The government may therefore face additional costs inraising the capital of the banks to prudent levels. 2.34 Credit from SOCBs has been growing rapidly recording an expansion of 28 percent in 2003. The ratio of SOCB credit to GDP stood at 37 percent at end-2003. The quality of lending however remains a concern. To an extent, fast expanding credit may be viewed as a build up of contingent liabilities with its quality representing the risk of their realization. Improving the quality of new credit i s the key to ensuring genuine banking reform and limiting fiscal costs. As noted above, there was progress in resolving the stock of past N P L s of SOCBs, but the quality of new loans remains largely unknown. 2.35 Recapitalizing the SOCBs through the issuance of shares or other equity-like instruments would relieve the fiscal burden of financial sector reform and could also, potentially, help strengthen SOCB corporate governance. Due to the high savings rate of Vietnam, and the limited range of long-term savings instruments available to the population, the issuance of shares or convertible bonds could allow the SOCBs to raise a considerable amount of resources in order to meet prudent capital-adequacy ratios. But 22 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment the protection of shareholders requires that reliable information be provided on the situation of the SOCBs, and that plans be devised to redress their most important weaknesses, especially in relation to their stock of NPLs. A comprehensive evaluation of the SOCBs should thus be a pre-condition to their listing in the stock market or their issuance of convertible bonds and other equity-like instruments. 2.36 One source of potential fiscal risk has been the extension of bank credit to SOEs. As Figure 2.9 illustrates, commercial bank credit (state and non-state banks) to SOEs has risen from about 10.5 percent of GDP in 1998 to 17.2 percent of GDP in 2003. However, the share of SOEs in total bank credit has been falling: from 52 percent in 1998 to 36 percent in 2003. Correspondingly, the share of SOEs in SOCB credit has fallen from 58 percent to 45 percent over the same period. It thus appears that hardening of the budget constraint on SOEs i s becoming increasingly effective in relation to banking credit. However, SOEs' investment plans of allegedly strategic importance are still supported through policy lending, especially by DAF. 2.37 Funds to be on-lent through DAF are mobilized from ODA sources, and domestically through borrowing from pensions and social funds (VSS), postal savings, and issuance of government bonds. Bonds with a 15 year maturity and carrying a coupon of 9.4 percent have recently been sold, with the proceeds to be transferred to the DAF. VND7-8 trillion are planned to be mobilized in this manner in 2004. Domestically mobilized funds for DAF represent another large accumulation of contingent liabilities for the government. The on-lent funds could carry interest rates that are low5r than the cost of raising them. Assuming that debts are serviced in a timely manner, the gap between the interest at which the funds are borrowed and lent represents the level of subsidy to be provided by the government. Figure2.9: CommercialBank Credit to SOEs 60% 50% 40% I -W- GDP 30% "/oOf +% of totalcredit 20% v K 10% i 0% 1998 1999 2000 2001 2002 2003 2.38 The quality of the fast expanding policy lending through the DAF i s a matter of concern. In 2004, its outstanding stock of loans i s expected to be equivalent to nearly 12 percent of GDP (Figure 2.10). A little more than half of this amount corresponds to on- lending from domestic sources. O f the resources mobilized domestically, about 80 percent has been on-lent to SOEs. Domestic capital sources of the DAF amount to roughly 9 percent of annual GDP in 2004, making it the biggest financial institution in the country. This i s despite the fact that many operational aspects of the DAF are still at an early stage of development, including credit risk management, accounting and reporting standards, information systems, and supervision. 14% 12% i ODA on-lending 10% Non-ODA on-lending 8% 6Yo 4% 2% 0% - Fiscal Trends and Sustainabilitv 23 Figure 2.10: DAFOn-lending (% GDP) I 2000 2001 2002 2003 2004 Rojection 1 2.39 In addition, new regulations and policies have been promulgated by the National Assembly and the Government with a view to improving the quality of state investment in general and of state credit in particular. These include the National Assembly's Resolution No. 36/2004/QH11 dated December 3, 2004 on state capital investment, regulations on investment inspection, verification of works completed and unit costs. Proper implementation of these regulations may help improve the quality and effectiveness of state investment, thus contributing to the reduction of the Government's contingent liabilities. 2.40 The issuance of Decree 106/2004/ND-CP, which regulates the operation of DAF, i s a step towards providing a sound regulatory framework for policy lending. It significantly narrows the list of borrowers eligible for DAF lending. The Decree states that only projects that are capable of direct repayment, have socio-economic efficiency, and have feasible business or production plans, are eligible for DAF support. The Decree also stresses that project assets are to be used as security, and those assets cannot be assigned, disposed, mortgagedor pledgeduntil loans are fully repaid. 2.41 An element of risk-sharing, through co-payments by other government bodies or by commercial banks, i s also built into the Decree. The new framework limits the number of support mechanisms and leads to a partial reliance on commercial lending. Those mechanisms are project loans, interest rate subsidies and credit guarantees. Only the former will be extended by DAF itself. Both interest rate subsidies and guarantees apply to financial support by commercial banks. And subsidies apply only to projects that have been completed and put into commercial operation and whose debts have been fully repaid. Investment lending and credit guarantee from DAF cannot exceed 70 percent of the total capital of a project. Interest rate subsidies cannot exceed 85 percent. 2.42 Success in putting DAF operations on a sounder footing will crucially depend on the effective implementation of these measures. Effective implementation would decrease 24 Vietnam Public Exuenditure Review and Integrated Fiduciarv Assessment the probability that debts would not be serviced and would ensure fiscal costs are minimized. FUTURE EVOLUTIONOF GOVERNMENT DEBT 2.43 Under a base case scenario, public debt (including off-budget items) i s projected to rise from a current level of about 36 percent to over 44 percent in 2010 (Figure 2.11). Broadly, this assumes a continuation of recent trends for main macroeconomic variables such as GDP growth, inflation, interest rates and the government's primary surplus. Thus GDP growth is assumed to average 7.2 percent over the period while inflation i s contained at 3.5 percent. Domestic interest rates average 7-7.5 percent and foreign borrowing i s assumed to remain largely concessional. The share of foreign debt in total debt is assumedto decline from a current level of 68 percent to 55 percent. The exchange rate continues to depreciate at its current slow rate. The government i s able to maintain a primary (non-interest) deficit of 1.1percent of GDP. The underlying assumption in terms of revenues i s that these would remain around 22 percent of GDP as presently observed, Figure 2.11: A Base Case Scenario for Government Debt (% GDP) (including realized contingent liabilities) 50 45 40 35 30 25 20 2000 2002 2004 2006 2008 2010 2.44 As with the recent evolution of debt, off-budget items are also introduced as drivers of debt. This includes off-budget bond issuance amounting to 0.9 percent of GDP per year on average. Also included i s the realization of contingent Figure 2.12: Alternative Debt Scenarios liabilities of 5 percent of current (including realized contingent liabilities) GDP spread out over 6 years. Importantly it i s these latter items 5 5 1 I that are subject to greater risk. ~ 50 - Factors that would tend to mitigate 45 - these risks include plans to equitize 40 - SOCBs, improving the quality of 35 - credit and the NPLs resolution 30 -- 1 process, especially where SOEs are - -Higher contingent liabilites involved, and the new regulatory framework for DAF. Vietnam's 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 present debt i s at manageable levels. Effective implementation of these policies would be a key factor in ensuring the sustainability of the government's debt burden in the foreseeable future. Fiscal Trends and Sustainability 25 2.45 Inthe event that the realizationof contingent liabilities, inthe form of fiscal costs of financial sector reform, i s higher or if growth i s lower the debt burden would be accordingly greater. Figure 2.12 presents two alternative scenarios. In the first, the realization of contingent liabilities i s twice as high as in the base case. In the second, the higher realization of contingent liabilities i s associated with a growth rate that i s one percentage point lower on average relative to the base case. The debt level in the latter case i s nearly 50 percent of GDP, the government's notional ceiling. STRENGTHENINGMANAGEMENT OF PUBLICDEBT AND FISCAL RISKS 2.46 Current public debt management practices in Vietnam are aimed at controlling the levels of debt stock and debt service payments. Insufficient attention i s currently paid to management of risk-cost trade-offs and to analysis of wider fiscal risks. However, the Government i s developing a strategy to enhance its capacity to analyze and manage the risks to its public debt portfolio and its wider fiscal risks. 2.47 Debt related information i s currently fragmented with Government. Currently, external debt data gets recorded in several constituent units with the Ministry of Finance, the State Bank of Vietnam and the Development Assistance Fund.The External Finance Department in M O F records the official bilateral and multilateral loans, while SBV's Foreign Exchange Department records data on external debt contracted by all enterprises (including SOEs) and financial institutions. Both these units have already installed the UNCTAD Debt Management and Financial Analysis System (DMFAS) for this purpose. However, it i s not yet fully operational and various duplicate and satellite databases remain in operation. Recording of domestic debt remains fragmented and weak. These multiple databaseshave a narrow focus that meet only the business needs of the particular unit handling the specific type of debt data. This makes it difficult for the authorities to get a comprehensive picture on the indebtedness position of the government at any point intime. It also results inunnecessary duplication of data within Government. 2.48 The Government has initiated a program of reform as part of its Public Financial Management Reform Initiative. Cross-country experiences in setting up fully functioning debt management institutions show that it i s time consuming and will require dedicated resources throughout the process and a continuous skills development effort. Several necessary steps will need to be accomplished in this endeavor. These include: (1) the identification of debt management objectives for the Government; (2) the establishment of a risk management framework; (3) the identification and management of different sources of risks (such as currency risk, interest rate risk, liquidity risk, among others); and (4) the creation either of a debt management office or a coordination mechanism between the various departments involved in debt management, to lead the organizational and regulatory changes that will be needed to get the debt management modernization program moving on a sustained basis. 2.49 Early priorities for the Government include strengthening arrangements for the recording and monitoring of fiscal risk, starting with domestic debt and credit through DAF. 26 Vietnam Public Expenditure Review mid Integrated Fiducian Assessment FISCAL TRANSPARENCY 2.50 Fiscal transparency plays an important role in ensuringfiscal sustainability. Fiscal transparency enables accountability to citizens regarding the use of government finances and the quality of fiscal management. It also facilitates the monitoring conducted by government agencies, social organizations and grass-root communities of the allocation and spending of public resources by national and subnational governments. As well as encouraging sustainability, this can enhance the effectiveness of public financial management. 2.51 Vietnam has made continuous efforts since 1997 to enhance fiscal transparency and publicity. These efforts have followed two tracks. The first i s announcement of policies regarding allocation norms, expenditure norms, fiscal assignments and monitoring arrangements. The second i s the publication and dissemination of figures, data and materials relating to annual budget planning and final accounts of the State Budget at every level of Government and at spending unit level. 2.52 In 1998, for the first time, the budget revenue and expenditure figures were published in the General Statistics Office's Statistical Yearbook. In the same year, again for the first time, Vietnam's State Budget revenues and expenditures were published in the I M F ' s Government Finance Statistics Yearbook. To further the progress of fiscal transparency, in November 2001 the Prime Minister issued Decision N o 182/200l/QD- TTG on the amendment of the Rules of Fiscal Publicity. Decision 182 made clear the stipulated content, the timing and format of the publicity of the state budget. 2.53 To ensure the legality of the extended publicity of finance and state budget, it i s stipulated in the revised State Budget Law of 2002 that the budget plans, budget final accounts and auditing results of the central budget, the local budgets and other organizations financed by the State Budget must be made public. To implement this stipulation, the Prime Minister in November 2004 replaced Decision 182/2001 with Decision 192/2004/QD-TTG on Fiscal Transparency, which further details and specifies the rules of fiscal transparency and publication (see Box 2.1). 2.54 Annually, the Ministry of Finance prepares a comprehensive report on the publicity of the State Budget at central and provincial level. For the year 2002, 58 out of the then 61 provinces sent their publicity reports to the Ministry of Finance. Of these, 56 provinces made public the 2000 final accounts and all of them made public the 2002 budget plans. Most provinces closely adhered to the sample forms stipulated by the Ministry of Finance. 2.55 In addition, in 2002, the Ministry of Finance received publicity reports from 75 out of 84 sector ministries and central agencies. 73 of these made public their 2000 final accounts and all made public their 2002 budget plans. All of these agencies followed the publicity regulations. 2.56 Nevertheless, the Government has identified considerable scope to improve fiscal transparency further. As Chapter 4 describes, the Government i s taking major steps to simplify and streamline the budget process, and to strengthen Treasury and State Budget financial management information. As Chapter 5 describes, the Government i s also Fiscal Trends and Sustainability 27 implementing a program of reforms to strengthen wider public financial transparency and accountability, including strengthening financial reporting and audit. Box 2.1: Key Contents of Decision 192/2004/QD-TTg on Fiscal Transparency ipplicability 1 The regulation i s applicable to all budget levels, state budgetary units, spending units, funds financed by state budget resources (including DAF, Vietnam Social Security and Urban Investment Funds) SOEs and organizations financed by contributions from the public -a wider range of entities than the earlier regulation. )isclosure Requirements 1 The requirements for disclosure have been substantially increased. Inparticular: 9 Allocation norms approved by the National Assembly and the People's Councils must be publicized at each budget level. 9 Organizations funded from the State Budget must disclose their State Budget support, budget plans and final accounts, and the methods used to determine their State Budget support. k For state capital investment projects, annual budget allocations, budget plans, final accounts and biddingresults for procurements must be disclosed. 9 Fundsfinanced by the StateBudget must makepublic their charters, financial arrangements. annual financial plan and operational performance. 9 For SOEs, mandatory disclosure includes their financial position, operational performance. provision for and use of welfare funds, remittances to the State Budget, per employee income and return on state equity. Methods of Disclosure Methods of disclosure stipulated in the Regulation include: 9 Announcement atpublic meetings. 9 Circulation inprinteddocuments. 9 Posting at offices for at least90 days. 9 Informing inwriting. >> Posting on websites. Publicity through the mass media. Reporting and Monitoring of Disclosure Agencies, organizations and units and their heads are responsible for sending the disclosed information and the compliance status of their subordinate bodies to their higher level. Ministries and Provincial People's Committees are responsible for monitoring compliance of their subordinate bodies. The Fatherland Front and other mass organizations will also oversee the implementation of the Regulation. There is a normal requirement to respond to questions regardingdisclosures within 10days. 2.57 From a purely fiscal perspective, a key priority for increasing transparency include avoiding further off-budget bonds issuance, channeling all Government borrowing in future through the budget; and taking urgent action to resolve the current expenditure arrears and to prevent their further build-up. These major off-budget operations undermine the integrity and credibility of fiscal management information. Another key priority i s to align fiscal reporting with the internationally recognized 28 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment "Government Financial Statistics" standards of the IMF - an issue which i s being addressed as part of the implementation of the new Treasury and Budget Management Information System (discussed in Chapter 4 RECOMMENDATIONS 2.58 To help secure fiscal sustainability, the PER-FA makes the following recommendations: Rapidly implement the plan to prepare a realistic and sustainable Medium-Term Fiscal Framework (MTFF). This should be done as part of every year's budget cycle. This should take a rolling three-to-five year perspective and should include forecasts of major items of revenue, expenditure and deficit financing, including guidelines for spending by sector. Aim to defend the share of revenues inGDP at its current reasonable target of 21- 22 percent. This will require ongoing attention to the impact of oil price fluctuations and trade tax reform on expected revenues; and re-doubled efforts to strengthen tax administration. Clarify and rationalize assignment of responsibility for management of debt and strengthen arrangements for the monitoring and management of other fiscal risks. Keep the budget deficit (measured net of amortization) below 3 percent of GDP. Restrict further off-budget bond issuance, channeling all Government borrowing infuture through the budget. Take urgent action to resolve the current expenditure arrears (particularly in the transport and agriculture sectors) and to prevent their further buildup. To support fiscal transparency, align Government's fiscal reporting with the internationally recognized "Government Financial Statistics" standards of the IMF. 3. TRENDS INCOMPOSITIONOFPUBLIC EXPENDITURE INTRODUCTION AND OVERVIEW 3.1 As Chapter 2 noted, the year 1999 marked a turning point in the size of overall government spending. After declining as a share of GDP in the years preceding, government spending (excluding on-lending, carry-over and expenditure from retained revenue) bottomed out at 20.5 percent of GDP. B y 2003, this ratio was boosted to over 25 percent. Between 1998 and 2003, total government spending rose at the internationally remarkable average annual rate of 16 percent innominal terms. 3.2 This rapid growth presented Vietnam with an extraordinary opportunity to restructure the composition of its expenditures: for example, to ensure more efficient allocation between economic types of expenditure (capital expenditure versus recurrent expenditures such as wages and salaries, operations and maintenance, debt interest, etc); to ensure more effective allocation between functions of government (education, health, transport, agriculture and rural development, etc); and to ensure more equitable or "pro- poor" allocation (between regions and social groups, etc). The exceptionally buoyant fiscal situation made it potentially less painful to restructure expenditure, offering the opportunity to reshape expenditure in a way that need not require absolute cuts in expenditure in any area. 3.3 This chapter considers the extent to which these opportunities were seized. The first section considers trends in the allocation of expenditure by economic type. The second section considers trends in the allocation of expenditure by function (or sector). The third section considers the extent to which spending was "pro-poor". A final section presents recommendations. More detailed data on trends in composition of public expenditure are presented in Appendices 3.1-3.6. ECONOMIC COMPOSITION OF EXPENDITURE Expenditureon Capital,OperationsandMaintenance 3.4 Capital expenditure has a vital contribution to make to socio-economic development. Vietnam has always accorded high priority to capital expenditure, including in the years of economic difficulty. The Government has followed the ruling that the growth rate of capital expenditure should be higher than the growth rate of recurrent expenditure. 3.5 During the review period, capital expenditure has been maintained at a high and growing level. Including purchase of fixed assets, the share of capital expenditure 29 30 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment increased from 31 percent in 1997 to 40.5 percent in 2002. Capital expenditure has increased significantly faster than recurrent expenditure and total public expenditure. 3.6 Alongside the growth of State Budget capital expenditure, there has been a shift in terms of its sectoral composition, with more focus on sectors identified by the Government as priorities, including transport, irrigation, education and infrastructure for poor communes. 3.7 It is also noteworthy that although State Budget capital expenditure has increased rapidly, its share in the country's total capital spending has not increased at the same pace. This indicates the larger contribution that the enterprise and the private sectors make to total investment in Vietnam. 3.8 As noted above, the ongoing growth in the share of capital expenditure reflects a requirement that the rate of growth of capital expenditure should be higher than the rate of growth of recurrent expenditure. Having been carefully adhered to for some time, this policy now needs urgent review, to come up with a better and more appropriate balance between those two types of expenditure. In the longer term, an ever-growing share of government capital expenditure i s not a sustainable strategy. It must be recognized that new capital assets carry recurrent cost implications-for example, recurrent budget i s requiredto maintain roads. Furthermore, beyond a point, government financing of capital investment may crowd out the private investment which i s so critical to economic success. 3.9 Vietnam, like many developing countries, has not paid proper attention to operation and maintenance (O&M) expenditure. Consequently, in several sectors the rates of return on such O&M expenditures are estimated to be higher than those on many new capital investments. This i s certainly the case in the irrigation sector, where i t i s estimated that some 50 percent of large scale irrigation infrastructure i s out of commission due to lack of maintenance. In this sector, a recent study has strongly suggested that operation and maintenance expenditure yields higher returns than new capital investment (see Box 3.1). 3.10 In the transport sector, too, there is growing evidence to indicate an impending maintenance crisis. A recent model for the World Bank-funded Road Network Improvement Project (discussed further in Chapter 11) shows that if road maintenance expenditure stays at its current level over the next ten years, the condition of the network will substantially deteriorate, with 34 percent of roads falling into poor condition, including 55 percent of the high traffic volume network. The percentage being in good condition would fall to just 10 percent of the network. From an economic cost-benefit perspective, this i s likely to be inefficient: an earlier World Bank study estimated that around US$160 million annually was imposed on the economy in the form of higher vehicle operating costs due to poor road maintenance. Even from a simple fiscal perspective, evidence from around the world shows that neglect of road maintenance i s highly inefficient: roads will eventually have to be rehabilitated at a cost many times higherthan routine operation and maintenance. Trends in Composition of Public Expenditure 31 Box 3.1: Returns to New Investment versus Operation and Maintenance Spending: The Case of Large-Scale Irrigation The Ministry of Agriculture and Rural Development (MARD) invests heavily in improving irrigation and drainage systems. The bulk of this amount goes to improvement of large-scale infrastructure such as putting in place large and expensive pump stations, expensive concrete-lined canals and other associated delivery systems (`the hardware'). But in recent years, some of the money, inparticular from donor-funded projects, was also spent on improving the management, operation and maintenance of existing infrastructure (`the software'), including the establishment of local-level participatory institutions for water management. To compare the impact of these different types of irrigation interventions on household welfare, the ADB and World Bank conducted ajoint study during the course of 2003. The study focused on areas of interest in three large-scale irrigation schemes, each representing one type of intervention. The study results showed a very different impact for each type of intervention. Investment in improving infrastructure only (Quang Nam) showed the lowest impact. This intervention did not lead to a higher cropping intensity of the farm, nor did it lead to a reduction in exposure to drought and flooding or in the variability inrice yield; rice yields were also not increased and total labor input and costs of producing rice were not reduced. In contrast, the highest impact was achieved in the scheme area where efforts concentrated on both upgrading infrastructure and strengthening the operation and management of the system (Thanh Hoa). This significantly increased rice yields, significantly reduced exposure to drought and flooding and also lessened the variation in rice yield; it also reduced farmer irrigation costs. Total labor costs and production costs for rice were also reduced and farm income was up by 142 percent. However, non-farm income was significantly higher among the `control' group, which may have led to a lower interest of the household in farming. Nevertheless, total household income was 22 percent higher for the farmers with the intervention. Significant positive impact was also seen in the scheme where all efforts concentrated on improving operation and maintenance of the irrigation system (Ho Chi Minh City/ Tay Ninh). Here the intervention led to a significantly higher whole-farm cropping intensity, as well as an increased rice yield, reduced costs for irrigation and led to a lower production cost of rice. In short, the study provided clear evidence that the highest impact of public spending on improving irrigation systems can be achieved under a combination of strengthening system operation and management and improving the infrastructure, and the lowest impact when it only pays attention to upgrading infrastructure. Source: Aldas Janaiah and Mekong Economics (2004) budget was squeezing out necessary operations and maintenance expenditure. 1~~ Between 1997 and 2002, the relative share of the O&M budget has continued to IWagss&Wanes decline, from 23.4 percent to less than 17 percent of total 056 expenditures (see Figure 3.1). 1997 1998 1999 XXX, X I 1 xo2 This has happened at a time 32 Vietnam Piiblic Expenditlire Relieu,cind Integrated Ficlirciar?*Assessment 3.12 As Figure 3.2 illustrates, the share of recurrent expenditure in total State Budget Expenditure varies significantly by sector. "Social" sectors such as Education and Training and Health are less capital intensive; while "economic" sectors such as Agriculture, Transport and Industry are heavily capital intensive. The higher growth rate of capital expenditure has resulted in a slowing down in the growth rate of recurrent expenditure. The trend has not been uniform across all sectors: sectors such as education and health managed to hold their recurrent expenditures roughly constant as a share of their total State Budget expenditures. Nevertheless, on average the share of recurrent expenditures in the State Budget i s sharply down. Figure 3.2: Recurrent Expenditureby Sector, 1997-2002 (percent of sector's total State Budget expenditure) I 100% -Social Insurance 80% -Health Education & Training 60% AVERAGE -Science, Techn & Env 40% -Culture. Sport 20% +-Agriculture, For & lrrig -Transport, Storage & Comm 0% ----Industry 1997 1998 1999 2000 2001 2002 Expenditureon Wages and Salaries 3.13 The share of Wages & Salaries (including government pensions) in total expenditure fluctuated between 27 and 33 percent of total expenditure (43 and 48 percent of recurrent expenditure), between 1997 and 2002. For 2003, budget figures show a further increase, accounting for 36.5 percent of total expenditure, largely due to a substantial increase in minimum wages. In 2003, the National Assembly adopted a proposal from the Steering Committee of Salary Reform to increase the minimum salary across the board b y 38 percent. The implementation guidelines were later spelled out in Decree 3/2003/ND-CP. 3.14 There are clearly many inadequacies in the civil service salary structure. The current salary system i s extremely complex, with around 20 salary scales, 200 grades and more than 2000 steps in the salary system. Other inadequacies appear to include pay compression (insufficient differences between lowest and highest pay), lack of parity with the private sector in certain classes of employment, limited recognition of educational attainment, salary based on seniority rather than meritocracy and an opaque, non-monetizedbenefit system. The work of the Steering Committee for Salary Reform i s an attempt to address these inadequacies. 3.15 However, the salary reform strategy needs further clarifying. There have been several attempts to date to assess the extent to which civil servants are underpaid in comparison with the SOE sector and the private sector. However, this i s a complex issue. According to an estimate by the Ministry of Home Affairs (MOHA), salaries represent only part of the remuneration actually received by civil servants. Other forms of remuneration include allowances (for housing, for instance), and various supplementary Trends in Composition of Public Expenditure 33 payments for services rendered. Taking these other forms of remuneration into account, there i s no firm basis for concluding that civil servants are consistently underpaid, although there are grounds to assume that civil service pay bears little relationship with labor market alternatives. 3.16 There i s some evidence to suggest that higher ranking and more skilled officials may be somewhat underpaid compared with labor market alternatives. However, the 2003 salary increase made no distinction between the ranks of the civil service, no allowance for special skill groups and no distinction between retirees and non-retirees. Treating all civil servants the same i s perhaps attractive from an egalitarian perspective but could impede productivity gains as there is no clear link to performance improvement. The Government has indicated that it would like to link pay better to performance improvement as part of its Public Administration Reform (PAR). However, this statement of intentionneeds to betranslated into strategy and then action. 3.17 Furthermore, civil service pay reform needs to be planned in parallel with public employment reform (since both pay and headcount determine the total civil service payroll) and as part of the wider PAR Master Plan. Vietnam's civil service i s not particularly large by international standards, nor does the payroll account for a particularly large share.of spending. However, pay increases will create pressure on the budget unless they are off-set by selective reductions in staff and there may be a need to make room for younger and more qualified personnel. The Government has a program to reduce the numbers on the direct payroll by around 15 percent. However, to date only around 25,000 civil servants have been retrenched. In practice, most of the retrenchment to date has taken place in the context of delegation to spending units under Decision 192 and Decree 10 (discussed further in Chapter 7), and not as part of a centrally coordinated strategy. There i s clearly an urgent needto strengthen and integrate pay and employment reform across the civil service. 3.18 The Government recently passed a new set of salary scales for civil servants, Government officials, pensioners and enterprise employees. This salary reform i s expected to have a significant fiscal impact and will need to be carefully analyzed and monitored over the coming months. FUNCTIONAL COMPOSITION OFEXPENDITURE 3.19 Looking at the functional classification of expenditure, some striking changes have taken place since 1997. Figure 3.3 compares the functional composition of the budget in 1997 and 2002. Figure 3.4 charts the changes over time in four key sectors. 3.20 Inline with the national priorities, the shares intotal expenditure of Educationand Training and Science have shown a rising trend. The share of Education and Training increased from 14.1 percent of the total State Budget expenditure in 1997 to 16.7 percent in 2002, while the share of Science, Technology and Environment rose from 1.3 to 2.2 percent. The National Assembly raised the Education and Training sector's share of the budget to 18.65percent for 2004, higher than the original target set by the Government. Further increases in the share of Education and Training in the budget may be needed in coming years, as lower-secondary enrollment rates increase. However, these cost ~~ This percentageis calculated basedon total expenditure excluding amortization. 34 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment pressures will slow somewhat, as demographic changes reduce primary enrollment numbers.Education and Training sector expenditure i s discussed indetail inChapter 10. Figure 3.3: Functional Composition of Budget Expenditure 1997 and 2002 - (percentages of total State Budget) 1997 2002 0 Agriculture,Forestry,Irrigation Fisheries 0Transport,Storage& 5%no/. Commulication 0 Industry w Education&Training 0 Health 4% Social hsurance 0 Culture-Sport 7% I#Science Technology& Environment I# Administrationekpenditure 0 Interestpayment Others Figure 3.4: Trends inKey Sectors 1997-2002 - (percentages of total State Budget) 18% , 1 14% 10% t 6% 2% J 1997 1998 1999 2000 2001 2002 1 --e Agriculture+Transport -A- mucation+Health1 3.21 The share of the Transport sector also increased, from 9.7 to 13.8 percent. However, it should be noted that-as discussed in Chapter 2 and in more detail in Chapter 11-there i s evidence to suggest that, including the unpaid arrears accrued each year, the amount of spending on transport may be considerably higher than indicated by these cash figures. Total accrued expenditure in the transport sector each year may well now be greater than education expenditure. A major investment in infrastructure has been necessary in Vietnam to complete reconstruction aild to respond to the demands of economic growth. However, government investment in road infrastructure may now need to focus on quality not quantity, to take account of the maintenance costs of the existing road network and to reflect the fact that many of the highest return investments may already have been made. Transport sector expenditure i s discussed in detail in Chapter 11. Trends in Composition of Public Expenditure 35 3.22 The share of Agriculture, Forestry & Irrigation showed no clear direction, fluctuating between 5.2 and 6.3 percent of total expenditure. However, it should be noted that, given the fast overall growth of total public expenditure, a constant expenditure share still implies an extremely large annual increase in absolute terms. Irrigation accounts for the largest share of public expenditure in the sector. As Box 3.1 discussed, lack of maintenance i s a growing concern and there i s evidence to suggest that the rate of return on such investments is not always as high as it might be. Agriculture and rural development sector expenditure i s discussed in further detail in Chapter 13. 3.23 The share of Health also fluctuated without a clear direction, between 6.1 and 7.1 percent of total expenditure. Again, it should be noted that, given the fast overall growth of total public expenditure, a constant expenditure share still implies an extremely large annual increase in absolute terms. Healthcare spending i s increasing through health insurance and social welfare funds. Nevertheless, the system depends to a great extent on private contributions, including by the poor. Demographic changes are raising the population shares of the elderly and epidemiological shifts are increasing the proportion of the population suffering from certain chronic diseases. Both of these trends increase the need for public funding to protect the public's health and to ensure healthcare for the poor. At the same time, economic development and rising incomes i s leading to an increase in the demand for high quality healthcare and special services. Currently, the share of health in public spending i s low in Vietnam by international standards, and the health sector i s perhaps the one major sector in Vietnam that i s clearly going to need a larger share of the budget in the coming years. Health sector expenditure i s discussed in further detail in Chapter 12. 3.24 Social insurance expenditure decreased slightly in relative importance. However, i t increased substantially in absolute terms. The Government has paid more attention to the improvement in living standards and provision of insurance for those who have no capability to work, for the elderly and all children under 6 years old. For pensioners receiving payment from the State Budget, the payments have been increased at the same rate as civil servants, and even faster in certain cases. 3.25 The category "Other" showed a declining trend, but with wide fluctuations from 29 percent in 2000 to 22 percent in 2001. This category includes expenditure on military, security and other related functions. The Government has announced its intention, with effect from 2005, to make the details of these "other" expenditures transparent and public. HOW PRO-POOR WAS THE COMPOSITION OF PUBLIC SPENDING? 3.26 Vietnam i s committed to a strategy of poverty reduction. Astonishing progress has been made. Over the period 1993-2002, the proportion of people below the poverty line has fallen from 58 percent to 29 percent. B y 2004, the Government estimates, this proportion has been reduced to 26 percent. Thus, Vietnam has lifted perhaps 25 million people out of poverty in less than a decade. Public expenditure allocation has played an important role in this achievement. 3.27 The extent to which public spending i s "pro-poor'' i s a complex question. In order to assess to what extent public spending i s resulting in a transfer of resources towards 36 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment poor households, it i s necessary to consider not only the composition of public expenditure, which i s the subject of this chapter, but also the effectiveness of public spending within each sector (the subject of later chapters). Nevertheless, it i s clearly the case that pro-poor budget allocations are an important step towards pro-poor service delivery. 3.28 Regarding the allocation of public spending and poverty reduction, the Government has three important tools. The first i s the mainstream budget transfer. The second i s the allocation of State Investment. The third i s the National Target Programs. We consider each of these tools in turn below. Budgettransfers 3.29 Inthe increasingly decentralized context of Vietnam, a key question to consider is whether resources are reaching the poorest provinces. As Chapter 6 discusses, the budget transfers to provinces (or revenue remittances from richer provinces to the State Budget) are based on an assessment of the gap between the revenues a province i s likely to generate and its recurrent spending needs. Recurrent spending needs are aggregated up on the basis of transfer norms for key sectors. The Ministry of Finance has made an ongoing effort to make these pro-poor. For example, in 2003 Government Decision No. 139/2003/QD-TTg was issued, requiring the use of the number of children of school age to calculate education norms, rather than the number of children enrolled in school. Since a lower proportion of school age children are enrolled in school in poorer provinces, this benefited poor provinces disproportionately. Similarly, this Decision establishes that the educational expenditure norm for a school age person living in a mountainous or remote area shall be 1.7 times higher than that for an urban area; and that the health expenditure norm for shall be 1.8 times higher. transfers resources from richer Figure 3.5: Budget Transfers and Poverty across to poorer provinces. This i s Provinces in2002 illustrated in Figure 3.5, which i s taken from the Vietnam Percapita transfer (000 VND) Development Report 2004 (VDR 2004). The horizontal axis in this figure shows the provincial poverty rate, based **** on the 2002 Vietnam -1,500 - I Household Living Standards 1 I Survey (VHLSS).The vertical axis measures the size of the I * * * net transfer per person from the ,... 0.0 20.0 40.0 60.0 80.0 central government (or Poverty rate ("4) negative transfer back to central government, through Source: Vietnam Development Report 2004, based on datafrom 2002 VHLSS Trends in Composition of Public Expenditure 31 collected by each province, including revenue collected on behalf of the central government, and actual provincial spending. It provides a reasonable approximation to the actual transfer when provincial budget deficits or surpluses are not too large. 3.31 These calculations suggest that three of the richest provinces-Hanoi, H o Chi Minh City and Haiphong-transfer more than three million VND per person each. But most provinces, and especially the poorest ones, receive a net transfer. In some cases, the amounts can be considerable. Bac Can province, the most favored, receives the equivalent of 1.19 million VND per person per year. As VDR 2004 concluded, this i s not too far from the food poverty line of 1.38 million VND per person per year, and amounts to roughly 62 percent of the full poverty line of 1.92 million VND per person per year. 3.32 This generally encouraging story needs to be qualified in a number of ways. First, the progressive nature of the transfer system is in part simply compensating for some regressive features in the apportionment of revenues between provinces. In particular, the treatment of VAT and corporate income tax as shared taxes, with revenues shared with the provincial jurisdiction where these revenues are actually collected, creates some serious equity problems which the transfer system needs to work hard to offset. This issue i s discussed further in Chapter 6. 3.33 Second, although transfers to and from provinces are based on spending norms, in practice provinces have considerable discretion as to how the resources they obtain are distributed-both between districts and communes within the province and between sectors. As Chapter 6 discusses further, provinces set their own policies for assigning revenues and expenditures between districts and communes and data for examining the extent to which this i s pro-poor remains weak. 3.34 At the sectoral level, education and health are generally thought to be sectors with disproportionate potential to benefit the poor. Evidence on spending by function, discussed above, suggests that in aggregate these sectors have increased their budget share (in the case of education) or else held it constant (in the case of health). Nevertheless, data i s poor as to whether or not this i s true in the poorest provinces, districts and communes. These issues are discussed further in the sectoral chapters of this PER-FA. State investment 3.35 The pattern of state investment expenditure presented in Figure 3.6, once again taken from VDR 2004, shows the average State investment per capita at the provincial level in 2000 (the most recent data), relating it to the provincial poverty rate from the 2002 VHLSS. There i s a striking contrast between Figure 3.6 and Figure 3.5 on recurrent budget transfers per capita. Net transfers for recurrent expenditures are skewed towards to poorest provinces, whereas resources for investment are skewed towards the richest ones. The correlation between budget transfers per capita and State investment per capita i s -0.72. 3.36 Clearly, it could be argued that this distribution of State investment i s "pro- growth"-and that in the longer term economic growth will benefit the poor. Infrastructure might have a smaller impact on economic growth in mountainous or remote areas than in more developed urban areas. Against this, it might be argued that 38 Vietnam Public Expenditure Relie\\' and Integrated Firluciun Assessrnent poorer regions need a greater share of State investment, to enable them to catch up and to promote inclusive development. Investment issues are discussed further in Chapter 8. Figure 3.6: State Investment and Poverty Rates across Provinces 1IPer capita investment (000 VND) 4,000 I 1 *** 2,000- ** * * * * 1,000 - '* 6. *.*A *** **$$ 8 + 8 t * * . * * 04 0.0 20.0 40.0 60.0 80.0 Poverty rate Source: Vietnam Development Report 2004, constructed using datafrom GSO and the 2002 VHLSS NationalTarget Programs 3.37 A third mechanism used by Government for directing public expenditure to the poorest localities and households are the National Target Programs. These are streams of funding over and above the norm-based allocation, intended to support and ensure the delivery of Government's national targets at local level. Many, though not all, of these targets relate to poverty reduction, including the Hunger Eradication, Poverty Reduction and Job Creation program (HEPR-JC) and Program 135 for particularly impoverished and remote communes. 3.38 These NTPs play an important role in addressing pockets of extreme poverty. However, it i s important to recognize the funding through these programs i s very small in comparison with main funding, and as such their impact i s necessarily limited. As i s common with targeted poverty programs around the world, not all the households identified as poor are included in the list of HEPR beneficiaries. As Table 3.1 details, analysis using the Vietnam Household Living Standards Survey (VHLSS) for 2002 suggests that for the main components of HEPR, mis-targeting (defined as the fraction of non-poor households among beneficiaries) stands at around a quarter (see Table 3.1). This i s not too high by international standards. However, coverage (defined as the fraction of poor households who benefit from the program) varies considerably across components. A significant proportion of poor households (between 12 and 20 percent, depending on the method of estimation) have access to exemption of school fees. However, little over 5 percent of poor households had access to subsidized credit in 2002. Trends in Composition of Public Expenditure 39 Table 3.1: Targeting Effectivenessof HEPRin2002 Coverage (percent) Mis-targeting (percent) Component (A) (B) (A) (B) Credit for the Poor 5.8 5.7 25.1 33.1 Health Care Certificate 2.6 27.3 Health Insurance 9.9 9.8 28.6 38.1 Educationfor the Poor 20.0 12.1 30.2 37.7 Note: (A)and (B) are two alternative estimates. Both (A)and (Bj are estimated using datafrom the 2002 VHLSS. Source: MOLISMUNDP (2004)for column A; Paul Shaffer and Nguyen Thangfor column B. Povertydata and targeting 3.39 One serious concern that emerges with respect to the poverty focus of public expenditure refers to the data used to underpin expenditure decisions. If those who are allocating resources are to take poverty into account while malung those decisions, it i s necessary that they have access to good data on poverty. For decisions made at the national level, this requires that the Ministry of Finance and the Ministry of Planning and Investment have reliable estimates of the numbers of poor people in different provinces and that they have some sense of the depth of their poverty. Similarly, if provincial People's Committees are to allocate resources across districts in a way that ensures the needs of the poor are catered to, then officials at this level need access to credible data which tells them where the poor are living. They also need the skills to use this data to plan and budget for growth and poverty reduction. 3.40 The production of the official poverty data of the Government i s currently under review. Collected by the Ministry of Labor, Invalids and Social Affairs (MOLISA), the official poverty data currently produced have a number of limitations when used as a basis for planning or for analyzing the poverty focus of expenditure (VDR, 2004). MOLISA i s currently calculating a new poverty line which will produce estimates that are likely to give a more accurate picture of poverty in Vietnam. MOLISA i s also piloting new methodologies to update the poverty figures on a regular basis. Used in conjunction with the Vietnam Household Living Standards Survey (VHLSS) data and poverty mapping techniques, this work should offer opportunities for MOF, MPI, sector ministries and provinces to have access to more reliable data about the distribution of the poor population across the country. RECOMMENDATIONS 3.41 Regarding the composition of expenditure, the PER-FA makes the following recommendations : 0 A better balance should be struck between capital and recurrent expenditure. Attention should be shifted from the quantity to the composition, quality and sustainability of capital expenditure. The need for balance between capital and recurrent expenditure should be recognized, with value for money being the guiding principle in allocating between capital and recurrent spending in each sector. Inthe view of the World Bank, the Government should now end the policy of increasing capital expenditure faster than recurrent expenditure. 40 Vietnam Public Exvenditure Review and Intenrated Fiduciary Assessment 0 Efforts should be made to create an enabling environment for private investment, rather than relying excessively on public investment to generate growth. Public investment in SOEs should be curtailed as SOEs are equitized. Options for increasing private finance in the provision of public assets should be chosen where analysis suggests this i s most cost-effective. 0 Recurrent expenditure will need to take an increased share of the budget in some sectors. However, care should be taken to ensure that increases in the level of wages and salaries do not end up squeezing out other operations and maintenance. 0 Planning of pay and employment reform needs to be strengthened and integrated. Wage increases need to be financed through implementation of a clearer strategy to shift non-core government employment to the private sector. 0 A sustained effort should be made to bring the functional and economic allocation of expenditure closer into line with national and sector-level strategies. This effort should be based on the policy analysis in the aggregate Medium-Term Fiscal Framework (MTFF) and sector and province-level Medium-Term Expenditure Frameworks (MTEFs), rather than on arbitrary and inflexible rules on budget shares for priority sectors. 0 Once finalized, Government agencies should be encouraged to use new MOLISA and GSO poverty data to assess whether resources are allocated in line with targets to reduce poverty and promote growth. As more disaggregated expenditure data becomes available for analysis, agencies should be encouraged further to investigate how allocations are benefiting various sub-groups of the population. Trends in Composition of Public Expenditure 41 Appendix 3.1: Composition of State Budget Expenditure, Final Accounts, 1997 Budgetexpenditurebyeconomictypesandbysector,bycentral andlomlgovernment 1997 State BudgetEpenditure StateBudget Expmditure Current Emenditwe Total capital of LGhich' Central Local Expenditure m d i t u r e **Total Salariesand o&M Wages 70,749 21,924 48,825 21,580 16,589 42710 28,039 Total inBillion Dong Agriculture, Forestry,Irrigation 3,712 2,709 1,003 141 694 2,071 1,641 Fisheries 103 60 42 8 33 42 60 Transport,Storage&Ccim-unicE 6,890 6,150 740 112 578 4,136 2,754 Industry 2,533 2,397 186 2 135 1,399 1,183 - oectriaty 514 498 15 0 3 227 287 -Water 699 694 5 1 4 23 676 Mucation&Traing 9,979 2,212 7,767 4,308 2,270 2,340 7,638 Health 4,329 925 3,404 939 2,308 1,489 2,839 sodal Insurance 9,372 206 9,166 4,037 601 8,547 825 Culture& Sports 2,578 1,374 1, a 4 253 883 1,211 1,367 Science,Technology & Environm 890 259 651 112 525 700 193 MMnistrationewnditure 7,138 0 7,138 1,583 5,555 NA NA Interest payn-tmt 1,916 0 1,916 0 0 Others 21,260 5,652 15,608 10,086 3,006 Shm?in totalsectoral Shareof totalbyeaxlomctype, % expenclitum,% Total inBillion Dong 100.0 100.0 100.0 100.0 100.0 60.4 39.6 Agriculture, Forestry, Irrigation 5.2 12.4 2.1 0.7 4.2 55.8 44.2 Fisheries 0.1 0.3 0.1 0.0 0.2 41.4 58.6 Transport, Storage& Ccim-uniw 9.7 28.1 1.5 0.5 3.5 60.0 40.0 Industry 3.7 10.9 0.4 0.0 0.8 54.2 - 45.8 Oectriaty 0.7 2.3 0.0 0.0 0.0 44.2 55.8 -Water 1.o 3.2 0.0 0.0 0.0 3.3 96.7 Education&Traing 14.1 10.1 15.9 20.0 13.7 23.5 76.5 Health 6.1 4.2 7.0 4.4 13.9 34.4 65.6 sodal Insurance 13.2 0.9 18.8 18.7 3.6 91.2 8.8 Culture&Sports 3.6 6.3 2.5 1.2 5.3 47.0 53.0 Science,Techndogy& Environm 1.3 1.1 1.3 0.5 3.2 78.6 21.4 Mm'nistrationexpenditure 10.1 0.0 14.6 7.3 33.5 NA NA Interestpayment 2.7 0.0 3.9 0.0 0.0 Others 30.0 25.8 32.0 46.7 18.1 Source:Mnisttyof Finance Note: `not includingewditure for Social support,pricesubsidies,additionaltransfersto the lower levds, interestpaymentandother wrrent emiture. ** lduding dataon major repair. NA Datanotavailable - 42 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment Appendix 3.2: Compositionof State BudgetExpenditure,FinalAccounts, 1998 Budgetexpenditurebyecomictypesand bysector,bycentraland localgovernment 1998 State Eudget Eqenditure StateBudget Eqenditure Current Emenditure Total capla of vhichf Central Local Eqmditure Eqenditure** Total Salariesand o&M Wages Total inBillion Dong 73,419 23,421 49,998 24,220 15,678 41,611 31,808 Agriculture, Forestry, Irrigation 4,591 3,493 1,099 168 708 2,578 2,014 Fisheries 484 453 30 6 19 426 58 Transport,Storage& C c " i c z 8,065 6,959 1,106 113 927 4,852 3,214 Industry 2,775 2,708 67 2 12 1,759 1,016 - oectriaty 422 394 29 0 4 61 362 -Water 612 604 7 1 4 227 385 Education&Traing 12,732 3,551 9,199 5,544 2,653 3,393 9,357 Health 5,207 1,342 3,864 1,060 2,553 1,930 3,277 Soaal Insurance 8,713 203 8,510 5,135 69 7,902 810 Culture&Sports 2,745 1,448 1,297 285 885 1,012 1,732 Science,Techndogy & Environm 1,067 353 714 111 583 836 232 AdMnistrationeqxmditure 6,741 0 6,741 1,743 4,998 NA NA Interestpayment 2,050 0 2,050 0 0 Others 18,231 2,911 15,320 10,054 2,272 Stmeintotalsectml shareof totalby e " i c type, % expenditure, % Total inBillionDong 100.0 100.0 100.0 100.0 100.0 56.7 43.3 Agriculture, Forestry,Irrigation 6.3 14.9 2.2 0.7 4.5 56.1 43.9 Fisheries 0.7 1.9 0.1 0.0 0.1 88.1 11.9 Transport, Storage& Corm-unicz 11.0 29.7 2.2 0.5 5.9 60.2 39.8 Industry 3.8 11.6 0.1 0.0 0.1 63.4 36.6 - oectriaty 0.6 1.7 0.1 0.0 0.0 14.4 85.6 -Water 0.8 2.6 0.0 0.0 0.0 37.1 62.9 Education&Traing 17.4 15.2 18.4 22.9 16.9 26.6 73.4 Health 7.1 5.7 7.7 4.4 16.3 37.1 62.9 sodal Insurance 11.9 0.9 17.0 21.2 0.4 90.7 9.3 Culture&Sports 3.7 6.2 2.6 1.2 5.6 36.9 63.1 Science,Techndogy& Environm 1.5 1.5 1.4 0.5 3.7 78.3 21.7 Adm'nistrationexpenditure 9.2 0.0 13.5 7.2 31.9 NA NA Interestpayment 2.8 0.0 4.1 0.0 0.0 Others 24.8 12.4 30.6 41.5 14.5 Souroe:Mnistryof Finance Note: 'not includingeqxmditurefor sodal support,pricesubsidies,additionaltransferstothe lowerlevels,interest paymentandother current expenditure.** lndudingdataon major repair. NA Eatsnot adlable - Trends in Composition of Public Expenditure 43 Appendix 3.3: Composition of State Budget Expenditure, FinalAccounts, 1999 84,817 51,551 TotalinBillionDong 33266 23,275 161% 3,w PgricuRue,Forestry,lnipticn 5,051 4,070 981 172 579 2,337 2,714 fisheries 275 239 46 8 31 181 94 Trarqmt, Storage&CcrrmmiE 10,615 9,892 723 29 605 5,903 4,715 1-v 3,499 3,403 90 4 19 1,995 1,504 - oectriaty - water 645 606 39 1 2 63 582 627 618 10 1 3 20 608 Mucation&Traing 13,232 3,429 9,803 5,883 2,545 2,558 10,674 wth 5,912 2,092 3 , a 1,108 2,290 2,144 3,768 M a l Insurance 10,072 217 9,= 6,w3 505 9,035 1,056 Qltue &Sprk 2,w 1,588 1,405 316 953 1,017 1,976 science,Tednology& Envirm 1,113 350 763 131 593 824 288 Wrrinistraticne w t u e 8,743 2,263 6,481 2,357 2,517 N4 N4 Interestpaymrn 2,a 0 2,327 0 0 ahers 20,985 5,727 15,258 7,234 5,564 slm?intdalsedasll s l m ? o f t o t d t y ~ C t Y p e , % sfplt3tLtq% TotalinBillionDong 103.0 1m.o 1a.o 103.0 103.0 54.0 46.0 Pgricuttue,Forestry,Inigaticn 6.0 122 1.9 0.7 3.6 46.3 53.7 fisheries 0.3 0.7 0.1 0.0 0.2 65.8 34.2 Trarspoct,Storage&Comnnia 125 29.7 1.4 0.1 3.7 56.6 44.4 Industry 4.1 10.2 0.2 0.0 0.1 57.0 43.0 - Bectriaty 0.8 1.8 0.1 0.0 0.0 9.7 90.3 -water 0.7 1.9 0.0 0.0 0.0 3.1 96.9 Mucation&Traing 15.6 10.3 19.0 25.3 15.7 19.3 80.7 wth 7.0 6.3 7.4 4.8 14.1 36.3 63.7 MalInsurance 11.9 0.7 19.1 25.9 3.1 89.4 10.6 Cultwe&SpMs 3.5 4.8 2.7 1.4 5.9 34.0 66.0 science,Tednolcgj &Envirm 1.3 1.1 1.5 0.6 3.7 74.1 25.9 Pm-inistraticne p x i t u e 10.3 6.8 12.6 10.1 15.5 N4 N4 Interestpaymrn 2.7 0.0 4.5 0.0 0.0 ahers 24.7 17.2 29.6 31.1 34.3 Sxm:Mnktryoffirrance M e : *rdincllwlnge w t u r efor sodals w , pricesbides,additional transferstothelmerI d s ,interestp a warddher clrrent e w t u e . Iddingdatacn@or r@r. N4-M ami awlable 44 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment Appendix 3.4: Compositionof State BudgetExpenditure,Final Accounts, 2000 Budgetexpenditurebyeconomic types andbysector, bycentraland localgovernment 2000 State Budget Expenditure State Budget Expenditure Current Expenditure Total Capital of hhich* Central Local Expenditure Expenditure** Total Salaries and O&M Wages Total inBillionDong 103,151 35,106 68,045 29,523 23,881 58,069 45,082 Agriculture, Forestry, Irrigation 5,647 4,436 1,212 249 665 1,759 3,889 Fisheries 157 106 51 12 34 50 107 Transport, Storage & C o m n i a 11,375 10,055 1,320 37 1,156 6,391 4,984 Industry 3,791 3,703 88 5 28 2,143 1,647 - Electricity 799 782 18 1 1 127 672 -Water 1,248 1,219 28 2 10 136 1,112 Education &Traing 16,344 4,470 11,874 7,707 2,591 2,798 13,546 Health 6,550 1,726 4,824 1,650 2,264 1,327 5,223 soclal Insurance 11,385 210 11,175 7,905 113 9,901 1,484 Culture&Sports 3,577 1,963 1,614 403 1,029 1,184 2,394 Science, Technology & Environm 1,423 482 941 140 758 1,046 376 Adm'nistration eyxnditure 9,228 2,481 6,747 3,094 3,215 NA NA interestpayment 3,514 0 3,514 0 0 Others 30,160 5,474 24,686 8,323 12,029 Sharein totalsectoral Shareof totalby emomic typ, % expenditure, % TotalinBillion Dong 100.0 100.0 100.0 100.0 100.0 56.3 43.7 Agriculture, Forestry, Irrigation 5.5 12.6 1.8 0.8 2.8 31.1 68.9 Fisheries 0.2 0.3 0.1 0.0 0.1 32.0 68.0 Transport, Storage & C o m n i a 11.0 28.6 1.9 0.1 4.8 56.2 43.8 Industry 3.7 10.5 0.1 0.0 0.1 56.5 43.5 - Electricity 0.8 2.2 0.0 0.0 0.0 15.9 84.1 -Water 1.2 3.5 0.0 0.0 0.0 10.9 89.1 Education &Traing 15.8 12.7 17.5 26.1 10.9 17.1 82.9 Health 6.3 4.9 7.1 5.6 9.5 20.3 79.7 social Insurance 11.0 0.6 16.4 26.8 0.5 87.0 13.0 Culture&Sports 3.5 5.6 2.4 1.4 4.3 33.1 66.9 Science, Technology & Environm 1.4 1.4 1.4 0.5 3.2 73.6 26.4 Adm'nistration expenditure 8.9 7.1 9.9 10.5 13.5 NA NA Interest payment 3.4 0.0 5.2 0.0 0.0 Others 29.2 15.6 36.3 28.2 50.4 Source:Ministryof Finance Note: 'not includingexpenditure for social support, pricesubsidies,additional transfers to the lower levels, interest payment and other current expenditure. ** Includingdataon major repair. NA Datanot available - Trends in Composition of Public Expenditure 45 Appendix 3.5: Compositionof State BudgetExpenditure,Final Accounts, 2001 Budgetexpenditurebyeconomictypesand bysector, bycentral and localgovernment2001 StateEudgetExpenditure StateBudget m d i t u r e Current Wnditure Total Capital of hhich* Central w Expenditure Expenditure** Total Salariesand o&M Wages Total inBillion Dong 119,403 44,319 75,084 35,931 20,906 a,= 56,043 Agriculture,Forestty, Irrigation 7,044 5,653 1,390 292 700 1,808 5,236 Fisheries 376 300 76 15 52 102 274 Transport, Storage& Communica 14,991 13,588 1,403 41 1,037 6,589 8,402 Industry 6,605 6,483 123 5 43 4,658 1,947 - Electricity 942 930 12 1 1 360 582 -Water 535 527 9 2 6 14 522 Education&Trahg 19,505 5,357 14,148 9,231 2,191 3,159 16,346 Health 8,475 2,888 5,586 2,187 2,831 2,143 6,332 social Insurance 13,958 267 13,691 9,730 160 12,179 1,779 Culture&Sports 4,642 2,776 1,866 461 1,216 1,622 3,021 Science, Technology& Envirmm 2,579 1,179 1,400 357 974 1,985 595 Administrationexpenditure 10,973 1,944 9,029 3,684 4,155 NA NA Interestpayment 4,485 0 4,485 0 0 Others 25,770 3,883 21,888 9,928 7,547 share in totalsectoral shareof totalbyeconomictype, % expenditum,% Total inBillion Dong 100.0 100.0 100.0 100.0 100.0 53.1 46.9 Agriculture,Forestry, Irrigation 5.9 12.8 1.9 0.8 3.4 25.7 74.3 nsheries 0.3 0.7 0.1 0.0 0.2 27.1 72.9 Transport,Storage& Communics 12.6 30.7 1.9 0.1 5.0 44.0 56.0 lndustly 5.5 14.6 0.2 0.0 - Electricity 0.2 70.5 29.5 0.8 2.1 0.0 0.0 0.0 38.2 61.8 -Water 0.4 1.2 0.0 0.0 0.0 2.6 97.4 Education&Traing 16.3 12.1 18.8 25.7 10.5 16.2 83.8 Health 7.1 6.5 7.4 6.1 13.5 25.3 74.7 soclal Insurance 11.7 0.6 18.2 27.1 0.8 87.3 12.7 Culture& Sports 3.9 6.3 2.5 1.3 5.8 34.9 65.1 Science,Technology& Environm 2.2 2.7 1.9 1.0 4.7 76.9 23.1 Administrationexpenditure 9.2 4.4 12.0 10.3 19.9 NA NA Interestpayment 3.8 0.0 6.0 0.0 0.0 Others 21.6 8.8 29.2 27.6 36.1 Source:Mnisttyof Finance Note: `not includingexpenditurefor social support, pricesubsidies,additionaltransferstothe lower levels,interest paymentandother current expmditure.** Includingdataon major repair. NA Datanotaw'lable - 46 Vietnam Public Expenditure Review and Intenrated Fiducian, Assessment Appendix 3.6: Compositionof StateBudgetExpenditure,FinalAccounts, 2002 Budgetexpenditurebyeconomictypesandbysector,bycentraland localgovernment2002 State Budget Expenditure State Budget Expenditure Current Expenditure Total Capital of Mich* Expenditure Expenditure** Central Local Total Salaries and O&M Wages Total inBillionDong 135,490 54,924 80,566 37,050 22,758 70,917 64,573 Agriculture, Forestry,Irrigation 6,993 5,352 1,641 292 892 1,454 5,539 Fisheries 478 384 94 16 60 104 373 Transport, Storage & Comnica 18,720 17,389 1,331 42 907 8,305 10,415 Industry 5,690 5,570 121 9 44 3,967 1,723 - ElKtriUty 543 534 10 1 1 21 522 -Water 416 400 15 2 12 5 411 Education&Traing 22,596 6,047 16,548 10,317 3,889 3,310 19,285 Health 8,616 2,013 6,603 2,288 3,510 1,997 6,619 social Insurance 13,177 425 12,752 8,603 199 11,198 1,979 Culture & Sports 5,140 2,724 2,416 522 1,383 1,679 3,460 Science,Technology & Endronm 2,949 1,327 1,622 260 1,297 2,346 602 Addnistration expenditure 12,478 3,855 8,623 3,945 3,695 NA NA Interest payment 6,008 0 6,008 0 0 Others 32,645 9,837 22,807 10,754 6,881 Sharein totalsectoral Shareof totalby economic e,% expnditure, % Total inBillionDong 100.0 100.0 100.0 100.0 100.0 52.3 47.7 Agriculture, Forestry, Irrigation 5.2 9.7 2.0 0.8 3.9 20.8 79.2 Fisheries 0.4 0.7 0.1 0.0 0.3 21.9 78.1 Transport, Storage & Comnica 13.8 31.7 1.7 0.1 4.0 44.4 55.6 Industry 4.2 10.1 0.1 0.0 0.2 69.7 30.3 - ElKtridty 0.4 1.o 0.0 0.0 0.0 3.8 96.2 -Water 0.3 0.7 0.0 0.0 0.1 1.I 98.9 Education&Traing 16.7 11.0 20.5 27.8 17.1 14.6 85.4 Health 6.4 3.7 8.2 6.2 15.4 23.2 76.8 social Insurance 9.7 0.8 15.8 23.2 0.9 85.0 15.0 Culture & Sports 3.8 5.0 3.0 1.4 6.1 32.7 67.3 Science,Technology & Environm 2.2 2.4 2.0 0.7 5.7 79.6 20.4 Administration expenditure 9.2 7.0 10.7 10.6 16.2 NA NA Interest payrent 4.4 0.0 7.5 0.0 0.0 Others 24.1 17.9 28.3 29.0 30.2 Source: Ministry of Finance Note: "not includingexpenditure for sOciai support,pricesubsidies, additional transfersto the lower levels, interest payment and other current expenditure. ** Includingdataon major repair. NA Datanot available - 4. INSTITUTIONS FOR PUBLICEXPENDITURE MANAGEMENT INTRODUCTION AND OVERVIEW 4.1 Good public expenditure management plays a central role inpro-poor, pro-growth service delivery; bad public expenditure management can do much to undermine it. The budget should be a financial mirror of a society's economic and social choices, and budget execution should adhere closely to the budget plan. Without strong public expenditure management, however, budget plans are unlikely to allocate resources responsively, efficiently and effectively inline with national plans and strategies. 4.2 Over the past decade, Vietnam has steadily but surely improved its institutions for public expenditure management. The legal and institutional foundations for sound budgeting have been put in place and institutions for both budget planning and budget execution have been considerably strengthened at every level of Government. Nevertheless, the Government has identified great scope for further improvement and has put inplace a Public Financial Management Reform Initiative to support further reforms. 4.3 The first part of this chapter discusses the legal and institutional foundations for budgeting in Vietnam. The second part considers budget planning. The third part considers budget execution. Recommendations are presented at the end of each part. Wider issues in public expenditure management-institutions for financial accountability and transparency; decentralization to subnational governments; delegation to spending units; public investment management; and public procurement-are the topics of subsequent chapters. LEGAL INSTITUTIONALFOUNDATIONSBUDGETING AND FOR 4.4 The legal and institutional foundations for budgeting have been considerably strengthened in recent years. The budget has considerable legitimacy, conferred first by the 1992 Constitution and subsequently by the State Budget Law. The State BudgetLaw 4.5 The 1996 State Budget law was the first legal document in Vietnam to provide explicit regulation for public financial management. The law defined the role and powers of each level of government in the budget process, reflecting the incipient decentralization of public finance. The law created a framework for financial relations between the national government and lower tiers of government. 4.6 A revised State Budget Law was approved by the National Assembly in December 2002 and came into force in January 2004. This Law builds on the 47 48 Vietnam Public Expenditure Review and Integrated Fiducialy Assessment achievements of its predecessor, strengthening the legal platform for budget management infour importantregards: 0 Clarifying powers and responsibilities: The State Budget Law gives the National Assembly the power to approve the most important aspects of the State Budget plan, including at subnational level; the allocations of central budget to ministries and agencies; and the equalization and special purpose transfers to each province/city. In addition, the new law further clarifies the division of responsibilities between the Ministry of Finance, the Ministry of Planning and Investment and sector ministries. MOF will be the lead ministry responsible for fiscal planning, budget formulation and debt management. MPI remains the lead ministry for capital budget formulation. Within the MOF, the State Treasury Department will be the lead agency responsible for budget execution and for government budget management information. Strengthening decentralization: The new law further clarifies the roles and responsibilities of central and subnational tiers of government. In particular, provinces are given more explicit powers and duties to prioritize resources, including determining allocations to different sectors and transfers to lower tiers. They also are given more explicit powers and duties to mobilize resources, with some taxes that were assigned to central government now shared. Local planning i s beingpromotedthrough the agreement of transfers from the center to provinces for stable periods of three to five years. Promoting administrative reform: A major effort i s made to simplify and strengthen financial management and reporting requirements. Spending units at every level are given greater budget flexibility, with a reduction in the number of controlled blocks of expenditure from nine to four. Duplicative and low value- added stages in the budget formulation and allotment process are eliminated or consolidated. 0 Strengthening transparency and accountability: The new law consolidates and builds on recent efforts to promote fiscal transparency and accountability. It mandates that budget plans, final accounts and audit reports should be publicized. Budget coverage 4.7 B y the standards of many developing countries, budget coverage in Vietnam i s reasonably comprehensive. The major expenditures of all spending units of government at central, province and district level are reflected in the State Budget, as are the bulk of the expenditures at commune level. Nevertheless, budget coverage i s incomplete and needs to be expanded to comply with internationally accepted norms. 4.8 As pointed out in the 2000 PER, fees and contributions paid by citizens are not adequately covered in official budget documents. This raises important questions regarding transparency and accountability and impairs an appropriate assessment of the equity of the fiscal system. Over the review period, progress has been made: the 2001 Ordinance on Fees and Charges helped systematize which kinds of fees can be collected at which level and which kinds of fees must be reflected in the State Budget. Decree 24/1999/ND-CP dated April 16, 1999 further clarifies the rules regarding which types of lnstitirrionsfor Public ExpenditureMuriugeinent 49 contributions must be reflected in the State Budget and which types of contribution can be kept off-budget (for example, community contributions in labor and kind to local construction works). The 2002 State Budget Law formally legalized the requirement that all revenues collected at the commune level have to be fully reflected in the budget. However, there remain weaknesses in implementation at local level, reflecting weak accounting systems and processes, as well as weak accountability and transparency institutions. 4.9 Another source of incompleteness in budget coverage emerges from the limited availability of information regarding donor financing. Despite harmonization efforts, most donors have their own disbursement mechanism, and sometimes more than one. Project Management Units often do not inform the Ministry of Finance about their donor- financed disbursements. Consequently, Government has extremely limited information regarding grants directly executed by donors. This lack of information on donor projects in various sectors constrains ability to plan and coordinate total resource allocation6. From a macro-fiscal management perspective, the major off-budget items include the Development Assistance Fund (DAF) and the new off-budget bonds (discussed in Chapter 2). From a public expenditure management perspective, unconsolidated user charges and off-budget donor-funded projects are also a significant concern. In recent years, the Government's strategy of raising off-budget bonds for infrastructure and education projects has significantly increased the proportion of the government's operations which are off-budget. Institutionalresponsibilities National Assembly and People's Councils 4.10 The 2002 State Budget Law substantially increases the powers of the National Assembly in Public Expenditure Management. The plenary of the National Assembly i s given the power not only to approve the overall budget (total revenue and expenditure), but also its composition and, in particular, the allocations to sector ministries, central agencies and the transfers to sub-national governments. Prior to 2002, the National Assembly would only approve total revenues and total expenditures while budget allocation would be discussed and decided upon in the standing committee of the National Assembly without involving the plenary. 4.11 The 2002 law thus increases the degree of involvement of individual parliamentarians in the budgetary process and the breadth and transparency of discussion on budget allocations. The members of the National Assembly are increasingly questioning ministers during their appearances at the legislature and proposing amendments to proposed legislation, including budgets. For example, members have raised questions on the efficiency of capital expenditure as well as regarding important issues like education. Moreover, during the last budget cycle, the National Assembly significantly increased the policy focus and allocations devoted to priority programs such as Program 135 to support disadvantaged communes. The dialogue between the sector ministries and the National Assembly has also increased. The International Federation of Accounting Public Sector Committee is currently drafting a standard that will provide guidance on the treatment of donor contributions in accounts. 50 VietnamPublic Expenditure Review and Integrated Fiduciary Assessment 4.12 Provincial People's Councils have also strengthened their role in public expenditure management, gaining responsibility from the center of government and gaining greater responsibility over lower tiers. The 2002 budget law also increases the powers of the legislative bodies at the provincial level, the provincial People's Councils. Provincial People's Councils now have the power to determine the allocation of the budget between the three tiers of government at the sub-national level-province, district and commune. They can also decide on revenue and expenditure, including on types of fees and contributions that can be collected at the local level and on the financial norms to be followed by the various levels of sub-national government. Finally, the new legislation allows provinces to borrow domestically under conditions stipulated by central Government but without further approval. People's Councils, however, remain weak in terms of their accountability and their capacity to perform the tasks assigned to them in the 2002 State Budget Law. Dual subordination 4.13 Vietnam has a unitary system of government, and the National Assembly remains ultimately responsible for approval of the unitary "State Budget". Nevertheless, the four tiers of government-central, provincial, district and commune-have significant and increasing authority over public expenditure in their locality. A system of "dual subordination" exists, whereby provincial departments, district bureaus and communes report both horizontally across to their local People's Council and vertically up to the relevant parent sector ministry. The reporting chain has each level of government reporting up to the next, so that communes report to districts, districts report to provinces and provinces report to central government-a system sometimes referred to as the "matrouchka system". 4.14 The 2002 State Budget Law eliminates certain aspects of dual subordination: provincial People's Councils are given more explicit powers and duties to prioritize resources, including determining allocations to different sectors and transfers to lower tiers. Nevertheless, there remains some confusion and overlap, with both the National Assembly and the provincial People's Councils entitledto approve the provincial budget. Furthermore, a wide range of "physical norms" continue to be set by sector ministries for their reporting departments and bureaus at local level, with clear budgetary implications. The 2002 State Budget Law makes clear for the first time that such norms are no longer mandatory, but this i s not yet widely understood. Dual budgeting 4.15 Vietnam has a single "State Budget", and the Ministry of Finance i s named in the 2002 State Budget Law as the lead agency responsible for overall budget preparation and allocation. Nevertheless, a system of "dual budgeting" exists whereby the Ministry of Finance (or Department of Finance at subnational levels) prepares the fiscal framework and the recurrent estimates, and the Ministry of Planning and Investment (or Department of Planning and Investment at provincial level) prepares the Public Investment Program and the investment budget, including capital expenditures and donor funded projects. 4.16 Many countries around the world have the "dual budgeting" system. However, in other countries i t i s often associated with poor coordination of capital and recurrent Institutionsfor Public Expenditure Management 51 spending. In Vietnam, as Chapter 3 illustrated and sector chapters further detail, there are serious and growing imbalances between capital and recurrent spending. For example, in the education sector, there i s probably too much recurrent spending relative to capital (with more money for primary school construction a greater priority than additional teachers). Conversely, in the transport sector and the agricultural sector, there i s probably too much capital spending relative to recurrent (with completion, maintenance and rehabilitation of existing assets a more cost-effective strategy than new construction projects). 4.17 If a "dual budgeting" system is to be sustainable, many people believe, Vietnam will need to strengthen the mechanisms for coordination between plans and budgets; between the recurrent side of the budget and the capital side; and between Finance and Planning and Investment functions at each level of government. Such mechanisms in other countries include Inter-Ministerial Working Groups andjointly sponsored Medium- Term Expenditure Frameworks (MTEFs). Vietnam i s committed to introducing such approaches on a pilot basis. Recommendations 4.18 Regardingbudget coverage, the PER-FA recommends the following: a Budget coverage should be expanded to comply with internationally-accepted norms. In particular, budgets should include all fees and contributions, and all government debt, including on-lending and the currently off-budget bonds. Budgets should also include as detailed as possible an estimation of guaranteed debts and other contingent liabilities. a Donors should be encouraged to accelerate their work on harmonization. The Government should encourage donors to channel aid through the State Treasury and should ensure that all such aid i s recorded on budget. STRENGTHENING STATE BUDGETAND INVESTMENT PLANNING 4.19 A number of steps are being taken to strengthen budget management so as to promote prioritization that supports the achievement of poverty reduction and growth goals. Coordinatingplanningof recurrentand investmentexpenditure 4.20 Much of the early progress in strengthening public expenditure management in Vietnam was focused on strengthening processes for the planning of public investment projects, coordinated by the Ministry of Planning and Investment. The Public Investment Program has often been seen as the key mechanism through which Vietnam's poverty reduction and growth objectives will be delivered. Socially inclusive development will require new investment and will have important implications for the allocation of capital expenditure. The government i s aware, however, that public investment management needs to be further strengthened (see Chapter 8); and that greater attention needs to be paid to strengthening public expenditure management on the recurrent side of the State Budget, given the great importance of recurrent expenditure in development. 52 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 4.21 Perhaps most centrally, the government believes that the planning of public investment and recurrent expenditure needs to be better coordinated. Vietnam currently has separate ministries of Finance and Planning & Investment; and separate budgets for recurrent and capital expenditures. In the absence of an integrated ministry and a unified budget, it i s imperative that strong steps are taken to coordinate the work of the two functions at central and provincial level, as well as within sector ministries and departments; and to coordinate planning of the recurrent part of the State Budget and the Public Investment Program. 4.22 Some such mechanisms already exist. Articles 21 and 23 of the 2002 State Budget Law'describe the respective roles of M P I and M O F in the budgeting process, and this i s further elaborated in Decree 60/2003/ND-CP, which outlines a formal division of responsibilities between planning and finance functions at central and local levels. However, the scope of Decree 60 i s basically to divide responsibilities between planning and finance, rather than establish any kindof mechanism for dialogue andjoint working. 4.23 The Government recognizes that mechanisms for dialogue and joint working are also necessary. To this end, it has recently established an Inter-Ministerial Working Group on Medium-Term Expenditure Planning, comprising senior officials of both M O F and MPI; and similar inter-ministerial groups are being established in four pilot sectors (Education, Health, Agriculture & Rural Development and Transport), comprising representatives of the finance and planning functions of the relevant sector ministry, plus the relevant sectoral expenditure specialists from M O F and MPI. Both M O F and MPI have been closely involved in the preparation of this 2004 PER-IFA. 4.24 The Government has initiated a program to pilot Medium-Term Expenditure Frameworks (MTEFs) in four sectors and four provinces. The piloting of sector-level MTEFs was a recommendation of the 2000 Public Expenditure Review. However, implementation has only started in the past 12 months, as part of the Government's new program of public financial management reform. 4.25 An MTEF involves coordinating capital and recurrent expenditure budgeting within a single, forward looking expenditure framework, with a single, forward-loolung macroeconomic framework and a single, forward-looking set of development goals (Box 4.1). MTEFs will help right the balance between capital and non-wage recurrent expenditure. They will help ensure fiscal realism in the planning of public investment. They will help bring policy analysis and development focus to a recurrent expenditure planning that remains overly driven by mechanistic norms. 4.26 The first of these pilots has been the Education sector. This pilot i s at an early stage. Work to date has focused on consolidating the government expenditure data on the sector at each level of government and on both the recurrent and capital sides of the budget; on developing simple, analytical tools for costing objectives and forecasting expenditure needs; and on building mechanisms for policy dialogue between MOET, M O F and MPI about sectoral goals, strategies, priorities and pressures and the implications of these for government spending. Institutions for Public Expenditure Management 53 Box 4.1: Objectives of a Medium-termExpenditureFramework Improving macroeconomic balance by developing a single, consistent and realistic resource framework. Improving the allocation of resources to strategic priorities between and within sectors by building budgets around a single, consistent and realistic set of policy objectives. Improving coordination and balance between capital and recurrent expenditures by integrating the planning for both of these within a single, forward-looking budgetary process. 0 Increasing commitment and predictability both in policy and funding so that ministries and provinces can plan ahead and programs can be sustained. Providing ministries and provinces with a hard budget constraint and increased autonomy, thereby increasing incentives for efficient and effective use of funds. 4.27 Nevertheless, as Box 4.2 outlines, powerful insights are emerging from the work which strongly vindicate the forward-looking integrated approach to recurrent and capital budget planning. It i s expected that the education MTEF work will become increasingly integrated with investment planning, State Budget formulation and political dialogue and that similar approaches will be taken in other sectors. Box 4.2: Preliminary Findingsfrom the Education Sector MTEFpilot The policy analysis and scenario modelling conducted as part of the education sector MTEF pilot has emphasised the need and opportunity to rationalise the teaching workforce. Falling school age populations and continuing recruitment have caused teacher utilisation to fall to low levels by international standards in the primary sector. Pupilcontact hours are also low in Vietnam by international comparisons. This is a difficult issue, but important to the efficiency of expenditures. It needs to be tackled on a number of fronts to bring about this rationalisation and free up further resources to contribute to improving the quality of service. Service standards remain low and variable. The MTEF work has identified the following as potential priority areas for focusing funds available for service improvement, basedon evidence of impact: Steps to accelerate movement towards full day working in schools, by freezing teacher numbers (on the recurrent side of the budget) and prioritizing primary school construction (on the capital side of the budget), which i s the binding constraint. Steps to accelerate upgrading of teacher skill levels. Bringing non-staff expenditures in schools across the country up to levels identified as necessary to achieve a quality of service and effectiveness consistent with national targets. As the buildingstock expands, taking measuresto provide for appropriate levels of maintenance. The introduction of regular, consistent and objective tests of educational attainment as a means of monitoring performance and managing improvement, 4.28 The Government has designed its MTEF pilots to take account of Vietnam's increasingly decentralized public expenditure management institutions (discussed further in Chapter 6). A key objective of the MTEF pilots is to strengthen public expenditure planning capacity at provincial level, and to strengthen the linkages between national sector-level planning and provincial planning. An effort will be needed to establish positive "vertical" linkages between resource planning at all levels (sector, ministry, province, district, spending unit) while respecting decentralization objectives and the need for appropriate flexibilities and incentives at every level. As part of the sector-level 54 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment MTEFs, it is envisaged that tools will be developed to assist provincial level budget planning, and to serve as a focus for constructive dialogue between national policy makers and provincial decision takers. Province level MTEFs will also be developed in four pilot provinces. It i s envisaged that this work will build gradually outwards from the norm process that that i s deeply embedded in existing systems to a system that, in time, i s more output rather than input oriented. 4.29 The Government has also initiated the piloting of a "top down" Medium-Term Fiscal Framework (MTFF), which i s expected to include aggregate expenditure projections, based on realistic and sustainable assumptions about economic growth, revenue collection and borrowing, as well as indicative totals for expenditure by province and ministry for both the current and capital sides of the budget. This will both inform and be informed by the "bottom up" sector- and province-level MTEFs. MTEFs in the three remaining pilot sectors are expected to get underway in early 2005, with provincial level pilots starting later in 2005. Insights from the education MTEF are already being reflected in the 2005 budget discussions; and the education MTEF should also be reflected in the Public Investment Program and in the forthcoming Socio-Economic Development Plan. 4.30 In the longer term, the pilot MTEFs may evolve into a "bridge" between the capital and recurrent sides of the budget, and between planning and budgeting processes. Alternatively, the Government may decide on the basis of the MTEF pilots to move beyond dual budgeting and establish a formal forward budget. In either eventuality, the forward looking and integrated policy analysis tools and skills being developed through these pilots have an important contribution to make. Recommendations 4.31 Regarding the coordination of recurrent and investment expenditure, the PER-FA recommends the following: The recently established Inter-Ministerial Working Group on Medium-Term Expenditure Planning should be given elevated status. Ministers and Directors should participate in meetings as well as senior officials. Implementation of the pilot MTFF and pilot MTEFs in four sectors and four provinces should be accelerated. In developing its new Treasury and Budget Management Information System (TABMIS-discussed further below), the Government should take steps to ensure this provides timely, accurate and immediately accessible data to MPI and provincial planning departments on project financial execution, classified using MPI'sproject codes. Currently, officials inthe planning functions spend too much of their scarce time simply trying to find out how much has been spent on different projects, leaving little time for monitoring of physical execution and impact. The Government should continue to explore options for selected merger of planning and finance functions at subnational and sectoral level. It i s notable that some of the finance and planning functions closest to the "front line" of service delivery are using the administrative flexibility granted to them to merge their Institutions for Public Expenditure Management 55 planning and finance functions. This i s observed in key sector ministries such as Health and Education and Training (where the unified finance and planning department has ledthe MTEFpilot); districts also have unified bureaus of finance and planning. CoordinatingBudgetingwith PlanningandPerformanceManagement 4.32 Successful public expenditure management cannot be divorced from social and economic planning. The budget should be a financial mirror of society's economic and social choices; at the same time, goals and strategies for poverty reduction and growth need to be realistic, and therefore devised with reference to opportunity costs and fiscal constraints. Coordinating the capital and recurrent sides of the budget i s an important step, but the goal should also be to coordinate budgeting with planning and performance management. 4.33 The Prime Minister has recently issued Directive 33 which initiates the preparation of the 2006-2010 Socio-Economic Development Plan (SEDP). This Directive calls for a more open, participatory planning process than has taken place to date. It also encourages agencies to frame their strategies around the achievement of desired outcomes and to focus less on inputsas targets. 4.34 Increasing emphasis has been put on the importance of "pro-poor'' and "pro- growth" policies, and on measuring the delivery of services and development outcomes. This is reflected in the sometimes quite specific performance indicators and targets established in the Vietnam Development Goals (VDGs), prepared as part of the CPRGS process. However, only towards the end of the CPRGS preparation process, and to a limitedextent, was attention given to methodologies for costing these targets. Over time, methodologies for identifying and costing strategic goals need to be developed in a way which i s fully-integrated with the MTEFs and the recurrent and capital budget process. 4.35 Traditionally, the government's approach to monitoring the impact of its expenditures has focused on input control, rather than on outputs or outcomes. Inputs have been controlled through the issuance of detailed expenditure norms formally bindingon all levels of government. Outputs and outcomes are monitoredthrough routine data systems (e.g. on enrollment rates or use of health facilities) and through periodic household and demographic and health surveys, but the extent to which these output and outcome indicators are used to evaluate the impact of government policies and the degree of feedback of monitoring results into policy-making has been limited. 4.36 However, with the 2002 State Budget Law and the gradual delegation of financial discretion to spending units, the Government i s increasingly stepping away from an input-focused system of budget management. It i s becoming increasingly important to strengthen performance management information, and to gradually integrate this with planning and budgeting. The ongoing plan to build performance-oriented, medium-term expenditure frameworks in four sectors and four provinces should provide an excellent opportunity to pilot and foster greater performance orientation and monitoring. 4.37 The Government i s currently working to devise ways to improve monitoring and evaluation for capital investment projects, focusing on systematic tracking of budgeted and executed expenditure and implementation progress. 56 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 4.38 The 2001 Public Administration reform strategy states that "efforts will be made eventually to abolish the staff number-based budget allocation and adopt a mechanism for calculating budget requirements on the basis of outputs and quality of operations, ensure the monitoring of outputs, quality of targets and objectives of administrative organs, and reform the expenditure norms setting system to make sure that it will be simple and ensure the ownership of budget user^."^ However, other countries, including developed countries, are still struggling to make such methodologies operational. Vietnam will need to make progress gradually on this front. Recommendations 4.39 Regarding the coordination of budgeting with planning and performance management, the PER-PA recommends the following: e The MPI should become an active partner with MOF, sector ministries and provinces in the preparation of MTEFs; and M O F should play a major role in the development of the new SEDP. The MTEFs could emerge as a powerful common analytical tool and vehicle for coordination between the planning cycle and the budgeting cycle. e MOF and MPI should increase their efforts to strengthen the links between performance indicators and budgetary decision. For the short-term, this i s likely to be a question of attempting to "fit" the existing Vietnam Development Goals together with existing expenditure programs within ministries and provinces, in order to strengthen ex post evaluation of spending. For the longer-term, the challenge will be to strengthen the definition and monitoring of the indicators themselves; and to forge processes which identify, cost and prioritize the goals in parallel with departmental and provincial expenditure programs. e A cross-government effort should be made to strengthen monitoring of service delivery. In this regard, the recent initiative to pilot a citizen report card survey exercise in four cities (see Box 4.3) i s an encouraging one. I t would also be desirable for the Government to encourage civil society organizations to become more actively involved in service delivery monitoring at the local level. Clarifyingthe role of norms 4.40 Financial and physical norms continue to play an important role in determining resource allocation in the budget process. These norms are used to estimate needs and determine budget allocations for the various sectors and provinces depending on criteria such as population, poverty rates, degree of remoteness and presence of disadvantaged population groups. Reform of these norms has played an important role in making the recurrent budget pro-poor. For instance, Decision 139/2003 establishes that "the educational expenditure norm for a person living in mountainous and/or remote areas i s 1.7 times higher than that of urban areas; similarly the health expenditure norm i s 1.82 times." Such reforms have provided disadvantaged areas with more funds for local socio- economic development. 'Section 4.3 of the Public Administration Reform Strategy. Decision No 136/2001/QD-TTg on 17 September 2001. Institutionsfor Public Expenditure Management 57 Box 4.3: Piloting of the Citizen Report Card Survey inFour Cities A "citizen report card" represents an assessment of public services from the perspective of citizens. Citizens are the users of these services and can provide authentic feedback on the quality, efficiency and adequacy of the services, as they perceive them, and the problems they face in their interactions with the service providers. Citizens may not be able to comment on the technical features and standards of the services or to evaluate the overall performance of a provider. However, they are eminently qualified to say whether the service meets their needs, and whether the agency i s responsive, corrupt, reliable, etc. There is a growing attention in Vietnam to the performance and accountability of public service delivery agencies. Most accountability measures for these agencies tend to focus on input such as number of personnel, facilities, and expenditures. Little is, however, known about the outcome and quality of service delivery. Inthis context, the Citizens Report Card aims at: providing municipalities with an analytical tool for self assessment and to enhance transparency and accountability in service delivery; generating specific citizeduser feedback on services; converting widespread private problems into collective issues; facilitating comparison of performance of service delivering agencies (for same service); and initiating dialogue and practical problem solving in service delivery To assess the performance o f service, the People's Committees and People's Councils of Ho Chi Minh City, Danang, Nam Dinh and Haiphong are piloting a Citizens' Report Card Survey. The Statistical Offices in the four cities carried out a survey in five service delivery areas: Education; Health; Solid Waste Management; Water Delivery; and Administrative Service Delivery including notary, land registration, housing construction permits and household registration. The findings are currently discussed through public workshops in the four cities. The findings will be published in both national and local media once the results are validated. 4.41 Whereas prior to 2002 these financing norms used to be sent by circular from the Ministry of Finance, the government now needs to submit them to the standing committee of the National Assembly. Moreover, the norms according to which budget allocation i s determined are now made public to sector ministries and the provinces. This publicity i s an important contribution to the transparency of the budgeting process while their submission to the National Assembly increases accountability. 4.42 Whereas in the past these norms were binding on implementing units, this i s no longer the case. Since the 2002 State Budget Law, norms are only used to determine the total amount to be transferred to the sub-national levels of government and the distribution across sectors can be changed from what i s indicated by the norms. Moreover, according to Decision 139/2003, it i s the province that determines the distribution of expenditure across districts. As a result, each province now issues its own norms to determine revenue sharing and resource allocation at the sub-national level. 4.43 The extent to which the elimination of the binding power of centrally-determined financing norms will represent a departure from past practice remains controversial. While some governmental and non-governmental actors believe norms were previously followed and hence the system did work as an effective input-control system from the national to the local level, others believe norms were never binding de facto and hence the 2002 budget law only formalizes actual practice. 4.44 Inaddition to the financial norms, there are a number of technical norms usedby sector ministries for planning purposes. These norms establish, for instance, the number 58 Vietrinni Public Expenditure Review nnd hregrrited Fidirciav Assessriiei7t of teachers and resources per class or pupilheacher ratios and are developed jointly between sector ministries and the Ministry of Finance. These norms will continue to be established following this procedure, but links with financial norms and hence with resource availability are sometimes tenuous. Partly as a result, some provincial governments attempt to achieve ambitious targets in various sectors by establishing high revenue targets which in turn force resource-constrained communes to overly burden their communities by raising own-resources and contributions. 4.45 The changing role of norms and the ongoing decentralization process have significantly reduced the role of sector ministries in the budgeting process. Whereas, at one time, sector ministries were responsible for establishing the allocative norms within their sectors, the new system provides for provincial People's Councils and provincial People's Committees to play this role for spending managed at'the subnational level. Subnational levels of government are expected to follow the sector plan in determining budget allocations, but there i s no system in place to ensure this consistency between plan and budget as well as no monitoring mechanisms. This disconnect will render the task of sector strategy implementation a very difficult one for sector ministries. Recommendations 4.46 On the role of norms, the PER-FA recommends the following: 0 Regarding the norms used to assess public expenditure needs in the formula for transfers to provinces and allocations to sectors, options for further reform should be considered. Inparticular, the use of simple percentages of budget to determine the recurrent budget for economic services and the capital budget should be reviewed. This method lacks a sound rationale and tends to favor richer provinces. The relationship between rates of provincial poverty and transfers per capita should be closely monitored. If necessary, regional adjustment coefficients should be increased to compensate more adequately for the cost disadvantages and lesser fiscal capacity of poorer areas. Regarding the remaining "physical" norms, it should be made clearer that these norms are indicative and not mandatory. Mandatory "budget allocation" norms should be used as sparingly as possible, and only where compliance can be monitored. STRENGTHENING BUDGETEXECUTION MANAGEMENT AND INFORMATION Strengtheningbudget execution 4.47 Budget execution i s a process coordinated between the State Treasury Department, Tax Administration Department and Customs Department, in cooperation with ministries, spending units and lower level departments and bureaus. On the spending side, budget execution i s the responsibility of spending units, working with the State Treasury Department of the Ministry of Finance which i s responsible for budget control, payment and financial management information. In addition to its central offices, the State Treasury Department has offices in every province and district. Payments requested by spending units must be cleared b y the local branch of the Treasury before the payment i s endorsed. Institutionsfor Public Expenditure Management 59 4.48 The existing computerizedTreasury system (separate installations of which exists in each Treasury office) includes some quite effective mechanisms to prevent over- spending and misappropriation of resources. Before a payment can be cleared, the payment i s assigned to a budget line and the system checks there i s sufficient free balance under that budget line before approving the payment, booking it and deductingit from the free balance. Nevertheless, the process of in-year adjustment, combined with the large amount of carry-over between years, has made i t hard to compare budgeted and actual expenditure. 4.49 Perhaps most strikingly, formal commitment controls are almost entirely absent. This is illustrated by the substantial payment arrears which have built up in the transport and agricultural sectors, as discussed in Chapters 2, 11and 13. The Government intends that the new Treasury and Budget Management Information System (TABMIS - discussed below) should help create processes ensuring much tougher controls on expenditure commitments. 4.50 A number of measures to simplify and expedite budget execution have been recently implemented and some more are under way. Up to 2003, spending unitsreceived their resources through a system of quarterly allocations, requiring the approval of either sector and finance ministries or local sector departments and financial agencies. Since January 2004, in accordance with the 2002 State Budget Law, spending units are free to execute their budgets according to their quarterly expenditure plans by issuing payment instructions directly to the Treasury. 4.51 This system will certainly expedite budget execution and allow for improved planning and more flexibility by spending units. The fact that resources will no longer go through sector ministries and departments has the potential to further weaken their role and the amount of information available to them regarding expenditures in their sectors. It i s arguable that it was never appropriate for sector ministries and departments to be playing a quasi-Treasury function. However, it i s clearly the case that they need access to information regarding expenditures in their sector. It i s the Government's intention that the new TABMIS system will help ensure this. 4.52 Also as of January 2004, and in accordance with the 2002 State Budget Law, spending units have more freedom to re-allocate resources across sector items. Under the previous State Budget Law, spending was controlled in nine line items. The new law gives spending units, at every level, greater budget flexibility, with a reduction in the number of controlled items of expenditure from nine line items to four broad blocks of items. Again, this system will allow responsible spending units a greater degree of flexibility and, potentially, optimization in their use of resources. Recommendations 4.53 To further strengthen budget execution, the PER-FA recommends the following: 0 The impact of single annual allocations and of greater budgetary flexibility should be closely monitored to ensure that it results in an improvement and not deterioration inthe composition of expenditure, outputs and services provided. 60 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 0 The new TABMIS system should be implemented so as to ensure that sector ministries are given direct access to Treasury data on execution of sector spending at every level of government. Strengthening Treasury and budget management information 4.54 Historically, Vietnam like many other countries has left the development of its financial management information systems to its individual departments and units, with little attention to the critical flows of information between system components and beyond the system. As a result, the current systems are fragmented technologically and have overlapping and conflicting functionality. The end result has been a lack of integrity in overall fiscal data, transparency and control. The Government has made a start in modernizing its core treasury and budget management systems. The systems put in place have proved broadly effective at preventing overspending and misappropriation of resources. Nevertheless, major weaknesses remain. 4.55 On the institutional side, the lack of a common account code structure results in financial reports which are inconsistent and hard to compare. Paralleling the three separate accounting systems, separate financial management information systems currently operate in the financial agencies (the State Budget Department at central level, and the departments and bureaus of finance at local level), in the State Treasury Department and in the spending units. Most donor financed projects, off-budget bonds and many user charges and fees are currently not consolidated into the State Budget or recorded in the State Treasury Department's main accounting system. The lack of a fully consolidated and integrated budget makes it difficult to monitor total revenues and expenditures, distorting perceptions of the true fiscal position, with potentially serious consequences for fiscal stability. It makes i t hard to assess how resources are being allocated as a whole, and to poverty alleviation in particular; and to compare budget plans with budget outturns. It also acts as an impediment to the successful management of further devolution and delegation. 4.56 Article 61 of the 2002 State Budget Law designates the State Treasury as the lead agency responsible for maintenance of financial management information. In addition, the Government has committed to integrate the Treasury accounting system, the budget accounting system and the spending units accounting system within a Unified Chart of Accounts. 4.57 Technological weaknesses also persist, despite progress in recent years. In 1999, Vietnam's key fiscal agencies began to invest in a united information infrastructure in their various agencies. The purpose was to create a network backbone connecting their individual networks. Since then, there has been a very substantial investment in ICT hardware, software and communications infrastructure. There are now Local-Area Networks (LANs) connecting the key financial agencies (the Treasury department, finance department, taxation department and customs department) in all 61 provinces. Wide-Area Networks (WANs) have been established in some of the larger cities. Investment i s currently underway to complete a national network backbone, which will linktheseLANs andWANs by the endof 2004. Institutions for Public Expenditure Management 61 4.58 The Government has announced its wish to build on institutional reforms, the improved ICT environment and the various legacy systems, to procure an integrated Treasury and Budget Management Information System (TABMIS),which will eventually replace the current myriad of systems and support budgeting, control, monitoring and accounting at every level of government. Recognizing the need to draw on international experience and global best practice to identify and document their functional and technical specifications, the Government of Vietnam appointed consultants to assist them. Following extensive technical analysis and dialogue, the Government finalized its functional and technical specifications (see Box 4.4) and launched a procurement process for a TABMIS solution in February 2004. It i s anticipated that a solution and implementation partner will be selected early in 2005 and implemented over a four year period starting in early/mid 2005. Box 4.4: Key Characteristics of TABMIS The functional and technical specifications for TABMIS are closely aligned with the World Bank- IMF"Treasury Reference Model". The detailed tabulation of functional and technical specifications run to many pages, but key requirements in summary are that TABMIS will: be centered on the Treasury Department, and will operate in the first instance at central, provincial and district levels, with data entry at the Treasury level and in some cases (on a pilot basis) at the commune or spending unit level; operate in the first instance on a cash basis, but will be based around an account code structure and a systems design that will enable commitment accounting and in due course full accruals accounting; to this end, include basic accounting and cash management modules as follows: (i) budget executiodgeneral ledger based on a double-entry system and (ii) treasury modules; and additional modules to establish an upgrade path to modified accruals or full accruals accounting including (iii) payables, (iv) receivables, (v) assets and (vi) procurement, forming the core structure; have the fields and character spaces sufficient to report by institutional (ministry/province) classification, project/program classification, geographical classification and functional and economic classifications consistent with the IMF's "GFS " standards; along with capacity to cater for future development of output/outcome budgeting in line with Government's future vision; be capable of generating consolidated reports and end-year accounts in a timely and accurate way and warehousing the information for subsequent analysis; have open architecture, including the capacity to interface with (i)a medium-term budget preparation module, to be developed at a later stage; (ii) management modules, existing and debt future; and (iii)such other modules as the Government may have or wish to develop at a later stage, including a tax revenue module, a customs module and a human resources module; have the capacity to exchange information and reconcile with the inter-bank transfer system, facilitating the move to a fuller Treasury Single Account approach in due course; include strong security features and will generate a computer-assisted audit trail to facilitate modern sample-basedauditing; and be capable of interfacing and operating concurrently with the existing systems during the transition period. Source: Public Financial Management Reform Project, Project Appraisal Document, April 2003 4.59 But before it can implement such an integrated system, the Government recognizes the importance of finalizing a Unified Chart of Accounts (UCOA). This integrated coding structure will form the "brains" of the new Treasury database, and i s 62 Vietnam Public Expenditure Relie\\*and Integrated Fidircians Assessment what will enable all users across Government to get the information they need from a single system. The State Treasury Department, the State Budget Department and the Accounting Policy Department of the Ministry of Finance have been working with consultants and with representatives from across Government to ensure that this UCOA i s inplace and approved over the coming months. Inamajor recent accomplishment, adraft UCOA has been prepared which integrates the three account code structures and which i s consistent with the IMF's Government Finance Statistics Manual (GFSM) and the International Public Sector Accounting Standards (IPSAS). However, it still needs to be refined and formally adopted. Recommendations 4.60 Regarding the strengthening of Treasury and budget management information, the PER-FA welcomes the strong progress that has been made since the 2000 PER and recommends the following: e Refine and formally adopt the new UnifiedChart of Accounts, for implementation as part of TABMIS. 0 Continue to give sustained attention to change management and organizational readiness as TABMIS implementation moves forward. Once implementation starts, a high priority should be given to training in the use of the new system and inthe application of the UCOA. 4.61 Implementing the new UCOA and TABMIS as a whole will place major demands on government. The existence of a legacy system means that Treasury offices, where most of the data input will take place, already have a critical mass of staff familiar with using electronic databases for budget execution and recording. Nevertheless, ensuring that these skills are updated and extending such skills to other agencies at central, provincial and district level will involve a major training effort which the Government i s fully committed to. Ensuringuser acceptance will require not only training but carefully devised change management processes. 5. INSTITUTIONS FORFINANCIALACCOUNTABILITY AND TRANSPARENCY INTRODUCTIONAND OVERVIEW 5.1 Experiences around the world suggest that public expenditure i s unlikely to result in cost-effective service delivery without strong institutions ensuring transparency and accountability. Transparency requires strong accounting and strong budgetary and financial reporting processes in the public sector. Accountability requires effective oversight, not only by government bodies, but also by elected bodies and the wider public. Both transparency and accountability require that budgets, accounts and audit reports be made available to the public in a timely manner, and that mechanisms exist for soliciting the concerns of the public and, where necessary, adjusting the way that public funds are appliedinresponse. 5.2 Over the past few years, as Box 5.1 details, Vietnam has taken significant steps in improving institutions, laws and regulations and practices to ensure greater financial accountability and transparency in public expenditure. A number of recommendations of the 2001 Country Financial Accountability Assessment (CFAA) have been implemented, or else are in the process of being implemented. For example, oversight by the National Assembly and the provincial People's Councils over public finances has been substantially increased. Public access to financial information continues to improve. Nevertheless, challenges remain. Public sector accounts need to be made more timely, accurate and consistent with international standards. Oversight by the National Assembly over public finances and the effectiveness of the State Audit of Vietnam need further strengthening. Audit reports need to be published. Financial accountability and transparency at the sub-national level need to be strengthened. Financial reporting, audit, and corporate governance of SOEs require significant improvement. 5.3 This chapter presents an update of the 2001 CFAA. It starts by reviewing accounting and reporting processes in the public sector. It then discusses the oversight functions in the public sector. The next section considers financial reporting and auditing in State-Owned Enterprises (SOEs). The chapter then considers the other steps to be taken to strengthen public oversight over the way public funds are used. ACCOUNTING AND REPORTINGPROCESSESINTHE PUBLICSECTOR 5.4 As Chapter 4 described, the 2002 State Budget Law designates the State Treasury as the lead department responsible for government financial reporting. The Government i s implementing an integrated project to strengthen budget execution and management information as part of its Public Financial Management Reform Initiative. This work includes streamlining and expediting budget execution; consolidation of extra-budgetary funds; establishment of a Unified Chart of Accounts (UCOA) for budget monitoring, 63 64 VietnamPublic Expenditure Review and Integrated Fiduciary Assessment Treasury management and accounting; and implementation of an IntegratedTreasury and Budget Management Information System (TABMIS), extending to central and provincial levels of government. A tender i s underway for the new TABMIS system and implementation i s expected to beginin early/mid 2005. Box 5.1: Progressin ImprovingFinancial Accountability and Transparency inVietnam Over the last few years, Vietnam has made significant progress in improving financial accountability inboth the public and private sector. This includes: A Public Financial Management Reform Initiative to integrate core treasury and budget management information systems, and reinforce the links between budget management and the government's developmental goals, including implementation of an integrated Treasury and Budget Management Information System (TABMIS) and a Unified Chart of Accounts (UCOA). A Budget Law that provides for more transparent allocation and procedures for disclosure of budget information, more flexibility in disbursement and use of funds to permit discretion and encourage initiatives in savings, and extensive budgetary powers to all levels of spending agencies.* An Accounting Law that defines applicable principles and requirements and the scope of accounting; prescribes the types of accounts, chart of accounts, financial statements, their disclosure and publication, their audit and inspection; and assigns to MOF the responsibility of formulating accounting standards in conformity with international financial reporting standards. A Decree that clarifies the responsibilities of State Audit of Vietnam, introduces operational auditing for evaluation and improvement of the effectiveness and efficiency of operation and permits the hiring of private audit firms to assist SAV in its audit. A Decree that assigns to the Government InspectionOffice the role of an anti-corruption agency. The issuance of twenty-seven new Vietnamese Standards on Auditing and sixteen Vietnamese Accounting Standards that have been developed on the basis of international standards. A Decree on independent audit that regulates the status of auditors and audit firms, defines the values of audit results and enumerates the compulsory cases that should be subjected to an independent audit. Publication for three successive years of audited State Budget Final Accounts along with the State Budget Plan approved by the National Assembly. A Decree on Grassroots Democracy, requiring communes to post their budgets and final accounts on notice boards. Launching of diagnostic work on corruption, following the resolution of the Central Committee ol the Communist Party inDecember 2003 to step up the fight against corruption. 5.5 Inparallel, the Government has taken important measuresto strengthen further its accounting and reporting. Measures to strengthen accounting 5.6 To ensure a consistent framework on accounting in both the public and private sectors, an Accounting Law was passed by the National Assembly". This law: (i) defines applicable principles and requirements and the scope of accounting; (ii) prescribes the types of accounts, Chart of Accounts, financial statements, their disclosure Resolution 387/NQ/UBTVQH11 dated Mar 17, 2003 and Decree 73/2003/ND-CP dated Jun 23, 2003 on formulation and allocation. Decree 128/2004/ND-CP and 129/2004/ND-CP dated May 31,2004. loAccounting Law promulgated by Resolution No. 03/2003/QH11 on June 17 2003 and effective January 1 2004 Institutions for Financial Accountability and Transparency 65 and publication, their audit and inspection; and (iii) to M O F the responsibility for assigns formulating accounting standards in conformity with international financial reporting standards. An important provision of the new law i s the requirement for all spending units to publish their annual financial statements within 30 days of the date of approval by relevant authorities, and non revenue-earning entities within 30 days of such an approval. 5.7 More autonomy and responsibility i s given to spending units in maintenance of accounts and preparation of reports through the Ministry of Finance's Decision 130/2003/QD-BTC dated August 18, 2003, on "Accounting Policy for State Budget and Treasury Transactions". This Decision was followed by Instruction No.1193KB/KT, issued by the State Treasury on September 11, 2003, which provided a detailed guideline on: (i)content, recording method and flow of accounting documents; (ii)accounting method for major economic transactions; (iii)methodology for recording accounting transactions; and (iv) methodology for preparation of financial statements and managerial accounting reports. The Director and Chief Accountant of the State Treasury's offices are responsible for checking financial reports before submission. Clear deadlines for submission of different types of reports are also prescribed. A system for quick reporting for management purposes i s established to facilitate monitoring by higher levels of authority. 5.8 The accounting system at sub-national levels has also been strengthened, with the issuance of Decision 141 of December 2001 on accounting at the commune level. This provided for a streamlined and simplified book-keeping method for small communes with limited resources and capability. Measures to strengthen reporting 5.9 The reports included in the existing Treasury system have been improved with more focus on budget spending by sectors and national programs. Some key reports will be issued on a daily, ten-day and monthly basis to ensure regular monitoring. An encouraging feature of the new reporting system i s the introduction of reports designed for managerial monitoring, so that these reports can be used as management tools for analysis and decision making.l1 Recommendations 5.10 Despite significant progress, challenges remain such as duplicative checking and verification at various levels. Financial reports require streamlining the number of reports and also developing reports showing impact of the expenditures. In addition, the decentralization and delegation of budgetary responsibility points to the need to strengthen accountability and internal control at subnational and spending unit level; the need to strengthen ex-post control; the need to complete the automation of accounting and financial reporting; and the urgent need for training of financial management staff, as well as the agency heads and the Chief Financial Officers who have increased responsibility and accountability for the financial performance of the agency. "DecreeNo. 10ofJanuary 162002andDecisionNo. 192ofDecember 17,2001. 66 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 5.11 As part of the implementation of TABMIS, efforts should focus on: redesigning the reporting system, adopting a modem internal control framework, strengthening the monitoring and evaluation system, and implementing a comprehensive training program. TABMIS should ensure that the information flow between agencies at the same level and between agencies at the lower and higher levels i s simplified and streamlined. I t should also ensure that the number of reports i s reduced while focusing on exceptions and on monitoring needs. The system should also embody a modern internal control framework, including a control environment that promotes openness and i s based on an integrated and automated system of verifications. The modern internal control system should be able to report on and evaluate the impact of budgetary expenditures. The training program should include staff at all levels, particularly at sub-national levels, increasing their understanding of these reports as management tools. 5.12 The PER-PA recommends: 0 moving progressively towards adopting International Public Sector Accounting Standards (IPSAS) and International Auditing Standards.l2 OVERSIGHTFUNCTIONSTHE PUBLIC SECTOR IN The OverallFramework 5.13 There are three agencies responsible for audit and inspection in the public sector: the Government Inspection Office, State Audit of Vietnam, and Financial Inspection Directorate of MOF. However, the roles of these agencies remain imprecise and overlapping. Due to inadequate capacity and resources, each agency currently focuses mainly on compliance with rules and regulations and on integrity of the financial statements with insufficient attention to economy, efficiency and effectiveness in use of public resources. Risk assessment at the agency level needs to be improved. External audit i s still heavily relied upon, since government agencies have no internal audit function. The StateAudit of Vietnam 5.14 To clarify the functions and powers of the State Audit of Vietnam (SAV), the government issued Decree No. 93/2003/ND-CP of August 13, 2003. This replaced Decree No. 70/CP of July 11, 1994. The new decree maintains the general principles that the SAV i s an agency attached to the Government; that the State Auditor General i s appointed and dismissed by the Prime Minister and i s responsible before the Government and the Prime Minister for all activities of the audit; and that the State Auditor General submits to the Prime Minister the SAV's long-term, medium-term and annual plans for approval and reports the results of audit to the Prime Minister, the Government or the audited ministries according to the Government's regulations. The responsibilities and authority of SAV are clarified as follows: 0 certify: (i)the accuracy and legality of the settlement of state budget at all levels and the financial reports of agencies using the state budget ; and (ii) compliance l2The World Bank will support implementation of IPSAS through an Institutional Development Fund grant. Institutionsfor Financial Accountability and Transparency 67 with the law and economy in the management and use of the state budget and public property; 0 propose to the audited units to correct errors, reorganize and improve their economic and financial management and accounting system; 0 contribute expertise to important schemes and projects within its professional domain, at the request of the Government; 0 enter into international cooperation inthe field of state audit; and 0 hire outside auditors to assist it in the audit, provided that SAV monitors and checks the results of such audits. 5.15 Decree N o 93 has introduced a number of initiatives that pave the way for further progress. These include: 0 introduction of performance auditing (although the principle of economy was recognized, the effectiveness and efficiency of operation were not clearly spelled out); 0 early engagement by the auditors in the public expenditure cycle, thus contributing to internal control related aspects of the project; and e employment of private audit firms to assist SAV in its audit. This will permit exchange of expertise and fresh ideas from the private sector (an open and progressive approach). 5.16 International good practice suggests that the supreme audit institution should be independent of the executive branch of government and should report directly to the National Assembly. At its last session in 2004, the National Assembly decided to make SAV a fully independent organization established by the National Assembly. A new State Audit Law i s being drafted for submission to the National Assembly. The new State Audit Law i s expected to enhance SAV's roles and responsibilities and provide a close linkage with the new Accounting Law (issued on June 17,2003). 5.17 The issue of increased transparency, however, remains unresolved for the time being. Article 2 of the Decree No. 93 refers to "according to the legal provisions" a context that could open the way for more transparency in the form of laws and regulations to be introduced later. At present, audit findings are not made public and can only be made public with Government authorization. For example, the audit of the State Bank of Vietnam financial statements for FY 2003 began in April 2004, and it i s the intention of the Government to submit a copy of the audit report to the National Assembly after its completion. 5.18 The Government i s addressing this issue through policies aimed at improving wider transparency and participatory aspects of the budget process. The new Budget Law-promulgated in 2002, and made effective from January 2004-provides for parallel reporting of the audit results on the budget to the National Assembly, the Government and other agencies, including the People's Councils. It i s up to the latter to disseminate the report to its members. Despite these improvements, direct public access to the audit reports prepared by SAV remains limited. Once the Audit Law i s passed and implemented, there i s expected to be improvement in this area. 68 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment The Government Inspection Office 5.19 The organization and functions of the Government Inspection Office (GIO) are governed by the Inspection Ordinance (1990), the Law of Complaint and Denunciation (1998) and the Anti-Conuption Ordinance (1998). The General Inspector of the Government Inspection Office (GIO) holds the rank of minister. He i s appointed by the President of Vietnam on the basis of a recommendation by the Prime Minister and approval b y the National Assembly. The Chief Inspector of a ministry i s appointed by the Prime Minister on the basis of a recommendation by the Minister and consent of the General Inspector of GIO. Thus, both the executive authorities and the inspection authorities have a voice on the selection of the inspector. 5.20 The GI0 performs the following main functions: (i) compliance with the inspect law by the executive branch in all its areas of operation, including financial, economic, army, police and national defense, etc; and (ii) resolve all legal complaints; and (iii)assist the government in resolving the people's complaints in administrative areas. In this last respect, GI0 performs the functions of an Ombudsman Office. 5.21 There i s a recognition that the Inspection Ordinance of 1990 i s no longer suitable and that a legal framework for a new role of inspection i s needed in the context of a market economy. The government i s the process of streamlining the inspection function. A new Inspection Law was approved by the National Assembly in 2004. Its main objective i s to more clearly define the GIO's role and responsibilities.It aims to provide a clearer demarcation between administrative complaints (those that could result in administrative sanctions) and accusations (denunciation by the people that could result in court proceedings as a crime); and provide a transparent framework for anti-corruption measures. 5.22 Previously, the Government Inspection Office and the State Audit Office (SAV) had similar and overlapping functions. In the future, the SAV will be given more legal authority, beyond the traditional role of ensuring compliance with rules and regulations. The GIO's role will be to evaluate whether national goals are being met. The GI0 itself recognizes the need to unify the present legal framework and clarify roles and responsibilities of inspectorates. The introduction of the new inspection law aims at addressing these problems through clear delegation of duties to the various inspectorates. The government also requires inspectorates to have good planning and coordination, once-a-year inspections and a maximum of thirty-day inspection engagement per enterprise per year. 5.23 Inaddition, by Government Decree No. 46/2003/ND-CP of May 9, 2003, the GI0 has been assigned the role of an anti-corruption agency. The Office will (i) the review entire legal framework to reduce the incentives for corruption; (ii) on prevention; focus and (iii) be in charge of declaration of assets by civil servants and speedy handling of denunciations by the people. 5.24 The above measures indicate a new direction for the inspection function. However, this new task i s a major challenge for the Office, due to limited capacity of staff, and a lack of modern means of investigation and resources for dissemination and training. The Office i s also assisting the sector ministries in strengthening their inspection Institutionsfor Financial Accountability and Transparency 69 function. This includes the establishment of an inspection unit within the Ministry of Planning and Investment that will conduct compliance inspections, investigation of complaints and accusations, and eventually serve as an internal audit function. The FinancialInspectionFunctionof theMinistryof Finance 5.25 The Financial Inspection Directorate i s a department of MOF. Its main functions are to conduct inspections of financial matters of all government and non-government organizations that use public funds and to review and resolve all complaints and denunciations relating to matters involving public finance. The Financial Inspection Directorate has wide scope of operations similar to those of the Government Inspection Office and the State Auditor General's Office. 5.26 In its effort in streamlining this function to avoid duplication of efforts, the government has taken the following measures: (i)adoption of distance inspection through risk assessment followed by rating and selection of samples to be inspected; (ii) change from ex post review to review of current operations and focusing on prevention of waste and inefficiencies; and (iii) emphasis on the role of the inspector as a advisor to management for improvement of systems and procedures, rather than detection of errors and sanction of the culprit. Further efforts should focus on establishing the system to monitor and follow up on the implementation of audit recommendations and resolution of irregularities by the executive branch, including introduction of incentives for speedy resolution and penalty for delay. Recommendations 5.27 The PER-FA recommends the following key measures to strengthen the State Audit of Vietnam: 0 Establish an effective public accounts oversight function at the National Assembly, perhaps by strengthening the capacity of the Economic and Budgetary Committee. 0 Ensure independence for SAV by converting it into an independent organization established by the National Assembly. 0 Improve the transparency of the oversight process by making audit reports available to the public, and promote demand for accountability by encouraging media and public debate on public financial management issues. 5.28 To improve wider oversight functions in the public sector, the PER-FA recommends the following: Further rationalize the roles and responsibilities of the audit and inspection functions: for example, consider establishing internal audit functions, initially at ministerial level, to gradually replace the present inspection function at each ministry. 5.29 Further efforts will be required to review and revise all existing manuals and guidelines on audit and inspections in order to make them consistent with the revised scopes and objectives. In addition, auditors and inspectors will need to be trained in 70 VietnamPublic Exuenditure Review and Intenrated Fiduciary Assessment performance audit techniques, and in reviewing the economy, effectiveness and efficiency of the audited entity's operation over time. FINANCIAL REPORTINGAND AUDITING INSOES 5.30 Financial reporting and auditing arrangements for enterprises are relevant to public financial management to the extent that the public sector includes State Owned Enterprises (SOEs). Major reforms have been undertaken in this area, although significant challenges remain. VietnameseAccounting Standards 5.31 Currently, enterprises operating in Vietnam have to comply with Vietnamese Accounting Standards which are driven by tax reporting considerations and comprise a standard chart of accounts and reporting format. The M O F has been issuing revised standards, but their development and adoption have not progressed as fast as originally planned. So far, 27 Vietnamese Standards on Auditing and 16 Vietnamese Accounting Standards have been developed and promulgated.l3 In addition, The Ministry of Finance has provided implementing guidelines for the first 10 Vietnamese Accounting Standards which were issuedin 2001 and 2002. TheAccountingLaw 5.32 In 2003, the National Assembly approved the first accounting law in Vietnam14. The law i s applicable to public and private sectors as well as SOEs. It also established a legal framework for setting accounting standards by assigning to the Ministry of Finance the task of developing and promulgating Vietnamese Accounting Standards on the basis of International Accounting Standards. The law made a distinction between financial accounting and management accounting and specified that the Ministry of Finance would provide guidance on the application of management accounting. Decree 105/2004/ND-CPon IndependentAudit 5.33 To promote the transparent financial reporting of enterprises, the government has issued Decree 105/2004/ND-CP of March 30, 2004, on independent audit. The objectives of the decree were to regulate the status of auditors and audit firms, define the values of audit results and enumerate the compulsory cases requiring an independent audit. This decree put SOEs and private enterprises into one group as they are now subject to the same requirements. 5.34 For instance, the financial statements of the following enterprises and other organizations must be audited by independent auditors: stock companies and limited liability companies, SOEs, enterprises having foreign-invested capital and enterprises engaged in credit, banking, insurance and developmental activities. This i s not however, applicable to SOEs that are included in the annual audit plan of the State Audit of l3The World Bank has supported this effort through an InstitutionalDevelopment Fund grant in the last 3 years. j4Accounting Law No. 03/2003/QH11promulgatedon June 17 2003 and effectiveJanuary lst 2004 Institutionsfor Financial Accountability and Transparency 71 Vietnam and listed companies that are subject to laws and regulations on securities trading. 5.35 In the same decree, the government encouraged all organizations to have their financial statements audited annually by independent auditors. However, SOEs and investment projects that have been included in the State Audit of Vietnam's annual audit program for any one year are not requiredto be submittedto an independent audit for that particular year. This will prevent the audit of the same SOE by several different oversight entities duringthe same fiscal year. 5.36 The decree defines the qualifications of independent auditors and their rights and obligations by clearly establishing the framework in which an independent auditing firm would operate: the type of audit firm, its establishment, organization and management, the kind of services provided, its functions and obligations. It also enumerates the forbidden activities of an audit firm, including collusion with the audited entity, corruption and other unlawful activities. It describes the required features of an audit report and prescribes the conditions in which a foreign audit firm i s authorized to do business inVietnam. 5.37 The decree provides a framework for strengthening the oversight of enterprises and other organizations, such as NGOs and non-profit organizations, and has put private enterprises and SOEs on the same footing. The role of the government in setting the strategy, plan and development of the independent audit profession in the country was defined. But in the future, this role shouldbe transferred gradually to a professional body. There i s also the need to strengthen the professional association of accountants. In addition, the internal audit function in private enterprises and SOEs i s weak or non- existent. 5.38 The new SOE Law defines an SOE's legal status and operating modalities, allowing SOEs to be regulated under the Enterprise Law if they choose to give up their SOE status and convert their legal status to one member limited liability company, and pave the way for SOE privatization. However, the Law does not explicitly require independence between Board of Directors and Management (the General Director can be a member of the Board of Directors). The corporate governance of SOEs requires significant improvements. There i s a need for a comprehensive training program on modern corporate governance for management and directors of SOEs." Recommendations 5.39 The PER-FA recommends the following measures: 0 Ensure application of improved accounting and auditing standards to large SOEs (General Corporations). 0 Draft a combined law for SOEs and non-SOE enterprises so that they are all subject to the same requirements in terms of roles and responsibilities of line management, general directors, boards of director and control committees; and l5The World Bank is providing assistance to the Academy of Finance to strengthen its training program on modern corporate governance. 72 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment strengthen the internal audit function in SOEs and establish a professional association of internal auditors. 5.40 Implementation of the above recommendations will require particular attention to strengthening the role of the Vietnamese Accounting and Auditing Association as a professional body; and also, a comprehensive training program for directors and executives of enterprises. The monitoring function of the Financial Enterprise Management Department of M O F needs to be strengthened through the establishment of clear requirements concerning the corporate governance regime, accounting policy and financial reporting, rate of return and dividend policy. This special unit should act as a resource center to assist lower-level government agencies in monitoring their local SOEs and improving accounting and financial management skills locally. Guidelines also will need to be issued on the conversion of legal status of audit firms to partnerships and on the type of partnership (limitedor unlimitedpartnership) for audit firms. PUBLICOVERSIGHT 5.41 International good practice suggests that government oversight i s more effective when public financial information i s easily accessible by the public. The Public Administration Reform Strategy states that "democracy and transparency in the area of public finance will be exercised, hence all financial expenditures will be made public."16 5.42 In 1998, the State Budget was made public for the first time in the Annual Statistics publishedby the General Statistics Office as well as in the Governance Finance Statistics of the IMF. Decision 182 (2001) made more explicit the timing, format and content of publicity of the state budget. In March 2002, the Government published the audited State Budget Final Accounts for 2000 along with the State Budget Plan for 2002, and this practice has continued on an annual basis in 2003 and 2004. 5.43 The 2002 State Budget Law, effective since January 2004, further stipulates that "budget plans, budget final accounts, and auditing results of the central budget, the local budget, budget planners, and organizations financed by state budget, must be made public." The 2002 law requires that final accounts, in particular, be published within 18 months of year-end and the government has so far complied with this requirement. 5.44 Despite significant variation across and within provinces and some progress in recent years, much less transparency exists at the local level. According to Decision 225, budgets need to be made public. To monitor the degree of implementation of this requirement, the Ministry of Finance prepares a yearly report detailing how many provinces, ministries and central agencies comply with publicity requirements and whether the content and format of the publicity i s up to the required level. According to this report for 2002, most provinces did send their reports to the Ministry of Finance. However, there i s no comprehensive publication of province level data. This contrasts with a much stronger record for sector ministries and central agencies, 90 percent of which publishedtheir budgets and final accounts. l6 Section 4.6 Public Administration Reform Strategy. Decision No 136/2001/QD-TTg on 17 September 2001. Institutionsfor Financial Accountability and Transparency 73 5.45 The Grassroots Democracy Decrees 29/1998/ND-CP and 79/2003/ND-CP establish the need for communes to post their budgets and final accounts on notice boards, discuss budgets and budget reports in the commune's People's Council and broadcast the discussions on the radio. Although some communes have followed these directives, concerns remain regarding the extent of compliance across the country. Overall, much more attention needs to be paid to transparency and accountability at the sub-national level. In future, both the demand as well as the supply-side for information provision should be taken into account in fostering an environment for improved information-sharing and accountability. 5.46 It i s important to note that at the spending unit level some progress has been achieved, at least on the legal front. For instance, schools are mandated to post their budgets, detailing their revenues, fees and contributions as well as the use of these resources. However, the extent to which this mandate i s being implemented and how many schools are indeed posting their budgets and with which degree of detail and accuracy i s not clear. Whether citizens have the capacity and are empowered enough to use the available information to holdtheir local government accountable i s also unclear. 5.47 The Prime Minister has issued a revised Regulation on Fiscal Transparency (Decision 19YQD-TTG dated November 16, 2004) to broaden and further detail the information that should be made public. It i s noteworthy also to mention the development of the Ministry of Finance's website, in which all of published financial data and regulations issued can be found.17 The revised Budget Law requires that audit reports of government entities to be published. However, the implementingrules have not yet been issued. Recommendations 5.48 The PER-IFA recommends the following: * Continue systematic monitoring of publicity of budgets and accounts at sub- national level, and also encourage provinces and districts to monitor how many schools, health providers and other spending units publish their budgets and accounts. Publish an annual report with detailed and consistent data on public spending in each province. Produce each year a "Budget White Paper" explaining in an accessible way the contents of the Budget and its links to national development goals and strategies. ACTION TO PREVENT CORRUPTION 5.49 The Government and Party have recently decided to step up the fight against corruption. The Central Committee of the Communist Party passed a Resolution to that effect in December 2003. As a result, the Government decided to carry out a review of corruption, with support from donors. The diagnostic work i s meant to identify the causes of corruption and the strengths and weaknesses of previous anti-comption initiatives, and to propose specific measures to reduce corruption. l7This website is: www.mof.gov.vn. \ 74 VietnamPublic Expenditure Review and Integrated Fiduciary Assessment Recommendations 5.50 The PER-FA recommends that the Government should: a Complete the corruption review as soon as possible and ensure appropriate follow-up action. 6. DECENTRALIZATIONTO SUBNATIONAL GOVERNMENT INTRODUCTIONAND OVERVIEW 6.1 The management of decentralization to subnational government i s a challenge in many countries around the world. Decentralization of fiscal authority to subnational government presents some major opportunities: subnational management of the budget can result in better mobilization and allocation of resources, and in delivery of services which are more appropriate and responsive to the needs and wishes of local people and more efficient given local conditions and circumstances. Equally, there are significant threats: of duplication, weakened coordination, and even growing inequity and the deterioration of service delivery in critical sectors. 6.2 Over the past decade, Vietnam has embarked on an extensive decentralization program. A cornerstone of this program was the approval of the 1996 State Budget Law. The result has been a substantial and growing level of decentralization, with the share of subnational governments in total expenditures increasing from 26 percent in 1992 to 43 percent in 1998; and to 48 percent in 2002. Despite progress, after five years of implementation the Government recognized that some parts of the system needed improvement. In 2002 a new State Budget Law was approved, which allowed for significant continuity in the decentralization system, but also represented in many ways a marked improvement over the 1996 Law 6.3 This chapter assesses the strengths and weaknesses of Vietnam's current system of decentralization and proposes measures to increase further both the efficiency and equity with which the system may be able to operate in the future, building on the 2002 State Budget Law. The chapter analyzes several key issues including: the vertical structure of government; the assignments of expenditure responsibilities and revenue sources; the system of transfers; and the budget process at subnational level, including issues of transparency, accountability and subnational borrowing. STRUCTURE OF GOVERNMENT AND ASSIGNMENTOFEXPENDITURE RESPONSIBILITIES Current vertical structure 6.4 Vietnam has four levels of government. The country i s divided into 64 provinces18,ranging in population between approximately 5.5 million and 0.3 million. Included in these 64 are the five largest cities which are granted provincial status: Hanoi, H o Chi Minh City, Hai Phong, Danang and Can Tho. The provinces are subdivided into 643 districts, which, in turn, are subdivided into 10,602 communes. l8Three new provinces were created in2003. 75 76 Vietnam Public Expenditure Review and Integrated Fiduciaiy Assessment 6.5 At the central level, legislative power resides in the National Assembly which, according to Article 84 of the Constitution, approves the State Budget (which comprises the central government budget plus the consolidated provincial-district-commune budgets). On the executive side, the PrimeMinister i s the head of the government and he and the rest of the government are appointed by the National Assembly. At each lower level of government there i s an elected assembly, the People's Council, and an executive authority, the People's Committee, which i s appointed by the People's Council. 6.6 The budget structure i s hierarchical: the budget at each level has to be approved not only by the People's Council at that level but also by the upper level government. Each administrative level operates under a system of dual subordination and accountability: at the same level of government, the People's Committee i s accountable to the People's Council and ultimately to local residents (the People's Councils at the subnational level are elected by universal suffrage). The same officials are also accountable to the upper level and ultimately to the central authorities. Whatever authority subnational governments exercise through the People's Council and the People's Committee, including its approved budgets, i s always under the supervision of the National Assembly, which has the power to suspend inappropriate decisions of the People's Councils. Significant policy issues 6.7 An important issue i s the highly hierarchical relationship among the different levels of government, following the "Matrouchka doll" model of budget structure, whereby communes report to districts, districts to provinces and provinces to the center. The reforms brought in this area by the 2002 State Budget Law give provincial governments considerable freedom to organize the budgets of district and commune governments. This reflects the need to adapt to the different needs and circumstances in which provincial governments find themselves vis-&vis their districts and communes. In the longer term, when accountability, transparency and capacity at subnational level strengthens, the Government may decide to consider options for moving away from the Matrouchka doll model and provide for more autonomy and independence of budgets at each level of government. 6.8 Discretion for provinces in fiscal arrangements between the sub-provincial levels at this time has positive features. It allows such arrangements to be tailored to reflect local diversity and the differences between more urban areas and rural areas in the mountain and plain regions. Potentially, this could allow more efficient and equitable resource mobilization and allocation. However, this approach i s not without risks. More discretion for the provincial governments may mean fewer guarantees that central government policies will be implemented as desired through the districts and communes. For example, central government objectives of geographical equalization and the reduction of poverty may be compromised if provincial governments decide to retain many of the revenue sources over which they have discretion, and shift down expenditure responsibilities in many important service areas to districts and communes, thus making them more dependent on transfers from provincial governments. A preliminary review of 2002 Budget Law implementation in 2004 has not found clear evidence of such a practice. However, this important issue needs to be examined carefully inthe future. Decentralization to Subnational Government 77 6.9 A related issue i s whether or not some communes are too small efficiently to deliver public services due to their inability to take advantage of economies of scale. However, the problems arising from small scale can be addressed through amalgamation or consolidation (voluntary or forced) of smaller communes, by promoting the associations of communes to deliver certain services, or by outsourcing public service delivery to the private sector. The other possibility i s to reassign certain responsibilities, subject to economies of scale, to the higher level governments. Currentexpenditureassignments 6.10 The previous assignment of public expenditure responsibilities hadbeen approved in the State Budget Law of 1996.19 The 2002 State Budget Law did not fundamentally alter the assignment of expenditures between central and subnational government. The current expenditure assignments are shown in Table 6.1, The assignment of expenditure responsibilities, which has evolved in practice over the years, i s in general terms consistent with the theoretical principles of expenditure assignments. Services are assigned at levels of government commensurate with the geographical area of benefits. 6.11 However, the new law changed the whole approach to expenditure assignments in Vietnam within the subnational level. The 2002 State Budget Law essentially leaves it to the provinces to organize expenditure assignments for the districts and communes inside the provinces.20 This follows the principle of granting a great deal of flexibility to provincial governments to adapt to their diverse specific conditions. 6.12 In this regard, the 2002 State Budget Law improves on the 1996 Law in some tangible ways. In particular, it strengthens the prohibition of unfunded expenditure mandates from higher to lower levels of govemment.21 Such a prohibition existed in the 1996 Law, but i t i s now complemented by the new freedom for the provinces to assign expenditure responsibilities to the lower level governments. Furthermore, the 2002 law explicitly states that new policies can only be introduced where any necessary new resources are available. The current expenditure assignments are contained in articles 31 and 33 of the 2002 State Budget Law. 2o The only exception i s contained in article 34 of the new Law on State Budget, which requires that townships and cities under a province must be assigned responsibilities for the construction of public *'schools,lighting, water supply and sewerage, urban traffic, and other public infrastructure. These are reassignments or new assignments of responsibilities to subnational governments where there i s no associatedincrease in funding to local governments to implement the new responsibilities. P P P E 2E D Decentralization to Subnational Government 79 6.13 The 2000 PER concluded that local government (provinces, districts and communes) spent an increasing share of total State Budget expenditure: total expenditures undertaken at the sub-national level increased from 26 percent in 1992 to 43 percent in 1998. This trend continued in this review period. As a result, the share of sub- national level expenditure reached 48 percent of total spending in 2002. This share in total expenditures, to judge from this single dimension, positions Vietnam among highly decentralized countries." 6.14 As Figure 6.1 illustrates, the downward trend for central spending is shown in most sectors, though it was particularly pronounced in Agriculture. In Health, Education & Training and Agriculture the largest share of spendingis carried out at the sub-national level. The shares of the central spending in each of these three sectors reduced to less than a quarter of the total expenditure in 2002. Figure 6.1: Central GovernmentShare inBudgetExpenditure 1997 2002 - - ---_ 70% i 60% - 50% - 30% . 20% - 10% 0% J I 1997 1998 1999 2000 2001 2002 ---Total+Health +Agriculture -Education 1 Signifcant policy issues 6.15 A remaining and significant problem with expenditure assignments in Vietnam is lack of clarity. The statement of expenditure assignments in the 2002 Law on State Budget i s a starting point but requires further clarification. Currently, more detailed functional assignments are scattered across a wide range of sectoral regulations, which remain unclear and unstable. 6.16 As in all countries, the exclusive assignment of each responsibility to a single tier of government i s not feasible. Some responsibilities need to be shared by two or more levels of government, as may be the case in general education or primary health (although this should be avoided where possible, since the reduced clarity and accountability can often weaken service delivery). One way to establish clarity in the case of co-responsibility i s to be explicit in the law about the attribution of competences for the regulation, the financing, and the implementation of the responsibility. For example, ''Decentralization i s a multidimensional concept and the share of subnational governments in total expenditures i s only one of those dimensions. Whether a country i s decentralized also depends, among other things, on the level of autonomy exercised by subnational governments on the composition of expenditures and their ability to raise their own revenues. As discussed below, looking at some of these other dimensions, Vietnam i s less decentralized. 80 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment inthe case of general basic education this will demand a clear statement of the level(s) of government with competence to regulate (introduce standards and so on), which may be the Ministry of Education and Training or the People's Councils; the level(s) obligated to finance the service; and the level in charge of actually implementing or delivering the service, This type of information i s not yet available in Vietnam. 6.17 One of the toughest issues in expenditure assignments i s how to deal with capital expenditure needs of subnational governments. One approach followed in some countries i s to keep all capital expenditure responsibilities at the central level. This i s fundamentally wrong because it leads to inefficient decisions with poor matching of needs and also to lack of maintenance and interest by subnational governments on the centrally built infrastructure. In general, subnational governments should be responsible for the infrastructure they need to deliver the services for which they are responsible. This principle seems to be accepted in the 2002 State Budget Law. In fact, the only specific expenditure assignments to provincial, city and town governments are for infrastructure investments in public schools, lighting, water supply and sewerage, urban traffic, and other public infrastructure (Article 34). 6.18 However, most district and commune governments lack the capital funds for their needs for rehabilitation, replacement and new construction of infrastructure. An important task for the Government will be to study how to mobilize the resources needed to start filling the gap for these infrastructure needs. As discussed below, part of the solution may lie in subnational government borrowing for capital investment purposes, but with careful monitoring and strict ruleS.23 6.19 Another task in this area for the Government to evaluate i s the lack of maintenance of the infrastructure. Many public facilities, including clinics, schools, roads and irrigation schemes, are reported to be in severe disrepair across the country. It i s not possible at this time to establish how much of this problem arises from expenditure assignment issues and how much i s due to the separation of capital and recurrent budgets. Recommendations 6.20 First, the Government should monitor and evaluate the effects on public service delivery of granting provincial government discretion over the fiscal arrangements of districts and communes. The focus should be to establish what effects the added flexibility at the provincial level may be having on the Government objectives of greater overall efficiency of public expenditures and a fair distribution of public resources. Depending on the findings, the Government should consider the introduction of a legislated fiscal structure, establishing clearer requirements regarding the fiscal arrangements between provincial governments, districts and communes. 6.21 Second, the Government should enhance administrative capacity at the local level of government by implementing programs to upgrade and maintain critical levels of technical and administrative capacities at the district and commune levels. 6.22 Third, the Government should consider the need in the longer term to introduce explicit expenditure assignments for districts and communes, and the need to allow for 23 For example, in 1994the GovernmentallowedHo Chi Minh City to issue bonds to finance a road. Decentralization to Subnational Government 81 asymmetric assignments below the province level to adjust for different levels of administrative capacity among districts and communes. 6.23 Fourth, the Government should issue regulations to strengthen collaboration and coordination among agencies at different levels of government that share a particular expenditure responsibility; and to facilitate contracting with the private sector. REVENUEASSIGNMENTS Currentassignments 6.24 Both the 1996 and the 2002 State Budget Laws distinguish three types of revenue: taxes assigned 100 percent at the central level, taxes assigned 100 percent at the provincial level, and shared taxes between the central and provincial governments. In terms of revenue assignments, the major difference between the 1996 and the 2002 budget laws i s that, whereas the 1996 law specified revenue assignment at all four levels of government, the 2002 law leaves provinces considerable discretion in their revenue assignments to districts and communes.24 6.25 The tax revenues assigned 100 percent to the central government include export and import taxes, VAT and excises on imports; taxes and other revenues from the petroleum industry; and corporate income tax on enterprises with uniform accounting. The tax revenues assigned 100 percent to the subnational level include land and housing taxes, natural resource taxes excluding those on petroleum activities, license tax, tax on transfer of land use rights, fees on land use, land rent, revenues from the leasing and sale of dwellings publicly owned, registrationfees and most other fees and charges. 6.26 Shared taxes between the central and provincial governments include all VAT receipts with the exception of VAT on import goods; corporate income tax with the exception of receipts from enterprises under the whole-unit accounting system; personal income tax; special consumption tax on domestic goods and services; and gasoline and oil fees. "Sharing rates" are determined which establish the proportion of total revenues from shared taxes which a province may retain. 6.27 The State Budget Law does not specify the "sharing rates" between the center and the provinces. Instead, this i s left for the Standing Committee of the National Assembly to decide for each period of three to five years. These sharing rates are uniform for all shared taxes for each province but they differ by province and are calculated as part of the budget process at the start of each stability period of at least three years. Before the 2002 State Budget Law the shared taxes were VAT, enterprise income tax, personal income tax, remittance tax. The 2002 Law added two additional taxes-special consumption taxes on domestic goods and services; and gasoline and oil fees-to the list of sharable taxes.25 24But article 34 of the 2002 Law provides some general principles and minimumstandards the provinces need to follow indesigning revenue assignments for local governments Revenue sharing arrangements between the provinces and local governments (districts and communes) for these taxes i s left to the discretion of the provincial governments. 82 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 6.28 Vietnam has a revenue sharing system that i s different from those in other countries. As indicated above, the revenue "sharing rate" i s the same for all shared taxes but it differs by province. In addition, the revenue sharing rate i s determined by formula. The formula estimates the gap between expenditure needs (estimated on the basis of norms) and revenue capacity. In the stability period ending in 2003, the sharing rate (which again applies to all sharable taxes) for the group of 56 poorest provinces was 100 percent. For the other 5 provinces the sharing rates were increased between 1996 and 2003 and now range from 24 percent to 53 percent.26For the stability period startingfrom 2004, with more revenue assigned to subnational government, the number of provinces that have a 100 percent sharing rate has reduced to 49. 6.29 InVietnam, all tax collections are centralized. The General Taxation Department collects all domestic taxes with offices that extend through the provinces and the districts, and the Customs Department collects all taxes falling on imports. Only minor fees and charges are collected by financial agencies and service providers. This has many advantages: it simplifies cash management and facilitates coherent implementation of tax policy throughout the country. 6.30 One problem with centralized tax collections i s the potential lack of incentives that central government bureaucrats may have to mobilize and collect subnational revenues. But in Vietnam, there i s defacto dual subordination of tax administrators to the central administration and the subnational authorities, meaning that provincial and district officials can have a recognizable influence on the decisions and activities of tax administrators.27 The important role played by subnational incentives for tax collection i s captured by the practice of letting subnational authorities retain a share of the collections above the targeted amount for taxes. The "surplus" collections retained at the subnational level include only subnational and shared taxes.2~Currently, the retention rate for surplus revenues i s 100 percent for Hanoi, 79 percent for and Ho Chi Minh City and up to 30 percent for all other provinces. 6.31 The actual collections from the different revenue sources of subnational governments for 2002, the most recent year available, are shown in Table 6.2. The figures show considerable differences across provinces in their abilities to raise taxes. The best off province raises 2,113,000 VND per capita (HCM City) and the worst off 439,000 VND (Ha Nam), or a multiple of almost 5. ~~ 26This generalized increase in the sharing rates for 2001-2003 was to compensate the provinces for the increase in public employee salaries. The two additional taxes (special consumption taxes and gasoline and oil fees) added to the list of shared taxes in the 2002 State Budget law and effective in fiscal 2004 increased the number of "surplus" provinces from 5 to 15. 27For example, it is not unusual for the provincial authorities to provide bonuses to the tax administrators with a better performance in collections. Provincial authorities also usually provide office facilities, housing and other amenities to tax inspectors and other officials of the central government tax administration agency. 28This system of incentives i s a realistic measure that has contributed to more revenue mobilization at the subnational level and obviated the need to create subnational tax administration agencies, as has been done in China. But the system of incentives has not been without its own problems, such as the tendency to underestimate revenues inthe proposed budgets. I 84 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 6.32 In terms of consolidated State Budget revenues (central and subnational governments combined), the subnational governments' share for 2002 was 25 percent.29 This share has oscillated mildly up and down 25 percent in the 1997-2002 period (see Table 6.3). The expectation i s that subnational revenues will grow faster than central revenues and that by the end of 2004 subnational governments will account for about 30 percent of State Budgetrevenues.30 Significant policy issues 6.33 The first issue with current revenue assignments in Vietnam i s the current lack of any material revenue autonomy (in the sense of being able to increase revenues at the margin) b y subnational governments. The only form of revenue autonomy for district and commune governments that currently exists i s the ability to introduce certain fees: for example, waste collection fees and school tuition fees. The issue of tax autonomy i s a difficult one for many governments, especially those in unitary political systems. There i s often political resistance to devolve taxing powers below the center and there may be even the perception that the equal treatment of citizens requires a completely uniform tax system in the entire country. However, the benefits of fiscal decentralization in terms of greater expenditure efficiency can only arise if subnational governments become accountable and responsive to the needs and preferences of taxpayers. It can be argued that an effective way to promote budgetary accountability and responsiveness i s to grant subnational governments a meaningful degree of tax autonomy. 6.34 Against this, not all forms of tax autonomy are desirable. For example, it may be counterproductive for the development of the market economy to allow subnational governments to introduce their own taxes or to be able to modify the structure of existing taxes. This may create tax competition and create barriers to the flow of goods and services. The simplest and best form of revenue autonomy may be to let subnational governments select tax rates for at least one significant source of revenue, probably between minimumand maximum rates stated by the National Assembly.31 6.35 A second set of issues is related to the mechanisms currently usedfor tax sharing. At present, revenues are split between the central government and the provincial jurisdiction where revenues are actually collected (the so-called "derivation principle"). This creates problems of fairness, especially for the VAT and the enterprise income tax. The problem with the VAT i s that i t tends to be credited and debited in the jurisdiction where the headquarters of the enterprise i s located and not necessarily where the tax i s actually incurred. This helps large official cities (Hanoi or H C M City) and a few more industrially developed provinces. The use of the derivation principle for sharing VAT revenues also can lead to market protectionism by creating artificial barriers to internal domestic trade. 29Of course, this figure excludes central government transfers to provincial governments. 30This will be helped no doubt by the two additional taxes, special consumption taxes and gasoline and oil fees, added to the list of sharable taxes starting in 2004. 31Providing a meaningful degree of tax autonomy at the local level i s the simplest and most effective way to address the issue of vertical imbalances (the mismatch between expenditure responsibilities and funding made available to different levels of government). 86 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 6.36 A solution to this problem i s complicated. Some countries have opted for centralizing the VAT; others use instead a direct formula approach to the sharing of VAT, for example, on a per capita basis. In the case of the corporate income tax, sharing on a derivation basis also presents similar problems. Some countries share this tax with subnational governments where the business operates using an apportionment formula (for example, a weighted average of several factors, which include the enterprise's share of labor, sales, and even assets in that locality vis-a-vis the entire country). The apportionment of the personal income tax at the subnational level may also present problems in urban areas where workers live in one jurisdiction and work in a different one, as may be the case for example between HCM City and Binh Duong Province. The main issue here i s that the personal income tax (or at least part of it) should be paid in the place of residence because that i s where taxpayers consume most local public services. 6.37 The elimination of explicit revenue assignments at the district and commune levels in the 2002 State Budget Law provides provincial governments with budget flexibility and the ability to adapt to the particular circumstances in the province. It may allow for higher levels of equalization than would be possible were there a concretely specified revenue assignment. But the lack of concrete revenue assignments also imposes costs on lower-tier governments. It deprives them of revenue autonomy and it may reduce revenue certainty and predictability, This negative impact on certainty and predictability i s partially controlled by the practice of "stability periods." Recommendations 6.38 First,the Government should study the feasibility of increasingrevenue autonomy at the provincial level in the medium term, with the benefits of addressing existing vertical imbalances and more significantly, increasing the efficiency and accountability of subnational budgets. Local tax autonomy should be confined to choosing the tax rate levels (possibly among minimum and maximum rates chosen b y the National Assembly) of a closed list of provincial/local taxes with common tax bases throughout the country. Although the Government should carefully consider this issue, the most obvious choice to provide subnational governments with some degree of tax autonomy would appear to be to develop a piggyback flat rate personal income tax for provincial governments. The Government should also explore the feasibility of usingpiggyback taxes or surcharges on the special consumption (excise) taxes on alcohol and tobacco. The likely best choice to develop tax autonomy at the district and commune levels i s to consider the introduction of a modern real estate property tax, a motor vehicle tax, and "betterment" or improvement levies.32 6.39 Second, and for the longer term, the Government should explore ways to improve the apportionment of shared taxes between central government and provincial governments: in the case of the VAT, by using an equal per capita basis, or the sharing, inproportion to estimates of final consumptionineachjurisdiction. For the profit tax, this would work through the introduction of a formula based on the geographical distribution of the enterprise's payroll, but also, as information becomes available, the geographical 32 The latter are surcharges to the property tax that local governments may approve within their jurisdictions, as one-time or multi-year charges, for improvement directly benefiting certain homeowners, such as improvements in street lighting, sidewalks and so on. Decentralization to Subnational Government 87 distribution of assets andor sales. The Government will also consider adopting the principle of sharing personal income tax revenues with district and commune government according to the place of residence of workers, where most local services get consumed. 6.40 Third, the Government should monitor and evaluate how the flexible unwritten approach to revenue assignments for districts and communes i s performing and what the added costs are for lower-tier governments. Depending on the findings, the World Bank recommends that the Government might mandate more explicit revenue assignments at the local levels as part of any further revisions to the State Budget Law. THESYSTEMOFTRANSFERS Current system of transfers 6.41 N o decentralized system of finance ever reaches a perfect balance between expenditure assignments and revenue assignments. Like most other countries where decentralization i s taking place, Vietnam suffers from fiscal imbalances. Horizontal imbalances can be caused by differences in local economic activity, wealth or resource endowments or differences in needs, because of either different costs of services (for example, mountain versus coastal regions) or from diverse demographic profiles (because of groups of population with special needs). Horizontal imbalances can be enlarged from physical and institutional impediments to population migration or the mobility of capital across provinces, and from government policies that favor some areas of the country over others. The typical measure of horizontal fiscal imbalance involves the comparison between fiscal capacity measures and expenditure needmeasures. 6.42 Horizontal and vertical imbalances are generally addressed through a system of transfers. However, transfers are also used by central governments in the pursuit of other objectives, including, for example, national priorities in particularly targeted sectors or addressing spillovers or externalities across subnational boundaries. Vietnam's system of transfers, though not very conventional, does nevertheless play a key role in the final outcomes from decentralization. 6.43 The typical remedy for horizontal fiscal imbalances i s a system of equalization grants. As described below, Vietnam has a system of equalization transfers (the "balancing transfer"), which plays an important role in reducing otherwise significantly larger horizontal disparities. The balancing transfer system i s considerably strengthened by the 2002 State Budget Law, which more clearly specifies allocation norms. However, not all such disparities have been rectified after transfers. As shown in Table 6.2, the large disparities in public funds arising from the assignments of shared taxes (coefficient of variation in per capita revenues of 1.21 and a difference between maximum and minimumper capita values of 47-fold) and from taxes assigned 100percent (coefficient of variation in per capita revenues of 0.86 and a difference between maximum and minimum per capita values of 20-fold) are significantly reduced after transfers. Total revenues per capita (also in table 6.2) which include revenues from transfers, show a coefficient of variation of 0.42 and a difference between maximum and minimum per capita values of less than 5-fold. Thus, significant equalization takes place. Nonetheless, it i s clear that substantial disparities remain. 88 Vietnam Public ExDenditure Review and Integrated Fiduciary Assessment Equalization transfers 6.44 The "balancing transfers" (equalization transfers) from the central government to the provinces in Vietnam are designed to increase the financing viability of poor provinces. These are unconditional grants, determined using a formula, and which remain fixed in nominal terms for the stability period of three-to-five years, as stipulated by the State Budget Law. The current system of equalization transfers in Vietnam represents a significant improvement over the "gap filling" ad hoc negotiated transfers used in the past. An important improvement i s in the explicit methodology or "formula" used to arrive at the amount of the transfers, which cuts down the role of bargaining and makes outcomes more objective.33 The formula uses the difference between estimated expenditure needs and revenue capacity or potential. Expectedrevenues for the provincial government are determined b y the provincial branch of the tax administration on the bases of the actual revenue collections of the previous years, taking into account any tax policy changes applicable in that year and the expected economic growth during the year. The minimumexpenditure needs of the provincial government are derived on the basis of the prevailing system of expenditure norms, and are supposed to cover all current expenditures (salaries, operation and maintenance, and so on) and (some minimum amount of) capital expenditures. The norms are adjusted for different regions depending on geography and remotene~s~~. 6.45 Another improvement over the ad hoc negotiated equalization grants of the past i s the use of fixed transfers during a three-to-five year stability period. The approach of using fixed transfers helps most with controlling negative incentives to revenue mobilization by provincial governments and b y providing more certainty and stability. For the most part, provinces implement with their lower-tier governments a system of balancing or equalization transfers that i s similar to that of the central government with the provinces. However, between the provincial and lower levels, the provinces set their own norms and so effectively set their own transfers. Conditional or targeted transfers 6.46 According to Article 29 of Decree 60, there are several types of conditional or targeted transfers in Vietnam. In one group there are transfers for the implementation of "National Target Programs," such as those for the poorest communes, reforestation, or the various national health programs. These National Target Programs are so-called because they are a key tool by which central government can ensure the delivery of key national targets at local level in a fiscally decentralized system. 6.47 There are also conditional transfers that are specifically designed for a particular province. In contrast to national targeted programs, the special targeted program i s typically solicited by proposal of the beneficiary province and, after consultations with 33 Note that this formula used for the determination of equalization transfers to the poor provinces is also the same formula used, in theory at least, to derive the customized tax sharing rate for the rich provinces, which we discussed under revenue assignments above. 34 Before the 1996 State Budget Law, there were no budget transfer norms and expenditure needs were negotiated with each province. In the 1996 State Budget Law, budget transfer norms were introduced but not clearly defined. In the 2002 State Budget Law, a transfer formula based on clear transfer norms was articulated. Decentralization to Subnational Government 89 the Ministry of Finance and National Planning Agency, it needs to be approved by the National Assembly as part of the annual budget. 6.48 The third type of targeted grants are the "matching" grants designed to implement important projects for socio-economic development. With matching grants, the central government leverages its resources by covering only a share of the project costs with the rest to be covered by the subnational government. This practice of matching i s not very common yet in Vietnam. But in recent times the government has started to make use of this tool. For example, the central government has recently agreed to finance 50 percent of a provincial road in Hai Phong. 6.49 Another type of conditional transfer i s an emergency grant. B y their very nature, these transfers cannot appear itemized in the annual budgets. The funds come from the central government's Contingency Budget and are incorporatedinto end-year accounts. Capitaltransfers 6.50 Vietnam does not have an explicit separate system of capital transfers. The bulk of the funds that can be used by subnational governments for investment in capital infrastructure comes from the balancing or equalization transfer, although some of the conditional grant funds (for example, the school construction program) are also for investment in infrastructure. Funds for capital investment are factored into the equalization transfer by grossing up the expenditure needs factor in the calculation of the equalization grant. The coefficient for the gross up i s determined by the National Assembly in the consolidated budget as the overall capital expenditure needs. For 2004, this was fixed at 28 percent of the budget on average, with individual allocations negotiated province by province. These funds (notionally for capital investment) are unconditional, as are the rest of the equalization grants, and the provincial governments can use them for any purpose besides capital infrastructure. Significant policy issues 6.51 Unlike other countries, in Vietnam equalization i s performed in integration with the tax sharing system and with the determination of capital transfers. Despite its peculiarities, the system of transfers has performed in many ways satisfactorily. However, certain policy issues and areas for potential improvement still arise. 6.52 First, in the case of targeted transfers, there has been a substantial degree of negotiation and ad hoc adjustment in the computation of expenditure needs, when it i s much more desirable to have a completely objective approach.35 The 2002 State Budget Law has introduced a reform requiring that targeted transfers be approved by the National Assembly, which may act as a check and balance on such negotiation. However, there i s still no system of capital norms. 6.53 Another issue in the estimation of expenditure needs i s that, although considerable progress has been made by the introduction of per capita expenditure norms to approximate the expenditure needs at the central-provincial level, many provinces still 35The adjustments are generally done to achieve greater equalization, such as th? priority given to the highlands, or they may be in response to particular needs, but the system would be more transparent if it were to use an explicit formula and only the formula for the calculation of needs. 90 Vietnam Public ExDenditure Review and Intenrated Fiduciary Assessment use the old physical standards or norms (such as number of staff) in the measurement of expenditure needs for districts and communes. This type of norm leads to negative incentives in expenditure decisions. 6.54 On the revenue side of the equalization methodology, the element of negotiation appears to be particularly acute. Revenue estimates have tended to be cautious, with insufficient allowance for the greater revenue buoyancy of richer provinces. This reflects the lack of a system of revenue norms or a reliable methodology for estimating revenue collections. 6.55 A second set of issues pertains to Vietnam's nascent conditional grant program. Up to now, conditional grants have been limited to special budget allocations for the "National Target Programs." An important task ahead for the Government i s to build more fully a conditional matching and non-matching grants program, for both recurrent and capital expenditures, based on well defined explicit sectoral policies with assigned roles for design and supervision of the sector ministries. Capital grants, in particular, may need to be better targeted by taking into account measures of capital infrastructure deprivation across provinces. Recommendations 6.56 In the shorter-term, the Government should consider augmenting the existing system of transfers with clearer revenue norms, based on stronger revenue-forecasting capacity; and by introducing explicit norms for capital expenditure. 6.57 The Government should consider in the mediumterm the reform of the system of transfers in Vietnam in two complementary ways. The first way i s the upgrading of the current "balancing transfer" methodology to a full formula-driven system of equalization grants, including a stable and explicit formula used to distribute the funds and possibly an explicit and stable rule to establish the pool of funds. The formula should equalize fiscal disparities among provincial governments on the bases of pro-poor expenditure needs and fiscal capacity. It will be important to apply methodologies that allow the quantification of revenue potential as opposed to actual revenues collected in order to avoid discouraging revenue mobilization at the subnational level. 6.58 The second activity will consist of strengthening the conditional grant system for pursuing national objectives through well defined sectoral policies in sectors such as education and heath, with and without matching provisions. Greater use of conditional grants could also be made to fund capital infrastructure programs. BUDGETARYAUTONOMY 6.59 The different aspects of the budget process (at all levels of government) are regulated in the 2002 State Budget Law. B y constitutional mandate, the National Assembly needs to approve the entire State Budget, which includes the central government budget and the consolidated budgets of all subnational governments (provinces, districts and communes). In principle, all subnational budgetary authority in Vietnam i s delegated by the central government. Decentralization to Subnational Government 91 6.60 However, the delegation of authority can be performed with quite different degrees of autonomy, The Government recognizes that the benefits of decentralization will not accrue unless an adequate degree of budgetary autonomy i s granted to subnational governments. Budgetary autonomy i s a multidimensional issue, which requires significant institutional development and which may take some time to accomplish. Therefore, the Government plans to develop a medium term strategy for reform in this area. This section considers some of the most important elements to consider in that strategy. Separation of budgets 6.61 In a fully decentralized system, subnational governments would be able to plan and approve their own budgets separately from those of the central government or any other upper-level government. This would enable full accountability to be transferred to the local level at which, in practice already, local expenditure decisions are largely made. If Vietnam moves in this direction, as seems likely at some stage, the constitutional mandate that the National Assembly must approve the entire State Budget, which includes the budgets of the provinces, as well as districts and communes, would need to be adapted accordingly.36 However, this will require considerable preparatory reform, including review of the optimal number of budget levels, reassignment of certain expenditures between central and local level, strengthening of National Target Programs and strengthening of local transparency and accountability. 6.62 The National Assembly and the national government have a critical role in determining national budget plans and overall inter-governmental fiscal transfers (including National Target Programs), and in monitoring implementation. However, it unclear to what extent they can realistically add value to the detailed elaboration of budgetsat commune and district level or even, arguably, at province level. The issue here i s not only timing inconsistencies at different levels, but also that the separation of budgets empowers subnational governments and promotes fiscal responsibility and horizontal accountability at the subnational level. Eliminating the current inconsistencies between provincial and central budget plans would certainly eliminate ambiguity and strengthen accountability. Sufficient autonomy and predictability inrevenues 6.63 Subnational budget autonomy requires that local governments be able to increase and decrease the size of their budgets b y using their own revenue sources, including taxes. The fundamental concept i s that subnational budgetary autonomy i s not possible without some meaningful level of discretion granted to subnational governments to change the rates of some taxes. Autonomy does not require that subnational governments be able to finance 100 percent of their expenditures from their own sources. Equity considerations demand that at least the poorer provinces continue to be supported in part from the center. However, it does require that subnational governments be able to 36 Currently, the budgets for the communes are approved by the districts, those of the districts by the provinces, and the provincial budgets by the National Assembly. In this way the budgets of the communes are approved four times. This methodology still follows the nested Matrouchka doll model of Soviet budgeting. 92 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment increase or decrease their revenues at the margin; for example, the ability to increase budget revenues by 10 percent for some special purpose, by raising the rates of one or several taxes. 6.64 To enjoy budget autonomy, subnational governments also need a sufficient degree of stability and predictability of revenues and expenditures. The introduction of stable rules and formulas, with explicit and stable revenue assignments at the sub-national level, are critical to increase budget predictability and certainty. This needs to apply to provincial governments as well as to districts and communes. Sufficient discretion in expenditure decisions 6.65 Subnational budget autonomy also requires that subnational governments have an appropriate degree of discretion in making expenditure decisions according to their own priorities and in deciding what i s the most cost efficient combination of inputs to deliver public services. The more explicit expenditure flexibility given to provinces in the 2002 State Budget Law i s a step in this direction. However, this needs to be balanced against providing incentives to make local choices consistent with national strategies and development goals. 6.66 At the same time, in a workable decentralized system of finance, the central government needs to find ways to induce subnational governments to pursue national objectives which are not invasive of subnational budgetary autonomy. One non-invasive way to address this type of concern i s to use conditional matching grants, whereby the central government induces subnational governments to spend more on particular items of national importance. However, this may be inequitable unless sufficient allowance i s made for the extent to which richer provinces are better able to afford matching contributions. Another tool, the imposition of ring-fenced minimum expenditure requirements on subnational governments in certain priority sectors, should be reserved only for sectors of the utmost national priority. 37 Recommendations 6.67 In the long run, the Government should consider developing a new inter- governmental fiscal system whereby full autonomy to approve subnational budgets will rest with People's Councils of corresponding level (Le. complete elimination of the "Matrouchka" structure). This would not, however, involve elimination of intergovernmental fiscal transfers in the budget system. 6.68 More authority should be delegated to subnational governments, especially provincial governments, in determining their own expenditure norms within a range set by the center. 37Currently, there are only two types of minimumexpenditure requirements imposed on local governments inVietnam: (i) governments will need to spend 18 percent of their budgets on education by 2005 and local 20 percent in 2010; (ii)local governments need to spend 2 percent of their budgets on science and technology. These levels are enforced at budget approval time by the National Assembly, although they are not easily monitored. These minimum expenditure requirements are not particularly invasive, given the special importance of education. Decentralization to Subnational Government 93 TRANSPARENCYFINANCIALACCOUNTABILITY AT SUBNATIONALLEVEL AND 6.69 For fiscal decentralization to work fully in Vietnam and deliver the expected increase in the overall efficiency of the public sector, the Government recognizes that it will be required to complement and in some cases replace the systems of vertical accountability (to central government authorities) with systems of horizontal accountability (to subnational government councils and to local residents themselves). In order to increase this horizontal accountability, i t will be required to increase the coverage and transparency of government budgets. All proposed and executed budgets, not only at the central but also the subnational level, will need to be fully accessible to citizens through newspapers and other media, and all budget decisions need to be made in public sessions open to the press and overall public scrutiny. This proper scrutiny of subnational budgets and the budget process itself will help ensure that government officials are accountable and eventually more responsive to the needs and preferences of residents, which represents the essential advantage of a decentralized system of governance over centralizedmodels. 6.70 There has been a significant ongoing effort in Vietnam to enhance the transparency of government budgets and policies, thereby increasing accountability. This trend has been apparent since the approval of the 1996 State Budget Law and has been enhanced with the new 2002 State BudgetLaw. The new law states that the budgetplans, budget final accounts and auditing results of the entire State Budget must be made public.38 It has become common practice to publicly announce budgetary policies, such as expenditure and allocation norms, budget processes, budget figures, data, and materials related to annual budget planning39 The final accounts for executed budgets at all levels of government are also made public. 6.71 In addition, the powers of the National Assembly and the People's Councils have been enhanced concerning approval of budget and monitoring of their execution. Another area where considerable progress has been made i s in the control and external audit of the public accounts. Under the 2002 State Budget Law, the Treasury i s now the lead agency responsible for expenditure control and financial management information at all levels of government. The role of the State Audit of Vietnam has been strengthened to conduct external audits covering all State Budget revenues and expenditures. The audit results are reported to the National Assembly and the auditing reports must be made public in accordance with the Government's regulations. 6.72 The Government recognizes that, despite these significant advances in budget transparency and publicity, further action may be required. It may be necessary to issue regulations on procedures for the publicity of auditing reports prepared by the State Auditing Agency. The area that will require most attention i s how to increase 38 To implement this stipulation, the Ministry of Finance has been preparing the draft of the (amended) Rules on FinancePublicity to be submitted to the PrimeMinister for his decision 39 In 1998, for the first time, the budget revenue and expenditure figures were published on the Annual Statistics Book issued by the General Statistics Office. Decision 182 of 2001 stipulates the content, timing and format of the publication of the state budget. The Ministry of Finance prepares annually a comprehensive report on the budget publicity implemented by budget planners of the central and provincial budgets. For 2002,58 provinces sent their publicity reports to Ministry of Finance. 94 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment participation and accountability to local residents, and how to increase monitoring of provincial, district and commune budgets. 6.73 Important reforms were initiated in 2003 in the area of external audit, when the National Assembly was given greater authority over the external ex post audit function. The Government plans to extend the coverage of external audits at the provincial, district and commune levels. Strengthening transparency and accountability at local level will take some time. Meanwhile, the PER-FA recommends that the Government should not consider further increasing the total share of subnational government in the State Budget untilcapacity and institutions for accountability andtransparency have strengthened. SUBNATIONAL GOVERNMENTBORROWING 6.74 Disciplined access to credit can be an appropriate source for financing subnational government capital investments for those expenditure responsibilities that have been assigned to them. Borrowing to finance good projects i s needed for various reasons, including bulkiness of some projects and lack of liquidity of subnational governments. In addition, the repayment of credit over time represents a fairer distribution of infrastructure costs among the different cohorts of users during the useful life of the infrastructure. However, borrowing at the subnational level i s risky and there i s a need for balance between access to borrowing by subnational governments with institutional mechanisms that preserve fiscal discipline (rules and regulations and, even better, private credit markets discipline). 6.75 The current rules for subnational government borrowing are set out in Article 8 of the 2002 State Budget Law and the Ministry of Finance monitors and enforces these rules on subnational borrowing through several mechanisms (including Treasury reports for the province, and the province's report to the Ministry of Finance during the meetings on budget estimates). So far, provincial governments have borrowed domestically from two sources-bond issues and institutional loans-and the level of debt at the subnational level i s quite low by international standards (0.43 percent of GDP in 2003.) The regulations make clear that provinces have to repay both principal and interest from their own future budget. 6.76 At first site, there appears to be little immediate threat of domestic financial instability caused by subnational government borrowing at this time. But given the serious deficiencies and needs for infrastructure investment at the subnational level across the entire country, the Government expects this level of debt to increase in the future. It i s critical that this occur in an orderly fashion, preserving fiscal discipline. This will require stronger systems for recording domestic debt stocks and flows and stronger processes for analyzing the associated risks. 6.77 To this end, the PER-FA recommends strengthening the current borrowing rules for subnational government. In particular, the debt limit rule that for a given year a province's stock of outstanding debt cannot exceed 30 percent of its capital budget may be substituted with rules using the more stable benchmarks common in international practice. For example, a better rule might be that total debt cannot exceed some percent of subnational revenues in any year, or expenditures on debt service must not exceed Decentralization to Subnational Government 95 some percent of subnational government revenues in any year. The Government should ensure the inclusion of all bond issuance by provinces in the State Budget. 7, DELEGATIONTO SPENDINGUNITS INTRODUCTIONAND OVERVIEW 7.1 A principle of good public expenditure management i s that accountability and authority need to go hand-in-hand. International experience suggests that, in order to deliver services efficiently and responsively, spending units need to be given a clear mandate; be assigned appropriate resources; and be held accountable for the way they use the resources to deliver their mandate. Experience suggests that if spending units are given resources but not held accountable for their use, funds are unlikely to be used properly and effectively. But experience also suggests that spending units can only be responsive and efficient if they have a degree of flexibility, and can only successfully be held accountable for the use of resources over which they have a real degree of control. Balancing authority with accountability i s a key challenge in managing the relationship between finance functions and spendingunits. 7.2 Increased delegation to spending units4' i s one of the key developments in public expenditure management in Vietnam in recent years. In parallel with the program of decentralization from central to subnational government, and with the wider Public Administration Reform (PAR) program, the government has delegated increasing amounts of budget discretion from the finance departments at different levels of government to spending units. Such delegation i s proceeding on parallel but separate tracks for Administrative Units (under Decision 192/2001/QD-TTg dated December 17, 2001) and for Service Delivery Units (under Decree 10/2002/ND-CP dated January 16, 2002). 7.3 The first part of this chapter provides an overview of delegation measures to date inVietnam. The second part describes their implementation to date. The third and fourth parts report on and discuss some of the initial impacts and risks associated with implementation of Decision 192 and Decree 10, respectively. OVERVIEW OF DELEGATION MEASURES INVIETNAM The H o ChiMinh City experiment 7.4 Reforms to delegate to administrative units began in December 1999, with Decision 230/1999/QD-TTg which authorized the province of H o Chi Minh City (HCMC) to pilot the introduction of "block grant" or "lump sum" budgeting, along with more flexible personnel arrangements, in 10 selected districts and departments inHCMC. These included four urban districts (Districts 1, 3, 5 and 11),three rural districts (Cu Chi, 40Inthis report, the term "Spending Unit" is used to refer collectively to administrative units and service delivery units. 97 98 Vietnam Public ExDenditure Review and Integrated Fiduciary Assessment BinhChanh and Nha Be), and three city-level departments (the Department of Transport, Communications and Public Works, the Department of Labor, Invalids and Social Affairs and the Department of Justice). 7.5 Under the 1996 State Budget Law then prevailing, finance functions at each level of government were required to follow a control regime which involved their placing some tight restrictions on reporting spending units.Spending units were required to agree their budgets with the finance function to line item detail; and they were then not allowed to reallocate spending between nine blocks of controlled expenditure without formal approval for such a "virement". In addition, spending units were obliged to comply with physical and allocative norms specified by sector ministries. 7.6 For the pilot administrative units, the prevailing control regime was replaced with a block granting approach somewhat similar to the "running cost" regimes of many OECD countries (see Box 7.1 on Australia). Under the block grant approach, administrative unit managers were given a significantly greater degree of budgetary flexibility within a fixed recurrent budget constraint, replacing the prevailing approach in Vietnam of approving and controlling detailed budgets for spending units. Box 7.1: Australia's Running Cost Systemfor Administrative Units Australia's running cost system allocates to line agencies a single block appropriation approved by Parliament. This appropriation covers all recurrent and minor capital costs, including salaries, administrative expenses, rental, and purchase of goods and services. Line managers have the freedom to allocate inputs as they see fit, including the number of employees and their salaries, subject to some limits. They can also keep funds from user charges, rental income from property and proceeds from the sale of under-performing assets. To optimize resource utilization, managers are allowed to carry over up to 10 percent of their operating budget into the next year or borrow up to 6 percent of next year's appropriation. Source: World Bank staff ~~ ~ ~ ~~ 7.7 The stated objectives of the pilot were to: 0 restructure administrative units and streamline administrative procedures; 0 rationalize administrativecosts; 0 reduce over-staffing; \ 0 raise incomes of employees by usingsavings from the measures above; and 0 increase levels of transparency. 7.8 Under the pilot scheme, budget appropriations to the 10 administrative units were converted into block grants, whose amounts were fixed for a three year period. Within these fixed budget constraints, administrative units could reallocate expenditure between line items, without having to seek special permission for a "virement" (a transfer of expenditures between approved budget lines). Administrative units were free to reduce staff numbers and to reprioritize between categories of administrative expenditures (with a few limited exceptions). Any savings could be retained and used to increase staff incomes through additional salary or bonuses. Thus administrative units could choose how to reduce material expenditures and eliminate unproductive staff; and the direct link between cost reductions and pay increases provided a powerful incentive to do both. Delegation to Spending Unit 99 However, administrative units were required to maintain previously specified levels and standards of service. 7.9 Significant financial impacts were quickly felt. All administrative units reported considerable reductions in expenditure on administrative items, with reported financial savings in the first year of the pilots ranging from 13 percent up to 29 percent. The majority of these arose from savings on communication expenses, utility bills such as fuel, water and electricity, routine repairs and maintenance, purchasing of goods and services, printing and production of documents. Nearly all administrative units also reduced their staff numbers below the official quota-most units reduced their staff by around 15 percent (compared to the staff quota), with some makingbigger cuts. 7.10 Although evidence remains quite limited, there i s little evidence to suggest that service quality has declined in the pilot administrative units. Official administrative data suggests indicators of service standards, such as waiting times, have if anything improved (with unnecessary layers of bureaucracy eliminated in the pursuit of savings). However, there i s some evidence to suggest that some of the savings in the pilot units may have reflected deferral rather than elimination of expenditures. Almost all of them reported sizeable reductions in expenditure on routine repairs and maintenance (although major repairs were not included in the block grant). Decision 192/2001and Decree 10/2002 7.11 On the basis of its experiences in HCMC, Government decided to extend the block granting experiment to administrative units in other provinces. To this end, in December 2001, the Prime Minister issued Decision 192/2001/QD-TTg. A month later, inJanuary 2002, the Government promulgated Decree 10/2002/ND-CP to delegate wide- ranging flexibilities to revenue-producing Service Delivery Units (SDUs). Under Decree 10,block grant budgeting for recurrent spending i s extended to all SDUs. 7.12 The essential provisions for Decision 192 and Decree 10 have some strong similarities. Both measures are aimed at delegating authority and creating incentives to encourage more efficient behavior and greater responsiveness in the spending units. Both measures convert State Budget appropriations to the spending units into a block grant that i s fixed (although with possible increases) for three years. Within this resource envelope, managers of the agencies are given substantial allocative control and encouraged to reorganize administrative structures and streamline procedures. To provide strong incentives to find efficiencies, the agencies are authorized to use the bulk of the savings to increase staff remuneration, within proscribed limits. 7.13 Nevertheless, the differences between the two measures are significant. First, they differ in scope. Decision 192 targets administrative units; it i s essentially part of a reform of bureaucratic administration. In contrast, Decree 10 targets service delivery units that provide the core public services to citizens, in education, health, culture and media, scientific research, sport, transport, agricultural services, etc. Its implementation will have far greater social impact because it will affect the delivery of these services, along with access to the services. At the same time, the potential gains from enhanced operational efficiency and effectiveness under the reform are also great, and are orders of magnitude greater than those from Decision 192. 100 Vietnam Public Expenditure Review and Intenrated Fiduciary Assessment 7.14 Second, Decree 10 changes the basis of financing for SDUs and will reduce their dependence on the state budget. For the first time, SDUs will be allowed to borrow from commercial banks to expand and improve services. They will be able to retain depreciation on capital, as well as receipts from liquidated assets, which had been remitted to supervisory ministries in the past. They will have more autonomy for setting user charges for non-core services (those not on the state regulated list, such as parking and laundry fees in hospitals, as well as new services). 7.15 Third, Decree 10 will change the incentive structure guiding SDUs by encouraging them to expand revenue resources and allowing them to use the receipts for reinvestment and staff remuneration. UnderDecree 10, SDUs will have greater discretion for widening wage differentials. While Administrative Units may grant pay increases up to 2% times the basic salary, SDUs managers may grant pay increases up to 3% times the basic salary, subject to review and agreement by the relevant union and subject to disclosure within the SDU. 7.16 Fourth, Decree 10 allows SDUs to hire additional staff on a contractual basis as needed. As such, it amounts to a step away from strict establishment control. 7.17 In sum, Decree 10 goes quite far in introducing market-like incentives to encourage SDUs to mobilize resources and improve efficiency. Altogether, these provisions confer allocative control to SDUs over both revenues and expenditures that far surpasses that given to administrative units under Decision 192. B y giving SDUs access to factor resources (capital and labor) outside of government control, these autonomies could potentially allow SDUs to change the scale, scope and orientation of their services. In short, Decree 10 will affect what services government provides, and for whom. By reducing the dependence of SDUs on the State Budget, Decree 10 will also significantly change the relationship between SDUs and their supervisory agencies. The 2002 State BudgetLaw 7.18 At the time of the HCMC experiment, and when Decisions 192 and Decree 10 were promulgated, the counterfactual to block granting was the 1996 State Budget Law. However, in considering the future of these reforms, it i s important to recognize the extent to which delegation of budgetary flexibility to spending units has since been embodied in the 2002 State Budget Law (which was approved by the National Assembly inDecember 2002 andbecame effective inJanuary 2004). 7.19 The 2002 State Budget Law has, to some degree, extended the principle of recurrent budget flexibility to all spending units. Specifically, the number of controlled blocks of expenditure has been reduced from nine items to just four broader blocks, and to only three broader blocks in the case of recurrent expenditure: "personnel expenditures" and "operations and maintenance expenditures" and "other expenditures". Furthermore, allocative norms have been revised (although many physical norms remain unchanged). As such, it has introduced more budgetary flexibility for spending units on a wide scale basis. This i s a more limited budgetary flexibility than under Decisions 192 and Decree 10: under the 2002 State Budget Law, three recurrent expenditure blocks remain controlled separately. Nevertheless, it i s important to recognize that Government Delegation to Spending Unit 101 has, in effect, already mainstreamed an important part of the HCMC approach; and to ensure that we evaluate Decisions 192 and Decree 10against the changed counterfactual. IMPLEMENTATIONOF DELEGATION DATE TO Implementation of Decision 192 7.20 The differences between the two measureshave to some extent shaped the pace of their implementation. Decision 192 has been rolled out slowly and deliberately. Considering that the HCMC pilot began in 2000, the block grants to administrative units are now in their fifth year of piloting. Por the time being, the Government's decision i s that this remains a pilot that will be reviewed for ongoing improvement. 7.21 During 2002, the first year of roll-out beyond the HCMC pilot, 22 provinces and cities introduced block grants to 116 administrative units. In a May 2003 meeting, the Ministry of Finance called for a target of at least 1 to 10 Administrative Units under every ministry and locality to have adopted the pilot block grant on staff number and administrative expenditure by the end of 2003. Data available indicate that the target was not met, with the central ministries being the chief laggards. Table 7.1 shows that as of December 31, 2003, only three ministries and three agencies at the central level had implemented block grants. Much of the implementation has been in the Ministry of Finance (MOF). Under the MOF, in 2002 block granting was extended to 122 subordinate agencies under the General Tax Department and the Treasury. In 2003 it was extended to another 33 subordinate units under Customs, bringing the total under block granting to 155. . Table 7.1: Implementationof AdministrativeUnitDelegationat Central Level (December31,2003) 1. Ministry of Finance (total 155 subordinate agencies piloting) General Department of Customs General Department of Taxation State Treasury 2. Ministry of Transportation Vietnam Road Administration River Road Administration 3. Ministry of Labor, War-Disabled and Society Social EvilPrevention Administration 4. Vietnam Television (Implementation carried out as per Decision 87, 01/06/2001 of the Prime Minister) 5. Vietnam Radio Station (Implementation carried out as per Decision no. 85, 06/05/2003 of the Prime Minister) 6. Vietnam Social Insurance Source: MOF implementation reports 7.22 At the provincial level implementation has gone somewhat faster, with 35 percent of 1417 Administrative Units involved. However, implementation i s unevenly distributed across regions: while 6 provinces have implemented block granting in all of their departments, only 15 provinces have implementation rates of 50 percent or better, and only 22 have implementation rates of 30 percent or higher (see Table 7.2). As of year-end 102 Vietnam Public ExDenditure Review and Intearated Fiducian,Assessment 2003, 12 provinces still had not introduced block grants to any department. In 2004 it i s expected that 748 provincial agencies are piloting block granting (MOF report 2004). Table 7.2: Implementationof Decision 192 at ProvincialLevel (December 31,2003) Province Number of Number Percentage of total Departments Implementing % BenTre 22 22 100 BinhDinh 24 24 100 BinhDuong 22 22 100 BinhPhuoc 23 23 100 BinhThuan 23 23 100 DongThap 21 21 100 An Giang 19 18 95 Vinh Long 22 18 82 Yen Bai 30 23 77 Thai Nguyen 21 13 62 QuangNam 25 15 60 Vinh Phuc 22 13 59 NgheAn 24 13 54 NinhThuan 23 12 52 GiaLai 28 14 50 Ho ChiMinh 22 11 50 DongNai 21 9 43 CanTho 23 9 39 TienGiang 20 7 35 PhuTho 23 8 35 Cao Bang 23 7 30 DaNang 23 7 30 Source: MOF implementation reports 7.23 At the district level, implementation rates are lower. On average only 22 percent of district offices have implemented block granting. Even in HCMC, only half of the 22 districts had implemented block granting by the end of 2003. Implementation i s only just beginningat the commune level, where the rate of block granting i s near zero. Implementationof Decree 10 7.24 Unlike Decision 192, Decree 10 i s not formally being piloted. Rather, the responsibility for implementation i s assigned to sector ministries, which are tasked to issue implementation guidelines for SDUs in their sector. This has resulted in an extremely uneven pace of implementation since capacities differ greatly across ministries. As of 2003, as Table 7.3 illustrates, some ministries had fully embraced the decree and reached 100percent implementationwhile others have not yet begun. 7.25 B y the end of 2002, 29 of the 45 central ministries and agencies had reportedly finished classifying their SDUs into revenue-producing and non-revenue producing units, issued implementation guidelines and begun implementation of Decree 10. Ministries that had completed Decree 10 implementation in all their SDUs included the Ministry of Industry, the Ministry of Construction, the Ministry of Trade, Ministry of Labor, Invalid and Social Affairs; and the General Department of Quality Measurement. Not far behind were the Marine Department of Vietnam (92 percent), the Ministry of Fishery (90 percent), the Ministry of Agriculture and Rural Development (80.2 percent), and Ministry Delenation to Suendinn Unit 103 of Education and Training (77.2 percent). For understandable reasons, given the major challenges for their sectors posed by Decree 10, the Ministry of Education and Training, and Ministry of Health, were slower in issuing implementation guidelines. Table 7.3: CentralMinistriedAgenciesImplementingDecree 10(2003) Ministriesand Sectors Number Number % of SDUs Implementing Implementing Ministry of ForeignAffairs 4 4 100 Ministry of Justice 4 4 100 Ministry of Trade 16 16 100 Ministry of Lab, War-Inv. & SOC.Affairs 24 24 100 Ministry of Construction 27 27 100 Ministry of Industry 51 51 100 Ministry of Scienceand Technology 19 19 100 Office of the Government 4 4 100 GeneralTourism Administration 7 7 100 GeneralDepartmentof Statistic 4 4 100 VietnamNews Agency 1 1 100 State Audit 1 1 100 Ministry of Agriculture & RuralDev. 98 96 98 Ministry of Fishery 10 9 90 Ministry of Nat.Resources.& Environment 24 20 83 Ministry of Educationand Training 57 45 79 Ministry of Finance 15 11 73 Ministry of Transport 56 40 71 All central ministries 622 388 62 Source: Ministry of Finance 7.26 At the subnational level, the pace of implementation was similarly uneven, as Table 7.4 illustrates. By December 2003, 35 provinces had finished the classification process and begun implementation of Decree 10 in 3,937 SDUs, including 251 fully self- funded SDUs and 3686 partially-funded ones. Altogether 4,361 income generating- service units, or 27.2 percent of the total number were implementing Decree 10, among them424 under the control of ministries and central agencies, and 3,937 under the control of subnational governments. 7.27 In May 2003, MOF estimated that, nationwide, 70 percent of all revenue generating SDUs were implementing delegation, and it was expected that by the end of 2003, the remaining 30 percent would have begun. INITIAL IMPACTS AND RISKSINTHE IMPLEMENTATION OF DECISION 192 Impacts 7.28 Implementation results under Decision 192 are only just being collected and analyzed. Broadly, they appear to be quite similar to those inthe HCMC pilots. 7.29 Pilot Administrative Units have clearly been given a strong incentive to use resources more economically. Of the 107 pilot agencies reporting results for 2002, 104 Vietnam Public ExDenditure Review and IntePrated Fiduciarv Assessment savings under block granting totaled VND14,604 million in 2002, equal to 13 percent of the total expenditure block grants.41 7.30 Savings came from staffing reductions and other administrative cost reductions. On average the agencies reduced staffing levels by 7 percent, with higher rates in some localities.42 Table 7.4: Provinces Implementing Decree 10 (December 2003) Province Number of SDUs No. implementing Percentage Bac Ninh 48 48 100 BinhDuong 29 29 100 Lang Son 405 405 100 NamDinh 341 341 100 Quang Binh 65 65 100 Thai Binh 359 359 100 Thai Nguyen 52 52 100 Thanh Hoa 786 786 100 H a Noi 798 790 99 Ben Tre 590 582 99 Hoa Binh 103 97 94 Hai Phong 72 67 93 Quang Ninh 79 71 90 Bac Lieu 143 119 83 H a Tay 453 343 76 Long An 28 21 75 NinhBinh 370 253 68 Hung Yen 452 243 54 Ha Tinh 19 10 53 Gia Lai 73 37 51 Da Nang 75 37 49 BinhDinh 33 16 48 Dong Nai 240 113 47 Nghe An 60 25 42 Lam Dong 149 52 35 Kien Giang 19 6 32 Total 12,474 5703 46 Source: Ministry of Finance 7.31 On the basis of these savings, they were able to increase staff incomes by some VND 55,000 to 507,000 per month. Among the localities achieving high salary increases were Ha Giang-with average increase of VND120,OOO per month; Quang Nam- VND150,OOO per month; Tien Giang-VND186,OOO per month; Gia Lai-VND203,OOO per month; and Hanoi-VND319,OOO per month. 7.32 Results for 2003 showed incomes continuing to rise through the piloting period. For examples, in Hanoi average annual incomes increased by VND2.2 million in 2002, and rose further in 2003, to VND3.5 million. In Danang average incomes rose by 41Reports from some provinces were: Dong Thap (6 percent), Hanoi City (6.1 percent), An Giang (12.7 percent), Ben Tre (12.8 percent), Gia Lai (14 percent), Tien Giang (18 percent) and H C M City (26.4 $?cent). This meant a reduction of 357 staff from the total of 5046. Reduction rates for reporting provinces were: Tien Giang (6.8 percent), Ben Tre (7 percent), and Ho Chi MinhCity (14 percent). Delegation to Spending Unit 105 VND3.4 million in 2002 and by VND3 million in 2003. In HCMC, they rose on average by VND4.8 million in 2002 and VND3.9 million in 2003. In Baria Vung Tau they rose by VND2.5 million in 2003 (MOF 2004). Risks 7.33 At the time of the HCMC experiment, there were some concerns that larger-scale roll-out might result in less striking savings, to the extent that the pilot administrative units were volunteers who anticipated substantial savings; and to the extent that avenues for cost-shifting would be closed down. However, the experience of Ben Tre Province does not support this concern. Ben Tre, which was the first province to implement Decision 192 in all Administrative Units (including 22 provincial departments and 8 district and town agencies), reported budget savings of VND3,262 million, equal to 12.8 percent of the block grants for all pilot units. These savings came from a substantial streamlining of organizational structures, which resulted in a reduction of 11 provincial departments and divisions. Staff reductions were also reported: among the 103 staff who were cut, 27 retired, 52 took early retirement and 24 were transferred. Average increase instaff incomes came to VND205,OOO per month. 7.34 Nevertheless, many of the impacts of the Decision 192 pilots are not yet clear. First, pilot administrative units could have made cost savings by reducing the level or quality of services provided to their clients, in ways that have not become apparent through the current mechanisms for monitoring service delivery. Little evidence exists to suggest that the performance of Administrative Units has declined. But as discussed in Chapter 4, Vietnam's current systems for monitoring the outputs, impacts and outcomes of public spending remain very weak. Vietnam has only recently started piloting the "citizen score card" approach (see Box 4.7 in Chapter 4). 7.35 Second, wider roll-out of the reforms poses a significant challenge to the coherence of national pay policy and increase the risk of salary costs squeezing out other expenditures. This i s particularly the case given the pent-up pay pressure that exists in parts of the public sector in Vietnam. 7.36 Third, apparent gains could have been offset by hidden or longer-term costs. One possibility i s that apparent savings have been generated by economizing on maintenance of capital assets. The item for major maintenance and repairs was excluded from the block grants for precisely this reason, but the potential remains for skimping on routine maintenance, which would be at best a deferral of expenditure rather than a genuine saving. Assessmentand recommendations 7.37 The introduction of block granting in Administrative Units under Decision 192 i s a modest reform in comparison with Decree 10. On balance, the benefits appear to be non-negligible and risks manageable. Nevertheless, more detailed and sustained evaluation i s required.As such, the PER-IFA recommends that the Government maintain the pilot status of Decision 192, with continued systematic monitoring and evaluation. 7.38 In order to strengthen this monitoring and evaluation, it is important that the Government obtains better performance information on the quality of administrative 106 Vietnam Public Exvenditure Review and Intewated Fiducianl Assessment service provided by such units. The PER-FA recommends that the Government use the "citizens score card" approach to help assess the impact on citizen satisfaction of block granting in the pilot agencies; and that this should be supplemented by strengthened systems for administrative monitoring of performance. The results of such performance monitoring should be made public. 7.39 In conducting monitoring and evaluation of Decision 192 pilots, the Government should bear in mindthree key issues. First, i s clear that Decision 192 has been associated with an increase in the share of wages and salaries in Administrative Unit spending; and with a widening of pay differentials between pilot and non-pilot units. It could be argued that neither of these developments i s cause for concern; that wages and salaries as a share of the total budget are rather low in Vietnam by international standards; and that the pay differentials between pilot and non-pilot units would be reduced were all Administrative Units operating under Decision 192. Nevertheless, experience with Decision 192 has certainly suggested that there may be pent up pay pressure in the public administration sector in some parts of Vietnam. Chapter 3 has already noted the need for a clearer public sector pay and employment strategy. The PER-IFA also recommends that the Government consider constraining the flexibility that managers have over salaries and remunerations more tightly, whilst still allowing sufficient flexibility to create an incentive for efficiency. 7.40 Second, it i s possible that Decision 192 may be putting excessive downward pressure on operations and maintenance expenditure. Serious consideration should now be given to whether block granting separately for wages and salaries may be preferable given the current inadequacies of the public sector pay structure in Vietnam. Until such time as the public sector pay structure i s reformed, and performance information strengthened, it may be desirable to constrain the potential for pilot units to pay themselves higher salaries by scrimping on routine maintenance. 7.41 In this regard, it should be recognized that the 2002 State Budget Law has effectively mainstreamed the key principle of the Ho Chi Minh City experiment, extending considerable budgetary flexibility to all spending units in Vietnam (as discussed above). 7.42 Starting from the counterfactual of the 1996 State Budget Law, the implementation of Decision 192 resulted in a move from nine separate blocks of control, with complex procedures for virement between them, to block granting for all recurrent costs together. Eleven blocks of control under the 1996 State Budget Law was almost certainly excessive. 7.43 Under the 2002 State Budget Law, however, the implementation of Decision 192 will result in a move from three separate block grants ("personnel", "operations and maintenance" and "other") to a unified block grant for recurrent expenditure. The PER- F A concludes that the optimal choice between these options remains an open question requiring ongoing and deeper evaluation. Delegation to Spending Unit 107 INITIAL IMPACTS AND RISKSINTHE IMPLEMENTATIONOFDECREE 10 Impacts 7.44 Decree 10 has much wider objectives than Decision 192, and the impacts are also much wider. Decision 192 i s focused on creating flexibility and incentives for cost saving and productivity improvement. B y contrast, the primary objective of Decree 10 i s to create flexibility and incentives for introduction of new and better services, and for mobilization of additional resources to finance these developments. Cost saving ,and productivity improvement i s also a goal of Decree 10, but the reform goes well beyond this. 7.45 Like Decision 192, Decree 10 implementation started only in 2002. Although the Decree has been implemented for two years now, it i s still too early to make a full assessment of its impacts. Results are only beginning to be collected and analyzed. Nevertheless, it seems quite clear that the flexibilities and incentives provided under Decree 10 are having powerful impacts, some apparently positive and some worrying. Mobilization of revenues 7.46 Revenues appear to have increased dramatically in the pilot SDUs. In 2002, receipts from service delivery for the 13 SDUs under ministries and central agencies that were implementing Decree 10 were 15.2 percent greater than planned, at VND2997 billion, with increases of 5.1 percent in the 4 SDUs under the Ministry of Justice; 38.2 percent in the 4 SDUs under the Ministry of Industry; and 38.3 percent in the 5 SDUs under the Ministry of Transportation. Compared to revenues in 2001, the increases were 13.6 percent in the 4 SDUs under the Ministry of Justice; 130.7 percent in the 4 SDUs under the Ministry of Industry; and 41.6 percent in the 5 SDUs under the Ministry of Transportation. 7.47 At the provincial level, revenues in 2002 for the 575 SDUs in Ben Tre, Dong Thap, Tien Giang and Vinh Long provinces that were implementing Decree 10 increased by 16.9 percent, with increases of 21 percent in the 16 units in the health sector; 20 percent in the 529 units in the education and training sector; 9 percent in the 4 units inthe scientific research sector, and 14 percent in the 26 units in other sectors. 7.48 User charging for public services i s not a new phenomenon. Hospitals have for more than ten years relied on user charges, including the sale of drugs, as supplementary support. In the late 1980s the public health system was facing a severe financial crisis, with inadequate funds to maintain regular activities, and with staff who went unpaid for extended periods. In 2000 estimates from the national health accounts showed fee revenues accounting for 22.5 percent of hospital revenues nationwide-substantial additional resources for the public health system. The growing role of fee revenues in financing hospital services i s shown inFigure 7.1. 108 Vietnam Public Exuenditure Review and Intenrated Fiduciarv Assessment Figure 7.1: The Growing Role of Fees inHospital Finance 5000 4500 g 4000 3500 B 3000 2500 2000 1500 1000 500 0 0 State Budget IUser fees Source: Phuong (2003) 7.49 Nevertheless, it i s clear that Decree 10 creates powerful new incentives for revenue mobilization. A dramatic illustration of this i s provided by St. Paul's Hospital in Hanoi. With the introduction of "on demand services" (see Box 7.2), revenues from this source rose rapidly from VND1.67 billion in 2000 to VND29.43 billion in 2003, quickly overtaking State Budget allocations and insurance reimbursements to account for over 50 percent of the hospital's total revenues (Table 7.5). Table 7.5: Sources of Revenue inSt. Paul's Hospital, Hanoi Sources of revenue 2000 2001 2002 2003 State budget allocations 49.9% 40.4% 23.6% 21.5% Insurance reimbursements 42.5% 39.0% 28.8% 28.2% Revenues from "on-demand health services" 7.6% 20.6% 47.6% 50.3% Total (millions VND) 21,877 31,458 52,095 58,476 Source: PER-IFA inputfrom St. Paul's Hospital, Hanoi Introduction of new services 7.50 Given the strong incentives to generate revenues, schools and training centers have quickly introduced a variety of new training programs, including full-time and part- time training courses, and distance-learning programs tailored to clients' needs. Some large universities such as Hanoi National University and H o Chi Minh City National University have set up joint programs with foreign universities to offer courses and even joint degree programs taught by foreign professors, some of which include exchange programs to send students abroad. 7.51 Likewise, hospitals and health centers are vying to offer new services and treatment programs. For example, two months after Decree 10 was implemented, the Saint Paul Hospital in Hanoi introduced a program of "on-demand health care services" that provides high quality health care to the private sector on a cost-recovery basis (see Box 7.2). Cost reduction andproductivity improvement 7.52 Implementation of Decree 10 appears also to have brought savings in recurrent expenditures of about 3 percent to 10 percent. These savings, together with additional Delegation to Spending Unit 109 revenues have enabled the SDUs to raise wages and salaries by sometimes substantial amounts, but these amounts vary significantly across ministries and regions. Box 7.2. Introduction of Cost-recoveryPricinginSaint Paul's Hospital, Hanoi Health services on-demand: By Decision of the Hanoi Health Department, with approval by the Hanoi People's Committee, St. Paul's Hospital was approved for providing health services on-demand in the areas of medical consultation (including hospital stays), diagnostic tests, X-ray and ultrasound, electrocardiogram, electro encephalogram and surgery. Onpricing: The decision authorized the head of the hospital to set fees and charges "under the principle of reasonableness, transparency, and equality, ensuring to cover all costs and have accumulation," and to report the fees and charges to the Health Department. Source: PER-IFA inputfrom St.Paul's Hospital, Hanoi 7.53 For the 13 SDUs in the Ministry of Justice, Ministry of Industry and Ministry of Transportation, the increases averaged 52.3 percent compared with the salary levels set for the work positions and grades, varying from 32.7 percent in the Ministry of Justice, 87 percent in the Ministry of Industry, and 41.8 percent in the Ministry of Transportation. Other SDUs achieved even bigger increases in staff incomes. For example, Hanoi University of Law had increases of 33.2 percent, Industry Advanced College No 4 had increases of 146.5 percent, and the Transportation Hospital No 1 was able to double wages for its staff. 7.54 Increases in staff wages seemed on average more modest for local SDUs, but variations across localities were also large. According to the reports from 575 local SDUs, salary increases averaged 4 percent in 2002, with increases of 3 percent in Ben Tre province; 15 percent in Dong Thap province; 23 percent in Tien Giang province; and 46 percent in Vinh Long province. 7.55 As for Administrative Units piloting Decision 192, too little is known about the impact of Decree 10, and the associated cost reduction and salary realignment, on service delivery. As such, it i s not possible to say at this stage whether these amounted to a genuine productivity improvement. Such evidence as there i s does tend to suggest that outputs from SDUs piloting Decree 10 have increased, but i t i s difficult to assess whether unit costs have fallen, or whether this merely reflects the dramatic increase in revenue mobilization. 7.56 It i s important to note that Decree 10 i s to a large extent simply formalizing and creating a regulatory regime around practices that have existed for some time. In many middle schools, the practice of two-tiered education began more than ten years ago (Box 7.4). Before Decree 10 in BinhDuong General Hospital, for example, a group of doctors had loaned money to the hospital for equipment purchases under profit-sharing arrangements (see Box 7.3). Even the existence of on-demand, high quality services at high prices, of the type offered by St. Paul's Hospital, predates Decree 10 in many provinces. The status quo i s no longer an option. Implementation of Decree 10 may create an opportunity to introduce a framework for rationalizing and managing these practices and abolishing questionable ones. 110 Vietnam Public Expenditure Review and Integrated Fiduciaiy Assessment Box 7.3: DoctorsInvestinMedicalEquipmentinBinhDuong Under a profit-sharing arrangement, doctors in the BinhDuong General Hospital made loans for the purchase of equipment and took partial ownership (with a limit of 40 percent ownership). As an example, in 1999 they helped the hospital buy a much needed ultrasound machine under this joint- ownership arrangement. The purchase price was VND300 million. The agreed arrangement was that the doctors would be repaid within three years with a monthly interest rate of 1.2 percent (compared to 0.7 percent for a bank loan). Thereafter, profits were shared among the owners. Since hospitals did not have the formal right to borrow, this seemed an innovative solution to meeting the hospital's need for equipment. The equipment purchases were approved by a hospital committee, which also set the loan terms and user charges. However, the conflicts of interest are serious in such practices. Giventhe information asymmetry in medical services, and given the lack of alternative providers, the doctors could easily abuse the set-up by recommending too many ultrasound scans. With the fee set at VND30,OOO per use and a planned utilization rate of 30 times per day, the arrangement was highly profitable for all43. Source: PER-ZFA field visit ~~ Box 7.4: Two-tieredEducationinHanoi The Phan Dinh Phung Senior Middle School in Hanoi operates two shifts for its 2800 students, divided into 53 classes. Aside from the separation into morning and afternoon sessions, the students are also divided into Grade A and Grade B categories. Grade A students are those admitted under the examination system with the required scores. Grade B students are "additional" admissions who had not met the scoring requirements or residential requirements. Grade B students are "extra-plan" and "extra-budgetary". They are admitted by the school with special approval of the Hanoi Provincial Department of Education and Training (DOET), to help pay the schools expenses. Currently they account for 30 percent of the student body. When the Hanoi People's Committee allocates budgetary funds to the school, it uses per-student allocation norms that count only the Grade A students. Likewise, the staff quota for teachers i s also basedon Grade A students. The school itself "stretches" these resources to cover the Grade B students. Grade A students pay tuition fees of VND30,OOO per month, 20 percent of which are remitted by the school to the DOET. Grade B students pay tuition fees of VND90,OOO per month, only 3 percent of which i s remitted. The fees they pay have thus become an important source of funds for the school to supplement the inadequate state budgetary appropriations. In 2003, when the Government raised minimum wages for the teaching staff by 38 percent (from VND 210,000 to 290,000 per month) for the minimumpay but did not appropriate sufficient additional funds, the school had to meet the gap through its own resources - tuition fees. The practice of admitting a second tier of students began almost ten years ago. At that time, schools were able to set the level of additional intake as well as the tuitions charged. Since Decision 73 was issued by the Hanoi DOET in August 2000), tuition has been set at a uniform level of VND90,OOO. The size of Grade B intake is negotiated between the school and DOET. Source: PER-IFAfield visit Risks 7.57 It i s clear that Decree 10 involves far greater risks than Decision 192. As Decree 10 enables and incentivizes SDUs to mobilize revenues through offering new services, 43 If the utilization rate of ultrasound scan is real, the revenue from this equipment can be over VND 300 million. Delenation to Svendinn Unit 111 the case of St. Paul's Hospital illustrates the powerful effect this can have. Inquick order, this could shift resources dramatically inthe affected sectors. 7.58 A primary concern i s the risk of deteriorating accessfor the poor and vulnerable. Decree 10 concerns service delivery units that provide core public services such as education and health: services with critical implications for Vietnam's development and major impacts on its poor. How the decree i s implemented could profoundly affect the quantity, quality and access to these services enjoyed by citizens. If well managed, Decree 10 can be used to increase user charge collection for non-core services, but also to ensure that extra State Budget resources are freed up for core public goods and to support access for the poor. If less well managed, Decree 10 has the potential seriously to undermine Vietnam's social progress and the egalitarian and cohesive structure of Vietnamese society, by creating uniform incentives for SDUs to neglect social mandates and focus on provide ever better services for those able to pay. 7.59 Vietnam has been remarkably sensitive to protecting the vulnerable groups in society. For example, the introduction of Health Funds for the Poor under the Prime Minister's Decision 139/2002/QD-TTg dated October 15, 2002 was a significant step toward providing access to health care for the poor, supplementing the wide range of fee exemptions already in place. The Government has also decided to provide free healthcare for all children under 6 years of age. At the level of health care facilities, too, there are some informal arrangements to ensure cross-subsidization from fee-rich activities to support fee-exemptions. For example, the Binh Duong General Hospital undertakes a variety of activities that are explicitly pro-poor, including a welfare kitchen that provides free meals to patients who are poor; fee exemptions to poor patients (in 2003 this amounted to 1.8 billion dong); and formation of a mobile team of doctors who go each month to remote rural areas to provide care to patients. In St. Paul's Hospital, 2 percent of revenues from on-demand services i s designated to fund fee exemptions and reductions for the poor. 7.60 Nevertheless, the effectiveness of these measures depends critically on the adequacy of funding available, and therefore the introduction of measures such as Decision 139 and the policy of free healthcare for children under 6 years old must be closely coordinated with the program of fiscal decentralization. The costs of Decision 139 are shared between the central and provincial governments, and the financing burden for the fee exemptions are jointly shared by the central government, the provincial government, and the health facilities. These cost-sharing arrangements tend to leave the populations in poor and mountainous regions more vulnerable, where local resources (be they local fiscal resources or own revenues of the SDUs) are more likely to fall short of expenditure needs. This i s a longstanding problem that had produced large regional differences in service standards in the past, and may worsen with the recent decentralization. 7.61 In addition, Decree 10implementation adds urgency to the need to expand health insurance coverage for everyone. At present the coverage of medical insurance schemes remains limited, and most of the population in Vietnam pays for health care directly out of pocket. This exposes them to risks of major - even catastrophic - health costs and impoverishment. 112 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 7.62 A second and related concern i s the need to strengthen mechanisms for monitoring and enforcing delivery of core services. The weakness of performance information on service delivery i s a generic problem of public expenditure management in Vietnam which has already been noted. However, Decree 10 will increase the flexibility and the incentives for SDUs to diverge from their core missions. This i s a problem that i s particularly salient in the health sector, where user charging i s applied to curative care but not generally to preventive care. Under the current incentives provided by Decree 10, the fact that user charging can contribute directly to staff incomes could bias providers towards reallocating resources to fee-producing services and neglecting public health and preventive health care services. 7.63 A third concern i s protecting against price gouging in monopoly markets. When the SDU i s the sole provider of its service in an area, local residents have little choice and user charges can become less "prices" for voluntary transactions and more like quasi- taxes. Competitive markets for many of the services of SDUs have not been developed in Vietnam and institutions for regulation remain often weak. In sectors such as health, where consumers are often dependent on suppliers to tell them what they need, potentially perverse incentives exist: health care costs could escalate as providers prescribe too many drugs and diagnostic tests, as has happened in China (see Box 7.5). ~ ~~~ Box 7.5: The Unraveling of Public Provision of Health Care inChina Like Vietnam, under the planned economy China had built a system of low cost provision of health care through publicly supported health institutions. Under that system, health care was available free of charge to all, and subsidies were allocated more or less on a needs basis. Starting in the early 1980s, however, the government reduced budgetary support to health institutions and pushed them to rely more on user charges. Institutions were offered greater autonomy for expanding into new areas of service and allowed to use a part of their self-raised revenues to pay bonuses and spend on staff welfare-incentives much like those offered under Decree 10 in Vietnam. Under this decentralized system, health care services grew rapidly with economic growth. Health spending now accounts for 5.5 percent of GDP. The health sector remains overwhelmingly dominated by state-owned institutions, which account for more than 90 percent of all hospitals and hospital beds and 100 percent of maternal-child health centers, although a significant portion of outpatient care i s provided by the private sector. The anomaly i s that government funding accounts for only 15 percent of total health expenditures today, out-of-pocket payments account for 60 percent, and the largely unregulated, albeit public, providers are reliant on fees for the bulk of their income. This has turned out to be an unhappy formula for delivering health care in China. Despite the government's efforts to control service fees, the health system faces severe problems of cost escalation. Hospital costs grew at annual rates of 16percent through the 1990s. Symptoms of supplier-induced demand are pervasive, with doctors over-prescribing drugs and prescribing unduly costly-and possibly inappropriate-drugs. Providers are acquiring equipment to compete with one another in performing revenue-producing tests, and offering services they are not mandated-and perhaps not qualified-to deliver. The result is not only an inefficient service mix, questionable quality of care, and a high degree of duplication, but also the displacement of essential activities such as basic curative care, public health programs, outreach and support and supervision of rural institutions. Most notably, the collapse of the rural Cooperative Medical System (CMS) has left the bulk of the rural population exposed to the risk of impoverishment through high out-of-pocket payments for health care. Medical care is increasingly out of reach for the poor-survey studies find that the poor tend more often to forego recommended treatments, get themselves released from hospital stays against advice, and resort to more self-medication. Source: World Bank staff Delegation to Spending Unit 113 7.64 In the health sector, both the opportunities and the risks seem clear. Positive impacts are already being observed: hospitals are energetically striving to increase efficiency and pursuing new avenues for attracting revenue-generating patients. In the process they are innovating and competing with each other to offer more and better services. But the current incentives remain largely for providing high cost services to those who can pay. The high costs could in turn reduce access to health care by the poor and other vulnerable groups. It could lead to unhealthy competition where hospitals keep the easy cases and shunt difficult cases to other hospitals. 7.65 Inthe education sector, Decree 10 is currently not being implemented at primary and lower-secondary school level. However, evidence from Participatory Poverty Assessments suggests that existing charges in such basic education remains a barrier to education in poorer areas. As such, it i s likely that further incentives for revenue collection in basic education would be extremely unhelpful for poor people. Well designed user charges (including perhaps for higher education) can contribute to economic efficiency and poverty reduction, by reducing subsidies to the better off; badly designed user charges can be economically inefficient (for example, by discouraging socially beneficial participation in basic education) and can aggravate inequalities across households and regions. Assessment and recommendations 7.66 Decree 10 adds up to a radical change in how SDUs are managed in Vietnam, albeit a change which builds on a long-standing direction of reform. Experience to date suggests that i s likely to have powerful impacts. Given more flexibility and stronger incentives, local governments have shown some significant capacity to innovate, both in terms of responding to local demands and conditions and in terms of achieving cost savings. However, experience also suggests some major risks which will need to be carefully managed. 7.67 Given the substantial risks and the limited evaluation work conducted to date, the PER-FA recommends that mechanisms need to be further developed to protect the access of the poor and near-poor to essential social services. However, the World Bank team recommends that the Government should move forward cautiously with the implementation of Decree 10, allowing time for experience to accumulate, detailed evaluation to be conducted and measures developed and implemented to mitigate potential risks; whereas the Government team feels that such "socialization" needs to pushed forward faster for those services that are affordable and acceptable to the public, with SDUs converted so as to operate under SOE law. Any expansion of socialization should be conditional on having in place a mechanism to protect access of the poor and the near-poor to basic social services. 7.68 As with Decision 192, there should be continuous review and assessment of the impact of Decree 10, using the "citizens score card" approach to help assess the impacts on citizen satisfaction; and that this should be supplemented by strengthened administrativemonitoring, with such performance metrics made public. 7.69 For those SDUs that do operate on a Decree 10 basis, the PER-FA recommends that Government should increasingly seek to tie State Budget resources to the "purchase" 114 Vietnam Public Expenditure Review and Integrated Fiduciaiy Assessment of specific public goods and specific services for the poor. This in a sense i s what Decision 139 on Health Care Funds for the Poor attempts to do (see Box 7.6).However, for this to be effective, mechanisms for monitoring and enforcing delivery against such contracts will need to be developed. 7.70 The proper implementation of Decree 10 requires reaching a new consensus on the role of government in public service provision. The degree to which the costs of public services should be "socialized" and how closely government should monitor and supervise their provision depend on the nature of the services-can market forces improve the efficiency and equity of outcomes? In introducing market-like incentives to revenue-producing units, Government must be clear about where to draw the boundary between the public and private spheres and why. Public resources could then be focused on the provision of core public services while non-core services could be gradually socializedor commercialized. Box 7.6: Decision 139/2002on Health Care Fundsfor the Poor Decision 13912002lQD-TTg was issued in 2002 with the goal of securing access by the poor to health care, in particular for in-patient treatment. I t created province-level Health Care Funds for the Poor (HCFPs) to finance the out-of-pocket costs of eligible individuals who use publicly provided health services. With province-level governments expected to pay a major role, Decision 139 directs HCFPs to either buy health insurance cards for the poor, or to pay directly for the out-of-pocket costs of services supplied to beneficiaries by hospitals and possibly other public providers. Decision 139 i s designed to bring benefits to the poor on two fronts: increased utilization of health services, and reductions in the extent of "income" poverty. Decision 139 could also increase efficiency, especially in those provinces in which responsibilities will be increasingly divided between financing and service delivery. This split in key functions provides opportunities for downward pressures on costs and upward pressureson service quality. (For further details, see Chapter 12.) 8. PUBLIC INVESTMENTMANAGEMENT INTRODUCTION AND OVERVIEW 8.1 Public investment constitutes an important and growing part of Vietnam's total public expenditure. As Chapter 2 noted, between 1998 and 2003, State Budget capital expenditure grew at the remarkable average annual rate of 20 percent in nominal terms, rising to average one third of all public expenditure over the last three years. State Budget capital expenditure has grown as a share of GDP, from 5.7 percent of GDP in 1998 to 8.4 percent of GDP in 2003. Furthermore, only part of public investment i s included in the State Budget. Indeed, if the Government's full Public Investment Program (PIP) were implemented, it would represent a claim of almost one fifth of Vietnam's GDP each year. 8.2 There are two important implications. First, improving the efficiency and effectiveness of such a large volume of investment would necessarily have a substantial impact on poverty reduction and growth. Second, to the extent that achievement of Government's poverty reduction and growth objectives requires shifts in the balance of investment (e.g. between sectors, regions or levels of government), such shifts should be made now: re-balancing of investment should be much easier to accomplish during a period of growing investment, since adjustments are less painful when they do not require absolute cuts in spending. 8.3 This chapter reviews public investment management in Vietnam. The focus of the chapter i s on the quality of public investment planning in general: the PIP document i s only the most visible output of national investment planning activities, although it reveals a lot about the wider system to which it contributes. The first section of this chapter describes the context for investment planning in Vietnam and reviews the current PIP and its implementation in some detail. The second section deepens the analysis of key issues, including the PIP and overall resource management, managing a more decentralized PIP, sustainability, appraisal and allocation criteria, program management and institutional implications. The third section summarizes recommendations. REVIEWOF INVESTMENT PLANNINGIN VIETNAM Trends inpublicinvestment 8.4 As Vietnam continues reforms in the direction of a "market economy with socialist orientation" it becomes increasingly important to distinguish between different categories of investment. Total investment in the economy includes both private sector and public sector investment. Conceptually, public sector investment includes investment by State Owned Enterprises (SOEs) as well as the capital expenditures of central and local governments that are funded by the State Budget and by state credit. Further complications are that some government capital expenditures do not appear in the State 115 116 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment Budget, and documents also do not always clearly distinguish between investment financed by government (which may be carried out by non-government bodies) and investment undertaken by government bodies (which may have non-government as well as government sources of finance). 8.5 The state sector share of investment (56 percent in 2003) has actually increased since the mid 1990s, offsetting a decline in the share of foreign investment (though in absolute terms foreign investment has also been increasing since it dipped in 1998 and 1999 during the Asian crisis). Table 8.1 summarizes the sectoral composition of investment over the four years 2000-2003. State investment i s relatively more important in the main infrastructure sectors (transport, electricity, gas and water) as well as public services (e.g. health and education), while non-state investment predominates in the other service sectors. Over 40 percent of state investment was for transport, storage and communication plus electricity, gas and water supply, with a further 25 percent for manufacturing (16 percent) and agriculture/forestry (9 percent). Only 5 percent of state investment was on education, and 2.4 percent on health. Table 8.1: Composition of Total Investment 2000-2003 (billion Dong, constant 1994 prices) ~~ State Sector Total Inv. Sector State Inv. Share of Share of Share of State Inv. Total Total (%I Inv.(%) Inv. (%) Total 303,555.4 100 536,985.60 100 56.5 Transport storage and communication 71,222.0 23.5 86,052.60 16.0 82.8 Electricity, gas and water supply 55,048.4 18.1 58,382.60 10.9 94.3 Manufacturing 49,177.1 16.2 121,746.10 22.7 40.4 Agriculture and forestry 26,726.4 8.8 46,173.20 8.6 57.9 Community social and personal service act. 17,288.4 5.7 78,396.10 14.6 22.1 Education and training 15,304.2 5.o 18,474.60 3.4 82.8 Construction 13,618.2 4.5 25,860.30 4.8 52.1 Wholesale & retail trade, repair of motor vehicle, motor cycles 10.888.2 3.6 26,945.80 5.O 40.4 Public administration and defense: compulsory social security 10,608.2 3.5 11,378.20 2.1 93.2 Health and social work 7,253.8 2.4 8,843.40 1.6 82.0 Recreational, cultural and sporting act. 6,496.6 2.1 8,495.80 1.6 76.5 Mining and quarrying 5,109.5 1.7 6,761.20 1.3 75.6 Scientific activities and technology 3,505.2 1.2 3,994.70 0.7 87.7 Fishery 3,501.0 1.2 9,290.30 1.7 37.7 Hotels and restaurants 2,490.7 0.8 11,635.50 2.2 21.4 Real State, renting business activities 2.446.2 0.8 8,483.80 1.6 28.8 Activities of party and membership organizations 1,320.3 0.4 1,480.70 0.3 89.2 Financialintermediation 1,214.8 0.4 4,218.40 0.8 28.8 Activities of international organizations 335.7 0.1 338.60 0.1 99.1 Private households with employed persons 0.4 0.0 33.70 0.0 1.2 Source: Derivedfrom General Statistics Ofice, Statistical Yearbook2003, Tables 156 and 162 8.6 Figure 8.1 shows the breakdown of state investment between the State Budget, public lending, equity of SOEs and other sources of finance. With the economy and public expenditures growing so rapidly, there have been absolute increases in investment Public Investment Management in Vietnam 117 from all these sources, but the most striking trend in the composition of state investment i s the increasing share of loans, even if this trend has declined recently. Figure 8.1: Composition of State Investment 1995-2003 90,000 0 80,000.0 70,000.0 0 others 60,000.0 50,000.0 9equty of state owned enterprises 40,000 0 loan 30,000.0 20,000.0 0 state bLdgeted 10,000.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: General Statistics Ofjice, Statistical Yearbook2003, Table 160. 1994constantprices (2003 data preliminary) 8.7 Figure 8.2 shows that there has recently been a significant increase in the proportion of state investment managed at local level, which i s now approximately half of all state investment. Figure 8.2: State Investment 1995-2003 under Central and Local Management I 100,000.0 90,000.0 80,000.0 70,000.0 60,000.0 Q Total 50,000.0 local 40,000.0 0central 30,000.0 20,000.0 10,000.0 1995 1996 1997 1998 1999 2000 2001 2002 2003 ( P 4 I Source: From Statistical yearbook 2003, Table 158. Current Frameworkfor Investment Key Laws and Regulations 8.8 The main laws and regulations that directly and indirectly govern the process of investment planning and management are: (a) The StateBudget Law 118 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment The State Budget Law of 1996 was the first to provide a comprehensive codification of the budgeting system, but important further reforms are embodied in the revised State Budget Law of 2002 and the amplifying decree of 2003.44 Important features of the revised Law are highlighted in Chapter 4 of the PER- F A . (b) The Regulation on Investment and Construction Management45 Decree 52/1999/ND-CP (as subsequently amended by Decree 12/2000/ND-CP and Decree 7/2003) i s the principal guide to the detailed allocation of responsibilities for planning and management of investment in general in Vietnam, and of public investment in particular. It includes the classification of investments into Categories A, B and C, depending on the size, sector and significance of the project and assignment of responsibilities for authorizing and managing investments in categories A, B and C. Category A essentially comprises the largest projects and other projects of national importance, with categories B and C comprising projects that are progressively smaller andor of more localized significance. In addition (Article 17), this Regulation describes the procedure for investment planning in which the plans of individual agencies, local governments, and enterprises are nested into the synthesized investment plans at higher levels; it sets out the system of pre-feasibility and feasibility studies to be undertaken, prescribes the contents of investment plans and stipulates regular reporting schedules (quarterly, but monthly in the case of national Group A projects). (c) The Circular on Investment Supervision and E v a l ~ a t i o n : ~ ~ This i s a detailed circular from MPI which spells out requirements under the applicable legislation. It i s the nearest thing to a manual on evaluation and supervision of investment projects ('evaluation' covers ex ante appraisal and selection of projects) explaining functions and procedures for monitoring and evaluation for MPI itself, for branch ministries, People's Committees, and investment owners, including SOEs. It commits MPI to produce six-monthly comprehensive reports, plus quarterly reports on nation-wide Group A projects, and stipulates that "projects shall not be budgeted for next year if they do not have adequate reports on supervision and evaluation of the previous year. The project's final accounts shall not be approved if it fails to comply fully with reporting and evaluation requirements." (d) Decision 192/2001/QD-TTg and Decree 10/2002/ND-CP 44 Law No. 01/2002/QH11,Law on State Budget, 16 December 2002,, and Decree No. 60/2003/ND-CP, Detailing And Guiding TheImplementation Of The State Budget Law, 6 June 2003. 45 Regulation on Investment and Construction Management, Decree No. 52/1999/ND-CP, 8 July 1999; as amendedby Decree No. 12/2000/ND-CP,5 May 2000 andDecree No. 07/2003/ND-CP,30 January2003. 46 MPI Circular NO.03/2003/TT-BKH, Guiding Investment Supervision and Evaluation, 19 May 2003; MPI Circular NO.04/2003/TT-BKH, Guiding the Verification and Evaluation of Investment Projects; Amending and Supplementing Several Points on Project Evaluation Dossiers, Investment Reports and Total InvestmentAmounts,17June 2003. Public Investment Management in Vietnam 119 These provisions, regarding delegation to spending units, are dealt with in detail in Chapter 7 of the PER-IFA. These developments have important implications for investment management because Decree 10 which allows SDUs to apply user- charges and to borrow for investment, makes them less dependent on the budget. Decision 192 affects the recurrent budgets of administrative units, but may have implications for spending on operation and maintenance (as indeed does the revised State Budget law, which extends recurrent budget flexibility more widely). (e) Development Znvestment Credit47 The Development Assistance Fund (DAF) i s the main channel of investment credit from the state. Decree 50 guides the operations of DAF and Decree 106 regulates the uses of development investment credits. As described in Chapter 2 of the PER-IFA, these regulations represent a major step towards a sound regulatory framework for policy lending. It significantly narrows the list of borrowers eligible for DAF lending. The Decree states that only projects that are capable of direct repayment, have socio-economic efficiency, and have feasible business or production plans, are eligible for DAF support4*. An element of risk- sharing, through co-payments by other government bodies or by commercial banks, i s also built into the Decree. Total support by DAF, all mechanisms combined, cannot exceed 85 percent of the total investment capital of a project. MPI, in collaboration with MOF, i s responsible for preparation and submission to the Prime Minister of an annual plan for state credit and interest subsidy. DAF i s required to submit monthly reports to the Prime Minister, copied to MPI, M O F and the General Department of Statistics (GDS). Funding of Provincial Investment 8.9 The State Budget Law establishes in general a formula-based system of transfers to the local level (see Chapter 6 of the PER-IFA). This includes equalization transfers that are adjusted for different regional circumstance, with overall calculations based on expenditure norms. The norms are a basis for the calculation, but are not binding on provinces' budgets, and capital expenditure requirements are allowed for simply by grossing up the transfer by a capital investment factor, which i s determined province by province and which currently averages 28-30 percent across provinces. In principle, then, provinces have a great deal of discretion over the grants that they receive. Their ability to plan ahead i s enhanced by a three-year stability period, during which the transfer i s unchanged and the province i s allowed to keep above-target revenues that it raises. However, the mechanism by which a province's capital grant i s determined i s rather less transparent than the mechanismby which the general purpose transfers are determined. 4'Decree No. 106/2004/ND-CP, On Development Investment Credit of the State, 1 April 2004. 48 2004 Budget Instructions (Resolution No. 01/2004/CP). These state: "The amendments shall focus on limiting the concessional loans only to essential projects and goods of the national economy and important projects for the socialization process of the education and training, and health sectors. The time frame of projects to be entitled to concessional loans shall be clearly and specifically defined and the favorable interest rates shall also be reduced." 120 VietnamPublic ExpenditureReview and Integrated Fiduciary Assessment 8.10 In addition, there are various targeted transfers (in essence, conditional grants), mostly for investment, under National Target Programs (NTPs). Often the recipient Province has significant discretion on exactly where NTP funds are deployed, within the overall guiding framework of each NTP. Recent reforms which have increased the responsibilities of Provincial authorities have undermined the traditional role of sector ministries: they no longer have such direct influence over sector expenditures at provincial level, and their ability to provide sector-wide analysis and advice on provincial level expenditures i s hampered by lack of information on budgets and expenditures-this information now accrues mainly to the finance bodies and i s not well shared with sector bodies. Institutional Allocation of Responsibilities 8.11 As regards the broaddemarcation of responsibilities for planning and finance: e The Ministry of Planning and Investment has traditionally taken the lead in economic planning. Under the State Budget Law of 2002, MPI retains the responsibility to: submit to the Government the draft plan on socio-economic development of the whole country, and the major balances of the national economy, which will serve as a basis for the elaboration of the financial and budgetary plans. At the same time, it makes M O F the lead Ministry for budget planning, budget estimates and debt management. In addition, as noted in Chapter 4 of the PER-FA, Decree 60/2003/ND-CP, outlines a formal division of responsibilities between planning and finance functions at central and local levels. However, the scope of Decree 60 i s basically to divide responsibilities between planning and finance, rather than establish any kind of mechanism for dialogue and joint working. Clearly, effective planning and management of Vietnam's investment and other public resources i s highly dependent on close collaboration and good working relationships between MPI and MOF, and their counterparts at lower levels of administration. (In many provinces, such collaboration has been facilitated by the creation of combined Departments of Planning and Finance.) e The regulation on Investment and Construction Management provides a more detailed description of the allocation of responsibilities related to investment, described as follows:49 9 MPI": investment policy, the balance between foreign and local investment (and issuance licenses under the Law on Foreign Investment); appraisal of certain projects for Prime Minister decision; preparation of annual and five year plans on development investments; coordinate with M O F on supervising implementation of investment plans funded by State-managed capital sources; prime responsibility in guiding and inspecting the implementation of the biddingregulation. > MOF: policies and regimes for mobilization of investment capital from various sources; coordinate with MPI in planning allocation of investment 49This is a selective summary from the much more detailed original document. 50 Decree 61/2003/ND-CP dated June 06, 2003 stipulating functions, tasks, powers and organization structure of MPI. Public Investment Management in Vietnam 121 capital to ministries, localities and important projects; unified financial management of loan capital and aid for development investment; financial inspection of state-funded projects. P State Bank of Vietnam (SBV): policy on banks and credit supervision; guarantees for foreign capital borrowings by credit institutions. The first PublicInvestment Program(1996-2000) 8.12 The PIP document i s a relatively recent addition to the Government planning system, The PIP i s a centerpiece of investment planning, and i s highly visible to Vietnam's external partners. However, as we shall show, the PIP i s less rigorous and decisive than it may appear. It i s an important document, and reveals much about Vietnam's investment planning system and policies, but it cannot necessarily be taken at face value. 8.13 Vietnam's first PIP document covered the period 1996-2000. It was prepared by MPIwith technical assistance provided through UNDP, and was published inJune 1996. Publication of the PIP was a response to a request from the Vietnamese leadership and from donors, both seeking an overview of investment proposals linked to a statement of the government's investment policies. According to the PER 2000 assessment, this first PIP represented a step forward in investment planning as "a preliminary effort to introduce more systematic appraisal and programming techniques into the investment decision-making process." It provided a broad overview of the government's medium- term intentions regarding public investment. The presentation of the PIP provided a basis for an international dialogue on important aspects of Vietnam's development strategy. It began to develop a logical approach to public sector investment programming by setting overall investment targets in relation to likely resource availability; b y making a broad estimate of capital requirements to achieve growth targets; and, at least in principle, by relating investment projects to sectoral strategies. 8.14 The PIP 1996-2000 also introduced an element of decentralized decision-making, as it coincided with the introduction of the ABC classification of projects, in which allocations to smaller (B and C) projects were left to line ministries, other agencies of government, and provincial authorities, based on guidelines issued by the MPI. However, the PIP 2001-2005 review of PIP 1996-2000 does not provide a meaningful analysis of the central/local distribution of responsibilities for investment. The figures presented simply assume a 60140 central/local division throughout. 8.15 PIP 1996-2000 was substantially revised mid-term, particularly to adjust to the Asian regional economic crisis, but also to address some problems identified during implementation. Implementation was tightened to concentrate funding on those large projects which could be completed within the five years. As a result, the number of projects was significantly reduced. Authority to approve and allocate funding to B and C group projects had been delegated to Ministers and Chairmen of People's Committees. While this empowered ministries and provinces to adjust their programs to their needs and priorities, it had also led to a tendency for them to spread resources over too many projects, partly reflecting pressures from districts and communes to have their "fair share". To remedy this problem, 70 percent of C group projects were required to be 122 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment completed within one year and all within two years. In addition, there was an increased emphasis on domestic sources of investment funds. 8.16 In practice, it is not easy to compare investment outtums against the detailed targets of PIP 1996-2000. PIP 2001-2005 begins with a review of investment during the first PIP period, but only one of its tables explicitly compares actual with planned investment. It indicates above-plan domestic capital offsetting below-plan foreign capital, with total investment reaching 98 percent of the planned figure for the five-year period. SOE investment was 17 percent above the planned figure and state credit 25 percent below. Less than two thirds of anticipated ODA was utilized. The second Public Investment Program (2001-2005) Contents and approach 8.17 PIP 2001-2005 was prepared in parallel with the CPRGS and not finalized until late 2002. It was approved by the Prime Minister (not the National Assembly, which has to approve the State Budget), by Decision No. 6579/VCP-QHQT, on 5 December 2002, nearly two years into the period it covers. As noted earlier, its preparation and implementation have coincided with important reforms in State Budget practice and significant advances in decentralization. 8.18 Compared with the first PIP, there are some changes in policy concems and emphases (notably on poverty, operation and maintenance, and on the financing of SOE investment) but the same basic approach i s followed. The methodology used, set out in the PIP document, emphasizes the prioritization of project lists against priorities set b y national, sectoral and provincial plans. However, responsibility for selecting most projects (even those funded by the State Budget) i s decentralized. The sector investment targets set in the plan are therefore a guideline for PIP implementation, including project selection, during the period, but do not actually represent a bottom-up compilation of investment projects. Thus, although the methodological emphasis i s on compiling and prioritizing lists of projects, Vietnam's PIP is, in practice, a set of investment targets, not a compendium of projects. Furthermore, the PIP-like the Socio-Economic Development Plan-should be recognized a directional document, particularly as far as enterprise investment i s concerned. 8.19 The PIP does include a substantial "List of Key Projects" but the status of this list (summarized in Table 8.2) and its consistency with the main text of the PIP i s rather problematic. The PIP does not itself provide any summary of project numbers and costing (although these have been compiled as part of working papers for this PER-FA). There are about 1,159 projects in total: 223 A projects with an average required investment during 2001-2005 of VND838 billion; 569 B projects with an average value of VND 109 billion; a further 320 projects in the industry and tourist infrastructure sectors are not broken down into AB classification, and have an average value of VND861 billon. The average size of A projects i s largest in the transport sector (VND1,575 billion) and smallest in the science sector (VND48 billion). m w r - m N - m m - 4 2 L W Ru. Y VI V a, e .C a I Y Q cfl .-6I! a, a 3v1 2 -i 3 - 0. 3 W N m .C 5wlE' L ep .-*uE' v1 Q) L c 4b x G 3h 4 5 .-6 a .-8 d ' 124 VietnamPublic ExpenditureReview and IntegratedFiduciary Assessment 8.20 Industry projects account for more than half of the 2001-2005 costs of the Key Projects, but it i s not possible to tell how much of the funding of industry projects i s expected to fall within the PIP definition of "public investment capital". "Total development investment capital" for industry and construction i s expected to be VND369.6 trillion, with VND197.5 trillion included in "public investment capital"; the total "investment capital" for key industry projects amounts to VND268.8 trillion. 8.21 The Key Projects list i s also not very informative about the financing of projects, including the expected requirements for ODA. There are columns to indicate whether State Budget, investment credit or SOEs' own capital features as a financing source, but no amounts are given. 8.22 The Key Projects list indicates a major effort to relate the PIP investment targets to specific major investments, but i t has some weaknesses, both as an analytical device and as a tool for managing the PIP. There are no project descriptions beyond the titles; projects are not given any unique identification number through which they could be tracked or linkedwith budget allocations; donor involvement i s occasionally mentioned, but not systematically covered; in a number of cases a project title appears with no costs given for the 2001-2005 period; and the breakdown between State Budget, investment credit and own capital i s not quantified (just an x or a blank in the respective columns). Annual breakdowns are not given for individual projects, nor by sector, nor for the total list, and there i s no prioritization among projects. Sometimes projects included in the plan are divided into more than one project at the implementation stage. This makes it difficult to monitor implementation on an annual basis. 8.23 The PIP does not limit itself to public investment that forms part of the State Budget. Rather it i s described, in its first sentence, as: the core of all investment activities in the economy. It plays an important role in guiding and attracting various sources of funding to achieve socio-economic development targets of the country during the planned period. Although it does distinguish between total investment and public investment, it i s not very clear what definition i s used, and figures are inconsistent between different sections of the PIP document. The PIP i s clearly not in any sense a budget: the allocations it shows are to 'sectors' not institutional budget holders, and, even for major projects, inclusion in the PIP i s not a necessary condition to receive funding. The PIP i s better understood as an exercise in indicative planning: i t shows targets and intentions for the level and balance of investment expenditures to be achieved, but the mechanisms for realizing these targets and intentions are not very clear. This reflects the fact that responsibilities for making individual investment decisions are increasingly diffused, notably between government and SOEs and between different tiers of government. 8.24 Understood in this light, the PIP i s nevertheless a significant document in spelling out the Government's approach to investment and its broad priorities. It shows which sectors and regions the Government proposes to focus investment on, and comparison of the PIP 2001-2005 with its predecessor reveals some significant new concerns and changes in emphasis. I t proposes a broad balance between sectors that reflects the general strategies set out in the SEDP and in various sector planning documents, but the treatment of each sector i s at a very broad level, without much detail or prioritization among sub-sectors. (For education and health, in particular, the PIP provides no breakdownbetween sub sectors to indicate relative priorities among them.) Public Investment Management in Vietnam 125 8.25 It does, however, indicate which broad sectors should be most dependent on public investment (as opposed to investment from other sources): this i s illustrated in Table 8.3 which ranks sectors from most to least dependent on public investment (see middle column). Because of the high level of aggregation in the PIP, and because there i s not a more detailed breakdown of financing sources by sector, it i s not clear precisely how these public investment shares come about. Table 8.3: PIP2001-2005 Sector Breakdownof PlannedInvestment (VND trillion,at 2000 prices) Total PublicInv. PublicInv. Share of Share of Dev. Inv. Capital Capital(% Total Public Capital of total) Inv. % Inv. % Science and technology 7.6 7.2 94.7 0.9 1.4 Health care & social activities 29.3 26.4 90.1 3.4 5.O Transport & Telecommunication 142.0 111.0 78.2 16.4 21.2 Agriculture, forestry, fisheries, irrigation 133.8 97.6 72.9 15.5 18.6 Education & training 45.O 30.0 66.7 5.2 5.7 Culture, information & sports 15.7 9.2 58.6 1.8 1.8 Industry & construction 369.6 197.5 53.4 42.7 37.7 Urban infrastructure & services 121.6 44.5 36.6 14.1 8.5 Total 864.6 523.4 60.5 100.0 100.0 Source: Derivedfrom PIP 2001-2005 (chapter by chapterfigures) 8.26 Given the increasing emphasis on decentralization, it i s also surprising that PIP 2001-2005 actually has very little quantitative analysis of the breakdown of public investment between central and local responsibilities. As already noted, the reported figures for this breakdown in the implementation of PIP 1996-2000 are actually an assumed 60:40 ratio, and the corresponding figures for PIP 2001-2005 are not provided. As regards regional priorities, the PIP does provide a breakdown of total investment by region (Table 8.4). As would be expected, the State Budget i s planned to be a higher proportion of total investment in the less developed regions. However, this i s once again a highly aggregated summary which does not explain how the figures have been arrived at; nor does it explain whether there i s any management mechanism to turn these breakdowns into targets that can be pro-actively pursued. Analysis by province i s not available. Table 8.4: PublicInvestment Capital2001-2005 by Regions VND Trillion at 2000 Prices Total O.W. State % Budget Northern Mountains 48.4 27.7 57.2 Northern Central Region 47.0 26.0 55.3 Western South (Mekong Delta) 81.0 37.1 45.8 Central Coastal Region 57.7 24.1 41.8 Central Highlands 25.4 10.0 39.4 RedRiver Delta 115.1 43.5 37.8 Eastern South Region 123.5 43.3 35.1 Total 498.1 211.7 42.5 Source: PIP 2001-2005, p 33 (excludes O&M investment) 126 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 8.27 As illustrated in Table 8.5, the most significant positive change envisaged in the financing of public investment i s a reduced on the State Budget of SOEs, with more of their investment needs financed from own-resources. The share for state credits i s projected to stay roughly constant, with an increasing emphasis on interest subsidy rather than direct credit. "Other" sources of financing for public investment, which account for up to 10 percent of projected annual totals, presumably include the recent bond issues for the education and transport sectors. All of this i s against a backdrop of continued rapid growth in annual public investment levels. Although availability of ODA i s included among the factors taken into account in setting investment targets, the PIP i s remarkably uninformative about details of anticipated sources and uses of ODA funds. Table 8.5: PIP2001-2005 PlannedFinancingof PublicInvestment 2001 2002 2003 2004 2005 Total 2001- 2005 VND trillion at 2000 prices 41.5 42.5 43.5 44.0 44.1 215.6 State budget 20.6 22.7 24.0 25.4 26.7 119.4 State credit 24.4 28.9 33.0 36.2 40.0 162.5 Investment of SOEs 9.4 10.5 10.0 10.0 10.5 50.4 Other 95.9 104.6 110.5 115.6 21.2 547.8 Total Shares (%) Statebudget 43.3 40.6 39.4 38.1 36.3 39.3 State credit 21.5 21.7 21.7 22.0 22.0 21.8 Investment of SOEs 25.4 27.6 29.9 31.3 33.0 29.7 Other 9.8 10.0 9.0 8.7 8.6 9.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 Source: MPI 8.28 Chapter IV of PIP 2001-2005 mentions operation and maintenance of public investment. I t envisages substantial increases in O&M expenditure in all sectors compared with the previous five-year period. However, the definition of O&M appears quite narrow, and it i s not entirely clear where in the Government budget the expenditures cited are to be found. O&M issues are discussed further in the Key Issues section of this chapter, under sustainability. 8.29 Poverty reduction i s also explicitly addressed in a separate chapter of the PIP. The chapter i s mainly in narrative form. It highlights the investments in main sectors that are seen as especially relevant to poverty reduction, and it also provides a table of year by year "requirements" for the various National Target Programs but it i s not very clear how these requirements are expected to be financed. Although the publishededition of the PIP makes repeated references to the CPRGS, one criticism that has been made of the PIP i s that i t fails to reflect the emphasis and prioritization impliedby the CPRGS.51However, the figures for sector allocation of investment requirements that appear in the final versions of the CPRGS and the PIP are very similar. The total public investment projected in PIP 2001-2005 i s very close to the "low" scenario of the CPRGS. Thus, 5' See, for example, Politics and the PRSP Approach: Vietnam case study, Tim Conway, Overseas Development Institute Working Paper 241, May 2004. Public InvestmentManagement in Vietnam 127 rather than arguing a disconnect between the CPRGS and the PIP, there may be a stronger case to argue that (a) neither document adequately follows through the implications of policy statements in terms of expenditure requirements, and (b) neither i s very operational-in terms of a clear relationship between targets and the budgeting process. The high level of aggregation in the PIP'S investment targets (the absence of useful breakdowns either by sub sector, or by tier of responsible government) and its pervasive lack of prioritization, also make it difficult to assess its likely contribution to pro-poor investments. Monitoring and implementationof PIP 2001-2005 8.30 Like its predecessor, the 2001-2005 PIP has some strong features. It represents the most detailed attempt to date to identify the medium-term financial implications of Vietnam's development strategy. Nevertheless, caution i s needed in reviewing implementation of the PIP. Nearly two years of the period had elapsed by the time the PIP was approved, and, as noted above, it does not provide very clear benchmarks. According to regulations on investment supervision, implementing agencies are supposed to provide at least quarterly reports on implementation. The limited availability of data in practice indicates that there i s room for improvement both in the provision of information and inMPI's ability to keep up with compilation and analysis. 8.31 Table 8.6 provides MPI's recent estimate of PIP implementation during the first three years, and projections to the end of the PIP period. The table suggests that total levels of investment are in line with PIP projections; total public investment i s projected to exceed the five-year target (by 3.7 percent), with total nationwide development investment even more above target (14.4 percent). So far there has been a significantly greater than planned reliance on state credit financing. Table 8.6: Three-YearStatus of PIP2001-2005Implementation ~ 2001- Three-year Five-year Percent Estimated 2005 Implemen- Estima- after 3 percent Plan tation tion years after 5 years VND Trillion Total nationwide development inv. 840.0 522.9 960.9 62.3 114.4 of which PIP excluding maintenance exp. 498.2 288.9 516.6 58.0 03.7 Financed by: -Investment from state budget 216.5 117.8 214.0 54.4 98.8 -Investment from state credit 119.4 82.3 134.9 68.9 13.0 -Investment from SOEs 162.3 88.8 167.7 54.7 03.3 Percentages PIP as share of total dev. Inv. 59.3 55.2 53.8 Financing of PIP -Investment from state budget 43.5 40.8 41.4 -Investment from state credit 24.0 28.5 26.1 -Investment from SOEs 32.6 30.7 32.5 Source: MPI; D r Thai's report, July 2004 (Annex A19) 128 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 8.32 A continuing concern has been that rates of project implementation are slow, reflecting capacity constraints, administrative difficulties (associated with ODA projects in particular), andor a tendency for decentralized bodies to spread investment funds too thinly. A special review of major projects in the transport sector was therefore undertaken, to see whether allocations for 2004 were enough either to complete the project or to cover 20 percent of its costs-consistent with completion within 5 years. Although the analysis i s crude, it does indicate that only about half the projects in the transport sector were on course for completion within five years. 8.33 The PER-FA team also conducted a review of information from the most recent monitoring reports compiled by MPI. The data are incomplete, but they illustrate MPI's approach to inspection. The focus i s on procedural aspects: numbers of projects started, completed and evaluated, the number which have "violated investment procedures" in various ways, and the number which have been adjusted in various ways. The data presented are not correlated with financial information on the projects concerned, making it hard to judge the significance of the deviations discovered. Nevertheless, there is concern that the commencement rate of projects by provinces substantially exceeds the completion rate, again implying that resources are being spread too thin. About 5 percent of all projects (rising to 7 percent of provincial projects) are identified as "violating investment procedures". In the overwhelming majority of cases, however, this means late approval (0.4 percent of projects) and late implementation progress (4.6 percent of projects). Only 50 provincial projects out of more than 5,000 reviewed were described as being "wasted" or of poor construction quality. 8.34 The data also shed some light on the question of how many public investment projects there are altogether. The 16 provinces reporting had over 5,000 projects comprising approximately 60 Group A projects, 1,000 Group B projects and 4,000 Group C projects. Overall, approximately 7,000 C projects, 2,000 B projects and 340 A projects are included in the detailed monitoring reports. Since not only the majority of provinces but several major ministries (including MOET and the Ministries of Industry and Construction) are omitted from the data, the true number of projects i s far higher. KEY ISSUES 8.35 The analysis above has demonstrated that, in practice, the PIP i s only one element inthe resource allocation system inVietnam. Although often described as a compendium of projects, it i s not a comprehensive screening and approval process even for large projects, and its main role i s as a summary of Government's investment policies and priorities. Provinces are performing an increasingly important role in public expenditures: most provincial funds are provided in block grant form, and sector ministries, as well as MPI and MOF, now exert less direct influence on their budgets. This year's Government Resolution 8 on accelerated decentralization (discussed below) i s likely to accentuate this trend, and i s a further reason for fundamental review of how Government seeks to influence and manage investment, and of the roles played by MPI and the PIP document in this process. This part of the chapter is therefore concerned to suggest broad strategic directions for the future planning and management of investment in Vietnam. Key issues are discussed under a number of thematic headings. Public InvestmentManagement in Vietnam 129 The PIP and Overall Resource Management 8.36 The PIP i s not intended to be an end in itself. It needs to be assessed in terms of its contribution to an overall system and style of planning and budgeting that serves Vietnam's objectives. It i s important to recognize the strengths as well as the weaknesses of public expenditure management in Vietnam. There have been high levels of investment, rapid growth and expansion of public services. There have also been impressive reforms and adaptations as systems of economic management and public administration have developed. Nevertheless, the Government itself recognizes continuing weaknesses and issues that needto be addressed(see Chapter 4). 8.37 Internationally, stand-alone PIP documents are increasingly being superseded by integrated approaches (variations on the principles of Medium Term Expenditure Frameworks) because of the disadvantages of focusing too narrowly on capital expenditure and investment aid requirements. I t should not be assumed that the PIP in its present form will be an essential part of the Government planning system indefinitely. I t s present main function of providing a comprehensive overview of public investment policy i s important, but does not necessarily require a separate document. As discussed below, there i s a need to develop more rigorous techniques and systems for the allocation of state investment resources, to disseminate better investment selection and management techniques to all levels of Government and also to be more rigorous in capturing and monitoring the costs of projects. 8.38 Both the PIP and the CPRGS were a response to requests for more accessible and comprehensive statements of Government policies and priorities that could serve as a basis for dialogue as well as a guide to the allocation of external resources. Both have drawn directly on the pre-existing national planning system, are clearly owned by Government, and have helped to clarify and enhance Government approaches to planning and resource management. Neither i s yet sufficiently operational, in terms of a clear and monitorable link to the budgetary resource allocation process. Managing a More Decentralized PIP 8.39 Decentralization i s taking place in multiple dimensions: to line ministries, to provinces, and to spending units. Both the last two are very profound: the amount of discretion to provinces breaks the lines of control by sector ministries; and spending units are gaining a huge degree of autonomy via Decree 10 (and to a lesser extent Decision 192). This requires a fundamental shift in the system of management (including investment management) from command and control to management through incentive structures. 8.40 Decentralization has been approached in a pragmatic way, taking place in stages that draw on previous experience, and using practical devices such as the ABC categorization of investment responsibilities and the block granting innovations. There i s still a tendency however, to rely on traditional approaches to top-down management and rule-based supervision that are much less appropriate or effective in a decentralized setting. There needs to be more progress towards establishing decentralized processes and incentive frameworks that can lead to efficient and effective investment choices without requiring detailed top-down interventions. 130 VietnamPublic Expenditure Review and Integrated Fiduciary Assessment 8.41 Inthe meantime, there are signs that the decentralization process has outpaced the development of appropriate management frameworks. Symptoms include: (a) indications that block granting may be leading to a neglect of routine maintenance; (b) continuing concerns that provincial decentralization i s leading to an excessive fragmentation of investments, with too many started at once, leading to delayed implementation, over commitment of budgets and build up of arrears to contractors; (c) as noted in Section Review of Investment Planning in Vietnam of this chapter, the existing supervision system i s struggling to cope, so that supervision reports are compiled from very incomplete information. 8.42 There needs to be more stress on providing decentralized bodies with a better, and more rigorous, methodology for allocating and managing investment funds within a comprehensive resource management framework. Quality and timeliness of reporting, on expenditures in particular, needs to be strengthened so as to allow the central and sectoral agencies to be more analytical in their monitoring and evaluation of investment performance. 8.43 At the same time, with the recent adoption of Government Resolution 8, the pace and depth of decentralization seems set to increase. Some of its key provisions are highlighted in Box 8.1. However, the Resolution will not take effect until its principles are translated into detailed amendments of existing decrees and regulations. MPI's role in this process provides an important opportunity for a thorough review of the planning process of which investment management forms part. Sustainability 8.44 Concerns about sustainability relate to the maintenance of completed projects (an issue which i s highlighted in the PIP) but extend to other issues which the PIP perhaps could address better. These include the wider relationship between recurrent and capital expenditures, and questions of aggregate sustainability and risk associated with public expenditure plans. The wider problems are potentially more serious: failure to provide for the O&M of one project can jeopardize the returns from that project; over-ambitious expenditure plans or neglect of large-scale risks can jeopardize the effectiveness of public expenditure as a whole. 8.45 The O&M problem appears in two ways: (a) through obvious failure to maintain completed projects-so that projects fail to function as planned, assets deteriorate and rehabilitation projects become necessary; (b) general shortages of O&M funds inbudgets. Both symptoms appear in Vietnam. Chapter 3 of the PER-FA draws attention to the declining relative share of O&M in the budget, despite increasing levels of capital expenditure. Irrigation and roads are frequently cited as sub-sectors where O&M i s problematic. Failure to ensure proper maintenance of assets i s both a financing issue (can maintenance be afforded?) and a management issue (are funds actually allocated and the necessary arrangements in place for maintenance to take place?). 8.46 It i s important to consider systemic reasons why recurrent costs may not be provided as required. Two particular issues in Vietnam are dual budgeting and the block granting system. The block granting issue i s discussed in Chapter 7 of the PER-FA, which notes that some pilot agencies reported sizeable reductions in expenditure on Public Investment Management in Vietnam 131 routine repairs and maintenance; it suggests that allowing agencies to raise salaries by cutting back on maintenance may be unwise, and, in effect, some ring-fencing of O&M expenditures should be considered. Box 8.1: Relevant Provisionsof Government ResolutionNO'S (De~entralization)~~ Main directions 0 Decentralization in management of planning and development investment a- Decentralization in management inplanning and socio-economic schemes The central Government unifies the planning management in the whole country. The central Government and the Prime Minister directly decide national planning, including: overall socio-economic planning in the whole country and in centrally-run provinces/cities, planning strategic infrastructure works that are of national importance, overall sectoral planning, overall planning towns (those are from category I1 upwards), planning the development of economic zones and high-tech zones. Line ministries, in cooperation with the Ministry of Investment & Planning develop sectoral planning and submit to the Prime Minister for approval. Ministers of line ministries decide detailed planning within the relevant lines as assignedby the Prime Minister. Based on overall national, sectoral and economic zone planning, chairmen of provincial people's committees develop local planning, including: provincial socio-economic schemes, those of sub-regions within the provinces, and those of administrative units under province; detailed sectoral planning within the provinces, planning of development of urban cities/towns (from category I11downwards) and rural areas to be submitted to provincial people's councils for approval. 5-year socio-economic development plan of a given level i s decided by the government of the same level; Provincial people's committees prepare and submit socio-economic schemes to the people's councils for approval. b- Decentralization in investment management. Based on features, scope of work, investment field, scale and investment source, decentralization i s made irrespective of whether the project in question belongs to category A, B or C. Delegation from senior to junior levels i s not applied. Provincial administrations can decide investment projects that come from local budget; and can issue licenses to foreign investment projects suitable to local socio-economic development level and management capacity. Decentralization in the management of State Budgei Defining clearly rights and responsibilities in managing and using State Budgets of every level. The Government manages the central budget which i s approved by the National Assembly. Provincial people's committees prepare, adjust, allocate and finalize local budget and submit it to provincial people's committees for approval. Based on resolutions of people's councils, provincial people's committees decide several regimes and cost norms in budget expenditure as in line with legal regulations; and decide some fees, people's contributions and their levels in accordance with local specific conditions and characteristics. 8.47 More generally, dual systems of planning and budgeting make it more of a challenge to ensure that O&M i s adequately provided for. Dual systems of planning and budgeting tend to perpetuate a bias towards new investment. The common assumption that investment should be maximized, and recurrent costs kept to a minimum, i s too simplistic and a recipe for serious inefficiency in resource management, and hence lower rates of return on projects. Instead of focusing first on investment and then checking that the recurrent cost implications have been factored in, recurrent cost projections should play a much more prominent role in expenditure planning: they should be taken as a crucial guide to how much future investment i s sustainable. 52 Resolution No. 08/2004/NQCP On Continued Acceleration of State Management Decentralization between the central Government and centrally-run provincial government, 30 June 2004. 132 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 8.48 Moreover, a broad concept of recurrent costs i s required, that includes all the operating costs, including salaries, of providing public services (the salary costs are relatively small where physical infrastructure i s concerned, but dominate services such as education and health). Because the cost structure and dynamics of different sectors are so different, figures for the expenditure composition of the budget as a whole have limited value. Sector by sector analysis of recurrent cost issues i s required. The sector chapters of the PER-FA show strikingly different trends across sectors (whereas lack of recurrent funding for maintenance i s critical in transport and agriculture, in education it i s capital funding for school constructionthat appearsto be the immediate constraint). 8.49 Risk i s also an important sustainability issue in relation to the PIP. Chapter 2 of the PER-FA draws attention to the risks associated in particular with state credit provided through the Development Assistance Fund, and with the issuance of bonds to finance off-budget investments. 8.50 Finally, it was noted earlier in this chapter that the intergovernmental transfer system channels resources from the more wealthy provinces to the poorer ones. Chapter 3 of the PER-FA demonstrates that net transfers for recurrent expenditures are skewed towards to poorest provinces, whereas resources for investment are skewed towards the richest ones. This reflects a conscious strategy to prioritize growth, and hence to make investments where the prospective returns seem highest. However, legitimate questions have been raised as to whether a system that requires perpetual transfers from poorer provinces to richer will be politically or socially sustainable, and the Government will need to consider whether to alter its investment allocation criteria so as to limit the inequality between provinces. This would need to be reflected in the appraisal and allocation criteria discussed next. Appraisal and Selection Criteria 8.51 In the context of decentralization it is important to strengthen investment appraisal and selection criteria and capacity at every level of government. Allocation issues can usefully be considered at two levels: e Criteria for choosing among individual projects. Note (a) that the same criteria will also serve as a guide to good project design; and (b) that inpractice it i s much easier to apply the same ranking criteria to projects within a sector than across sectors (because of well known difficulties of assessing social projects in the same way as infrastructure ones, and so forth). e Criteria for deciding the balance of investment between sectors or regions. 8.52 Improving the quality of the public investment portfolio i s not only a matter of applying robust appraisal (and evaluation) criteria to individual projects; it also requires procedures to ensure that proposed investments are properly related to sector policies and mediumterm sustainable recurrent expenditure plans, that a comprehensive view i s taken of competing claims for investment funds, and hence that projects are prioritized within reasonably hard budget constraints. The MTEF approaches which are being piloted are relevant to both inter-sector and intra-sector choices. In each case a hard budget constraint (provided by a firm planning guideline ceiling) i s extremely useful in stimulating prioritization within and across sectors. The process aspect of an MTEF i s Public Investment Management in Vietnam 133 equally important: it requires the claims of different sectors and projects to be considered simultaneously, in a way that should ensure consistency not only with relevant sector policies but also with realistic levels of resource availability. Prior to detailed assessment of individual projects, there needs to be clarity about the proper role of Government in financing or delivering services inthe sector concerned. 8.53 A number of criteria for project selection need to be strengthened because they are under-emphasized in the planning traditions Vietnam has inherited.These include: a Growth effects (more systematic and rigorous economic analysis techniques, which can have the added benefit of promoting harmonization between Government and donor appraisal requirements). a More explicit and systematic analysis of poverty and environment issues (at sector and policy level as well as for individual projects) (The Vietnam Development Report 2004 discusses some possible approaches). a Recurrent cost issues (maintenance and runningcosts)-see sustainability section above. 8.54 MPI has issued various instructions concerning the selection of projects, but it does not have an overall manual on project appraisal. (TA support for the preparation of such a manual a few years ago was unfruitful and the outputs were never translated into Vietnamese.) Preparation of such a manual, aimed at sectoral and provincial planners as well as MPI itself, could provide a useful focus for MPI's review of planning techniques and project appraisal criteria. It would need to take into account the need for much simpler techniques for small and local level investments than for major national projects. Any such guidelines need not only to disseminate appropriate techniques, but also to standardize costings and formats-so as to facilitate comparability across projects and to provide a standard basis for monitoring individual projects and aggregate programs. 8.55 Improvements in quality (and transparency) of appraisal and allocation are inherently necessary, but should have the added benefit of making it more possible for donors to use non-project modalities and to harmonize around government procedures. Program Management 8.56 Stronger analytical techniques need to be complemented by strengthened systems for investment program management. Key elements include: a project documentation and formats a procedural requirements a supervision and monitoring. 8.57 Visible stresses and strains in the supervision system indicate a need for strengthening the other elements too. The general trend should be towards more performance-oriented monitoring, but this i s additional to, not instead of, strong fiduciary standards, for which financial monitoring i s a vital foundation. Effective monitoring depends on establishing, from the outset of a project, proper benchmarks, and targets against which implementers can be held accountable. 134 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment 8.58 Chapter 4 of the PER-FA notes that, officials in the planning functions spend too much of their scarce time simply trying to find out how much has been spent on different projects, leaving little time for monitoring of physical execution and impact. In the medium term, the proposed Treasury and Budget Management Information System (TABMIS) could make the compilation of expenditure information much more straightforward, but its ability to do so will depend on appropriate standardized classification of expenditures; and in order for users to relate expenditure to the original plans and project authorizations, these too need to employ the same expenditure classifications. All investment-authorizing bodies should be required to keep standard project registers, in which planned, authorized and actual expenditures for projects are kept systematically up to date. This would help to ensure discipline by focusing on the complete implementation costs of projects when they are authorized, and would make it more practical to ensure that the sponsoring agency bears the consequences if cost overruns occur. 8.59 Review of investment management procedures should include a reassessment of the allocation of investment responsibilities, to ensure that responsibilities for investments are located at the appropriate level, and, where investments have implications across jurisdictions, there are procedures for coordination and collaboration amongst those affected (for example the problem of ensuring that provincial decisions concerning transport infrastructure are taken within a regional planning framework. Further attention to the problems of fragmented local investments and the build up of project arrears i s also required. 8.60 The system of investment supervision i s struggling to cope with the ramifications of much more decentralized investment, and expenditure management generally. It needs to move further towards systems for indirect supervision (Le. where MPI and M O F set criteria and standards, and provide appropriate techniques and training for decentralized managers and investment owners). The current supervision system i s too focused on process, and the patchy coverage of supervision reports i s very worrying: it suggests a serious danger that delegation of responsibility i s proceeding faster than systems of accountability can keep up with. As part of its review of inspection functions in general (discussed in Chapter 4 of the PER-FA) Government should consider whether it i s helpful to continue to combine the fiduciary responsibility of invigilating procurement systems with MPI's more strategic and analytic concerns. Inthe longer term it may make sense to group such functions under the Ministry of Finance and the audit bodies, leaving MPIwith a clearer focus on policy and strategic resource allocation issues. Infuture MPI should put less of its efforts into checking the legal compliance of subordinate bodies and more into analytic and thematic reviews (such as the PER-FA). More emphasis on thematic reviews of investment and post-evaluations of projects and programs would also be appropriate. Institutional Implications 8.61 The directions in which investment management should develop will clearly have institutional implications, both for MPI and for the relationship between MPI and MOF. MPI, like MOF, needs to consider simultaneously how it can update its structure, staffing and work programs to reflect the changing demands of managing an increasing market- Public Investment Management in Vietnam 135 oriented and decentralized economy and administrative system, and how it can ensure effective collaboration with M O F and other agencies at the interface between planning and budgeting. Irrespective of the precise division of responsibilities, investment planning needs to become more integrated with overall resource management and budgeting, and the development of Medium Term Fiscal Framework (MTFF) and MTEF approaches i s promising. These cannot succeed without the full engagement of both MPI and MOF. They will also need to collaborate closely in considering how best to rationalize the present array of planning process and documents, of which the PIP i s one, into a more integrated system. SUMMARY OF RECOMMENDATIONS 8.62 In summary, regarding public investment management, the PER-FA makes the following recommendations: The MPI should actively participate with MoF, sector ministries and provinces in the preparation of MTEFs.These could emerge as a powerful common analytical tool and vehicle for coordination between the planning cycle and the budgeting cycle. (See also Chapter 4.) The Government should build the analysis of recurrent costs into investment selection, so that recurrent costs become a key factor in developing the investment program, not just an after effect of it; and should use the MTEF approaches as a way of ensuring consistency between recurrent and capital costs. The Government should use the opportunity provided by Government Resolution 8 to undertake a strategic review of the allocation of responsibilities for public investment between tiers of Government and levels of management, and of the guidelines issued by MPI to guide investment by other levels and tiers of government. The guidelines should be strengthened to focus more on the criteria and techniques for selection between projects and balance across sectors, and on minimum standards for initial documentation and subsequent monitoring of project budgets, expenditures and outputs. MPI should continue to strengthen its capacity to guide, train, support and supervise bodies with decentralized authority for investment planning and implementation. The Government should develop and adopt an investment planning manual that deals both with techniques for individual project selection and with techniques for assess appropriate balances between sectorshegions. The Government should systematize more explicit analysis of risks linked to investment, including analysis of the contingent risks associated with state credit lending. The new TABMIS system should be implemented so as to ensure that MPI and planning departmentsbureaus are given direct access to Treasury data on execution of public investment projects at every level of government. (See also Chapter 4.) 136 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment e All investment-authorizing bodies should be required to keep standard project registers, in which planned, authorized and actual expenditures for projects are kept systematically up to date, drawing on TABMIS data as soon as feasible. 9. PUBLICPROCUREMENTMANAGEMENT INTRODUCTION AND OVERVIEW 9.1 Sound public procurement i s a central pillar of good public expenditure management in a market-based economy. A transparent, fair and efficient procurement system has a critical role to play in ensuringthat development objectives are met and that value for money i s achieved, in accordance with the budget plan. A poor procurement system results in higher costs to the Government and public, increases the scope of corruption, discourages the creation of competitive markets and delays budget execution, economic growth and poverty reduction. 9.2 In Vietnam, the importance of market-based mechanisms such as contracts in public spending i s rapidly rising. The value of procurement financed through the State Budget more than doubled between 1999 and 2003 from US$2.0 billion to US$4.9 billion, according to Department of Public Procurement (DPP) statistics. A significant portion of public spending occurs through the public procurement process and even modest efficiency improvements in the workings of the system could potentially yield substantial savings-perhaps of the order of 1-2 percent of GDP. It i s therefore no surprise that there i s increased Government and public awareness of the need for procurement reform. 9.3 In 2002, the Government of Vietnam and the World Bank conducted and published a Country Procurement Assessment Report (CPAR). The main objective of this chapter i s to review and update the implementation of the 2002 CPAR. The chapter takes stock of the main developments in the areas identified as critical by the previous CPAR and, in light of the progress achieved and remaining challenges, recommends a set of actions to be taken to improve the efficiency and effectiveness of the procurement system. 9.4 The performance of any procurement system i s determined b y the degree to which it ensures that public funds are used effectively and efficiently. The 2002 CPAR identified a large range of issues that negatively affected the performance of the Vietnamese procurement system, ranging from concerns over the legal framework to the organization and management of the procurement function. Vietnam has started to address these issues and has created much of the foundation for establishing a procurement system that meets the needs of the country. Experience suggests that the transition to a well-performing procurement system will be a lengthy process, responding to changes introduced within the procurement system as well as progress made in other areas of public financial management and market development. At present, the procurement system continues to create a significant fiduciary risk that Government and donor funds will not be put to best use. This chapter focuses attention on the sources of 137 138 Vietnam Public Expenditure Review and Intezrated Fiduciary Assessment fiduciary risk and the critical need to build upon the foundation that has been created by puttinginto practice new procedures and practices. 9.5 Just as this chapter i s part of a larger study, it i s important to place the discussion that follows in its proper context. Improving the process of procurement can lead to very significant benefits for Vietnam. At the same time, procurement i s correctly seen as one element of the public financial management system as well as one part of the contracting process. The chapters in this study examine public financial management in Vietnam, while contracting-a process that begins with a decision to contract, followed by the drafting of a contract, the contract award and its fulfillment-is largely outside the realm of this analysis. While this chapter focuses attention primarily on the procedures used to award public contracts, contract outcomes are influenced by a number of factors-the prudence of expenditure choices, the skill applied to drafting technical and legal specifications, the existence of competitive supplier markets, the reliability and timeliness of budget execution, the degree to which courts are able to interpret and enforce commercial contract provisions, as well as the mechanisms used to determine how contracts are awarded. Establishing an effective contracting process in the public sector represents a much broader challenge than getting procurement right since it i s deeply connected to Vietnam's transition to a market-based economy that operates in accordance with the rule of law. OVERVIEWOFPROCUREMENT REFORMS 9.6 Following the 2002 CPAR, the Government has made progress in improving the procurement system. However, that progress has not been as comprehensive as planned under the 2002 CPAR and has not yet resulted in improved performance of the system. The main areas of progress have been: a. strengthening the functions of DPP (in MPI) as the focal procurement policy unit and strengthening the capacity of the department to carry out its functions; b. taking important steps towards increasing transparency of public procurement by mandating the creation of a Public Procurement Bulletin; c. taking steps towards increasing standardization of procurement through mandating the development of Standard BiddingDocuments (SBDs) to guide procurement action; and d. enacting selected procurement procedures consistent with international good practice. 9.7 Important constraints, however, still remain: a. the overall legislative framework i s fragmented and therefore not conducive to fully efficient, economic, fair and transparent procurement; b. procurement practices which are not conducive to competition are still widely used (e.g. merit point system, direct contracting, limitedbidding) and there are restrictions to foreign bidders' participation; Procurement Management 139 c. a procurement complaint review system and an independent audit processes, cornerstones for a modem public procurement system, do not exist or are in an early stage of development; d. corruption and collusive practices inpublic procurement remain and continue to undermine quality and impact of public investment; e. while progress has been made towards improving market incentives, there i s still more to be done to level the playing field between the private and state sector; f. the ability of buyers and sellers to administer contracts inaccordance with the agreed terms and conditions i s not well developed; g. the local consulting industry i s relatively weak; and h. capacity within government to undertake procurement needs substantial further development. 9.8 Overall, the Government has recognized the importance of procurement and begun to create a set of rules to guide the public sector's contractual relations. Those rules reflect the government's desire to create an effective market-based procurement system. The procurement issues that this report highlights indicate the difficulty of trying to adopt market practices while at the same time creating administrative procedures that limit flexibility and the role of competition. Utilizing traditional forms of state control and intervention to respond to current capacity limitations in the public and private sector can restrict the long-term development of public procurement. 9.9 As the reform progresses, it will be essential for the Government to promote the development of competencies in the public and private sector necessary for effective procurement. Such an approach would seek to increase standardization of procedures and enhance competition and transparency while strengthening oversight and monitoring through introducing a system for handling complaints and subjecting procurement transactions to meaningful internal and external auditing. 9.10 Early steps are needed to ensure the implementation of the procurement procedures stipulated in existing legislation. This will likely encompass: i. the production of a procurement bulletin and a phase-in plan to ensure transparency for all contracts of significant value; ii.theproductionofstandardbiddingdocuments andaphase-inplantoensure their utilization in priority areas; iii.the creation of guidance on competitive procedures and the use of more effective evaluation methods and a phase-in plan to ensure their early adoption in the areas of greatest need; iv. the creation and implementation of a system for receiving and acting on complaints raisedin the procurement process; and 140 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment v. the establishment of a mechanism for monitoring procurement practices and performance that creates information on practices used in capital and recurrent transactions. 9.11 Success in this effort i s likely to require disciplined prioritization due to the decentralized nature of procurement. In the medium-term, the Government needs to establish a unified basis for all procurement actions including consultant selection and purchase of goods and works; regardless of whether procurement i s being done at central or subnational level; and in the context of both capital projects and recurrent spending. This i s a pre-condition for a modern public procurement system and essential if Vietnam's procurement system i s to be in line with international integration requirements. 9.12 Several of the issues identified in this study are not exclusively related to "procurement" or within the sole purview of MPI'sDPP's responsibilities, but rather are cross cutting issues requiring action by, and collaboration between multiple agencies. For example: strengthening the Government oversight capabilities through internal and independent audit, combating corruption, SOE reform and market development, and administrative simplification. Vietnam has a tremendous opportunity to take advantage of i t s integrated approach to improving the impact of public spending to make progress in these shared areas. 9.13 While in the short term, the Government's strategy emphasizes implementation and enforcement of existing legislation, it i s clear that establishing a high-performing procurement system requires long term efforts in drafting and enacting new legislation and ensuring adherence to the required principles and rules. LEGAL REGULATORYFRAMEWORKFORPROCUREMENT53 AND 9.14 The 2002 CPAR identified as a key shortcoming the existence of multiple normative acts with overlapping, conflicting provisions. It also identified conflicts with related legislation, such as the Economic Contract Ordinance and the Trade Law. Inorder to address the above shortcomings the CPAR recommended that Government enact a new ordinance to address the identified weaknesses and ensure consistency with the associated legislation. The eighth draft of the ordinance was submitted by MPI to the Office of the PrimeMinister on August 12, 2004, and shared with donors in August 2004. Joint consolidated comments of the World Bank, ADB, JBIC, KFW and AFD were sent to MPI in October 2004. MPI sent a ninth draft of the Ordinance to the Office of the Prime Minister on September 14, 2004. A tenth draft of the Ordinance was submitted to the Prime Minister by MPIon December 7, 2004. 9.15 In the absence of a comprehensive ordinance, the Government has adopted an incremental approach to improving the procurement system by amending and supplementing the existing legislation: e.g. Decree 66/2003 (amending Decree 88 and 14) and Circular number 1/2004 (providing guidelines on the implementation of Decree 53 In addition to procurement legislation, a number of other laws have significant impact on the effectiveness and efficiency of public sector purchasing. Existing or proposed provisions in the Civil Code, the Contract Law, the Construction Law, the draft Competition Law, the draft State Audit Law, the draft Arbitration Ordinance, not to mention the draft Law on Associations all impact on the contracting process. Procurement Management 141 66). The results have been mixed. On one hand, important principles of good procurement, including some of the CPAR's proposals have been enacted. On the other hand, the piecemeal approach to procurement legislation continues, with additional complexity added to an already cumbersome legal framework. Although Decree 66 introduced significant improvements, the fact that Decree 66 does not unify all procurement processes, negatively affects public resource management and performance. 9.16 While Decree 66 represented a step forward in unifying procurement regulations, other actions have increased fragmentation of the legal framework for procurement and introduced significant inconsistencies. The Construction Law, which came into force in July 2004, contains procurement provisions for the construction industry that are at odds with some of the core procurement principles that have been adopted by the Government of Vietnam. In particular, provisions in the law create floor prices for bids in an attempt to guard against dumping or severe under-biding. Well-performing procurement systems deal with such concerns by creating clear standards in biddingdocuments and effectively sanctioning poor performance-approaches discussed in later sections in this report. The current Construction Law provisions restrict the ability of Vietnam to capture the benefits of competition and are not in accordance with establishing a unified national procurement system that serves the needs of the nation. 9.17 When passing the Public Procurement Ordinance, attention should be given to ensuring that the ordinance rules apply universally to all procurement transactions. Sector specific procurement provisions should come as instructions issued under the public procurement law rather than as sector specific legislation. 9.18 The short term focus on implementation of current regulations does not diminish the pressing need for unified procurement legislation that establishes a solid foundation for an effective procurement system. There i s awareness of this need in the Government and among other procurement stakeholders. It i s uncertain, however, whether there i s sufficient political momentum to support its passagein the short term. 9.19 DPP hopes that the final ordinance will be generic enough to secure compromise and adaptability to the changing environment, but specific enough to make sure that the principles of what i s generally recognized as good procurement practice are clearly stated. The Guidelines and standard bidding documents based on the new ordinance would provide the specific tools needed to implement the ordinance. TRANSPARENCYFAIRNESSPROCUREMENT AND IN 9.20 The Government has followed up on the 2002 CPAR recommendation concerning the need for a single authoritative source on all bidding opportunities, contract awards and other procurement related matters. CP66/2003 has mandated the establishment of a Procurement Bulletin for advertising procurement opportunities and other procurement related information including relevant legislation, award results, and the list of firms that are prohibited from receiving state contracts due to past behavior. DPP has already launched a test pilot website, and the new system i s expected to be fully implemented and operational by the end of 2005. 9.21 The Government has decided to move gradually toward a fully electronic Bulletin by first launching a paper Bulletin in the interim before the electronic version i s ready. 142 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment This approach seems reasonable as the experience accumulated with the paper Bulletin could be quite useful when implementing the e-solution. A Procurement Bulletin titled "Vietnam Public Procurement" has been published twice a week since December 14, 2004. This Bulletin has already proven useful for parties to public procurement, contributing to transparency and publicity. 9.22 The creation of a Bulletin i s an important step towards building the transparency necessary for an effective procurement system. It will be important for the Government's implementation strategy to ensure that the Bulletindoes not contain so much information that i t i s difficult for clients to recognize opportunities or that the bulletin fails to include a large proportion of the contract opportunities that are of potential interest. Both of these problems can be avoided by the adoption of a phased-in strategy in which the initial postings capture all contracts of a value significant enough to attract national interest (irrespective of the investment category). The threshold floor can be lowered in time once the DPP i s convinced that it has systems in place to allow it to manage a larger flow of information without compromising on transparency. 9.23 A second best alternative would be to ensure that all contracts above a given value were available on provincial governmenVministries websites which could be accessed directly from the DPP central website. Such an implementation strategy would be consistent with the Government's policy to provide equal access to the significant market opportunities represented by public contracts. 9.24 The Bulletin also enables increased transparency regarding the performance of the procurement system. The design of the system should allow for the wealth of information gathered on biddingmethods, short lists, awardees, sanctions etc, to be further processed in order to obtain indicators of the health of the procurement system. Such indicators could be used by the policy and oversight agencies in assessing the strengths and systemic weaknesses of the public procurement system and determining necessary improvements 9.25 In designing its electronic system, DPP should examine the extent to which the new Treasury and Budget Management Information System (TABMIS) could provide relevant information to supplement its own database. The TABMIS system will include details of transactions by vendor and there i s likely to be significant value to ensuring that, at a minimum,the DPP database can easily inter-link with TABMIS. 9.26 Under CP66/2003, the Government has mandated that the new procurement website should include a bidders' database. Registration i s a condition for participating in bidding. The process is described in the Circular number 1/2004. DPP plans to proceed cautiously and only enforce the registry requirement when it i s confident that its information system i s operating effectively and will not work to disadvantage eligible firms. 9.27 Experience in public sector agencies around the world indicates that registries have a tendency to create needless time and staff costs for potential bidders and large time demands on those bodies tasked with registry management and upkeep (see Box 9.1). In many countries, registries that were established based on a purported need for information have degenerated into time consuming application processes that operate to restrict the entry of new domestic and foreign firms, especially when registration i s a pre- Procurement Manaaement 143 condition for bidding. InVietnam, the value of the information that would be obtained in a registry needs to be balanced against the potential risk that the registry will act to reduce competition. Prior to including a registration requirement in the final draft of the Procurement ordinance, the government may wish to consider whether its informational needs can be satisfied through malung using of existing sources of data, such as company registries, or through a data entry process designed to capture information from firms that bidfor and/or win contracts. If the government determines to move ahead with a registry, it will be essential for it to establish clear instructions and to buildin safeguards to ensure that the system does not operate to restrict competition. The easiest way to reduce concerns over the impact on competition of a registry would be for registration to occur simultaneously with the submission of a bid, or even upon the award of a contract. Regardless of when registration occurs, the registry should be explicitly identified as having a purely information purpose and only require the provision by bidders of basic data necessary for firm identification. The Procurement Ordinance should spell out the principles of a good registration system: fair, continuous access by all bidders, timely registration; and have simplified and streamlined registration procedures and processes described in detail in the procurement implementation guidelines. Linkages with other systems of registration should be forged to the greatest extent feasible in order to reduce time requirements for compliance and registry management. For example, in some countries the bidding registry establishes links to business, tax and licensing registries in order to reduce duplication of paper work and document checking. Box 9.1 Costs Associated with Registry Management: An Example from the World Bank Some agencies around the world, including the World Bank's Corporate Procurement, have tried to manage and update a bidders registry on a continuous basis but found it very expensive and time consuming. Thus, at the World Bank, a decision was taken to let the bidders register themselves, relying only the information they provide in the first stage. Applications are reviewed and validated only if the Bank is going to make an award based on a procurement process. Once the application i s approved, the vendor i s added to the Bank's "approved vendor list". In order to minimize costs and avoid the information overload, Bank Corporate Procurement removes from the vendor list those companies who have not been awarded contracts over the last three years. 9.28 Effective data management represents a challenge for DPP especially given the decentralized nature of the registration process and DPP may wish to consider whether the capacity needed to operate the system will be developed within DPP or obtained by hiring consultants with relevant data management experience. DPP can monitor the operation of the registry through regular consultation with the leading business associations including domestic and foreign firms. 9.29 Bulletin and Registry implementation may be uncoupled, giving first priority to the Bulletinwhich is more urgent for transparency reasons. For both the Bulletin and the Registry the Government should consider learning from the experience of other agencies in countries who are already implementing this type of mechanisms including how to implement such a bulletidregistry it in a decentralized system and the possible benefits from outsourcing the registry. 144 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment PROCUREMENTMETHODSBIDDING DOCUMENTS AND Procurementmethods 9.30 One of the deficiencies identified under the 2002 CPAR was the fact that open competitive bidding was not the default procurement method and justification was not required for the use of less competitive methods. Recent legislation (i.e. CP 66) has clearly demarked "open competition" as the "main method" of procurement. In practice open competition i s still limited: DPP statistics show that in 2002, 32 percent of the public bidding in terms of value was subject to open tender, 42 percent to limited bidding, 10 percent shopping and 16 percent to direct contracting. Statistics collected from Bac Ninh province suggest a much more limiteduse of full competition since out of 77 bidding packages in 2003, only one was procured through open bidding, while 45 contracts were awarded based on limited bidding and shopping, and 31 contracts came about as a result of direct contracting. Both statistical and anecdotal information point to the excessive use of less competitive methods and inadequate advertising under open bidding. 9.31 At the same time there i s a lack of equality afforded to all potential bidders. Foreign bidders' participation i s restricted by (i)using International Competitive Biding (ICB) only when there i s no local qualified bidder; (ii)mandating associations with local bidders; and (iii)enacting additional requirements under the 2004 Statute Governing Foreign Contractors. Such provisions may minimize the Government's benefit from potential lower prices and better quality associated with an enhanced competitive environment and may retard the development of a strong and efficient private construction industry. 9.32 Some skepticism about the benefits of competition remains at some levels of the bureaucracy. Open competition i s sometimes associated with buying cheap, low quality goods and services. However, there are studies (e.g. a study commissioned by MPI on recurrent cost procurement practices in August 2004) which clearly demonstrate the positive impact of competition in obtaining quality goods at reasonable prices in a province where competition was organized annually for health supplies. This i s contrasted with higher prices in provinces where competition was not considered. 9.33 In order to maximize the benefit resulting from competition, the following steps may be considered: i. Enforce advertising in mass media under existing open bidding until the Bulletin becomes operational, in particular for contracts of national interest. Revise advertising rules in mass media to ensure that interest in transparency i s not compromised by the associated costs (e.g. requesting the procurement agencies to advertise in three consecutive issues in mass media may be a disincentive to advertising). .. 11. Define the situations when direct contracting may be used irrespective of threshold with clear justification, such as extension of contract if there i s no advantage from further competition, equipment being proprietary, emergency such as response to national disasters, need for compatibility with existing equipment or a consultant uniquely qualified to provide certain services. Procurement Management 145 iii. Restrictlimitedbidding:CP88article3/2stipulatesthatoneoftheinstances limitedbiddingmay be used is when it i s deemed "more advantageous due to special circumstances". This statement i s open ended and subject to abuse and therefore we would recommend that this method be used only when a limited number of bidders can perform the contract (and the bidders can be identified in an exhaustive manner) or whenever the benefits of advertising are outweighed by the cost of the process (a monetary threshold could be set to limit its use). It i s also generally recognized that limitedbidding is a method prone to collusion and corruption and anecdotal information, such as some recent procurement in the oil and gas sector reported in the newspapers, strongly points to this type of misuse. iv. Secure equal access for all Vietnamese bidders to public funds, for example by reassessing the use of provincial preferences to find out whether they meet the objectives for which they were established or whether possible negative consequences (collusion, increase in local bid prices, unfair treatment of bidders in the neighboring provinces) outweigh the expected benefits. The recommendation i s to remove provincial preference altogether unless it i s meant to protect the very poor areas and the benefits are not outweighed by the costs. v. Gradually remove the above mentioned restrictions to international competitive bidding. With the Vietnamese economy becoming more open, it should capitalize on the benefits of global competition. Besides the general belief that exposing the local economy to international competition strengthens rather than weakens it, allowing access to foreign bidders may result in a similar treatment of Vietnamese bidders abroad. The U.S.-Vietnam Bilateral Trade Agreement contains a provision on removing restrictions on ICB precisely for this reason. As indicated before, special attention should be given to ascertaining that the registry and the regulations on foreign bidders who operate in Vietnam (as per CP66, article 23/57b) are not barriers to entry for foreign bidders. 9.34 The Government may also want to consider including in the Procurement Ordinance a provision allowing for the use of the full range of e-procurement actions (bidding, contracting) in order to create the legal basis for such transactions when the technology permits. Ensuringtransparentand effectivebiddingdocuments 9.35 Decree 66/2003 made an important step forward b y mandating the use of Standard Bidding Documents (SBDs) in line with the 2002 CPAR recommendations. A set of harmonized bidding documents were prepared and some Government agencies already use them (see Box 9.2). However, their development has been interrupted by the passage of new legislation. As per CP66, DPP i s currently preparing standard bidding documents for the procurement of goods and works and the selection of consultants, consistent with the existing decree. It i s strongly recommended that the preparation of the bidding documents, while being consistent with the current decree, incorporate the 146 Vietnam Public Expenditure Review and Integrated Fiduciaiy Assessment benefits and good practices gained under donor-funded projects and in the harmonization process. Box 9.2: Vietnam: Good PracticesinProcurement There are agencies in Vietnam who are at the forefront of promoting good procurement. Some good practices have been identified. For example, Electricity of Vietnam: has issued standard documents for goods, works and consulting services; i s publishing on its website invitations to bid and procurement related regulations; displays on the intranet its procurement manual where the rules for less competitive methods are described in more detail; and publicizes procurement issuedmistakes to its subsidiaries via the intranet. 9.36 One of DPP's highest priorities in the short to medium-term will be to ensure that SBDs are used once they have been issued. This effort is likely to require clear guidance on using the SBDs, an extensive training program to ensure knowledge of the standard procedures i s widespread both in government and the private sector, and monitoring of the use of SBDs in executing agencies. EVALUATION AWARD AND Bidevaluation 9.37 The quality of Bid Evaluation Committees (BECs) was identified as an issue in the previous CPAR. As a result, the recent legislation includes more specific requirements on the composition of the BECs (CP66/Art.11).Further attention should be paid to integrity requirements for evaluation committee members, as well as to their qualifications. Efforts should also be made to define more clearly the specific roles and responsibilities of the committee members so as to enhance accountability. For example, the scope of work of the evaluation committee including the individual and collective responsibilities of its members needs to be defined. Consideration should be given to establish parameters for minimizing repeated participation of the same members in bid evaluation committees. At the same time, more transparency can be brought into the process by including in such committees, particularly for community projects, representatives of the community/beneficiaries, as proposed in recent Government initiative to involve the communities/beneficiaries inproject supervision. 9.38 An important step forward was made in CP66/2003 by introducing the padfail method for technical evaluation in the procurement of goods and works. This innovation also made possible the use in the evaluation process of factors other than price, quantifiable in monetary terms. Although the government has aspirations to use the padfail method widely in the future, the method i s currently prescribed only for small value packages, and as an option in larger value contracts where technical specifications can be clearly defined. Unless steps are taken to promote the pass-fail method, the merit- point system i s likely to remain the dominant option in the evaluation of large value contracts which i s a source of great concern given its subjectivity and vulnerability to manipulation. Procurement Management 147 9.39 It i s strongly recommended that Vietnam establish a plan to guide the adoption of p a d f a i l evaluation as the dominant evaluation mechanism in the procurement of goods and works. Such a plan might begin with ensuring that pass-fail i s universally used in practice for small value packages. This may be followed by promoting the use of the pass-fail method in larger value contracts for the procurement of all goods and works whose technical specifications can be easily defined in advance. As Vietnam gains experience with the padfail method and in the development of technical specifications, the pass-fail method can be extended to all contracts with the exception of those where the merit-point system can be justified based on clearly defined circumstances (requiring special approval). The merit point could be used, for instance, for assignments where the intellectual input i s critical (e.g. consulting services, certain IT services or complex turn key contracts). Implementation of a strategy for pre-dominant use of the pass-fail mechanism would require setting and monitoring yearly targets based either on number of contracts or, preferably, volume of expenditures. 9.40 The implementation of the pass fail method will require a deliberate program to phase i t in and publicize the impact to various stakeholders, including: (i) identifying agencies who are already applying padfail and encouraging them to share their experiences with the others; (ii) identifying areas were there i s an urgent need to improve procurement such as the National Target Programs and/or the Public Investment Program; and (iii)monitoring the results of using the padfail system in these initial pilots in order to refine the broad roll-out plan. 9.41 It i s recognized that the pasdfail method requires an increased capacity in designing technical specifications. Improving the skills of consultants and technical staff working within executing agencies, along with training in evaluating bids using the path- fail method i s a necessary component of a program to make the padfail method standard practice. 9.42 In addition, the SBDs should clearly define the use in evaluation of criteria other than price and the way they are quantified in monetary terms, and should better distinguish between the technical evaluation of bids and evaluation of bidders' qualifications in order to ensure objective evaluations. 9.43 While CP66 has formally eliminated the use of trade names and country of origin, this practice continues through exploiting loopholes in the current legislation. There is genuine concern among practitioners that by removing such requirements the quality of the goods purchased will suffer. Box 9.3 provides an example of how, through adequate specifications, the desired quality can be obtained without specifying the trade name or the country of origin. When the use of a trade name i s unavoidable, it should always be followed by the words "or equivalent". Approval and award process 9.44 Vietnam has an award approval system which may include approval from three parties: the purchaser, the project owner and the competent authority. With CP66/2003 the process has been increasingly decentralized, devolving more authority to the lower levels. Further delegation i s expected to occur in the future. There i s concern that the potential for inefficient or ineffective use of public funds will increase unless care i s 148 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment taken to ensure that such delegation of authority i s coupled with increased capacity to conduct proper post reviews and increased accountability for the use of funds. Box 9.3: Examples of Specificationsto Buy HighQuality Computers without Indicating Trade Name or Country of Origin The equipment proposed shall be manufactured in facilities that apply quality management systems. This must be demonstrated in the bid by including a legible certificate (or certificates) issued by an independent authority, such as a certificate according to ISO-9000 (IS0 9001 andor 9002) or a newer I S 0 standard, or documentation certifying equivalenthetter TQM or quality design and manufacturing. The bid must include valid and legible certificates issued by the proper authorities in the manufacturing countries of the proposed equipment, for at least one of the following: FCC compliance, ULlisting, TUV approval, CE approval, or equivalent. IfWindows XP or 2000 is specified as the operating system, the specifications couldrequire proof of the offered workstation having certification by Microsoft Corp. for the required operating system. A particularly small footprint can be required for the desktop, or a very small box size for a tower. However, this should still be a size which typical major manufacturers known for quality are actually producing. Main computer unit, monitor and keyboard must carry the same brand name, and the brand must have been in existence for reliable workstations over at least the last two years. 9.45 It is important that the need for effective oversight and monitoring of decentralized decision-making in procurement not serve as a rationale for further increasing the burdenon executing agencies. Vietnam already has a wealth of bodies that have the right and power to conduct inspections and examine transactions, often before contracts are finalized. Unfortunately, this confusing control environment often works only to add delay into the procurement process. An oversight and monitoring process that supports decentralization's greater flexibility and responsiveness to local needs will primarily strengthen the ex post review of transactions in order to ensure that executing agencies have proper procedures in place and monitor adherence to those procedures. Reviews that examine compliance with rules can be supplemented with external analysis to ensure that individual procurement transaction represent good use of public funds. This type of oversight, especially connected to mechanisms to ensure follow-up or corrective action i s taken on problems or misuse identified, does not currently exist but i s beginning to be explored by such as groups as the Inspectorate of MPI and the State Audit Authority. Enhancing technical and managerial capacity at decentralized levels for performing approval and award processes, along with greater capacity for review b y oversight bodies i s essential in order to minimize the potential fiduciary risk of delegation. 9.46 When multiple levels of review are involved, a system of establishing and monitoring response times should be considered to reduce delays in order to promote economy, efficiency and transparency in the procurement process. DPP has a role in establishing standards for monitoring and oversight and carrying out independent evaluation to ensure oversight i s conducted effectively. 9.47 With the Vietnamese economy becoming more open to global markets, price fluctuations are bound to become more significant and the new legislation should take this into account. While the ability to determine a realistic owner's estimate should be Procurement Management 149 improved, the owners' estimate should not be used as an absolute/automatically applied benchmark in the evaluationlaward process and contract price escalation. While discouraged under CP66 and application guidelines, in practice such procedures are still applied. 9.48 The use of ceiling and floor prices to reject bids without further analysis is strongly discouraged in public procurement as it removes the benefits of competition from the procurement process. Each bidder should be free to quote its own price without fear it will be rejected if higher or much lower than the estimate. However, when deemed necessary, the Government agency should conduct a price/cost analysis, and seek, in writing, justification from the bidders in order to verify that the bidprice i s realistic. Only if this analysis shows that the unit prices are unrealistic should the bidbe rejected. This practice of seeking clarifications (e.g. when the price significantly deviates from the estimate) i s mandated under the CP66 and needs to be strictly enforced. 9.49 It is also recommended that studies be carried out to assess the reasonablenessof the Government's cost-norm system in estimating construction costs. Evidence suggests that the norm system sometimes results in inflated costs in Vietnam. Also, the establishment of a data-base of prices of various categories of goods and works, based on awarded contract information, may be considered for reference purposes. CONTRACT IMPLEMENTATION 9.50 The discussion so far has concentrated on the procurement process andtouched in passing on its substance, which i s the objective of contract implementation. Ensuring efficiency and economy in the use of public funds obviously does not end nor begin with successful procurement. The procurement process i s a means of selecting a contractors and forming a contract, which i s a promise between two parties to perform and can be enforced under law. The parties are equal before the law and neither may impose its will on the other. Such certainty i s critical in ensuring that both parties will not face unforeseen risks and suffer unanticipated losses, thus ensuring reasonable value in contract price. This safeguard i s central to the commercial transactions of a well functioning market economy. Contract formation i s an intermediate step in the contract management process that commences with the planning of the contract and establishment of appropriate terms and conditions, and i s followed by objective administration of the agreement. 9.51 Ordinarily, the quality of the contract agreement and effectiveness of its administration should be critical to achievement of the desired results. The previous sections have emphasized the importance of clear contract terms and conditions consistent with Vietnamese law. The World Bank advocates in its SBDs the use of Federal Deposit Insurance Corporation (FDIC) standard terms and conditions as an example of best practice in meeting fairly the interests of both parties. However, as mentioned earlier, the Vietnamese legal system i s incomplete and in practice it i s doubtful that disputes, however the contract i s framed, can be readily or equitably resolved through the courts. Moreover, no formal arbitration mechanisms are available for contract disputes to which the government i s a party. It i s also important to recognize that in many cases in Vietnam, neither buyer nor seller and certainly not the domestic legal profession andjudiciary has much experience in interpretation of a FDIC contract. 150 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment Hence, all parties are more likely to rely on familiar administrative practices prevailing for traditional transactions between state entities. Under these circumstances, it i s not unusual for third parties to interfere in contract administration. Thus, the nascent private contractor in particular faces considerable uncertainty in controlling costs and receiving due payment, in addition to the many operating disadvantages faced vis-a-vis state enterprises. 9.52 Vietnam does not enjoy a fully functional market for construction goods and services. The dominance of the state-owned enterprises as well as the absence of an experienced private construction industry and the unfamiliarity of government agencies in the administration of legally binding contracts all add to the distortions in the market for construction services. Under these conditions, sound procurement procedures alone cannot ensure best value for money and risk-adverse behavior such as collusion among bidders should not be surprising. Contract prices or due performance in the absence of objective contract administration cannot be considered reliable. 9.53 Other contract implementation issues exist. First, the market for public service contracts i s affected by delays in payments for works performed, frequently as result of contracts signed by budgetary organization that do not have funds either to initiate or complete projects. Second, a process of extensive reviews (requested even when they are not required) and detailed inspections do not seem to achieve the preventive effect expected. Third, the quality of contract supervision i s widely regarded as weak. 9.54 Better contract management i s to a significant extent linkedto the evolution in the market environment. Thus, it i s expected that the SOE reform will lead to a level playing field of commercial entities subject to the enterprise law. Furthermore, improved public expenditure management will result in better budgetary discipline so that agencies can not commit funds that were not appropriated; it i s expected that the new TABMIS system will progressively introduce better commitment controls to mitigate such risks. At the same time, contract specific reforms should include improved contract law, enhanced contracts management and supervision skills, flexible price adjustment mechanisms, and effective independent oversight. INSTITUTIONAL FRAMEWORK Procurement policy formulation 9.55 The 2002 CPAR noted a lack of an appropriate and autonomous agency responsible for procurement policy formulation. Such an agency has not been created, but the Government has strengthened and enlarged the functions of the Department for Public Procurement (DPP) within the Ministry of Planning and Investment. DPP serves as the focal point for public procurement policy and administration. DPP i s also the policy formulation agency, in that all legislation related to public procurement i s meant to originate from its office. Although DPP has been designated as the lead procurement policy organ, there remains some inconsistency in the actual setting of procurement policy in Vietnam. Passage of the previously cited Law on Construction, which contains inconsistent and conflicting regulations on procurement in the construction industry, signals that ministries are continuing to develop their own procurement practices. Establishing and enforcing the primacy of DPP in defining public procurement policy i s Procurement Management 151 fundamental to improving the effectiveness of public procurement. The Government should determine what steps are necessary in order to ensure that the authority of DPP i s sufficient to enable it to carry out its function. 9.56 While not completely "autonomous" in the sense of a stand alone agency under the Prime Minister or National Assembly, the DPP performs the key functions of a public procurement agency including formulation of public procurement law, development of SBDs, issuing implementation guidelines, compliance oversight, administration of a National Procurement Bulletin, preparation of an Annual Report, development and maintenance of a procurement data base, support and coordination for the development of new training systems and networking with public procurement agencies from other countries. In addition, DPP has a prior review control function for all bidding packages under the PrimeMinister's authority. 9.57 In order to be able to deliver its multiple mandates, DPP will need significantly to enhance its capacity. It i s thus recommended that a strategic plan be prepared outlining how the agency will support the many functions it has recently acquired, how much it i s possible and desirable to contract out to other agencies and what activities would need to be carried out exclusively b y the DPP with staff and resources available. In this context and with the continuing decentralization of procurement the prior review function of DPP i s rightly gradually decreasing except for all but the most critical contracts. 9.58 Monitoring the progress of procurement reform will be one of DPP's core tasks, especially in this period of implementation of Decree 66. Experience in other countries indicates the importance of defining a simple and constructive framework for monitoring progress in order to determine what has been achieved. A possible such framework i s set out inBox 9.4. Box 9.4: Possible Framework for Monitoring Progress inProcurement Pillar 1: The Legislative and Regulatory Framework (indicators: procurement methods, advertising, evaluation, tender documents, complaint review process, implementing rules and regulations) Pillar 2: Central Institutional Framework and Capacity (indicators: linkages with the budget process, functional procurement normative body at the center). Pillar 3: Procurement Operations and Public Procurement Market Performance (Indicators: efficient procurement operations and practices, functionality of the public procurement market, contract administration and dispute resolution) Pillar 4: The Integrity and the Transparency of the Public Procurement System (indicators: control, audit, appeals mechanism, access to information, ethics and anticorruption measures.) Source: World Bank Mechanismfor dealingwith complaints 9.59 A cornerstone of a well functioning public procurement system i s the availability of a robust mechanism for bidders to file complaints and challenge decisions, preferably with independent administrative and judicial review bodies. Vietnam has recently taken significant steps in recognizing the importance of complaints across the public sector, through the adoption of the Law on Administrative Complaints, as well as the inclusion of sections on complaints in fiscal transparency regulations. The possibility of complaints in procurement has been recognized (CP66 24/25). However, there currently exists no 152 Vietnam Public Expenditure Review and Integrated Fiduciary Assessment formal mechanism for addressing them in Vietnam. There are no statistics available as to the magnitude of the complaints at the national level and no monitoring process concerning the source of the objection or how it has been resolved. DPP i s aware of the need for an effective complaint system for procurement and has included in the proposed draft procurement ordinance, provisions that allow for filing of complaints. However, details on the design and implementation of a system for addressing procurement complaints are yet to be developed. As a first step, the system for addressing complaints should specify the levels of review and should include in its structure an entity independent from the procuringklearing entity. The details of the complaint mechanism should also be provided in the bidding documents, including to whom the complaint should be made and the time frame for responses, additional recourses at higher level without stopping the procurement process. 9.60 Establishing an effective procedure for addressing complaints in procurement by a fully independent administrative and judicial review body i s likely to be a lengthy process that i s strongly influenced by the development of a robust administrative complaint procedure in the nation. Box 9.5 sets out the features of a mature procurement complaint process that the Government may wish to consider as it builds its own system. ~ Box 9.5: The Complaint System inthe Republic of Slovenia The Complaint Process was Established by the Act on the Review of Public Procurement. A complaint can be raised anytime inthe procurement process-from prior to bids being opened to after the awarding of a contract. Complaints can be raised by bidders, as well as the Public Procurement Office, the Office of the State's Attorney, and the Office of Competition. Complaints are first addressed to the contracting entity itself. The entity appoints a Review Expert from a list maintained by the Ministry of Finance. The Expert reviews the complaint and the entire procurement process, and must provide his opinion within 8 days. The contracting entity must decide how to act upon the opinion within 3 days of receiving the opinion. If the contracting entity does not amend its position, the complainant has 3 days to submit the compliant to the National Review Commission. The Commission is an independent body composed of a judge who acts as president and 4 members appointed by Parliament. The NRC i s supported by a staff of 8 experts and an administrative personnel. The NRC provides its judgment within 15 days (except in exceptional circumstances where an extension of up to 20 days i s possible). The NRC can reject the claim or annul the procedure in whole or in part. I t provides a reason for all of its judgments. There is no appeal of an NRC judgment. The NRC does not have the authority to award damages since this authority is vested exclusively inthe courts. Procurement oversight 9.61 Decree CP66 adds additional specificity to the process of conducting inspections, including identifying the agencies involved, the sanctions applied and the areas of risk. It further stipulates that sanctioned bidders are to be publicized in the Bulletin. In addition to the MPI's oversight role (DPP and inspectorate) mentioned under Decree 66, other agencies are involved, to certain degrees, in procurement and contract implementation oversight, whether the financing sources are capital or recurrent costs. Examples include the State Inspection Office, the Ministry of Finance's Financial Inspectorates, the State Audit of Vietnam and the inspectorate departments in the ministries and provinces. As discussed in Chapter 5, this creates a complicated and less than fully effective control environment (not only for procurement). Procurement Management 153 9.62 The need to avoid duplication and enhance coordination i s recognized by Government, which i s preparing a new Inspection Law, Procurement and contract management oversight should be seen as part of this overall oversight mechanism which i s being rationalized in order to maximize its impact. In modernizing the oversight system special attention should be given to moving from a transaction-based approach to a systems-based approach. This would entail balancing policing with greater attention to prevention through the adoption of a risk based approach to oversight that focuses limited resources on area of high risk and a gradual move from exhaustively scrutinizing contracts prior to their signing to reviewing signed contracts to ensure adherence to law. We have earlier described how post-review, combined with a process for addressing complaints, forms the basis for effective oversight and accountability. The deliberations on the draft ordinance within the National Assembly give some insight into the costs of the existing approach and evolving attitude to the inspection function in Vietnam. Delegates voiced concerns regarding overlapping inspection functions and frequency of inspections. Some enterprises were inspected as many as 10 times. In one case, the enterprise received28 visits from inspectorates duringthe year. 9.63 Efforts should be made to increase the involvement of communities, beneficiaries and NGOs in the procurement process and contract implementation supervision, to promote transparency and accountability. The Government has shown its commitment to step up the involvement of communities by promoting a resolution on community supervision of investment projects. 9.64 The future ordinance should be more precise in differentiating between the roles of different oversight agencies so that duplication of responsibility i s avoided. The new law also should be clearer as to the due diligence process in sanctioning the civil servants andor the bidders violating the procurement regulations (e.g. review steps, right to appeal, who formulates sanctions). The process, like in the case of the complaints mechanism, should rely on the judgment of an independent entity, which has a reputation for fairness and balanced decision making. 9.65 Another related development i s strengthening of the role of the State Audit Agency. The agency has responsibility for conducting audits to ensure that procurement transactions are in accordance with applicable laws and regulations. It i s important that staff at the Audit Agency be provided the training necessary in order to fulfill their function and to establish a close working relationship between DPP and the Audit Agency. Ideally, the work of the State Audit Agency should help DPP identify agencies most in need of support, and the rules that are most inneed of revision or clarification. Procurement staff development 9.66 The Government has become increasingly aware of the need to build capacity of the procurement staff. There is a sense of urgency as the decentralization process i s underway. In this context, a Public Procurement Capacity BuildingProgram was recently prepared by MPI and donors. This study identifies systemic weaknesses in public procurement including: (i) shortage of qualified procurement staff; (ii) contract weak implementation supervision capacity in procuring entities; (iii) lack of necessary capacity for provision of education and training on public procurement; and (iv) no systems of incentives (such as accreditation) for procurement professionals. After review by the 154 Vietnam Public Exuenditure Review and Intenrated Fiducian,Assessment Government the priorities and financing sources to implement a sustainable capacity buildingprogram should be identified, which should take into account and bring together the various capacity building initiatives that are presently underway with the support of different development partners. 9.67 Main recommendations regarding procurement management capacity include: i.strengtheningtheDPP'scapacitythroughadequatestaffingandtechnical assistance; ii.providingtrainingmanagementknowledgeforrelevantDPPstaff; iii.identifyingapoolofkeytrainersoutsidetheagencyinordertoconductthe capacity building program; iv. identifying learning institutions and other entities able to provide the required facilities; v. liaising with universities in order to include intheir curricula procurement and contract management courses; vi. defining an accreditation program based on the programs prepared. 9.68 Given the limited resources there i s a necessity of determining core priorities and adopt a sequenced approach to capacity building. Further work i s required to refine the breakdown of target groups for national training, confirm the content of the training curriculum, and develop institutional partnerships. The resulting project if successful will become a self sustaining, on going national program inpublic procurement training. THECONSULTINGSECTOR 9.69 The work on procurement as part of the PER-FA included a review of procurement practices and capacity in the consulting services sector. This exercise, the first of this kind in Vietnam, looked at the strength and weaknesses of the consulting profession and underscores the importance of intellectual capital formation in the economic progress of the country. A key finding of this review i s that the operating environment in which consultants are selected and employed lacks sound professional practices. Transparency and accountability problems are reported throughout the process of selection and during execution of such services, which i s generating a substantial waste of resources and undermining development of the sector. The presence of numerous public sector affiliated consulting organizations constitutes an obstacle to high quality, competition-driven services and to the formation of independent and viable private consulting firms. In addition, the legislative framework i s fragmented and not aligned to international good practices. Finally, there i s a lack of familiarity with the role and purpose of independent consulting services and insufficient professional and policy dialogue in this area. 9.70 The key recommendations of the above review are as follows: i. Mechanisms need to be established for more effective dialogue between ministries, consultant associations, academic and scientific institutions and other stakeholders in the consulting services sector. Procurement Management 155 ii.Basedon feedback from stakeholders, the Government shouldreview and update the legislative and policy framework governing the consulting services sector, with an eye toward gradually removing the barriers hindering the emerging domestic consulting services market. iii.DPPstaff capacityinpolicyformulation andregulationfor the selection of consultants, and the development of national capacities should be increased. iv. Plans should be defined for gradual introduction of fair competition among consultants for government funded assignments by gradually cutting all undue advantages of public sector affiliated consulting organizations. v. Government should strengthen, by adequate training and incentives, the capacity and integrity of government staff assigned to select consultants and manage consulting contracts. PROCUREMENT ENVIRONMENT 9.71 Public procurement needs to be seen in the context of Vietnam's market economy. Vietnam still maintains an economically important public sector. The private sector i s constantly growing: more than 1,200 new private enterprises are registered on average every month and the SOE commercial sector i s gradually decreasing, although it still accounts for 40 percent of GDP. While the Government i s keen to continue to retain ownership in areas considered strategic, it tries to implement a level playing field which i s essential for a fair and transparent public procurement environment. 9.72 The objective is still to be attained. There i s strong evidence of preferential treatment vis-a-vis the SOE sector. For example, the Government often requests waivers from the World Bank's rule that dependent SOEs are not eligible to participate in World Bank financed procurement. Such rules are not applicable in Government procurement, and therefore it i s unlikely that the competitive environment i s level if such preference i s given. The PER-FA'Sreview of the consulting services sector in Vietnam, for example, points to strong preference for SOEs' services. This represents a source of inefficiency and a barrier to the emerging private consulting professions. Priority should be given to those areas which can be outsourced without causing significant disruptions. Transitional arrangements should to be put in place (see Box 9.6) Box 9.6: Transitional arrangements for procurement When Canada privatized some Government functions, transitional arrangements were set in place. Competitive advantages were thus given to the affected agencies such as: (a) set-aside from competition for defined functions; a n d or (b) a 5-10 percent advantage (of total technical and financial points) over other bids. This was granted, however, for a limited period of time (2-5 years), with a clear end-date. At the end of the transition period, preferences were completely removed and the competitive field was level with former state owned companies competing on an equal footing with the private sector. 9.73 The new SOE Law of November 2003, effective July 1, 2004, has allowed for the creation b y SOEs of new types of holding groups called "General Corporations established and invested by companies" which are market based commercial entities operating under the Enterprise Law. However, the relationship between these companies and the State-owned parent company continues to create concern regarding conflicts of interest. There i s even more concern regarding the traditional forms of industrial 156 VietnamPublic Expenditure Review and Integrated Fiduciary Assessment groupings-the "General Corporations established by the State"-which will continue to have unreformed dependent and independent accounting SOEs intheir structures. 9.74 Issues regarding discipline in budget execution and commitment control are an added aggravating factor for securing a level playing field in procurement. Preliminary studies indicate that provinces frequently commit more funds than appropriated through the budget process for procurement purposes, and require contractors to wait long periods of time to receive payment. These financial practices create a preference for companies such as SOEs or well established private companies who have the resources to pre- finance contracts in the medium term. In this manner, small and medium-sized private companies are restricted in their ability to compete for state contracts to the detrimentof economic growth and the state budget. 9.75 Government i s committed to implementing SOE reform, and key steps have been agreed with donors under the Poverty Reduction Support Credit (PRSC) programs. In addition, the reform of the budget process and the establishment of the new TABMIS system will strengthen budget discipline and have a beneficial impact on the equal access to public contracts by the different Vietnamese companies irrespective of ownership. 9.76 The Competition Law, now under preparation could potentially improve the competitive environment. A key role in improving the business environment (including the environment for public procurement) in Vietnam i s played by the Chamber of Commerce. One of the objectives of the Chamber of Commerce i s to secure an open the communication link between the business community and the Government. In this capacity, the Chamber of Commerce could step up its role in public procurement and serve as a facilitator of the dialogue in this area as well. It could gage the pulse of its constituency by including procurement related questions in its periodical surveys and the key findings could be usedinits annual report to the Government. TACKLING CORRUPTION INPROCUREMENT 9.77 As noted in Chapter 5, the fight against corruption i s an essential component of the Government's agenda and the public procurement reform i s one of its pillars. Thus, Decree CP66 addresses new concepts relevant to this agenda: collusion, conflict of interest, bidders' debarment and sanctioning of civil servants involved in fraudulent and corrupt practices. 9.78 The Government has recently taken a more strategic approach towards addressing corruption and integrity issues. In January 2004, the Prime Minister instructed the State Inspectorate to start preparing a comprehensive anti-corruption strategy, seeking technical assistance from the international community. 9.79 In procurement, many of the reform initiatives which have been described above are conducive to a more transparent, accessible, competitive public procurement system with fewer opportunities for corruption and more accountability. These include: creating a sound legal system with clear implementation regulations; adopting an effective complaint system; ensuring equal access by promoting competition and broad advertising; adopting SBDs; strengthening internal and independent oversight; providing adequate training of the procurement corps; and promoting a culture of integrity and accountability. These actions will be most effective at reducing corruption if they are Procurement Management 157 combined with efforts to involve and empower project beneficiaries through increased disclosure, as well as providing opportunities for greater citizen participation in the procurement process and in contract implementation. 9.80 Other more targeted measures may include the following: i. adoptingariskmanagementapproach; ii.training staff in corruption awareness and in identifying red flags of corruption (abuse of less competitive methods, narrow contract specifications, incomplete files, lower bids declared non-responsive without justification); iii.traininggovernmentstafftoidentifyindicatorsofcollusion(identicalformat, unit prices, grammatical and arithmetical errors etc); iv. strengthening the capacity of DPP and MPIInspection Department in this area and enhancing their collaboration among themselves and with the State Audit Agency; v. making the bidding documents accessible through the website, when available to minimize intermediation; vi. incorporating in the new ordinance provisions on the ethical conduct of procurement staff and contract administrators; and vii. building in the procurement data base analytical tools to identify suspicious or highrisk transactions. 9.81 While emphasis should be on systems and prevention, sanctions should be enforced when needed, proportionate to the violation and with strict observance of due process. SUMMARY OF KEYRECOMMENDATIONS 9.82 It is essential to ensure that the procurement regulations that are in force are fully and consistently implemented. In order to implement Decree 66, DPP should focus on providing the necessary support and instructions to guide and direct proper application of rules. Specifically, in this regard, the PER-FA recommends the Government should: 0 issue instructions and standard bidding documents to guide staff in executing agenciesin implementing Decree 66; 0 supplementthe new paper-basedProcurement Bulletin with a website version; 0 train staff in implementing required procedures; 0 establish a way to monitor the application of procurement rules and procurement outcomes; and 0 establish a mechanism for receiving and responding to complaints. 9.83 The existing procurement regulations do not provide an adequate basis for a high- performing procurement system. In order to create a unified regulatory basis for procurement that encompasses effective and efficient procedures, the PER-FA recommends the Government should: 158 VietnamPublic Expenditure Review and Integrated Fiduciary Assessment 0 finalize the procurement ordinance/law as a unified legal document, addressing consultant selection as well as procurement of goods and works, and submit it to the Standing Committee of the National Assembly. 9.84 Current procurement practices will need to be modified in order to implement either Decree 66 or the Procurement Ordinance (when it i s enacted). The PER-FA recommends that: 0 DPP should define a multi-year capacity development strategy to ensure that managerial and technical staffs have the capacity required to perform their required tasks and functions. 9.85 The impact of improvements in the legal basis for procurement will have limited effect on procurement outcomes unless the Government of Vietnam significantly improves its contract administration. To this end, the PER-IFA recommends the Government should: 0 develop a program to develop enhanced capacity to manage contracts; 0 establish guidelines and instructions to enhance the quality of design and contract supervision; and 0 strengthen enforcement of contractual provisions and respect for formal processes to resolve disputes over commercial contracts.