Report No. 21021 -VN Vietnam Managing Public Resources Better Public Expenditure Review 2000 (In Two Volumes) Volume 2: Annexes December 13, 2000 Poverty Reduction and Economic Management Unit East Asia and Pacific Region Document of the World Bank Vice President * Jemal-ud-din Kassum Sector Director : Homi Kharas Task Manager : David Shand TABLE OF CONTENTS ANNEX A: THE PUBLIC INVESTMENT PROGRAM ................. ............................. 1 INTRODUCTION .....................................................................1 THE FIVE YEAR PLAN, THE PUBLIC INVESTMENT PROGRAM AND THE ANNUAL ECONOMIC AND SOCIAL PLAN ...................................................................... 1 THE 1996-2000 PIP .................................................................... 2 SOME LESSONS FROM THIS PIP ..................................................................... 3 MID-TERM REVISION ..................................................................... 4 TIMING OF THE PLAN AND THE FINANCING THE INVESTMENT PROGRAM .................................. 5 PIP PREPARATION .................................................................... 6 TECHNICAL APPRISAL AND POLITICAL CHOICE .................................................................... 6 THE RELATIONSHIP BERWEEN PUBLIC INVESTMENT AND RECURRENT EXPENDITURE .............. 7 DECENTRALIZED INVESTMENT DECISION-MAKING .................................................................... 8 MPI COMMENT ON THE PIP PROCESS ..................................................................... 9 Concepts .................................................................. 9 The Content of Public Investment Program 1996-2000 ....................................................... 10 The mid-term adjustment ................................................................... 12 Successes and limitations of the investment program .......................................................... 13 Continuing problems in preparation of public investment projects ..................................... 14 The PIP 2001-2005 .................................................................. 14 ANNEX B. FISCAL DECENTRALIZATION AND EFFICIENCY AND EQUITY IN PUBLIC SERVICES ................................................................ 17 I. INTRODUCTION ..................................................................... 17 II. FISCAL ARRANGEMENTS IN VIETNAM .................................................................... 17 Fiscal Assignment .................................................................. 18 Budgetformulation process ................................................................... 19 Trends in expenditure decentralization .................................................................. 21 III. EXPENDITURE ASSESSMENT, ALLOCATIVE AND TECHNICAL EFFICIENCY ........................ 23 Subjectivity in Allocations .................................................................. 23 Lack of Maneuverability ................................................................... 24 Inadequate Revenue Assignments ................................................................... 24 National Programs .................................................................. 27 IV. FISCAL ARRANGEMENT AND INTER-REGIONAL EQUITY .................................................... 28 V. REFORMING INTERGOVERNMENTAL FISCAL ARRANGEMENT IN VIETNAM ......................... 33 Independent Revenue Sources, Incentives and Efficiency in Service Delivery .................... 33 Needfor a formula based transfer system ................................................................... 34 Improving the Design and Implementation of National Programs ...................................... 35 ANNEX C : COMMUNITY-LEVEL FEES AND CONTRIBUTION ......... ...................... 43 BACKGROUND TO THE STUDY .................................................................... 43 STUDY FfNDINGS .................................................................... 44 Households pay a large number offees and contributions ................................................... 44 Households receive little information about the basis of charges. The system is confusing and non-transparent .................................................................. 45 Contributions andfees are a very important source of commune financing ....................... 46 Contribution-payers know little about how their money is spent ......................................... 47 Contributions are not strongly linked to ability to pay ................................................I ...... 47 i Certain fees and contributions are particularly unpopular ................................................. 49 Recommendations .......................................................... 50 ANNEX D. AGRICULTURAL SECTOR ....................................................... 55 ROLE OF AGRICULTURE IN VIETNAM'S ECONOMY ............................................ ............. 55 PUBLIC AND PRIVATE ROLES IN THE AGRICULTURAL SECTOR ..................................... ........... 56 PUBLIC EXPENDITURES IN THE AGRICULTURAL SECTOR ......................................................... 57 The Government's Priorities in the Agricultural Sector ...................................................... 57 Level and Composition of Public Expenditures in the Agricultural Sector ......................... 58 The composition ofpublic expenditures in agriculture ........................................................ 60 The composition of subsidies and current transfers ......................................................... 61 The regional distribution ofpublic expenditures in the agricultural sector ........................ 61 Public Expenditures on Agricultural Research ......................................................... 62 Public Expenditures on Agricultural Extension ......................................................... 65 Public Expenditures on Forestry ......................................................... 67 Public Expenditures on Irrigation ......................................................... 68 Public Expenditure in Agricultural Production and Processing Activities ......................... 72 Agricultural Expenditures and Gender Equity ......................................................... 75 BUDGET DECISION-MAKING AND INSTITUTIONAL ISSUES ....................................................... 76 A SET OF REFORM OPTIONS ......................................................... 76 FUTURE RESEARCH NEEDS ......................................................... 77 BIOLOGRAPHY .................................................................. 78 ANNEX E - HEALTH SECTOR ....................... ................................ 87 OVERVIEW OF THE HEALTH SECTOR ......................................................... 87 Infant Mortality ......................................................... 87 Morbidity and Its Causes ......................................................... 88 Child Malnutrition .......................................................... 90 Issues in a Market Economy ......................................................... 90 Provision and Utilization of Health Services ......................................................... 91 Users of Public Health Facilities ...................... 95 The Public Expenditure Decision-Making Process in Health ............................................ 96 Levels of and Trends in Health Expenditure ............................................ 98 Public Expenditures on Health ................................................ 99 Cost Effectiveness of Public Spending ................................................. 100 Composition of Government Health Expenditure ............................. ................... 102 Provincial Distribution of Health Expenditure ................................................ 108 Public and Private Spending on Health ................................................ 109 Provincial Spending Disparities over Time ................................................ 110 Health Financing Policies ................................................ 110 A Menu of Reform Options ................................................ 117 REFRENCES ................................................ 122 ANNEX F. EDUCATION SECTOR ................................................ 123 A. BACKGROUND TO EDUCATION PROVISION IN VIETNAM ................................................ 123 Current Education Policies and Priorities ................................................ 125 Frameworkfor Delivering Government Policies ................................................ 127 B. LEVEL OF AND TRENDS IN PUBLIC EDUCATION SPENDING ................................................ 131 Share of GDP and Public Spending Devoted to Education ............................................... 131 Shares of Education Expenditure in Total Public Expenditure .......................................... 133 Recommendations - Resource Assignment ................................................ 133 C. COMPOSITION OF EDUCATIONAL SPENDING ................................................ 134 Subsectoral Shares ofAllocations ................................................ 134 ii The Composition of Education Expenditures - School Education ..................................... 135 Expenditure Norms ....................................................... 137 D. EQUITY OF RESOURCE DISTRIBUTION . ......................................................... I41 Geographical Distribution ....................................................... 141 Benefit incidence of Public Expenditure ....................................................... 143 Cost sharing with Parents ....................................................... 145 Gender Balance ....................................................... 149 Recommendations - Equity of Resource Distribution ....................................................... 149 E. EFFICIENCY OF RESOURCE USE ........................................................ 150 Staff Utilisation ....................................................... 151 Salaries. ....................................................... 151 Pupil Teacher Ratio (PTR) ....................................................... 152 Higher, Technical and Vocational Education ....................................................... 153 The Quality of Teaching and the Motivation of Teachers .................................................. 154 Infrastructure ....................................................... 155 Higher Education ....................................................... 156 Planning and Management ....................................................... 156 Recommendations - Efficiency of Resource use ....................................................... 157 F. OUTPUTS AND ACHIEVEMENTS .................................................... . .. 158 Enrolments ....................................................... 158 Repetition ....................................................... 159 Retention ....................................................... 159 Quality of Education ....................................................... 160 Recommendations - Improvement of Service Quality ....................................................... 162 G. POLICY IMPLICATIONS OF THE ANALYSIS ...................... .................................. 162 Recommendations - Development of Education Policy ...................................................... 163 ANNEX G. TRANSPORT SECTOR ................................................... 175 SECTOR OVERVIEW AND PERFORMANCE ........................................................ 175 Sector Overview ....................................................... 175 Sector Performance ....................................................... 177 GENDER AND TRANSPORT IN VIETNAM ........................................................ 178 GOVERNMENT OBJECTIVES AND PRIORITIES ....................... .................................. 179 EXPENDITURE LEVELS AND TRENDS IN PUBLIC TRANSPORT ................18............................... I 0 Quality of Information ....................................................... 180 State Central and Local Transport Expenditure ....................................................... 181 Sub-Sectoral Allocations and Capital Expenditure ....................................................... 182 Spending by Regions ....................................................... 183 Current Expenditure ....................................................... 184 User Charges for Non-revenue Earning Road Transport Infrastructure ........................... 185 The 1996-2000 Public Investment Program (PIP): Plan vs. Actual ................ .................. 188 Ministry of Transport Expenditure Information ....................................................... 188 IMPROVING THE QUALITY AND MANAGEMENT OF TRANSPORT INFRASTRUCTURE PROJECTS 190 Quality of Proposals .................................................... 191 Project Implementation .................................................... 194 Program and Project Monitoring .................................................... 195 CHANGING THE ROLE OF THE PUBLIC SECTOR IN TRANSPORT ............................................... 195 Organization of the Transport Industry .................................................... 196 Role of the State .................................................... 196 Divestiture Program for Transport SOEs ..................................................... 196 Increasing the Scope for Private Sector Participation .................................................... 197 Future Expenditure Strategy .................................................... 199 iii BOXES ................................................................................................. Box G. 1. A Road Fundfor Vietnam? ............................................................. 187 Box G. 2: Coordinated Infrastructure Planning in Malaysia ............................................. 192 Box G. 3. Manila International Container Terminal (MICT) ............................................. 198 Box G. 4: Broad Strategy for Future Spending ............................................................ 200 FIGURES............................................................................................... Figure B. L. Share of Central and Local Governments in State Expenditures .......... ........... 22 Figure B.2. Expenditures in Provinces - Plan and Actual (1998) ........................................ 26 Figure B. 3: Revenue Equalization in Vietnam, 1998 ........................................................... 29 Figure B. 4: Distribution of Per Capita Expnditure Among Provinces, 1998 ...................... 31 Figure D. 1: Per capita Agric Spending vs. Per capita Ag GDP by province, 1998 ............ 62 Figure E. 1. Infant Mortality Rate, 1960-98 ............................................................ 87 Figure E.2. Infant Mortality Rates and Real GDP per Capita Across Asian Countries, 1997 ............................................................. 88 Figure E.3. Major Causes of Morbidity, 1976-97 ............................................................ 88 Figure E. 4: Prevalence of Underweight Children Aged 0-59 Months, Selected Countries in Asia,1975-95 ................................................................. 90 Figure E. 5. Annual Public-Sector Health Contacts per Capita, Asia, 1990-95 .93 Figure E. 6: Annual Inpatient Admissions per 1, 000 Persons, Asia, 1990-95 .93 Figure E. 7. Economic Composition of Users ofDifferent Health Providers, 1998 . 95 Figure E. 8: Lorenz Distributions of Service Contacts with Public Hosipitals and Commune Health Centers in Comparision to Lorenz Distribution of Aggregate Consumption Expediture, 1998 .96 Figure E. 9: Total (Public Plus Private) Health Spending per Capita, Selected Asian Countries, Latest Year (1991-98) .98 Figure E. IO: Total Public Spending on Health, Vietnam 1991-98 .99 Figure E. 11. Sources of Health Financing (%0 of Total Expenditure on Health Contributed by Each Source), 1998 ............................................................ 100 Figure E. 12: Source of Funding Public Expenditure on Health, 1991-98 ........................ 100 Figure E. 13: Productivity of Public Health Spending, 1992-98 ....................................... 101 Figure E. 14: Life Expectancy and Real Public Spending on Health per Capita across Low-and Middle-Income Countries, 1995 .......................................... 102 Figure E. 15: Persons per Hospital Bed, Selected Asian Countries, 1995 ......................... 108 Figure E. 16: Total Public Spending on Health per Capita in the Provinces and Provincial GDP per Capital, 1998 Public and Private Health Spending per Capita, by Province, 1998 ............................................................ 109 Figure E. 17: Public and Private Health Spendingper Capita, by Province, 1998 ........... 109 Figure E. 18: Percentage Annual Increase in Real Public Spending on Health per Capita across Provinces, 1991-98, as Related to the Level of Real Public Spending on Health per Capita in 1991 ........ . ................................... 110 Figure E. 19: User Fee Collections, 1991-98 ......................................11...................... I Figure E.20: Enrollment in Health Insurance, 1993-98 .................................................... 113 Figure E.21: Share of Insurance and User Fees in the Financing of Total Public Expenditure Health, 1992-98 ............................................................ 114 Figure E. 22: Share of Insurance and User Fees in the Financing of Total Public Expenditure Health, 1992-98 ............................................................. 115 Figure E.23: Health Insurance Coverage by per Capita Expenditure Quintile, 1998 ...... 115 Figure F.1: Evolution of Higher Education Revenues ....................................................... 148 Figure F.2: Teaching Staff Utilisation in Higher Education ......................... ................... 154 Figure F. 2.1. Evolution of Higher Education Revenues .................................................... 168 Figure F.2.2: Revenues per Student in Higher ............................................................. 169 iv Figure F.2.3: Expenditures per Student in Higher Education ...................... ..................... 170 Figure F.2.4: Teaching Staff Utilisation in Higher Education .......................................... 170 Figure F.2. 5: Staff Sudent Ratios ............................................................ 171 Figure F.2.6: Average Expenditures per FTE Student by Size of Institution .......... ........... 171 Figure F.2.7: Composition of lnstitutions 'Revenues, 1998 .............................................. 172 DIAGRAMS ............................................................................................. Diagram F. 1: Key Issues Identified in Earlier Analysis .................................................... 130 Diagram F. 2: An Example of a Calculation Flow for a per Student Based Allocation ..... 141 Diagram F. 3: Public and Private Average Expenditures on Primary Education .............. 147 APPENDIX TABLES................................................................................... Appendix Table B.1: Revenue assignment in Vietnam ......................................................... 36 Appendix Table B.2: Expenditure assignment in Vietnam ................................................... 38 Appendix Table B.3: Budget transfer norms ............................................................ 40 Appendix Table B. 4: Budget allocation norms ........................................................... . 41 Appendix Table C. 1: Fees and Contributions Levied in Six Study Communes .................... 52 Appendix Table D. l: Gross Value ofAgriculiural Production at Constant 1994 Prices, 1994-97 ............................................................ 80 Appendix Table D.2: Public Investment Program Allocation 1996-2000 to the Agricultural Sector ............................................................. 81 Appendix Table D. 3. Subsidies and Current Transfers in 1998 ........................................... 82 Appendix Table D. 4: Distribution ofAgricultural State Expenditure at the Central and Local Levels ............................................................ 83 Appendix Table D. 5. Source of Funding for Selected Research Institutes ........................... 84 Appendix Table D. 6: Agricultural Extension Expenditures for Selected Provinces ............ 85 Appendix Table D. 7: 1999 Budget Information for Bac Hung Hai IMC (Controlled by MARD) ............................................................ 86 TABLES................................................................................................ Tables A.: Sectoral Composition of the PIP ............................................................ 10 Tables A,2: Total planned and actual investmentfor 5 years 1996-2000 ............................ 11 Tables B. 1: Expenditure Decentralization in Vietnam ........................................................ 22 Tables B.2: Regression Equations, 1998 ............................................................ 24 Tables B.3: Excess of Actual Expenditure over Planned (Budgeted) Expenditures ............. 26 Tables B.4: Excess ofActual Expenditure over Planned (Budgeted) Expenditures ............. 28 Tables B.5: Local Expenditures on Different Services, 1998 (Capital and Recurrent) ....... 30 Tables B.6: Income elasticity of revenues and expenditures ................................................ 31 Tables B.7: Distribution of public and private investments in provinces 1996-99 ............... 32 Tables B.8: Composition of tax Revenue in Vietnam, 1998 ................................................. 34 Tables C. 1: Basis for Calculating Fees and Contributions in Six Communes ..................... 45 Tables C.2: Examples of Differences in Fees and Contributions Between Areas ............... 45 Tables C. 3: Percentage of commune budget raised through fees and contributions ........... 47 Tables C.4: Annual (1999) Taxes, Fees and Contributions paid by seven households in Nhan Loc communes, Can Loc District, Ha Tinh Province . ...................... 48 Tables C. 5: Other Contributions to Schools (in VND per pupil/year) ................................. 50 Tables D. 1: Public Expenditures in Agriculture . ........................................................... 59 Tables D.2: Average growth of agricultural sector in various countries ............................ 59 Tables D.3: Capital and Current Expenditures by Agricultural Sub-sector, 1997- 1998 ............................................................. 60 Tables D. 4: State Budgetfor agricultural research (in current VND million) .................... 63 Tables D. 5: Shares in Total Expenditures and Research Expenditures ............................... 64 Tables D. 6: The Public Agricultural Extension Service ....................................................... 65 v Tables D. 7: MARD Expenditures on Agricultural Extension ..........................................,.,.66 Tables D. 8: Extension Expenditures by Agricultural ......................................... 66 Tables D. 9: Public Expenditures on Irrigation .69 Tables D. 10. Number of SOEs in Agriculture and Forestry ..................................... .......... 72 Tables D. I 1: Budget Support to SOEs .73 Tables D. 12: Information on Agricultural SOEs Owned by the Central .74 Tables E. 1: Distribution of Health Service Contacts by Providers and per Capita Expenditure Quintiles, 1998 .94 Tables E. 2: Distribution of Public Health Spending by Source and by Economic Type, 1991-98 ......................................................... 103 Tables E. 3: Functional Composition of Public Expenditure on Health, by Financing Source, 1991-97 ......................................................... 104 Tables E.4: Ministry of Health Expenditure on Health Programs, 1993-97 .106 Tables E.5. Recurrent Health Sector Expenditures, 1991-98 .107 Tables E. 6: Percent of Users Who Reported no Out-of-Pocket Payments for a Visit to a Government Health Facility, 1998 ..........................................................,.112 Tables E 7: Service Use and Health Expenditures of Health Insurance Enrollees, 1993-97 .116 Tables F 1: Structure of Education Provision .123 Tables F.2. Enrolment Trends . 124 Tables F.33: Gross and Net Enrolment Ratios .125 Tables F.4.- Gender Balance of School Enrolments .125 Tables F. 5: Regional Variations in (net) Enrolment Ratios .126 Tables F. 6: Growth of Public Expenditure on Education .................................................. 131 Tables F. 7: International Comparisons of Education Expenditure .131 Tables F. 8. Education Expenditure and Years of Study .132 Tables F. 9. Subsectoral Shares of Education Expenditure ................................................ 135 Tables F. 10: Input Cost Mix .136 Tables F.11: Teachers 'Salaries as a Multiple ofper Capita Output ............................... 136 Tables F. 12. Regional Differences in Average School Expenditure .142 Tables F. 13: Regional Shares of Expenditures and Enrolments .142 Tables F. 14: Net Enrolment Ratios by Quintile .143 Tables F. 15: Lorenz Curves for Education .143 Tables F. 16: Regional Benefits From Education Spending .144 Tables F. 17: Benefits from Public Expenditure by Region (Primary Education) 145 Tables F. 18: Composition of Private Expenses, by Level (% of Total Private Costs) . 145 Tables F. 19: Shares of Expenditure in Total Expenditure .146 Tables F20: Exemptions for Charges for School Education .148 Tables F.21: PTRs in Selected Countries .152 Tables F.22: Pupil Teacher Ratios and Class Sizes .152 Tables F.23: Raising Pupil Teacher Ratios .153 Tables F. 24. Repetition Ratios .159 Tables F.25. Transition and Completion Rates .160 Tables F.26: Examination Pass Rate .160 Tables F. 27: Recent Attainment Test in Mathematics and Language .161 Tables G. 1: State Central and Local Transport Expenditure .182 Tables G. 2: State Transport Expenditure by Transport Mode, Locality and Expenditure Category .183 Tables G. 3: Public Investment Program 1996-2000 vs. Actual 1996-1998 .189 Tables G. 4: Expenditure by the Ministry of Transport 1994-1999 .190 vi Currency Equivalents CURRENCY UNIT = DONG US$ = 14459 DONG (November 2000) Government Fiscal Year January 1 to December 31 Abbreviations ADB Asian Development Bank AFTA Asian Free Trade Area AMC Asset Management Company BOT Build - Operation - Transfer CG Consultative Group DANIDA Danish International Development Agency DFID Department for International Development FAO Food and Agriculture Organization FDI Foreign Direct Investment FTE Full Time Equivalent GCOP Government Committee for Organization and Personnel GDP Gross Domestic Product GER Gross Enrolment Ratio GFS Government Financial Statistics GNP Gross National Product GOV Government of Vietnam GSO General Statistics Office GTD General Taxation Department HCMC Hochiminh City HEPR Hunger Eradication and Poverty Reduction IMCs Irrigation Management Companies IMF International Monetary Fund IUD Intrauterine Device JICA Japan International Cooperation Agency LSMS Living Standard Measurement Survey LSS Lower secondary MARD Ministry of Agriculture and Rural Development MCTPC Ministry of Communication, Transport, Post and Construction MICT Manila International Container Terminal MIS Management Information System MOF Ministry of Finance MOH Ministry of Health MOLISA Ministry of Labor, Invalids and Social Affairs MOT Ministry of Transport MOTE Ministry of Education and Training MPI Ministry of Planning and Investment MTEF Medium term expenditure framework MVUC Motor Vehicle User Charge NCFAW National Committee For the Advancement of Women NCPFP National Committee of Population and Family Planning NER Net Enrolment Ratio NFEP National Fund for Employment Promotion NGO Non-government Organization NPL Non-Performing Loans NTPE National Target Program on Employment O&M Operations and Maintenance ODA Official Development Assistant PER Public Expenditure Review PHP Phillipine Pesos PIP Public Investment Program PPA Phillipine Ports Authority PPC Province People's Committee PTR Pupil Teacher Ratio SBV State Bank of Vietnam SDI Strategic Development Institute SEC Securities and Exchange Commission SMEs Small and medium-scale private enterprises SOE State-Owned Enterprise SSR Staff Student Ratio T.1B Tuberculosis UNDP United Nations Development Programme UNICEF United Nation Children's Fund USS Upper secondary VAT Value Added Tax VHI Vietnam Health Insurance VHIA Vietnam Health Insurance Authority VHSR Vietnam Health Sector Review VICT Vietnam International Container Terminal VITRANSS Vietnam Transport Sector Study VLSS Vietnam Living Standards Survey VR Vietnam Railway VRA Vietnam Road Administration VSI Vietnam Social Insurance VTET Vocational Technical Education and Training VWU Vietnam Women's Union Annex A - The Public Investment Program (PIP) ANNEX A - THE PUBLIC INVESTMENT PROGRAM (PIP) Introduction 1. Investment spending is a crucial element in public expenditure and investment programming is best integrated into a comprehensive framework of public expenditure programming. This is not only because investment spending is a significant component in total public expenditure but also because of the inter-relationships between recurrent and capital spending. Capital spending is an important determinant of future recurrent spending requirements, while capital spending can only generate the required outputs if recurrent funds are available to utilize the additional capacity. 2. In fact, for many expenditures the distinction between investment and recurrent spending is recognized as unclear and possibly misleading. Thus recurrent spending on education and health is just as much an investment in human capital as spending on school buildings and hospitals. In some cases, items treated as recurrent expenditures, such as routine road maintenance, are efficient substitutes for future investment on road rehabilitation. 3. In Vietnam a clear distinction is made between recurrent and capital spending. There is a provision that the rate of growth in recurrent spending should not rise above the rate of growth in investment spending, based on the view that raising the rate of public capital accumulation is a key growth objective. 4. However, the Vietnamese authorities recognize the need for effective coordination between investment and recurrent spending programming, as stated clearly in the attached memorandum from the MPI. It is difficult to judge how far that principle is translated into practice in the consultations between the MPI and MOF. This is an area of work which will require ongoing attention by the two ministries, particularly in the context of the formulation of the coming PIP. The Five Year Plan, the Public Investment Program and the Annual Economic and Social Plan 5. The key documents through which MPI co-ordinates investment programs are the Five Year Plan and the five year Public Investment Program (in essence a summary of the macroeconomic investment targets and main projects in the Plan) and the Annual Economic and Social Plan, which is the annual tool for implementing the Five Year Plan. 6. The formulation of the Five Year Plan is timed so that it can be presented to the five-yearly party congress. The MPI is responsible for formulating the Five Year and Annual Plans, and within the MPI responsibility for co-ordinating that work lies with the General Department of Economic Issues, supported by inputs from other departments of the MPI (e.g. the Development Strategy Institute provides inputs on long-term goals and the macroeconomic framework). Vietnam Public Expenditure Review 2000 - Annexes 7. The MPI is currently formulating the Five Year Plan for the period 2001-2005. Proposals are being formulated by sectoral ministries, provinces and major cities and lower level bodies in response to guide-lines issued by the MPI. Based on those proposals, a draft plan will be presented to the government in mid-2000. Following discussion and revisions, the Plan will be discussed by the National Assembly and presented to the Party Congress early next year. 8. While the Five Year Plan is a serious document of the Party and Government, with economic and social performance being reviewed in light of the adopted targets and strategy, the five year investment program is not "carved in stone". There is flexibility during the Plan period to adjust the investment program in light of economic conditions, availability of funds and other changes. The 1996-2000 PIP 9. The 1996-2000 PIP was the first such effort in Vietnam, and was largely derived from the established procedures for the formulation of the five year plan. It was completed to meet a condition of an ESAF loan. In many ways it was a preliminary effort to introduce more systematic appraisal and programming techniques into the investment decision-making process. 10. The 1996-2000 PIP made a number of positive contributions: * It provided a broad overview of the government's medium-term intentions regarding public investment, particularly in relation to large projects. * 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 i by setting overall investment targets in relation to likely resource availability; ii by making a broad estimate of the capital requirements to achieve growth targets; and iii at least in principle, by relating investment projects to sectoral strategies. 11. One important aspect of the 1996-2000 PIP was the effort to further clarify the respective roles of the government, foreign investors and households in implementing investment programs. 12. The 1996-2000 PIP was a transitional exercise that left a number of issues unresolved. As a first attempt at producing a PIP, it represented a new experience for many officials. Current MPI assessment suggests that the capacity to appraise and program investment activities remains weak at all levels, with this weakness reflecting the limited technical capabilities of staff and pressures from the sectors and provinces to accept projects. 13. One innovation in the PIP involved an attempt to clarify the appropriate use of differing modes of State funding: non-reimbursable funding, reimbursable State credits at varying interest rates (more concessional rates applying to priority areas and sectors) and 2 Annex A - The Public Investment Program (PIP) State investment in joint ventures. The basic principle adopted was that the mode of funding should be related to the potential of the project to generate revenues. The obligation to service the debt on external borrowing was covered by provisions for on- lending and repayment of State credits. 14. The targets for the various sectors in the PIP were painted with a broad brush. In practice data available were not sufficient to assess current investment activities of the domestic private and household sectors, and there was certainly no firm basis for projection. The estimates for private and household investment in the PIP were therefore no more than indicative. 15. Nevertheless, despite the need to adjust targets mid-plan in response to the 1997 regional economic crisis and climatic shocks, the aggregate outcome for the 1996/2000 period will not be too out of line with program targets. Actual PIP spending of VND. 214 thousand billion (estimate March 2000) compares to a target of VND. 239 thousand billion (1995 constant prices) in the PIP. 16. One reason that projected investment levels were not achieved in the first half of the plan period was that although foreign assistance was in line with expectations, there were severe constraints in short-term absorptive capacity, partly resulting from the need for the Vietnamese to learn required disbursement procedures, and in some cases for donors to adjust to Vietnamese conditions. Achievement of the overall investment targets also was predicated on a leading role being played by private foreign investment. Initial results (e.g. licenses issued) justified that projection. However by 1998 there were a number of negative factors affecting foreign private investment. 17. Interestingly, program targets tended to overestimate the relative importance of external as compared with domestic funding. ODA played a lesser role than expected, largely because of long delays in implementation of externally funded projects. 18. A particular weakness of the financing scheme for overall investment projections included in the PIP was the neglect of the potential for mobilizing domestic non-State investment, from households and small businesses. This was partly because of the lack of concrete data about such activity, but may also have reflected unclarity and unstability on policy on encouraging of domestic private investment. However, a number of factors now make such investment increasingly important, including a greater role for a non-state sector and a greater emphasis on smaller scale rural investment. 19. The MPI expresses some disquiet that household savings are channeled to activities that do not reflect plan priorities (e.g. housing and small trading) and are not available for the funding of larger scale economic activities. However, the degree to which small-scale household investment contributes to production should not be under- estimated, given the fast growth in small and medium scale enterprises. The possibility of tapping household savings to fund larger scale investment will depend on the development of financial markets, something which is proceeding rather slowly. Some lessons from this PIP * during the development of the PIP, training was provided to officials on modern methods of project appraisal (initially mainly for officials at the national level). Since the completion of the PIP, training has been extended to the provincial level. UNDP's 3 Vietnam Public Expenditure Review 2000 - Annexes in-depth evaluation of the training efforts it supported suggests that insufficient progress was made in transferring the required skills. * weaknesses in capacity at the time of formulation of the PIP meant that many projects included in the PIP were not subject to systematic appraisal. It is unclear how far the guidelines issued by the MPI and the training efforts made in recent years will increase the application of systematic appraisal techniques in the formulation of the new Five Year Plan. * the composition of the 1996-2000 PIP demonstrated some weaknesses in the govermment planning process. The sectoral allocation of investment spending for the 1996-2000 period did not adequately reflect Government stated priorities and was too heavily weighted towards investment in capital intensive industrial projects orientated towards the domestic market. This inconsistency became more evident when, during plan implementation, the goals of the Party and Government began to give even greater prominence to the need to develop rural areas, to generate more employment, reduce poverty and to promote exports. Some of these weaknesses are recognized in the MPI's own assessment of the 1996-2000 PIP. * while some effort was made to analyze projects in relation to sectoral objectives, this work was not reflected adequately in investment decisions. In particular, too little attention was given to the likely demand for the outputs expected to result from industrial investment. Plans to invest in some industries (e.g. cement, sugar, coal) were over-ambitious in light of market limitations, and the resulting excess capacity in those sectors has created difficult problems for the economy and the government • nevertheless, many of the investment decisions in the 1996-2000 Plan were sound. To some degree this was because in some key sectors the choices were relatively straightforward. For example in the transport sector the main projects related to the obvious need to rehabilitate and upgrade the major communication system, including the trunk roads and main ports. Similar projects were undertaken in relation to the main irrigation and flood control systems, all of which suffered from the effects of the war and the subsequent embargo. As these programs are completed, investment choices may become more complex and the risks of mistakes greater. * despite the implementation bottlenecks which have held up the implementation of aid programs, the list of completed projects is impressive. The Mid-term revision 20. During the 1996-2000 Plan period substantial changes were made to the PIP in 1998 in light of the regional economic crisis and the need to solve some problems which had been identified during the implementation of the PIP. 21. Key mid-term revisions of the PIP were as follows: ' In fairness to the government planners, it should be noted that the record of private investors, who should have been attuned to the need for realistic assessment of market prospects, was no better. Over investment in hotels and offices led to a glut of accommodation and private investment in some industries producing for the domestic market (notably automobiles) was totally out of line with any plausible estimate of the likely market. 4 Annex A - The Public Investment Program (PIP) a. Implementation was tightened to concentrate funding on those large projects which could be completed within the five years. As a results, the number of projects was significantly reduced. b. Authority to approve and allocate funding to B and C group projects was 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 have their "fair share" to remedy this problem, each C group project was required to be 70% completed within one year and the entire project within two years. c. There was an increased emphasis on domestic sources of investment funds. 22. The adjustment involved difficult choices regarding which projects should continue and which should be stopped. The adjustment also involved the problem of retrenching the labour force involved in the discontinued projects. 23. The MPI estimates that 95 percent of the revised PIP will be implemented by the end of the plan period, but with a higher proportion (55 percent) of total investment being funded domestically than in the original PIP projections. Timing of the Plan and the financing the investment program 24. The National Plan and the PIP are presented in terms of a five year program, corresponding to the established five year planning period. Five years is a reasonable time period for medium term programming-long enough to implement large and complex projects, but short enough to make broad judgements about likely economic conditions and resource availability. 25. However, any five year program is based on economic projections and assumptions which can quickly change. Key factors that may affect the feasibility of funding investment include: * Changes in the international economic environment (notably in the recent period, the East Asian economic crisis). * Other exogenous shocks, such as the natural disasters Vietnam has recently experienced. * Uncertainties on the availability of external funding. * Delayed program implementation (e.g., resulting from the difficulties in implementing aid). 26. One way of handling such financial uncertainty is through a Rolling Plan, involving the production of a revised medium term (say three years) program each year, modifying the previous year's plan in the light of experience and changed economic circumstances. Such an approach also recognizes that economic activity is continuous-it 5 Vietnam Public Expenditure Review 2000 - Annexes does not start at the beginning of a five year plan period or end with its completion-and project activities spill over at the beginning and end of the "Plan". 27. However, National Plans cannot realistically be rolled over annually. But MPI has stated that during the coming five year period it will conduct an annual review of the investment program, to adjust it in the light of experience with implementation and changing economic circumstances. PIP Preparation 28. The preparation of the PIP involves two distinct processes-the agreement about individual projects, required before the project can proceed, and the preparation of the annual economic and social plan by the MPI, for presentation to the Government and the National Assembly. 29. The procedure for approval of projects depends on size, with large (A) projects being subject to the full process of approval, while (B) and (C) projects are subject to a simplified system which delegates authority to appraise, approve and implement B and C projects to the Ministers and Chairmen of People's Committees. Large and important projects have to be reviewed by the State Council for the Evaluation of Investment Projects, a body with inter-ministerial representation which submits projects to the Prime Minister for final approval. 30. The aggregate five year investment "target" set out in the PIP and the Five Year Plan should be seen as essentially indicative, suggesting an order of magnitude. Proposed implementation (and therefore achievement of the target) should be set annually, in the light of the evidence of likely resources and implementation progress. Formulation of the Annual Economic and Social Plan allows MPI the opportunity to adjust the overall investment program. As mentioned above it is now the intention of MPI to conduct an annual review and updating of the PIP. Technical appraisal and political choice 31. The MPI identifies that staff capability in relation to the "methodology" of investment appraisal is limited and there is a need to provide them with more knowledge and skills in the formal techniques of project appraisal. However international experience suggests that most project failures do not result from weaknesses in the formal methodology (or criteria) applied to appraisal-although they can be important. Failures more typically result from unrealistic forecasts of the data to which the criteria are applied. This presents the greatest challenge to the MPI and government decision-makers. It is extremely difficult for those working in central agencies to assess the empirical validity of data supplied by agencies promoting projects, who are often interested parties. 32. MPI staff should be encouraged to undertake field assessments whenever possible. Infornation should be sought from a wide range of knowledgeable people in the industry in question, in the local community and from those directly involved in the project, and used in the project appraisal. 33. In the face of weak professional capacity, professional assessment can easily be displaced by an entirely political process. Obviously, political judgements have a crucial 6 Annex A - The Public Investment Program (PIP) role to play (e.g. in such issues as giving preferences given to backward areas or disadvantaged groups, or the emphasis placed on poverty alleviation). However, those political judgements should be made in the light of the best available factual evidence and economic appraisal. The relationship between public investment and recurrent expenditure 34. One of the most common weaknesses in public investment programming relates to the connection between investment expenditures and the recurrent budget. The importance of this connection and the need to co-ordinate the investment and recurrent budgets is recognized by the government. The potential risks are at the macroeconomic, sectoral and project levels. 35. At the macro-level, the failure to take proper account of the recurrent cost implications of investment can result in strain on the national budget, while the costs of external borrowing can create strains on the balance of payments. These major macroeconomic risks have not yet emerged in Vietnam, although under-funding of some maintenance expenditures has imposed costs on the economy and created some build up of expenditure pressures, and the situation could change quite fast, as external borrowing increases, and if there is a period of decline in foreign earnings growth. 36. Sustainability issues emerge at the sectoral and project levels, as well as in the context of the overall budget. When the recurrent implications of investment are neglected * the operation of programs and projects after the completion of the investment stage may be impaired because of the scarcity of funds to provide the necessary inputs; and * it may not be possible to fund the repair and maintenance of the capital created, so that facilities quickly deteriorate. 37. One view expressed by the MPI is that this is not yet a big issue for Vietnam in that the level of public investment is still only 20-25 percent of total state budget spending. This will probably be true when the growth of GDP is buoyant, in the 8-10 percent per annum range experienced beti5re 1998-1999 . However, the problem could easily emerge if growth were to fall to the 4-5 percent per annum range, particularly in light of the decline in the proportion of GDP captured in government revenue. 38. But external funding creates the possibility of breaking the link between domestic resource availability and public investment. There is little evidence that appraisal by donors takes appropriate account of the potential impact of aid programs on the recurrent budget and neither the GOV nor the donors have addressed the issue of financial sustainability of the aid program, apart the sustainability of debt servicing. The only donor focus so far has been on some specific aspects of the inter-relationship between investment and recurrent funding issue, e.g. in work done by the World Bank on repair and maintenance requirements in the transport sector. 39. The recurrent cost sustainability issue will also need to be addressed at the local level, as donors seek to increase their support for decentralized investment, particularly in the poorer communities, which have weak resource and tax bases. 7 Vietnam Public Expenditure Review 2000 - Annexes 40. The tendency of those responsible for project formulation to be optimistic about the recurrent cost issue is understandable. Added to this is the fact that the external agency is often responsible for the project only until "completion", and does not include the operational phase, when recurrent funding constraints come more clearly into play. Decentralized investment decision-making 41. An element of decentralized decision making was introduced into the PIP by leaving control over allocations to smaller (B and C) projects to line ministries, other agencies of government, and provincial authorities, based on guidelines issued by the MPI. The increased GOV emphasis on the development of rural development and donors' interest in supporting such development is likely to increase the importance of investrnent decision making at the provincial and lower levels of government. 42. Decentralization brings with it issues of quality control and co-ordination. Decisions about the investment program at the provincial level are guided by an institutional structure, which mirrors arrangements at the national level. The Chairman of the Provincial People's Committee acts as the head administrator. The Department of Planning and Investment co-ordinates the socio-economic development plan and the PIP. The sectoral departments are responsible for provincial level activities in areas corresponding to the national sectoral ministries. The Provincial People's Council acts as the elected body supervising the provincial program. Provincial authorities also manages SOEs which operate separately from national SOEs and undertake investment expenditure. 43. In principle, the MPI provides direction to the work of the provincial DPIs, and the sectoral ministries provide guidance to the provincial sectoral departments. However, provincial level officials are recruited by the provincial authorities, and in most cases pursue careers within the province. Although in principle they have a responsibility vertically to their national parent ministries, in practice their primary responsibility and accountability is horizontal, to the provincial administration. There is therefore a high degree of provincial autonomy in practice. 44. For the PIP, the provinces prepare and submit five year plan proposals to the MPI for processing. Based on the projected resource envelope, MPI consolidates these proposals into the overall investment plan and summits them to higher authorities for review and approval. However, given the huge amount of funding and large number of projects requested by provinces, consolidation of the proposals has been quite difficult and the final outcome tends to be a compromise. As a result, the quality control of projects may not be adequate 45. Although in principle provincial investment projects should be under the supervision of the relevant central sectoral ministries, that supervision is often weak. For example, MPI informs MARD of provincial level agricultural projects, but MARD has does not control the capital budget for those projects, which is allocated by the provincial authorities. Conversely, provincial authorities sometimes show little interest in national level projects being implemented within their area. 46. MPI attempts to assert some control over the B and C projects by limiting their number, by not allowing sub-division into component parts (a potential means of 8 Annex A - The Public Investment Program (PIP) avoiding the controls exercised over large projects on the A list), and by restricting the time period over which projects must be implemented (4 years for B projects and 2 years for C projects). 47. While a high degree of decentralization of part of the PIP has the potential to provide the flexibility to respond to local needs and to encourage local initiative it also has some unwelcome consequences. The capacity to identify and appraise projects at the provincial level is highly variable. For example in some provinces observation suggests that there has been a diversion of scarce funds to non-productive activities such as renovation or new construction of provincial and district headquarters. 48. There are also problems of co-ordination, with provinces inefficiently using resources by seeking to proliferate activities. Thus in the transport sector most coastal provinces want their own major port, and some of the excess capacity in State industry has resulted from the proliferation of industrial projects at the provincial level. 49. There are also difficulties of co-ordination of activities which cross provincial boundaries. This can relate to situations where a province requires access to resources (e.g. water) from a neighboring province, and also where infrastructure investment decisions at the provincial level have regional implications requiring co-operation between neighboring provinces. MPI comments on the PIP process PIP 1996-2000 50. The five-year Public Investment Program 1996-2000 was formulated and implemented as follows: Concepts 51. The State investment program includes the investment projects formulated in the State plan and financed by the State budget, State credits and capital from SOEs The Public Investment Program is financed from the following sources: * from local revenues of the State budget, used as grants channeled to projects that have little or no cost recovery. For projects that can recover part of their costs, financing from the State budget is complementary. * State investment credit at preferential interest rates is invested in projects in high priority sectors, especially in SOE projects in key sectors. The organizations that borrow State credit are responsible for repayment of principal and interest. This type of financing accounts for a high proportion of total State investment. * capital generated by SOE's, such as depreciation funds, after tax income, and unused land and housing; plus contributed capital in joint ventures or co-operative ventures with international or local partners. * ODA channeled to projects supported by donors; part of ODA borrowing is included in the State credit fund. 9 Vietnam Public Expenditure Review 2000 - Annexes 52. The mode of State funding of public investment projects is based on the socio- economic characteristics of projects. One or more mode of financing can support a project. Projects that use credit funds are appraised to assess their capacity for repayment of the funds, especially those that use ODA loans. The content of Public Investment Program 1996-2000 53. In June 1996 the Prime Minister approved Vietnam's five-year Public Investment Program 1996-2000. The preparation of this first PIP prepared by Vietnam was supported by funds from UJNDP and the Government of Netherlands, participation of World Bank consultants, the co-operation of relevant departments in MPI, central ministries and some localities. 54. The Public Investment Program 1996-2000 was based on the development objectives of the five-year plan 1996-2000 passed at the VIII Party Congress. These objectives included an annual GDP growth rate of 9-10 percent, industrial growth of 14- 15 percent, agricultural growth of 4.5-5 percent and service sector growth of 12-13 percent. The objectives were based both on the capacity and urgent need to improve all living standards of all groups of people and prepare the ground for the country to compete and integrate into the international economy. 55. To achieve the growth goals, the total investment requirement for the 5 years (at 1995 constant price) was estimated at US$ 41-42 billion, twice the level of the previous five-year period. The investment/GDP ratio of 30 percent in 2000 also increased from 25 percent in 1995. The Public Investment Program accounted for 38 percent or US$ 15.6 billion (at 1995 constant price) of the total investment target and focused on the following main sectors: Table A.1: Sectoral Composition of the PIP Sectors % in total % of State Investment contribution Transport and water supply infrastructure 24.2 22.1 Energy and industry 16.7 1.9 Post and telecommunication 7.1 0.9 Science, technology and environment 1.7 3.1 Education and training 4.2 7.5 Health and social capital 3.7 6.8 Culture and information 2.8 5.0 Others 4.1 7.2 Source. MPI 56. Since mid-1997 Vietnam has had to cope with a difficult economic environment. Weaknesses in the economy, the regional financial crisis and natural calamities (drought and floods) affected production and reduced growth. 57. The macroeconomic performance after 4 years of plan implementation and based on the forecast of the performance for 2000 is expected to be: GDP growth on average 6.7% Industrial growth 12.2% Agricultural growth 4.9% Service growth 7.0% 10 Annex A - The Public Investment Program (PIP) Table A.2: Total planned and actual investment for 5 years 1996-2000 (billion dong at 1995 constant price). Planned Five-year Disbursed Disbursed Disbursed Disbursed Committed Likely 1996-2000 1996 1997 1998 1999 2000 outcome Total 460,000 73,050 83,630 76,480 77,920 88,450 399,530 Domestic 232,000 42,513 46,713 44,007 48,516 56,558 238,307 resources External 227,500 30,537 36,917 32,473 29,404 31,892 161,223 resources PIP 239,200 33,030 40,200 40,930 48,000 51,920 214,080 Domestic 23,373 29,433 27,737 32,766 35,128 resources External 9,657 10,767 13,193 15,234 16,792 resources 1. State budget 96,800 15,180 17,780 16,420 19,500 18,480 87,360 fund Of which, 4,440 4,773 5,328 5,550 5,938 ODA 2. State credit 76,400 7,640 10,960 11,740 14,250 17,620 62,210 Of wvhich, on- 3,752 4,329 5,977 7,675 8,998 lending ODA 3. SOE own 66,000 10,210 11,460 12,770 14,250 15,820 64,510 funds Private 76.500 19,140 17,260 16,270 15,750 21,430 89,850 domestic investment Foreign Direct 144.300 20,880 26,150 19,280 14,170 15,100 95,580 Investment 58. Total investment spending for 5 years (1996-2000) at 1995 constant price is 400 thousand billion dong, equivalent to US$ 36 billion, nearly twice the period 1991-1995. During the final years of the period 1991-2000, the policy of mobilising internal resources in line with the Party 4th Plenum was reflected by domestic resources accounting for over 60 percent of the total resources. 59. The allocation of investment by region is estimated to have been as follows (percent): Northern mountainous area 7.6 Red River delta 26.0 North central 9.0 Central coastal area 11.7 Central highland 4.7 South east region 27.6 Mekong delta 15.6 Source: MPI 60. A number of the objectives of the investment program have been achieved, e.g. upgrading national highways, irrigation, flood control in Mekong delta, rural electrification, substantial investments in profitable enterprises and achievements of educational, cultural and social objectives. 11 Vietnam Public Expenditure Review 2000 - Annexes 61. The Sectoral allocation of five-year estimates is as follows: % of PIP % of Capital Budget Expenditure Agriculture, forestry and 18 45 acquaculture. Industry 45 4 Transport and communications 26 40 Education and Training 4 10 Cultural, health, social affairs and 6 17 sports Science, technology and 1 4 environment Source. MPI 62. Poverty reduction objectives were integrated into other programs such as employment creation, afforestation, cultural, social, education, and health care programs implemented in poor areas. Poor communes receive funding for six types of infrastructure: vehicle-accessible roads to the commune center, clean water supply, power supply, primary and junior secondary schools, health clinics, commune or inter-commune market places. The mid-term adjustment 63. In mid 1997, the Vietnamese economy was faced with the impact of the regional crisis, and by unfavorable climatic conditions, with prolonged storms and floods. In particular, there was a decline in new foreign investment proposals, so that it became evident that more reliance would have to be placed on domestic funding of investment. The regional crisis also resulted in heightened competition in international markets. A substantial adjustment in the PIP was required. 64. The table above indicates that during the three years 1998-2000, the total spending was about 60 percent of total for the five year period, of which the domestic fund accounted for about 55-60 percent, i.e. the level of investment was broadly maintained, but with an increased emphasis on domestic funding. 65. The adjustments in the program reflected the following priorities: * An emphasis on agricultural development, national food security and poverty reduction through rural economic development. * Priority to processing industries, aqua-culture, export production, and consumer goods. Electronics and software services received some attention. The focus shifted to labour-intensive, instead of capital-intensive industries. * Development of such services as telecommunication, tourism, transportation, trade, science and technology, finance, banking, insurance, serving both the domestic market and the international market. 12 Annex A - The Public Investment Program (PIP) * In infrastructure priority given to critical projects that serve the economy as the whole and expand trade, and to key infrastructure in rural, mountainous, and remote areas. * Regional emphasis continued to be place on the three economic growth zones, plus continuing attention to the development of mountainous areas, border areas, islands, ethnic minorities, and former revolutionary bases. * State budget and ODA allocation for projects under central government control sought to achieve an even regional allocation in building infrastructure. Local budget funding aimed to allocate about 30 percent to the three economic growth zones and 70 percent to other regions. Successes and limitations of the investment program. 66. While the 1995/2000 period saw solid achievements and the main thrust of the investment program was sound, there were weaknesses which need to be corrected. The limited funds available should have been used in a more focused and effective manner. In some areas, resources were scattered over too many projects, with a loss of effectiveness, especially in the group C with projects. Often the quality of project preparation was inadequate and feasibility poor, delays in appraisal and approval, as well as in bidding and bid evaluation processes badly affecting the progress of project implementation. The government has issued new Decrees, Resolutions and Directives to improve the investment process and is reviewing the investment program, reassessing the effectiveness and urgency of projects. 67. With respect to the State budget, although efforts have been made to generate savings to fund the investment program, revenues are limited and recurrent expenditure requirements great. Therefore, investment funding from the budget is still small, equal to only about 20-21 percent of aggregate investment funding, which has added to the difficulty of completing investment projects. 68. The state credit fund is also limited, usually meeting 50 to 60 percent of the demand. State credits are intended to support focal areas and key industries with favorable repayment prospects. However the preparation of investment projects is still poor. 69. Investment in state owned enterprises has continued at a high rate, but the number of enterprises earning profit is not large and therefore the mobilization of their own capital or after tax profit for further investment is still limited. 70. The role of private savings in the PIP is limited for two reasons. Private savings are still a low proportion GDP. Also, the average household has only modest capital and potential investors are unwilling to commit their capital to longer term investments in medium and larger scale projects. At the same time, policies are not attractive enough to encourage households to expand their investment in production and business activities. Much household investment is in building houses, shops and small kiosks, which is not in line with plan objectives. 71. Foreign direct investment grew at a high rate in the early years, and the sectors that received that investment continue to depend on foreign partners. The mobilization of this source of capital has become increasingly difficult due to international competition in 13 Vietnam Public Expenditure Review 2000 - Annexes the aftermath of regional monetary crisis. Though foreign direct investment has made an important contribution to economic growth, technology and transfer of modem management expertise, many foreign investment activities have been concentrated in protected industries producing import substitutes. 72. Most investment projects have been chosen on the basis of supply side considerations (e.g. availability of raw materials), without sufficient consideration of demand or of production costs or of possible competition. Sectoral and location choices have been frequently irrational, not promoting potential comparative advantage that is needed for effective integration in the international and regional economy. There has also been indecisiveness between focussing on domestic or foreign markets, between protection and enhancement of competitiveness, and between location in dynamic growth areas or poorer regions with inferior facilities. 73. The policy and incentive framework for foreign investment has been insufficiently attractive and lacks consistency. The quality of investment preparation and planning has been limited, with insufficient emphasis on high quality products and export goods and too much emphasis on import substitutes. Continuing problems in preparation of public investment projects. 74. Capacity in project formulation and appraisal remains quite weak and the implementation of the investment program is constrained by administrative bottlenecks. Despite the adjustments made in the targets for 1996-2000 program, in light of external and climatic shocks, slow implementation has meant that there will be short-falls in the program. 75. Weaknesses in the public investment process relate to three general areas: * Project preparation, project evaluation and project assessment is often problematic. * Planning and forecast work is still limited and highly subjective. In particular, many important projects have been selected without adequate analysis of market demand. Project planning often has been supported by insufficient data to provide a basis for effective decision-making. The location of projects has often been haphazard. * The ability to identify, design and supervise investment projects at the sectoral and local levels is still limited. Investment is not used to efficiently promote the comparative and competitive advantages of the country in international scale. 76. The 1996-2000 PIP represented the first effort at formulating an investment program in the context of the market-oriented economy. For many officials, the formulation of the PIP involved a new approach. The training for staff working in research and public investment program development under the sponsorship of UNDP and the WB has proved fruitful, but there is still a need for much more training, especially at the provincial and sub-provincial levels. The PIP 2001-2005 77. In order to implement the likely investment programs, Vietnam will need to mobilize an average of over thirty percent of gross domestic product for investment, of 14 Annex A - The Public Investment Program (PIP) which the PIP will account for almost forty-five per cent. This is an ambitious goal but not unattainable. Policies for to foreign in the coming years will need to be more open and attractive. 78. The following are some key issues for the formulation of Public Investment Programs (2001-2005): a. The PIP is one of the core elements of the social economic development plan. Therefore it should be widely disseminated, primarily to those involved in the planning process from the central government level to local levels. Program formulation has to involve Ministries and agencies and cities and provinces. b. The program will be will be updated each year, adding projects in accordance with the level of preparation and the availability of capital. c. Based upon the experience of the formulation of the 1996-2000 PIP, cities and provinces need assistance and supervision to contribute to formulation of the PIP. d. The core need is to have good projects. The formulation, evaluation and approval of projects is therefore very important. Training that helps staff acquire necessary expertise in preparation, formulation and approval of projects, especially at local levels is an important need. e. Given the limited funds available, there is a need to define priorities and to set criteria for the inclusion of projects in the PIP. For this purpose, it would be useful to compile a project appraisal manual to disseminate to all sectors and levels. f. The MPI wishes to set up uniform procedures for the formulation of the PIP whereby the work steps are arranged in a standardised manner and the relationship among relevant agencies, among the central government and local authorities is clearly defined. g. International assistance is therefore required for training and dissemination of methodology for investigating, forecasting, calculating and facilitating the implementation of the coming five year PIP. 15 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services ANNEX B - FISCAL DECENTRALIZATION AND EFFICIENCY AND EQUITY IN PUBLIC SERVICES i. Introduction 1. As Vietnam's economy makes the transition from a planned to a market economy, the delivery of public services should be efficient, should meet the requirements of different groups of population and respond to the requirements of different regions. Efficient functioning of a market economy calls for fiscal decentralization, to ensure efficient provision of public services. Over the years the role and functions of central and local governments in Vietnam in providing physical infrastructure and social services have evolved to meet changing requirements. However further changes are desirable, as discussed below. 2. There are several ways in which fiscal decentralization impacts on efficiency and equity in the delivery of public services. Proximity of the policy makers to the people may reduce information and transaction costs of identifying specific requirements of different regions and population groups and ensure better design of appropriate policies and programs. This helps to provide public services according to the diverse needs and requirements of different regions and population groups, and also enables harnessing of varying resource endowments of different regions. The "decentralization theorem" or the "principle of subsidiary" states that decentralized provision of public services enhances welfare, particularly when there are significant localized differences in preferences. Decentralized provision of public services should be preferred so long as any additional organizational costs and loss of economies of scale do not outweigh the welfare gains. 3. To the extent the system ensures linking of revenue and expenditure decisions, decentralized provision of public services improves accountability as well. In such a system, local authority is made accountable to both the taxpayers below and higher levels of government above. 4. It is necessary to examine whether recent changes in Vietnam are adequate and if not, identify the constraints and measures needed to improve the system. In this chapter an attempt is made to examine fiscal decentralization in Vietnam, analyze its equity and efficiency implications for the delivery of services and identify policy and institutional changes necessary to improve the service delivery while ensuring greater accountability. The next section provides the description of the prevailing system of intergovernmental fiscal arrangements. Efficiency and equity implications of the prevailing system are examined respectively in section III and IV. Policy and institutional changes required to meet emerging challenges are summarized in section V. ii. Fiscal Arrangements In Vietnam 5. The prevailing fiscal arrangement in Vietnam can be characterized as a de- concentrated system. Public services are provided within the unitary system of government in which centralized state authority is extended to local administration in 17 Vietnam Public Expenditure Review 2000 - Annexes provinces, districts and communes. The center assigns the responsibility of providing certain public services and of implementing policies and programs to subordinate local levels under its supervision and control. Local participation is relied upon to ensure that public service provision responds to local needs. In this 'double subordination' system, the local executive is accountable both to the higher level government and to the elected bodies at the same level. 6. An essential complement of the de-concentrated system in Vietnam is the "unified" budget. The annual budget approved by National Assembly includes revenues and expenditures of both central and local governments. All policy decisions relating to revenues and expenditures are taken at the national level, to be implemented uniformly throughout the country through the line agencies at the center and local authorities. The local administration in Vietnam consists of 61 provinces constituted by 598 districts, which in turn are made up of 10.500 communes, wards and district towns 7. There have been significant attempts to reform fiscal arrangements to keep pace with changing requirements. With the implementation of the Budget Law in 1997, a legal system for fiscal transactions between the different levels of government in Vietnam has been instituted. It also lays down the framework for the formulation, imnplementation and monitoring of budgetary decisions. Fiscal Assignment 8. The budget law lays down the responsibilities of different agencies in the process of budget formulation, implementation, monitoring and control. The Ministry of Finance (MOF) is responsible for revenue forecasts and consolidating the recurrent expenditure estimates of all levels of government. While in principle, the MOF is required to negotiate with the line ministries and provincial finance departments to finalize the expenditure allocation to them, in practice final budgetary allocation is a "top-down" decision. 9. The starting point in budgetary formulation is the assignment of revenues and expenditures. The Law defines expenditure responsibilities and revenue sources of different levels of government as shown in Appendix B. 1 and B.2. Expenditures are assigned to central and local levels, but the assignment follows a functional separation based on the principle of geographical spread of benefits, size of the projects and volume of spending. Operation and maintenance of central projects is undertaken by the center; local governnents undertake projects benefiting their respective jurisdictions. Education and health are shared responsibilities with a functional division of responsibility; school education is a local responsibility and the center looks after training (higher education). Large hospitals are in the central domain, provinces administer provincial hospitals, districts administer district hospitals, and communes look after the health centers. Construction and maintenance of national highways and large arterial roads is a central responsibility, provincial roads are maintained by provinces and communes look after rural roads. 10. In the case of revenues, the law specifies the taxes exclusively assigned to central and local levels and those that are shared between them. The policy determining the base and the rate structure is exercised by the center. The Customs Department collects import-export taxes and the General Taxation Department (GTD), which is a central government agency, collects all taxes through its branches in provinces and districts. 18 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services However, in respect of taxes assigned to them, the local governments can retain the entire collections in excess of the targets and even in the case of taxes assigned to Central government, the provinces are allowed to retain part of the collections in excess of targeted revenues. This provides incentive to the provinces to help in better enforcement of tax collections. Thus, provinces have a vested interest in facilitating collection of taxes by the GTD and are known even to provide bonus incentives to the tax officials to collect more than the targeted revenues. (It may also encourage under-estimation of budget revenues, as discussed later). 11. Local governments have been given powers to raise revenues from some fees, charges and tolls. Collection of fees from schools and hospitals, tolls from roads under their jurisdictions, raising voluntary contributions from the members of the community are some examples (further discussion of local fees and contributions is contained in Chapter 2 and Annex C of the PER). Provincial governments have domestic borrowing powers, and many local services like waste disposal, water and electricity supply and local transportation are provided by local SOEs, which can borrow funds from both internal and external sources. All external borrowings are subject to approval by the central government (which provides a sovereign guarantee). Budget formulation process 12. In June every year, Government issues a decree for the preparation of the budget for the ensuing year. Based on the decree, the MOF issues a budget circular detailing the projected fiscal, macroeconomic and balance of payment environment and targets. The circular also details the guidelines for the preparation of budget estimates for the ensuing year for each of the spending departments and provinces (and cities directly under the center). It also provides indicative targets of recurrent expenditure to each of the central ministries and provinces. The guidelines in the budget circular contain detailed norms for estimating expenditures both for the central and local departments (Appendix B.3). The provinces in turn send guidelines to districts and the latter, 13. At each level, the spending departments prepare their expenditure estimates based on these guidelines and norms and submit them to the finance department at that level which, after consolidation are submitted to the respective People's Committees for approval. Such approval, however, is subject to prior clearance from Standing Committees of People's Councils.. Simultaneously, the estimates are communicated to the finance department at the higher level. The process of preparing the budget extends all the way down to the commune level. The finance department at each level consolidates the draft budgets of spending agencies at that level as well as those below it. At the center, the MOF consolidates the estimates received from the line ministries and provinces. 14. The MOF has specified detailed norms used in determining transfers to the provinces - "budget transfer" norms. These are detailed in the Budget Circular 38 issued by the MOF in 1996 and set out in Appendix B.3. The norms, in most cases are based on an allocation per head of population, but adjusted on the basis of nature of terrain to take account of the cost variations in providing public services in different areas. Education and health allocations for example, are determined on the basis of population. and for administrative services, the number of employees. 19 Vietnam Public Expenditure Review 2000 - Annexes 15. The allocation of funds by provinces once they receive their funding within each of the functional heading has to be done according to "budget allocation" norms. These are laid down in the guideline 562 issued by the MOF in 1998 and set out in Appendix B.4. These are different from the budget transfer norm. In the case of education, for example, budget transfer norm is based on population whereas, the budget allocation norm is number of students. Similarly, the total allocation for health services is determined on a per capita basis whereas allocation of health expenditures to various hospitals is based on the number of hospital beds. Budget allocation norms too vary to take account of cost disabilities in mountainous and remote areas. 16. Thus expenditure estimates are based on allocative norms, which are supposed to reflect the quantity and cost of public services. However in practice expenditure allocations are finalized by the higher level of government. Although negotiations may play some part expenditure determination is essentially a top down process with the norms being used as guidelines. This process is not very transparent and in some instances many lead to arbitrariness and subjectivity. However the allocation then remains fixed for three to five years, with adjustments made each year to take account of changes in economic conditions, prices and policy. 17. MPI is responsible for the investment budget. The total figure for public investment is estimated by MPI (consistent with the envisaged growth rate of the economy) which provides allocations to individual ministries and provinces which then prioritize from among their approved projects. Investment projects classified into three groups A, B and C depending on their nature, size and strategic importance. Group A projects are large national projects, which are approved by the Prime Minister. Apart from this ministers, heads of agencies and Chairmen of Peoples Committees in the provinces and provincial cities make decisions on the smaller Group B and C projects, which are subject to a simpler evaluation and approval process. 18. Investment funds are allocated among the provinces based on socio-economic conditions, level of development of the province, its population size, number of centrally managed projects in the province, and counterpart funding required for the ODA. Although the relative bargaining strength of provinces does matter the system is predominantly "top-down", as with negotiations with MOF on recurrent expenditures. The provinces consolidate investment budgets of various provincial departments as well as those of lower level govermments (districts and communes) communicate them to the MPI. Similarly, the MPI consolidates the investment budgets of the provinces and the central government ministries. To safeguard against having too many projects and spreading resources too thinly, 70 per cent of Group C projects included in the investment plan must be completed within the first year and the entire project within two years, and for Group B, 70 percent within two years. 19. After determining revenue assignment, estimation of expenditures is critical to determining cash transfers to provinces made by MOF. Some taxes are entirely assigned to provinces and some are shared with the center. Revenues from both sources projected for each of province and are matched with its expenditure requirements. The provinces which are left with gaps after tax assignment and sharing are given cash transfers. 20 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services 20. In addition to these general expenditures, there is a separate mechanism to determine expenditures under the National Programs, as discussed later in this Annex. These cover services of national importance which are delivered locally but funded centrally. 21. Besides the estimate of expenditure, revenue assignment and cash transfers, and National Programs, the budget also contains the latest estimates of revenue and expenditure of the previous year and the finalized actuals of the year before. These are presented to the National Assembly for approval at its end of year meeting. The process of allocating central and local budgets, bringing about budget balance at each of the levels is required to be completed before December 31 to enable the budget implementation from the financial year beginning January 1. 22. After it is approved, the Government communicates the budget to the spending ministries and provinces and the Treasury to facilitate implementation. The finance departments at different levels have the powers to supervise and control the implementation of their budget as well as the levels below them. The spending units prepare the spending plan and based on this the finance departments approve the quarterly spending plans for each spending unit. The Treasury ensures that the disbursements are made according to the spending plan. 23. Based on the guidance received from the MOF, the budgetary units have to prepare their accounts at the end of the fiscal year and submit them to the higher level. The State Treasury at the relevant level certifies the accounts. The finance department at each level examines and approves the accounts of various departments at that level, consolidates them with the accounts of those below it for the People's Committee to submits to its People's Councils for ratification. The report is also sent to the finance department at the next higher level. The MOF considers and approves revenues and expenditures of central agencies and consolidates them with the local accounts for the Government to submit to the National Assembly for ratification. The state audit agency is responsible for auditing the accounts to ensure correctness and legality. Trends in expenditure decentralization 24. Expenditure decentralization as measured by the share of local administration in total expenditures has shown a steady increase from 26 percent in 1992 to over 43 percent in 1998 (Figure B.1). While this can partly be attributed to the more comprehensive coverage of local expenditures the general trend towards greater expenditure decentralization is clear. Local governments implement 43 percent of total national expenditure including about 4 percentage points on account of National Programs (Table B.1). These are essentially central programs implemented through the local governments. Thus, the expenditures falling within the discretion of local governments are just about 38 percent. The local role is particularly important in human development. Almost three-quarters of education expenditures are incurred at local levels and in health expenditures, the local share is about two thirds. The proportion for social expenditures overall is 50 percent. 25. However, in reality, the power of local governments to influence expenditure decisions is much lower than indicated by these ratios. This is because inadequate funding coupled with stagnant revenues makes it difficult to alter expenditure allocation between and within different functions and economic categories. This is compounded by 21 Vietnam Public Expenditure Review 2000 - Annexes the fact that local governments have very little independent revenue raising powers to enable them to change expenditure pattems. The problem is particularly severe for those provinces for which revenue assignment is inadequate to meet their expenditures and which therefore receive cash transfers from the center. Figure B.1 Share of Central and Local Governments in 100% l 80% 1992 1993 1994 1995 1996 1997 1998 19§§ Years Source. MOF Table B.1: Expenditure Decentralization in Vietnam Per Capita%Expenditure Per Cent of GDP Share of Local (Th. VND) ExPerCet Central Local Central Local Total (Per Shnt) Total Agriculture and 39.2 27.1 66.3 0.8 0.6 1.4 40.8 Fisheries Transport 61.9 41.3 103.2 1.3 0.9 2.2 40.0 Industry 23.0 13.3 36.2 0.5 0.3 0.8 36.6 Economic Services 124.1 81.6 205.7 2.6 1.7 4.4 39.7 Education 44.3 122.2 166.5 0.9 2.6 3.5 73.4 Health 25.2 42.8 68.0 0.5 0.9 1.4 62.9 Social Subsidies 103.2 10.6 113.8 2.2 0.2 2.4 9.3 Culture and Sports 13.2 22.6 35.8 0.3 0.5 0.8 63.1 Science and 10.9 3.0 13.9 0.2 0.1 0.3 21.7 Technology Total - Social Services 196.9 201.3 398.1 4.2 4.3 8.4 50.6 Other Expenditures 222.5 132.6 355.1 4.7 2.8 7.5 37.3 Total Expenditures 543.5 415.5 959.0 11.5 8.8 20.3 43.3 Note: a -Excluding interest payments. Sourcey Ministry of Finance 22 Annex B - Fiscal Decentralization and Efficiency and Equity in Pnblic Services III. EXPENDITURE ASSESSMENT, ALLOCATIVE AND TECHNICAL EFFICIENCY Subjectivity in Allocations 26. In the de-concentrated system prevailing in Vietnam, the method of determining expenditures is an important factor the allocative and technical efficiency of public spending. As mentioned earlier in the Vietnam budgetary system local governments estimate expenditure requirements on the basis of specified norms. In addition to spending norms, there are physical norms and staffing norms stipulated by the relevant ministries. In the case of education, for example, there are as many as forty different norms which are applied in varying degrees by different levels of government2. 27. Existence of multiple norms is only a part of the problem. Equally important is their relevance to achieving equity and efficiency. In most cases, population is taken as the principal factor, but provision is made for cost differences in the provision of services due to different terrain. Additional cost disabilities in major cities is also considered. However, the norms are used only to build the initial estimates. Ultimately expenditure estimates are finalized on the basis of negotiations, during which supplementary allocations are often added to adjust for specific local conditions that are not fully taken into account when calculating the norms The budgetary transfer norms stipulated for determining expenditures on education, health and administration are summarized in Appendix B.3. 28. The inappropriateness of some norms is illustrated by education where the budget transfer norm is based on population This provides no incentive to increase enrolments. A more appropriate norm might be the population in the relevant school going age. A per pupil norm favors regions with a higher enrolment ratio. In the case of medical expenditures, the number of hospital beds is the principal allocative norm and this tends to favor those which already have health infrastructure. Similarly, per employee norm for administrative expenditures favors regions with more employees and is biased against those with poorer administrative infrastructure, or operating more efficiently. 29. It is thus not surprising that the actual expenditures incurred by the provinces are distributed randomly and not related to any of the need factors used in the regression equations (Table 13.2). The estimated regressions show that primary and secondary education expenditures per child in the relevant age group, for example, are not related to the enrolment ratio or teacher-student ratio. It is not related to per capita GDP either. Expenditure on higher education is related only to enrolment ratio and not to per capita GDP or the teacher-student ratio. Thus, in provinces where the enrolment ratio is higher by one per cent, expenditures were higher by 0.7 percent. 30. The distribution of health expenditures among the provinces is even more difficult to explain. Actual expenditures are related neither to hospital beds nor to the number of medical personnel. They are positively related only to per capita GDP, which indicates that more affluent provinces have higher per capita expenditures with income elasticity of expenditures estimated at 0.7. 2 Peter Brooke Duong Thi Thanh Mai, Nguyen Huu Tu and Kitty Stewart, "Education Financing Study: Primary Education in Vietnam", 1999. 23 Vietnam Public Expenditure Review 2000 - Annexes 31. These illustrative examples indicate that in practice expenditure allocation to provinces is not related to any of the indicators of need taken in the regressions, nor do they show equitable regional distribution. The combination of norms and negotiations thus often results in a distribution of expenditures, which may not be fully intended, let alone desirable. Table B. 2: Regression Equations, 1998 Dep. Constant PGDP Enrol. Teacher Hospital Medical R Bar F - Variable Ratio -Student Bed! Th. Personnel Sq Statistics ratio Persons per Th. persons PE -3.3968 0.2104 0.4421 -0.330 - - 0.0929 2.742 -(2.514) (1.087) (1.755) -(1.088) SE -2.6935 0.2283 0.442 -0.1356 - - 0.079 2.458 -(2.665) (1.1395) (1.938) (0.543) TG 3.2319 -0.1306 0.726 0.0404 - - 0.8023 10.134* (6.452) -(1.689) (11.22) (1.301) HLT 8.434 0.667 - - -0.054 -0.561 0.435 13.591* (8.340) (4.656)* -(0.331) -(3.56)* PE - Primary education expenditure per child in age group 5-9. SE - Secondary education expenditure per child in the age group 10-14. TG - Training (higher education) expenditures per capita. * Significant at 1 per cent level. (The estimated regressions are in log-linear form). Lack of Maneuverability 32. Another constraining factor in achieving efficiency in public expenditure allocation is the lack of flexibility in expenditure allocations. This is particularly true of poorer provinces with low amounts of independent revenues. The local governments receive expenditure appropriation with priority already determined for various sectors and sub-sectors. In most cases, salary expenditure has the first call and it constitutes the bulk of expenditures particularly in social services. Often, after meeting salary expenditures, there may be little left for other items. Thus, in poorer localities, bulk of the expenditures is on salaries and very little may be left for medicines and equipment, without additional user charges being levied. Similarly, in the case of education, after meeting teachers' salaries very little may be left for books and teaching materials. In the event, there appear to be wide variations in the standards of services provided depending on the ability of the communes to tap their own resources, including through voluntary contributions. Inadequate Revenue Assignments 33. A key determinant of the quality of services provided at the local level is the ability of local governments to mobilize their own revenue sources. In Vietnam's revenue assignment system local governments have very little revenue raising powers. All decisions on tax rates and bases are decided by the central government. The only independent source of revenues to the provinces consist of various fees and charges, 24 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services revenues collected in excess of budget targets and voluntary contributions made by the community. This arrangement has both efficiency and equity implications. 34. In terms of efficiency first, it is felt by some provinces that in respect of taxes with immobile tax bases and a local spread, the center does not have adequate information on the tax bases and hence does not fully collect the revenue - and that local goverrnents would have a comparative advantage in collecting them Second, in the absence of significant independent revenue sources, local governments have not had the flexibility to adopt the resource allocation desired by their residents, as discussed above. Thirdly in the weak link between revenue and expenditure decisions lessens the sense of ownership of local expenditure programs, reducing both efficiency and accountability. 35. The fees and charges levied by local governments since the implementation of the Budget Law in 1997 is a welcome step in providing them with some independent revenues. Local governments can now determine school and hospital fees, market fees, tolls on roads under the jurisdiction of local administration and such other minor revenues. However, these constitute a very small proportion, less than 5 per cent of total expenditures. Thus, own revenue sources do not constitute a major element in financing local expenditures. 36. The main independent revenue source at the provincial level is actual taxes collected in excess of the budget target. In respect of taxes assigned exclusively to local governments, the excess collections accrue entirely to them. In the case of shared taxes, excess collections accrue according to the sharing ratio. Even in respect of taxes accruing exclusively to the center, the center allows provinces to retain some of excess collections. This is provides incentives to local governments to assist in increasing tax collections even through the GTD, which collects all tax revenues, and the Customs Departrnent, which collects import-export taxes, are central government departments. There is also an inherent incentive for all stakeholders to underestimate revenues. Underestimation helps local governments gain additional independent revenues, the GTD can show that it has over-achieved the targets and tax collectors get additional incentive payments. 37. Thus there are inherent incentives in the system to under budget revenues and to exceed the revenue targets. Current regulations reinforce this as the new assignment remains unchanged for three to five years. 38. These additional revenues can be spent by local governments on social sector expenditures excluding salary payments, on capital investment or maintenance or can be deposited in the Financial Reserve Fund. Thus, excess collections add to the provinces' maneuverability in expenditure allocation. Expenditures which the provinces consider important, but cannot finance from their regular assignment of revenues can be funded. 39. A comparison of budgeted or planned with actual provincial expenditures (Table B.3) brings out some interesting features. In 1998 for example, except the Hanoi- Haiphong region, actual expenditures were much higher than the budget figures, varying from about 20 percent in Red River Delta to 43 per cent in HCMC environs. In all provinces taken together, actual expenditure was higher than the budgeted by about 27 percent. Second, the difference between actual and budgeted expenditures has shown a steady increase since 1996, not only in the country as a whole but also in all individual regions except in Red River Delta. It is also seen that percentage difference between actual and budgeted expenditures was the highest in the high-income provinces. 25 Vietnam Public Expenditure Review 2000 - Annexes Table B.3: Excess of Actual Expenditure over Planned (Budgeted) Expenditures Percentage Excess ofActual Per Capita Excess of in Actual Provinces Over Planned Expenditures over Planned Expenditures (Th. VND) 1996 1997 1998 1996 1997 1998 Regions Hanoi-Haiphong -0.6 -2.4 -2.9 -3 -16 -20 Northern Mountains 6.0 9.9 25.9 19 36 100 Red River Delta 16.4 9.2 19.7 28 20 45 North Central Coast 13.1 16.3 21.9 27 41 58 South Central Coast 16.3 27.0 28.0 44 87 98 Central Highlands 14.7 15.2 21.7 53 67 101 HCMC Environs 13.0 15.5 43.2 57 81 235 Mekong Delta 14.0 18.5 33.3 33 52 100 Income Groups High Income Provinces 12.9 15.0 30.5 51 66 140 Middle Income Provinces 10.8 15.5 23.9 26 45 74 Low Income Provinces 9.0 11.1 23.8 21 32 72 Total 11.2 13.9 26.8 33 48 98 Note: Classification of provinces in terms of income categories has been done on the basis ofper capita GDP in 1998. High income provinces are those with per capita GDP of more than VND 3.5 million, Middle Income Provinces have per capita GDP between VND. 2.5 - 3.5 Million, and those with less than VND. 2.5 Million are the Low Income Provinces. Source: Ministry of Finance. 40. In the case of middle and low-income provinces the differences between budgeted and actual expenditures were about the same. This is true also in per capita terms. But provinces with a higher per capita GDP had higher percentage excess of actual expenditures over budget estimates (Figure B.3). Thus, while the system has provided an incentive for additional resource mobilization to provide greater flexibility to local governments, richer provinces have been able to take greater advantage of this incentive. Clearly this has equity implications. Figure B.2: Expenditures in Provinces - Plan and Actual (1998) ZE1400.00 1200.00 A 6 00.00 o.1 400.00 00.00 0 00 10000 15000 200 2500 26 Source :Anistry offinance Per Capita GDP (Th. VND) Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services 41. There are however other reasons to explain these differences, apart from revenue under-estimation. First, there is a budget contingency allowance held by the center which may be allocated to provinces for spending. Secondly, provinces may carry forward surplus funds to the following year. 42. As already mentioned, in the prevailing system of budgeting, with bulk of expenditures already earmarked for salaries, local governments have limited flexibility in their expenditure allocation. Their difficulties further increase when the revenues show a downward trend. In such cases, the lowest level of government usually bears the shortfall and may be forced to raise community contributions to provide the minimum required service levels. 43. Thus, efficiency in expenditure policies at local level depends on the availability of independent revenue sources. However, availability of independent revenues vary with income and wealth position of the locality. This has important implications on equity in the provision of public services. National Programs 44. Local governments also implement various National Programs which cover services considered to be of national importance. Program design and the amount of spending to be incurred are decided by the center, but implementation is through the local governments. These are akin to specific purpose transfers, or to the local authority undertaking the activity as the agent of the central government. The procedure for determining National Programs is specified in Article 4 of the Prime Minister's decision 531 (August 8, 1996). 45. There are two types of National Programs- "National Target Programs" which cross sectors and localities and have clearly specified targets and "sectoral programs" which are confined to individual sectors. At present there are seven National Target Programs and sixteen Sectoral Programs and within each there are a number of sub- programs. The programs cover both current and investment expenditures. The National Target Programs include hunger eradication and poverty reduction (HEPR), family planning, eradication of diseases and epidemics, prevention of HIV, provision of clean drinking water, promotion of sports and creation of employment opportunities. The list of the programs and their budget allocation during the last few years are detailed in Table B.4. 46. In 1999, 3.7 per cent of the State budget was allocated to National Programs (Table B.4). In addition, some resources are made available from ODA, predominantly in the form of loans, and from local contributions. However the amount of local contributions is decided by the Peoples' Committee in each local authority. 47. National Programs have served a useful purpose in ensuring expenditure on important government objectives. This is particularly true of programs targeted to alleviate poverty, spread literacy and prevent and cure contagious diseases. These programs have been especially helpful to disadvantaged groups in mountainous and remote areas. There is therefore, a case for strengthening the programs to enable equitable delivery of essential services. 27 Vietnam Public Expenditure Review 2000 - Annexes Table B.4: Expenditure on national programs Targetedprograms 1997 1999 Expenditure Percent of Expenditure Percent of (Bn. VND) Total Total Afforestation in bare land, hills and mountains (327) 350 14.8 316 10.3 Afforestation in riverside and coastal bare land (773) 150 6.3 120 3.9 Employment and job creation 104 4.4 50 1.6 Education and training 630 26.6 567 18.5 Health 423 17.9 430 14.1 Culture 45 1.9 45 1.5 Nationwide radio broadcasting 20 0.8 18 0.6 Television broadcasting to mountainous areas 45 1.9 45 1.5 Population and family planning program 315 13.3 400 13.1 Information technology development 100 4.2 44 1.4 Drug control 40 1.7 36 1.2 Anti-HIV/ AIDS program 55 2.3 50 1.6 Prostitute control 30 1.3 18 0.6 Clean water and rural sanitation 60 2.0 Hunger eradication and poverty reduction 714 23.4 Total 2,365 100.0 3,057 100.0 Source: Ministry of Finance 48. However to improve effectiveness, some modifications in design and implementation have been instituted recognizing that there may be too many small programs. Some programs have been consolidated to permit provinces to determine the allocation of the total amount to better reflect local priorities, e.g., combining previously separate health programs covering malaria and T.B. Both further changes may be desirable. .Under current arrangements the local authority receiving the funds under specified National Programs can substitute its own expenditure on the service. It may therefore, be useful to formally require provinces to make matching financial contributions, varying with their level of per capita GDP. Finally, it is also important that national programs be regularly evaluated to ensure that they are appropriately targeted and are still needed. IV. Fiscal Arrangement and Inter-regional Equity 49. In a de-concentrated system the center has adequate control of the fiscal system to bring about the desired equalization. Inter-regional redistribution can be achieved either by the central government directly allocating expenditures to different regions or by determining the spread of expenditures between different local government units. It is difficult to trace the distribution of central government expenditures in different regions and therefore, most analyses of redistribution are limited to examining the distribution of provincial expenditures. 50. The distribution of revenue collections and revenue allocations in different regions in Vietnam shows that despite the weaknesses of the system mentioned above, a substantial portion of the revenues collected in richer provinces is redistributed to poorer provinces for spending. Figure B.4 presents the pattern of revenue collections and allocations in provinces with different levels of per capita GDP. Analysis shows that in 1998, while the revenue collections were appreciably higher in more affluent provinces, 28 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services the revenues ultimately allocated to them were significantly lower and vice versa. Thus, when per capita GDP in provinces was higher by one per cent, revenue collections were higher by 1.8 per cent, but they were assigned only 0.9 per cent of extra revenues. The difference between the revenue collection and assignment represents revenue equalization. It is seen that degree of revenue equalization has remained broadly the same since 1996 (Table B.5). Figure B.3: Revenue Equalization in Vietnam, 1998 12 1 0 Revenue Cole > 8 . 6 eveue Assinmot 42 2 0 6 7 8 9 10 11 Source: MOF Ln Per Capita GDP 51. However the issue is whether this level of equalization is adequate to offset the revenue and cost disabilities of poorer provinces. Thus, standards of services and cost vary from one province to another even if revenue allocation is identical. To ensure equalization of public services, it is necessary to take into account cost variations as well. 52. The analysis of per capita expenditures in 1998 brings out features of inequity in the expenditure levels across the provinces in Vietnam. It is seen that per capita expenditures in high-income provinces are significantly higher than in middle and low- income provinces (Table B.5). In the three highest income provinces (Hanoi, Ho Chi Minh and Ba Ria-Vung Tau) per capita aggregate expenditure was twice that of the average of the country and 2.4 times that of middle and low-income provinces. A similar pattern of distribution of expenditures is seen in the case of individual sectors - education, health and transport. There is a significant difference in per capita expenditures in the three provinces and other high income provinces on the one hand and the latter and middle and low-income provinces on the other. However the expenditure levels in middle and low-income provinces, however, are similar. 53. Thus the pattern of local expenditures across provinces in 1998 shows that per capita expenditure are significantly higher in provinces with per capita GDP as shown in Table B.6. It is seen that per capita expenditures were higher by 0.3 per cent for every one per cent higher per capita GDP. Further, planned expenditures were distributed more equally among the provinces as compared to actual expenditures. In 1998 for example, 29 Vietnam Public Expenditure Review 2000 - Annexes planned expenditure were higher in provinces with higher GDP only marginally-it was higher by only 0.17 per cent for every one per cent higher GDP. Thus, in formulating the budget, the intention may be to achieve substantial equalization of expenditures, in actual practice this does not result for the reasons described earlier. It is also seen that provinces with higher per capita GDP had significantly higher per capita capital expenditures than current expenditures (Figure B.4). This feature is seen in all the years since 1996 and suggests increasing regional inequality in the provision of physical infrastructure. 54. The reasons for the distribution of per capita expenditures in favor of more developed provinces are not difficult to see. While the indicative norms used to make initial estimates of expenditure are mainly population based, in some cases they actually favor the provinces with better existing infrastructure. These norms play only an indirect role for, in the final analysis, expenditures are adjusted to match revenues. Thus, the process does provide an adequate mechanism to offset revenue and cost disabilities. In addition, giving provinces the right to retain revenues in excess of budget targets favors better off provinces. The greater capacity to levy fees and charges as well as voluntary contributions also ensures higher expenditures for the richer provinces. Thus, to generate comparable levels of services, poorer localities (particularly at the commune level) need to collect higher "voluntary contributions", which makes the system more regressive. Table B. 5: Local Expenditures on Different Services, 1998 (Capital and Recurrent) Education Health Agriculture Transport Total i. Per Cent of GDP Hanoi, Ho Chi Minh and Ba Ria - Vung Tau 1.4 0.8 0.2 0.7 9.3 Other High Income Provinces 2.8 1.0 0.3 1.1 13.5 Middle Income Provinces 3.6 1.2 0.3 1.2 13.0 Low Income Provinces 5.2 1.4 0.2 1.3 16.8 Vietnam 3.0 1.1 0.3 1.0 11.3 2. Per Capita (Th. VND) Hanoi, Ho Chi Minh and Ba Ria - Vung Tau 177 98 20 86 934 High Income 120 43 32 46 464 Middle Income 105 35 26 35 383 Low Income 118 32 24 29 382 Vietnarn 122 43 26 41 460 Note: See, Table B. 3. Source: Ministry of Finance 30 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services Table B. 6 Income elasticity of revenues and expenditures Revenue/Expenditure Constant Elasticity R bar Sq 1. Current Expenditure 1996 4.2584 0.1652 0.0330 1997 4.3662 0.1582 0.0454 1998 3.6033 0.2714** 0.1343 1999 5.0900 0.0680 0.0063 2. Capital Expenditure 1996 0.3138 0.5987* 0.2332 1997 0.4101 0.5244* 0.1504 1998 0.9152 0.4826* 0.1579 1999 1.2434 0.4227* 0.2135 3. Total Expenditure 1996 3.3992 0.3113* 0.1345 1997 3.7798 0.2731* 0.1244 1998 3.4991 0.3256* 0.1639 1999 4.6026 0.1693 0.0493 4. Revenue Collection 1996 9.2134 1.9029* 0.8223 1997 9.2138 1.8813* 0.8215 1998 8.5532 1.7806* 0.7573 1999 11.3135 2.0871* 0.7670 5. Revenue Assignment 1996 -3.1871 1.0746* 0.6820 1997 -3.2393 1.0903* 0.6661 1998 -2.0184 0.9330* 0.5581 1999 3.7422 1.0845* 0.6487 Note: * Significant at I per cent level,. ** Significant at 5 percent level Figure B.4: Distiibution of Per Capita Expenditure Among __4------ Provinces, 1998 14 - loipil xpenmilure- S 12 10 w 8 Capital Expenditure 6 -iure ;, 4 2- 0 6 7 8 9 10 I1 Ln Per Capta GDP 31 Vietnam Public Expenditure Review 2000 - Annexes Table B. 7: Distribution of public and private investments in provinces 1996-99 Local Central Foreign Other Total Budget Budget Direct Investment Expenditure Expenditure Investment 1. Per Cent of GDP Hanoi, Ho Chi Minh and Ba Ria - Vung 0.83 3.11 15.07 21.41 40.42 Tau Other High Income Provinces 1.09 2.77 12.24 19.86 35.96 Middle Income Provinces 2.94 0.74 1.85 10.20 15.73 Low Income Provinces 3.68 1.97 2.35 9.69 17.69 Vietnam 2.04 2.15 7.45 15.55 27.19 2. Per Capita (Th. VND) Hanoi, Ho Chi Minh and Ba Ria - Vung 99 370 1790 2544 4803 Tau High IiPWe : MOF 82 209 923 1498 2713 Middle Income 80 20 51 278 429 Low Income 80 43 51 211 385 Vietnam 81 111 435 791 1418 Source: Strategic Development Institute, Ministry of Planning and Investment. 55. As mentioned earlier, local public expenditures constitute about 40 per cent of total State expenditures. Therefore, judgements on the equity of expenditure distribution can be made only when regional distribution of central expenditure is also considered. Analysis by the Strategic Development Institute (SDI) of MPI of the distribution of investment expenditures for the period 1996-99 shows that the principal source of unequal distribution in State investment is the direct investment by the Center (Table B.7). The resulting unequal distribution in infrastructure is accompanied by an inequitable spread of foreign investment as well as other private investment - although of course the overall pattern may well be efficient in terms of promoting economic growth as discussed in the following paragraph. Thus, in the aggregate, per capita investment in the three high-income cities was almost 12 times the investment in low-income provinces. Investment in other high-income provinces was about 7 times the investment in low-income provinces. In fact, middle and low-income provinces received very little investment during the four years. 56. Since the commencement of Vietnam's economic reform initiatives the Hanoi- Haiphong region and Ho Chi Minh City environs (including Ba Ria- Vung Tau where there is oil based activity), have served as growth centers. Besides infrastructure, proximity to markets has attracted significant private investment in these regions. Consequently there has been a heavy demand for state investment in infrastructure and the government has had to make disproportionate investments in these regions maintain the economic growth momentum. The fast growing regions will continue to demand large investments in infrastructure and regional distribution of State investment in the coming years will exhibit serious tension between equity and efficiency. Resolving this will be a major challenge. 32 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services v. Reforming Intergovernmental Fiscal Arrangement in Vietnam Independent Revenue Sources, Incentives and Efficiency in Service Delivery 57. In spite of the recent changes in the system of intergovernmental fiscal relations, further reforms may be desirable to empower local governments and to improve their capacity to efficiently deliver public services. In the absence of independent revenues, local governments do not have the flexibility and autonomy to allocate expenditures to achieve maximum efficiency. Secondly, in a "top-down" system of expenditure allocation and revenue assignment such as in Vietnam, achieving efficiency in spending is more difficult. Therefore, it is desirable not only to make the system more participatory but also simple, transparent and formula based. A formula based system should allow for revenue and cost disabilities to enable the poorer localities to provide comparable standards of public services. However, this should not be achieved at the cost of reducing public services in the growth centers which need adequate independent revenue sources to meet their infrastructure requirements. 58. As the demand for public services increases, local authorities need more flexibility and autonomy in allocating expenditures. At present, their only independent revenue sources are some minor fees and charges, excess collections over targeted revenues and contributions from households and communities. There are significant variations in independent revenues not only between provinces, districts and communes but also among each of the three layers of local administration. At the level of provinces, independent revenues do not constitute a significant source of revenue. In 1998 for example, non-tax revenues constituted about 17 percent of total budget revenues and in provinces, they constituted less than 6 percent. Excess revenue collection over the budget target is also a source, but may dry up in times of economic stagnation. This source, besides adverse consequence on equity, adds some degree of uncertainty to the finances of local governments. 59. The need for more independent revenue sources is particularly felt in the economic growth regions - Hanoi-Haiphong belt in the north and Hochiminh environs in the south. Empowering these cities with adequate revenue sources is particularly important for, these are the engines of growth for the economy as a whole. They generate economic activity, provide markets for rural goods, and create income-earning opportunities to large number of rural poor who migrate to these regions in search of livelihood. Greater independent revenue sources may also create a sense of ownership and enhance accountability at the local level. 60. Any taxes to be assigned to the provinces should have a regional spread be relatively immobile. The taxes that satisfy these criteria include land and housing tax, license tax, tax on property transfers and agricultural tax. There are also some indirect levies such as wholesale tax which have local bases and are not significant contributors to revenue. These taxes could be assigned to provinces. The revenue loss to the center would be small. As total revenue collected from these taxes in 1998 constituted only about 6 per cent of total revenues (Table B.8). 61. Assignment of tax powers to the provinces implies that the provinces should have a small tax department of their own to levy and collect these taxes. Given that the taxes 33 Vietnam Public Expenditure Review 2000 - Annexes assigned to them have a regional spread, they would have better information on tax bases and thus a comparative advantage over central government in mobilizing revenues from these taxes. Because at the provincial level there is a closer correspondence between tax payments and benefits of public services, tax compliance will be better as well. However an important precondition would be that the provinces should also develop the necessary capacity to administer these taxes. Therefore, as a practical solution, it may be preferable to make the assignment of tax powers initially to larger cities with capacity to design, implement and enforce these taxes and later to extend them to other provinces as the capacity to administer these taxes develops. Table B. 8: Composition of tax Revenue in Vietnam, 1998 Tax Revenue Revenue Bn. VND Per cent of Total Corporate income tax 13099 18.5 Individual Income tax 1782 2.5 Capital User Charges 1724 2.4 Excises 5573 7.9 Trade taxes 14873 21.0 Natural resources tax 3282 4.6 Value Added Tax 11750 16.6 Land and housing tax, license tax, tax on 2017 2.8 property transfer Agricultural tax, 1956 2.8 Other indirect taxes 2898 4.1 Total tax revenue 58954 83.2 Non tax revenue 11868 16.8 Total Revenue 70822 100.0 Source. MOF Need for a formula based transfer system 62. The most important reform issue is the method of determining funding allocations to local govermments. One option is to reforn the detailed system of expenditure norms. Even with the information already available with the MOF, it is possible to determine expenditure needs based on more objective norms. The recurrent expenditure needs in respect of a service can be estimated by taking expenditure per potential beneficiary and then weighing it with cost disability factors. Thus, in the case of primary education, expenditure per child in the school going age group could be weighted by variables like geographical terrain, population density and urbanization. Similarly, in the case of health expenditures separate norms can be developed for preventative and curative health care with proper weights assigned to cost disability factors. However, maintenance expenditures will have to be determined by applying engineering norms on existing stock of infrastructure (roads, buildings etc.). In the case of investment expenditures, to begin with it is necessary to exclude all commercial investments in favor of the private sector. Infrastructure investment estimates, however, will have to be made on the basis of the targeted growth rates in different provinces. Once these are estimated on the basis of appropriate and transparent norms there is less need to for formal or detailed negotiations to determine the expenditure requirement of the provinces. 34 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services 63. Another option is to determine the transfers to be made to each of the provinces directly in the basis of a simple formula. Given that the objective of the transfer system is to offset revenue and cost disabilities, total revenues available for distribution to the provinces can be allocated on the basis of a simple formula in which the transfers will be positively related to population but inversely related to per capita GDP. Additional weight can be given to cost disability factors such as area and population living in highlands and remote areas. Additional requirements of Hanoi and Ho Chi Minh City because of urbanization can also be allowed for. 64. The total amount of transfers by way of revenue shares and cash transfers may be decided by the central government at the beginning of the budget process. As regards revenue assignment, it is preferable to decide on a share of total revenue collections rather than different shares of revenues from individual taxes. The ratios accruing to individual provinces can be worked out on the basis of the formula designed to offset revenue and cost disabilities. This will preserve the incentive for the provinces in total revenue collections irrespective of the source from which revenue is derived. At the same time, the revenue assignment will assist in achieving of equity as the formula used to determine the shares of individual provinces is designed to offset revenue and cost disabilities. 65. The revenue shares of individual provinces as well as the cash transfers may be decided once in three years to provide the necessary stability in the arrangements. Cash transfers may be adjusted for inflation according to a pre-determined formula. There can also be a provision to give extra transfers to take care of the unforeseen expenditures and contingencies. 66. Designing the fiscal decentralization system on the above lines would promote expenditure efficiency as well as inter-regional equity. Providing the provinces with some independent revenue sources would provide them with incentives to generate revenues to finance priority expenditures. Formula based sharing of revenues maintains the incentives and at the same time, targets the distribution according to expenditure requirements. It also avoids the potential for subjectivity and arbitrariness of the negotiated system. Improving the Design and Implementation of National Programs 67. National Programs have served a useful purpose in ensuring expenditures on important government objectives. To ensure the targeted outlay on these services is achieved, it may be necessary to require the local governments to make matching contributions. The matching ratio may, however, vary with the level of per capita GDP. It is also important that National Programs be regularly evaluated to ensure that they are appropriately targeted and are still needed. 35 Vietnam Public Expenditure Review 2000 - Annexes Appendix B.1: Revenue Assignment in Vietnam 1. Central Revenues II. Revenues of Provinces III. Revenues of Districts Revenues of Communes and Townships (Article 34) (Article 28) (Article 30) (Article 32) 1. 100% Assigned 1. 100% Assigned 1. 100% Assigned. I. 100% Assigned a. Value Added tax on imported a. Land rent a. Trade license tax excluding a. trade license tax paid by individuals and small business groups; goods b. Proceeds from leasing and the tax paid by individuals b. animal slaughtering tax; b. Import tax, export tax sale of State-owned and small businesses in c. Fees, charges and contributions to the commune or township in accordance with the c. Special Consumption tax residential houses communes and townships; law; excluding: c. Registration fees collected b. Slaughtering tax paid by d. Revenues from the use of public land and from other public property; the tax on home-made in urban and rural districts slaughtering enterprises e. Revenue collected from non-business activities managed by the commune or township playing cards; votive papers excluding those on houses located in wards; administration; and objects; services related and land. c. Fees and charges collected f. Voluntary contribution to the communes and townships; to dancing halls; massage d. Revenues from lotteries from activities of the g. External aid given directly to communes and townships in accordance with the parlors; or karaoke bars; golf e. Extemal grants given agencies under the provisions of law; membership cards and directly to the provinces in management of the district h. Balance from the previous year; playing tickets; casino accordance with the Law. level administration; i. Additional allocations from the higher level budget; business; jackpot games; f. fees, charges and other d. revenues collected from j. Other revenues as prescribed by law. horse racing and car racing revenues to be remitted to non-business activities of betting tickets. the provincial budgets as units under the d. Income tax on enterprises stipulated by the management of district under the whole-branch Government. level administration; accounting system g. Revenues mobilized from e. External aid given directly e. Taxes and revenues from organizations and to districts in accordance petroleum industry to be individuals for investment with the provisions of Law; remitted to the central budget in infrastructure projects f. Contributions from as stipulated. according to govemment organizations and f. Income from State controlled stipulation. individuals for investment capital, capital recovered from h. Voluntary contributions in infrastructure projects economic establishments, from domestic and foreign according to govemment recovery of State loans and organizations and stipulations; interest on them, income from individuals g. Voluntary contributions the State reserve fund. i. balance from the previous from organizations and g. Govemment borrowing; year; individuals inside and external grants j. additional allocations made outside the country to h. Fees, charges and other in the central budget; district budgets; revenues as stipulated by the k. Other revenues prescribed h. Balance from the previous government. by Law. year; i. Balance from the previous year i. Other revenues prescribed by k. Other revenues prescribed by Law; law. .. _ 36 Annex B - Fiscal Decentralization and Efficiency and Equity in Public Services 2. Revenues Shared with Local 2. Revenues shared with the 2. Revenues shared with the 2. The revenues shared with provincial and district budgets as prescribed in Clauses 2, and Budget. Center according to Article 28 provincial -level budget as 4 of Article 30 of this Law. a. Value Added tax excluding 1 (2). The division of these prescribed in Clauses 2, 3, and (a). revenues to different levels of 4 of Article 30. b. Enterprise income tax local government will be excluding I (d). decided by the provincial c. Personal income tax administration. d. tax on remittance of incomes abroad e. Tax on the use of State capital 3. Revenues shared with 3. For provincial capitals and district, commune or towns, some share of township budgets; registration fees (excluding a. tax on the transfer of land house and land registration use right; fees) co]lected in their b. House and land tax; respective territories to set up c. Land use levy. investment funds according to Government stipulations. 4. Revenues shared with province, district, commune, township or ward budgets; a. Tax on the use of agricultural land; At least 20% of this tax should be assigned to commune b. Natural resources tax; c. Registration fees on house and land; d. Special consumption tax excluding those specified in Article 28 (Ic); 37 Vietnam Public Expenditure Review 2000 - Annexes Appendix B.2: Expenditure Assignment in Vietnam I. Regular spending on: 1. Regular spending on: Regular spending on: Regular Spending on: (a) Non-commercial activities in education, training, (a) Non-commercial activities in the field of (a) Non-economic spending on economic, (a) Social work and activities in the fields of health-care, social affairs, culture, information, economy, education, training, health-care, cultural, communication, physical training culture, information and sports managed physical training and sports, science, technology culture, information, physical training and and sport and social activities managed by by the commune or township; and environment and other operations managed sports, social affairs, science, technology, district administration; particularly spending (b) Financial support for extracurricular by the govemment; environment and other non-economic on education, training, and health as assigned education, child care centers and (b) non-commercial activities managed by the operations managed by the provincial by the provincial govemment; kindergartens managed by the commune central agencies; agencies; (b) The defense, security and social safety and or township; (c) National defense, security and social order and (b) National defense, security and social order order maintenance tasks assigned to the (c) health care activities in commune or safety, excluding those assigned to local and safety assigned to the provinces; district; township; govemments. (c) Activities of the agencies of the state, the (c) The activities of the district agencies of the (d) management, repair or maintenance of (d) Operations of the central agencies and the Communist Party of Vietnam, and socio- state and the Communist party of Vietnam architectural projects, public properties, Communist party of Vietnam and other socio- political organizations at provincial level; and other socio-political organizations; welfare projects, and roads managed by political organizations; (d) Support to social and socio-economic (d) Support to district social organizations, the commune or township; (e) Price subsidies in accordance with State policies; organizations at the provincial level socio-professional organizations in (e) Activities of the agencies of the state, (f) National Program managed by the Central according to the provisions of Law; accordance with the provisions of Law; the Communist Party of Vietnam and govemment; (e) Implementation of social policies managed (e) Other expenditures prescribed by Law; socio-political organizations of the (g) Support for social insurance fund as provided by by the province; (f) In addition to the above, provincial capitals commune or township; Militia and self- the government; (f) The national programs assigned by the and towns shall also assume other tasks of defense work; maintenance of social (h) Subsidy to eligible individuals according to the Government to provincial management; spending on management, repair and order and safety in the commune or social policies implemented by the central (g) price subsidies in accordance with state maintenance of public utilities and urban township; government; policies; non-business and administrative activities. (f) Other spending in accordance with the (i) Support to social and socio-economic (h) Interest on borrowing for investment as provisions of Law. organizations at the central level according to the allowed by the Government; provisions of Law. (i) other expenditures stipulated by Law. (j) Interest payments on borrowing; (k) Aid; Other expenditures stipulated by Law. 2. Investment Expenditures on: 2. Investment expenditures: 2. Investment expenditures: 2. Spending on Investment: (a) Socio-economic infrastructure projects of non- (a) Socio-economic infrastructure projects of Socio-economic infrastructure projects as 3. Construction of socio-economic commercial nature managed by the Central non-commercial nature managed by assigned by provincial level. Provincial capitals infrastructure projects in accordance with govemment; provinces; and towns shall be assigned the task of building the assignment and responsibilities by the (b) Investment in State enterprises, contribution of (b) Investment and capital support for State schools, of all levels and public utilities such as province. capital to joint ventures with enterprises in areas enterprises in accordance with the electric lighting, water supply, and drainage, warranting State participation in accordance with provisions of Law; urban traffic, traffic safety, and urban sanitation. the Law; (c) Expenditures on National Investment Reserve Fund and Development Support funds for development programs and projects; (d) The State reserve; _ __._.__ 38 Annex B - Fiscal Decentralization and Efficiency andEquity in Public Services 3. 3. Repayment of Loans; 3. Contributions to lower-level budget (a) Repayment of loans 4. Contribution to Financial Reserve Fund; (b) Contribution to financial reserve fund; 5. Contribution to lore level budgets. (c) Contribution to lower level budgets; 39 Vietnam Public Expenditure Review 2000 - Annexes Appendix B.3: Budget Transfer Norms (Attached to MOF Circular No. 38 TC/NSNN of 18th July 1996) NORMS FOR EXPENDITURES ON EDUCATION, CULTURE, INFORMATION, SPORTS, RADIO AND TELEVISION AND HEALTH SERVICES VND/person/year No Sector Education Culture and Sports Radio and Health Information Television Prevention Curative 1 Cities 65000 4200 3100 2700 5200 14500 2 Plains 46000 2600 1800 1700 3800 11000 3 Mid-lands 57000 2700 1700 1700 3800 11000 4 Deep lying, low 62000 3200 1300 1800 4800 13860 mountains 5 High mountains 82000 3800 1300 2200 6000 17600 Note: Ha Noi and Ho Chi Minh City are entitled to a 20% increase for education and 50% increase for culture and information, sports, radio and television and health services (vs. the standard scales). NORMS FOR EXPENDITURES ON HEALTH SERVICES Million VND/hospital bed/year No Type of Service/Sector Hospital Sanitarium Clinic i Ministry of Health 19 4.7 4 2 Labor, invalids and social affairs 5.3 1.7 3 Other sectors 10.5 4.7 4 NORMS FOR EXPENDITURES ON TRAINING Million VND/trainee/year Level of Training/Type of Training Tertiary Secondary Vocational education Education Training '1. Arts and Sports 8.0 6.5 2. General and teacher training 6.3 4.0 3. Geology, hydrology and mining engineering 5.5 4.1 5.4 4. Marine navigation 6.4 4.0 4.7 5. Agriculture, forestry and fisheries 5.9 3.54 6. Medical science and pharnacology 6.0 3.6 4.5 7. Food and foodstuff technology 5.6 3.4 4.2 8. Metallurgy, mechanical, thernal, electrical and 5.9 3.5 4.3 civil engineering 9. Storage engineering for goods and materials 5.5 3.0 4.1 10. Electronics and posts and telecommunications 5.3 3.2 3.9 11. Culture, information and tourism 5.4 3.2 4.0 12. Economic management, enterprise 5.2 3.1 management, public administration and law 13. Re-training at the average scale of 3.0 million VND/fellowship (equivalent to 10 working months /trainee) 40 Annex C: Commune-Level Fees and Contributions Appendix B.4: Budget allocation norms (MOF Circular No. 562,3 March, 1998) Unit: Thousand VND/student/year High mountainous areas Average Cities Plains Mid-land Deep High and lying and mountai central low -nous coastal mountain areas, areas ous areas islands 1. Pre-School Education + Nursery Schools 490 390 429 507 585 780 + Kindergartens 411 290 319 377 435 580 2. Primary Schools 290 220 242 286 330 440 3. Lower Secondary Schools 390 320 352 416 480 640 4. Upper Secondary Schools 450 380 418 494 570 760 5. Special Upper Secondary 800 710 781 923 1065 1420 Schools for Gifted people 6. Practical Technical Training 80 120 132 156 180 240 Centers 7. Continued Education Centers 85 180 198 234 270 360 8. Schools for Disabled Children 800 800 800 800 800 800 9. Boarding Schools for Ethnic 0 0 0 0 Minority Children + District level (primary and lower 2200 2200 2200 2200 2200 2200 secondary) + Provincial level (upper 2700 2700 2700 2700 2700 2700 secondary) + National level (bridging for 3000 3000 3000 3000 3000 3000 tertiary education) 10. Literacy Education 70 60 66 78 90 120 NORMS FOR DETERMINING LOCAL HEALTH EXPENDITURES - 1998 a. Applicable to local levels: Unit: Million VND/hospital bed/year Ha Noi Ho Chi Other Delta Mid- Deep High Minh Cities lands lying and mountainous City low areas, mountain islands ous areas Provincial and 17 16 13-16 10-13 11-14 12-15 14-17 City Level District Level 13 12 8-11 6-9 7-10 8-11 7-10 41 Vietnam Public Expenditure Review 2000 - Annexes b. Applicable to centrally managed provincial/city health departments Unit: Million VNDthospital bed/year No. Sector Type of Service Hospital Sanitorium Clinic 1 Ministry of Health 19 4.7 4 2 Labor, invalids and social affairs 5.3 1.7 3 Other sectors 10.5 4.7 4 NORMS FOR DETERMINING LOCAL GENERAL ADMINISTRATION EXPENDITURES - 1998 a. Applicable to local levels: Unit measurement: Million VND/Staff member/Year Administrative Centrally Provinces Provinces Provinces Provinces level administered and (districts) in (districts) in (districts) in cities districts in mid-land deep lying and high the delta and coastal low mountainous areas mountainous areas areas Provinces and 11.4 10 11.4 1 7 18 centrally managed cities Provincial cities 8.5 8,5 9.8 15.5 17.3 District and 8.0 8.0 8.7 14 16.3 equivalent level Notes: Exclusively Ha Noi and Ho Chi Minh City are entitled to a 20% increase (vs. the standard scales) b. Applicable to the central level: 14.0 million VND/staff member/year. Centrally managed agencies shall apply the scales of expenditures corresponding to their respective levels 42 Annex C - Commune-Level Fees and Contributions ANNEX C - COMMUNE-LEVEL FEES AND CONTRIBUTIONS Background to the Study I. Article 25 of the Law on State Budget (dated March 20, 1996) assigns "responsibilities and rights" to People's Councils and People's Committees. This includes the right to determine 'fees, charges and other contributions of people ". A number of recent studies have indicated that there was concern and confusion by farmers about the range of fees and contributions they were required to pay3. These fees and contributions are authorized largely at the local level and are distinct from the nationally- mandated taxes, which by contrast seem to cause little confusion or anxiety. 2. Information available at the time of the study suggested that the high level of fees and contributions was a more pressing problem in the Red River Delta, the north central and central regions than elsewhere in the country. In some parts of the country, the level and number of fees and contributions may not be an issue. Had the study taken place in the Mekong Delta region or the south-east region, the findings may have been quite different. The study looked uniquely at rural areas, again because earlier studies suggested that this was more of a problem in rural areas than in urban areas. 3. It was agreed with the Ministry of Finance that it would be useful to examine this issue systematically and that such research could usefully inform the Public Expenditure Review (PER). The World Bank office in Vietnam commissioned a team of researchers to conduct field visits to three rural Provinces of Vietnam to interview local officials and individuals about the amounts various households were paying in fees and contributions and about the process surrounding the levying and collection of such charges. This annex summarizes information presented by a team of researchers in a longer study report. 4. The researchers visited Ha Tay, Quang Binh and Ha Tinh Provinces. The first two provinces were chosen because they had been nominated by the Ministry of Finance as focal points for other PER work. Ha Tinh was not one of the PER Provinces, but earlier surveys had indicated that the high level of fees and contributions was a real problem for the poor and it was agreed that it would be helpful to include Ha Tinh in the study. 5. In order to gauge the extent to which different processes are employed by different communes within the same district, one district and two communes were selected in each province: Thuong Loc and Nhan Loc communes in the Can Loc district (Ha Tinh Province); Van Trach and Hoan Trach communes in the Bo Trach district (Quang Binh Province); and Hop Tien and Phu Luu Te communes in the My Duc district (Ha Tay Province) The team then conducted interviews: 3Oxfam UK & Ireland, 1996; Nguyen Manh Huan, 1998; ActionAid Vietnam, 1999; Government of Vietnam-donor-NGO Poverty Working Group, 1999. 43 Vietnam Public Expenditure Review 2000 - Annexes * with provincial Departments of Finance (except in Ha Tinh), in order to gather information from each Province about its policy on taxes, fees and contributions and its official decisions on the subject, particularly any instructions issued to the Districts; * with district People's Committees, usually with Chairman of the People's Committee and representatives from the Department of Finance. Again the team sought information about official decisions and instructions to communes; * with two communes per selected district. At this level, the team met the Party Secretary of the Commune (usually also the Chairman of the Commune People's Council), the Chairman of the People's Committee, his deputy, and the cadre responsible for finance. There the team asked for the commune budget (both revenues and expenses), and gathered information about guidelines governing partial and full exemptions from fees and contributions, about procedures for collecting payments and about the information provided to citizens by the commune administration and leadership; * with village managers (or in the case of Ha Tay, with cooperative brigade leader) and households. 80 households were interviewed, the bulk of whom were categorized as poor. Approximately 60% of respondents were "poor" (ngheo), 30% "average" (trung binh), and 10% "wealthy" (kha) households. 6. The study was carried out during February and March 2000 by a combined team of one local foreign consultant and two national consultants. Three field missions were organized with the help of the World Bank, the Ministry of Finance and the international NGO ActionAid. Study Findings Households pay a large number of fees and contributions 7. The team collected information about more than 32 different fees and contributions levied in the six communes visited (excluding school fees). In addition they collected details of a further 14 education-related contributions. Not all of these are levied in all the communes, but most are levied in most of the communes. These fees and contributions are in addition to the nationally-mandated agriculture and residential taxes, which households are also paying. The fees and contributions, excluding those which relate directly to child education, are listed in Appendix C. 1. This table also illustrates some other features about the fees and contributions levied on the farming households in the six communes. First, there are range of ways in which the fees and contributions are calculated by the local authorities - gome on a per worker basis, some per sao of land, some per capita, others per household, others still per pupil and one or two per woman (Table C. 1). 44 Annex C - Commune-Level Fees and Contributions Table C.1: Basis for Calculating Fees and Contributions in Six Communes Thuong Nhan Loc Van Trach Hoan Hop Tien Phu Luu Loc Trach Te per worker 11 6 3 6 2 5 persao 6 9 2 5 6 5 per household 5 3 2 2 6 2 per capita 3 4 perpupil 16 16 15 15 16 14 per woman 2 1 1 1 1 1 Other 2 1 1 1 1 1 TOTAL 40 38 27 29 31 27 8. Secondly, there is considerable variation on the number, level and calculation of different fees between the communes. There are 23 official contributions in Thuong Loc, 20 in Nhan Loc, 13 in Van Trach, 15 in Hoan Trach, 16 in Hop Tien and 13 in Phu Luu Te; In addition, in the 6 communes, there are between 4 and 8 fees, and between 11 and 13 "voluntary" contributions for schools. Table C.2 illustrates some of these differences. Table C. 2: Examples of Differences in Fees and Contributions Between Areas FeelContribution Differences Irrigation fee From 13,5 kg/sao up to 25 kg/sao Agriculture production Does not exist in 3 out of the 6 communes visited by the team encouragement fee Plants and cattle protection fee Does not exist in Quang Binh and Ha Tay Duty for public work 15 days in one commune, 10 days in the 5 other From VND 3.400/day up to 1 0.000/day Commune budget fund Exists in two communes only Village fund Exists in three communes only 9. These variations arise because each level of the administration is, quite legally, taking the instructions of the level above and adding to them - either adding to the range of contributions levied, or increasing the charge for these contributions and keeping the difference for their own expenditure needs. So the range of fees and contribution charges that any one household will pay reflects a series of decisions, decrees and ordinances issued by various levels of Government. Households receive little information about the basis of charges. The system is confusing and non-transparent 10. This list of fees and contributions in Appendix C. 1 was compiled directly from the commune officials. Households themselves generally do not receive information at this level of detail. In general, demands for payment would be broken down into five or six categories which would represent the sum of several unspecified contributions. Because there is such a large range of fees and contributions each one of which might be 45 Vietnam Public Expenditure Review 2000 - Annexes calculated in one of several ways, calculating the total payment due from any one household is an obscure and error-prone process. The complexity of this process means that only one or two people in each commune would be aware of what the full range of charges are and how they are calculated. The total bill for any individual household will depend on which province it is located in, which district of that province, which commune of that district, how many people are in the household, how many laborers are in the household, how many pupils are in the household, how many women are in the household and how much land they have. It will also depend on whether that household is eligible for any full or partial exemptions for any of the fees and contributions. The guidelines surrounding eligibility for exemptions are also not always clear. 11. As a result of this lack of transparency, many of the respondents during the study were not clear at all how much they were paying for any specific contribution. They were unable to verify their demand for payment and therefore unable to question or appeal against it in an informed way. It is also very difficult for household to gauge whether they are receiving the services they are due. If households are paying an agricultural extension fee, then they might reasonably expect to receive some kind of agricultural extension services. If they are not aware that they are paying such a charge because it is not itemized on their demands for payments, then it makes it difficult for them to hold authorities accountable for providing these services. 12. Part of the reason for the confusion may lie in local officials trying to simplify the demand for payments in the belief that households will not be able to understand a more detailed breakdown of charges. They may be right that households would have difficulty understanding the detail, but the practice of consolidating several different charges relating to different services does not lead to greater clarity. Officials themselves might also have difficulty understanding the system of levying charges and there is possibly a need to develop stronger capacity in designing and managing simplified systems of revenue raising. Contributions and fees are a very important source of commune financing 13. Interviews with commune and district officials illustrated the crucial role which these fees and contributions play in boosting the commune budget. In three out of six communes, these funds were larger than the transfers coming from higher levels of govermnent. In four of the six communes, locally-raised contributions and fees constituted 23%-61% of the total commune budget (table C.3). 14. In principle, contributions raised in the name of delivering a specific service or good are earmarked for this particular activity. In practice, however, the team found sometimes that funds raised for various activities were being used not for the specified purpose, but were being used fungibly to finance general commune expenditure. For example, in one case contributions raised for a specific, education-related purpose were being used only partially to finance this expenditure. The residual was being used to finance allowances for commune officials who were not otherwise remunerated. Some communes levy some charges which are simply channeled directly into recurrent commune expenditure Such a commune fund was found in two of the six communes and a parallel village fund was found in three communes. 46 Annex C - Commune-Level Fees and Contributions Table C.3: Percentage of commune budget raised through fees and contributions4 (VND millions) Province Commune Total commune Locally % of total revenue raised revenue Quang Binh Van Trach 543 314 57% Hoan Trach 193 27 14% Ha Tinh Nhan Loc 450 328 58,5% Thuong Loc 1202 737 61,3% Ha Tay Hop Tien 447 103 23% Phu Luu Te 587 24 4% Source.: Calculatedfrom 1999final accounts of communes Contribution-payers know little about how their money is spent 15. Government Decision No 29/1998/ND-CP - the "Grassroots Democracy Decree" - requires commune officials to publish their budgets and actual expenditures. With the exception of one commune in Ha Tay, there was little evidence that this was taking place. Indeed, most of the households indicated that they had no knowledge of the commune budget and what money was spent on. Current practice means that households have no satisfactory means of holding local authorities accountable for delivering services in return for contribution payments. 16. The combination of lack of information about commune-level expenditures and lack of transparency in calculating the complex bundle of contributions and fees creates an environment in which misunderstandings could easily arise. The absence of appropriate mechanisms for addressing grievances and clarifying misunderstandings intensifies the situation. Contributions are not strongly linked to ability to pay 17. It is difficult to gather information on exactly how much households have spent in taxes fees, and contributions. Many households are unable to break down their payments in this way and information sometimes came from the village manager. While it is possible to calculate the percentage of household revenue spent on taxes, fees and contributions, it is not possible to establish unequivocally whether certain elements are progressive or regressive. 18. Figures calculated by the research team indicate that poorer households may sometimes be paying a greater proportion of their annual income in taxes, fees and contributions than wealthier families (Table C.4). In the households interviewed for these calculations, the wealthy household was paying less than all but one of the poor households. The total taxes, fees and contribution of the wealthiest household were two- thirds that of a poor household which had half the annual income enjoyed by the wealthiest household. The household paying least in fees and contributions was the 4 It has not been possible to separate the service fees and charges from other fees and contributions in communes in Ha Tinh These service fees and charges are excluded from the fees and contributions in the other communes, where they represent between 1-4% of the total revenue. 47 Vietnam Public Expenditure Review 2000 - Annexes smallest household, with no children in school. For some households, the overall burden of taxes, fees and contributions was extremely high. 19. It seems most likely that it is the fees and contributions which lead to a situation where poor households pay more than better-off households. The research team found little evidence of dissatisfaction with the agriculture tax, which people found to be clear and fair. Table C.4: Annual (1999) Taxes, Fees and Contributions paid by seven households in Nhan Loc communes, Can Loc District, Ha Tinh Province5 (in VNDOOO) Hanh Tho Tan Vinh Hoa Ky Quy Poor poor Poor poor average Average wealthy Status Persons 6 4 7 8 3 11 6 Workers 2 2 2 2 2 2 2 At school 4 2 4 3 0 4 4 Annual Income 6400 4000 4440 3700 6800 7900 7500 Total taxes, fees and 659 700 1167 1500 142 1675 1063 contribution %/income 10% 18% 26% 41% 2% 21% 14% 20. At least two factors underlie this rather random relationship between the socioeconomic status and fee/contribution payments of a household. Table C. 1 identifies the range of ways in which the fees and contributions are calculated. Of these perhaps two might be loosely linked with a household's earning potential, namely contributions calculated according to the number of workers and contributions calculated in relation to the size of landholdings. These do not make up the majority of fees and contributions charged, which are more closely linked to the number of children in school a household has. Finding a way to more directly connect earning potential and these contribution payments would lighten the burden on the poor. 21. Irrigation fees show considerable variations across communes It would be appropriate to relate these as closely as possible to the volume of water consumed, to prevent wastage. But in redesigning the system, the objective should be to make it more progressive overall. 22. A second factor which mitigates against a more progressive system of fees and contributions is the issue of exemptions from fees and contributions. Eligibility criteria for full and partial exemptions from specific fees and contributions have been set down in various decisions by central, provincial, district and commune level authorities. Just as a household's contribution bill will depend on central decisions, plus additions from the Province, plus additions from the District, plus additions from the commune, so a household's eligibility for exemption from a given contribution will be dependent on decisions from this series of administrative levels. The range of contributions which 5 The gross annual income amount was calculated by the team member who interviewed the households and a staff from ActionAid specialised in micro credit activities who knew them personally. Total amount of taxes. fees and contributions was calculated based on information obtained from household interviews and verified with village manager in case the household did not remember well. 48 Annex C - Commune-Level Fees and Contributions various groups may be granted exemption from is not extensive, given the very long list of potential contributions. In the areas visited by the team, fees and contributions which certain groups may claim exemption from include: school fees (but not contributions); the natural disaster fund; the defense and security fund, the child protection fund, the "Den On Dap Nghia6" fund; the public work contribution; and the social evils fund. Groups which are often listed as eligible for exemptions or reductions in these charges include: the disabled; orphans; those living in areas officially designated as poor; "poor" or "hungry" households; civil servants suffering from injuries; disabled and ill veterans; families of soldiers who died for the Fatherland; students; military and police personnel; victims of industrial accidents; single parents; pregnant women; women with children under the age of 3 years; people recently released from the army; registered clergymen; non-commissioned officers and soldiers; and the elderly. Not all these groups are necessarily poor. Not all of the groups are eligible for exemption or reductions for all the contributions and fees listed above. Indeed, most of the "vulnerable" groups are eligible for exemptions from only one or two of the listed fees and contributions which does not really make a great impact on the overall amount a household must pay. 23. A system of exemptions which is clear to understand, well-publicized and linked to households' economic status could help to make the fee and contribution bill more progressive. Certain fees and contributions are particularly unpopular 24. Interviewees did not reject all fees and contributions as unfair. Where fees and contributions were linked to the provision of services that the households received, and considered good value for money, there was acknowledgement of the need to pay the required charge. Other fees and contributions were regarded as more unfair. 25. Education-related contributions were regarded as particularly burdensome. School fees were the same in all communes and followed central guidelines. In addition to school fees, there are a wide range of contributions which were required and the level of these varied considerably according to location (table C.5). In many instances, the amounts demanded at the commune level were higher than set out in provincial guidelines. For example in Thuong Loc commune the school construction charges exceeded the provincial guidelines by a minimum of 50% (for grade I students) to a maximum of 350% (for grade 5 students). The team calculated that educating three children (an average family size for the communes visited) would cost VND390,000 per annum. 26. In Quang Binh and Ha Tay Provinces, the irrigation fee was causing dissatisfaction because it was not related to the amount of water they were able to use on their crops. In Ha Tinh, there were complaints about the agriculture extension fee, which respondents felt was high relative to the value of the service they received. They felt that the extension messages broadcast by television and radio were more useful, but they were 6 This fund does not translate well, but roughly means duties towards those who have participated in the Revolution, or died for their country, or families of these people 49 Vietnam Public Expenditure Review 2000 - Annexes unable to afford to buy this equipment which would allow a regular flow of extension messages. Table C.5: Other Contributions to Schools (in VND per pupil/year) Contributions Thuong Nhan Van Hoan Hop Tien Phu Luu Loc Loc Trach Trach Te School kindergarten 10.000 10.000 40.000 36.000 5.000 20.000 construction Lower secondary 55.000 25.000 60.000 35.000 40.000 40.000 Upper secondary 45.000 25.000 45.000 35.000 30.000 30.000 Parent's fund 15.000 15.000 10.000 17.000 15.000 20.000 Insurance 5.000 5.000 10.000 10.000 12.000 12.000 Books & class decoration 5.000 5.000 5.000 5.000 6.000 6.000 Chairs (for ceremonies) 5.000 5.000 5.000 5.000 Poor pupil Assistance 2.000 2.000 2.000 3.000 Electricity 1.000 1.000 2.000 2.500 4.000 6.000 Teacher's Day 1.000 1.000 2.000 2.000 3.000 3.000 Lunar calendar's New year 2.000 2.000 2.000 1.000 2.000 2.000 Labour fee 18.000 18.000 18.000 18.000 18.000 18.000 Supplemental study 36.000 36.000 36.000 36.000 36.000 36.000 Year end party 2.000 2.000 1.000 2.000 2.500 2.500 School protection 3.000 3.000 3.000 1.000 2.000 Examination fee 10.000 10.000 5.000 5.000 10.000 10.000 Total 105.000 105.000 96.000 102.000 115.500 120.500 27. Households often expressed the view that some of the expenditure financed by contributions was not appropriate for poor households. In one commune, the commune had levied a contribution for the construction of a cultural house, which was not a priority expenditure for the households interviewed in the research. Similarly with some of the funds, the poorer households felt they should be benefiting rather than contributing. Households resented, for example, paying a child protection contribution to protect other children when they could not afford to feed, clothe and educate their own children. Recommendations The following recommendations arise from the study: * simplification, by reducing the number of separate fees and contributions * clearer criteria for exemptions from fees and contributions * for utility charges such as electricity and water, relating fees as far as possible to household consumption, to encourage efficiency of use * for other charges, restructuring the charging system so that overall the fees and contributions system becomes significantly more progressive * ensuring transparency of how individual household assessments for fees and contributions are calculated, of the total amounts raised and the way they have been spent through enforcement of Decree 29, possibly raising this to the level of an Ordinance, to ensure it is fully observed 50 Annex C - Commune-Level Fees and Contributions * strengthening the implementation of Decree 225, which provides a mechanism for households to raise questions about commune financial issues * establishing a mechanism for households to seek further clarification from appropriate authorities outside the commune, if these questions are not satisfactorily answered. 51 Vietnam Public Expenditure Review 2000 - Annexes Appendix C.1: Fees and Contributions Levied in Six Study Communes (excluding education) Ha Tinh Province Quang Binh Province Ha Tay Province (Can Loc District) (Bo Trach District) (My Duc District) Thuong Loc Nhan Loc Van Trach Hoan Trach Hop Tien Phu Luu Te Duty for public work 15 days/W/y 1O days/w/y 1OOOOOVND/w/y 10 days/w/y 10 days/w/y 20kg/w/y Irrigation fee 12 kg/sao/y 12kg/sao/y 25 kg/sao/y 25kg/sao/ y 20kg/sao/ y 20kg/w/y Plant-cattle protection fee 3 kg/sao/y 6 kg/sao/y 6kg/sao/ y 6kg/sao/ y Field forest protection fee 2 kg/sao/y 2kg/sao/y Defense & Security 6 kg/W/y 2 kg/w/y 2 kg/w/y 2 kg/w/y 2 kg/w/y 6kg/household ly Defence militia 8kg/w/y Natural disaster 1,5kg/W/y 1 kg/w/y IOOOVND/w/y 5-l1 OOOOVND/h/y 2kg/w/y "Den on dap nghia" 2 kg/W/y 2 kg/w/y 5000VND/w/y 3000VND/W/y 5000VND/h/y Voluntary Child protection 1,5kg/W/y I kg/w/y 1OOOVND/capita/y 1 kg/sao/y Study encouragement 1 kg/W/y lkg/w/y 1 OOOVND/capita/y 24000VND/p/y Social evil fund 1,5kg/W/y Elderly people assistance fund 3000VND/h/y 2000VND/h/y Commune budget 6 kg/W/y 4 kg/capita/y School construction 7 kglh/y 6000VND/capita/y 35000VND/P/y 12kg/h/y Village fund 6kg/sao/y 5 kg/sao/y 7 kg/ sao/y Commune basic infrastructure 5 kg/W/y 5000VND/W/y District basic infrastructure 3 kg/W/y Village basic infrastructure 2kg/W/y 50009/W/y 1 OOOOS/W/y Electrical station 30000®/h/y 52 Annex C - Commune-Level Fees and Contributions Ha Tinh Province Quang Binh Province Ha Tay Province (Can Loc District) (Bo Trach District) (My Doc District) Thuong Loc Nhan Loc Van Trach Hoan Trach Hop Tien Phu Luu Te Construction of secondary levee system 5kg/sao/y 0,9kg/sao/y Cooperative management 4kg/lOOkg SL 4kglsao/y 4kg/capita/y 6kg/sao/y 4,5kg/sao/y Dyke construction 2kg/h/y Construction of Concrete canal 12 kg/sao/y 14kg/sao/y Tertiary and on farm Irrigation fee (Thuy 4 kg/sao/y 3 kg/sao/y 7,4kg/sao/y 6kg/sao/y 4kg/sao/y 2kg/sao/y loi "Noi Dong") Socio-cultural fund 4kg/capita/y 8kg/sao/y Transport fund 3kg/capita/y Poor woman assistance fund IOOOOVND/ IOOOOVND/ 5000VND/ 2000VND/woman/y 5000VND/woman/y IOOOOVND/ woman/y woman/y woman/y woman/y Pest eradication 1,5kg/sao/y 2kg/sao/y War cemeteries maintenance 10 kg/h/y Support to central provinces (suffering 5-lOOOOVND/h/ 5000VND/h 2000VND/h 2000-5000VND/h 5000VND/h from flood) Contribution to construction of the 50000VND/capita/y Church Lighting fee IOOOVND/h lOOOVNDlh/month 500VND/h/month 1OOOVNDlh/month /month Note. w- working person, h-household, sao- is a measurement of area and equal to 360m2, p-pupil 53 Annex D - Agricultural Sector ANNEX D - AGRICULTURAL SECTOR Role of Agriculture in Vietnam's Economy 1. Vietnam's agricultural sector provides approximately one-quarter of the country's GDP, generates more than a third of its exports and employs over two-thirds of its labor force. About 80 percent of the population and 90 percent of the poor live in rural areas and 45 percent of the rural population still lives below the poverty line (World Bank 1999a). 2. The gross value of agricultural production in 1997 was VND 88,663 billion in 1994 constant prices (see Appendix Table D. 1). Agricultural production is dominated by rice. Although its share has been gradually declining as farm households diversify their production base, rice still accounts for almost half of the gross value of agricultural output. Other food crops account for another 15 percent, industrial crops for 16 percent and livestock for 17 percent. Agricultural production hit record highs in 1998 and 1999 with total grain production increasing by about 2 million tons per year to reach about 34 million tons in paddy equivalent in 1999. Vietnam has moved from being a net food importer in the mid 1980s to being the second largest exporter of rice in the world. Exports of rice have increased from about 1.4 million tons in 1989 to 4.5 million tons in 1999. 3. The agricultural sector has been performing well despite the latest slow down in the national economy. The average annual growth rate in the past five years was between 4 and 5 percent fuelled by diversification into high-value crops (e.g. coffee, cashew nuts, rubber), increases in rice productivity and expansion of cultivated area including irrigated rice. As a consequence, agricultural income rose by 61 percent between 1993 and 1998 and was the main source of poverty reduction in rural areas (World Bank 1999a). During this same period rural income from non-farm enterprises increased by 30 percent while income from wages and other sources declined slightly. In 1998 average annual income of a rural household was estimated at VND 9.84 million (or about US$ 700) (World Bank 1999a) with income from agriculture providing about 47 percent of total income, non- farm enterprises 19 percent and wage and other income 34 percent. 4. Despite these achievements the gap between rural and urban incomes widened between 1993 and 1998, mainly because urban incomes rose twice as rapidly as rural incomes (60 percent versus 30 percent) (World Bank 1999a). Although the Government of Vietnam (GOV) has a stated goal of rapid "industrialization and modemization," I it 'Based on excerpts from the Government's Report presented by the Prime Minister at the 4h Session of the 10"' Assembly on October 28"', 1998. In addition, the Public Investment Program (PIP) 1996-2000 stated that the goal of the GOV is to reduce the agricultural share of GDP from 29 percent of GDP in 1995 to 19- 20 percent in 2000. This was based on a GDP growth scenario of 9-10 percent, which was not achieved due to the slow down in the economy in the past 3 years. Therefore, this structural shift will not be attained by 55 Vietnam Public Expenditure Review 2000 - Annexes recognizes the important role of agriculture in the sustainable development of the country and as an engine for broad-based poverty reduction and income growth. In particular the GOV considers agricultural processing and off-farm rural service activities as an important way of achieving its long-term development goals. Public and Private Roles in the Agricultural Sector 5. Determining the roles of the private and public sectors in agriculture is a difficult task, particularly in transition economies such as Vietnam that are undergoing rapid structural change. Public spending in agriculture, if properly managed, can have positive impacts on agricultural productivity and can provide beneficial spill-over effects, inducing more private sector investment. For example, long-term investment in research, extension, irrigation and rural infrastructure seem to have contributed significantly to the productivity growth of several Asian countries (Fan and Pardey, 1998). In India in particular, Fan, Hazell and Thorat (1999) find that public spending on rural roads, and agricultural research and extension have had the largest impact on agricultural productivity growth and poverty reduction compared to other types of investments. 6. On the other hand if public spending is poorly targeted it can crowd-out private investment and result in large inefficiencies and a misallocation of resources hindering the growth of the sector. Examples would include subsidizing inefficient crop sub-sectors or investing in marketing and processing activities that can be more effectively managed by the private sector. 7. From an economic efficiency point of view, there are four main justifications for government involvement in the agricultural sector: i. Public goods, which have two key characteristics-"nonexcludability8" and "'nonrivalness.9 In other words they are goods which one cannot exclude other consuners from using and consumption by one person does not preclude consumption by another person. The private sector will invest in public goods below the socially optimal level because private returns tend to be lower than the overall benefits to society as a whole. Examples of public goods in the agricultural sector include basic research, extension, large irrigation systems and rural infrastructure. ii. cases of market failure, due to high transaction costs and poorly developed legal institutions. This occurs often in rural credit markets in developing countries where the absence of adequate land ownership rights, lack of contract enforcement mechanisms, as well as poor information on credit worthiness mean that private institutions are reluctant to lend to farm households that do not possess a collateral. Therefore the government could intervene to develop institutions that would reduce these transaction costs. the set date. In the Party Congress of 1996, the GOV stated that in 2020 Vietnam should become an industrialized country. 8 Goods are excludable when the consumer will be excluded from consuming them if he or she does not pay for them. 9 Goods are rival when their consumption by one person precludes consumption by another person. 56 Annex D - Agricultural Sector iii. redistribution measures or safety nets for the poor, or for those who may be temporarily hurt by reforms, natural disasters, and personal risks such as illness and accidents. iv. provision of a legal and regulatory framework that facilitates private sector investment. This would include a functioning judicial system, regulations concerning market conduct (e.g. anti-trust law, contract law, consumer protection law, bankruptcy law, property rights law), etc. Public Expenditures in the Agricultural Sector The Goverdment?s Priorities in the Agricultural Sector 8. According to the Public Investment Program (PIP) (1996-2000), there are some weaknesses in the agricultural sector, which include (i) agricultural output remains vulnerable to natural uncertainties (e.g. resulting from climate and from insect infestation); (ii) product quality remains low; (iii) there is limited adoption of technology in certain areas and sectors (particularly in remote and mountainous areas); (iv) infrastructure and agricultural services are still weak (the rural transport system is particularly weak in mountainous and remote areas, limiting access to markets); (v) irrigation is limited mainly to rice production (80 percent of which is irrigated) - only 20 percent of industrial crops are irrigated; (vi) rural electrification has a limited coverage; and (vii) rural capital markets are very underdeveloped and agricultural credit is still limited. 9. To address some of these weaknesses, the GOV has stated that its future priorities in the agricultural and rural sectors will be in the following areas: * rural infrastructure (irrigation and flood control, rural roads, and electricity); - post-harvesting and processing activities; * reforestation and forest protection; * increasing agricultural productivity; * increasing the quality of many crops, especially export crops; and 10 * redistribution and poverty alleviation (ethnic minorities; mountainous areas)' 10. Several of these priorities go beyond the boundaries of agriculture and relate to overall rural development. Recent initiatives such as the 5 Million ha Reforestation Program, the Hunger Eradication and Poverty Reduction Program (Decree 133/1998), '0 Based on Excerpt from the Government's Report presented by Prime Minister Phan Van Khai at the 4th Session of the Ioth National Assembly on October 28th, 1998, and on the Report on the Review of the Recommendations from Constituencies - Presented by Mr. Vu Mao, member of the NA Standing Committee, to the deputies of the 10th NA at its 4th Session on 1Oa' November, 1998, as well as the 1996- 2000 PIP. 57 Vietnam Public Expenditure Review 2000 - Annexes and the 1715 Poorest Communes program (Decree 135/1998), translate these priorities into ambitious programs that will have a large impact on public spending. 11. The following sections focus on the level, composition, and effectiveness of public expenditures in agriculture: Level and Composition of Public Expenditures in the Agricultural Sector 12. As Table D.1 demonstrates, public expenditures in agriculture in 1994 constant prices almost doubled from 1992/1993 to 1997/98. This may reflect the renewed focus of the GOV on agriculture and the rural sector. However, public expenditures on agriculture absorb between 5 and 6 percent of the total government budget (6.3 percent in 1998). This is lower than for other South and East Asian countries such as China, India and Thailand, where the budget share for agriculture was between 8 and 16 percent in 1990- 93 (Fan and Pardey, 1998 - Table 13, p. 66 and Fan, Hazell, and Thorat, 1999 - Table 1, p. 8). As a share of GDP, public expenditures on agriculture are about 1 to 1.5 percent, even though agriculture contributes about one-quarter of total GDP. 13. As a share of agricultural GDP, agricultural public expenditures constitute between 3 and 6 percent of total agricultural output which is again lower than for other East and South Asian countries which averaged 8 to 9 percent in 1990-93 (Fan and Pardey, 1998 - Table 10, p. 64). Again in 1990-93, average government expenditures on agriculture per unit of agricultural labor in East and South Asia were about $130 in 1985 international dollars (Fan and Pardey, 1998 - Table 11, p. 65) compared with about $67 in 1989 international dollars in Vietnam. 14. These comparisons do not necessarily constitute an argurnent for a higher level of public spending in agriculture. The performnance of the agricultural sector is related also to public expenditures in transport, education, and health. To justify a higher level of public expenditures, one would also need to show that agricultural spending has a higher marginal return than other types of spending. This exercise would require a substantive study that is beyond the scope of this paper. 15. It is noteworthy that Vietnam's agriculture has also performed extremely well when compared to other countries in the region. With the exception of China, Vietnam's agricultural growth has been the strongest among a wide range of developing (and developed) countries (see Table D.2). 16. Although the GOV has declared agriculture a priority sector until recently this priority has not been fully reflected in public investmnent in agriculture. For example, in the 1996-2000 PIP, the agricultural sector was allocated 23 percent of total state investment (see Appendix Table D.2), but actually received much less around 12 percent in 1997 and 15 percent in 1998. However, since 1998 Vietnam has significantly increased public investment in agriculture and the preliminary figure for 1999 is 28 percent of total state investment. 58 Annex D - Agricultural Sector Table D.1: Public Expenditures in Agriculture 1992 1993 1994 1995 1996 1997 1998 Public Expenditures in Agriculture (in current VND billion) 1.125 1.283 2.238 3.495 na 3.712 4.591 Capital Expenditures 861 896 1.690 2.782 2.455 2.709 3.493 CurrentExpenditures 264 387 548 713 na 1.003 1.099 Public Expenditures in Agriculture (in 1994 constant VND billion) 1.358 1.480 2.238 2.884 na 2.567 2.855 Capital Expenditures 1.039 1.033 1.690 2.295 1.742 1.873 2.172 Current Expenditures 319 446 548 588 na 694 683 As a percentage of the State Budget 5.4% 4.5% 6.1% 7.2% na 5.2% 6.3% As a percentage of total GPD 1.00/O O.90/0 1.3% 1.6% na 1.2% 1.3% As a percentage of agricultural GDP 3.00/o 3.1% 4.6% 5.6% na 4.6% 4.90/o Per agricultural person (in 1994 VND) na na na 57.113 na 49.221 54.020 Per unit of agricultural labor (in 1994 VND) na na na 119.590 na 103.465 112.437 Sources: Data from 1992 to 1995 are from UNDP (1996), several tables. Datafrom 1996 arefrom the World Bank (1999b) - Table 5.5 Datafor 1997 and 1998 arefrom the Ministry ofFinance. Table D .2: Average growth of agricultural sector in various countries Average Growth Rates (1990-97) Low income 2.5 Lower middle income 1.3 Upper middle income 1.7 Low & middle income 1.6 East Asia & Pacific 3.7 Thailand 2.4 Indonesia 2.8 Philippines 1.8 China 4.6 India 2.5 Vietnam 4.4 Source: World Bank Development Indicators CD 59 Vietnam Public Expenditure Review 2000 - Annexes The composition of public expenditures in agriculture (Table D.3) 17. The most important sub-sectors for agricultural public expenditures in 1997 and 1998 were irrigation (which absorbs 50 percent of the agricultural budget), two national programs on reforestation and land reclamation (14-17 percent), and forestry services and forestation (10 percent). Table D .3: Capital and Current Expenditures by Agricultural Sub-sector, 1997-1998 (in million VND) ; (shares of Total Expenditure appear below each entry) 1997 1998 Total Total Total Total Total Total Expenditure Capital Current Expenditure Capital Current Expenditure Expenditure Expenditure Expenditure Total 3,712,494 2,709,279 1,003,215 4,591,435 3,492,647 1,098,778 100% 73% 27% 100% 76% 24% Of which: Cultivation 130,725 72,758 57,967 189,373 111,623 77,750 4% 2% 2% 4% 2% 2% Animal husbandry 85,448 52,669 32,749 77,416 49,143 28,272 2% 1% 1% 2% 1% 1% Veterinary activities 65,970 17,898 48,072 68,917 18,114 50,804 2% 0% 1% 2% 0% 1% Forestry and related activities 315,235 104,775 210,460 363,703 130,682 233,021 8% 3% 6% 8% 3% 5% Forestation 48,815 26,459 22,356 85,863 75,806 10,057 1% 1% 1% 2% 2% 0% Fixed cultivation and related activities 270,289 194,414 75,875 188,721 66,188 122,533 7% 5% 2% 4% 1% 3% Irrigation and related activities 1,806,371 1,651,112 155,259 2,469,736 2,307,102 162,634 49% 44% 4% 54% 50% 4% National programs on reforestation and land Reclamation 642,702 489,576 153,126 642,702 489,576 153,126 17% 13% 4% 14% 11% 3% Services for cultivation and animal husbandry, including extension services 280,000 405,000 7,5% 9% Other national targeted program and expenditure 66,939 100,104 1,5% 1 % Source: Ministry of Finance (MOF) 18. On average, the capital and current expenditure mix is about three to one. Irrigation receives the lion's share of capital expenditures (about 50 to 80 percent) whereas current expenditures are more evenly distributed between various subs-sectors with the largest share going to forestry and related services (about 20 percent) 60 Annex D - Agricultural Sector 19. In 1998, current expenditures comprised for goods and services (about 64 percent), salaries and wages (around 15 percent), subsidies and transfers (around 18 percent), and other uses (3 percent). The composition of subsidies and current transfers 20. Subsidies and transfers amounting to VND 195 billion (see Appendix Table D.3) were allocated mostly to sedentarization (resettlement) programs (VND 65 billion), crops (VND 31 billion), livestock (VND 23 billion), and irrigation (VND 21 billion). Subsidies and transfers equaled 18 percent of total current expenditures and 4 percent of total public expenditures in agriculture. 21. A more detailed breakdown of these subsidies and transfers in terms of the beneficiaries, and their actual uses would be useful. The data indicate strong budgetary support for activities such as crops and livestock that might not constitute public goods. They also indicate high support for sedentarization (resettlement) programs which need careful management to avoid possible negative effects on the environment. The regional distribution of public expenditures in the agricultural sector 22. In 1997 and 1998, 56 to 60 percent of total agricultural expenditures were spent at the central level; the rest was spent by local government in the provinces. For irrigation, fixed cultivation and sedentarization (resettlement) , and the two national programs on forestation and land reclamation, a higher percentage is spent at the central than the local level. However for other sub-sectors, such as cultivation or animal husbandry, the largest is spent at the local level (See Appendix Table D.5). 23. As Figure D. 1 demonstrates, there seems to be a positive, albeit not very strong relationship between per capita agricultural GDP and per capita agricultural expenditures by the provinces"l. This suggests that the state budgeting process does not significantly contribute to redistributing expenditures among the provinces for alleviating poverty. For example in 1998 agricultural expenditures in Quang Binh were VND 35,875 per agricultural person (about $2.5 per person) whereas in Ho Chi Minh City (HCMC) agricultural expenditures were VND 187,600 per agricultural person (about $13.4 per person). This suggests that richer provinces spend more on agriculture than poorer ones. If the main objective of public expenditures in agriculture is to promote growth this might be appropriate. But if the main objective is to promote equity between regions, then provincial allocations are not fulfilling this role. However, even if the government allocates more funds to poorer provinces, these funds will not necessarily reach the poorest communes or districts. 24. Poor provinces, such as Quang Binh, are dependent on the state for a large share of their expenditures and resources and therefore have little control over inter and intra- sectoral expenditure allocation, except through voluntary contributions. Richer provinces such as HCMC, which generate surplus revenues above the budget targets, have more fiscal flexibility because they can allocate their surplus revenues according to their own l The tre'id is the same for 1997, therefore the chart for 1997 is not shown. 61 Vietnam Public Expenditure Review 2000 - Annexes priorities. Thus budgetary flexibility at the local level depends on the ability of the localities to generate independent revenues. This also means that the quantity and quality of agricultural and rural services vary significantly between provinces. Another important issue is that once the provinces receive their transfer from the state budget, the trickle down to the districts and communes is not very transparent. 25. However, to get a total picture account also needs to be taken of the distribution across the provinces of expenditures by the center which as discussed above cover mainly irrigation and forestry activities. This information is not readily available. 26. However, other mechanisms are currently in place to target poverty, taking into consideration regional disparities. The recently initiated 5 Million ha Reforestation Program, the HEPR (Hunger Eradication and Poverty Reduction) Program and the Poorest Communes Program tackle poverty in less favored regions through more broadly based, rather than just agricultural sector programs. Figure D.1: Per capita Agricultural Spending vs. Per capita Agricultural GDP by province, 1998 ('000 VND) 10 oo- _ -t__ _ _ _ _ - . I e0 to ,;60- E ,240 _ ____ 20L . fi______ _ ,-,20 f X~~~0 1o0 1.0 - __ ______ 0 500 1.000 1.500 2.000 2.500 3.000 3.500 per capita Agricultural GDP Source: Ministry of Finance Public Expenditures on Agricultural Research 27. The agricultural research system in Vietnam is very fragmented. There are about 30 agricultural research institutions, 18 of which are under the control of MARD. The rest are owned by commodity SOEs (rubber, tea, coffee, sugar, etc.) or are semi- independent institutions that receive government support but are also dependent on external sources of finance. Each institution has a fairly narrow research focus, small budgets and staff, and there is little coordination with other research centers. 62 Annex D - Agricultural Sector 28. Returns to investment in agricultural research are known to be very high worldwide. For example, Fan and Pardey (1998) show that public investment in agricultural research has accounted for about 20 to 40 percent of the agricultural production growth in 9 Asian countries. Investing in research is crucial to raise agricultural productivity and enhance competitiveness in world agricultural markets. In 1996 average paddy yields in Vietnam were still smaller than those in China or Indonesia (3.6 tons/ha versus 6.1 and 4.5 respectively) and Vietnarn still has potential to increase its crop productivity. 29. In 1996 the GOV decided to re-organize the national agricultural research system as follows: (i) keep or merge some institutions but continue their full state funding (both salary and research funds); (ii) move some institutions or centers to SOEs and gradually reduce direct state funding to these centers; and (iii) spin-off some institutes to become independent self-financing organizations within 5 years (if unable to become self financing, they would be dissolved). However, the re-organization scheme has run into several difficulties in identifying which centers to merge, and resistance from some of the institutions. 30. The level of state expenditure on agricultural research is low. As Table D.3 below shows, total public expenditures for agricultural research in 1997, 1998 and 1999 were about VND 80 billion. This is equivalent to 1.7 percent of public expenditures in agriculture and 0.08 percent of agricultural GDP. In comparison, China spends about 6 percent of its agricultural expenditures on research, and Malaysia, Pakistan and Thailand about 10 percent; other Asian countries spend at least 3 percent (Fan and Pardey 1998, Table 16, p. 70). Table D.4: State budget for agricultural research (in current VND million) Sub-sector 1997 1998 1999 Total Research Salary Total Research Salary Total Research Salary FoodCrops 20,434 2,691 14,288 23,016 6,368 15,340 22,701 6,551 15,289 Forestry 7,883 2,433 4,869 10,232 3,494 5,208 9,215 3,241 5.187 Fruitand Vegetable 4,168 930 2,958 4,564 1,270 3,244 3,162 1,912 1,151 Industrial and Export Crops 9,048 1,416 7,418 6,677 1,370 4,882 4,363 1,535 2,648 Livestockand Veterinary 6,421 2,454 3,967 8,498 4,253 4,245 8,015 3,615 4,400 Plant Protection and Land Management 5,112 1,718 3,084 6,120 2,267 3,318 6,438 2,235 3,317 Post Harvest Technology 2,572 825 1,047 0 0 0 1,501 1,484 0 WaterResources 10,248 5,273 3,065 11,660 5,882 3,223 10,201 4,396 3,360 Other Sub-sectors 7,656 2,285 4,194 13,613 4,202 7,141 12,340 4,204 6,532 Total 73,542 20,025 44,890 84,380 29,106 46,601 77,936 29,173 41,884 Source: MARD 31. The expenditures in Vietnam cannot sustain an effective research program to develop improved crop and breed varieties and increase the quality and marketability of Vietnam's agricultural exports, even taking into account the additional funding that research institutes can raise from local and other sources. As Appendix Table D.5 shows the central level funding of several research institutes was between 50 and 60 percent of 63 Vietnam Public Expenditure Review 2000 - Annexes total expenditures between 1996 and 1998; local sources contributed between 10 and 30 percent. 32. In 1999, the number of agriculture research staff was about 4,874. However, only 4,114 were approved for funding by the Ministry of Finance (MOF). More than half of agricultural research expenditures are used to cover salaries and other administrative costs. Total spending on salaries and current expenditures for research staff for 1999 was VND 43 billion, equivalent to VND 10.5 million per staff per year. The budget share for research activities proper is only around one third of total agricultural research (see Table D.4). Most of the research was devoted to food crops and livestock breeding, with little attention devoted to activities such as post-harvest research, plant protection, and horticultural products. Table D.5: Shares in Total Expenditures and Research Expenditures 1997 1998 1999 Share of Share of Share of Share of Share of Share of subsector research subsector research subsector research in total in in total in in total in research subsector research subsector research subsector budget budget budget budget budget budget Food Crops 27% 23% 28% 27% 30% 28% Forestry 10% 22% 12% 34% 12% 29% Fruit and Vegetable 5% 22% 5% 28% 4% 60% Industrial and Export Crops 11% 16% 8% 21% 5% 35% Livestockand Veterinary 13% 29% 15% 37% 16% 32% Plant Protection $ Land Management 7% 28% 3% 38% 4% 30% Post Harvest Technology 3% 32% 0% 2% 99% Water Resources 13% 51% 13% 50% 13% 43% Other Sub-sectors 10% 30% 16% 31% 15% 34% Total 100% 28% 100% 33% 100% 35% Source: MARD 33. The current agricultural research system in Vietnam needs both institutional and budgetary reform. State funds to agricultural research should be significantly increased. Because of its low base, such an increase would still be small in absolute terms when compared to the overall agricultural budget. 34. Increased public funding would also need to be accompanied by more rapid implementation of the reorganization program determined by GOV in 1996, to ensure that the additional funds are well used. 64 Annex D - Agricultural Sector Public Expenditures on Agricultural Extension 3 5. Despite a long history of agricultural extension in Vietnam, the national agricultural extension service was created only in 1993. The current national system consists of central, provincial and district level service organizations. Under the responsibility of the MARD, a Steering Committee for Extension evaluates and allocates funds for extension projects. Each province has an extension center with an average staff of 15 to 24 workers (see Table D.6). At the district level, district extension stations were created to provide training and demonstrations to farmers with an average of 1 to 6 workers per station. However not all districts have an extension station. At the commune level, village extension agents are hired on a contractual basis for various extension projects. 36. The national extension service suffers from a scarcity of staff (including well- trained extension agents) and limited resources. Only 70 percent of districts have an extension station and only 30 percent of communes have an extension unit. A total of 2,757 extension workers at the province and district level serve a farming population of 10 million households (see Table D. 6 below). 37. MARD expenditure on agricultural extension consists mainly of current expenditures to cover the operating and administrative costs at the central level and part of the operating costs at the local level. The salaries of extension officers at the provincial and district level organizations are paid from the provincial budgets, meaning that local spending on agricultural extension is largely dependent on the resources of the province. In fact most of the expenditures on agricultural extension occurs at the province level. Information on selected provinces indicates that around 75 percent of total expenditures on extension comes from local budgets (see Appendix Table D.7). Table D.6: The Public Agricultural Extension Service Region Number of Average staff per Number of Average staff per Provincial provincial center District Extension district station Extension Centers Stations NW 12 16.6 94 4.9 NE 4 23.3 27 2.3 RRD 9 22.5 50 1.4 NCC 6 21.0 61 4.6 SCC 8 17.8 29 2.5 CH 4 15.2 37 2.8 NES 6 20.5 24 4.3 MRD 12 28.6 55 6.2 Vietnam 61 23.1 337 4.0 Source: MARD 38. As Table D.7 shows MARD spending on agricultural extension has declined since 1997. In 1998 MARD spent VND 27 billion on agricultural extension, equivalent to 2.5 percent of total current expenditures in the agricultural sector and only 0.6 percent of total agricultural spending. It also represents about $0.2 (20 cents) per farm household. 65 Vietnam Public Expenditure Review 2000 - Annexes However in addition the provinces also allocate funds to agricultural extension. According to MOF total public expenditure on agricultural extension in 1998 was VND 405 billion. However reflecting classification problems, this figure is known to include expenditures on other activities and the exact amount relating to agricultural extension is unknown although it comprises the bulk of this amount. Table D.7: MARD expenditures on Agricultural Extension (in million VND - Current and Constant 1994 Prices) 1994 1995 1996 1997 1998 1999 Total Expenditures in 13,575 15,828 24,079 29,820 27,143 27,664 current prices Total Expenditures in 13,575 13,059 17,089 20,622 16,880 na constant 1994 prices Source: MARD 39. A disaggregation of the 1999 MARD extension budget by sub-sector is presented in Table D.8. This indicates more emphasis on the dynarnic areas such as livestock, industrial crops, and aquaculture. This allocation is not proportional to the respective shares of these sub-sectors in agricultural GDP, perhaps indicating a strategy that gives priority to those sub-sectors with higher potential. The limited amount of resources devoted to training is noteworthy. Also virtually absent is extension to promote marketing of agricultural products. 40. The relatively recent establishment of a national extension service means there is an on-going need to strengthen its capacity to transfer new technologies to farmers. The link with research institutes also appears weak and often competitive rather than cooperative. Table D.8: Extension Expenditures by Agricultural Sub-sectors in 1999 Sub-sector Share in the extension budget 22.3 % Food crops Industrial and fruit crops 23.6 % Livestock 18.8 % Forestry 11.3 % Aquaculture 10.1 % Information and training 8.8 % Other 5.2 % Source: MARD 41. The findings above suggest that as with to the agricultural research system, extension services in Vietnam need both institutional strengthening and increased spending. The institutional strengthening should eliminate the current practice of public extension staff to being involved in commercial activities such as input distribution to farmers, and involve participation of farmers and agro-industry representatives in extension advisory boards. 66 Annex D - Agricultural Sector 42. Extension services are key for the adoption of technology by farmers and to improving the productivity and returns of the farm sector. The extension budget should be increased at least to expand extension services in the districts and communes without an extension center, with priority given to the poorest communes. Public Expenditures on Forestry 43. The upland areas have long been seen as impoverished and "backward," but they are increasingly viewed by the central government as a storehouse of natural resources necessary for national economic development. The last decade has seen a proliferation of policies and programs related to natural resource management, especially forest, watershed and soil management (Nguyen 1998). Although these programs aim to contribute to national wealth and the well-being of those living in the uplands, it is important that they be managed to avoid natural resource degradation and to improve sustainable agricultural practices. 44. In 1993 the government initiated Program 327 "Re-Greening the Barren Hills", covering 7 million hectares, including 60-65 percent of the hills in the Northern Uplands and 25-33 percent of the land in the Central Highlands. This program was allocated US$68 million, a large share of central government transfers to the provinces. About 20 percent of these funds were earmarked for interest-free loans to households. The remaining 80 percent was for government investment in infrastructure, scientific and technical facilities, public services, reforestation, and initial support to settlers. By the end of 1998, the GOV had invested about US$270 million (VND 2,980 billion) in this program (GOV, 1999). Of this about 14 percent was interest-free credit and the remainder was "government direct investment". 45. Under the program the government planned to grant user rights to parcels of barren land and to protect certain tracts of remaining forest land. Parcels of forest land, usually forest under the control of state-owned forestry enterprises, were to be contracted to households. Average forest allocation was to be about 20-30 hectare per household and the government planned to pay VND 50,000 annually to households for each hectare of forest they were protecting. However, as a result of limited resources the implementation was incomplete. For example, in 1998 a survey team led by staff of the Center for Natural Resources and Environmental Studies of Vietnam National University found that Decree 327 had been implemented in only 2 villages out of 16 in Dakto Kan Commune in Kon Tum Province. The forest protection payments had initially been made in these villages in 1996, but not subsequently in 1997 or 1998. It was not clear whether in the absence of expected payments, the new "owners" of the forest land would begin to exploit the resource. 46. According to the World Bank (1995), some weaknesses of Program 327 included: * A lack of planning and prioritization on the basis of land use and economic returns. * Virtually all projects were on-going, selected for their readiness for implementation rather than by in-depth assessment of priority needs, constraints and institutional capacity. 67 Vietnam Public Expenditure Review 2000 - Annexes * Funds were spread over too many projects, and in many cases were not sufficient to achieve their objectives. * High administrative and operating costs (about 8 percent of the total). * State-directed approaches to project development and implementation were used with virtually no participation of affected units or households in resource planning. 47. In 1998 the government launched the ambitious Five Million Hectare Reforestation Program, absorbing Program 327. The main objective is to speed up reforestation activities, with the target of raising Vietnam's forest cover from 28 percent to 43 percent, the rate at the time of national independence. At the same time the program hopes to establish areas for production of fuel-wood for domestic consumption and raw materials for processing and export, thus raising incomes of people in the mountainous areas. According to MARD (GOV, 1998), implementation difficulties facing the program include: * Land use planning: The actual area of land to be forested has not been defined and it is unclear whether the land is available. Some of the land is already being farmed, and even land under management of enterprises or state forestry organizations is already being otherwise used. * The results of plantation planting in Vietnam are unclear and in some cases not efficient, partly because the timber grown has not yet found a market. It is not yet clear whether the processing capacity and the market for forest products is large enough to justify such a large forested area. * How to develop an appropriate incentive policy and mechanisms to achieve reforestation without excessive costs. The lesson from Prograrn 327 is that uncertainties still exist on how to provide equitable and effective forest protection incentives to local populations. 48. Draft recommendations for effective implementation of the program include that land use planning and allocation to different economic sectors should be done quickly; that the present forest contracting system should be clearly defined so people can enjoy benefits from protecting forest; and that land for reforestation should be determined based on the requirements of agricultural production; and allowance should be made for agro- forestry activities and diversified forest products (GOV, 1998). Public Expenditures on Irrigation 49. About 80 percent of the 7 million hectares of cultivated land area in Vietnam are equipped with some sort of irrigation, the majority of which is dedicated to rice. MARD is usually responsible for primary irrigation infrastructure, while local governments (provinces, districts and communes) are responsible for secondary and tertiary infrastructure. Large and medium scale irrigation works are managed by 172 Irrigation Management Companies (IMCs), three of which are under the control of MARD and 169 are under local government control. Together they employ a total of 20,000 staff. 68 Annex D - Agricultural Sector 50. As mentioned earlier, irrigation absorbs about 50 to 55 percent of the agricultural state budget, 80 to 90 percent of which is capital expenditures (see Table D.9). Total public spending on irrigation has almost doubled since 1992 but the share dedicated to recurrent expenditures has been shrinking markedly (from about 20 percent in 1992/93 to 10 percent in 1997/1998). This reflects that about 50 percent of operation and maintenance (O&M) costs are now covered by the IMCs, the expenditures of which are not included in the state budget (for an example of the costs and revenues of one IMC, refer to Appendix Table D.7). However, the allocation of the state budget to cover the remaining portion of O&M costs is not enough to effectively maintain the existing systems, resulting in deteriorating infrastructure and low capacity utilization. As described below and in Section 4 following, the imbalance between current and capital expenditures in irrigation reduces the efficiency of new investments. Table D.9: Public Expenditures on Irrigation (in billion VND, current and 1994 constant prices (in parentheses)) 1992 1993 1994 1995 1996 1997 1998 Total 558 814 1,542 1,804 1,806 2,470 Expenditures (674) (939) (1,542) (1,488) (1,249) (1,536) Current 102 191 302 288 155 163 Expenditures (123) (220) (302) (238) (107) (101) Capital 456 623 1,240 1,516 1,355 1,651 2,307 Expenditures (551) (719) (1,240) (1,251) (962) (1,142) (1,435) Sources. UNDP (1996), World Bank (1999b) and Ministry of Finance 51. According to the MARD, the total estimated value of irrigation works in 1994 prices was VND 60 trillion. The estimated regular O&M cost needs of the existing irrigation system are VND 1,200 - 1,500 billion per year (excluding costs of depreciation, major upgrades and repairs, and rehabilitation of damaged infrastructure due to natural disasters, which together would cost an additional one trillion VND). Irrigation fees collected from water users are about VND 600 billion per year. The shortfall between irrigation O&M costs and revenues from water fees is partially covered by state subsidies of about VND 22 billion per year that go directly to finance IMC losses. The rest of the funds needed are either collected through donor grants or more often than not irrigation systems are left to deteriorate. In many areas, especially poor ones, communes rely on voluntary contributions from farmers (either in labor or in cash) to maintain and repair tertiary canals in the villages. 52. Because basic IMC costs such as salaries, administration, and electricity have to be paid, under-funding of O&M costs results mainly in poor maintenance and under- utilization of existing capacity. According to the World Bank (1996), "of the 4 million ha of cultivated paddy (60 percent of agricultural land), 3 million ha are equipped with some kind of irrigation. However, due to incomplete systems, planning or design deficiencies, deterioration, lack of water and poor operations, only 2 million ha are actually irrigated." Discussions with several government officials also suggest that capacity utilization of the existing irrigation infrastructure is only about 60 percent. Capacity utilization in 69 Vietnam Public Expenditure Review 2000 - Annexes mountainous and remote areas is particularly low and cost recovery is lower than for the rest of the country. 53. Inadequate cost recovery and deteriorating infrastructure are a thus major concern. Although current cost recovery through water fees is less than 50 percent, Vietnam fares better than other countries such as Thailand and Malaysia, where O&M cost recovery is very low (the irrigation fee in Thailand is about $2 per ha per season compared to an average of about $30 in Vietnam). On the other hand cost recovery in the Philippines is high and pays for a large portion of capital investment (Abernethy and Heim 1999). 54. The basic guidelines for charging water fees were set in Vietnam in Decree # 112 (1984) at the rate of 3 to 8 percent of paddy yields (payable in paddy or in equivalent cash terms). The rates vary by type of irrigation and drainage system used (gravity or pumping), quality of service delivery and by season. The actual rates charged by the IMCs in each province are approved by the Province People's Committee (PPC) and vary quite substantially between regions. The level of cost recovery can differ markedly among IMCs as well; while most IMCs incur losses some have much higher levels of cost recovery than others'2. The level of the water fees among the IMCs, local governments and the communes also vary substantially by region. 55. According to MARD's experts, the irrigation fee would have to be doubled just to cover the current needed level of O&M costs. Meanwhile most depreciation costs and the cost of investment in new construction works or pumps as well as major rehabilitation and repairs are provided as a subsidy by the state. There is scope for improving cost recovery and system performance in Vietnam without necessarily increasing the water fee, which could be detrimental to poor farmers,. 56. There are four major sources of inefficiencies with the current system that if addressed could substantially improve cost recovery. (1) The official fee is often not fully paid by the farmers or the communes and districts do not pay the full amount owed under contract to the IMC. Many farmers are not willing to pay the official fee because they complain that the systems are unreliable and the quality of service delivery is very poor. Some IMCs are even reluctant to sign contracts where farmers could pay the fee in advance because they fear they will not be able to deliver the amounts requested by farmers. (2) The irrigation fee is based on the volume and type of crop production'3 rather than on the volume of water use, encouraging excessive and inefficient water usage. The fee 12 According to a MARD official, a pumping station in Ha Giang Province was recently sold to a private company. In 1998, this company turned in substantial profits by raising the water fee and improving its fee collection. This could be an example to set for other IMCs which could be privatized or their management contracted out to private enterprises. 13 The water fee for other crops (such as vegetables or sugar cane) is priced at 40 percent of the fee charged for paddy production. 70 Annex D - Agricultural Sector for other crops (such as vegetable or sugar cane is set at 4 percent of the fee charged for paddy production). (3) Related to issue #2, commodity targets, under which farmers are encouraged to grow certain crops such as paddy through increased public investment and support, may prevent the allocation of water to those crops with the highest returns to water use. Many farmers complain that they cannot afford current water prices, mainly because they are getting little returns from their water use. Low crop profitability reduces the willingness of farmers to pay. (4) The water fee is collected in paddy or its equivalent in cash. The PPC determines the price at which measures of rice are converted to cash and sets the price at 10 percent below the market price to encourage payment in cash. This means that the fee varies with the production and the price of paddy rather than being determined by the supply and demand for water. 57. In short, the irrigation system in Vietnam suffers from both financial and organizational problems. On the financial side, several measures could be considered. First, the irrigation fee should be based on volume of water used rather than on paddy production14. Second, the water fee should be paid in cash and not in paddy. Third, commodity targets should be reconsidered, as they limit the choice of farmers to allocate water resources in the most efficient way and reduce their willingness to pay for water. Fourth, the GOV should perhaps focus its investments in irrigation on rehabilitation and repair of existing systems rather than on construction of new systems which cannot be maintained in the long-term. Further deterioration in the irrigation system will impose a heavier budgetary burden on the GOV in the long-run. 58. Any necessary increase in the water fee should only be instituted once the current collection system is improved and once stronger institutions or organizational changes are put in place to provide better services. 59. On the organizational side, institutional reform is needed so that the primary beneficiaries participate in and contribute to operation and management. At present there is no incentive for the farmers to pay their fee in full, nor for the local authorities or the IMCs to cover all their O&M costs. An efficiently functioning system would require linking water users rights and responsibilities with the benefits and costs of the system. One way to do this is to have water user associations, where farmers are responsible for financing, operating, and maintaining the system and where the benefits from the schemes accrue to them directly. 14 In one province where irrigation fees were based on usage, farmers reduced water use considerably and costs were fully recovered, resulting in benefits to both farmers and the IMC. One implementation problem sometimes posited is installing and supervising water meters to measure the volume of water used. However, technical staff of the MARD suggest that consumption rates can be approximated through the calibration of sluices or through the volume of fuel used in case of electric water pumping. 71 Vietnam Public Expenditure Review 2000 - Annexes 60. Water user organizations have already been successfully implemented in parts of Vietnam and elsewhere in the world and could be replicated throughout the country. For example in Tuyen Quang Province irrigation systems are now successfully managed and maintained by a water user association which took over the responsibilities of the previously existing IMC. Similarly, a pilot project financed by the Asian Development Bank (ADB), under which water user associations were developed to manage the irrigation systems of North Nghe An and Son Chu, has received very high performance ratings. These schemes have been highly effective in increasing farmers' productivity through more reliable water supply, in reducing maintenance costs, in increasing collection of water fees, in improving water use and irrigation efficiency, and in reducing subsidies required from the state. Public Expenditure in Agricultural Production and Processing Activities 61. Rural farrn households provide the bulk of agricultural output in Vietnam. However agricultural processing and marketing activities are largely dominated by SOEs. The central and local governments own over 1,400 enterprises in the agricultural sector (excluding the IMCs) (See Table D. 10). Although the number of agricultural SOEs has declined since 1996 they still represent about a quarter of all SOEs. In addition, according to the 1996-2000 PIP (see Appendix Table D.2) the total projected investment of VND 95.5 trillion in agriculture would be provided by a mixture of state investment (23 percent), SOEs' investment (45 percent), household investment (25 percent) and FDI (7 percent). This implies that SOEs are still expected to play the major role in agricultural investment. The question is whether the investments by SOEs are money well spent or whether it would be more efficient to use these funds on other activities. Table D.10: Number of SOEs in Agriculture and Forestry 1996 1997 1998 1999 Total # of SOEs 6,200 5,816 5,718 5,462 of which: SOES in 1,591 1,538 1,454 1,424 agriculture & forestry (excluding IMCs) Source. Ministry of Finance 62. SOEs in the agricultural sector are involved in a variety of activities including the production and processing of sugar, rubber, coffee, tea, seeds, breeds, fertilizers, etc. It can be argued that many of these activities could be carried out by the private sector. However, an appropriate environment for such private operations is not yet available. 63. While these SOEs no longer receive direct budget subsidies, they receive indirect subsidies in various forms such as loans at concessionary interest rates, debt forgiveness, tax exemptions, and preferential access to land and infrastructure. For example provincial fertilizer companies may receive interest rate subsidies on outstanding loans, exemptions from depreciation costs, subsidies to transport fertilizer to remote areas and subsidies to store fertilizers to support the stability fund. Overall there is a lack of detailed information concerning the amounts and purposes of such indirect subsidies. 72 Annex D - Agricultural Sector 64. Table D. 11 shows state budget transfers to support certain SOEs. For example, fertilizer importing SOEs received an interest rate subsidy of VND 22 billion in 1998, an amount almost equal to the total MARD expenditure on agricultural extension in 1998. Total subsidies in 1997 for the three types of SOEs shown below were higher than MARD expenditure on agricultural research. State subsidies to SOEs are not always transparent and may be hard to trace and measure. For example the sugar industry is indirectly subsidized through local policies favoring conversion of land to sugar cane production and restrictions on imported sugar (which is cheaper than locally produced sugar). According to analysis by Goletti and Rich (1998) the government's protection of the sugar industry costs society about USD 92 million per year, mainly through higher consumer prices. Table D.11: Budget Support to SOEs (million VND) Type of Subsidy 1996 1997 1998 1. Interest rate subsidies for food 67,400 30,250 18,310 2. Interest rate subsidies for fertilizer 21,687 50,160 21,629 3. Payment for maintaining national 6,000 9,000 11,000 hybrid livestock and poultry Total 95,087 89,410 50,940 Source: MOF, Department of SOEs 65. Evidence from different studies suggests that many of the SOEs in sugar, rubber, or beer are loss-making and are inefficient. In the sugar industry for example, many refineries are uncompetitive and are located in areas where not enough sugar is being produced, resulting in under-utilized capacity. The GOV itself recognizes that "in the countryside, agricultural and industrial facilities operate inefficiently (sugar companies, cement from blast furnaces, tunnel brick making, beer companies, state-owned farms and state forestry farms)" (Socialist Republic of Vietnam, 1998). According to the IMF of the 50 most-indebted SOEs out of a sample of 1,044 state enterprises, food enterprises accounted for 24 percent of their total debt. Out of 5,429 SOEs studied by the MOF, about 40 percent were classified as profit-making enterprises, 44 percent were temporary loss-makers, and 16 percent were permanent loss-makers (IMF 1999). Within the central SOEs engaged in agriculture, about 17 percent were making losses in 1998 (see Table D.12). The average loss of these unprofitable SOEs was VND 8.4 billion, while the average profit of the profitable SOEs was VND 1.3 billion. This suggests that large losses made by a few SOEs are more than offsetting the profits made by the majority of SOEs. 73 Vietnam Public Expenditure Review 2000 - Annexes Table D.12: Information on Agricultural SOEs Owned by the Central Government Unit 1996 1997 1998 1- Total of enterprises No 476 472 497 1.1- Number of Public enterprises No 14 18 49 1.2 Number of Profitable enterprises No 315 351 326 1.3 Number of Self-Sufficient enterprises No 24 26 38 1.4- Number of loss-making enterprises No 123 77 84 2. Commercial capital bill. VND 10,741 12,438 12, 794 2.1- Capital from state budget bill.VND 5,152 5,770 7,205 2.2- Owned capital bill.VND 2,218 2,510 1,395 2.3- Loan bill.VND 2,973 3,780 3,826 2.4- Other bill.VND 398 378 368 3. Workers Persons 177,754 188,658 221,032 4. Revenues and Profits 4.1- Revenues bill.VND 25,129 28,437 30,390 4.2- Profits bill.VND 989 661 420 4.3- Losses Bill.VND (673) (695) (708) J. Contribution to state budget bill. VND 1,738 1,425 1,277 5.1-Turnovertax bill.VND 338 381 397 5.2- Profit tax bill.VND 347 200 118 5.3- Import-export tax bill.VND 578 464 377 5.4- Tax on Special commodity bill.VND 39 45 4 5.5- Tax on capital bill.VND 112 104 111 6. Debts bill.VND 9,947 12,535 14,943 6. 1- Debt to state budget bill.VND 636 667 661 6.2- Debt to the banks bill.VND 5,681 7,825 9,627 7. Amount owed to SOEs bill.VND 5,853 7,444 8,189 Source: MOF, Department of SOEs 66. Table D. 12 sets out information on agricultural SOEs owned by the central government. It can be seen that SOEs are still a significant source of revenue to the state. For example, in 1998 budget revenues from agricultural SOEs amounted to almost VND 1.3 trillion. However, the net contribution to society in supporting SOEs in a variety of ways including price support to farmers-net of efficiency losses to society and net of the various subsidies received by these SOEs, is an issue which. needs to be addressed in future studies. 67. Support given to loss-making SOEs also results in an uneven playing field between private enterprises and SOEs. This discourages the private sector from investing in areas where SOEs are dominant and reduces the incentives for SOEs to operate efficiently. A cost-effective means for the GOV to increase the levels of investment in the rural sector would be to provide a conducive environment to private sector investment. While some SOEs are now being equitized, privatized or liquidated, new SOEs are being formed, for example being in the sugar industry. In general, allocation of state and local funds to agro-food processing SOEs crowds out capital to small and medium scale rural enterprises. For example, the Vietnam Bank for Agricultural and Rural Development 74 Annex D - Agricultural Sector portfolio of medium and long-term outstanding loans to the corporate sector in 1998 was VND 1,506 billion. Of this amount, 85% was allocated to SOEs and only 15% to non- SOEs (see ANZDEC-IFPRI, 2000). 68. A common argument in support of SOEs is that they help generate employment. However SOEs in the rural sector do not seem to alleviate rural unemployment to any significant extent. Small and medium-scale private enterprises (SMEs) are usually more successful at absorbing rural labor and reducing rural-urban migration. According to the World Bank (1998) the labor to capital ratio of a private firm is 10 times that of a SOE. In addition a job created in a SME requires a capital investment of about $800 compared to $18,000 in a SOE. Greater encouragement of SMEs could thus be a major means of increasing off-farm employment in rural areas. Agricultural Expenditures and Gender Equity 69. The analysis of public expenditures in agriculture does not immediately lend itself to a assessment of the differential impact on men and women. First, public expenditures on agriculture are not reported by the type of recipient (poor or non-poor, male or female, etc.), neither are revenues (such as water fees) disaggregated by the type of payer. Second, little information is available on decision-making or allocation of resources within households, which makes it difficult to analyze the impact of public expenditures on individual household members. Although one can usually infer the gender implications of public spending by looking at whether some sub-sectors are traditionally dominated by either sex, in Vietnam gender bias in the agricultural sector appears minimal. For example, 70 percent of the female labor force and 68 percent of the male labor force work in agriculture and related services (VLSS 1998). And despite the fact that the hourly farm wage rate of a female is lower than for a male, the mean per capita income of a female-headed household in rural areas is 17 percent higher than for a male- headed household (VND 2.643 million versus VND 2.264 million). In the general economy, the female unemployment rate is lower than the male one (4.5 percent versus 6.2 percent) and poverty measures are lower for female-headed households than for males (World Bank 1 999a). 70. However, the analysis of other evidence suggests that two aspects of public investment in agriculture, extension and agro-industry, could be improved in order to promote gender equity. Concerning extension, it is reported that the existing public extension services are often targeted to men. This has important implications for the access to and adoption of modern technology by women in rural areas. As documented elsewhere (see ANZDEC-IFPRI 2000), "doi moi" has been accompanied by a "feminization" of agriculture, a process by which a greater number of women are engaged in agriculture, while men often migrate to urban areas or engage in non-farm activities. In spite of the increasing importance of women in Vietnam's agriculture, the extension service is still mainly oriented towards communication with men. An appropriate redesign of extension programs to target women would be appropriate. 71. Concerning agro-industry, the previous section has shown that a large share of public expenditures in agriculture is devoted to SOEs engaged in agro-processing. These are typically capital intensive enterprises. On the other hand, SMES are more labor 75 Vietnam Public Expenditure Review 2000 - Annexes intensive and female employment in this sector represents a large share of total employment. Therefore, by favoring capital intensive large SOEs, public expenditures may restrict the growth of SME and thus the growth of female employment. Budget Decision-Making and Institutional Issues 72. As in other sectors the Ministry of Planning and Investment (MPI) plans a key role in allocating investment expenditures to the agricultural sector, while recurrent expenditures are allocated by the Ministry of Finance (MOF). The coordination between these two ministries and MARD is not always transparent. This lack of coordination may reinforced the imbalance between capital expenditures and O&M funding. 73. Weaknesses in project preparation and evaluation referred to in Annex A of this report such as poor project formulation, over- optimism about benefits and implementation capacity and lack of adequate evaluation methodology are also reflected in agricultural expenditure. 74. MARD has little access to information on agricultural budget allocations or actual spending by the provinces which limits its ability to develop and monitor national policies. At the provincial level the departments of agriculture have little influence on or information about budget allocations or actual expenditures. This means that agricultural sector spending at the central and local levels are not well integrated and local spending in agriculture does not necessarily reflect national priorities set by the MARD. Moreover, the budget negotiation process is far from transparent and can result in an inequitable distribution of state expenditures among provinces. A more systematic formula mechanism for budget allocations would result in a more transparent and equitable system. 75. Another important issue is insufficient transparency in reporting and publishing public expenditures. For example, public expenditure data on agriculture are not disaggregated by function-such as extension, research, land reclamation, etc., agricultural SOEs accounts are not always audited, and information about transfers or subsidies to individuals or SOEs is not disaggregated by the type of beneficiary and the product covered. There is also lack of consistency of the data between different ministries, between the local and central levels, and between different departments of MARD. A set of reform options 76. Despite tremendous achievements in the agricultural and rural sectors, changes in Vietnam's public expenditures on agriculture could lead to even higher performance. * Increasing expenditure on agricultural research and extension. The GOV could consider significantly increasing the agricultural research budget. The extension budget should be increased to the extent at least to expand extension services in the districts and communes without an extension center, with priority given to the poorest communes. This increase could be reflected in the new 2001-2005 PIP and the 2001 76 Annex D - Agricultural Sector annual budget. The budget increase should be accompanied by mechanisms ensuring strong linkages between research and extension, the involvement of farmers and agro- industry representatives, and priority being given to services for the poorest communes. This additional expenditure could be funded by reallocating public spending from various forms of budget support to agricultural SOEs. * Base the irrigation fee on water consumption. Currently, water fees are based on paddy production levels rather than volume of water used. This results in excessive water use, problems in collecting the fee, and insufficient cost recovery. Basing the fee on water consumption has already been successfully implemented in some localities in Vietnam and could be extended to other areas. At the same time the system of commodity targets and collection of fees in paddy as well as cash should be changed. * Extend pilot projects to develop water user associations. The benefits of water user organizations in terms of water use efficiency, cost recovery, and improvement in service delivery have been proven in Vietnam and elsewhere. The GOV could extend the pilot programs developed in Tuyen Quang Province and North Nghe An and Son Chu to other areas within the next two or three years. Eventually the whole irrigation system in Vietnam could be managed by water user associations. Future Research Needs 77. Based on the above analysis, there are four main areas that need further investigation. These topics were not addressed thoroughly in the current report due to lack of data and the limited time-frame. * Further in-depth research is needed to evaluate the impact of public expenditures in the agricultural and rural sectors on the poorest segments of the rural population. For example, the performance of the Hunger Eradication and Poverty Reduction Program (HEPRP) in reducing rural poverty in rural areas might be examined. * The gender impact of public expenditures in agriculture needs more analysis and requires linking household level data with macroeconomic data. * A study on the performance of the SOEs in the agricultural and rural sectors is also warranted, covering the net contribution of these SOEs to the state budget and their impact on the economy. * Rural credit plays a very important role in the agricultural and rural economy and could be a major factor in meeting consumption, income and investment needs. A separate study on public expenditures and rural credit would be a useful addition. 77 Vietnam Public Expenditure Review 2000 - Annexes Bibliography Abernethy, C.L. and F. Heim (eds). 1999. Irrigators' Organizations: Government Actions Towards Effective Irrigators' Organizations with Special Reference to Lao PDR and Vietnam. Deutsche Stiftung fur Internationale Entwicklung (DSE). ANZDEC-IFPRI. 2000. ADB Agricultural Sector Program TA 3223. Inception Report: Current Status and Past Performance of Agriculture. Phase I Technical Report, March 2000. Desai, J. 1995. Vietnam through the Gender Lens: An Empirical Analysis Using Household Survey Data. Hanoi, Vietnam. Excerpt from the Government's Report presented by Prime MinisterPhan Van Khai at the 4th Session of the 10t' National Assembly on October 28th, 1998. "Socio-Economic Development in 1998 and Major Orientations for Socio-Economic Development in 1999." Vietnam 1998/1999. ThA Gioi Publishers, Hanoi. Fan, S. and P.G. Pardey. 1998. "Government Spending on Asian Agriculture: Trends and Production Consequences" Chapter 3 in Agricultural Public Finance Policy in Asia. Tokyo, Japan: Asian Productivity Organization. Fan, S., P. Hazell, and S. Thorat. 1999. Linkages between Government Spending, Growth, and Poverty in Rural India. IFPRI Research Report No. 110. Washington, D.C.: International Food Policy Research Institute. Goletti, F. and K. Rich. 1998. Policy Simulation for Agricultural Diversification. Final Report submitted to UNDP. Washington, D.C.: Intemational Food Policy Research Institute Government of Vietnam. Ministry of Agriculture and Rural Development. 26 March 1998. The Five Million Hectare National Reforestation Programme: Some Matters for Discussion. International Support Group Core Meeting. MARD. Hanoi. Government of Vietnam. Ministry of Agriculture and Rural Development. Department of Forestry Development. February 1999. Forest environment management in relation to poverty alleviation and hunger eradication in Vietnam. Country paper prepared for Mid-Term Review of RETA 5771. International Monetary Fund (IMF). 1999. Vietnam: Selected Issues. IMF Staff Country Report No. 99/55. Washington, D.C. International Monetary Fund. Nguyen, X.N. 1998. Review of rural finance environment Vietnam in the context of outreach. Vietnam- Canada Rural Finance Outreach Project (VCRFOP) Did-7001. Report on the Review of the Recommendations from Constituencies - Presented by Mr. Vu Mao, member of the NA Standing Committee, to the deputies of the I Oth NA at its 4th Session on 10"' November, 1998. Vietnam 1998/1999. The Gioi Publishers, Hanoi. 78 Annex D - Agricultural Sector Socialist Republic of Vietnam (SRV). 1996. Public Investment Program 1996-2000. Hanoi, June, 1996. Socialist Republic of Vietnam (SRV). 1998. Government Report for Consultative Group Meeting Paris 7- 8 December 1998. "Government Direction on Agricultural and Rural Development". Hanoi, November 1998. United Nations Development Program (UNDP). 1996. Vietnam Public Expenditures Review. Volumes I and II. UNDP VIE/94/025. World Bank, Asian Development Bank, FAO, UNDP, and the NGO Water Resources Group in Cooperation with the Institute of Water Resources Planning, Vietnam. 1996. Vietnam Water Resources Sector Review. Main Report. World Bank. 1997. Vietnam - Agriculture for Sustainable Rural Development. A Sector Report. Report No. 17278-VN. Washington, D.C. The World Bank. World Bank. 1998. Vietnam - Rising to the Challenge. An Economic Report. Report No. 18632-VN. Washington, D.C. The World Bank World Bank. 1999a. Vietnam - Development Report 2000 - Attacking Poverty. The World Bank. Consultative Group Meeting for Vietnam. Hanoi, December 14-15, 1999. World Bank. 1999b. Vietnam - Preparing for Take-Off? An Informal Economic Report of the World Bank. Consultative Group Meeting for Vietnam. Hanoi, December 14-15, 1999. 79 Vietnam Public Expenditure Review 2000 - Annexes Appendix Table D.1 Gross value of agricultural production at constant 1994 prices, 1994-97 1994 1995 1996 1997 (in billions of VND, at constant 1994 prices) Total 74,838 79,812 83,967 88,663 Food crops 51,465 54,034 56,814 59,559 Paddy and other cereals 40,220 42,110 44,654 46,953 Vegetables and beans 4,531 4,984 5,088 5,246 Fruits 5,431 5,578 5,688 5,943 Other 1,283 1,362 1,384 1,417 Industrial crops 10,331 12,149 12,806 13,989 Animal husbandry 13,042 13,629 14,347 15,206 (Share in total, percent) Food Crops 68.8 67.7 67.7 67.2 Paddy and other cereals 53.7 52.8 53.2 53 Vegetables and beans 6.1 6.2 6.1 5.9 Fruits 7.3 7 6.8 6.7 Industrial crops 13.8 15.2 15.3 15.7 Animal husbandry 17.4 17.1 17.1 17.2 Source: IMF (1999) - Table 5 80 Annex D- Agricultural Sector 81 Appendix table D.2: Public investment program allocation 1996-2000 to the Agricultural Sector (in Trillion VND) Investment Structure for the Period 1996-2000 Total State SOE Investment Household FDI Total % of State % of State Investment Total of which Investment % Investment Credits e State Credits Total Investment 460.00 96.80 142.40 76.90 76.50 144.30 100.00% 100.00% 100.00% Agriculture 95.50 21.90 42.90 22.50 24.20 6.50 20.76% 22.62% 29.26% General Agriculture 58.50 2.90 34.90 18.00 14.20 6.50 12.72% 3.00% 23.41% Water Management 14.50 14.50 3.15% 14.98% 0.00% Forestry 13.00 4.00 4.00 3.00 5.00 2.83% 4.13% 3.90% Storage 9.50 0.50 4.00 1.50 5.00 2.07% 0.52% 1.95% Source: Socialist Republic of Viet Nam - Public Invesment Program 1996-2000. Table 11I.2, page 20. Hanoi, June 1996. 81 Vietnam Public Expenditure Review 2000 - Annexes Appendix Table D.3 Subsidies and current transfers in 1998 (in Million VND) Total Expenditures Total Current Subsidies and Current Subsidies and Current Subsidies and Current in Expenditures Transfers to Transfers as Share of Transfers as Share of Agriculture in Agriculture Current Expenditures Total Expenditures Agriculture in Agriculture in Agriculture Total 4,591,435 1,098,778 195,174 17.8% 4.3% Of which: Cultivation 189,373 77,750 31,279 40.2% 16.5% Animal husbandry 77,416 28,272 23,807 84.2% 30.8% Veterinary activities 68,917 50,804 3,127 6.2% 4.5% Forestry and related activities 363,703 233,021 12,012 5.2% 3.3% Forestation 85,863 10,057 1,679 16.7% 2.0% Fixed cultivation & sedentarization 188,721 122,533 65,012 53.1% 34.4% Irrigrstinn A' reIntpeI %rtivitiPc 2,469,736 162,634 21,627 13.3% 0.9% Source: MOF 82 Annex D - Agricultural Sector 83 Appendix Table D4: Distribution of agricultural state expenditure at the central and local levels (in Million VND) 1997 1998 State Central Local State Central Local Total 3,712,494 2,071,250 1,641,244 4,591,435 2,755,649 2,013,787 Of which: Cultivation 130,725 50,537 80,188 189,373 48,257 141,116 Animal husbandry 85,448 42,259 43,189 77,416 38,012 39,403 Veterinary activities 65,970 11,745 54,225 68,917 5,106 63,811 Forestry and related activities 315,235 57,627 257,608 363,703 77,623 286,080 Forestation 48,815 2,161 46,654 85,863 45,613 40,250 Fixed cultivation & sedentarization 270,289 174,753 95,536 188,721 97,254 91,468 Irrigation & related activities 1,806,371 946,684 859,687 2,469,736 1,422,333 1,047,403 National Programs on reforestation and land reclamation 642,702 636,956 5,746 642,702 636,956 5,746 Source: MOF 83 Vietnam Public Expenditure Review 2000 - Annexes Appendix Table D.5 Source of funding for selected research institutes (in Million VND) 1996 1997 1998 Source Source Source Institute Total Central Local Other Total Central Local Other Total Central Local Other Food crop research institute 4,114 3,628 486 6,925 6,220 705 6,414 4,622 1,792 National institute of animal husbandry 17,351 5,368 9,971 2,012 27,072 9,594 14,882 2,596 14,021 9,041 2,406 2,574 National maize research institute 2,581 1,950 631 2,636 2,486 150 4,012 3,983 29 Institute of agricultural science ofiSouth Vietnam 8,635 5,135 1,021 2,479 8,116 5,548 784 1,784 10,090 4,778 1,882 3,430 Cuu Long rice research institute 4,490 2,090 300 2,100 4,699 2,049 200 2,450 5,888 2,313 550 3,025 0 0. 0 Total 37,171 18,171 11,292 7,708 49,448 25,897 15,866 7,685 40,425 24,737 4,838 10,850 Source. Personal Interviews With Above Research Institutes. 84 Annex D - Agricultural Sector 85 Appendix Table D.6: Agricultural extension expenditures for selected provinces (in Million VND) Provinces 1996 1997 1998 1999 Total State Local Total State Local Total State Local Total State Local Budget Budget Budget Budget Budget Budget Budget Budget 1. Ca Mau 1,238 151 900 1,281 158 1,100 1,461 151 1,100 2. Phu Yen 719 516 920 158 700 1,032 207 700 832 177 600 3. Dak Lak 942 175 767 1,177 201 976 1,437 217 1,220 1,578 246 1,332 4. Dong Thap 2,057 163 1,816 2,490 381 2,098 2,412 125 2,289 2,606 220 2,386 5. Lang Son 406 140 265 553 174 379 511 130 381 573 176 397 6. Nghe An 1,629 268 1,361 1,645 283 1,361 1,597 264 1,333 1,749 399 1,350 7. Nam Dinh 741 741 1,578 1,578 1,195 1,175 20 1,182 1,152 30 Total 6,494 2003 4,209 9,601 2,926 6,414 9,465 2,276 7,043 9,981 2,521 7,195 Source: Provincial Department ofAgricultural Extension 85 Vietnam Public Expenditure Review 2000 - Annexes Appendix Table D.7: 1999 budget information for Bac Hung Hai IMC (controlled by MARD) Cultivated area under irrigation 124,000 ha Provincial coverage 3 provinces and I city: Hung Yen, Hai Duong, Bac Ning, and Southern part of Da Long district of Hanoi Irrigation system 200 km of primary canals, 13 irrigation control works, more than 500 pumps, and several thous. kms of secondary and tertiary canals COSTS in VND billion REVENUES In VND billion Salaries and wages 1 Water Fees 7 - 8 Large repairs 3.9 Regular & small repairs 2.7 Electricity (580 VND/kw) 0.2 Management 0.92 Services for flood and storm 0.023 Transfer of technology or training 0.13 Compensation for unplanned 1.4 losses (i.e. in case fees are not fully collected) Depreciation 0.35 Other 1.32 Total 11.9 Source. Personal Interview The water fee charged by this IMC is 90 kg of paddy per ha for the spring season and 60 kg of paddy per ha for the summer and fall seasons. The provinces keep 60 percent of the fee and give 40 percent to the IMC. Costs that are not covered by water fees are transferred from the GOV to the IMC. However, the budgeted revenues of 7 to 8 billion are based on the assumption that the water fees will be fully collected. The manager of the IMC believes they will collect only about 60 to 65 percent of this amount. If they get less than their 7 to 8 billion, they are allowed to use their compensation fund (1.4 billion) and if that is not enough, they have to reduce expenditures such as repairs, maintenance, etc. Cost recovery in this IMC therefore does not exceed 50 percent. The budget of this IMC was approved by MARD. 86 Annex E - Health Sector ANNEX E - HEALTH SECTOR Overview of the Health Sector 1. Vietnam has made impressive progress in improving the health of its population during the last 3-4 decades. Vietnam's health indicators are much better than would be expected for a country at its level of income per capita. For instance, its average life expectancy at birth has been estimated at 68 years, which is considerably greater than the life expectancy for Cambodia (54 years)--a country having approximately the same per capita GDP as Vietnam. Vietnam's success is partly a consequence of its past centrally planned economy , when it invested considerable resources in establishing a large network of primary health facilities throughout the country and in developing a number of effective categorical health programs to deal with priority health problems, such as malaria, diarrheal diseases, and immunizable diseases. Another contributing factor to the generally impressive health indicators is the high rate of adult literacy achieved by Vietnam. Infant Mortality 2. Available data show a sharp decline in the infant mortality rate in Vietnam during the last four decades--from a Figure E. 1. level of about 160 per 1,000 Infant mortality rate, 1960-98 live births in 1960 to 90 in 160 1975-80, 75 in 1983, 45 in 1989, and to 44 in 1989-93 140- 0 (Figure E.1). Data from the 1 1997 the 1997 Demographic and Health Survey suggest too that the infant mortality rate i may have declined further to * only 28.2 per 1,000 live births g 60 for the period 1992-96, 40 although these estimates are likely underestimated. The 20 L 9 9 1 9 9 -T Ministry of Health has 1960 1975-80 1983 1989 1989-93 1996-97 estimated the infant mortality Source MOH rate in the country to be 39 per 1,000 live births in 1997 (Ministry of Health, 1997). 3. These achievements are not simply impressive--they have few parallels in the world. Not only has Vietnam experienced one of the most rapid infant mortality declines in the East Asia region, its current level of infant mortality is among the lowest in the region and comparable to countries that have substantially higher levels of per capita 87 Vietnam Public Expenditure Review 2000 - Anneres income. Indeed, as Figure E.2 indicates, if Vietnam were to follow the experience of other Asian countries, it would be expected to have an infant mortality rate roughly two times as large as it currently has, given its per capita income level. 4. Of course, the low overall level of infant mortality is not enjoyed uniformly by all population groups in the country. There are large rural-urban, regional and income disparities in infant mortality. For instance, the rural areas of the country have an infant mortality that is nearly 80 percent greater Figure E.2 than that of the urban areas. The Infant mortality rates and real GDP per capita infant morality rate in the across Asian countries, 1997 Central Highlands is more than 110 two times as high as that in the 100 Southeast and the Red River so Delta (GSO 1995). And the t 70 wealthiest quintile of households Q 60 have an infant mortality rate that - 50 is only one-half of that f 4_ !a 30 experienced by the poorest 2_ quintile (NCPFP, forthcoming). 0 _ These differences are indicative 0 of the large disparities in access 1,noo 3,000 5,000 7,000 9,000 l c000 13.000 to health services, child Real GDP per capita (PPPS). 1997 nutrition, and environmental hygiene (such as availability of safe drinking water and sanitation). Additionally, there is some evidence that regional disparities in infant mortality may even have widened in the recent past (GSO 1990, GSO 1995). Morbidity and Its Causes Figure E.3: Major Causes of Morbidity, 1976-97 _ l 00'4% -~90% .k 80% 65%1- x* - 9 0% 50% 40% X 30% a 20% I % - _% aT 1976 1986 1997 WCommu.icable diseases fNon-conmanicable diseases DAccidents, injuries and poisoniag 88 Annex E - Health Sector 5. Nationally-representative data on morbidity rates show 41 percent of individuals reporting an illness episode in a four-week reference period (Vietnam Living Standards Survey VLSS 1998), implying that, on average, a Vietnamese experiences 5.3 illness episodes per year. In most other developing countries at Vietnam's level of per capita income, individuals average 2-2.5 illness episodes annually. 6. Vietnam has experienced an epidemiological transition during the last 20 years, with a major decline in the share of communicable diseases in morbidity. While communicable diseases accounted for 50-56 percent of morbidity in 1976, the corresponding share had fallen to 27 percent by 1997 (Figure E.3). Much of this decline was achieved during the last ten years. Most dramatic has been the increased incidence of accidents, injuries and poisoning, whose share in morbidity increased several-fold between 1976 and 1997. The share of non-communicable diseases also increased considerably over this period. 7. These shifts reflect the success of communicable disease-control programs, especially the expanded program of immunization, which has reduced the incidence of vaccine-preventible diseases in the country. 8. In 1997, the five leading causes of morbidity in hospitals were infectious and parasitic diseases (accounting for 23 percent of all cases), respiratory ailments (17 percent), diseases of the digestive system (12 percent), genito-urinatory diseases (5.3 percent), and diseases of the circulatory system (5.1 percent) (Ministry of Health, 1997). 9. Despite the relative decline in the incidence of infectious diseases, the latter still remain a main public health problem, particularly some new or reemerging diseases, such as tuberculosis, HIV infections, dengue fever and Japanese encephalitis. For example, the incidence of dengue fever has increased dramatically in the last 4-5 years, with the infection rate increasing almost five-fold between 1994 and 1998. Likewise, the number of tuberculosis cases has also increased substantially in recent years. In 1997, the incidence of tuberculosis was about 90 cases per 100,000 inhabitants and 2,402 deaths from tuberculosis were recorded at hospitals (Ministry of Health, 1997). 10. HIV/AIDS is a new disease whose incidence has increased rapidly in the past few years. The number of new HIV cases have increased ten-fold between 1993 and 1998. It is estimated that there are 88,000 persons living with HIV/AIDS in the country. By the end of the year 2000, it is estimated that this number will rise to 186,000. Even though 65 percent of the HIV infections reported so far are associated with intravenous drug use, heterosexual transmission among young people is increasingly becoming more common (UNDP, 1999b). 11. Even though implementation of malaria control measures in the past 6-7 years have been very successful and have brought down the incidence of malaria from 16 cases per 1,000 persons in 1991 to only 6 in 1998, malaria continues to remain a major public health problem in the mountainous and ethnic minority areas of the country. 12. Accidents, injuries and poisoning are a major cause of death in hospitals. In 1997, there were 475,000 cases of and 5,424 deaths from accidents, injuries and poisoning. As noted earlier, this represents a dramatic increase from the situation twenty years ago. 89 Vietnam Public Expenditure Review 2000 - Annexes Many of these deaths are caused by traffic accidents. This is a trend that is characteristic of many developing countries experiencing economic growth, rapid urbanization and traffic growth. 13. Finally, although the incidence of diseases of affluence, such as cancer and heart disease, has increased over time, they still account for a small proportion (less than 1 percent) of total morbidity or mortality in the country. Of course, these diseases are much more prevalent in urban centers, such as Hanoi and Ho Chi Minh City, but their incidence, at least as of now, is confined largely to better-off populations in the urban areas. Child Malnutrition 14. While Vietnam has far superior levels of infant mortality than most other countries at its level of per capita GDP, its performance in the area of child malnutrition has been less than impressive. Indeed, child malnutrition rates in Vietnam are among the highest in the region. A recent national survey of nutrition conducted by the National Institute of Nutrition and UNICEF (1999) indicates that 39 percent of children under the age of 5 years are malnourished in terms of weight-for-age and 34 percent are undernourished in terms of height-for-age'5 The fact that the rate of malnutrition in terms of weight-for-height is significantly less (11 percent) suggests that Vietnam's major child nutrition problem is stunting from longer-term, chronic under-nutrition rather than wasting from short-term, acute food deficits. Figure E.4 15. Figure E.4 shows that Vietnamn's malnutrition rates have Prevalence of underweight children aged 0-59 nmonths, selected remained above the regional 3 60 average and above those of l Thailand and Indonesia ._ consistently from 1975 to 1995. 40 (As noted above, more recent data e from 1998 suggest that this picture t 30 has not changed) X_ 5 20 Issues in a Market Economy E 10 16. The roles of government in a a market economy are quite 1975 1980 1985 1990 1995 different from those in a centrally- planned economy. This is important for Vietnam as it makes its transition from a centrally-planned economy to one where markets play an increasingly important role. In a market economy, the government is neither the exclusive provider nor the only financier of all health services, as in a 15Another national survey--the Vietnam Living Standards Survey of 1998--indicates an underweight rate of 36 percent and a stunting rate of 35 percent. 90 Annex E - Health Sector command economy (in principle at least)16 As a large number of private health providers and private drug vendors enter the market for health services, the government needs to play an important new role that of regulating the private and public provision. More specifically the government needs to establish the legal framework for private (and public) provision, setting minimum quality-of-care standards, promulgating laws and regulations that govern appropriate behavior of private and public providers and implementing these regulations and laws through a system of regular inspections and monitoring. The regulatory role of the government becomes particularly important in the context of a rapidly expanding private sector with considerable variation across private providers in the quality of care (e.g. infection control, hygienic practices, and standards of practice protocols) and quality of drugs dispensed, as is the case in Vietnam. 17. In the transition to a market economy, out-of-pocket payments that individuals have to make for utilizing health services typically increase sharply. This often leaves the poor in a vulnerable position, unless there are well-functioning mechanisms for exempting the poor from the increase in health-care costs. It is therefore important for the government to take on a direct role in financing health-care expenditures of the poor by, say, providing them with pre-purchased (by the government) health cards that can be used to obtain free inpatient services and drugs at hospitals and clinics. Indeed, there may be a case for the government to reduce its role in the direct provision of (curative) health services and increase its role in the financing of health expenditures for the poor. 18. Private health providers in a market economy are naturally limited to the provision of curative health services. The government continues--and needs to--to remain as the exclusive provider of preventive health services, such as immunization and control of communicable diseases." One of the key messages in health economics is that preventive-care interventions should have the first claim on public resources due to the substantial externalities they create for society and due to the resulting under-spending on them by individuals. 19. Thus, in the transition to a market economy governments reduce their role in the direct provision of curative health services and correspondingly take on new roles, such as regulation of the private sector and protection of the poor from market pressures. Provision and Utilization of Health Services 20. Service Provision. Service delivery is organized along a four-tiered pyramid, with commune health centers at the bottom of the pyramid. Above these are the inter- communal polyclinics and their referral organizations (i.e., the district hospitals), both of which are run by the district health bureaus. Provincial hospitals form the second tier in the pyramid, and at the very top are central hospitals and specialty institutes, which are run directly by the Ministry of Health. 16It is important to note that, in practice, individuals in most centrally-planned economies financed non- trivial amounts of health-care expenditures themselves, as informal payments to supposedly-free public health providers were common. 17While it is certainly possible for the government to pay for these services and subcontract their provision to private parties, this is typically not done in most low- and middle-income countries. 91 Vietnam Public Expenditure Review 2000 - Annexes 21. In 1998, there were a total of 30 central hospitals and specialty institutes belonging to the Ministry of Health plus 196 provincial hospitals, 576 district hospitals and maternity homes, 929 polyclinics, and 9,920 commune health centers. One of Vietnam's greatest achievements over the last 30 years has been the establishment of an extensive network of commune health centers throughout the country. Indeed, at 7,019 persons served by a commune health center on average, Vietnam's primary health center/population ratio is significantly smaller than that of neighboring Cambodia (15,000) and comparable to that of Thailand (6,762)-a country having four times the per capita income of Vietnam. 22. Since private health services were forbidden until 1989, it is natural that government health facilities dominate the provision of health services in the country. However, there has been significant growth of private health services in the last ten years. A recent mail survey indicated a total of 36,442 private health facilities (29,400 not including traditional healers) in 44 (out of 61) provinces. This suggests that the actual number of licensed and unlicensed private health facilities in the country may well be close to 50,000. Most of these (about 18 percent) are drug commission sellers, followed by general practitioners' clinics (16 percent), traditional medicine clinics (14 percent), private pharmacies (14 percent), and nursing homes (12 percent) (VHSR 1999). 23. While most private health facilities are located in urban areas, many rural areas, especially in the North and the South, have significant numbers of private health providers as well. Even as far back as 1996-97, a Ministry of Health survey found that there were more private than public-sector health workers at the commune level in the North and the South of the country (Ministry of Health, 1998). This number has probably grown substantially since then. 24. While the private provision of primary health services has expanded rapidly, that of secondary and tertiary health services has increased relatively little during the last 10 years. It is estimated that there are only six private hospitals in the entire country. 25. Service Utilization. Clearly, what is ultimately important for improved health outcomes is not the provision of health services, but their utilization. Administrative data from health facilities indicate that an average Vietnamese has an average of 1.7 annual consultations with the health services, and that the annual inpatient admission rate is 68 per 1,000 persons. The annual consultation rate is likely to be an underestimate as it only includes consultations with the public health services, which forrm a relatively small proportion of consultations. 92 Annex E - Health Sector Figure E.5 Annual public-sector health contacts per capita Asia, 1990-95 Sri Larka ¢ 2.1 Singapore _ S - China 8 Vietnam i1 7 2 Pakistan [ ': I DZ In d ia 30 a t Malaysia = 7 Cambodia d0i4 Figure E.6 Annual inpatient admissions per 990-95 117 2 89 0 68 0 57 0 m V~~~~~~~~ 41 9 204 X, 136 150 2 LJ_~L Indonesia Hongkong Cambodia China S Korea Vietnam Japan Singapore Annual inpatient admissions per 1,000 persons 26. An annual public-sector contact rate of 1.7 per person compares favorably with those countries in the region for whom data are available. In Figure E.5, Vietnam's utilization rate is observed to be lower than that of Sri Lanka, China and Singapore-all countries that have substantially higher per capita income levels-but greater than that of India, Pakistan, Cambodia, and, surprisingly, Malaysia. 27. Vietnam's inpatient admissions rate is, in fact, among the highest in the region (Figure E.6). Indeed, the only countries in Asia having higher inpatient admissions rates than Vietnam are Japan and Singapore. If anything, there is a hint in all these numbers of over-utilization of inpatient, hospital-based services in Vietnam. 28. Distribution of Health Contacts Across Providers. Table E. 1 reports the distribution of health service contacts in a four-week reference period in 1998 by per capita expenditure quintiles. These data reveal three important findings. First, drug 93 Vietnam Public Expenditure Review 2000 - Annexes vendors are clearly the most commonly-used provider in Vietnam, accounting for two- thirds of all service contacts'8 29. Second, relatively few (14 percent of total) health service contacts take place with public health facilities. If consultations with drug vendors are excluded from the total, the public-sector share rises to 40 percent (However, as indicated in the Vietnam Health Sector Review, the public hospitals are the exclusive provider of inpatient services and the share of the private hospitals in total inpatient service contacts is negligible). 30. Third, the choice of health providers varies markedly across economic groups. There is a pronounced shift toward private clinics/doctors and public hospitals--and away from drug vendors and commune health centers--with increasing income (or, more accurately, consumption expenditure per capita). These shifts are observed in most developing countries, and represent shifts to higher-quality health services (at least as perceived by the patient) with increased affluence. Of particular importance for policy is the switch from commune health centers to public hospitals with increased affluence. The relative role of public hospitals in total service contacts is three times greater among the highest income quintile of persons than among the poorest quintiles. Conversely, the share of commune health centers in total health contacts is four times greater for the poorest quintile than for the highest income quintile. Table E.1: Distribution of health service contacts by providers and per capita expenditure quintiles, 1998 Per capita expenditure quintile Provider Poorest Second Third Fourth Richest Total Public hospitals 3.15 3.75 4.63 7.22 9.55 5.84 Commune health center 7.39 6.84 6.91 5.62 1.72 5.56 Otherpublicproviders 2.05 1.79 2.13 3.14 2.89 2.42 All public providers 12.59 12.38 13.67 15.98 14.16 13.82 Private clinics, doctors 14.93 14.46 17.38 16.00 21.38 17.02 Drug vendors 69.29 69.68 66.56 64.81 59.37 65.66 Traditional healers 3.19 3.38 2.39 3.21 5.10 3.48 Other providers 0.00 0.09 0.00 0.00 0.00 0.02 Total 100.00 100.00 100.00 100.00 100.00 100.00 Source: VLSS 1998 data. 18 Only visits to drug vendors without a prescription from another medical provider are considered in this analysis. Visits to drug vendors to fill a prescription obtained from another provider are assumed not to constitute a separate health service contact. 94 Annex E - Health Sector Users of Public Health Facilities Figure E.7 31. Another way of Econotnic coinposition of users of different health providers, 1998 looking at these data is to 100% - - examine the economic 90% composition of all of the ` 80% users of a particular health 7 0% provider. This information is 60% 1 important in knowing which o I 2 40%- types of public health 30% - 25 expenditures are likely to be B 20% more pro-poor than others. 10% _; 0% - _ 32. These data, shown in Public hospitals Traditional Private cilncs Drug vendors CHCs Figure E.8, clearly indicate healers Per capita expenditure quintile that the users of public [ElPoorest OSecond *Third *Fourth *R,chsi hospitals are largely drawn from the highest income quintiles, while commune health center users are drawn disproportionately from the poorest quintiles. For instance, in 1998, 36 percent of all public hospital users were drawn from the richest 20 percent of the population and 61 percent were drawn from the richest 40 percent of the population. On the other hand, the poorest quintile were vastly under-represented among hospital users relative to their share in total population, constituting only 8 percent of aggregate hospital users. 33. Several factors are responsible for the preponderance of more affluent individuals among users of hospitals. First is the urban location of the majority of hospitals, which improves geographical access to hospitals for urban dwellers, who are typically better-off than rural residents. Second, because the direct private cost of using hospital services is large, they tend to be used by individuals with higher purchasing power. Third, the spread of health insurance between 1993 and 1998 may actually have exaggerated the pro-rich bias of hospitals, since individuals with health insurance coverage tend to be better off than those without insurance coverage and the health insurance agency largely reimburses hospital (but not commune health center) costs. 34. Implications for Progressivity of Public Health Spending. Public subsidies to a particular level of health facility (e.g. commune health centers) will be weakly progressive if the share of the poor among all users of that facility is greater than their share in aggregate national consumption expenditure. In such a case, the distribution of conrmune health center users is more progressive than the underlying distribution of personal consumption; hence, public subsidies to commune health centers will be weakly efficient (in terms of targeting). On the other hand, public subsidies will be strongly progressive if the share of the poor among all users of that facility is greater than their population share. 95 Vietnam Public Expenditure Review 2000 - Annexes Figure E.8 Lorenz distributions of service contacts with public hospitals and 35. Figure E.8 suggests that commune health ersiters in comparison to lorenz distribution of public subsidies to commune 100 aggregate consumption expenditure, 1998 health centers and public 90 hospitals together are weakly pro- poor, since the distribution 7 l of users of these public health 6 O Xl facilities is more progressive l than the underlying distribution 40 of personal consumption °3l expenditure. However, only 20 public subsidies to commune 10 health centers (but not to public 0r hospitals) are strongly pro-poor, 0 20 40 60 80 100 as the share of the poor among commune health center users exceeds their share in the population. The Public Expenditure Decision-Making Process in Health 36. In Vietnam, the public health budget covers four levels of government: the central and the three levels of local governrnents: provincial, district and commune. Since some expenditures by communes from their own resources are not included in the State budget, a comprehensive database which aggregates the four levels into a consolidated statement of sectoral expenditure is lacking. However, central, provincial and district governments account for about 90 percent of the national budget. 37. The Ministry of Health has a limited role in deciding the overall budgetary allocation to the health sector. It is the Ministry of Finance (MOF) and the Ministry of Planning and Investment (MPI) that play deciding roles in this process, albeit with some discussion with the Ministry of Health. The overall level of recurrent spending on health is determined by projected growth rates of total revenue and of total recurrent expenditures (after subtracting estimated expenditures for new policies, such as increased salaries, etc.) and the share of health in total recurrent expenditures. This type of incremental budgeting pays little attention to health goals and health priorities set by the Ministry of Health. 38. The overall budgetary resources to the health sector are allocated on the basis of certain norms across two dimensions: functional (curative versus preventive) and sectoral (Ministry of Health, other sector ministries, and the provinces). These transfer norms are based either on the number of hospital beds belonging to a ministry (for curative expenditures) or on the population of the province (for both curative and preventive expenditures). These amounts are supposed to cover salaries for a centrally-determined number of health workers in the provinces, fees and drug exemptions for the poor, and other recurrent expenses. The allocation norms vary across five geographical regions, presumably to capture variations in the unit cost of delivering services in different geographical regions. In addition, Hanoi and Ho Chi Minh City have a norm that exceeds those of other cities by 50 percent. In general, the norm for curative expenditures is about 96 Annex E - Health Sector 3 times as high as that for preventive expenditures, reflecting the higher unit cost of curative interventions. 39. In addition to central government-disbursed funds, the provinces, especially the richer ones, can (and do) spend on health out of their own revenues. 40. Interestingly, however, in the annual budget allocation document sent to provinces, by Ministry of Finance only an aggregate recurrent budget is provided -- not its composition across sectors19. Thus, provinces are at liberty to make their own spending decision on each sector, including health. The provinces also can spend on health out of their own revenues20. The financing of health expenditure out of provincial revenues is largely what causes large disparities in government health expenditure per capita across provinces, as richer provinces are able to spend more than poorer provinces. 41. Although, in principle, provinces have to inform the Ministry of Finance about how they allocate their aggregate recurrent budget, this is an ex post formality. It is not at all clear how strongly, if at all, the Ministry of Finance can influence provincial spending on health out of the aggregate budget already approved for the provinces. The Ministry of Health has an even smaller role in this process. Although current regulations require provinces to notify the Ministry of Health of their budget allocation for health in practice this is not observed . Provinces are not required to provide the Ministry of Health with information on actual health expenditures or on results achieved. As a consequence, the Ministry of Health has very little or almost no information on provincial health budgets and expenditures. 42. In addition to the budgeted amounts for health, the provinces receive funds from the various national health programs (such as the Expanded Program of Immunization, the Tuberculosis Program, etc.). Until this year (2000), specific amounts were earmarked for each program, and provinces had no authority to reallocate funds across programs. This resulted in under utilization of some program funds and excess demand for other program funds in the same province. Starting with FY 2000, a new mechanism of fund allocation has been introduced whereby provinces are given block grants for all targets of the national health program and have discretion in allocating the funds across the different programs. 43. In principle, the national health programs should serve as vehicles for providing additional funds for disadvantaged areas and addressing various disease patterns specific to each province. However a recent study by Ministry of Health experts found that the allocation of program funds is more-or-less equally distributed across provinces and does '91n fact, provinces are provided with two minimum amounts for spending on education and science and technology and a maximum spending on administration. Other than these, no sectoral breakdowns are provided. 20A province can retain revenue that it collects in excess of its assigned target. It can choose to spend this retained revenue in any way it wishes, including on health. Likewise, provinces may have to reduce spending on health if their revenue performance fall short of the assigned target. 97 Vietnam Public Expenditure Review 2000 - Annexes not take account of the specific disease profile of a province.21 This suggests such expenditure may not be well targeted. 44. Districts receive most of their support from the provinces, although they have some own revenue sources. Provinces have almost complete discretion in the resources they pass on to district health services. While some provinces split their allocations evenly (i.e. 50 percent for the province and 50 percent for all the districts in the province), other provinces may spend two-thirds of their allocation on nationally- deternined staff salaries, which leaves little room for discretion. This results in even greater variation in per capita resource available at the district level, than that existing at the provincial level. 45. Before 1995 the communes were expected to be basically self-sufficient, on health expenditures, and these were not included in the integrated state budget. Since 1995, however, the government has taken over the payment of salaries to approved staff of commune health centers out of the state (provincial) budget. Virtually all communes supplement their small budgets through the sale of drugs. They buy drugs from the province or district pharmacy and resell it with a markup (typically between I and 10 percent). In richer communes households are sometimes assessed for contributions to defray the operational or capital expenses of the commune health center. Thus commune health centers in richer communes typically offer better-quality services than those in poorer communes. However households' willingness to make general contributions for the upkeep of the commune health center is limited and these taxes are not very popular. (See discussion of Fees and Contributions in Chapter 2 and in Annex C). Figure E.9 Total (public plus private) health spenaing per capita, seeccrea Asian Levels of and Trends in countries, latest year (1991-98) 408 Levels of and Trends In 40840 Health Expenditure 400 3 50 ' 3001 46. A consistent time : 300 series of public spending on -n 250 health in Vietnam is available 200 for the period 1991-98. In . 150 addition, nationally- 00 83 representative data on private 50 1 (household) spending on 5 9 10 12 t2 health are available for two . 3 years, 1993 and 1998, from e,ig° -° .¢ 5 the Vietnam Living Standards Survey (VLSS). Adding these two sources results in a total health expenditure estimate of VND 28,773 billion in the aggregate or VND 383,226 per capita for 1998. This is equivalent to US$27.41 per capita 21 Draft Report, Evaluation of Current Status and Impact of some National Programs. December 1999 - discussion of National Health Support Project sub-component 3.1.2). 98 Annex E - Health Sector or 8 percent of GDP. In comparison, total (public plus private) spending on health in 1993 was US$9.67 per capita or 5.2 percent of GDP. 47. A total health spending of US$27 per capita places Vietnam among those countries in Asia that spend the most on health care (Figure E.9). Indeed, only Malaysia, Thailand and South Korea spend more on health per capita than Vietnam. Public Expenditures on Health 48. The public expenditure Figure E. 10 data show an impressive increase Total public spending on health, Vietnam, 1991-98 70000 in public spending on health over the 1991-98 period (Figure E.10). 2 60,000 Both total and per capita public t 50000 spending on health increased more 1 6 than two-fold in real terms &..0 n representing an annual rate of . 30 growth of around 12-14 percent, while health spending as a share of 3 3 0 00 GDP and as a share of total f 2 20000 government expenditure also 1 1,0000 increased, albeit at more modest 0 rates. Such high growth rates in °9 1992 1993 1994 1995 199 1 99a public spending on health are impressive and reflect a high level of commitment by the government of Vietnam to the health sector. 49. Of course, these impressive rates of growth represent increases from what was a very low level of public spending on health in the early 1990s. Even after these increases, per capita public health spending in Vietnam (inclusive of user fees and health insurance contributions) was only VND 80,844 ($5.78) per capita or 1.7 percent of GDP and 8.5 percent of total government expenditure in 1998. In absolute terms, this level of government health expenditure is not overly high; for instance, it is lower than that in Indonesia ($6.2), Cambodia ($7.2), China ($8.1), Sri Lanka ($9.2), Philippines ($12.3) and Thailand ($28.7) (World Bank 1999, UNDP 1999). In addition, Vietnam's public expenditure on health of $5.78 per capita is less than one-half of the $12 recommended by the World Development Report 1993 as being needed to finance an essential package of health services, including public health, for a low-income country.23 To achieve this level of spending, Vietnam would need to spend an additional US$468 million annually. 22Government health expenditure includes spending from the state budget as well as from other public sources, such as user fees, health insurance reimbursements, and external donors. The state budget includes both the central state budget and the local (provincial) health budgets. Finally, the central state budget includes not merely the budget of the Ministry of Health but also the health budgets of other ministries and government agencies, such as Posts and Telecommunications, the army, and police. 23World Bank, World Development Report 1993, Oxford University Press, Washington D.C., 1993. 99 Vietnam Public Expenditure Review 2000 - Annexes 50. Sources of Financing. While the government is a large provider of health services in the country, its role in the financing of health expenditures is modest. In 1998, the public share in aggregate health spending was only about 20 percent, with households accounting for the remaining 80 percent of health spending. One consequence of this heavy reliance on private financing is that the poor have to bear the burden of high user charges at public health facilities Interestingly, a large portion of the private spending on health actually takes place at public facilities, even though services at many of these facilities are in principle free or nominally-priced. Figure E. 11 51. Figure E. 11 shows the Sources of health financing (% of total expenditure on health sources of health financinig contributed by each source), 1998 within the public sector. In 1998, the bulk of the public spending *Households E Provincial E] Central gov't on health was undertaken by ODonors * Insurance E Communes provincial governments, 2 followed by the central 2. government, health insurance, external donors and communes. Figure E. 12, which shows the 80.5 19.6 2.3 evolving shares of each financing 0.9 source in total public health ------__4__ spending, indicates that the share of provincial governments in total financing has fallen, while that of the central governnent has risen correspondingly, between 1991 and 1998. The most dramatic change has been Sources of funding Figure E. 12 calth, 1991-98 the rising share of health 100% insurance in financing public health expenditure. Starting from o0% a small base in 1993, it has come * 3 to contribute nearly 11 percent of § 60% public spending on health * currently. The share of donor 1. 4 M 9 financing of health declined from 1.i 14 percent in 1991 to only7 20 percent in 1996, but has since increased to 11.4 percent in 19 1998. 1991 1992 1993 1994 1995 1996 1997 1998 E Local badgets *Certral budget EHealth nsurance OUser fees ODA grants and loans *Conmnune budget Cost Effectiveness of Public Spending 52. What does public spending buy in the health sector and how has the 'productivity' of government health expenditure changed over time? Figure E.13, which shows the 100 Annex E - Health Sector number of public-sector service consultations and inpatient days 'produced' per million VND of public health spending (measured in constant 1994 VND) for the period 1992- 98, indicates that productivity of public health spending declined precipitously from 1992 to 1995, but has since trended upwards. Nevertheless, the number of service consultations and inpatient days per million dong of public health spending in 1998 were well below their respective values in 1992. F,islreF 13 Productivity of public health spending, 1992-98 41~~~~~~~~~~~~~~~~~~~~~2 39 - 35 3 3 a 16 4 3 4 S i ii ,B l l | | g - 12 v i ° 27 1992 1993 1994 1995 1996 1997 1998 53. Of course, the reduced productivity in public spending observed in Figure E.13 might simply reflect (i) an improvement in the quality of public-sector contacts (both inpatient and outpatient) over time, or (ii) an increase in the number of preventive health activities performed within the public health system. Quality improvements would serve to increase public spending on health without a commensurate increase in the number of service consultations or inpatient days, thus making it seem that productivity of spending was declining. Likewise, because the 'output' in Figure E.13 reflects only curative activities of the health system, it does not account for the possibility that public health spending may have been 'diverted' to more preventive health activities and preventive contacts. However, neither of these possibilities appears plausible, leading one to believe that the productivity of public health spending has indeed declined since the early 1 990s. 101 Vietnam Public Expenditure Review 2000 - Annexes Figure E. 14 54. How does the cost- Life expectancy and real public spending on health per capita across effectiveness of public health low- and middle-income countries. 1995 spending in Vietnam compare to that of other countries? 70 Countries that spend more 6 public resources per capita on health tend to have higher i levels of life expectancy, 5 controlling for literacy. Adult -2 5 literacy needs to be controlled 45 for in this relationship, as 40 there is a strong positive relationship between life 3 e O 150 200 2 0 350 0 50 100 150 200 250 300 350 expectancy and adult literacy. Real pubk. spending ox health per capata (PPP S) Figure E.14 plots the relationship between life expectancy and public health spending per capita for a cross- section of low- and middle-income countries from around the world.24 Vietnam is observed to be almost on the regression line in the graph, which suggests that Vietnam more-or-less enjoys a life expectancy rate consistent with its high level of public spending on health. However, there are a number of countries that are well above the regression line, which suggests that there may be further scope for Vietnam to improve the cost-effectiveness of its public health spending. 55. Thus, Vietnam's own experience in the past as well as the experience of other low- and middle-income countries suggests that it may be possible for Vietnam, with policy changes and expenditure reallocations, to improve the effectiveness of its public health spending and thereby obtain superior health outcomes without a large increase in health spending. Composition of Government Health Expenditure 56. Economic Composition. There are three important observations to be made on the economic composition of public spending on health. First, capital spending comprises a large share of the state health budget, although this share has declined from 28 percent in 1991 to 21 percent in 1998 (Table E.2). Nearly one-half of capital spending over the years (but rising to 82 percent in 1998) is on construction, which is again very large share in comparison to the health budgets of most developing countries. 24The sample of low- and middle-income countries in Figure 14 are from all over the world -- not just from Asia. The data are taken from UNDP (1999). 102 Annex E - Health Sector Table E.2: Distribution of public health spending by source and by economic type, 1991-98 Recurrent As % of total recurrent expenditure Capital As % of total capital Year expenditure as expenditure expenditure % of total as % of total expenditure expenditure Salaries and Goods and Subsidies Others Purchase of Capital wages services and transfers fixed assets construction State Budget on Health (Central plus Locall 1991 71.70 22.67 71.29 5.89 0.15 28.30 54.48 45.52 1992 70.73 16.34 76.79 6.34 0.53 29.27 67.70 32.30 1993 76.22 27.45 69.57 2.56 0.43 23.78 59.19 40.81 1994 77.99 32.92 62.32 4.74 0.03 22.01 53.07 46.93 1995 76.21 31.36 62.51 6.10 0.02 23.79 48.25 51.75 1996 77.51 29.22 65.75 5.02 0.00 22.49 40.38 59.62 1997 78.63 27.59 67.82 4.58 0.01 21.37 52.60 47.40 1998 79.08 29.38 59.38 4.02 7.22 20.92 17.73 82.27 Total Public Spending on Health (Including User Fees and Insurance) 1991 73.34 20.88 73.56 5.43 0.14 26.66 54.48 45.52 1992 72.42 15.04 78.64 5.83 0.49 27.58 67.70 32.30 1993 78.02 24.99 72.00 2.31 0.70 21.98 59.19 40.81 1994 80.07 29.41 65.84 4.18 0.57 19.93 53.07 46.93 1995 79.49 26.33 67.99 5.04 0.64 20.51 48.25 51.75 1996 81.39 23.47 71.91 3.96 0.66 18.61 40.38 59.62 1997 82.31 22.19 73.63 3.62 0.56 17.69 52.60 47.40 1998 82.77 22.94 67.80 3.08 6.18 17.23 17.25 82.75 Notes: Figures in bold are percentages of total health spending. The remaining figures are expressed as percentages of total recurrent or capital expenditure. Source: Ministry of Finance data. 57. Second, only about 29 percent of the recurrent health budget is spent on salaries and wages, while 59 percent is spent on goods and services. Since health insurance reimbursements and user fees are typically not used to pay for salaries, the share of salaries and wages in total government health expenditure (i.e. state budget plus insurance plus user fees) is even smaller-23 percent. 5 In this sense, Vietnam is unlike other developing countries which spend the majority share of their recurrent health budget on salaries and wages, and have very little to spend on consumables such as medical supplies and drugs. 58. Functional Composition. Table E. 3 below shows the composition of government health expenditure by source and by function (i.e., preventive, curative, and family planning). Preventive expenditures include monies spent on communicable disease- control programs and on the training of village-level health workers, while curative interventions reflect spending on hospitals, health centers and health facilities providing 23User fees may be used, however, for paying bonuses to health workers. 103 Vietnam Public Expenditure Review 2000 - Annexes ambulatory care. It should be noted that the budgetary classification of public expenditure into the three functional categories is far from perfect.26 Table E.3: Functional composition of public expenditure on health, by financing source, 1991-97 Year Central State Budget Year Local (Provincial) State Budget Preventive Family Preventive Family Curative Planning Other Total Curative Planning Other Total 1991 56.07 9.56 2.96 31.40 100.00 1991 71.29 12.55 4.55 11.61 100.00 1992 73.27 20.73 5.90 0.10 100.00 1992 69.85 13.79 4.96 11.41 100.00 1993 83.28 6.60 9.53 0.59 100.00 1993 66.74 13.68 5.51 14.08 100.00 1994 52.98 24.62 20.81 1.59 100.00 1994 68.75 13.99 4.27 13.00 100.00 1995 44.07 23.16 20.70 12.07 100.00 1995 70.26 13.17 3.29 13.27 100.00 1996 38.74 35.90 18.58 6.78 100.00 1996 70.97 11.74 4.14 13.15 100.00 1997 33.17 27.30 22.00 17.53 100.00 1997 79.82 10.59 3.04 6.55 100.00 1998 31.22 22.92 25.73 20.13 100.00 1998 81.53 9.82 3.09 5.55 100.00 Year ODA Grants and Loans Year All Public Spending (including User Fees and Health Insurance) Preventive Family Preventive Family Curative Planning Other Total Curative Planning Other Total 1991 85.29 13.74 0.97 0.00 100.00 1991 70.55 11.29 3.52 14.64 100.00 1992 51.28 48.61 0.12 0.00 100.00 1992 70.12 19.02 4.39 6.48 100.00 1993 65.38 34.26 0.36 0.00 100.00 1993 73.85 13.02 5.76 7.37 100.00 1994 59.24 40.31 0.45 0.00 100.00 1994 65.59 18.37 8.90 7.13 100.00 1995 75.76 15.17 0.39 8.68 100.00 1995 66.22 14.88 8.10 10.80 100.00 1996 62.50 19.39 8.20 9.92 100.00 1996 65.82 17.30 7.90 8.99 100.00 1997 46.55 20.89 13.98 18.59 100.00 1997 67.86 14.20 8.47 9.47 100.00 1998 31.11 17.63 34.60 16.66 100.00 1998 69.72 11.42 10.31 8.55 100.00 Source: Ministry of Finance data. 59. There are three observations that stand out from an examination of Table E.3. First, curative care takes up a disproportionately large share of public spending on health (70 percent in 1998). On the other hand, preventive interventions, such as communicable disease-control programs, account for merely 11 percent of public spending on health. The share of family planning is about 10 percent. 60. While this budgetary emphasis on curative activities is not unusual (and is observed in many developing countries), it is both inefficient and inequitable. One of the key messages in health economics is that preventive-care interventions should have the first claim on public resources due to the substantial externalities they create for society and due to the resulting under-spending on them by private agents. This pattern of spending is inequitable, as most curative spending is focused on hospitals and survey data for Vietnam show that a disproportionately large share of hospital users tend to be drawn from the affluent sections of the population (see Figure E.7). 26The separation of expenditures into preventive and curative is not perfect. Until recently, for example, one of the national programs included under preventive expenditures involved upgrading equipment at public hospitals. 104 Annex E - Health Sector 61. Second, the curative and preventive shares in total public health spending have not changed much over time. The curative share has hovered in the range of 66-74 percent over the period 1991-98, while the preventive share has had a wider range -- 11 percent to 19 percent. However, there has been an increase in the share of family planning in total public health spending -- from 3.5 percent in 1991 to 10.3 percent in 1998. 62. Third, the relative constancy of the curative share in total public health spending masks important changes that have taken place in the curative bias of central versus local government expenditures. The proportion of central health spending on curative interventions fell very sharply between 1993 and 1998 -- from 83.3 percent to merely 31.2 percent. Correspondingly, the shares of preventive programs and family planning increased - from 6.6 percent to 22.9 percent in the case of preventive programs and 9.5 percent to 25.7 percent in the case of family planning. Indeed, over this period, central spending on hospitals and curative care declined even in nominal terms. In contrast, the proportion of local health spending on curative interventions increased from 66.7 percent in 1993 to 81.5 percent in 1998. 63. It is possible that the growth of alternative financing sources for curative expenditures - in particular, user fees and health insurance - may have allowed the central government--in particular, the Ministry of Health--to cut back its allocation to the hospital and curative sector and focus instead on preventive programs and family planning. 64. Greater disaggregation of Ministry of Health expenditure on (vertical) health programs is shown in Table E.4. While overall spending on disease-control programs increased by 15 percent annually in real terms over the 1993-97 period, there were some important reallocations within this spending category. The malaria control program saw a 40 percent real decline in its, while funding for EPI, tuberculosis and leprosy control programs increased sharply in real terms expenditures. (Actually, the shortfall in budget spending for malaria is compensated by funding from the World Bank financed National Health Support Project). 65. Composition by Service Level. Unfortunately, budgetary data in Vietnam are not organized by service levels, so it is difficult to know how much of the state health budget is devoted to primary, secondary and tertiary care. A rough decomposition of budgetary health spending into spending by levels is shown in Table E.5. This decomposition is not perfect, as it is not easy to allocate many items of expenditure, such as training and communicable disease programs, to specific service levels. Therefore, the estimates shown in Table E.5 should be treated with caution and are only indicative. 66. These suggest that hospitals take up a disproportionately large share (75-87 percent) of the recurrent state health budget. There are two additional caveats to these numbers. First, the figures shown in Table E.5 only reflect recurrent health expenditures financed by the state budget (both central and provincial) and exclude the contributions of donors, user fees and health insurance. Since virtually all of the spending out of user fees and health insurance is on hospital care, it is likely that the share of hospitals in total 105 Vietnam Public Expenditure Review 2000 - Annexes recurrent public health spending is even greater than that shown in Table E.5. Second, first referrals to district hospitals are very much part of an integrated primary care system. But even if the spending on district hospitals is excluded, the share of the state health budget spent on secondary and tertiary hospitals (i.e., provincial and central specialty hospitals) remains very high-at about 50 percent. Table E.4: Ministry of Health expenditure on health programs, 1993-97* Iodine de- Leprosy Nutri- Year Malaria ficiency EPI AIDS TB tion Other** Total Expenditure in nominal terms (million VND): 1993 44,799 27,235 10,738 1,362 84,134 1994 46,693 17,518 32,961 9,880 107,052 1995 35,277 30,419 37,103 4,770 930 108,499 1996 43,232 28,796 40,690 6,766 22,073 2,689 1,904 14,606 160,756 1997 37,642 25,087 55,618 33,897 6,685 42,828 201,757 Expenditure in real terns (millions of constant 1994 VND): 1993 50,223 30,533 12,038 1,527 - - - - 94,321 1994 46,693 17,518 32,961 9,880 - - - - 107,052 1995 29,719 25,627 31,258 - 4,019 783 - - 91,406 1996 34,669 23,092 32,630 5,426 17,701 2,156 1,527 11,713 128,914 1997 27,780 18,514 41,046 - 25,016 4,934 - 31,607 148,898 % share of each program in total MoH expenditure on health programs: 1993 53.25 32.37 12.76 1.62 100.00 1994 43.62 16.36 30.79 9.23 100.00 1995 32.51 28.04 34.20 - 4.40 0.86 100.00 1996 26.89 17.91 25.31 4.21 13.73 1.67 1.18 9.09 100.00 1997 18.66 12.43 27.57 - 16.80 3.31 - 21.23 100.00 Notes: * exclusive of national program expenditure spent by provinces as authorized by the center **Includes the hospital equipment upgrade program. Source: Ministry of Health estimates. 67. The per capita availability of hospital beds in Vietnam is among the highest in the region (Figure E. 15). The only countries having a lower ratio of population to hospital beds than Vietnam are Taiwan, Hong Kong, South Korea, Singapore and Sri Lanka, all of which have considerably larger GNP per capita than Vietnam. Indeed, even Malaysia, Thailand and the Philippines have higher population to hospital bed ratios. Prima facie this suggests that there may be further scope for reducing the per capita number of hospital beds in Vietnam, primarily by holding back any expansion in the aggregate supply of hospital beds. 68. In contrast to hospitals, commune health centers account for a much smaller share of the state health budget. However, the commune health center share in the recurrent state health budget increased considerably over time--from 7 percent in 1991 to 12 percent in 1996, reflecting the fact that commune health center worker salaries began to be paid out of the state budget from 1994-95. 106 Annex E - Health Sector Table E.5: Recurrent health sector expenditures, 1991-98* Estimated actual expenditures Budget (planned) Program Category (million VND) (mill. VND) 1991 1992 1993 1994 1995 1996 1997 1998 Central administration and 2,730 3,450 4,100 4,500 4,600 5,900 7,700 7,700 management (0.40) (0.43) (0.41) (0.39) (0.34) (0.35) (0.44) (0.44) Commune health centers 50,000 50,000 70,000 70,000 150,000 200,000 200,000 240,000 (7.30) (6.22) (7.03) (6.02) (11.03) (11.79) (11.51) (13.83) Intercommunal polyclinics 14,942 21,347 22,185 29,974 24,798 23,565 24,000 (1.86) (2.14) (1.91) (2.20) (1.46) (1.36) (1.38) District health centers 86,100 46,100 36,960 83,880 156,200 273,643 231,143 (12.57) ((5.73) (3.71) (7.21) (11.48) (16.14) (13.31) All hospitals: 546,000 690,000 864,000 983,080 1,019,602 1,191,422 1,274,603 1,464,200 (79.73) (85.77) (86.71) (84.48) (74.95) (70.26) (73.38) (84.35) District hospitals 128,900 248,400 311,040 318,120 363,773 402,357 458,857 527,112 (18.82) (30.88) (31.22) (27.34) (26.74) (23.73) (26.42) (30.37) Provincial hospitals** 324,300 324,300 406,080 488,330 477,032 599,063 548,390 688,174 (47.35) (40.31) (40.75) (41.97) (35.07) (35.33) (31.57) (39.64) Other hospitals 92,800 117,300 146,880 176,630 178,797 190,002 267,356 248,914 (including central)*** (13.55) (14.58) (14.74) (15.18) (13.14) (11.20) (15.39) (14.34) Total 684,830 804,492 996,407 1,163,645 1,360,376 1,695,763 1,737,011 1,735,900 (100.00) (100.00) (100.00) (100.00) (100.00) (100.00) (100.00) (100.00) Notes: Numbers in parentheses are percentages of the total. * Includes state budget (central andprovincial) expenditures only. Does not include ODA sources. ** Includes both general and specialty hospitals. ***Includes teaching hospitals, specialty hospitals and hospitals for special populations. Source: Calculations by Le Duc Chinh, 1999. 107 Vietnam Public Expenditure Review 2000 - Annexes Figure 15- Persons per hospital bed, selected Asian countries, 1995 3839 3265 1500 1503 1522 1582 935 137l 791 579 591 208 215 276 283 360 380 427 I7I iall gx4yt< -O9fioRt s $ ", ,/ 2P 69. The large share of hospitals in the state health budget has major equity implications. As noted earlier, the major users of hospital-based services in Vietnam, as in many other developing countries, are individuals belonging to the upper consumption quintiles. In 1998, 36 percent of all hospital users were drawn from the highest income 20 percent of the population, while only 7 percent were drawn from the poorest quintile. The facts that (i) more than three-quarters of the recurrent state health budget is spent on hospitals and (ii) over a third of all hospital users in Vietnam belong to the highest income 20 percent of the population, implies that public health subsidies disproportionately benefit the most affluent sections of the Vietnamese population. 70. Obviously, this is not an intentional outcome of government policy. Indeed, it is completely contrary to the government's objective of targeting public health subsidies to the poorest segments of society. However, as noted earlier, hospitals throughout the developing world tend to attract more affluent patients because of their location (typically urban), perceived high quality, and high costs. Therefore, as long as government health budgets provide large general (i.e. non-targeted) subsidies for hospitals, public health spending will continue to primarily benefit the rich. Provincial Distribution of Health Expenditure 71. Equity in Provincial Allocations. Provincial per capita government spending on health varies greatly across provinces--from a low of VND 24,000 in Hung Yen to a high of about VND 133,000 in Ho Chi Minh City in 1997, but with the vast majority of provinces (83 percent) having a public spending level of VND 24,000-50,000. The high levels of public health spending in Ho Chi Minh City, Da Nang and Hanoi might reflect the fact that these provinces are also partly responsible for provision of public health services in surrounding provinces. 108 Annex E - Health Sector Figure E.16 Total public spending on health per capita in the provinces and 72. - Figure E. 1 6 shows that ~~~~provincial GDP per capita, 1998 public health spending per capita 20 in a province is strongly related to provincial domestic product _ l per capita, even after Ho Chi e Minh City, Da Nang and Hanoi -g are excluded owing to, the X special nature of their provincial 2 'IO health services. The positive association between government 40 health expenditure and provincial GDP per capita is not 20 1,000 2,000 3,000 4,000 5,000 6,000 7.000 9.000 entirely unexpected, since a Princ0al domic product pet capita (000 VND) portion of government health expenditure in a province is financed by provincial governments out of their own revenue sources. This allows local governments in richer provinces to spend more on health than those in poorer provinces. It is also likely that central government health expenditures in provinces, especially in the form of central hospitals and research institutes located in the provinces, are biased in favor of better-off provinces because these provinces have higher absorption capacity for funds owing to their better infrastructure, facilities, and administrative resources. Public and Private Spending on Health Figure E.17 73. An important policy question is the relationship Public and private health spending per capita, by province, 1998 between public and private (household) spending on health across provinces. Figure E.17 l 10 shows that provinces having 5 higher levels of private spending on health per capita also have higher levels of goverment health 6 expenditure per capita, which 5 implies that the inter-provincial 40 distribution of public spending on 30 health reinforces-rather than 20 compensates for-interprovincial 0 1,000 2,000 3,000 4,000 5,000 6,000 7.000 8,000 9,000 10,000 disparities in private health Private health spending per capita (000 VND) spending. In turn, this means that public spending on health is not being used appropriately as a policy tool for reducing the inequities inherent in private health-care spending. 109 Vietnam Public Expenditure Review 2000 - Annexes Provincial Spending Disparities Over Time 74. Mean provincial Figure E.18 government health expenditure Percentage annual increase in real public spending on health per per capita grew by about 14 capita across provinces, 1991-98, as related to the level of real public percent per annum in real terms spending on health per capita in 1991 between 1991 and 1998. Provinces that had lower levels 25 of public health spending in 1991 experienced more rapid 2g growth in public health ~ spending than provinces that .l 5 had lower levels of public c I n c spending in 1991 (Figure E. 18). As a result, provinciales disparities in per capita public g spending on health inarrowed anc0 ise 20ae5 B30h 35ple h40 9 45 50 io for Ptbhe spe.dmg on heapeta ('000 1994 VND) somewhat between 1991 and 1998. While this is a positive trend, it needs to be accelerated further by reviewing the norms used in allocating state health expenditures across provinces. Health Financing Policies 75. In the last decade, the goverment has vigorously pursued two policies for improved financing of the health sector-greater cost recovery via higher user fees and expanding health insurance coverage. Both policies have had important implications for the sector. 76. UJser Fees. Basic health care was fonr-erly free for all, regardless of a patient's ability to pay some or all of the cost of care. In 1989, a fee system was introduced in the three higher (district, provincial and national) levels of the health-care delivery system,"s requiring those patients who are able to do so to pay at least a minimal part of their health care. The handicapped, orphans, families of health personnel, individuals able to produce certification of indigency from their neighborhood or village People's Committee, and patients suffering from certain diseases are to be treated free of charge. 77. In 1995, the Ministry of Health issued user fee schedules for each type of consultation and each type of diagnostic test and procedure performed in clinics and hospitals. For inpatient services, there is an additional daily bed charge. User fees vary by hospital level (i.e., first-class hospital, second-class hospital, fourth-class hospital and polyclinic. etc.)2 In addition, the schedule specifies a range of charges -- not a single 27 Primary health care at commnune health centers continues to be free (in principle), but fees are paid at inter-communal po lyclinics. 2tThere are guidelines on how the user fees can be used, especially in hospitals. Of the total user fee collection. 70 percent is supposed to be used for improving the supply of consumables (drugs, blood supply, chemicals, X-ray materials, etc.) and medical equipment in the facility, 25-28 percent should be used to reward health workers in the form of bonuses, and 2-5 percent should be returned to competent 110 Annex E - Health Sector fixed charge -- for each type of service. All fees are required to be paid in advance by non-insured patients and patients not eligible for fee exemption. For the most part, patients are responsible for the purchase of drugs themselves, either from a private pharnacy or a pharmacy run by the public facility on the public premises. 78. User fee collections Figure E.19 have increased dramatically User fee collections, 1991-98 over time, but th 35,0 acceleration has been 300000 _7 particularly rapid since 1994 -6 (Figure E.19). Wvhile 250,000 E ! aggregate user fee 5 o collections increased at a 4 0 0 compounded real rate of 6.4 g 150 000 percent from 1991 to 1994, they increased at a real rate 2000 of 29 percent between 1994 50,000 2 and 1998. The increase in . t ,. . the ratio of user fees to total 0 0 1991 1992 1993 1994 1995 1996 1997 1998 public spending on health has been less meteoric. Indeed, the share of user fees in total public spending on health was declining from 1991 to 1994, but has been increasing since 1994. By 1998, user fees accounted for about 7.4 percent of total public spending on health.29 79. There are four important observations regarding user fees. First, because the price schedule issued by the Ministry of Health only indicates a range of fees that can be charged. Hospitals have discretion over the user fees they can charge, especially for technical services and laboratory tests. Central hospitals normally apply the ceiling price indicated in the price schedule, while lower-level hospitals charge the floor price. There are also differences in fee levels across regions. In the South, user fees tend to be higher than in the North for comparable services. 80. Second, the user fee schedule that is in use today dates back from 1995. Although the general price level has increased by about 23 percent since then, user fee rates have not been adjusted. As a result, user fee levels are regarded as being too low. Of course, total user fee collections have grown rapidly over time, in part because of an increase in the number of patients seen at public facilities but also in part because of an increase in the number (and the quality) of laboratory and radiology tests performed per patient. authorities to build a supporting fund for hospitals. Hospitals are not supposed to use any of the user fee proceeds for construction activities. 29Since user fees are largely collected by hospitals, the cost recovery rate of 7.4 percent underestimates the true role of user fees in financing hospital activities. In the last 5 years, user fees have become a very important source of hospital financing, accounting for nearly a third of all hospital revenue in 1998. This has allowed hospitals to significantly reduce their reliance on the state budget (from 76 percent of total revenue in 1994 to 47 percent in 1998). 11l Vietnam Public Expenditure Review 2000 - Annexes 81. Third, even though user fee levels may be low, average out-of-pocket costs for public health services--particularly, public hospitals--are large and onerous for the poor (World Bank 2000 Vietnam: Attacking Poverty). For instance, single visit to a public hospital takes up 22 percent of all non-food expenditure for a year for a typical person in the poorest quintile. A visit of standard (national average) quality would be even less affordable; it would constitute 45 percent of annual nonfood expenditure for an individual in the poorest quintile.30 In contrast, a person in the highest income quintile would pay only 4 percent of his or her nonfood expenditure on a public hospital visit of standard quality. 82. Fourth, although there is a formal mechanism in Vietnam for exempting certain classes of people, such as the poor, handicapped, war veterans, orphans, and individuals suffering from certain ailments (such as tuberculosis and leprosy) from user fees, it does not work well in practice. One study found virtually no correlation between fee exemptions and household income nor between the number of fee exemptions given and the general level of affluence of the commune (Ensor and San, 1996). Data from the VLSS 1998 show that exemptions from payment for drugs are non-existent in both commune health centers and public hospitals (Table E. 6) although MOH receives reports from hospitals indicating that such exemptions are granted. Additionally, even though a larger percentage of the poor than the better-off are exempted from user fees at both types of facilities, the variation in the exemption rate across economic groups is not as large as would be expected a priori from a program that is supposed to be mainly targeted to the poor. For instance, as many as 38 percent of individuals in the highest income quintile were fully exempted from user fees (but not drug charges) at public hospitals in 1998, which is an inordinately large leakage of benefits intended for the poor. The Government has decided to conduct a review of how the exemption system is working in practice. It has also issued an inter-ministerial circular, outlining a consistent mechanism for providing free medical services to the poor. Table E.6: Percent of users who reported no out-of-pocket payments for a visit to a government health facility, 1998 Per capita consumption expenditure quintile Health Provider Poor Second Third Fourth Richest Average est Government Hospitals % paying no fees 50 42 42 41 38 42 % paying nothing for drugs 0 0 0 0 0 0 Commune Health Centers % paying no fees 90 83 78 79 75 82 % paying nothing for drugs 0 0 0 0 0 0 Source: VLSS, 1993 and VLSS, 1998. 30The poor typically pay less for a health service contact than the rich, because the average quality of health services they purchase is lower. Using a national average level of out-of-pocket expenditure controls for quality variations across income groups. 112 Annex E - Health Sector 83. Health Insurance. Formal health insurance began in Vietnam in 1993, following a decree issued in 1992 that set up two separate insurance schemes: (i) a compulsory scheme covering primarily current and retired civil servants and employees of state and large (i.e. having more than 10 employees) private enterprises, and (ii) a voluntary scheme aimed at the remainder of the population, viz., farmers, school students, and family members of compulsory scheme enrollees. 84. The coordinating agency for health insurance is the Vietnam Health Insurance Authority (VHIA). Each province has a provincial health insurance office, which until recently was under the control of the Provincial Health Bureau. However, since 1999, the provincial insurance offices have been placed under the direct control of VHIA, as several of the local schemes were in financial difficulties. 85. Premiums in the compulsory scheme are paid from a 3 percent payroll tax, which is shared by employers (two-thirds) and employees (one-third). Enrollees in the voluntary scheme pay fixed annual premia that vary across provinces and according to benefit coverage chosen. In some provinces, annual premia are as low as VND 10,000, while in others they are VND 50,000. 86. Membership in VHI grew at double-digit rates from 1993 to 1996, but has since slowed down considerably (Figure E.20). In 1998, there were a total of 9.8 million enrollees, of whom 38 percent (3.7 million) were enrollees in the voluntary program. This means that health insurance covers about 12 percent of the country's population. Figure E.20 Enrollment in Health Insurance, 1993-98 Voll iE 4 nt roll 2 u t! t - ° 'Voluntary . ~~~~~~~~~~10 lolntar of tetl (i 35e 1993 1994 1995 1996 1997 199S 87. The compulsory scheme covers more than three-quarters of its target population (viz., salaried workers). However, the voluntary program is still in its infancy, and covers a mere 5 percent of its target population. Thus, most of the growth of social health insurance in the future will have to come from expansion in coverage of the voluntary scheme. Within the voluntary scheme, over 90 percent of members are school children. So far, the scheme has had very limited success in enrolling farmers and other rural residents. 113 Vietnam Public Expenditure Review 2000 - Annexes Figure E.21 88. Despite its relatively late Share of insurance and user fees in the financing of total public start, the health insurance 10 10 9 program has come to play an eInsurance important role in generating v 8 l Ue- fees 7 additional resources for the l7 health sector. In 1993, its first S 6 6 year of operation, it financed 5 about 3 percent of total public > I; spending on health in the c country. By 1998, this share had c i i increased to 11 percent (Figure E.21). Meanwhile, the share of user fees in total public spending 1992 1993 1994 1995 1996 1997 1998 grew from 5 percent in 1993 to only 7 percent in 1998. 89. Even these figures understate the very important and significant role that health insurance expenditure has played in financing hospitals in Vietnam. As hospitals have been forced to reduce their reliance on state budget subsidies, they have increasingly relied on health insurance and user fees to finance their activities (Figure E.22). Indeed, by 1998, health insurance reimbursements constituted 15 percent of total hospital revenue. 90. There are two main issues related to the health insurance program. First, the health insurance program primarily benefits the more affluent population in Vietnam, as enrollees in the compulsory health insurance scheme are civil servants and salaried employees of state and large private enterprises, most of whom are middle-class urban residents. This is borne out by Figure E.23, which shows that not only are health insurance coverage rates much smaller among the poorer consumption quintiles than among the richer quintiles, but that health insurance enrollees in the country are drawn disproportionately from the more affluent groups in society. For example, while only 8 percent of insurance enrollees are drawn from the poorest consumption quintile, as many as 37 percent are from the richest quintile. Thus, if the social health insurance program is to benefit the poor in Vietnam, there will need to be substantial expansion in coverage of the rural population by the voluntary insurance scheme. 114 Annex E - Health Sector Figure E.22 Figure E.23 The share of health insurance expenditure and state budget subsidies in total hospital revenue, 1994-98 H-g per pa epmot- qumbie, 19 80- 76 *Health 37 70 [User 35 - % of ndwidmls -th inane ca OState2) 60- 30 0 % of all insanc em,rlle dr,n finm qarnile 54 55 52 -o 26 %50- 47 -925- of tot 40 -2 al 1 _ 2 reY30j 2125 1 9& 10 11 Potst SODO Thid Forth Riche 1994 1995 1996 1997 1998 PrWcpta a,ticn eqxcture qtirwie 91. One implication of this income bias in membership of the health insurance program is that a large portion of insurance reimbursements go toward treating diseases and ailments of affluence, such as heart disease, cancer and diabetes (Dunlop, 1999).31 Health insurance reimbursements for cancer and heart disease treatments have lately averaged VND 1-3 million per case, while reimbursements for kidney dialysis have averaged VND 50-60 million per case (Dunlop, 1999). These diseases are not representative of the mortality or morbidity profile in the province. 92. Second, the health insurance program has run into major financial problems in part because of uncontrolled growth of insurance reimbursements. In 1996 and 1997, many of the provincial health insurance funds were overdrawn in comparison to the paid premiums during that year. In 1996, nearly 25 percent of provinces had insurance expenditures exceeding insurance revenues, while in 1997 about 10 percent of the provinces ran a deficit (Dunlop, 1999). While the overall financial integrity of the insurance fund was not in jeopardy, such problems are likely to keep cropping up in the future. One reason for the increase in insurance expenditures is the existence of consumer and provider moral hazard.32 Table E.7 shows that, while the annual outpatient service contact rates of health insurance enrollees were not significantly different from those of the general population in 1997, the inpatient admission rate was significantly higher (by about 70 percent). In addition, both the outpatient contact rate and the inpatient admission rate have grown much more rapidly over the period 1993-97 for the insured than for the 3"Actually, data on reimbursements by type of diseases treated are not reported routinely by the VHIA from its reimbursement records. However, the data reported by Dunlop (1999) were collected by the author for Phu Tho province in 1998-99. 32Moral hazard is the practice of excessive utilization of covered services by the insured. This can occur because (i) the insured demand more health services as the effective price of care for them is lower (consumer moral hazard), and (ii) providers deliver more services to the insured because they have a financial incentive to do so through the fee-for- service system (provider moral hazard). 115 Vietnam Public Expenditure Review 2000 - Annexes general population. Indeed, the inpatient admission rate for the general population has been more-or-less stagnant over the entire period, while that of the insured has grown by about 41 percent. The growth rate of inpatient admissions among the insured is probably unsustainable. 93. Table E.7 also indicates that the average expenditure on an inpatient admission for insurance program enrollees has been increasing much more rapidly than the average expenditure for an outpatient service contact. In 1993, the cost of an inpatient admission for an enrollee was about 9 times the cost of an outpatient service contact. By 1997, the ratio had increased to a factor of 12. The increase in inpatient expenditures probably reflects provider moral hazard, as hospitals respond to reduced public subsidies by increasing the number of billable and reimbursable services to insured patients. Table E.7: Service Use and Health Expenditures of Health Insurance Enrollees, 1993-97 Health Insurance Program General Health Program Enrollees Enrollees Population Insurance Year Average Average Ratio of Average Average Ratio of Average exp. Average exp. Ratio of average number of number of inpatient number of number of inpatient per outpatient per inpatient exp. per outpatient inpatient admissions outpatient inpatient admissions to contact ('000 admission outpatient contact contacts per admissions to out- contacts per admissions outpatient VND) ('000 VND) to average exp. enrollee per enrollee patient capita per capita contacts per per inpatient contacts per capita admission enrollee 1993 0.48 0.07 0.14 0.84 0.09 0.11 11.67 103.06 8.83 1994 0.95 0.12 0.12 0.98 0.08 0.08 14.12 122.11 8.65 1995 0.97 0.11 0.11 1.18 0.07 0.06 17.63 169.27 9.60 1996 1.20 0.12 0.10 1.49 0.07 0.05 18.08 193.95 10.73 1997 1.36 0.12 0.09 1.52 0.07 0.05 18.56 215.71 11.62 Source.- Dunlop, 1999. 94. In response to these financial difficulties, the provincial health insurance offices have been placed under the direct control of VHIA since 1999, and VHIA has tightened the guidelines regarding the control and use of funds. Several changes to the health insurance package have been implemented, including: * a 20 percent co-payment for those insured under the two schemes (with school children exempted), up to an annual maximum of six months' basic salary (about VND 860,000); * limitation on maternity benefits for only the first two children; * extension of waiting period (for benefits to commence) from one month to three months. 95. In addition, many provinces have placed their own rules to limit reimbursements. For example, Ben Tre has limited outpatient reimbursements to health facilities to 55 116 Annex E - Health Sector percent of the total facility charges, effectively raising the co-payment by individuals. The province has also placed a ceiling of VND 102,000 per case on reimbursements for inpatient care. A Menu of Reform Options There are several major issues emanating from the above analysis 96. At about US$6 per capita per annum, government health expenditure per capita is about one-half of the US$12 recommended for low-income countries for implementation of a basic package of health services. To achieve this level of spending, Vietnam would need to spend an additional $468 million annually on health. While increased levels of user fees and expansion of health insurance could meet some of this increase (see discussion below), the bulk of the increase would have to come_from the state health budget. Currently, households account for nearly four-fifths of total spending on health; an increase in public spending would also serve to reduce the financial burden of medical expenditures on households, especially those that are poor. 97. The efficiency and equity of public health spending are as important as its quantity. The current pattern of public spending on health is inefficient as public goods like preventive care, which should have the highest priority in public spending because of their substantial positive externalities, are under-funded relative to private goods (e.g. curative care for non-communicable diseases). The pattern of public spending is also inequitable as the public subsidies to hospitals disproportionately benefit the more affluent, while facilities more used by the poor, such as commune health centers, are subsidized to a small extent by the state budget. Additionally, cross-country data suggest that the public health-care system in Vietnam is overly hospital-intensive. Vietnam has more hospital beds in relation to its population than even Malaysia, Thailand and the Philippines-countries whose per capita GDP is several times larger than that of Vietnam. Not surprisingly, Vietnam has among the highest inpatient admission rates in the Asia-Pacific region. Shifting patients from high-cost, hospital-based services to more cost-effective commune health centers- and clinic-based health services will greatly enhance the cost-effectiveness ofpublic health spending. 98. An important and urgent step for the government would be to improve the composition of the public spending on health in favor of preventive care, especially at the secondary (provincial) and tertiary (central) levels. To some extent, the government has already begun this process. As noted in this Annex, the central government has already sharply reduced the proportion of its state budget going to curative care from 83 percent in 1993 to 31 percent in 1998. This has been possible because of the growth of health insurance and hospital user fees, both of which have become important alternative sources of revenue in the health sector and which are almost entirely spent on curative care expenditures. The curative share in the local (provincial) state budget, which is more than twice as large as the central health budget, has increased from 67 percent to 82 percent over the same period. Thus, a major task will be to control and reduce local government spending on secondary (provincial) hospitals. 117 Vietnam Public Expenditure Review 2000 - Annexes 99. The high degree of curative bias in provincial government health expenditures is symptomatic of a more general problem in the manner in which health expenditure decisions are made in Vietnam. While the Ministry of Health is in charge of setting health priorities and goals for the country, it has little control over the extent to which public spending on health is oriented toward these priorities. This is because more than two- thirds of the state health budget spending takes place at the local (provincial) level. Not only does the Ministry of Health not have any influence over the manner in which this spending is allocated across various functions and service levels, but the provinces have no legal obligation to even notify the Ministry of how much they spend on health. While the high degree of fiscal decentralization achieved by Vietnam is desirable, there is an urgent need for a stronger mechanism for ensuring that provinces make their health spending patterns consistent with national health priorities and strategies. 100. To reduce state subsidies going to the hospital sector, the government may need to explore the feasibility of making public secondary and tertiary hospitals into financially autonomous entities that receive some, but progressively declining, subsidies from the state budget. This would put additional pressure on hospitals to diversify their sources of finance and further reduce their reliance on local budgetary subsidies. 101. The issue of user fees is a complicated one. While the user fee schedule has not been revised upwards since 1995 despite an increase of 23 percent in the general price level, the volume of user fee collections has increased very rapidly in real terms, largely because of increases in the number of hospital patients, average length of hospital stay, .nd the number of diagnostic and other tests performed on patients. There is also a great dcal of evidence showing that actual out-of-pocket payments made by poor patients are \ cry large in relation to their incomes and substantially exceed the official levels of user [>es that are supposedly in effect. While a formal mechanism for exempting poor patients n-om user fees does exist, it does not work well in practice. Not only do a proportion of h. poor not receive these exemptions, there are large leakages of benefits intended for Aie poor to the non-poor. Thus, even though a strong case could be made on efficiency k,rt)ndsfor an immediate increase in hospital userfees, equity considerations necessilate 1ki, such an increase be implemented only after the fee exemption mechanism is ,C ealmped and significantly strengthened Otherwise, the economic burden of user fees on _l 'noor will only increase, and they could be priced completely out of hospital services. l; Improving the affordability of health care for the poor should be a matter of high ri.ority for the government. While there are a number of active programs to subsidize the ~ca]ih care costs of the poor, the fact remains that assistance to the poor remains very low i .hsolute terms. In addition, there is evidence of poor targeting of user fee exemptions I o he poor. The main problem is that the Ministry of Health does not have a uniform and .>l1sQistent mechanism to deal with the provision of health care to the poor. Nor is there is fconsistent criteria for identification of the poor throughout the country. This results in Pl: ovinces have virtually complete discretion over social assistance to the poor, with some provinces using traditional exemptions to help the poor and others relying on newer methods, such as subsidized health insurance cards. The fact that the financial assistance to the poor comes out of the local budget also means that more affluent provinces with I 18 Annex E - Health Sector larger budgets are able to spend more on assistance to the poor than disadvantaged provinces. 103. The government needs to establish a national strategy for assistance to the poor with their health-care costs. The government has already begun moving in this direction-the Ministry of Health, the Ministry of Labor, War Invalids and Social Affairs (MoLISA), and the Ministry of Finance have agreed on the terms of an inter-ministerial circular, to be issued soon, outlining a consistent mechanism for providing free medical services to the poor.33 Under the terms of this decree, health insurance cards will be granted to "hunger- stricken families" as well as about 30 percent of the poorest in all areas according to the standards set by MoLISA. The number of people to be assisted will be about 3 million annually. Around VND 120 billion will be used from the state budget to cover the cost of this program.34 While this is a very positive step in the right direction, the intended beneficiary group of 3 million persons represents less than 4 percent of the population, which implies that the scheme would cover only a small fraction of the poor in the country. 104. As soon as a program for assisting the poor with their health-care costs is in place, the government should consider raising hospital user fees, since this will allow the government to reduce its subsidies to the hospital sector. As noted earlier, reduction of subsidies to hospitals, especially at the secondary and tertiary level, will not only increase the efficiency and cost-effectiveness of public health spending, but will also improve the equity of these subsidies. While user fees are onerous for some users, such as the poor and people living in remote mountainous regions, they are not unusually high for most of the rest of the Vietnamese population. (For instance, a single hospital visit constitutes only 4-5 percent of annual nonfood expenditure per capita for a person in the highest income consumption quintile.) There is evidence that because user fees have not been revised since 1995 and are perceived to be too low, hospitals have increased their gross revenue from user fees by increasing the number of diagnostic tests and procedures performed per patient and by increasing the average length of hospital stays for inpatients. From a societal perspective, this is a wasteful method of raising finance. 105. Another means of raising revenues for the health sector will be through the expansion of the health insurance program, especially the voluntary scheme. While the health insurance program has expanded very rapidly in the last five years, starting from a zero base, the compulsory scheme within the program now covers over three-quarters of the target population (i.e. civil servants and salaried employees of large private and state- owned companies). Since the voluntary scheme covers only 5 percent of the target population, future increases in health insurance coverage will have to necessarily come from this population. 331nterview with the Minister of Health, as reported in Vietnam Economic Times, Vol. IX (16), April 19-25, 1999. 34This would amount to VND 40,000 per person per annum, which is the approximate annual premium for the voluntary insurance scheme.35 GER-the ratio of all children, including children over school age, in school at a given level to.the relevant school age population; NER-the share of school age children enrolled in school, excluding over age children. 119 Vietnam Public Expenditure Review 2000 - Annexes 106. There is also a need to use public spending as a policy tool for reducing inter- provincial disparities in private health-care spending and access to health care. Currently, government health expenditure reinforces--rather than compensates for-- provincial disparities in private health spending. Some inter-provincial disparity in government health expenditure is unavoidable, since a portion of provincial health spending is financed by provincial governments out of their own revenues, which tend to be strongly correlated with provincial incomes. But, in principle, it should be possible for the central government to reduce this disparity by adjusting the formula by which the health budget that falls under provincial control is allocated to the provinces by the Center. Unfortunately, this budget is allocated by the provinces on the basis of norms such as the number of hospital beds in the province (for curative expenditures) and on the population of the province (for preventive expenditures). These norms do little to redress the inherent inequality in provincial health spending per capita. It may be worthwhile for the government to reexamine the formula used by the government in allocation the state health budget across provinces. The formula should incorporate additional factors, such as the level of private health spending per capita, incidence of poverty, and the level of health infrastructure in a province, so that disadvantaged and under-served provinces can receive disproportionately larger allocations of the state health budget. 107. In the longer run (next 5-10 years), the government should rethink its roles in the health sector. It may make more sense for the government to reduce its role in the direct provision of curative, hospital-based health services, and focus on (i) provision of high- quality preventive health services that have a clear public goods character, (ii) greater financing of health expenditures, especially for the poor, and (ii) improving regulation and monitoring of the large and growing private sector in health. 108. Finally, the information base for analyzing the functional and other composition of government health expenditures and the incidence of public health spending is weak. The current budget classification system does not permit disaggregation of government health spending by the level of service (i.e. primary, secondary and tertiary)--something that is essential for determining the benefit incidence of public spending. While health budget data are classified into broad functional categories of preventive and curative expenditures there is not full agreement among government agencies as to what individual items of expenditure are aggregated into these broad categories. For instance, until recently, one of the national programs included under preventive expenditures involved upgrading equipment at public hospitals. 109. In addition to the weaknesses in budgetary data collection, information on health- service utilization and health status of different economic groups in the population has been seriously lacking until the release last year of the VLSS 1998 data. The government has in the past emphasized the routine collection of facility-based (administrative) data on utilization of health services. While such data are important, they do not permit analysis of income (and other) differences in utilization of health services. This makes it impossible to know which groups in the population are benefiting from government programs and facilities. Regular household surveys are often the only means of obtaining such information. The forthcoming National Health Survey--the first nationally- 120 Annex E - Health Sector representative multi-purpose health survey planned by the Ministry of Health in 2001- will go a long way toward meeting this data gap. However, it is essential for the Ministry of Health to establish a regular system of household surveys to monitor the impact of its programs and its spending on different socioeconomic groups. 121 Vietnam Public Expenditure Review 2000 - Annexes REFERENCES ChinH, Le Duc, 1999, "Hospital Activities in Vietnam," unpublished working paper. Dunlop, David W., 1999, "Situation Review of Health Financing and Health Insurance in Vietnarn," paper prepared for the Asian Development Bank Rural Health Project TA, Hanoi, June. Ensor, Tim and Pham Bich San, 1996, "Access and Payment for Health Care: The Poor of Northern Vietnam," International Journal of Health Planning and Management, volume I 1, pp. 69-83. General Statistical Office (GSO), 1990, General Population Census of Vietnam 1989, Hanoi. GSO, 1995, Viet Nam Intercensal Demographic Survey 1994. Major Findings, Hanoi, May. GSO, 1999, Population Projections of Vietnam. 1994-2024, Project report VIE/97/P`14, Hanoi, May. GSO, 1999, "Vietnam Living Standards Survey 1997-98," Report No. VIE/95/043, August. Ministry of Health, 1997, Health Statistics Yearbook 1997, Health Statistics and Informatics Division. NCPFP, forthcoming, Vietnam Demographic and Health Survey 1997, Hanoi. UNDP, 1999, Human Development Report 1999, Oxford University Press, New York. Vietnam Health Insurance Agency (VHIA), 1999, Vietnam Health Insurance Statistics 1999, Hanoi. Vietnam Health Sector Review (VHSR), 1999. World Bank, 1993, World Development Report 1993, Oxford University Press, Washington D.C. World Bank, 1999, Health, Nutrition and Population Indicators: A Statistical Handbook, Human Development Network, Health, Nutrition and Population Series, Washington DC. 122 Annex F - Education Sector ANNEX F - EDUCATION SECTOR a. Background to education provision in Vietnam 1. There are four levels of schooling in Vietnam - pre-primary, primary (five years from age 6), lower secondary (LSS) (four years) and upper secondary (USS) (three years). Access to levels above primary is by examination. Information on the current structure and institutions is given in Table F.1. Table F.1. Structure of education provision 1994 1998 Pre-school Teachers 71,749 144,540 Institutions 6,959 5375 Primary Teachers 288,173 324,545 Institutions 18,187 12,606 Lower Secondary Teachers 142,166 179,512 Institutions 6,727 Upper Secondary Teachers 37,065 46,979 Institutions 1,269 1,522 Secondary Technical Teachers 7,728 9,808* Institutions 244 239* Vocational Training Teachers 4,618 6,425* Institutions 198 174* Higher Teachers 21,700 24,083 Institutions 109 136 Source: MOET 2. Vocational education is offered in a variety of institutions, and is available after primary and LSS. The VTET (Vocational Technical Education and Training) system forms part of the lower and upper secondary cycles, in addition to skills training for adults. There are many training providers at central and local levels of government, and different ministries are responsible for meeting the training needs of their sectors. 3. Higher education is offered in colleges and universities, the main distinction between the two being that college courses last for three years, and university courses for four years or longer. Entry requirements are the General Education Diploma (after USS), the Vocational Education Diploma (after SVS) and the Technical Education Diploma (after STS). In addition there is a continuing education system which provides a variety of facilities 4. The Ministry of Education and Training has policy responsibility for the education sector. In broad terms, the Ministry manages higher and some vocational institutions, Provinces manage secondary schools and Districts and Communes manage the Primary schools and pre-primary facilities. But in practice, there are variations in individual Provinces. 123 Vietnam Public Expenditure Review 2000 - Annexes 5. By far the greatest part of provision is in public schools. A small, but increasing element is in semi-public schools where government provides the facilities, but parents meet the operating costs, including salaries. The private sector remains a small scale provider. An analysis of the composition of provision is shown in Table F.2. Table E.2. Enrolment Trends Actual 1994 1998 Public Pre-school 1,614,803 1.270,710 Primary 10,024,633 10,218,536 Lower Secondary 3,529,679 5,328,294 Upper Secondary 692,040 1.110.436 Secondary technical schools 119,000 164.100* Vocational training 62,614 102,535* Higher 225,540 647.103 Non-public Pre-school 18,026 1,312,527 Primary 22,931 31,678 Lower Secondary 149,055 236,594 Upper Secondary 170,960 547,272 TOTAL Pre-school 1,632,829 2.583,237 Primary 10,047,564 10,250,214 Lower Secondary 3,678,734 5,564,888 Upper Secondary 863,000 1,657,708 Secondary technical schools 119,000 164,100* Vocational 62,614 102.535 Higher 225,540 647,103 Source. MOET 6. Enrolments have been rising rapidly at post-primary levels. Table F.3 shows the evolution of participation ratios. The primary Gross Enrolment Ratio (GER)35 has declined as less over age children enter school, while the Net Enrolment Ratio (NER) has risen. The post primary enrolment ratios have risen substantially. Vietnam has high enrolment ratios for primary education compared to countries of a similar level of economic development, and, indeed, compared to much richer countries. The Gross Enrolment Ratio is a measure of the fiscal effort required to maintain the school system, and the high GER signifies that Vietnam is allocating resources to cover full enrolment of school age children at current levels of resource utilization. As the school system develops the number of over age children can be expected to decline, and the proportion of school age children in school can be expected to increase, so that GER equals NER. This appears to be happening at the primary level. 7. At the lower secondary level Vietnam also has a relatively high GER. indicating a strong effort to enroll children in a basic cycle of nine years of education. 124 Annex F- Education Sector Table F.3 Gross and net enrolment ratios 1993 1998 Gross Enrolment Ratio Primary 120 115 Lower Secondary 42 78 Upper Secondary 9 36 Net Enrolment Ratio Primary 87 91 Lower Secondary 30 62 Upper Secondary 7 29 Source: MOET 8. The gender composition in primary and lower secondary education of school enrolments is given in Table F.4. There is a fairly even balance between male and female enrolment in primary ahd lower secondary education and girls' shares in upper secondary participation are rising (note that girls form a slightly higher proportion of the school age population than boys). Table F.4: Gender balance of school enrolments 1993 1998 Primary Male % 49.8 50.4 Female % 50.2 49.6 Lower Secondary Male % 51.8 49.7 Female % 49.2 50.3 Upper Secondary Male % 57.9 52.3 Female % 42.1 47.7 Higher* Male % 59.0 56.3 Female% 41.0 43.5 * 1996 and 1998 Source: MOET 9. However, there are significant variations in enrolment ratios around the country, as is shown in Table F.5, although there has been improvement at all levels in all regions. Current Education Policies and Priorities 10. MOET has prepared an education sector draft strategy paper covering the period 2001- 2010 which is to be submitted to the Government this year. The following are the key elements: Enrolment objectives * full primary enrolment * increase enrolment ratio of lower secondary pupils to 88 percent in 2010 * increase upper secondary enrolment to 45 percent in 2010 Teacher quality 125 Vietnam Public Expenditure Review 2000 - Annexes * 40 percent of primary teachers to have a college degree by 2010 * 60 percent of lower secondary teachers to have a university degree by 2010 * 10 percent of upper secondary teachers to have a masters degree by 2010 Table F.5: Regional variations in (net) enrolment ratios 1993 1998 Primary Northern Uplands 85.7 94.1 Red River Delta 95.1 95.6 North Central 88.1 92.5 Central Coast 84.4 88.1 Central Highlands 67.8 80.2 South East 93.5 93.4 Mekong River Delta 79.0 86.9 Lower Secondary Northern Uplands 22.2 56.4 Red River Delta 46.5 83.2 North Central 30.6 62.3 Central Coast 38.1 64.2 Central Highlands 15.0 43.9 South East 35.2 71.7 Mekong River Delta 19.4 45.0 Upper Secondary Northern Uplands 5.6 22.2 Red River Delta 10.3 45.2 North Central 5.8 29.6 Central Coast 11.6 31.8 Central Highlands 2.0 10.5 South East 9.7 36.3 Mekong River Delta 3.6 17.4 Higher and Vocational Northern Uplands 1.4 5.3 Red River Delta 4.4 16.3 North Central 2.0 8.7 Central Coast 3.8 9.4 Central Highlands 0.0 1.6 South East 4.2 14.0 Mekong River Delta 1.9 5.5 Source: MOET Infrastructure * each comtnune to have at least one well equipped lower secondary school * each district must have at least one upper secondary school 126 Annex F - Education Sector Finance * make full use of foreign aid for school infrastructure; curriculum and textbook reforrn; teaching method reform; improving quality of teaching staff; vocational and technical education * establishment of professional and higher education institutions with 100% foreign investment, and co-financing of high technology centers * adequate budget for overseas training * education budget to reach 20 percent of total expenditures by 2010 (15 percent by 2000) * increase in non-governrnent element of school financing. Pre-primary to be financed from local communities, with state help for the poorest. Primary education to be free of charge but discretionary collection by schools to provide additional services. Districts and communes to be responsible for basic school infrastructure, and provinces for upper secondary infrastructure, with state aid to poorest areas. Increase lower secondary enrolments in semi-public schools to 20-40 percent of total * banks to create education credit funds for low income children * allocate budget on a per pupil basis Management * decentralize sector management * establish a Council for allocating resources to education and training * restructure and improve education management Other - development of distance education * international research cooperation * An implication of Government's policies is a requirement for substantial and growing resources in the sector. Although some of the burden will be shared with the private sector, significant and rapid increases in public sector resources will be required by the need to meet underlying growth in student numbers arising from demographic pressures, the broader participation implied and the costs of enhancing service quality. The affordability and sustainability of these policies is, therefore, an important issue. Framework for Delivering Government Policies 11. There have been a number of diagnostic studies of the sector during the last decade. In addition to government's own analysis, there have been the previous Public Expenditure Review carried out in 1995, the Education Finance Study coordinated by the Bank jointly with 127 Vietnam Public Expenditure Review 2000 - Annexes government and there is currently work going on in the context of the proposed Primary Teacher Development project. As a result, there is a fairly well defined set of variables that need to be addressed in achieving governments policies. These are summarized in Diagram F. I. This chapter considers the key issues categorized as follows * Issues concerning resource assignment to the education sector. * Issues concerning resource distribution within the sector and the impact of that distribution on service users. * Issues relating to the efficiency of resource utilization within the sector. * Outputs and achievements achieved as a result of that utilization. * The impact and implications of this analysis for the delivery of Governments policy objectives. 12. It will be noted in this report that considerable progress has been made during the last decade across all these fronts. In particular: * The sector is receiving more resources. * Investment levels have improved, although the standard and stock of classrooms and other facilities is still deficient. * Net enrollment rates have improved considerably across the whole of the education system . * The greatest increases in participation have been among the lower income groups although remaining higher overall in the higher income groups. * The high levels of participation achieved in primary is beginning to feed through into higher grades-particularly lower secondary. * Drop out and repetition rates are falling. * The better gender balance in primary is beginning to feed through into higher grades- particularly lower secondary; * Variations in regional enrolment and service performance have narrowed a little, but remain large as does regional access to funds for * education in both absolute and per capita terms. - The balance of public sector subsidies have shifted more towards the early grades. * The incidence of those subsidies has begun to shift towards the poor in the lower grades, but remains to the advantage of the better off in the higher grades. 128 Annex F - Education Sector * The burden of costs met by parents has reduced a little in the lower grades and increased in the higher grades. However, for the poorest the burden is still high, where 75 percent of households are food poor. So far, there is no clear sign that charge levels have prevented improvements in participation although they may now be contributing to regional variations. 13. In the light of these positive trends, this report will focus on next steps in accelerating the progress and initiating action where it appears most appropriate. 129 Vietnam Public Expenditure Review 2000 - A4nnexes Diagram F.1: Key issues identified in earlier analysis NATIONAL POLICY H1ollCy Issues OBJECTIVES - National objectives and targets are realistic Natlonal sPolicy Dev. m nte rnational Well placed to take lessons from Asian growth Experience economies. urns Low rates of return on education generally-particularly on Education Rtgbahly a function of labour market structuire and may change. EXPENDITURE ui. _ . | R~~~~~~~~~~~~~'esource Distribution and Equity iDntralisaion 13111111Wpita expenditures higher in wealthy of resources W rtuitions in performance linked to variations in Per Capita 411i4S*Mo not guarantee a minimum standard of funding at per pupil level. Incidence of subsidies greater at higher grades and tend to be pro- rich. Resource Utilisation Expen ure IfPli1;1se of drop-outs etc, high cost per cffl viency . graduation. Reasonably high PTRs in lower grades, but quite low for higher Teacher Ratio | grds;cope for temporary increase to fund service growth. .4 [Cet Input t Low shares for teaching materials, maintenance, investment PoiMl IWelkl classroom defijcit. ff|nan.ag9em_n7 |Higher Education in small, less efficient Institutional lArrangements I Beu0nentation of control over HEts. < ITeacher Qualityl | Low teacher salaries Motivation | Low input hours and service required from _ _ilrand Achievement Issues ojpu No systematic review of attainment levels other than quality Exams. Service Numerical access being improved, but service quality quality suffering. gfParticipatlon I Relatively high in Primary, but falls rapidly Afwedftfrother Asian Countries at this stage of economic development. JAccess l_ Relatively good gender balance in Primary, but falls into higher grades. JRe-Ontio- 7 High drop-out rates Low graduation from Primary to Lower and then to Upper Secondary ept5_ High repetition rates. F KLJKUAc)lLI I fT gl I Resounce Asslonment Issues Based on GDP forecasts in 1996, plans and projected growth MllSdly affordable RSPivatemi Low levels of private sector supply. Poor standards in semi- ASSIGNMENT public suppl pCiiols. arge* §iRelatively high in relation to total lxpn6 t h greater share of total cost at lower grades than tgbr4avel of exemptions based on u! >-sector | B5MfYor shifting public subsidy to lower grades and away share fiurhtlr Education which is largely used by the better off. Edecor Scope for share of Education Sector to increase share p t-aniii) o xp Under pressure from falling revenue yields in relation to GDP share growth. conomic Growth Vietnam share of GDP allocated to Education historically (GDPJ) low. 130 Annex F- Education Sector B. Level of and trends in public education spending Share of GDP and Public Spending Devoted to Education 14. The share of education expenditure in national output rose significantly between 1992, when public expenditure on education was about 2.2 percent of GDP, and 1994, when the share rose to 3.5 percent, a share which has been maintained, as shown in Table F.6. Table F. 6: Growth of public expenditure on education ACTUAL (B VND) 1992 1994 1998 GDP Current 110.5 170.3 361.5 1994 144.6 170.3 244.7 Total Public Expenditure Current 21.9 42.8 73.4 1994 28.6 42.8 49.7 Education and Training Sector Current 12.8 1994 3.1 6.0 8.6 Percentage Sector Expenditure of GDP 2.2% 3.5% 3.5% Percentage Sector Expenditure of Public 10.9% 14.1% 17.4% Table F.7 shows international comparable data for 199536 Table F.7. International comparisons of education expenditure Country Primary Secondary Education Education Share of school GER GER Expenditure Public education in as % GNP Expenditure education % total exp budget 1993-95 Average High Human 107 59 4.1 18 70 Average Medium Human 113 58 4.6 15 76 Average all developing 101 50 3.8 16 71 Industrial countries 100 99 5.2 12 71 Selected countries Hong Kong, China 96 75 2.8 17 56 Malaysia 91 57 5.3 15 76 Rep of Korea 101 101 3.7 17 79 Thailand 87 55 4.2 20 72 Philippines 116 79 2.2 10 n/a India 100 49 3.5 12 64 SriLanka 113 75 3.1 8 73 Vietnam 1998 118 32 3.5 17 68 Source: UNDP Human Development Report; World Bank World Development Report 36 Note that international comparisons use GNP rather than GDP - GDP shares will be lower in most countries. In Viet Nam GDP and GNP are roughly the same. 131 Vietnam Public Expenditure Review 2000 - Annexes 15. There are many factors which influence the share of national output allocated to education. Some countries have longer compulsory education cycles than others, for example, and wages and salaries are higher in some countries than in others in relation to per capita output. The table shows Vietnam to be spending a lower share than the average for developing countries, but it is also likely that the average duration of studies in Vietnam is lower than that of comparable countries. The school year is shorter in Vietnam than in many comparator countries. Table F.8 collates data for shares of education in GDP in comparator countries since 1950, and compares those shares with average years studied. Table F.8: Education expenditure and years of study Country Share of Education in GDP Average Duration of Studies 1950 1960 1970 1980 1992 1950 1960 1970 1980 199 High Performing Japan 4 4.1 3.9 5.8 4.9 10.2 11.1 12 12.8 13.7 Rep of Korea 3.1 3.5 3.7 4.4 6.3 7.5 8.8 11.3 13.1 Singapore 4.2 3.1 2.8 5.4 5.1 8.2 9 9.8 11.1 Taipei, China 2.6 2.8 3 5 7 6 7.7 9.4 11 12.1 Other Asian Bangladesh 1.5 2.3 1.8 2.3 3.3 3.9 5.1 China 1 1.2 1.9 1.7 1.3 2.8 5.4 7.3 7.8 Hong Kong 1.5 2.1 2.6 2.5 3.1 3.8 8 9 10.3 11.3 India 0.8 2.3 2.8 3 3.9 1.6 2.7 5.1 5.9 7.8 Indonesia 1.2 1.6 2.8 1.7 2.2 2.6 4 5.5 7.9 9 Malaysia 3.9 4.4 6 5.5 3.8 5.5 7.1 8.6 9.5 Pakistan 0.4 0.9 1.7 2 2.7 1.8 2.3 2.8 2.8 3.6 Philippines 2.4 2.6 2.6 1.7 3 7.4 7.5 9.6 10.9 11.6 Sri Lanka 3 4.7 4 2.7 3.3 6 7.4 8.3 10.3 11.5 Thailand 0.6 2.6 3.5 3.4 3.6 4.8 5.8 6.1 8.2 8.7 Asia Mean 1.7 2.8 2.9 3.1 3.8 4.5 6 7.3 8.6 9.7 Vietnam: Average number of years of education 1992 1998 Urban 7.1 8.0 Rural 5.0 5.7 Male 5.9 6.8 Female 5.0 5.6 Notes and Sources: Mingat A, Towards Improving Our Understanding of the Strategy of High Performing Asian Economies in the Education Sector, ADB, November 1995, Table 13, p 33, derived from UNESCO data. Expenditure is public expenditure only, and appears to include capital expenditure, though the report is not clear on this point. Vietnam data from LSMS Surveys 1993 and 1998. 16. Another ratio for intemational comparison is the average expenditure per pupil in relation to per capital GDP. In 1998 the average expenditures per pupil and primary, lower secondary and upper secondary levels as ratios to per capital GDP in Vietnam were 6.8, 8, and 15 percent respectively. In East Asia in 1990 the average for primary per pupil expenditure was 8 percent, with Korea reaching 12 percent and Malaysia 15 percent. The average for Sub- 132 Annex F- Education Sector Saharan Africa was 14 percent and for Latin America was 11 percent.37 There are many reasons for the variations, including, for example, high salary costs in Sub-Saharan Africa as a percentage of per capita GDP. 17. Taking the primary and lower secondary GERs into account, as well as the need to increase total instruction time and to raise real salaries, it is likely that the share of national output allocated to education will have to increase. An appropriate target over the next ten years would be to raise the share of education in GDP to 4 percent. Shares of Education Expenditure in Total Public Expenditure 18. Total education and training expenditure (recurrent and capital) currently amounts to about 17 percent of total public expenditure (excluding amortization), having increased over the period, as shown in Table F. 6. The share of discretionary expenditure (defined as total government expenditure less the interest costs of public debt) allocated to education and training, is only a little higher, because of the relatively low debt service burden. Recurrent expenditure on education is about 19 percent of total discretionary recurrent expenditure. 19. International comparisons of sector shares in total budget are complicated by relative levels of debt servicing. The appropriate measure of fiscal effort is the share of discretionary expenditure, because debt costs have first claim on the budget, and sectoral choices are made once those claims have been accounted for. Table F.7 shows a wide variation between the countries. It is likely that much of the variation reflects different budget structures.38 20. Government has introduced a target of 15 percent of public expenditure for the education sector by 2000. On the basis of 1998 data it appears that this target (which includes amortization) will be met in 2000. However, the draft policy statement suggest a perceived need for this to increase further to 20 percent by 2010. This also would be within the broad range of international comparisons and may well be necessary in the light of salary pressures discussed later.39 Recommendations - Resource Assignment * International comparisons suggest that to sustain the development of the education sector planned by Government, education spending should continue to rise over the period to 2005 to around 4 percent of GDP. * Action is necessary to ensure that the higher salary levels that are planned and necessary do not squeeze out non-salary and investment expenditures. 37 See Mehrotra S, Education for all: policy lessons from high achieving countries, UNICEF Staff Working Papers, 1998. 38 Some countries (such as Kenya) expend over 30 percent of their discretionary budgets on education while such expenditures reflect less than 20 percent of total budgets. 39 This target should be expressed in terms of discretionary expenditure (net of debt interest), and probably in terms of domestic revenue expenditures (net of foreign aid). 133 Vietnam Public Expenditure Review 2000 - Annexes * In order to achieve this it may be necessary to allow education expenditure as a percentage of public expenditure to move towards the top end of the 15 - 20 percent range indicated as appropriate by international comparisons i.e. to around 20 percent. c. Composition of educational spending 21. This section is concerned with resource distribution within the education sector, and whether the method of distributing resources actually achieves its objectives. The section concludes with an analysis of the norm system of resource distribution and ways in which that system might be modified in order to distribute education resources more equitably. The following section, section D, pursues the issue of resource distribution in terms of its regional impact and in terms of its impact on different population groups according to their economic status. Following that, some components of the proposed modified allocation system are discussed in terms of efficiency of resource use in section E. The three sections are therefore closely related to each other. 22. The underlying argument is that, looked at from the points of view of efficiency, effectiveness and equity, the present norm based system may not be achieving the objectives of all the stakeholders in the allocation system. The population based formula clearly has the intention to distribute resources equitably, but it may be that although the system does have an equity effect, in terms of educational objectives the effect may not be strong enough. Moreover, as we show later, other outcomes should be included in an effective formula, including that of the efficient use of resources. We consider the various pros and cons of the present system in terms of its outcomes. Subsectoral Shares of Allocations 23. Primary education takes up about 36 percent of total education and training expenditure, and lower secondary about 18 percent: basic education therefore accounts for 54 percent of education and training expenditure. When upper secondary education expenditures are added, total school education accounts for over 62 percent of education and training expenditure, and training expenditures account for the remainder. The shares of total sectoral expenditure are shown as well in Table F.9. 24. In general, therefore, although international comparisons should be treated with care, it can be concluded that Vietnam's fiscal effort on education, as well as its budget structure with respect to education, are in line with international experience, particularly when associated with school enrolment ratios: from table F.7 it can be seen that shares of school education expenditure are frequently in the order of 70 percent of budget. The primary share has fallen since 1992, as the shares of post primary expenditure have risen, although it seems that pre- primary education is also competing for a greater share. Overall, the share of the budget for school education has risen. Clearly lower secondary is going to compete for a greater share of resources, while at the same time a reduction in the primary share is undesirable. This suggests that there may be relatively little room for reallocation between subsectors, and that 134 Annex F - Education Sector finding needed funding will depend on economic growth, revenue effort efficiency measures within the sector. Table F.9: Subsectoral shares of education expenditure (recurrent & capital) Actual shares 1992 1994 1998 Pre-school 3.2 4.4 6.3 Primary 40 29.6 36.4 Lower secondary 14.8 17.7 17.6 Upper secondary 6 8.6 8.3 Vocational 13.9 11.2 6.7 Higher 16 12.5 13 Other 6.1 4.6 100% 100% 100% Source: MOF The Composition of Education Expenditures - School Education 25. The composition of expenditures refers to the relative shares of salaries and wages, recurrent operational, and capital expenditures. It is normal in education sectors, particularly at the lower levels, for expenditures to be dominated by salaries, because education is a very labour intensive industry. Developing countries typically spend over 95 percent of their primary education, and 60 - 80 percent of upper secondary recurrent budgets, on salary costs. 26. The published data for expenditures in Vietnam suggests that the effort to maintain a reasonable level of non-salary expenditure in the recurrent budget has been successful. However, much of the expenditure recorded as non-salary recurrent expenditure is personnel related, and would be properly allocated to personnel expenditures. The share of salary expenditures is therefore likely to be much higher than in the published data. 27. Moreover, the composition of non-salary recurrent expenditure is important. It can include administrative costs, and costs more directly related to teaching and learning. From the point of view of effective classrooms, it is necessary to be able to analyse what non-salary money is being spent on. Although the new chart of accounts permits such expenditures to be classified more clearly than before, the may still not meet management's requirements . Many countries have sector specific methods of analyzing the components of budgets, and there is much to be said for that practice. The education system could construct its own budget format which can easily be linked to the government format as a whole, as well as to the accounting system. 28. Vietnam has maintained significant capital expenditure levels: many countries at a similar level of economic development depend entirely on foreign aid for capital expenditures. In spite of this, there is an acute shortage of classrooms, and major maintenance requirements throughout the system. 135 Vietnam Public Expenditure Review 2000 - Annexes Table F.10 Input cost mix Input Cost Mix Actual Percentage share 1993 1998 Primary 61 55 Salaries and wages 24 15 Operational Expenditure 15 30 Investment Expenditure Lower Secondary 65 62 Salaries and wages 18 20 Operational Expenditure 17 18 Investment Expenditure Upper Secondary 46 47 Salaries and wages 22 24 Operational Expenditure 32 29 Investment Expenditure Higher and Vocational 22 12 Salaries and wages 42 74 Operational Expenditure 36 14 Investment Expenditure Source: MOF 29. It is widely accepted that the current level of teachers' salaries is not sufficient to adequately remunerate and motivate teachers. The problem is reflected in short teaching hours, and teachers' various methods of supplementing income, for example through private tuition (often of their state school pupils) or through various entrepreneurial activities. A comparison of salary levels in terms of ratios to per capita GDP is given in table F. 11. Table F.11: Teachers' salaries as a multiple of per capita output Country/Year Teachers' salaries as multiple of per capita GDP Primary Secondary High Performing Countries Japan 2.1 2.1 Rep of Korea 3.2 3.3 Singapore 1.9 2.0 Taipei, China 2.8 3.0 Other Asian Countries Bangladesh 3.1 3.6 China 1.5 2.7 India 2.9 3.1 Indonesia 2.7 3.1 Malaysia 2.5 3.2 Pakistan 3.4 3.4 Philippines 1.7 1.8 Sri Lanka 1.3 1.6 Thailand 2.2 2.3 136 Annex F - Education Sector Asia Mean 2.4 2.7 Vietnam (1998) 1.7 2.4 Source: Mingat, op cit, Table 16 p 45 Vietnam data derivedfrom MOF 1998 data 30. As Table F.11 shows, average teachers' salaries in Vietnam expressed as a multiple of GDP per capita, have been low relative to other countries. The government has raised the basic minimum wage from VND 144,000 per month to VND 180,000 (25 percent). As public sector salary scales are set as differentials to the basic wage, teachers' salaries will rise by an average of 25 percent. This will raise the ratio of salary to GDP per capita to nearer the international average.40 Expenditure Norms 31. There are a large number of norms referred to in the budgetary process for the allocation of education funds to (transfer norms) and by (allocation norms) provinces. If they were all strictly observed the result would be a highly standardized and unresponsive system. The main norms and their respective roles are shown in Appendix F.1. The main transfer norm is based on population, rather than the actual or potential number of students as discussed below. Allocation norms set guidelines on PTRs and class sizes and on the ratio of salary to non-salary spending. In practice some are of a mandatory nature, and some appear to be more advisory in nature. In some cases the difference norms potentially conflict-for example, are not mutually achievable within overall resource constraints. This leaves the managers at each level to reconcile such conflicts as best they can. For example it is not possible, if resources are insufficient, to attain simultaneously the 70:30 salary/non-salary norm, the PTR and class size norms, and the per pupil spending norms. 32. The presence of contradictory or unachievable, norms suggests that there is an inconsistency of objectives between the different stakeholders who manage the education system. The assumption is that there is a general agreement that the purpose of a norm based system is to help relate resource allocation decisions to policy. In a norm based fiscal transfer system, norms should reflect education sectoral objectives as part of an overall formula for allocation from the center, and at provincial and sub-provincial levels those norms should continue to be reflected in actual allocations. Negotiation will always be important in final allocation decisions, but norms should provide meaningful benchmarks against which negotiation outcomes can be measured, so that policy objectives are clearly reflected. Such a continuity between policy, norms and out-turn is not easily seen in the Vietnamese budgetary system. In principle it should be possible to relate ex ante facto unit cost norms with ex post facto average expenditures and to explain variations. 33. One important explanation of the variation between norms and out-turn is to extent to which regions keep surplus revenues. Where regions retain an element of surplus, they are in a position to supplement their budgets and this in turn will affect comparison of average expenditure out-turns between provinces. 40 Clearly, however, such comparisons do not reflect purchasing power parity. 137 Vietnam Public Expenditure Review 2000 - Annexes 34. There are few norms for allocation used by the MOF specifically to the education sector. Indeed, it is important to note that the actual distribution of per pupil expenditures is not a main concern of the MOF. Guidelines on target expenditures per pupil in the five geographic regions (mountain areas, etc) are issued, but no data are collected to determine whether budget out-turns reflect the guidelines. The eventual allocation for education is a result of negotiation between MOF and MOET, but is primarily based on a general population per capita and unrelated to the number of actual students . 35. A recent review41 of how budgetary resources cascaded through the education system showed how decisions about the mix of resources may be affected by the allocative mechanisms used at the different levels of government rather than conscious decisions about what mix will produce the best service. Some discussion of how this works in practice is shown in Appendix F. 1. 36. In particular, staff related costs are treated at each level in the cascade as fixed by the current and planned number of teachers. Investment expenditure is subject to a separate, parallel, and sometimes unrelated allocative process from recurrent expenditure. As a result, other important expenditure such as teaching and learning materials, teacher aids, furniture and equipment and building maintenance are treated as residual budgets and highly vulnerable to overall resource constraints. 37. It is also clear that there are big disparities in both resource distribution and educational achievement around the country. International experience is that tackling these disparities is likely to be critical to converting the rapid gains in overall terms in recent years into a well-balanced system offering, adequate service levels and equal opportunity. 38. There are positive aspects to the current system: . The population-based assessment of spending need currently applied by the center is relatively simple for the Ministry of Finance to operate and calculate. * It potentially leaves greater freedom to make expenditure decisions at the local level. * It also has the effect of making available resources in lieu of eligible children who do not participate in the system. This money is theoretically available to fund development and improvement expenditure in those areas with low participation in order to encourage greater participation. On the other hand, there are some serious drawbacks: * The level of the current per student expenditure norms has more to do with balancing to overall resource constraints than a consistent notion of what is required to deliver an acceptable level of service. 41 Education Financing Study: Primary Education in Viet Nam, for DFID, September 1999. 138 Annex F- Education Sector * The current general per capita is based on population statistics. A census every ten years requires estimation in between, and at the local level migratory patterns are very hard to estimate. * It creates some potentially perverse incentives. In particular, higher pupil numbers do not attract additional resources. * It creates a disconnect between national resourcing systems and the conveyance of resources to the working level of a student in a school. A report prepared in 1999 demonstrated that the process of cascading those resources to the school/student level gave rise to more than40 variables. All of these variables may be treated differently at different levels of the system with the end result that the level of resources that reaches the operating level is highly variable around the country for reasons that might be quite unrelated to the needs of the student. * The result of that variability is that for many students, the amount of resources made available is clearly not enough to provide an adequate standard of service. * The system is not transparent, and is not seen as fair by managers. * The system fails to create adequate incentives for efficient management. 39. In a mature economy where social participation and service pattems are reasonably settled, the case for a relatively simple transfer system to support local expenditures is perhaps clearer. It is less appropriate to a rapidly developing countries which require a more targeted approach. It is suggested, therefore, that there is scope for considering options that sacrifice a little of the simplicity of current arrangements in favour of ensuring that appropriate incentives and perceptions are created at the local level. 40. It would be more in keeping with the steps being taken to decentralise service provision if the structure of norms was simplified and mandatory norms clearly distinguished from advisory guidelines. The resultant flexibility is also likely to produce a more efficient service at the local level. It is suggested that mandatory norms should be based around ensuring that a minimum level of budgetary provision per student finds its way from the top of the system to the level of the school/student. 41. It is further suggested that while such norms (reflecting as at present different geographical characteristics) might usefully include within them a minimum level of non- salary expenditure. local management should be empowered to improve on that minimum level if they choose to by reducing staff costs per pupil (through higher class sizes and pupil/teacher ratios). In other words, norms are guidelines, and so long as (a) education outcomes are achieved; and (b) total budgets are not exceeded, local management should be permitted some discretion. The norm is what is believed to be required to achieve desired education outcomes, but if those outcomes can be achieved in different ways within the budgets set via the norms, there is no reason to insist on strict adherence to budget lines. 42. If it is agreed that a consistent set of per student norms is to be the focus of resourcing, it would be appropriate for such norms to be based on a clear and transparent view of the main 139 Vietnam Public Expenditure Review 2000 - A nnexes components of expenditure and a balanced mix of resources that should, under most circumstances, be capable of producing an efficient and effective service. 43. A per student based allocation systern designed to create equity and appropriate incentives might include the following characteristics: 44. In calculating central assessment of education spending needs, a distinction between adequately funding students within the system on the one hand, and providing additional funds to help provinces attract new students. The criteria for these could be quite different. 45. For funding students in the system, a per student per capita on a consistent basis from the top of the system to the bottom to ensure that at least a minimum amount per pupil arrives at the school level throughout the country. This could form a core unit cost, around which adjustments are made according to local circumstance. That amount should be related to what is calculated as being adequate to achieve a reasonable standard of service (including building maintenance). As now, this amount may be different in different parts of the country to reflect physical and geographical characteristics. 46. Provide separate funds to facilitate the tackling of local problems of access, participation and enhancing service quality. These funds could be distributed by direct reference to low wealth factors recognizing that the wealthier areas have ready access to additional local revenue. They might also be contributed to by donors in view of their catalytic value. 47. Widening the entitlement to fee exemption based on relative wealth and compensating less wealthy areas through the distribution formula. 48. A package of measures along these lines would need to be carefully balanced. But it need not add to expenditure requirements overall. For example, if it was necessary to make the package resource neutral overall, this could be achieved by rebalancing the levels of 'surplus' revenue that some provinces are permitted to retain. The purpose of such a package or variants upon it would be to maximize the distributional efficiency within the system in ensuring improved access to an adequate level of service throughout the Country. 49. An example of an algorithm for the construction of a per student based resource distribution norm that takes these considerations into account is shown in Diagram F.2. 50. Efficiency factors could include such factors as planned moves towards higher PTRs, weighting for repeating students (so as to encourage reduction), rationalization of infrastructure, etc. Further analysis of the key variables is included later in this chapter. 51. For technical and vocational, as well as higher education, norms are not set in terms of the five geographical regions as they are for school education, but in relation to training areas. However, the same issues apply. As is shown later in this report, there is a variety of expenditures per student across the training system, and that variety is not necessarily related to the cost of providing different types of training, nor to the efficiency of resource utilization. 140 Annex F - Education Sector Diagram F.2: An example of a calculation flow for a per student based allocation Loc ~ational Target / |~~Average| / Isale IN. Effiaency . e____ __ __ __ PER STUDENT level I lex~~ipenditurle 52. As we discuss below, there have been improvements in efficiency and in equity in the distribution of education resources over the period. These improvements have taken place in the context of the current system of population based norms for school education, and training area based norms for training. However, those improvements may not be the result of the norm system, but rather the result of the limited resources available, and the reaction of managers to the limited resources they are given. That reaction has at all levels been to seek ways of enhancing revenues through fees and charges, sales of goods and services, and private students. The question is therefore whether the current system permits an efficient, effective and equitable distribution of all resources, and whether the way government resources is allocated creates the right incentives to achieve such a distribution. The next sections consider this question. D. Equity of resource distribution 53. There are two related aspects of equity that we consider in this section. First, in a decentralized system, are resources distributed fairly between and within the provinces? Second, is the distribution of public resources for education progressive, in the sense that poorer people receive their fair share of benefits? Geographical Distribution 54. There is significant variation between and within provinces in the share of school age children in the total population. There are also significant differences in enrolment ratios-the proportion of school age children actually enrolled in school (see Table F.14). It follows automatically that a fiscal transfer formula based on population will tend not to provide an equal level of per student resource in the absence of adjustment factors to compensate for those variations. 55. This is confirmed in Table F.12, where wide variations in average budgeted expenditure per pupil are seen between and within the provinces. Moreover, as the table shows, there is a wide variation between per capita provincial GDP allocated to school education. The variations in average expenditures will reflect to some degree the different weights given to each region via the population norms. The question is how far the variations 141 Vietnam Public Expenditure Review 2000 - Annexes in per pupil expenditure assist in maintaining an equitable, efficient and effective use of education resources. Table F.12: Regional differences in average school expenditure Province Average Expenditures GDP Per Education %GDP %GDP %GD Primary Junior Senior Sec capita as % Primary LS P US Sec '000 Dong GDP Red River Delta 4,010 2.6% 0.9% 0.9% 0.4% 340 417 809 North East 2,321 5.0% 2.3% 1.5% 0.6% 374 466 956 North West 2,113 6.6% 3.7% 1.3% 0.5% 490 444 783 North Central Coast 2,368 4.1% 2.0% 1.2% 0.5% 315 346 742 South Central Coast 2,998 3.5% 1.5% 1.1% 0.5% 348 459 1046 Central Highlands 3,355 4.5% 2.6% 0.7% 0.5% 421 325 1124 North East South 8.598 1.5% 0.6% 0.3% 0.2% 423 484 1365 Mekong River Delta 3,811 2.3% 1.2% 0.6% 0.3% 340 401 930 3,697 3.8% 1.9% 0.9% 0.4% 381 418 969 Source. calculatedfrom MOF data 56. Table F. 13 compares the relative shares of provincial expenditures on school education and enrolments. It can be seem that some regions have a greater proportional share of enrolments in relation to total expenditure shares than others. Table F.13: Regional shares of expenditures and enrolments A B C D E F G H I Province Primary Education Lower Secondary Upper Secondary Education Education Share Share Ratio Share Share Ratio Share Share Ratio Total Total A/B Total Total D/E Total Total G/H Expend Enrol Expend Enrol- Expend- Enrol- itures ments itures ments itures ments Red River Delta 15.0% 16.1% 0.93 23.6% 23.8% 0.99 20.8% 24.5% 0.85 NorthEast 15.8% 15.3% 1.03 17.2% 15.6% 1.11 15.0% 15.0% 1.00 North West 4.7% 3.5% 1.35 2.6% 2.5% 1.06 2.1% 2.6% 0.82 North Central Coast 13.0% 15.0% 0.87 13.4% 16.3% 0.82 11.8% 15.2% 0.78 South Central Coast 8.2% 8.5% 0.96 9.5% 8.7% 1.09 9.0% 8.2% 1.10 Central Highlands 6.0% 5.2% 1.16 2.6% 3.4% 0.77 4.2% 3.6% 1.18 North East South 17.0% 14.6% 1.16 14.6% 12.6% 1.15 21.0% 14.7% 1.43 Mekong River Delta 20.4% 21.8% 0.94 16.4% 17.1% 0.95 15.9% 16.3% 0.98 1.00 1.00 1.00 100.0% 100.0% 100.0% 100.0% 100.0% 1.00 57. The conclusion from the foregoing discussion is that consideration may be given to specifying more clearly what is expected from the budget allocation system in respect of per pupil expenditures. As discussed earlier in this chapter, do the variations reflect policy decisions? 142 Annex F- Education Sector Benefit incidence of Public Expenditure 58. This analysis of the impact of the distribution of central government allocations on relative provincial shares of public finance is reflected in the data from the LSMS.42 Table F. 14 shows that net enrolment ratios by consumption quintile have risen in the lower quintiles. Table F.14: Net enrolment ratios by quintile Level Consumption quintile Vietnam 1 2 3 4 5 Average Primary 1992 72.0 87.0 90.8 93.5 95.9 86.7 1998 81.9 93.2 94.6 96.0 96.4 91.4 Lower secondary 1992 12.0 16.6 28.8 38.4 55.0 30.1 1998 33.6 53.0 65.5 71.7 91.0 61.7 Upper secondary 1992 1.1 1.6 2.6 7.7 20.9 7.2 1998 4.5 13.3 20.7 36.5 64.1 28.6 Higher & vocational 1992 0.0 0.4 1.5 2.1 8.9 2.9 1998 0.4 0.6 2.9 8.3 28.9 9.3 Source. Vietnam Living Standards Surveys 1993 and 1998 59. The analysis of the survey data suggests that there is a movement towards a more 'pro- poor' balance in the benefits from public expenditure on education, corresponding to the trends in enrolment ratios. Table F.15 shows, for example, that the incidence of public spending on primary education on the poorest 40 percent of the population rose by 7 percent over the period, and that the poorest 40 percent benefited from over half the expenditure. There has been a shift towards the poor since 1992, though benefits at post primary levels still, on the basis of the data in Table F. 15, favor the better off on the whole. Table F.15: Lorenz curves for education Accumulative Benefit Incidence By Quintile 1 2 3 4 5 Primary 1992 19.9 43.6 64.5 82.6 100% 1998 26.1 50.7 71.4 87.6 100% Lower Secondary 1992 7.2 17.8 37.7 66.2 100% 1998 12.7 31.8 55.1 79 100% Upper Secondary 1992 2.1 7.8 15.6 39.7 100% 1998 4.4 15.1 31.9 62.2 100% Higher and Vocational 1992 0 3.3 14.6 32.5 100% 1998 0.6 2.4 8.6 28.3 100% Source: Vietnam Living Standards Surveys 1993 and 1998 60. Thus, taking Tables F.14 and F.15 together, not only have participation rates between the wealthy and the poor moved closer across all grades, but the balance of subsidy has also shifted towards lower grades where participation by the poor is strongest. Further improvements can be expected as pupils in the lower grades, containing larger numbers of students from poor families eventually move through the system, provided steps continue to remove the barriers to higher grades of education for the poor. Whereas, for example, in 1993 the poorest 20 percent of the population accounted for 2.3 million enrolments (23 percent of 42 See Nga N N, Vietnam: trends in the education sector during 1993-98. Evidence from the Vietnam Living Standards Surveys 1993 & 1998, draft, World Bank Vietnam, 2000 143 Vietnam Public Expenditure Review 2000 - Annexes total) in primary school, by 1998 this number had increased to 2.8 million (27 percent of total). 61. Standard benefit incidence analysis for education assumes a constant average expenditure per pupil by level across all consumption quintiles. However, there are significant geographical variations in average expenditure. There are also variations in average expenditures between consumption groups. Moreover, standard analysis assumes a constant share of school age children throughout the country. It is of interest to interpret the standard results taking into account these differences. 62. Table F.16 compares data across regions43, and shows the percentage share in public expenditure for education on basic education, as well as regional shares of total population, and of population of school age. Table F.16: Regional benefits from education spending Region 1993 Northern Red North Centra Central Sout Mekong Vietnam Uplands River Coast I Coast Highlands heast Delta Delta Central Primary 22 19 11 11 3 17 17 100 Lower secondary 10 30 19 13 1 12 14 100 % of pop. 16 22 13 12 3 13 22 100 % of school-age 17 21 14 1 1 4 12 22 100 1998 Primary 24 16 14 9 5 12 19 100 Lower secondary 24 17 13 11 4 11 20 100 % of pop. 18 20 14 11 4 13 21 loo % of school-age 21 19 16 10 4 10 20 100 Per capita 1,920 2,939 2,197 2,640 1,942 5,052 2,548 consumption Source .V ietnam Living Standards Surveys 1993 and 1998 63. Table F. 17 shows the change over the period in regional shares in public expenditure on primary education, making adjustments for the factors described above. The regions have been re-ordered according to per capital consumption in 1998. The poorest region by this measure, Northern Uplands received a declining share in public expenditure for primary education. The next richest, on the other hand, Central Highlands, received a considerable boost, with the next two receiving smaller increases. The richest three regions received declining shares. 43 Note that to compare regions between 1993 and 1998 requires standardization for the boundary changes. 144 Annex F - Education Sector Table F.17: Benefits from public expenditure by region (primary education) 1993 1998 Percentage Change adjusted for school age Standard Adjusted Adjusted Standard Adjusted Adjusted spending for for spending for for ipop population school age /pop population school age Northern Uplands 1.06 1.43 1.31 1.25 1.35 1.16 -12% Central Highlands 1.12 0.87 0.74 1.26 1.37 1.13 52% North Central Coast 1.06 0.84 0.79 1.17 1.01 0.90 14% Mekong Delta 0.98 0.78 0.78 0.96 0.90 0.96 23% Central Coast 0.91 0.92 1.03 0.91 0.82 0.86 -16% Red River Delta 0.97 0.90 0.91 0.88 0.82 0.85 -7% Southeast 1.00 1.31 1.41 0.71 0.98 1.26 -11% Vietnam 1.00 1.00 1.00 1.00 1.00 1.00 Source: Vietnam Living Standards Surveys 1993 and 1998 64. It may thus be concluded that although there have been substantial improvements in relative shares of benefit from public expenditure over the period, there is still a challenge to direct resources towards the areas which most need them. Cost sharing with Parents 65. Household expenditures may be divided between compulsory and voluntary expenditures, and there are in addition quasi-compulsory expenditures, where, for example, social or school pressures rather than legislation require parents to spend money on education (for example, uniforms). It should also be remembered that government derives its revenue largely from taxes, and that cost sharing is a term used to denote expenditure by households outside the tax system. Furthermore, in considering the extent and nature of cost sharing, it should be borne in mind that higher 'private' spending does not axiomatically entail better quality education, both for reasons described above, and because much household spending has nothing to do with quality (e.g. purchase of uniforms, access costs (transport, lodging), etc). Table F. 18 shows how households distribute their expenditures between items. Table F.18. Composition of private expenses, by level (% of total private costs) 1993 Fees PTA Uniform Books Transport Lodging All Primary 5.2 10.6 11.8 23.1 0.2 26.8 100.0 Lower-sec 12.8 6.2 13.6 19.8 1.1 19.7 100.0 Upper-sec 16.9 4.3 12.4 15.4 0.8 15.1 100.0 Post-sec 32.0 3.7 4.8 12.6 2.9 25.8 100.0 1998 Primary 3.9 13.1 10.9 13.4 20.1 14.1 100.0 Lower-sec 11.6 10.2 10.7 10.8 17.3 11.3 100.0 Upper-sec 18.1 8.0 7.5 7.9 17.5 7.1 100.0 Post-sec 33.8 1.1 0.7 5.3 38.4 2.9 100.0 Source : Vietnam Living Standards Surveys 1993 and 1998 145 V ietnam Public Expenditure Review 2000 - Annexes 66. A high proportion of the total cost of education at all levels is born by parents in the form of a variety of charges or directly in meeting costs such as textbooks. However, in termns of shares in total expenditure on education, tax and non-tax financed, government has assumed a greater share of basic education expenditures, while financing a lower share of post-basic expenditures (see also higher education section), as in shown in Table F. 19. Table F.19: Shares of expenditure in total expenditure 1993 1998 Primary Public Subsidies 45 61 Tuition Fees 3 1 Direct costs to the parents. 52 38 Lower Secondary Public Subsidies 34 42 Tuition Fees 9 7 Direct costs to the parents. 57 51 Upper Secondary Public Subsidies 40 33 Tuition Fees 10 13 Direct costs to the parents. 50 54 Higher and Vocational Public Subsidies 71 46 Tuition Fees 9 1 8 Direct costs to the parents. 20% 36% 67. Table F. 19 illustrates that official private sector involvement in supply remains low although it is well known that many teachers provide supplementary education on an informal basis for parents who pay. Over 11 percent of total expenditure per primary child appears to be accounted for by private tuition charges, with richer households paying on average twice that percentage. In many provinces some semi-public places may be occupied by children who fail graduation tests. Furthermore, post primary level fees are used by schools to supplement teachers' salaries. Parents may thus be paying for tuition directly as well as indirectly. 68. Diagram F.3 shows the evolution of relative shares of tax financed and household spending between 1993 and 1998 in each region. There is considerable variation between regions, though the variation in the levels of public spending per student seems to have decreased. 146 Annex F - Education Sector Diagram F.3: Public and private average expenditures on primary education 1200 *50 _- U Pr0l 199- ------ ----_0 1993 00 I - -- Publ- ---- -- 3x ---rivate7 --- 800 - 250 j- - - lbl 1 600 -- - -- - - *42° ° C 150 -- -- - | i------ -l :400 20(: --- ~ ~~~~100 -- --- - - - 200 -~~~~~~~~~~~5 - 0 0 II IV V VI \Al I 11 111~~~~~~~~~~~~~~~~~~I IV V Vl Vl V° Region I 11 Region VI Vll 69. However, expressed as a percentage of household non-food expenditure, primary costs to the lowest consumption group have risen, though lower secondary costs have fallen. Upper secondary costs have risen as a proportion of non-food expenditure. At the same time, 75 percent of the poorest words, even though public expenditures as a share of total costs per basic school pupil rose, the real burden of household costs also rose. 70. It is not possible from the LSMS data to determine whether households are forced to dissave in order to finance education or health charges. PPA studies suggest that poor households accumulate debt, most frequently to pay for medical expenses.44 It is important to recognize that education costs to households cannot be considered in isolation: while on the surface it may appear that costs are reasonable in that enrolments are rising but the reality is more complex; as poor households are paying a higher share of their income for their children's education because of the high priority they attach to it. 71. Table F.20 indicates that the incidence of exemptions from charges is still far from pro-poor without significant differences between many of the consumption quintiles. In part at least, this is likely to be due to the reliance of poor provinces and communities on charges as a source of finance that they are reluctant to give up. 72. It is quite probable that further narrowing of the participation gap around the country will be hindered unless measures to enhance the ability/incentive to waive charges for the less well off are taken. Proposals made later for a better balancing of resource distribution may make it possible to introduce enhanced and selective arrangements for waiving charges. 44Vietnam: Attacking Poverty, pp 28-29. See also Vietnam Living Standards Survey 1997-98, table 2.10.3 147 Vietnam Public Expenditure Review 2000 - Annexes Table F.20: Exemptions for Charges for School Education % of children 6-14 1998 Vietnam Consumption Quintile 1 2 3 4 5 Average Partial exemption 8.9 10.7 13.4 9.9 8.6 10.4 Full exemption 11.6 6.5 3.6 3.6 0.9 5.8 Total of both 20.5 17.2 17.0 13.6 9.5 16.2 Rural 9.1 Urban 17.5 Source: Vietnam Living Standards Surveys 1993 & 1998 73. The higher education and technical and vocational subsectors have also seen a rise in cost sharing. In common with higher education institutions, vocational technical education and training schools generate additional revenues through privately paying students, contract training and sales of goods and services. Some schools earn over 75 percent of their operating expenditures in this way, but there is a wide variety of practices. A significant proportion of fees goes to supplementing teachers' salaries. Fees are charged for all short courses, and are paid by employers or students themselves. Typical examples are reported to be VND 40,000 for a one month agricultural course; VND 100,000 for a one month sewing course; or VND 2.5 million for a 9 month driving course. In addition students are expected to provide their own books and learning materials. 74. Better data are available from the 1999 annual survey of higher education institutions on which to base conclusions about cost sharing in universities and colleges. Figure F. 1 shows how the relative share between fees and government revenues in colleges and universities evolved, with the share of fee income rising. Fuller details of the data and the analysis are to be found in Annex F.2. Figure F.1: Evolution of higher education revenues 2500 ~~~-Total revenues 2,500 - Revenues from governt 800,000 ___Revenues fromfees 2,000 FTE students - 700,000 2,000 2- - - - - - - - ~ 600,000 c 1,500, i500,000 400,000 . § 1,000 -- ~ + 300,000 H 500 __--- 1 200,000 500_---- 1 - 100,000 1996 1997 1998 75. At the same time, although the tertiary education samples are small, the LSMS data make it clear that higher education is severely restricted to those who can afford to pay. 148 Annex F- Education Sector Training scholarships are treated in many different ways. In some technical training institutions fees are charged according to ability of students to pay, and in others students are awarded grants according to academic performance. Other schools charge all students a flat fee, not distinguishing between quota and non-quota students. Scholarship expenditure in universities and colleges was in 1998 the equivalent of some 15 percent of fee income, but, on the evidence of Table F. 14 above, the impact of scholarships on access of the poor to higher education is probably slight. Although one criterion for scholarships is financial need (the other is merit), 'social scholarships' do not appear to have an impact on the distribution by consumption quintile of students in the system. It may be that family income is not given a heavy weight in determining eligibility, and that other criteria such as ethnic minority status, and war related factors have more influence on the decision. Gender Balance 76. The indication from the data contained in Table F. 4 is that access to primary education is reasonably well balanced in terms of gender. However, there are provincial variations. Gaps between male and female participation at higher levels have been closing, with females accounting for over 43 percent of university and college enrolments. No data are available on the percentage of females undertaking vocational training, though the relatively low proportion of female vocational trainers (about 20 percent, as opposed to 46 percent in secondary technical schools) would suggest that attracting females to vocational training schools is problematic. 77. Many of the reasons for these gaps are of broader social and cultural origin. There are no specific programs to tackle gender participation and the main instrument being used by Government is one of universalization of early grades of education as a way of priming higher participation in higher grades. To some extent this seems to be working. As larger cohorts move from primary to secondary, it clearly includes a higher proportion of females. 78. There are a number of ways that Government might seek to provide a further catalyst for greater female participation. For example it might seek to skew expenditure towards females in school, either by deliberate resource allocation or by measures such as streaming on a gender basis entry into higher grades. But such measures can be difficult to manage and are likely to have implications for overall expenditure efficiency. In addition, it is likely that, given the fairly even gender balance on average, imbalances are localized, and local initiatives the most suitable response. 79. It may be more appropriate to focus on creating positive discrimination through charges made to parents. A deliberate policy of charging less for female students might be easier to manage and would tackle in part one of the perceptions of parents concerning low returns on education for females by lowering the cost of that education. Recommendations - Equity of Resource Distribution * There is little scope to shift resources between sub-sectors without damaging service levels. Therefore, the Sector is critically dependent on economic growth and a greater share of that growth for additional resources. It will need to concentrate this growth in 149 VieInam Public Expenditure Review 2000 - Annexes . * resources to fund higher participation in Secondary and Higher grades and to service improvement at all levels. * There are two underlying elements to education spending need-funding for students in the system on the one hand and development of the system on the other. The resource distribution system should deal with these two elements differently. * Fusnding for students in the system should take account of geographic factors such as sparsity of population , and appropriate incentives for more efficient operation. Having done so, a minimum amount of funding per student should be guaranteed at the school/student level. * Funding for developmental and improvement expenditure should reflect a pro-poor bias (perhaps using Provincial GDP per capita). An element of this second component of funding should include a fund for measures to attract and retain new pupils. Donors should be approached to provide support for this fund in order to assist further improvements in the sectors performance. * Incentives should be created for education managers to achieve results. Consideration should be given to experimenting with a third element of funding which would be initially small and perhaps confined to Upper Secondary and Higher Education. This third element would provide additional investment resources based on the level and quality of graduates achieved rather than just students attending. * A more equitable funding base along the lines outlined above should enable more extensive waiving of fees at local level (for example, in Education Priority Areas). * Part of this more extensive waiver regime could be preferential treatment for female students entering higher education. This would create positive discrimination to overcome current gender participation disparities. * As a higher proportion of students move into upper secondary and higher education, care is necessary to extend fee waiver to avoid the system becoming two-tier at those levels . There is a danger that provision of fee paying places ( often of low quality and used largely by students who fail examinations) and the use of examinations as a device for rationing access to public provision, will have the combined effect of creating a two tier system of access based on wealth. e. Efficiency of resource use 80. Efficient utilisation of public resources is not only desirable for its own sake: measures which reduce unit costs make education more accessible to poor people. The allocation system can also have an important effect on the efficiency of resource use. The following paragraphs consider in more detail factors influencing the efficiency of sectoral resource utilisation, and provide further analysis of the variables described in Diagram F.2 above. 150 Annex F - Education Sector Staff Utilisation 81. Education budgets are dominated by staff related costs, as we have seen, more so at the lower levels of the system. The two critical variables which influence the unit cost per pupil are the average salary of a teacher, and the number of pupils taught by each teacher: these variables are themselves subject to a number of influences. Therefore, from a budgetary point of view, these are the two main issues of policy. From the point of view of educational outcomes, the issues revolve around how the budgetary variables and outcomes relate. Is the teacher remuneration structure consistent with incentives to teach well; are classes too large or not large enough? Is the curriculum too extensive, or too narrow? Salaries 82. The average teacher salary across any given location (country, region, province, commune, school), is the first critical variable. What are the influences on that average? The variations in average expenditure per pupil across regions suggests also that there may be a variation in the average expenditure per teacher. The government is currently considering the school teacher remuneration structure and in doing so will need to relate any changes to the effect on averages in regional budgets. At the higher levels of education the degree of salary supplementation is considerable, and needs to be taken into account when developing a remuneration policy. 83. Teachers' salary scales are relatively compressed, and generally speaking related to length of teaching experience and seniority, which is common in most countries. They are not related to teaching competence and learning outcomes, for example. When considering changes to the remuneration structure, the effects on average salary at a given location need to be considered. Average salary in any given location will be mainly affected by teacher qualifications, seniority and length of service. In a province or school any changes to the average will change unit costs per pupil, other things being equal. An ideal remuneration policy would be one that permits discretion to local managers in respect of the average salary (e.g distribution of experienced teachers), while at the same time providing financial (there are also non-financial) incentives to teachers 84. There will always be a variation around the average. The question is, how is that variation to be distributed? What should be the differential between rural and urban teachers' average salaries in a given location, for example? As there is trade off between the unit cost of a pupil and the number of enrolments that can be adequately financed, and as the critical issue is that of learning outcomes and how they relate to inputs, excessive rigidity and automaticity in teacher remuneration structures can create perverse effects at the local level. 85. The complex issue of salary policy for colleges and universities is discussed in Annex F.2. Salary supplements paid out of non-government income can double staff salaries, on top of which those staff who can work outside in other jobs can seek further supplements. The rate of supplement seems to vary according to location and size of university, and, in that in some institutions government income accounts for a higher share of total income than in others, the proportion of fee income devoted to additional personnel remuneration will also vary. 151 Vietnam Public Expenditure Review 2000 - Annexes Pupil Teacher Ratio (PTR) 86. Pupil teacher ratios are one of the most difficult, and neglected, aspects of education management, but they are at the heart of resource allocation issues: there is always a trade-off between the number of pupils per teacher, and the availability of resources to finance non- salary items essential for learning and good teaching. Where resources are severely limited, as in Viet Nam, management of the PTR is a critical factor. Table F.21: PTRs in selected countries Country 1985 1995 Primary Secondary Primary Secondary Korea 38 35 32 25 Malaysia 24 22 19 19 Philippines 31 32 35 35 Indonesia 25 15 China 23 16 25 17 India 45 26 47 32 Japan 24 18 18 Hong Kong 28 24 24 20 87. International experience is mixed. In earlier years, Korea's was higher: nearly 50 in 1980. Underlying the figures are many factors: at the secondary level, for example, the subject mix of teachers. In general, however, PTRs, where they have been fairly high, seem to decline slowly as resources become available. However, a more detailed look at PTRs around the world does tend to confirm that countries with low PTRs have high salary shares in their budgets, often accompanied by problems of financing teaching and learning inputs. 88. At the primary education level there is more or less one teacher per class: in other words the PTR is roughly the same as the pupil class ratio. This is because teachers are not subject specialists, but teach all subjects. At lower secondary in 1998 there was a ratio 1.45 teachers for every class, meaning that the student class ratio was 45 percent greater than the PTR. The equivalent ratio for upper secondary schooling was 1.6. In other words, post- primary PTRs are curriculum driven, as well as being influenced by total enrolments in any given school. The trends in pupil teacher ratios (PTRs) and average class size are shown in Table F.22. Table F. 22: Pupil teacher ratios and class sizes 1994 1998 Primary PTR 34.9 30.4 Class size 33.5 31.6 Lower Secondary PTR 25.9 28.7 Class size 40.4 41.6 Upper Secondary PTR 23.3 30.5 Class size 45.2 48.9 Source: MOET 152 Annex F - Education Sector 89. Overall the picture is one of increasing PTRs at post primary levels and relatively static ratios at the primary level. This tendency for class sizes and PTRs to increase in secondary education is likely to be the result of several factors. First, rapid enrolment growth may be outstripping teacher availability in those sub-sectors. Second, class sizes are increasing, so that subject teachers teach more students. The effect of rising PTRs is to reduce expenditure per student at any given level of real average salary. 90. One way of looking at the significance of class size and the PTR is to consider the potential effects of raising the PTR by one pupil. The calculation for a primary and lower secondary example is shown in Table F.23, using 1998 expenditure data. Table F.23 :Raising pupil teacher ratios Average Salary PTR Per pupil Enrolments Total VND, ,OO cost of a ,O0O Salary teacher Expenditure VND, ,OO Pimary 7,533 30 256 10,250 2,624,055 100.0% 40 192 10,250 1,968,041 75.0% Lower Secondary 7,680 29 265 5,565 1,473,766 100.0% 3 0 256 5,565 1,424,640 96.7% 3 1 248 5,565 1,378,684 93.5% 91. If in 1998 there had been two more pupils per lower secondary teacher, salary expenditures would have been over 6 percent lower. Of course, in reality, as teachers are not allocated on a per pupil basis, such savings might not have occurred, but as we suggest in this chapter, a move towards per pupil budgeting could permnit closer control of efficiency factors. Similarly, a rise in the primary PTR from 30 to 40 would have permitted a theoretical reduction in the salary bill of 25 percent. 92. In reality such savings would not have accrued, because at present, the budget is not linked actively to the PTR via an allocation criterion, just as it is not linked to average salary considerations. Nevertheless, the choice of raising PTRs to release funds for other uses remains at all levels. 93. The critical factor for post primary teacher utilisation is curriculum policy as it affects teaching loads, as well as class size for each subject. Secondary school teachers are trained to teach one subject. MOET strategy is to train multi-disciplinary teachers, and in due time the capacity of teachers to teach more than one subject will cause the PTR to rise. In addition, a closer analysis of the relation of the required curriculum and teacher demand, particularly in the context of permitting secondary school principals more discretion in the range of subjects offered, could assist fuirther in a clo'ser alignment between class size, teacher working hours and the average PTR. This analysis would need to take the examination system into account. Higher, Technical and Vocational Education 94. The higher education system presents a good case study of how more efficient utilisation of staff has 'paid for' enrolment growth. Figure F.2 shows how both teaching hours 153 Vietnam Public Expenditure Review 2000 - Annexes and student staff ratios have increased in response to the large rise in student numbers (partly caused by the need for institutions to get fee income). Figure F.2: Teaching staff utilisation in higher education 420 30.0 .~~~~~~ 410 ~~~~~~~~~~~25.0 ~400~ ,,3C .. 4 - -- T== 20 .0 0 370 --15.0 la J~~~ 2 _ _ _ _ _ _ _ _ _ _ _ _ ~~~~~~~~~10.0 rj.j 340 - Average standardised teaching hours per year P 330 - ....... Staff student ratio (FTEs) 5.0 320 - - 0.0 1996 1997 1998 95. At the same time, as is seen in more detail in Appendix F.2, there is a variation in these utilisation rates between types of course and according to size of institution (as measured by enrolment number). 96. In the vocational training institutions similar issues prevail. Reported weekly periods for theory teachers were low. Average student teacher ratios were about 16, about half what they are in upper secondary schools. The Quality of Teaching and the Motivation of Teachers 97. We have noted that the low level of public sector salaries in general and teacher salaries in particular has long been recognised as a problem affecting the efficiency and effectiveness of teachers. One consequence is that a relatively short academic year, a low number of contact hours and topping up earnings through private tuition, have become normal practice. Many of these practices may have an adverse affect on the quality of education provided through the state system. 98. There are two sets of policy required to approach this problem. First, salary policy, which has already been partially discussed. Teachers' salaries would need to be set at a level that enabled teachers to afford not to seek other work. In addition, appropriate measures would be required to give incentives to teachers in difficult conditions. Second, effective ways of developing teacher competence need to be developed and implemented. This is a public expenditure matter not only because it has cost implications. 99. Most of the literature on this subject in Vietnam is concerned with lengthening teaching time and pupil teacher contact hours. It is certainly the case that the school day is 154 Annex F - Education Sector short in Vietnam compared with other countries.45 While there is international evidence that learning outcomes are related to length of teaching time, it does not follow that lengthening the school day in Vietnam would improve learning outcomes sufficiently in relation to the associated costs. Clearly, however, many parents believe that it is worth spending more. Moreover, government clearly recognizes the trade-off in that implicit policy is to encourage parents to pay. 100. Many countries have extended their basic education cycles beyond the limits of fiscal affordability. Policy makers in Vietnam would need to consider carefully the costs of improving teacher productivity in relation to the benefits gained. As noted, there are two basic components of an improvement strategy (apart from those associated with financial productivity such as class size and the PTR). First, the quality of teaching time may be improved within the existing system; and second, teaching time may be lengthened. 101. The first strategy requires resources for school change and other teacher development activities, and would probably yield relatively high returns, depending on the technology chosen. The second strategy would be more costly. Teachers would need to be paid more; and physical infrastructure would need to be developed to eliminate shift systems in many areas. In practice, a mixture of the two strategies would probably be the most appropriate, for many teachers would benefit from programs to improve their competence, and some marginal changes to teaching time are probably possible. Moreover, investments in teacher housing in remote areas would be part of an incentive package. Infrastructure 102. In spite of investment levels, there are clear signs that infrastructure is failing to keep pace with growing participation in the system. The Living Standards Survey (table 2.11.2) shows high percentages of classrooms in poor condition or without basic facilities such as blackboards. There is considerable variation between regions and geographic regions. Basic schools (primary + lower secondary) seem to be a particular problem, and over all primary schools one third of classrooms are reported as inadequate, North Central Region being the worst affected with more than half of primary classrooms in poor condition. 103. From technical and cost points of view, construction becomes more complex and more expensive (in total and across per pupil averages) the higher the education level. It is natural that state finance will tend to gravitate towards higher level construction. Capital expenditures in the higher education system accounted for over 20 percent of the entire sectoral capital budget (see Appendix F.2). One particular issue in higher education is the demand for capital funds from smaller institutions. The question is, what are the best mechanisms to gain the most benefit from marginal revenues for construction? Given the magnitude of the capital requirement, and sustainability issues arising from over-dependence on non-domestic capital revenues, some thought may be given to how better to utilize capital funds. 104. One approach is to improve the leverage of construction funds. Where parents contribute to primary and lower secondary construction, for example, does the public budget system (a) give the best incentives to contribute to the levy; and (b) maximize total funds thus 45 Living Standards Survey table 2.9.2 155 Vietnam Public Expenditure Review 2000 - Annexes raised? Schools and communities that can borrow can use levy funds to service interest and principal of the loans, but it is unlikely that such practices are common. One option might be to consider this principle in relation to public funds: capital funds could be deployed as loan capital, serviced and repaid against an agreed schedule. Indeed, subject to appropriate regulatory changes, even foreign grants and loans could be deployed to create revolving funds. Appropriate national (e.g. are loans deficit financed) and local accounting methods would need to be developed. However the system would operate through the capital funds being placed in a revolving fund, which with initial and some subsequent funding from the Budget would then lend funds to schools and communities and receive the repayments for on- lending to other schools and communities. Higher Education 105. The cost structure of Higher Education is rather different from school based education and is separately analyzed in Appendix F.2. The analysis reflects the shift towards a greater part of finance for the service being based on charges to students and less subsidy from public finance. 106. The analysis suggests that there is a need to update the norms that are used for budgetary purposes (the current norms are more than 20 years old) and that updated norms should be applied more consistently. 107. There also appears to be considerable further scope for rationalising and increasing student/teacher ratios and also for rationalising the institutional infrastructure in the light of what appears to be major cost differentials between large and small institutions. Planning and Management 108. It follows from the above analysis that there are considerable planning and management challenges to be addressed in achieving the triple goals of an efficient, effective and equitable distribution of public expenditures on education. 109. Management roles within the sector have been undergoing change during the 1990's caused, particularly, by moves towards greater decentralisation. This process is ongoing and there is, perhaps, a need to consolidate around a clearer understanding of appropriate roles at each level in a decentralised system. 110. Institutional arrangements for managing education in Vietnam revolve around three sets of institutions: central, regional and provincial governments, and education institutions, all accountable in different ways to the National Assembly. Resource control in the sector has been increasingly deconcentrated. Broadly, districts and communes manage pre-school, primary and lower secondary education, provinces manage upper secondary schools and central ministries still manage higher and vocational education. There are, however, variations on this theme from province to province and even at Institution level. lll. The term 'deconcentrated' is used because control is devolved within what remains a unified national budgetary system. Furthermore, the majority of regions are reliant on 156 Annex F - Education Sector resource transfers from the centre to supplement local revenue. This tends to make those regions highly dependent on central assessments of spending need. As a result the system operates with some of the characteristics of a decentralized system, but is really transitional in nature. 112. At the national level, four ministries are key: the MOF, MPI, MOET and the Government Committee for Organisation and Personnel (GCOP). This type of structure is a common one in developing countries, though the role of manpower ministries has somewhat diminished. There are a number of particular issues that appear to arise in the education sector: 113. There is a need to clarify the role of the Ministry of Education and Training. With decentralisation the role of the Ministry becomes much more one of policy setting, standard setting/service quality assurance, and ensuring accountability for perfornance rather than rigid and detailed control. However, this role requires different techniques and skills which the Ministry needs support to acquire. It also needs to be given clearer levers to influence than appears to be the case at present. 114. There needs to be a clear accountability between each level of management as to both financial disposition and service performance/quality. Earlier management style based on control needs to be broadened out. There are a number of modern public sector management techniques for achieving this which managers at all levels might usefully be exposed to. Among possible innovations to the system could be the introduction of an element of performance targeting as part of the budget system, both from the centre to the regions, and within regions. 115. The planning process is also under challenge arising from decentralisation. Earlier systems based on strong central direction are under stress-there is an increasing gap between centrally mandated targets and norms, bottom up plans based on unrealistic resource assumptions and the reality of budget constraints. The result is that the planning process is increasingly failing to guide what happens on the ground and to resolve issues of priority. As decentralisation gathers pace, there is an increasing need to re-integrate the planning and budgeting processes on a basis that takes account of decentralisation. Recommendations - Efficiency of Resource use * There is scope for increasing pupil/teacher ratios at all levels. This seems important to the govemrnent's ability to sustain the cost of its education policies. Methods of achieving higher ratios at post primary levels include increasing the number of subjects a teacher can teach, and the number of periods taught. * The opportunity should be taken when increasing teacher salaries to relate increases to improved competence and to better working practices such as increases in school day/year. Recent significant increases in salary have not been accompanied by such gains. * It is suggested that minimum expenditure on non-salary expenditure be set within the allocations per student. Local managers should be given discretion to increase non-salary 157 Vietnam Public Expenditure Review 2000 - Annexes expenditures beyond these amounts within their overall resource ceilings. Nonns should be adjusted to give room for managers to reduce dependance on contract staff. * Consideration should be given to leveraging funds raised locally, by providing access to counterpart loans from the centre through a form of 'revolving fund' to be trialled on a pilot basis. This would provide both incentives and greater flexibility to undertake investment. * The role of MOET should be re-appraised in the context of increasing decentralisation. Its functions should be policy development, quality monitoring and accountability of service providers such as Provinces and Institutions for performance. * Within this decentralised system steps are necessary to ensure the close integration of planning, budgeting and perfornance monitoring systems supported by an adequate Management Information System. This integration must be both upwards and downwards through the management layers to be effective. * Useful 'glue' to such integration would be the introduction of 'performance agreements' between each management level. The role of these agreements would be to ensure that plans, budget allocations and required performance targets are linked to each other and are fully transparent. * There is further scope for rationalising resources absorbed by some areas of higher education where norms are out of date and not soundly based. Pupil teacher ratios in this sub-sector continues to be low and education is often provided in small and inefficient education establishments. f. Outputs and achievements 116. This chapter has shown that very considerable progress has been made in the education sector in respect of increasing enrolments and improving the efficiency and equity of the distribution of resources. The effectiveness of public expenditure on education is determined by the learning outputs and outcomes achieved. This is about much more than just reaching larger numbers of students. It is also concerned with the quality of service provided and its relevance to the needs of both the students and the country. 117. Many of the recommendations made concerning resource distribution and efficiency have implications for the quality of outputs and achievements. In particular, better utilisation and motivation of teachers and ensuring better levels of non-salary expenditures would have major benefits for overall service quality. Enrolments 118. In an ideal system every pupil graduates successfully to the next class. In reality some pupils leave the system early, for a variety of reasons, and others repeat classes because they 158 Annex F - Education Sector have not mastered sufficient content to proceed, again. How these problems are tackled depend on their underlying reasons. Repetition 119. The LSMS collected data on the number of enrolled children who repeated at least once. The data show therefore not rates of repetition, but ratios of repeaters, and that ratio indicates that the poorer the children the more they repeat. This is to be expected: poorer children are also likely to have lower attendance rates, which are to be distinguished from enrolment ratios. Although the children may be officially reported as enrolled, they are likely to be frequently absent from school, and therefore to fall behind in learning. They can also afford less extra schooling. 120. Table F.24 shows the percentage of enrolments that repeated in 1994 and 1998. From the data it appears that there has been some reduction in repetition at the primary level, though not at the post-primary levels. Table F.24: Repetition ratios 1994 1998 Primary 5.70% 3.6% Lower Secondary 2.30% 2.5% Upper Secondary 1.30% 1.3% Source: MOET 121. From 1998 MOET data it appears that the rates of repetition are highest in the entry classes (7.6 percent in primary), which is internationally a common experience. Where education budgets are not based on per pupil expenditure, there may be very little additional marginal cost associated with repetition. This is because repetition rates are not high enough to involve additional classes or teachers. When budgets are allocated on the basis of teacher numbers, or indeed of class numbers, the per pupil marginal cost to the state may be zero, or close to zero if the relatively low expenditures on teaching and learning inputs for the marginal pupil are taken into account. 122. Repetition becomes more significant in terms of budget when resources are allocated against pupil numbers, such as is proposed in this chapter. Schools with high repetition rates become more expensive than they should be. One of the possible advantages of per pupil budgeting is that there is that it is much more possible to build-in incentive to reduce repetition rates. One way of doing this was illustrated earlier. Retention 123. MOET does not gather data on early leaving by grade. The data are expressed in terms of the percentage of total enrolments of the previous year who have not returned in the current year. The data cannot show therefore what proportion of children drop out of school from one grade to the next, and it is not therefore possible to determine the composition of gross progression rates (grade promotion + repetition - early leavers). The transition rates between 159 Vietnam Public Expenditure Review 2000 - Annexes the primary, lower secondary and upper secondary levels has been rising, with a consequent rapid growth in post primary numbers. Table F.25 shows the trends. Table F.25: Transition and completion rates 1994 1998 Survival Rates Primary 68.00% 68.1% Lower Secondary 68.00% Upper Secondary 82.00% Transition Rates Primary to Lower Secondary 80% 98% Lower Secondary to Upper Secondary 62% 67% Source. MOET Quality of Education 124. There are several ways of considering the quality the education received by the increasing number of students enrolled in the system. On the one hand, the examinations applied at different levels of the system are a measure of how students in the system are meeting internally defined criteria of success. The trends in this respect are shown in Table F.26. Table F.26: Examination pass rate (percent) 1994 1998 Primary 95 93 Lower Secondary 90 87 Upper Secondary 89 93 Source. MOET 125. However, examination pass rates are generally recognised as a poor measure of attainment. There are many other influences on changes in pass rates. Some may be connected to environmental factors for the students themselves. Others may reflect a wide range of internally generated factors such as the quality of the examination setting and marking process, how far examinations are norm or criterion referenced, and a need to use the examination system to ration access. Therefore, such results may be a poor indicator of absolute and changing levels of 'value added' arising from service quality. 126. In recent years, a number of sample studies of student attainment have been carried out. The results of one recent test are shown in Table F.27. 160 Annex F - Education Sector Table F. 27: Recent attainment test in mathematics and language At Grade 3 (%l) At Grade 5(%) Male Female Male Female Mathematics Hanoi 70 74 72 78 Yien Bai 31 39 34 37 ThanhHoa 23 32 17 16 Quang Nam 18 20 22 17 Vinh Long 24 25 13 18 Vietnamese Language Hanoi 67 79 75 83 Yien Bai 47 50 50 58 Thanh Hoa 20 30 27 31 Quang Nam 25 27 16 24 Vinh Long 31 36 21 27 Source : Monitoring Vietnamese and Mathematics 1998 - Phase I Report 127, These results suggest that the level of attainment outside Ha Noi is both poor and very different in different schools and parts of the country. Unfortunately, there seems to have been no systematic follow up to any of these measures and, therefore, no time series to demonstrate whether levels attainment are improving or falling. It is clear, however, that standards have a long way to improve, a conclusion that is supported by observation of both teaching methods and the quality of infrastructure and teaching materials in schools. 128. The analysis of how resources are allocated to education thus far has concentrated on the type and composition of inputs. Teachers are required to teach, but there must be a return to expenditures on teachers in terms of improving pupil achievement, however defined; books are required, but the quality and utilisation of books must contribute to better achievement; and so on. In other words, education resource allocation methods should include an explicit set of measures of pupil achievement, and of system performance in general. 129. There is an additional element to be considered in relation to measurement of achievement band how that measurement contributes to better performance. Not only is large scale testing expensive: it takes time to process, and there is a tendency for the results to be analysed and used more at the higher levels of the educational bureaucracy, and less at the school level. Large scale test systems rely on reliable and valid test construction, which is also expensive. In other words, although there is value in undertaking nationally comparable sample tests, cost and complexity can make them a dubious instrument for performance budgeting, in relation to the impact they may have at the local level on improving pupil performance. 130. One of the potential impacts of pupil testing, particularly at the lower education levels, is that when feedback to schools, parents and communities is rapid, and when that feedback includes information on relative performance of classes within schools as well as relative performance between directly comparable schools (e.g. in the same commune, province, etc), that feedback can be more easily related to the causes of differential performance, and schools, communities, and authorities can take early action to address those causes. There is less need for tests to be reliable and valid (though it is preferable for them to be as robust 161 Vietnam Public Expenditure Review 2000 - Annexes possible), for their potential lies in as much in providing indicators of performance as a means for those most equipped to act on the results to do so in a timely fashion. 131. For performance measurement purposes, the main issue is what changes take place over time, rather than relative performance at a given time, though that is also important. Where there is, for example, intense community interest in primary school diagnostic and achievement testing, it is more likely that schools will have more incentive to improve. Proposals were made earlier for the introduction of a system of performance agreements to focus the attention of managers at all levels on achievement and to enhance accountability. 132. Appendix F.2 comments briefly on qualitative indicators in higher education, where there is low expenditure on research, and where, as at other levels, pass rates in examinations are so high as to render them poor indicators of teaching quality. Recommendations - Improvement of Service Quality * The amount of minimum funding per student recommended above should be based on what is required to achieve an adequate standard of service. * Improvements in teacher skills and motivation are vital at all levels. Teacher qualifications in areas such as Vocational need to be substantially improved. * Current levels of non-salary expenditures do not appear low by international standards (see also section C). However, this is misleading due to the analytical treatment of some items of expenditure. In practice there is evidence at the local level of very low non-salary expenditures in many provinces. In assessing the minimum funding per student, regard is necessary to adequate levels of non-salary expenditure. * Student attainment ( external objective tests of pupil skills in critical areas ), should be tested on a sample basis. These tests should be carried out frequently in order that a time series may be built up of trends in service quality. This objective test of impact is vital to effective policy development and planning. * The system of performance agreements recommended earlier should focus on both output (students served) and attainment (level and usefulness of skills achieved). g. Policy implications of the analysis 133. A critical issue that brings together the forgoing analysis is the affordability of current and emerging Government policies. The analysis undertaken in 1996 at the time of the Education Finance Study concluded that at the relatively high growth rates then being experienced, policies were likely to be broadly affordable. Since then however, Vietnam has experienced reduced rates of growth and the question needs to be re-addressed. To assist with this question, a simple model was developed. 134. The output of the model relates unit costs to actual projected enrolments, by showing 'affordable' enrolments defined as the number of pupils that could be afforded given standard 162 Annex F - Education Sector unit costs and the budget constraint: this is expressed in terms of recurrent unit costs, which is a simplification. In reality, capital costs do set a constraint on enrolments, but capital requirements vary. For school education the main recurrent issues are salaries and non-salary recurrent costs. Reductions in capital expenditure can finance increases in recurrent expenditure, and vice versa. 135. The model suggests that in addition to an increased share of national resources, other measures are required. In particular, pupil teacher ratios need to rise in order to permit higher enrolments at a given salary and non-salary cost, and in order to permit constant real rises in teachers' salaries. Teachers' salaries, like the PTR, can become more of a policy than they are at present. 136. The data do not show the age range of enrolments. The GER supports full enrolment, but not all children of school age are in school. Clearly, although the primary GER seems to have been declining towards 100 percent, maintenance of a GER over 100 percent (i.e. financing more children than of school age), causes some fiscal stress. A policy is suggested of linking the funding formula to attracting children of entry age to primary class 1, in order to over time increase the NER without increasing the GER. 137. Full enrolment at lower secondary is more of a problem, and much will depend on the macro variables, as good quality primary education is likely to remain a priority. In this respect, an aspect of some other countries' experience which is of interest is a steady erosion of primary expenditures per pupil in favour of lower secondary pupils. 138. Demand for upper secondary places, like demand for lower secondary places, will exceed supply. Some private provision will satisfy that demand, though the ability for households to pay is likely to be a constraining factor. Again, the student teacher ratio will need to be raised, and, as at lower secondary, one factor is the demands of the curriculum for specialist teachers, and the skill mix of teachers. Recommendations - Development of Education Policy 139. Tests of the affordability of current policies suggest that the following are the key parameters that need to be balanced: * macro assumptions relating to economic growth and public expenditures * salary levels * pupil teacher ratios * non-salary recurrent shares * capital expenditure levels 140. Tests have been performed with the aid of a computer model. These tests suggest that an appropriate mix of these variables that would align service expenditure on an affordable basis would be along the following lines: 163 Vietnam Public Expenditure Review 2000 - Annexes * the share of public expenditure in GDP rising slightly with improved revenue collection * the share of education in GDP rising to 4 percent over the next 5 years * the share of education expenditure in total government discretionary expenditures rising to 20 percent over the next 10 years * capital expenditures on education retaining its current share of aggregate Education expenditures. * salaries as a ratio to GDP per capita reflecting modest real rises in salary levels, at least maintaining the enhanced 2000 levels in real terms. The model could be used to assess the impact of further real increases in salary levels. * Based on better data, a rise in non-salary recurrent expenditure shares reflecting improved provision of those inputs with a direct impact on teaching and learning. * significant rises in pupil teacher ratios to release funds through efficiency measures to supplement those made available through real budgetary growth 141. In the light of the Sector's current policy and strategies it is likely that one of the responses to this situation of resource pressure would be to ration entry into Upper Secondary and Higher Education and to divert access for those students who fail entry criteria to semi- public (fee paying) provision. 142. It is true that private/fee paying provision is a possible part of a package of responses to the problem. But there are dangers in a situation in which semi-public provision is seen at present as of low quality. The overall efficiency and quality of the service is diluted and a two-tier system reinforced. 143. A more balanced strategy might include the following elements: * Broadening public provision in the higher grades of school education as far as possible by allowing pupil/ teacher ratios to rise to levels still consistent with maintaining adequate standards. * To shift the focus of fee paying provision to offering standards of service that it is attractive to parents to pay for. * An extended system of fee waiver for the poor, as discussed earlier. * The foregoing policy targets would be founded on an improved norm-based system which permits consistency between policy objectives and actual budgetary outcomes, such as that described in this chapter. 164 Annex F- Education Sector Appendix F.1: Examples of how norms and the cascade of resources to local level works in practice 1. There are very many norms used in guiding the resource allocation process. The main ones are listed in Appendix Table F. l.1 A distinction is made based on their apparent role between whether they are: * Obligatory/mandatory in their nature. * Largely advisory. * Directly inform the budget negotiations that take place. 2. As noted in the main report, in a situation of resource constraint, some of these norms become mutually exclusive and contradictory. 3. Resources reach the operational level in a school through a cascade process from the centre to Provinces and to District and schools (for those schools managed at District level). Different techniques are used at each level and in different ways. 4. The way in which norms work, the negotiation process that goes around them, and the application of the different cascade techniques ends up having a very significant impact on the amount of resources per pupil that reaches school level. An illustration of this using data from a sample of schools in 4 Provinces is shown in Appendix Table F.1.2. Although these examples relate to Primary, most of the different techniques and issues surrounding norms have been observed to apply to Secondary as well except that allocations in most Provinces are made direct to secondary school level. 5. It will be noted that the end result is often quite different than implied by the norms and often quite regressive. For example, much higher levels of resources per pupil reach the school level in Hanoi compared to Lam Dong although the original norms would imply the reverse. 165 Vietnam Public Expenditure Review 2000 - Annexes Appendix Table F.1.1: Role of different norms and criteria that feature in the budgeting process NORM Obligatory / Advisory To inform Mandatory negotiation Key Guideline Financial Allocation Norms / Criteria Expenditure per head of population X Expenditure per pupil X Proportion of salary / non salary X expenditure Ability to pay through surplus revenue X Local ability to raise fees and charges X Physical Norms Standard of building and asset X maintenance Availability of standard equipment & X materials Building capacity and location X Availability of basic facilities (electricity, X water etc) Land availability for construction X Text books per pupil X Staffing Norms Student / teacher ratio X Class size x Teaching contact hours X Administrative Staff levels X Notes: 1) A norm would be regarded as obligatory / mandatory if it automatically deter final allocation 2) A "key" advisory norms is one that seems to be used by the center in judging of Provinces 166 Annex F - Education Sector Appendix Table F.1.2: 1998 Current Spending on Primary Education, thousand VND per pupil National Province District School level norm Private contributions Total Salary Non-salary Total Salary Non-salary 286 Yen Bai 330 334 299 35 311 280 31 28 440 HaNoi 220 294 220 73 358 232 126 161 Soc Trang 242 229 206 23 222 194 28 4 330 Lam Dong 440 280 244 36 283 257 26 75 Total Non- salary Yen Bai 339 59 Ha Noi 519 287 Soc Trang 226 32 Lam Dong 358 101 Notes: All data other than national norms is for actual spending per pupil (total spending/number of pupils). 'Estimated total spending, public and private' is district average spending per Pupil plus estimated parental contributions for the school visited in that district. All private contributions are assumed to be spent on non-salary items. All four schools operate on a two-shift system except Soc Trang on three shifts. 167 Vietnam Public Expenditure Review 2000 - Annexes Appendix F.2: Higher education 1. The MOET has undertaken annual surveys of universities and colleges over the last three years. The following discussion is based on the survey published in 1999, which incorporates data in the years 1996-98. 2. The basic information derived from the survey on revenues and expenditures in the sub-sector, is summarised in Appendix Figure F.2.1. Full time equivalent (FTE) students46 have risen at an annual rate in the three years of about 16 percent, a total increase of about 55 percent between 1996 and 1998. The bulk of the increase is accounted for by fee paying students. 3. The Figure shows that * Total (nominal) revenues to higher education have risen at a slower rate that the rate of growth of enrolments * The growth of revenues from government has declined significantly * Revenues from fees have been the compensation for the declining share of government revenues. Fees in 1998 accounted for about 36 percent of total revenues, compared with 24 percent three years previously. Figure F.2.1: Evolution of Higher Education Revenues 2,500 - ---Total revenues 000 Revenues from government 800,000 -_Revenues fromfees ______ FTE students 700X000 2,000 600,000 1,500 - - 500,000 c Z - E 400,000I > 1,000 300,000 500 - ______ 200,000 100,000 1996 1997 1998 4. It is not easy to relate revenues reported by higher education institutions in the survey to the reported expenditures in the government budget reports, as table F. 1 shows. 46 Because not all students are full time, part-time student numbers need to be adjusted to make total numbers comparable. 168 Annex F - Education Sector Part of the reason for this is no doubt because while higher education institutions benefit from some vocationai and technical expenditures, these are not easily separated. However, a comparison of reported allocations to universities, colleges and higher education in 1997 and 1998 shows that universities report lower revenues from government than are reported in the government accounts. 5. Figure F.2.2 analyses revenues to higher education in terms of FTE students in 1996 prices: this permits a meaningful comparison over the three years. Average revenues per FTE student in total have declined. Government allocations per student also declined. Figure F.2.2: Revenues per student in higher education 4.000 3.500 3.000 - 2.500 2.000 1.500 - 1.000 j Total revenues per FTE student in 1996 prices | - - - Revenues per FTE student from government, 1996 prices 1996 1997 1998 6. The government at present does not take into account total numbers of students in universities and colleges when budgetary allocations are made. Instead, it draws up a schedule of per pupil norms. The amount per pupil varies, although it does not appear that the reason for the variation relates to the cost of offering each course. There may, however, be some relation, if actual recurrent expenditures per FTE student by faculty are compared with the norms: however, because the higher education survey data are collected in different categories from the norm categories, and because the survey does not specify the number of government financed students on which the actual allocations are based, it is not possible to ascertain directly how far the government norm system itself is reflected in the actual expenditures. 169 Vietnam Public Expenditure Review 2000 - Annexes Figure F.2.3 Expenditures per student in higher education 3,500 30.0 ° 3000 250 z 2,500 X. 20 ,2,000 - 00 &1,500 . 10.0 u1,000 ..Total expenditure per FTE student 1996 prices Recurrent expenditure per FTE student 1996 prices 5,0 - _ _ _ Staff student ratio (FTEs) I l I o0.0 1996 1997 1998 7. Figure F.2.3 shows the evolution of expenditures per FTE student. The staff- student ratio (SSR), adjusted for full time equivalence, has risen from just under 18 students per teaching staff to just under 25, an increase of about 39 percent but at the same time, teachers are teaching more hours. Thus average expenditure per student has declined, and this has been partially paid for by increased SSRs, and increased hours of instruction per teacher. The increased utilisation of staff is illustrated in Figure F.2.4. Figure F.2.4: Teaching staff utilisation in higher education 420 30.0 ,3 410 Y 400 n g 25.0 390~~~~~~~~~~~~~~~~~~ 20.0 ' c 370 - 15.0 r = 360 / 10.0 t 350 - 340 - Average standardised teaching hours per year ~~ 330 ~~Staff student ratio (FME) 320 i_. l0.0 199 1997 1998 8. Figure F.2.5 compares staff student ratios between subject areas and shows how they have evolved in response to the changing pattern of revenues. 170 Annex F - Education Sector Figure F.2.5: Staff student ratios 40.0 35.0 -_ - _ _ __ _ _ _ 30.0 -_ _ __ _ _ 25.0 _ _ _ __ _ _ _ _ _ _ _ _ 20.0 -_ _ _ _ _ _ _ 15.0 _ _ _ _ _ _ 10 .0 5.0 0.0 -d t~~~~~~~~~I 9. The data suggest a growing efficiency in the use of staff, which has permitted as growth in student numbers. It is not clear, however, how far efficiency improvements are a deliberate result of government policy, given the way resources are allocated. For example, the official staffing norms have not changed in 20 years. 10. Moreover, when the data are disaggregated, further variations are apparent. Figure F.2.6 comnpares average expenditures per FTE student (recurrent and capital). Figure F.2.6: Average expenditures per FTE student by size of institution 10,000 9,000 8,000 _ __ 7,000 6,000 5,000 4,000 3,000 2,000 <1000 1,000-2,000 2,000-4,000 4,000-10,000 >10,000 I1. The relative expenditures per student in the Figure reflect the fact that smaller institutions have lower utilisation of staff, in terms of SSRs and of teaching hours. Smaller institutions require considerable more capital expenditure per student than larger institutions. Other indicators in the table suggest that staff in smaller institutions are on 171 Vietnam Public Expenditure Review 2000 - Annexes average less well qualified and less experienced. It seems that the main influences on efficiency improvements are to be found in the larger institutions. 12. Larger institutions are less dependent on income from the government, or at least, they receive less in relation to total income, as is shown in Figure F.2.7. Figure F.2.7: Composition of institutions' revenues, 1998 80.00/o Sh f_ [ vhaieofgovemrnment revenues in total 70.0%/o Share of fees in total 60.0% __ 50.0% , __ .,_ X 40.00/o __ 30.00/o __ _ _ 20.00/o - 0.0%/ <1000 1,000-2,000 2,000-4,000 4,000-10,000 >10,000 Salaries 13. It is widely accepted that university teachers' salaries in terms of official rates are low in relation to the cost of living. The government grant is distributed in principle partly on the basis of the number of staff and their location on the pay scale. Salaries and related personnel expenditures as a share of recurrent expenditure are fairly constant across all sizes of universities and colleges, with the exception of institutions with less than 1,000 students, but larger universities provide larger salary supplements out of fees. Although they do this, salary subsidies as a percentage of fee income is lower, the larger the institution. This would follow from the fact that larger institutions collect a greater share of their total income in the form of fees. 14. The survey also permits disaggregation across regions. As is to be expected, most students are in the Red River Delta region (43 percent of total), followed by Southeast region with 30 percent. The percentage of staff with masters' and doctorate degrees varies, and is lowest in Northwest and Central Highlands. 15. There is some variation in average expenditures between regions, which will be accounted for largely by the relative size of institutions, the related staff student ratio, which is much lower in Central Highlands and Northeast regions, and the student-course level mix (postgraduate, teacher training, etc). There is also considerable variation in the share of fee income in total revenues. 172 I Annex F- Education Sector 16. One aspect of issues relating to the economies of scale of higher education institutions is that provision of higher education away from major centres will tend to be less efficient when measured in terms of resource utilisation. The question then becomes one of defining the benefits of such institutions, particularly in terms of their location. This in turn suggests a need for policies to maximise those benefits: higher education has been seen in several countries as one way to promote regional growth. Some indicators of academic quality 17. The benefits of higher education institutions can be considered in different ways: for example, as noted they can contribute to regional growth, and they also provide training for skills required in the national economy. Two common ways of identifying benefits from higher education institutions on which a judgement about their funding can be based, are in terms of their research output and in terms of how well they teach. Other issues of a manpower planning nature relate to differential government funding for specific types of course. Research 18. One function of higher education institutions is research. At the same time, in most countries some institutions are more research oriented than others. Research as a share of total expenditures on higher education in Vietnam is under 2 percent, and the share is greater, the larger the institution. Academic research appears to be financed almost entirely by government: that is, it may be that in the absence of earmarked government funding for research none would be financed out of other revenues. 19. Although research accounts for a low share in total expenditures, in principle it is still important, both for social and economic reasons, and as a way of informing funding decisions on the basis of performance. Degree Passes 20. As these are very nearly 100 percent, there seems to be little real difference in quality between institutions if passes are taken to be a proxy for quality. Drop-out and Retention 21. Drop out and repetition rates are not high. Drop-out is highest in Southeast, but lowest in Northwest. Repetition is highest in Southeast, the next highest being Northwest and Red River Delta. 22. Thus, according to these data, there is relatively little distinction in performance, defined in terms of degree passes and student completion, to be made between regions or between institutions of different sizes. 173 Annex G - Transport Sector ANNEX G - TRANSPORT SECTOR Sector Overview and Performance Sector Overview 1. Vietnam's transport sector comprises all modes of transport-roads, rail, inland waterways, sea and air, covering a narrow, elongated territory of more than 1,700km north to south: Due to Vietnam's geographical features and the concentration of the population and economic activity, the transport network is concentrated and traffic density is highest in and around the centers in the northern and southern delta areas. North-south connections are provided by roads, the railway, coastal shipping and air. Roads are the most important land access to the neighboring countries of China, Lao PDR and Cambodia. 2. Roads. The country has a well developed road network of over 209,000 km, of which 7 percent are national roads, 8 percent provincial roads and the remainder district, ,commune and village roads. Roads are by far the most important transport mode, accounting in 1998 for about 71 percent of the total land freight traffic (48 percent in ton- kim) and 80 percent of passenger traffic (65 percent in passenger-km). In terms of coverage the network compares favorably with its neighbors and the other ASEAN countries. With a road density of 0.3km/sqkm, Vietnam equals that of the more developed ASEAN countries (Thailand, Philippines, Brunei). However, in terms of quality Vietnam's road network among the worst: only 42 percent of the primary road network (national and provincial roads) and about 4 percent of the district and commune roads are paved. Surveys indicate that almost half of the national road network is in poor to very poor condition and one quarter of the bridges have reduced bearing capacity and need to be rehabilitated or replaced. The situation of the provincial, district and commune networks is even worse. Ninety percent of the rural roads are laterite and earth roads in poor condition, and may be subject to seasonal flooding. In addition, more than 600 communes still have no road access to district centers, 12 percent of the rural population has no access to motorized transport and more than 100,000 km of local roads can only be used during the dry season. 3. The vehicle fleet is still relatively small and consists mainly of utility vehicles. There are about 140,000 passenger vehicles (equal to 1.8 vehicles per 1000 people) and 4.3 million motorbikes (equal to 56 motorbikes per 1000 people). The fleet of utility vehicles is old and trucks have on average a low carrying capacity. However this situation is changing, as the trucking fleet is slowly being replaced by modem, higher capacity (and thus lower cost) vehicles. 4. The poor state of the roads imposes a high economic cost - estimated to increase vehicle operating costs by up to 80 percent, depending on the condition of the road-or about US$160 million per year in additional vehicle operating costs alone. Moreover, if 175 Vietnam Public Expenditure Review 2000 - Annexes roads are not maintained regularly they will have to be rehabilitated eventually at a much higher cost, (i.e. US$ 200,000 per km of paved road over a 15 year period compared with maintenance cost of US$ 60,000)47. 5. Thus the costs of poor roads to the economy are enormous. Evidence of this can be found in the increase in transport costs between 1986 and 1991 before the road rehabilitation program had started. During that period the cost of freight transport increased by a factor of 2.7 relative to general prices and, following an increase in road rehabilitation during the period 1991-1996, it slowed somewhat to 1.8. In addition, the additional cost is not uniformly distributed across the country. The areas with relatively high quality road networks (i.e. the two economic zones in the north and the south) have lower vehicle operating costs than those with low quality networks-the provincial and rural areas. Such higher transport costs have the same effect as the imposition of a tax: they tend to raise the costs of good and services with origin or destination in the areas with low quality road networks and thus tend to increase the economic disparities between the growth zones and the rest of the country. 6. Inland Waterways. In the northern and southern deltas inland waterways play an important role as a low cost mode of transport for bulk commodities such as coal, building materials and rice. About one quarter of the country's total land freight is transported by inland waterways. Inland waterways are complementary to road transport, rather than a substitute. In the Mekong Delta they are also a low cost mode of transport for the population in the rural areas which tends to be less well served by roads. There are almost 41,000 km of inland waterways available for informal and small scale commercial transport. Of these only about 10 percent are of sufficient depth to allow industrial type transport by vessels of 50 dwt or more. Similar to the road network maintenance has been inadequate, which has resulted in siltation and loss of water depth. This in turn has hampered the utilization of large vessels and thus increased transport costs by about 25 percent, particularly affecting the rural poor. 7. Vietnam Railways (VR). VR operates a network of 2,832 km from Lao Cai at the Chinese border and Haiphong in the north to HCMC in the south. Infrastructure and rolling stock have deteriorated and are in need of replacement/repair. The main immediate problem is the large number of deteriorated bridges (some of which are being repaired with assistance from Japan). VR's share of total land traffic in 1998 was about 5 percent for freight (17 percent in ton-km) and 1.3 percent for passengers (12.5 percent in passenger-km). The Government is financially responsible for the provision and upkeep of the rail infrastructure while VR is responsible for railway operations. VR pays the Government 10 percent of its revenue as a user charge for the rail infrastructure. 8. Sea Ports. Vietnam has seven national sea ports administered by MOT, 20 municipal and provincial sea ports, administered by localities and a number of ports that are operated by other ministries, including the Navy. The most important ports are Haiphong, Da Nang and Saigon which in 1997 handled more than 47 percent of all port throughput. Shallow water depths, long access channels (Haiphong and HCMC), narrow 47 Vietnam: Moving Forward, Achievements and Challenges in the Transport Sector, World Bank, Report Nr. 18748-VN, April 1999. 176 Annex G - Transport Sector port areas, outdated equipment and poor management all have all contributed to low port productivity, high operating costs and under utilization of the ports' potential capacity. 9. From 1994 to 1999, traffic on all land transport modes rose by an average of 9.9 percent a year for freight (10 percent in ton-km) and 3.2 percent for passengers (10.8 percent in pass.-km). Maritime traffic increased the fastest, at 26 percent a year for freight, followed by inland waterway transport at 20.6 percent a year (7.5 percent for passengers), road transport at 9.8 percent a year (7.2 percent for passengers), and rail transport at 3 percent a year (3.3 percent for passengers). Throughput of the country's seaports rose by an average 14.5 percent a year (1995-1999) reflecting Vietnam's successful re-integration into world trade. 10. An analysis of the period 1995-1998 reveals that the state sector transported around 80 percent of all freight in tenns of ton-km and 25 percent in terms of tons. This means that the state sector dominates long distance transport which is normally capital intensive, while the private sector is focused on short distance (road and inland waterway) transport which requires less investment. Sector Performance 11. Despite recent progress the productivity of the transport sector is still low. * In road transport the average speed per vehicle for freight is between 20 and 25 km/h and for passengers 30 km/h. The average carrying capacity of a truck is 5.7 tons per vehicle, up slightly from 5.2 tons per vehicle in 1991. Trucks with a higher carrying capacity (and thus lower unit costs) are concentrated in the economic growth areas of the north and south. Output per vehicle is also low-trucks produce 1,900 ton/km per day and buses 2,200 passenger seats per month on average. The average length haul for trucks is 53 km which means the average truck carries only about 18 tons in three round trips per day (assuming empty back haul). * Regarding sea-transport, the country's two main port complexes of Haiphong and HCM are inefficient, often congested in storage areas and poorly connected with the land-based transport network. Shallow drafts limit ship size and thus raise shipping costs. The present system of national/regional tariffs and charges, the cross- subsidization of the national shipping line by foreign shipping and the lack of commercial autonomy of port management hampers competition between ports and limits incentives for improved performance. This results in lower port productivity and reduces Vietnam's international competitiveness. A recent survey found that Vietnam's ports (HCMC, Danang, Haiphong) are among the highest cost ports in the Southeast Asian region: to transport a twenty foot container, for example, from Tokyo to one of the three economic zones in Vietnam can cost up to $1,000 more than to transport it to Manila or Bangkok.48 To increase productivity, selective investment- in equipment and upgrading of port facilities is needed. However, 48 World Bank, Country Framework Report on Private Participation in Infrastructure: The Case of Vietnam, 1999. 177 I'ietnam Public Expenditure Review 2000 - Annexes removal of the institutional and regulatory barriers to improving performance mentioned above is even more important. * Vietnam Railways (VR) is roughly comparable to most of the other smaller railways in the region. Although well run in an engineering sense VR's labor and locomotive productivity are among the lowest in the region, as is its average traffic density. Since VR has not yet introduced a commercial cost accounting system which reports total costs (including depreciation) and enables costs to be allocated across different activities and services , a true picture of its financial performance including comparisons with other railways in the region is not available. * In inland waterways the backlog of dredging and the lack of navigation aids to permit night navigation has affected even the most utilized corridors. This reduces the size and speed-and thus productivity-of inland waterway vessels. 12. All the above highlights that the potential for transport cost reductions is considerable, provided management and operations are improved and infrastructure and equipment continue to be modernized. Gender and Transport in Vietnam 13. Transport policy, regulations and public expenditure decisions appear to be gender neutral in Vietnam. For example, transport taxes, fees and tolls are based on the costs of providing a service and on the principle of what the traffic can bear. There are special subsidized tariffs for the transport of certain essential commodities (such as salt and rice) to disadvantaged remote areas Some public transport enterprises (e.g. the Railway) offer special tariffs for certain population groups such as war veterans and pensioners, regardless of gender. 14. As far as transport capital expenditure is concerned, the allocation process itself by the center and local entities is based primarily on economic considerations. However, equity and socio-economic aspects are also considered especially in relation to poor regions and rural and remote areas. Thus, there exist national programs administered by MOT to establish as a minimum year round road access for villages that are either only accessible by footpaths or during the dry season, transport investment programs with a focus on the mountainous and remote regions and programs with a transport element in the context of poverty alleviation (e.g. Program 135 for the 1800 poorest communes). These programs aim to establish minimum levels of access for those communes which cannot afford to finance their own local transport infrastructure. A transport master plan presently being considered by the Government proposes to increase investment in rural roads substantially over the next decade, from about 5 percent of total transport investment in the 1996-2000 national plan to over 12 percent. The donor community is actively supporting this policy with the funding of rural roads projects and technical assistance. Although these programs are not directly targeted in terms of gender, women should benefit at least as much as men from better access to such services as education and health as well as district and regional markets. 178 Annex G - Transport Sector 15. Recurrent expenditure for the maintenance of transport infrastructure is allocated by MOT and the localities in accordance with several criteria, including physical condition. While adequate funds for routine or short-term maintenance of the national and provincial road network are available, funds available for small, medium and large periodic repairs and maintenance have been estimated to meet only 40 percent of requirements. District and lower level roads do not receive funds for regular maintenance from central Government, and very little from provincial governments. This is relevant in the context of gender because road maintenance is an important source of employment for female workers. Thus, an increase in road maintenance funding, discussed this review, would open up new employment possibilities for women, especially in the rural areas. 16. Vietnam does not appear to take gender explicitly into account in the development of its transport sector, nor do its policies discriminate, directly or indirectly, against them. However, policies aimed at fostering access of the poor, of rural areas, of remote and mountainous areas to improved and well maintained transport infrastructure tend benefit to women by virtue of providing employment, better access to services, institutions and economic opportunity. Government Objectives and Priorities 17. The rehabilitation and modernization of the country's dilapidated transport infrastructure has been one of the main economic objectives of the Government's Public Investment Program (PIP) 1996-2000. Priority was assigned to the development of transport infrastructure in the three strategic economic zones i.e. the Hanoi-Haiphong corridor in the north; Da Nang in the center and the HCMC-Vung Tau corridor in the south ( in 1997 these generated about 75 percent of the country's GDP, attracted more than 80 percent of foreign direct investment and generated 75 percent of the country's exports/imports); the creation of links with neighboring countries; the integration of transport systems, including port development to facilitate international trade; and the development of urban infrastructure and public transport services to reduce congestion. For these projects economic efficiency and growth is the rationale and selection criterion. However a number of projects have been justified primarily on the basis of social welfare or equity considerations, including in the highlands and mountainous regions. 18. These objectives were very ambitious and implied a large investment program. It soon became obvious that this program could not be implemented during the 1996-2000 plan period alone and needed to be broken down into more manageable sub-programs which could be implemented over a period of 10-15 years. Consequently, the 1996-2000 priorities have mainly focused on improving the basic transport infrastructure. This meant some provincial and rural projects needed to be deferred. The main emphasis has therefore been on upgrading the trunk roads and other transport infrastructure and their linkages of the growth triangles in the north, center and south, In total, about 2,000 km of national roads and about 10 km of bridges have been rehabilitated or modernized. Thus, such major transport investment projects as the rehabilitation of highways no.5, 10 and 18 in the northern growth triangle, of highway No. 1 (the main trunk road between Hanoi and HCMs) as well as the modernization of ports in HCMC, Da Nang and Haiphong 179 Vietnam Public Expenditure Review 2000 - Annexes reflect the Government's priorities. In the south, the Government has commenced a program to rehabilitate and modernize the main inland waterway corridors between HCMC, Can Tho and Ca Mau, which are the most important mode of transport in the Mekong Delta. 19. Investment in the railway system has so far been small and is focused on keeping the system operating safely, pending a major reassessment of the railway's future and its adaptation to the market economy. A study has identified the areas that need to be addressed and Vietnam Railways has decided to implement its recommendations with technical assistance made available by a bilateral donor. 20. Apart from a few minor interventions, urban transport improvement has received little funding so far, despite the danger of serious gridlock if action is not taken soon. 21. Progress in improving the nation's rural roads network, however, has been slow and is a reflection of the resource constraints faced by the Government (central and local). Nevertheless Government entities, with the active assistance of the local population, have managed to improve about 11,000 km of rural roads. About 50 percent of the total costs were funded by local contributions and some donors are contributed. Unfortunately, the quality of the work has often been poor reflecting funds being spread too thinly. 22. To address this and other issues the Government (assisted by the donor community) is finalizing a comprehensive rural roads strategy whose basic elements are to: (i) rehabilitate rural roads as an integral element of rural development; (ii) ensuring basic all-weather road access to rural communities; (iii) establish sustainable decentralized rural road management, maintenance and finding structures. It is also considering increasing substantially (in the 2001-2005 PIP) investment expenditures for the improvement of the rural road network. Expenditure Levels and Trends in Public Transport Quality of Information. 23. It has been very difficult to gain comprehensive insight into the amount and quality of state spending on transport because: (i) not all state transport expenditure can readily be identified in the budget; (ii) the information produced by MOF, MPI and MOT (for seemingly the same spending items) cannot be reconciled; and (iii) there is little systematic exchange of information between agencies, both horizontally (e.g. MOF, MPI, MOT and other ministries dealing with aspects of transport expenditure) and vertically (i.e. between central and local government entities). 24. This means that MOT as the responsible ministry is hampered in its policy role for the sector. In addition, other ministries directly spend on transport related activities or projects. For example, the expenditure of the Ministry of Agriculture on rural roads, or of the Ministry of Industry for dedicated coal ports is not readily available from the budget data. Further, a comparison of the data for capital expenditure of the central government, for example for roads for the first three years of the 1996-2000 PIP shows 180 Annex G - Transport Sector the following discrepancies: MOF 9 trillion dong; MOT 10.9 trillion dong; MPI 6.8 trillion dong. Similar discrepancies also exist in the statistics of recurrent expenditure recorded by MOF and MOT. For example, MOF data show for 1998 482 billion dong in recurrent expenditure by the central government for roads and MOT data show for the same year 568 billion dong apparently because MOT appears to record actual cash flows per year (i.e. including short term borrowing and other inflows) and MOF does not. Even the data kept by the central ministries on donor assistance are not consistent. For example, MOT data show 1998 ODA funding for capital investment in transport of 3.13 trillion dong (actual), compared to MOF's 2.38 trillion dong. 25. These differences should not occur if all data are recorded in accordance with the principles of the unified budget. This lack of consistency, transparency and comprehensiveness of information is bound to impair the quality of the planning, programming and budgeting in the transport sector. 26. The following analysis reviews (on the basis of the available un-reconciled data) state central and local transport expenditure as recorded by MOF and, because of its key role in transport, expenditure by MOT, followed by an assessment of the implementation status of the five-year 1996-2000 Public Investment Program (PIP) for transport, based on MPI data. State Central and Local Transport Expenditure 27. The development strategy outlined above and the importance the Government attaches to the modernization of the nation's transport infrastructure is reflected in the overall expenditure patterns for the transport sector. There has been a marked shift of public resources in both nominal and real terms towards transport since the early 1990's. Between 1994 and 1998 capital spending on transport grew substantially faster (21 percent p.a.) than total Government capital expenditure (14 percent p.a.) and for current expenditure slightly faster (+13.5 percent compared with 12.3 percent p.a.). Thus, between 1993 and 1998 the share of capital expenditure for transport (central and local entities) increased from 18.8 percent to 29 percent of total State capital expenditure (Table G.1). During the same period capital and recurrent expenditure in transport as a share of total State expenditure increased from 6.5 percent to 10.8 percent and, as a share of GDP, after an initial surge from 1.7 percent in 1993 to 2.1 percent - 2.3 percent thereafter. The strong increase between 1993 and 1994 reflects the increased flow of ODA into Vietnam, and the priority attached by many donors to rehabilitating the transport infrastructure and substantially higher allocations of domestic funding. 28. The split of total transport expenditure between central and local spending units shifted slightly in favor of local spending entities, from a 64/36 split in to 60/40 in 1998. This reflects faster capital and recurrent-expenditure growth of local entities. Between 1994 and 1998 capital expenditure by local government entities grew faster (23 percent p.a.) than that of the central Government (MOT) which grew by 19 percent p.a. 181 Vietnam Public Expenditure Review 2000 - Annexes Table G.1: State Central and Local Transport Expenditure 1993 1994 1997 1998 perce perce perce perce nt nt nt nt 1. Capital Expenditure: Total Transport/State 18.8 27.2 26.3 29.0 Capital Expenditure. 2. Capital and Recurrent Expenditure: 6.5 9.0 9.5 10.8 Transport/as a share of total 3. Local Transport Expenditure: share of total 44 64 59 60 4. Transport Capital and Recurrent 1.7 2.3 2.1 2.2 Expenditure as a share of GDP Source: Ministry of Finance. 29. More than one half of 1997 and 1998 transport capital expenditure at the central level was funded by ODA, while local capital expenditure was funded mainly by the Government's own funds and contributions by the people, although the latter are not included in the figures for transport expenditures. Recently, ODA funding has been provided for large local A-type projects in agriculture and rural development, in which transport infrastructure repair comprises a large or even dominant role. The exact amount of such spending cannot be identified reflecting the fact that transport expenditure by other government departments is often not separately identified. Sub-Sectoral Allocations and Capital Expenditure 30. About 85 percent of all (reported) transport expenditure is for investment purposes (Table G.2). In allocations to the sub-sectors, the Government has consistently adhered to its strategy of giving priority to the rehabilitation and modernization of the country's trunk road network. In 1998, roads again received the bulk of transport funding and increased their share from 77 percent of total transport spending in 1994 to 94 percent, reflecting the acceleration of major road investment programs at the national level. Spending on rail fixed infrastructure (which is the central Government's responsibility) took second place, with about 5 percent of total spending in each of the years of 1994-1998. The other sub sectors-inland waterways, sea and coast, air, received between 2 and 3 percent of total transport expenditure. However, this does not necessarily show the full picture. Various transport SOEs may also receive funds from the banking system (at preferential interest rates) as well as the budget, but the extent of any such borrowing is not known. 31. With the exception of spending for inland waterways it appears that the sub-sector allocation is appropriate in relation to the role these modes play in the transport sector. Inland waterways as the second most important mode of transport still appears under- funded, despite the increases between 1994 and 1998. 182 Annex G - Transport Sector Table G.2: State Transport Expenditure by Transport Mode, Locality and Expenditure Category -Bill. VN dong- Mode Exp. 1994 1997 1998 Category Total Central Local Total Central Local Total Central Local Roads Capital 2,456 1,296 1,164 5,219 2,879 2,340 5,955 3,308 2,647 Current 513 331 182 668 425 244 814 482 333 Total 2,969 1,627 1,346 5,887 3,303 2,584 6,769 3,790 2,980 Inl. Capital 97 93 5 97 73 24 139 111 28 Water Current 36 29 7 36 29 8 85 29 56 Total 133 122 12 133 102 32 224 140 84 Rail Capital 223 213 10 128 126 2 312 310 2 Current 22 20 2 152 148 4 150 150 0 Total 245 223 12 280 274 6 463 460 2 Sea+ Capital 131 120 11 91 40 51 146 123 24 Coast Current 76 75 1 2 0 2 5 3 2 Total 207 195 12 93 40 53 151 126 26 Air Capital 272 272 0 100 99 2 133 133 0 Current 0 0 0 1 0 0 20 0 0 Total 272 272 0 101 99 2 153 133 0 Other Capital n.a. n.a. n.a. 123 100 23 117 66 50 Current 50 20 30 43 23 20 Total 173 120 53 160 89 60 Grand Capital 3,198 2,000 1,198 5,760 3,316 2,444 6,803 4,052 2,751 Total Current 661 464 197 909 622 287 1,098 686 412 Total 3,859 2,464 1,395 6,669 3,938 2,731 7,901 4,738 3,163 Source: Ministry of Finance, Statistical Unit. Spending by Regions 32. As regards the distribution of local transport expenditure between the provinces, about 20 percent of the total was spent in the southern growth area HCMC-Ba Ria Vung Tau and some 10 percent of the total in the northern growth corridor Hanoi- Haiphong (1998) which is less in proportion to the budget revenue raised in these growth areas. The majority of the remaining provinces spent between 1 and to 2 percent each of the total local transport expenditure. At the local level roads received more than 90 percent of all transport expenditure and almost 85 percent of that was capital spending. This is true even in those provinces where waterway transport dominates. 33. Although more detailed analysis is needed, some preliminary conclusions can be drawn from these local spending patterns. First, the economic growth areas spend less on local transport in proportion to their revenue generating power while many of the poorer provinces spend more. This may be an indicator that the revenue transfer mechanisms between Government entities (central and provincial) are functioning in that poorer provinces/districts receive funds to undertake higher spending. Also, the rural areas tend to benefit from transport infrastructure spending by the Ministry of Agriculture in the context of rural development projects. However, at the same time the growth areas benefited most from central Government transport expenditure in the form of spending on 'A' projects. Second, there was a tendency to distribute the funds transferred by the center equitably across the 'deficit' provinces, which raises the question of to what extent 183 Vietnam Public Expenditure Review 2000 - Annexes economic efficiency or growth as opposed to equity criteria, played a role in the re- distribution of these funds. Current Expenditure 34. In analyzing current expenditure it is useful to distinguish between non-revenue earning entities - road, inland waterways, the railway track for which the state is responsible - and revenue earning entities and SOEs - the railway, ports, airports and other transport enterprises. The non-revenue earning entities are funded by the state and the revenue earning entities are required to fund their own current expenditure which consequently is not included in the above figures. This exclusion of SOEs expenditure figures means that a comparison of state current expenditure between the modes would be misleading. Second, unlike other sectors, data on current maintenance of road in Vietnam is consistent with the GFS classification insofar as in Vietnam the expenditure category "big repair" (e.g. the resurfacing of a road) is considered a capital expense. While this practice should continue, for the purpose of comparing Vietnam's total maintenance spending with other international road organizations "big repair" should be recorded separately. Third, statistics of current transport expenditure on national roads compiled by MOF and MOT do not match. The difference amounted to about 86 billion Dong in 1998. The same discrepancies occur in respect of local entities but the extent of this is, however, not known. 35. Bearing in mind the data limitations, the following analysis focuses on current expenditure of non-revenue eaming entities, mainly for roads which comprise about 80 percent of all current expenditure in the transport sector. 36. Current expenditure (adjusted as outlined above) of the MOT for national roads in 1998 is estimated at about 698 billion dong. This includes 568 billion dong for routine maintenance and an estimated 130 billion dong for periodic maintenance (big repair). Provided that the amount estimated for periodic maintenance is reasonably accurate, the expenditure covered about 62-67 percent of the estimated required annual maintenance for national roads.49 37. Because of lack of sufficiently detailed information, however, it is more difficult to estimate total current expenditure by local entities. According to MOF about 333 billion dong were spent by local entities in 1998 on small and medium repair of local roads (i.e. province, district, commune and village level). It has been estimated that the entire allocation was spent on maintenance for the provincial road network. 38. There are about 36,000 km of district and 47,000 km of commune roads. The communes manage the commune and village level road budgets. Sources of funding are the resources available to the communes themselves, obligatory labor days, peoples' contributions and transfers from higher level administrations. ( It has not been possible to ascertain whether all these sources of funding are captured by the unified accounting system of the Government.) Except for a few targeted poverty reduction programs (e.g. 49 According to estimates of the MOT Transport Development Strategy Institute and World Bank estimates. 184 Annex G - Transport Sector the 1715 Poorest Communes program) there appears to be very little central government funding of local level transport infrastructure. 39. The total annual budget for district and commune roads has been estimated at about 750 billion dong per year.50 Between 5-10 percent of this amount is financed by the province, 50-60 percent by the districts and 40-50 percent by the communes. Recent surveys suggest that about 15 percent of the district and commune roads (i.e. 12,500 km) are in maintainable condition but are not regularly maintained, i.e. planned maintenance is not practiced. At present, the annual budget is spent on repair of damaged roads, the rehabilitation of seriously damaged roads and road upgrading; and the roads in good to fair condition are basically left to deteriorate until they need repair. This strategy is as inefficient for the rural road network as it is for national and provincial roads. A more cost efficient strategy (to be piloted by the Rural Roads II Project) would be to give priority to the maintainable network (about 24,000 km after the implementation of donor funded projects) and leave a balance of locally managed funds for repair, rehabilitation and upgrading of the remaining rural roads, which would gradually increase the maintainable network. 40. For inland waterways, according to MOF data current expenditure (mainly for maintenance) has been kept constant at 29 billion dong for the centrally administered waterways and at 7-8 billion dong for locally administered waterways (1994-1997), increasing to 56 billion dong in 1998 (Table G.2). However MOT data show substantially higher current expenditure for the centrally administered waterways - 60 billion dong for 1997 and 1998. Assuming MOT's figures are correct, even this amount would not have been sufficient to cover the current expenditure requirements of the inland waterways under the management of its Inland Waterways Administration (about 6,500 km) which have been estimated at about 190 billion. Improving inland waterways in the northern and southern deltas would be the fastest and most cost effective means to support the poor rural delta regions, because waterways are the main mode of transport in these regions. User Charges for Non-revenue Earning Road Transport Infrastructure. 41. The issue of developing a stable source of funding for the maintenance of the country's road transport infrastructure is under active consideration in Vietnam. Various studies of the Vietnamese road sector have shown that there is a potential for substantial savings of public funds through better cost recovery in the road sector, which at the same time would increase sector efficiency. The principle that road users should pay for the use of the roads provided by the state and the damage they inflict on the road is accepted by the Government and partially applied in the form of tolls charged for the maintenance of specific rehabilitated or new roads. In addition, the Government charges a transport fee or surcharge of 500 dong per liter of gasoline which yielded US$ 100 million in 1995, US$ 136 million in 1996 and US$ 161 million in 1997. Theoretically this revenue is 50 Source: World Bank, Second rural roads Project, Report No. 19755-VN, of November 29, 1999, Attachment 1, p. 33 185 Vietnam Public Expenditure Review 2000 - Annexes earmarked transport investment and maintenance, but in practice is part of general budget revenue. 42. The present toll system is inefficient, costly and covers only a small amount of the road network while the transport fee is not specifically allocated to road expenditure and therefore does not ensure the proper maintenance of the road network. 43. MOT is therefore considering to submit to Government a proposal based on the principle that the road user should pay for the maintenance of the road network and have a voice in the administration and use of the funds. The proposed instrument is a levy on gasoline (similar to the one already in place) that would replace the existing tolls and whose proceeds would flow into a separate road maintenance fund, jointly managed by the Government and the representatives of the road users. Examples of road funds in the Philippines and under consideration in Lao PDR are given in Appendix G.2. The fund would be gradually built up so as to minimize economic dislocation. Initially the Government would transfer from the budget to the road fund the resources it now allocates for road maintenance and the road levy would supplement these resources. In the final stage, the road fund would be entirely funded by the levy on road users and eliminate the need for transfers from the budget. The net impact on consumers and producers of introducing a road fund should be positive, given the substantial net savings that are expected from improved road conditions. 44. It has been estimated that the budget required to properly maintain Vietnam's roads would be in the order of US$118 million per year (excluding urban and district roads) or 190 million per year(covering the whole network).51 As set out in the following box (Box G. 1) there is a range of ways in which such a fund might be structured. In addition to a surcharge on heavy vehicles the proposed road user charge could add between four to seven US cents per liter of fuel to the current cost of about US cents 28 (diesel) US cents 35 (gasoline) per liter, depending on the approach adopted. It is estimated that this would impact on the general price level by less than half a percentage point in the rural areas and about one percentage point in urban areas. If the levy were introduced over a number of years, as recommended, the impact on the general price level should be negligible. However, it would be economically detrimental to burden today's road users with the cost of rehabilitating a road network that has suffered from decades of under-funding. Rather, this cost should be shared by the entire population and therefore financed through the general budget from tax revenues and ODA. Sl See World Bank Report: Vietnam, Moving Forward, Achievements and Challenges in the Transport Sector, Chapter 3 and Annex 3.1. World Bank, Report No. 18748-VN, April 1999. 186 Annex G - Transport Sector Box G. 1: A Road Fund for Vietnam? Road funds take various forms. Those that function best provide a regular flow of finance to support spending; ensure that the public gets what it expects by involving road users in the management of the funds and limiting the road program to what users are willing to pay for; keep revenues separate from the Government's consolidated accounts; and reduce the road sector's fiscal burden on the Government. The road fund recommended for Vietnam would in its final stage be exclusively supported by user charges, mainly through a levy on fuel and a license fee on heavy vehicles. The fund should initially be dedicated to road maintenance, rather than rehabilitation and other expenditures, until maintenance needs are fully met. At the same time, tolls on specific publicly funded roads should be abolished. Coverage. Two questions determine the coverage of road funds. First, which costs should be recovered from road users? Second, what types of road should the fund cover? The answer to the first question is that road users should at least pay for the damage they cause to the road and in aggregate they should cover the cost of maintaining and operating the road system which means they should pay for the variable cost of road maintenance and contribute towards the fixed maintenance costs. The answer to the second question depends on a consensus between Government and road users. Alternatives include main roads only (national and provincial), a network of designated roads or the entire road network. The cost of the road fund will of course vary accordingly. Most countries with a road fund cover at least the main road network. Road User Charges Model. In order to explore the cost implications of various coverage scenarios a road user charges model developed by the World Bank Group has been used. It was assumed that, at a minimum, the road fund would cover the variable costs of road maintenance for all road types and in all scenarios the fixed maintenance costs for the national roads. By contrast, the fixed maintenance costs of provincial, district and urban roads would be funded in various proportions by the local governments concerned or from Government transfers. Further, the accumulated backlog of maintenance was assumed to fall into the investment category and would thus be funded in principle through the budget (at least until all maintenance needs of the selected road network have been met by the road fund first). The model was run for the designated road network of 72,150 km. To illustrate the volume of funding needed and the road user charges required, seven hypothetical scenarios have been considered with different arrangements for sharing the fixed costs of road maintenance with the Provinces, District and Urban Governments. (Of course, the model can test any other alternative chosen by the parties concerned). The first three scenarios cover the entire designated network, but have different cost sharing arrangements-ranging from 20 percent of fixed maintenance costs to 100 percent-for provincial and district roads. Urban Governments are assumed to cover 100 percent of fixed road maintenance costs. In these three scenarios the variable and fixed maintenance cost would amount to US$ 190 million per year and the road fund would cover between US$ 1 10-US$ 156 million. The next three scenarios were assumed to cover only the designated national and provincial roads with cost sharing arrangements of 20 percent, 50 percent and 100 percent for provincial roads. Total maintenance costs would be US$ 118 million and the road fund would cover between US$ 92-112 million. The last scenario includes national, provincial and district roads and assumes that Provinces and Districts cover 20 percent of the fixed road maintenance costs. In this case total funding requirements would amount to US$ 156 million and the road fund would cover US$ 152 million. Depending on the arrangements for sharing the fixed road maintenance costs the charges road users would have to pay will also be different. For example, the first scenario which assumes that provinces and districts would only contribute 20 percent of the fixed road maintenance costs of their designated networks the road user charge would amount to 7 US cents per liter of gasoline and 5 US cents per liter of diesel fuel. In addition, heavy vehicles would have to pay US$ 98 per equivalent standard axle per vehicle unit per year. If the provinces and districts cover the entire fixed road maintenance costs from their own resources the user charge would drop to 4 US cents per liter of gasoline and diesel and heavy vehicles would have to pay US$ 115 per equivalent standard axle per vehicle unit per year. The road user charges model calculates the variable and fixed road maintenance costs as a function of road type, vehicle type and utilization and road length. For example, for Vietnam's national roads included in the designated road network the model calculates the following total maintenance costs: Paved roads: US$72,000 per km/year; Gravel roads: US$27,000 per km/year; Earth:US$6,000 per km/year. These costs comprise annual and periodic road maintenance. Source: Seefootnote no. I in text. 187 Vietnam Public Expenditure Review 2000 - Annexes The 1996-2000 Public Investment Program (PIP): Plan vs. Actual 45. Table G.3 below is an attempt to compare the PIP transport investments planned for the 1966-2000 period (as recorded by MPI) with actual spending between 1996-98, bearing in mind that for the reasons explained earlier the table does not cover all state spending on transport. In addition, a comparison of the MPI data with MOF's data (Table G.4) shows significant discrepancies in actual spending during the period 1996- 1998, which may be due to differences in reporting 'C' type projects. Ministry of Transport Expenditure Information 46. The total capital expenditure (excluding rail) during the three years 1996-1998 is estimated at 8.9 trillion dong. Of this capital expenditure amounted to 6.8 trillion dong for national roads and 1.2 trillion for provincial and other local roads. This left about 14.1 trillion dong or 64 percent of the 22.3 trillion dong road investments planned in the PIP to be spent over the remaining two year period. In other words, project implementation was much slower than originally anticipated. It is doubtful, however, whether the shortfall in spending can be made up during the remaining time. A large number of ODA assisted projects (i.e. all large road projects) show implementation delays due to reasons other than lack of funding, and are unlikely to be fully implemented before 2002. In contrast the implementation delays of such centrally managed and predominantly locally funded "A" projects as the railways and certain sea and air ports seem primarily caused by a shortage of domestic funds. Anecdotal evidence suggests implementation delays in the localities were less than for large central projects. The reason may be the small size and the decentralized implementation of most local projects: any single locality manages a less complex project portfolio which normally is subject to less cumbersome administrative procedures. However the available data does not allow a comprehensive analysis of the implementation status of all "B" and "C " transport projects 188 Annex G - Transport Sector Table G.3: Public Investment Program 1996-2000 vs. Actual 1996-1998 -bill. dong- Output Period Funding Exp. 96- Plan 99- Total 98 2000 PIP Total State ODA Nat. km 28,552 5,761 22,781 6,817 13,138 19,955 Roads Of which Hghw. 1 1,620 96-00 8,783 1,369 7,413 2,532 6,251 8,783 Other 96-00 3,165 659 2,497 1,395 1,761 3,165 Hgw. 1 Hai Van 8.5 98-02 2,121 200 1,921 53 1,320 1,373 Pass Hghw.5 106 95-99 2,862 887 1,975 2,204 658 2,862 Hghw. 10 156 98-03 2,750 412 2,338 2 1,430 1,432 Hgw. 18 401 95-05 6,676 1,139 5,537 368 692 1,060 Hghw. 22 148 97-01 1,584 484 1,100 37 700 737 Other Nat. 611 611 226 326 552 Hghw. Local 2,701 1,111 + 590 1,234 1,147 2,381 Roads(a) 1,000(b) Rail 1,730 95-? 11,872 10,706 1,166 1,882 n.a. n.a. Ports 9(c) 950-02 5,578 3,744 1,834 401 2,677 3,078 Inl. 862 155 707 31 849 880 Water Air 992 922 70 451 541 992 Grand 38,675 12,693 25,982 8,934 18,352 27,286 Total(d) Source: MPI and own calculations. (a) Provincial, district and commune, incl. Bridges; (b) peoples contribution 1, 000 billion dong; (c) including Dung Quat port; (d) excluding rail. 47. It should be noted that these conclusions are based on MOF and MPI statistics. As pointed out earlier in the introduction to this chapter it has not been possible to reconcile this information with the data provided by the MOT. If the MOT data are correct spending for capital investments would be significantly higher for 1997 and 1998 (i.e. plus 40 percent in 1997 and plus 64 percent in 1998. For capital investment the higher MOT figure would mean, for instance, that all the investments planned in the PIP for these years have been implemented (although there is evidence from project implementation delays of donor funded projects that this cannot be correct). Obviously, it would be highly desirable to resolve these data discrepancies. 48. To the extent that MOT data is correct, it is interesting to note that since 1996 the actual expenditure in both the capital and recurrent categories exceeded the budgeted amounts for these categories (Table G.4). The reasons are not clear. However, it appears that the budget process allows some flexibility to shift funds from one year to another and from one budget item to another. It is also noteworthy that ODA funding of capital expenditure has increased substantially since 1997, when it reached more than 50 percent of the total. Although ODA commitments for previous years were substantial, management structures and processes to absorb the resources promptly and efficiently were not yet in place. This situation has improved, but absorption problems and weak 189 Vietnam Public Expenditure Review 2000 - Annexes implementation capacity still hamper the effective use of donor funds as discussed in a later section. Table G.4: Expenditure by the Ministry of Transport 1994-1999 Item 1994 1995 1996 1997 1998 1999 Actual Actual Budget Actual Budget Actual Budget Actual Budget Actual Capital 1.34 2.28 2.18 2.34 4.35 4.63 5.23 5.68 6.77 6.63 Expenditure Roads 1.00 1.86 1.88 2.04 3.16 4.05 4.04 4.81 5.67 5.69 Railways 0.19 0.23 0.14 0.14 0.33 0.11 0.33 0.20 0.42 0.39 Maritime 0.07 0.09 0.03 0.03 0.18 0.16 0.74 0.49 0.49 0.36 Inland 0.01 0.06 0.05 0.05 0.03 0.03 0.05 0.05 0.10 0.09 Waterways Other 0.07 0.04 0.08 0.08 0.65 0.28 0.07 0.13 0.09 010 Funded from: Budget 1.27 1.89 1.25 1.64 1.90 2.29 2.26 2.55 2.98 3.13 ODA 0.06 0.38 0.93 0.69 2.45 2.34 2.97 3.13 3.79 3.50 Current 823 1057 947 1152 1073 1211 1081 1219 890 1003 Expenditure Roads 360 554 460 600 495 577 445 568 270 336 Routine 216 431 320 460 350 432 267 390 132 198 Medium R. 144 123 139 140 146 145 178 178 138 138 Railw. 272 265 200 269 244 307 288 288 272 310 Il. Water 29 43 50 50 60 60 60 60 64 67 Sea 106 127 153 148 167 160 187 194 189 189 Other 56 69 84 86 107 107 102 110 96 101 Source: MOT Note: Capital exp. in trillion dong; current exp. in bill. dong. 1999 provisional. Improving the Quality and Management of Transport Infrastructure Projects 49. In the transport sector there is ample room for efficiency gains, both in planning and programming processes and institutions, apart from the strategy and structure of resource allocation. This section identifies some of the issues in improving the quality and efficiency of public expenditure in the transport sector. However, most of the issues are generic and should therefore be addressed across other sectors and Government entities. This section is based on various sources, including reviews carried out by Government entities, donors and consultants. It is not intended to be a comprehensive analysis since that would require a substantially more detailed and lengthy study. For ease of reference the discussion is organized along the lines of the planning cycle practiced in Vietnam. 190 Annex G - Transport Sector Quality of Proposals 50. Issue: Investment programs (5-year) proposed by line ministries and by localities for the PIP are not prioritized on the basis of a coherent and consistently applied set of project selection criteria and do not take account of funding constraints. In particular, project alternatives are not sufficiently explored. For example, better management and operation of existing infrastructures and facilities could often reduce investment requirements. 51. Recommendation. The project proposals should be based on a list of projects ranked in order of priority. The priority of projects should be determined - at the technical level - in accordance with the PIP objectives/targets stated by the Government and on the basis of technical criteria (such as cost/benefit and cost effectiveness analysis, project implementation schedule) developed by MPI, and applied consistently for all sectors and projects. This implies that at least 'A'type projects should be fully evaluated. 'B' and 'C'-type projects should be subject to a simplified evaluation methodology, but also clearly ranked in order of priority. 52. Line ministries and localities should be given (by MOF/MPI) an estimate of the likely funding available over the five-year period, to enable them to formulate realistic investment programs. 53. The Transport Master Plan, as well as the complementary rural and urban transport development plans now being prepared and considered for adoption should provide the basis for a coherent and prioritized public expenditure program in the transport sector. 54. Issue: Vertical and horizontal coordination of programs and projects is weak Coordination of projects between the center and localities is unsatisfactory. For example, national, provincial and district roads are both substitutes and complementary, since they are part of the same road network. Yet there seems to be very little coordination horizontally or vertically within the Government. The same problem occurs between entities at the same level. For instance, MOT which is nominally responsible for all transport planning is rarely consulted on transport projects of other line ministries (such as a commercial port project of the Ministry of Industry or Defense, or a rural roads project under the responsibility of the Ministry of Agriculture and Rural Development). District or provincial planning entities are often not concerned with (or consulted about) plans and projects of neighboring local governments or central projects which will affect them. 55. Recommendation. All publicly funded transport projects should be coordinated between the Government entities concerned before they are included in the PIP (and the Five Year Plan) and funded by the budget. In Malaysia, for instance, government decision making involves different levels of government, as in Vietnam, with each level focusing on those issues for which it is best suited (Box G.2). 191 Vietnam Public Expenditure Review 2000 - Annexes Box G.2: Coordinated Infrastructure Planning in Malaysia The institutional framework for infrastructure development in Malaysia has ensured that public provision of infrastructure has reflected both broad national priorities and local needs. Similar to Vietnam, Malaysia blends both centralized and decentralized forms of infrastructure planning. At the central level, national development objectives and targets are formulated by the National Economic Council (chaired by the prime minister) and the National Development Planning Committee (composed of top civil service officers form federal ministries). Alongside these two groups, the Economic Planning Unit (in the Prime Minister's department) acts as a coordination and integrating agency rather than an initiator of sectoral plans. Following deliberations by these two groups, the federal govemment's development policies and sectoral priorities are conveyed to federal ministries, statutory bodies and state government, which are invited to submit their proposals for the next five-year plan. Federal agencies that are located in the states are obliged to discuss their development programs with the appropriate State Economic Planning Unit before submitting them to the relevant federal ministry. This ensures that the state governments are aware of the development proposals of the federal agencies operating within their boundaries. The National Planning Committee is the ultimate coordinator and has jurisdiction over the expenditure programs for the five-year plans. Source: World Bank Report No. 12778- VN, Vietnam, Transport Sector, Serving an Economy in Transition, August 1994, Annex 3. More examples of transport planning methods applied by China, India, Hong Kong, Indonesia, Korea, the Philippines and Thailand can be found in Annex 3 of that report. 56. Issue: The capital expenditure and current expenditure programs are not coordinated Capital expenditure is subject to a five-year planning cycle and current expenditure follows an annual planning cycle. Yet current expenditure and capital expenditure may be substitutes. For example under-funding of current expenditure has resulted in severe and widespread road deterioration which now requires costly investment. Thus, most of the capital investment program for roads is the result of decades of under-funding of road maintenance. Furthermore, when planning investments, the follow-on costs to the recurrent budget for the maintenance of the assets are not explicitly taken into account. 57. Recommendation. Capital investment and current expenditure planning should be linked and be subject to the same planning cycles. This would enable planners to take account of trade-offs between current and capital expenditure and to avoid the present bias towards capital expenditure For example MOT's agency for road maintenance, the Vietnam Road Administration (VRA) could formulate a five year road maintenance program concurrent with the PIP, separating its program into the existing categories of recurrent expenditure and periodic maintenance expenditure (big repair). Annual maintenance expenditure would continue to be allocated directly by MOF in the recurrent 192 Annex G - Transport Sector and periodic expenditure would be included by MOT in its proposals for the five year period for the PIP (under the category 'big repair'). 58. Issue: The capacity to design, evaluate and review projects and programs is low. Consequently, there is too much reliance on stop-gap measures such as ad hoc outside consultants (for large projects), which does not ensure continuity of the planning process and the accumulation of an 'institutional memory'. There are three aspects to this problem: First, the entities responsible for planning and programming (i.e. MPI; MOT and its subsidiary planning entities; local planning departments) lack qualified staff at all levels and are thus unable to adequately prepare and scrutinize the large number of projects. With the exception of large, donor-funded projects this has resulted in an emphasis on formal project preparation requirements and less scrutiny of substantive aspects (such as the realism of assumptions and projections). Second, too little importance (and funding) has been devoted to improving the skill levels of planning and programming staff. This is a particularly important problem at the lower levels of the planning, i.e. the MOT's lower planning entities and the planning departments of the localities. Third, project preparation in the initial stages is not supported by a budget allocation for this purpose. According to current regulations a project budget can only be allocated once a proposed project has been approved by the Prime Minister, i.e. after all evaluation processes have been completed. This lack of funding often results in an adequate project preparation. 59. Recommendation. A systematic effort should be made to upgrade the skill levels of planning and programming staff at the central and local levels. This requires both increased funding and the development of a training scheme. Poorly prepared and designed projects lead to poor output and this is ultimately more costly than upgrading skill levels. 60. In the longer term, a selective increase in the number of planning and programming staff should be considered. In addition, Government should consider establishing a special allocation for project preparation to which central and local entities could access in the pre-approval stage. 61. Issue: The project approval process is top heavy and repetitive. Although the Prime Minister approves, after prior review and approval by appropriate entities such as MPI, MOF and the line ministry (MOT), the five year public investment plan, each A- type project has to pass through the same chain of approval from line ministry to Prime Minister again before it is adopted for implementation and funding. Certain elements within an A-type project, such as technical assistance programs also have to be reviewed and approved again by the Prime Minister's office before they can be implemented. If project cost estimates need to be revised by more than 10 percent the entire project evaluation process has to be repeated. 62. Recommendation. MPI (and for 'A'-type projects the National Project Evaluation Committee) should review line agencies' and local level agencies' proposals and prepare the draft PIP based on the projected resources available for the five-year period, after input from the National Project Evaluation Committee and in consultation with the appropriate line agencies. The draft PIP would be submitted to the Prime Minister for 193 Vietnam Public Expenditure Review 2000 - Annexes approval. The PIP projects and their implementation schedules would be the basis for the preparation of the annual expenditure plan of the budget. Once approved by the Prime Minister, the appropriate line Ministries and the designated local entities would be responsible for implementation of the PIP. The requirement for prior approval of individual projects or project elements (such as T.A.) by the Prime Minister or MPI should be abolished. In return for this flexibility the implementing agencies should be held accountable for the delivery of the project within the Plan and on budget. The Plan's progress and relevance would be reviewed annually by MPI and the implementing agencies and modified if necessary. Projects proposed to be added should be subject to the same evaluation criteria as the original Plan. The annual implementation progress report and proposals for modification, if any, would be submitted to the Prime Minister for review and approval. 63. Issue: Project planning and programming is burdened by overlapping and often inconsistent project preparation requirements of different funding (e.g. donor) or reviewing entities This leads to duplication of efforts and waste of scarce staff time. Project preparation entities may be required to prepare different reports for the same project. If more than one donor is involved in a project, their separate procedures (e.g. for project justification, resettlement, procurement) must be followed and different reports must be prepared for each donor agency. In addition, a different report is required by the Government's internal reviewing entities in accordance with their own rules and procedures. 64. Recommendation: The requirements for project preparation and programming should be harmonized to the maximum extent possible. The Government should set up a committee consisting of central and local agencies and the donor community, to review the present project preparation and processing requirements to reduce, simplify and coordinate procedures and reporting requirements. Project Implementation 65. Issue: Implementation Delays and Work Quality. As regards donor-funded projects, a 1996 joint review by Government and major donors concluded that inconsistent regulations, weak institutional capacity and the lack of a long term plan for ODA management were the main causes of the absorption difficulties. While some progress has been made there still are such areas as conflicting policies and procedures on persons displaced by major projects, a weak and undercapitalized local construction industry and, related to the latter, price ceilings for construction cost estimates that are unrealistically low, which prevent the modernization of plant and equipment of the local construction industry and often lead to low quality output. Some large locally funded projects also suffer from implementation delays. The quality of implementation also appears to be low in some cases. Experience with the implementation of the 1996-2000 PIP has shown that there is a tendency of planners to be over-optimistic. 66. Recommendation. Proposals to deal with the resettlement issue have already been made above and have also been dealt with in the portfolio performance reviews. As regards the local construction industry the Government should remove the remaining regulatory price constraints, which are a remnant of the old command economy. Rules 194 Annex G - Transport Sector and regulations on local and international competitive bidding are now in place and if applied consistently and rigorously should be adequate to ensure the lowest competitive prices. The lessons to be learned for the preparation of the next PIP are to fully allow for the country's still limited absorptive capacity. This implies focussing on the core priority projects that can be implemented on time, even if more donor funding for further large complex projects might be available, and to shift some resources to small scale local projects. Program and Project Monitoring 67. Issue: Program and Project Monitoring are focused on expenditure and less on the quality of implementation and outcomes. The current control systems for investment programs and projects are based on monitoring expenditure, but are not designed to monitor whether the expenditure provided value for money and produced the desired outcomes. The present internal quality control at the central and local levels is useful but there is no independent audit mechanism. Consequently there is very little feed-back into the planning process to incorporate lessons learned. 68. Recommendation. MPI should develop a process and indicators to monitor the efficiency of program implementation outcomes. The line ministry and the local entities should do the same at the project level. This information should be compiled into a periodic implementation progress report which should also flag implementation issues that have arisen or risks, and be submitted to the Prime Minister for review. Procedures should be developed to ensure that the lessons learned are systematically incorporated in the ongoing programs/projects and future plans. In addition, an independent external body reporting to the National Assembly should audit all large (transport) projects and 'B' and 'C' type projects of the center and localities on a sample basis, in order to ensure consistency in project implementation. Changing the Role of the Public Sector in Transport 69. Vietnam is still in the process of transforming its economic system from the former command economy to a model based on the principles of a market economy combined with a strong state presence. Finding the right balance between state involvement and a market based approach is not easy. It is not surprising, therefore, that the public/private balance in the transport sector is still evolving. This has important implications for public expenditure. 70. On the one hand divestiture/equitization of public enterprises implies that payments which SOEs make to the Budget will drop initially, but in the long run be replaced by higher tax revenues. On the other hand the claim on the state's resources, for example for subsidies and investment capital, will also be reduced, releasing funds for core state activities . However, the current overall budget impact of SOEs is not known, nor is the potential impact of any divestiture. 71. There appears to be scope to re-balance the public/private sector mix in transport, thus increasing economic efficiency through reducing operating costs and the sector 195 Vietnam Public Expenditure Review 2000 - Annexes claims on the state's scarce resources, while preserving the essential role of the state as regulator and guardian of the public interest. The progress made so far and the remaining agenda as well as the scope for moving further towards market-based solutions for the transport sector are explored below. Organization of the Transport Industry 72. Vietnam's transport sector has undergone significant changes since the introduction of market principles into a socialist economy . This is reflected in road and inland waterway transport which are now predominantly operated by non-state entities. Thus, in 1998 about 98 percent of the freight capacity and 94 percent of the passenger capacity of road vehicles and 71 percent of the freight capacity of inland waterway vessels was provided by non-state entities. Of those, transport collectives carried about 21 percent of the freight volume and 22 percent of the passengers. By contrast, the railways, as well as air and sea transport are dominated by public enterprises and the participation of the non-state sector in transport infrastructure provision and operation is negligible. 73. This situation is partly a reflection of market realities. After the removal of barriers to private ownership the non-state sector has quickly taken hold in markets with low entry barriers, such as road transport and small vessels inland waterway transport. It also, reflects the still incomplete process of transformation. Although the Government has slated the majority of its transport State-Owned Enterprises (SOEs) for divestiture, only a few have been divested. Role of the State. 74. In keeping with the objective of developing a mixed economy, the Govermnent's policy is to retain in state ownership the country's basic transport infrastructure and those transport enterprises which it considers of strategic importance (e.g. ports, airports, sea transport). At the same time, the Government's aim is to improve economic efficiency. These two basic objectives should complement each other. The challenge is to ensure that the principle of economic efficiency is given full weight and applied in the management of transport enterprises, regardless of ownership and industrial organization. This means that: (i) the transformation of non-strategic enterprises from state to non-state ownership and management should be expedited; (ii) the SOEs remaining in the state sector should operate in a commercial environment and be subject to the same standards, incentives and disciplines of the marketplace as private sector companies, even if they provide a public service; and (iii) the role of the Government should be focused on establishing a regulatory framework that fosters efficient markets and fair competition between all market participants. Appendix G. 1 sets out the way these institutional arrangements would operate. Divestiture Program for Transport SOEs. 75. The divestiture process of transport SOEs has been slow to date. Out of a total of about 300 transport SOEs of varying size for which the Ministry of Transport (MOT) is 196 Annex G - Transport Sector still responsible, only about 40 had been divested by the end of 1999. The core activities of those enterprises considered of strategic (or security) significance are to be retained under State ownership. There are Vietnam Railway, the seaports and the three maritime shipping companies. Further, the national airline and the airports, under the direct supervision of the Prime Minister, are to be retained as well. The remaining 253 companies, comprising among others transport operating companies; construction, maintenance and repair companies; engineering and design companies; freight forwarding agencies, hotels and catering companies and so on, are slated for divestiture over the next few years. In addition Provincial Peoples' Committees own numerous companies/enterprise, such as bus and taxi companies, road construction and maintenance companies, consulting engineering firms, etc most of which are also awaiting divestiture. 76. There are a number of reasons for the slow progress. These include an incomplete or inappropriate legal framework; inexperience with the divestiture process; lack of interest and financing capability by the non-state sector; vested interests in the status quo within government organizations, SOE management and employees; concern about the social consequences of redundancy (i.e. lack of social safety net). However, it is expected that the divestiture process in transport will accelerate significantly over the next year, since the Minister has given it highest priority and instructed his staff accordingly. MOT has received technical assistance to support its divestiture work. Major cities, such as HCMC, Hanoi and Haiphong also seem to have shed a number of enterprises but little progress seems to have been made in the provinces. As mentioned in Chapter 1 the Government is about to announce a major program for SOE reform, which will include the transport sector. Increasing the Scope for Private Sector Participation 77. If the objective of retaining some SOEs is to exercise a certain influence and national control over activities whose proper functioning is considered of vital economic interest to the state, such as sea or air transport, ports and airports, outright ownership and/or operation is not the only solution, particularly if the equally important objective of economic efficiency is taken into account. Public ownership and operation may not always be necessary to ensure the state's legitimate public interest. For example Vietnam's regional neighbors, Thailand, Malaysia and Indonesia do not rely on outright ownership or operation to ensure adequate and competitive maritime or air transport for their countries' international trade. These services are usually provided by private (national and foreign) enterprises and based on appropriate national and international regulations to ensure fairness, safety, security and the preservation of the state's legitimate interests. 78. Efficiency can also be increased by retaining ownership but contracting out operations and exposing the latter to the commercial disciplines of the marketplace. Vietnam's major international ports and airports could increase their operating and commercial efficiency significantly while remaining in state ownership, if the private sector were allowed to operate them on a commercial basis and within a market oriented regulatory framework. A good example is the ICTSI container port in Manila (Philippines) which was leased out to a private company and which rose from an 197 Vietnam Public Expenditure Review 2000 - Annexes obsolete, loss-making port to one of the most efficient container ports in the region (Box G.3). China opened the management of its ports to the private sector in 1991. For example, a private Hong Kong operator (Hutchinson Whampoa) took over the development and management of container facilities in Shanghai in 1993 and in Yantian in 1994. By 1998 thirteen port facilities in China were managed by the private sector. Elsewhere in the region, Korea (Pusan, Kwangyan) and Thailand (Laem Chabang) granted leases to the private sector for the operation of government owned port facilities. Box G.3: Manila International Container Terminal (MICT) In 1988, the Philippine Ports Authority (PPA) decided to contract with the private sector for the operation of MICT. The successful bidder was a consortium of local enterprises and a foreign operator. The concession contract stipulated ICTSI investment of US$54 million in new cargo handling equipment and other facilities. Moreover, the consortium had to guarantee PPA at least US$550 million in revenue during the 25-year duration of the concession period. The consortium started its operations in June 1988. Since then, the operation of MICT has been a success story. For example, ship turnaround time was reduced by 60 percent almost immediately. ICTSI investment in MICT has turned the facility into the most important container port in the Philippines, capable of handling one million TEU (twenty foot equivalent units) per year. The revenue from ICTSI became the main source of income for PPA. ICTSI itself has been profitable every year since its inception. With explicit support from the Philippine Securities and Exchange Commission (SEC), ICTSI was listed on the Philippine Stock Exchange and the flotation of 33 percent of the consortium's stock was well oversubscribed. 79. Provided Vietnam's sea ports are allowed to compete with each other on a commercial basis,( i.e. having financial and other performance targets and the ability to set their own prices so as to achieve these targets) either outright private ownership of ports or parts of them (such as container or break bulk terminals) or BOT (Build-Operate- Transfer) type arrangements could also be considered, in particular for new investments for modernization, rehabilitation and expansion. Examples of this possibility in Vietnam are the small Baria-Serese port complex and the Thuon Dong container port-Vietnam International Container Terminal (VICT) near HCMC. 80. The country's international airports require substantial resources for their modernization and expansion which could be met by utilizing their potential for public- private sector partnerships in the form of concessioning of terminal operations and airside services. Experience elsewhere shows that publicly operated airports/terminals are less likely to maximize the commercial potential for the development of airport business opportunities than privately operated airports/terminals. An example of a successful public/private partnership in airport operations is the Subic Bay Airport in Manila which was developed from an abandoned military base into one of the most important airfreight hubs in Southeast Asia. 81. Other SOEs may require restructuring to enable them to focus on their core activities and thus increase their efficiency. A typical example is Vietnam Railway (VR) 198 Annex G - Transport Sector which could divest itself of a number of activities to its main mission (such as rolling stock repair, quarry, hotel and catering operations), operate services in accordance with strict commercial principles and receive budget funding for any public service obligations required of it by the state. 82. The relatively few activities in the transport sector which require exclusive public ownership and operation are reflected by the functions related to the state's exercise of its sovereignty or regulatory oversight duties such as customs control, the duties of the harbor master, traffic regulatory agencies, safety regulation and control. In most other cases, however, public/private partnership arrangements offer greater gains in economic efficiency as well as protecting national or strategic interests provided the roles, rights and obligations of the public and private sectors are well defined and adhered to. 83. Even non-revenue earning transport entities such as roads can increase their economic efficiency, and thus reduce their demands on state funding, by applying market-based solutions. Such studies as the VITRANSS Transport Master Plan have shown that the funding requirements for road rehabilitation, modernization and maintenance are a heavy burden on the state budget and can be only partly funded through the budget.52 For this reason the Government has attempted to find alternative sources of funding and has offered a number of roads to private sector funding and operation. So far only a few small domestic contractors are involved in BOT schemes for some small bridges and highway sections, for which they have the right to collect tolls. Foreign enterprises have so far been reluctant to invest in road infrastructure schemes because of low traffic levels and ability of users to pay, the lack of an enforceable legal framework and the risks thus involved, high up-front costs and the relatively low (risk adjusted) returns to be earned. In fact only one local/foreign partnership (between a Taiwanese company and the city of HCMC) has materialized so far in the form of a 17.8 km tolled highway between Tan Thuan Export Processing Zone and highway No. 1. Although the Government is seeking private financing for a number of new roads and bridges, foreign investors have shown little interest so far. Thus, the Government will most likely have to continue to meet the bulk of the funding requirements for the road network. Future Expenditure Strategy 84. As part of the preparations for the PIP 2000-2005 and the Government's long- term transport master plan several strategy documents are being formulated that will determine the policies and related funding requirements for the next decade. A transport master plan that covers all inter-urban transport is in the process of finalization.5 It identifies an overall level of spending for transport infrastructure, based on certain assumptions on resource availability; it addresses the proposed ranking of major projects and suggests appropriate levels of recurrent expenditure for maintenance (Box G.4). 52 The Study on the National Transport Development Strategy in the Socialist Republic of Vietnam (VITRANSS), Final Report, MOT/JICA, July 2000 53 Study on the National Transport Development Strategy in the Socialist Republic of Vietnam (VITRANSS), MOT/JICA, Final Report, Hanoi, July 2000. 199 Vietnam Public Expenditure Review 2000 - Annexes Separate, but coordinated strategies are being developed for rural and urban transport. Full cost estimates are incorporated in all these proposed strategies. Box G.4: Broad Strategy for Future Spending * About 20 percent of the budget envelope should be allocated to maintenance and minor improvements. * About 8-8.5 percent of the budget envelope should be allocated to improvement of accessibility in rural (remote and mountainous) areas. Communes which are presently inaccessible should be provided with adequate access. * Priority should be given to projects that remove traffic bottlenecks and strengthen the network to meet traffic demand. * Strategic transport infrastructure for land, water and air transport in and between the north, center and south growth corridors should be improved and integrated with transport links to regional and global markets. * Urban transport should be improved so as to reduce congestion and to support the urban centers' function as engines of growth and gateways to regional and international markets. Source Adaptedfrom VITRANSS Master Plan Study, Final Report, July 2000 85. These strategies are being reviewed at present by the Government and donors and other interested parties are being given an opportunity to present their views as well. The policy objectives are to achieve a transport sector that is competitive, equitable and responsive to demand, capable of sustaining a balanced economic development of the country, supporting the poor in rural and remote areas and facilitating integration in regional and global trade all in an environmentally responsible manner. Since these strategic are now being considered by the Govermment it would be premature to propose a detailed public expenditure program in this PER. However, it seems appropriate to set out the basic parameters being considered, which will likely guide future transport spending. * The Government should increasingly focus on policy and regulatory functions and reduce somewhat its role in funding transport infrastructure and equipment. This means that public expenditure should be concentrated on non-revenue earning transport infrastructure such as roads, inland waterways, the sub-structure of ports and the fixed infrastructure of the railways. In these areas, user charges should be levied for the maintenance of the transport infrastructure and, if appropriate, contribute towards the capital cost as well. * Revenue earning entities of the state, such as transport SOEs, port, airport and railway operations (or with the potential to earn revenue), should either be considered for divestiture or, if retained for strategic reasons, should be considered for public/private partnership arrangements. SOEs should operate under commercial principles, having financial targets and management autonomy to achieve these targets. The required investment and other expenditures should be funded by these entities from their own resources or borrowing on a commercial basis. If this principle were applied to the 200 Annex G - Transport Sector country's ports and airports, for example, a significant amount of the investment required for their rehabilitation, upgrading and expansion could be shifted from the budget to the private sector. * The rehabilitation and modernization of the country's transport infrastructure will continue to be a priority. However, the resources likely to be made available will not exceed 2.5 percent of GDP on average. The plan implies a public expenditure program for transport (capital and recurrent) by the center and local authorities of about US$ 1 - 1.2 billion per year for the five-year period. This is more than double the capital resources currently being absorbed by the sector. While the recommended investment volume may be appropriate for a long-term program of say 10 to 20 years, if viewed against the still limited capacity of the center and local entities to effectively implement such large investment programs such higher spending seems unrealistic for the 2001-2005 period. It should therefore be carefully phased in over a longer time period. For the forthcoming 2000-2005 PIP, explicit account should be taken of the implementation constraints of the center and the local government entities, even if all the proposed projects can be justified on economic or social grounds. Funding such a large program will not be easy. According to MOT figures, more than 50 percent (or US$ 220-250 million per year) of the central transport capital expenditure (by MOT) since 1997 was funded by donors. It is unlikely that donors will increase their transport assistance substantially, given their new emphasis on social sector funding. This would mean that the entire balance (of US$ 750 million per year on average) would have to be raised locally. In view of the overall funding constraints the Government is facing, it appears unrealistic to assume that the allocation of public resources to the transport sector will increase significantly, let alone to this high level. * Therefore, if the investment program were to be fully implemented complementary funding mechanisms, such as special contributions by the population or the issue of bonds will have to be increasingly utilized. This would need to be carefully considered by the Government in the light of spending demands in other sectors. * Whatever funding is ultimately decided the broad strategy outlined in the VITRANSS transport master plan seems appropriate. (Box G. 4). 86. Based on the strategy, the priorities for a transport expenditure program have been identified by VITRANSS as follows: * First priority should go to funding maintenance of existing (maintainable) and rehabilitated transport infrastructure, rural transport, as well as the completion of ongoing and committed projects. * Second priority should go to new projects with the highest ranked evaluation (based on multi-criteria analysis). For projects of public interest with (potential) high financial rates of return, private sector participation should be solicited. 201 Vietnam Public Expenditure Review 2000 - Annexes 87. These two project categories would comprise the core public expenditure program and exhaust the projected funding ceiling. * Third priority would be assigned to projects ranked below the second priority projects, but only be taken up in the PIP if additional funding and implementation capacity became available. 88. While this strategy is broadly appropriate much depends on the available level of future funding. Given the likely limitations on resources it is suggested that the Government should consider the following options to maximize the use of the limited transport budget. * Consider an increased role for private ownership or operation in the transport sector; * Accelerate the reform of transport SOEs, including divestiture, where appropriate, and a commercial performance framework for the SOEs remaining in state ownership; * Consider the establishment of a dedicated road fund financed by user charges in the form of a fuel levy (for diesel and gasoline) to fund future road maintenance; 89. Further, this review also suggests two other important policy issues for the government to consider. * Levels of maintenance for inland waterways should be further increased; * Expenditures for the construction and maintenance of rural roads should be increased. 202 Annex G - Transport Sector Appendix G.1: Key Elements of Efficient Provision of Infrastructure and Services under Alternative Institutional Options Option I: Public Ownership and Public Operation Examples: Railways; Ports * Government roles as owner, regulator and operator clearly separated. * No Government interference in day to day management. * Public enterprises subject to commercial law (including auditing standards) and operating under the same conditions as private enterprises. * tariffs set to achieve cost recovery, as appropriate, and enterprises subject to hard budget constraint (no subsidies to cover losses). * Public Service operations to be identified and compensated explicitly (i.e. on the basis of contractual agreements) by Government. * Managers selected on the basis of professional qualifications and track record and compensated in accordance with comparable private sector compensation. * Appropriate mechanisms in place to obtain feedback from users. * Separate activities and functions that can be unbundled (e.g. track maintenance, rolling stock maintenance) to allow private sector participation (e.g. through service contracts). * Private management skills should be obtained as needed (e.g. through management contracts for ports). * Ownership and control shared with private sector wherever feasible (e.g. in joint ventures, BOT). Option II: Public Ownership and Private Operation Exarnple: Port (Airport) concession to private operator * Basic legal framework of contract law, including credible enforcement mechanism, should be in place. * Contracts specify clearly monitorable performance targets; responsibilities of owner (i.e. public entity) and operator; mechanism to dissolve disputes; sanctions for non- performance. * Tariff autonomy or, in cases of monopoly, agreement on an appropriate pricing formula and processes for periodic review (to account for unforeseen changes in input costs). 203 Vietnam Public Expenditure Review 2000 - Annexes * Contracts awarded (and re-bid from time to time) on the basis of transparent selection process (i.e. competitive bidding). Option III: Private Ownership and Operation Example: Shipping (coastal, maritime, inland water); Airlines * Appropriate deregulation has taken place to permit competitive restructuring of sub- sector. * Statutory and practical barriers to market entry removed (e.g. restrictions on or preferential treatment of certain market participants to access to credit and foreign exchange). * Regulation(s) in place to safeguard public interest (where competitive discipline is inadequate) and to ensure access to network facilities on a non-discriminatory basis when relevant. Option IV: Communal and User Provision Example: Community road maintenance; road fund based on user charges. * Participation of users or community members from earliest stage of program preparation to ensure willingness to pay and ownership of scheme. * Participation of beneficiaries ensured through appropriate organizational means, and with contributions in kind or in cash. * User group supported by access to training and technical assistance from sector agency or NGO. * Appropriate consideration give to interconnection with primary and/or secondary network infrastructure, where relevant. * Service operators appropriately trained, compensated and held accountable. Source: Adapted from: Ch. Kessides, Institutional options for the provision of Infrastructure, World Bank Discussion Paper 212; World Development Report 1994: Infrastructure for Development. 204 Annex G - Transport Sector Appendix G.2: Main Features of Road Funds in the Philippines and in Lao PDR Philippines The road sector is a major economic factor in the Philippines. The 161,000 km road network has a replacement value of 1 trillion Philippine pesos (PHP), equivalent to 40 percent of the country's GDP. Road transport costs comprise about 330 billion PHP annually and road maintenance and development cost about 40 billion PHP per year. Road vehicles consume about 8.5 million liters of fuel per year. Road transport inputs generate about 60 billion PHP in taxes, equivalent to 20 percent of total government revenue. Despite the crucial importance of these assets for the Philippine economy years the road network deteriorated because of systematic under-funding of required maintenance. Currently, only 38 percent of the National Road network is in good condition and 58 percent is paved. It has been estimated that only about 40 percent of total requirements for road maintenance are funded, a situation very similar to that of Vietnatn. In addition, there is a large outstanding backlog of both preservation and upgrading works. The Government realized that a substantial national reform effort was needed to restore the road network to optimum efficiency which was seen as a crucial component of improving the country's overall competitiveness in the global market place. The two key elements of creating a sustainable road maintenance management and funding system identified were: (a) Commercialization of the road system, with roads viewed and managed as a business rather than as a social service; (b) Involvement of road users-as clients-in the management and funding of roads. Main features of the Road Fund. Funding instrument: A Motor Vehicle User Charge (MVUC), payable yearly. An additional fuel levy is under consideration. Administration: The MVUC is earmarked to a special Road Fund established by law. The revenue collected must be used, in specified proportions, for the maintenance of the primary national road network; secondary national roads and provincial roads; and road safety improvements. A Road Board has been established and is mandated with the prudent and efficient management and utilization of the Road Fund. The Road Board is composed of seven members, appointed by the President of Philippines, as follows: four ex-officio members from Government departments concerned and three members from transport and motorist organizations. Status of the Road Fund& The Road fund is distinct and separate from and in addition to any appropriation authorized and granted yearly to the Philippine Department of Public Works and Highways and the Department of Transport and communications for road 205 Vietnam Public Expenditure Review 2000 - Annexes maintenance. Revenue not spent in any one year can be carried over to subsequent periods. Implementing Rules and Regulations: To be promulgated by the Secretaries of the Department of Public Works and Highways and the Department of Transport an Communications. Lao PDR (under consideration) Funding Instrument: Revenue shall be derived from user charges, including among other sources to established, a heavy vehicle surcharge, tolls, international transit charges, funds from donors, and mainly a surcharge on the price of gasoline and diesel fuel (the fuel levy). Fuel used for off-road purposes shall be compensated for having to pay the fuel levy. Administration: The revenue derived from the fuel levy shall be collected independently and deposited directly into the road Maintenance Fund. The Road Maintenance Fund shall be set up as a separate account, operated and managed by a Road Board. A Secretariat will be instituted for the daily administration of the Road Maintenance Fund. The road board shall supervise, direct, and recommend on matters related tot he overall management and financing of road maintenance. The Members of the Board are appointed by the Prime Minister and include nominees from the Ministry of Finance; Ministry of Communication, Transport, Post and Construction; Provincial Authorities (MCTPC); Lao Chamber of Commerce and Industry; Association of Road Transport Operators; Road Users Association. The Chairperson o f the Board shall be elected by its members. The Board reports to the Minister of MCTPC. Application of Revenue collected: The revenue collected and deposited into the Road Maintenance Fund shall be applied in order of priority to: (a) routine maintenance, emergency and periodic maintenance works, and administration; (b) renewal works, provided that money shall not be allocated for renewal works before all maintenance works have been adequately provided for; and (c) road safety projects and activities when resources permit. Allocation of Revenue: Allocation of funds for maintenance works along the road network shall be based on need, functional classification and on economic considerations. The ultimate objective is the preservation of the road network asset. Audit and Reporting: The account of the Road Maintenance fund shall be subject to annual audits to be carried out by independent auditors and comply with Government Audit Regulations. The Board shall submit an annual report to the Prime Minister. Both reports shall be made available to the public. 206