66317 Tax Reform in Vietnam: Toward a More Efficient and Equitable System Gangadha Prasad Shukla Duc Minh Pham Michael Engelschalk Tuan Minh Le (Editors) September 30, 2011 Poverty Reduction and Economic Management Unit East Asia and Pacific Region Table of contents Foreword .............................................................................................................................................................. 11 Acknowledgments ............................................................................................................................................ 13 Abbreviations ...................................................................................................................................................... 15 Executive Summary by Martin Rama, Deepak Mishra, and Duc Minh Pham................................17 Chapter 1: The Legal Framework for Tax Administration by Duc Minh Pham, Tuan Minh Le, and Gangadhar Prasad Shukla ....................................................................................................................... 33 1. Summary ........................................................................................................................................................... 33 2. Introduction ...................................................................................................................................................... 37 3. Value-Added Tax .............................................................................................................................................. 41 4. Excises or the Special Consumption Tax (SCT) ......................................................................................43 5. Trade Taxes ....................................................................................................................................................... 45 6. Direct Taxes ...................................................................................................................................................... 47 6.1. Personal Income Tax (PIT) ..................................................................................................................47 6.2. Corporate Income Tax (CIT)...............................................................................................................51 7. Taxation of Natural Resources ...................................................................................................................... 53 8. Nontax Regimes (Fees and Charges) ..........................................................................................................55 9. Tax Administration.......................................................................................................................................... 57 Chapter 2: Tax Compliance and Sources of Revenue Leakage: Conceptual Framework and Assessment by Tuan Minh Le, Duc Minh Pham, G.P. Shukla, and Gregory Kisunko .................59 1. Summary ........................................................................................................................................................... 59 2. Introduction ...................................................................................................................................................... 60 2.1. Tax Compliance and Revenue Leakage: A Conceptual Framework............................................60 2.2. Measurements of Compliance and Noncompliance ........................................................................69 3. Conclusion ........................................................................................................................................................ 73 Chapter 3: Tax Administration by Duc Minh Pham and Michael Engelschalk ..............................75 1. Role and Challenges of Tax Administration in Vietnam .........................................................................75 2. The Need to Modernize the Tax Administration System ........................................................................76 3. Ensuring Compliance in a Changing Environment ..................................................................................81 4. Dealing with Small and Medium Taxpayers...............................................................................................84 5. The GDT Reform Initiatives ......................................................................................................................... 85 6. External Assistance .......................................................................................................................................... 87 7. The Critical Issue of Reform Management and Sustainability ...............................................................91 8. Can GDT Become a Model for Integrity Enhancement?........................................................................92 9. Stakeholder Involvement in the Reform Process ......................................................................................95 3 10. Summary of Key Findings and Recommendations ................................................................................97 Chapter 4: Forecasting Tax Revenue: Methods And Tools by Gangadhar Prasad Shukla, Daniel Alvarez, Tuan Minh Le, and Duc Minh Pham ...........................................................................101 1. Summary .........................................................................................................................................................101 1.1. Organizational Set Up and Forecasting Process .............................................................................101 1.2. Data Collection and Data Integrity....................................................................................................101 1.3. Present Status of Revenue Forecasting in GDT ..............................................................................102 1.4. Revenue Forecasting Techniques and Modeling ............................................................................102 1.5. Strategy for Revenue Forecasting in Vietnam .................................................................................103 2. Introduction ....................................................................................................................................................104 2.1. Why Revenue Forecasting? .................................................................................................................105 2.2. Present Status of Revenue Forecasting in Vietnam ........................................................................107 2.3. Organizational Structure .....................................................................................................................107 2.4. Legal Framework ...................................................................................................................................108 2.5. Reporting Procedure ............................................................................................................................108 2.6. Coordination within MoF ...................................................................................................................108 2.7. Linkage of Revenue Projection with the Budget Process..............................................................109 2.8. Data Collection, Data Maintenance, and Data Integrity ...............................................................109 2.9. Present Status of Database and Revenue Forecasting in GDT ....................................................109 2.10. Data from other Sources ....................................................................................................................111 3. Conclusion ......................................................................................................................................................111 3.1. International Practice in Revenue Forecasting ................................................................................111 3.2. Revenue Forecasting Techniques and Modeling ............................................................................112 3.3. Strategy for Revenue Forecasting in Vietnam .................................................................................114 3.4. Revenue Forecasting Model and Results for Vietnam ...................................................................115 3.5. Summary of Model Results .................................................................................................................127 3.6. Suggestions for Improving Revenue Forecasting ...........................................................................128 Chapter 5: Tax Reform and Revenue Assignments in Vietnam by Jorge Martinez-Vazquez and Minh Van Nguyen.............................................................................................................................................129 1. Introduction ....................................................................................................................................................129 2. Review of the Existing Revenue Assignments .........................................................................................130 3. What is to be Accomplished by Revenue Assignments? .......................................................................140 4. Desirable and Less Desirable Forms of Tax Autonomy .........................................................................143 5. Significant Policy Issues with Revenue Assignments in Vietnam ........................................................145 6. Guidelines for Reforming Tax Assignments in Vietnam .......................................................................147 6.1. What kinds of Tax Instruments are best suited for Subnational Governments? .....................147 6.2. Selecting Tax Instruments for Assignment at the subnational level ..........................................149 4 7. Recommendations for Reform ...................................................................................................................155 Chapter 6: Integration and Government Revenues: Assessing the Implications of Vietnam’s Accession to the WTO by Robert Warner................................................................................................159 1. Summary .........................................................................................................................................................159 2. Introduction ....................................................................................................................................................160 3. Vietnam’s system for Taxing and Managing Trade..................................................................................161 4. Analyzing the Revenue Impact of Tariff Bindings...................................................................................176 5. Industry and Tariff Protection in Vietnam ...............................................................................................182 Chapter 7: Value-Added Tax by Gangadhar Prasad Shukla, Tuan Minh Le, and Duc Minh Pham ...............................................................................................................................................191 1. Introduction ....................................................................................................................................................191 2. Exemptions .....................................................................................................................................................198 3. Taxation of Special Sectors of the Economy ............................................................................................204 4. Zero-rating and VAT Refunds.....................................................................................................................206 5. Policy and Administrative Recommendations ........................................................................................211 Chapter 8: Taxation In Vietnam: Who Pays What? by Jonathan Haughton .................................215 1. Introduction ....................................................................................................................................................215 2. Current Tax Structure ...................................................................................................................................215 3. The Structure of Household Income and Expenditure..........................................................................219 4. Tax Incidence..................................................................................................................................................224 5. Property Tax ...................................................................................................................................................236 6. Checking for Robustness .............................................................................................................................238 7. An Assessment of Incidence and Its Implications...................................................................................239 Chapter 9: Property and Land Taxes by Bruno Moser and Nicolaus Tideman ...........................243 1. Property and Land Taxes: Legal Background and Revenue .................................................................243 1.1. Summary of laws related to land taxes ..............................................................................................243 1.2. The House and Land Tax .....................................................................................................................243 1.3. Agriculture Land-Use Tax ...................................................................................................................244 1.4. Land Use Right Transfer Tax ..............................................................................................................245 1.5. Land and Water Surface Rental Charges ..........................................................................................245 1.6. Land Use Levies.....................................................................................................................................245 1.7. Registration Fee .....................................................................................................................................246 2. The Economics of Property Taxes..............................................................................................................247 3. Land Tax ..........................................................................................................................................................248 4. Equity ...............................................................................................................................................................251 5. Tax on Improvements ...................................................................................................................................252 6. Tax on Real Estate Transfers ........................................................................................................................253 7. Policy Implications ........................................................................................................................................254 8. Land Recovery ...............................................................................................................................................254 5 9. Settling Land Disputes..................................................................................................................................255 10. Land Fragmentation and Consolidation ................................................................................................256 11. Internalizing Externalities ..........................................................................................................................256 12. Land Taxes and the Poor............................................................................................................................257 13. Application and Exemptions.....................................................................................................................258 14. Equitization of State Owned Enterprises (SOEs) ................................................................................258 15. Taxes on Rental Income from Real Estate..............................................................................................259 16. Protecting against Sudden Increases in the Rental Value of Land .....................................................259 17. Pilot Projects ................................................................................................................................................259 18. Practical Application of Taxing Scarce Space: London’s Congestion Charge................................260 19. Administrative Issues of Land Value Taxation ......................................................................................260 20. The Optimal Duration of Land Use Rights ............................................................................................262 21. Revenue Potential of Land Rent ..............................................................................................................263 22. Sharing the Revenue from Rent Collection among Levels of Government....................................264 23. Should Land be Rented or Sold?..............................................................................................................266 24. Dealing with Land that Has Been Sold by the State .............................................................................266 25. Securing the Natural Value of Land by Bonds .......................................................................................268 26. Implementing Land Value Taxation ........................................................................................................269 Annex ...................................................................................................................................................................273 References ..........................................................................................................................................................295 Tables Table 1.1: Average Buoyancy of Tax System, 1997–2009............................................................................40 Table 2.1: South Africa: Tax and Accounting Costs as Percentage of Turnover .....................................63 Table 2.2: Revenue-Depleting Impact of Corruption ..................................................................................65 Table 2.3: The Bigger the Shadow Economy, the Greater the Tax Leakage .............................................69 Table 2.4: Estimation of Operating Surplus by Economic Sector, 2005 ..................................................71 Table 2.5: Estimation of CIT Tax Gap, 2005..................................................................................................72 Table 2.6: Tax Filing Compliance by Major Taxes, 2010 .............................................................................73 Table 3.1: Paying Taxes in Vietnam ..................................................................................................................79 Table 3.2: Registered Taxpayer Population: An International Comparison ............................................81 Table 3.3: A New GDT Results Monitoring Framework.............................................................................92 Table 3.4: Which Are the Most Corrupt Government Agencies in Vietnam? ........................................93 Table 3.5: Key Issues in Anticorruption Strategy ..........................................................................................94 Table 4.1: Budget Cycle ....................................................................................................................................105 Table 4.2: VAT Policy Impact Scenarios, 2007 ............................................................................................118 Table 4.3: Average Tax Rates for Different Levels of Income Tax Ordinance 2004 and Current Law 6 Enacted in 2009 ..................................................................................................................................................123 Table 4.4: Assessing Progressivity of Old (2004 and New (2009) PIT System ...................................124 Table 4.5: 2007 Revenue Forecasting Summary .........................................................................................127 Table 5.1: Tax Sharing Rates for the Provinces 2004–06, 2007–10, and 2011–15 Stability Periods....131 Table 5.2: Provincial Revenues by Revenue Sources and Summary Statistics, 1997–2010 ...............134 Table 5.3: Provincial Revenue per 1,000 Population and Summary Statistics, by Revenue Sources, 1997–2010 ..........................................................................................................................................................137 Table 5.4: Outcomes of State Budget Revenue Decentralization in Vietnam .......................................139 Table 6.1: Composition of Revenues .............................................................................................................161 Table 6.2: Summary of Vietnam’s Preferential or MFN Tariff Structure, 2005–09..............................162 Table 6.3: Summary of Vietnam’s CEPT tariff schedule ............................................................................163 Table 6.4: Imports by Region of Origin: ASEAN and the Rest of the World, 1995–2005 ................163 Table 6.5: Application of VAT to Goods since 2009 ..................................................................................165 Table 6.6: Special Consumption Tax on Goods (as at August 2007) .....................................................167 Table 6.7: Excise Tariff since April 2009 ......................................................................................................167 Table 6. 8: Vietnam’s Tariff Binding Transition............................................................................................172 Table 6. 9: Tariff Bindings for Countries Acceding to the WTO .............................................................173 Table 6.10: Value of Imports by Customs Processing Treatment ............................................................178 Table 6.11: Real Imports, GDP, Growth, and Elasticities .........................................................................179 Table 6.12: Trajectory of Collection Efficiency ...........................................................................................181 Table 6.13: Vietnam’s Tariffs and Tariff Bindings by Economic Activity ...............................................183 Table 6.14: Regional Distribution of Manufacturing Activities and Average MFN and Bound Tariffs...186 Table 6.15: Summary Measures of Changes in ERPs .................................................................................188 Table 7.1: VAT Rates and Revenues by Regions .........................................................................................192 Table 7.2: VAT in Selected OECD Countries ..............................................................................................193 Table 7.3: VAT in Asian Countries .................................................................................................................194 Table 7.4: Production and Consumption Efficiency Ratios for VAT by Region ..................................194 Table 7.5: Productivity and Consumption Efficiency in Vietnam and Neighboring Countries .......216 Table 8.1: Revenue from Taxes and Fees, 1997–2008 ...............................................................................218 Table 8.2: Summary of Main Taxes in Vietnam ...........................................................................................221 Table 8.3: Sources of Household Income for Vietnam Households, by Income Per Capita Decile, 2006 ..222 Table 8.4: Breakdown of Household Expenditure by Vietnamese Households, ..................................223 Table 8.5: Breakdown of Household Expenditure and Income, by Region and Urban/Rural, 2006 ...224 Table 8.6: Spending Comparison (in VND ’000 per Household per Year, 2006)................................225 Table 8.7: Tax Paid by Vietnamese Households in 2006, by Expenditure Per Capita Deciles ..........226 Table 8.8: Tax Paid by Vietnamese Households in 2006, by Income Per Capita Deciles ...................227 Table 8.9: Tax Paid by Vietnamese Households in 2006, by Region and Urban/Rural Residence ..229 Table 8.10: Measures of Tax Progressivity ....................................................................................................230 7 Table 8.11: Reconciling Survey with Budget Data......................................................................................231 Table 8.12: Structure of Current and Previous Personal Income Taxes .................................................232 Table 8.13: The Distributional Effects of Revising the Personal Income Tax Law ...............................233 Table 8.14: Income and Estimated Tax for High-Income Households, 2005 .......................................235 Table 8.15: Estimating the Net Effect on Tax Revenue of Personal Income Tax Changes, 2006 ......236 Table 8.16: The Distributional Effects of Introducing a Tax on Nonagricultural Real Estate ............237 Table 8.17: Tax Paid by Vietnamese Households, by Region and Urban/Rural Residence ..............238 Table 9.1: Revenue from Land-Related Taxes (million VND).................................................................246 Figures Figure 01: Change in Tax Revenue Structure ................................................................................................18 Figure 02: Changes in Revenue Structure ......................................................................................................19 Figure 03: A Progressive Tax Burden ...............................................................................................................24 Figure 04: Gap in Operating Process ...............................................................................................................26 Figure 05: Tax Administration Modernization Objective ...........................................................................27 Figure 2.1: Determinants of Noncompliance and Revenue Leakage ........................................................61 Figure 2.2: Higher Corruption Associated with Greater Revenue and Tax Loss ...................................66 Figure 2.3: Negative Correlation between Size of Shadow Economy and Tax-to-GDP Ratio ............68 Figure 3.1: Staffing Proportion in Hearquarters ............................................................................................77 Figure 3.2: Operational Gap .............................................................................................................................. 78 Figure 3.3: Taxpayer Segmentations: Population vs. Revenue Contribution .........................................82 Figure 3.4: Tax Administration Modernization Objective and Indicators...............................................88 Figure 4.1a: Typical Taxpayer Model Structure ..........................................................................................121 Figure 4.1b: Aggregate Model Structure .......................................................................................................121 Figure 4.2: PIT Progressivity (wage income) ..............................................................................................123 Figure 4.3: PIT Progressivity (all sources of income) ................................................................................125 Figure 5.1: Total Provincial Revenue, by Revenue Sources ......................................................................136 Figure 5.2: Coefficient of Variation of Provincial Revenue, by Revenue Sources ................................136 Figure 5.3: Average of Provincial Revenue per 1,000 Population, by Revenue Sources .....................138 Figure 5.4: Coefficient of Variation of Provincial Revenue per 1,000 Population,...............................138 Figure 5.5: Share of Local Budget Revenue in Vietnam.............................................................................139 Figure 5.6: Share of Local Budget Revenue in Vietnam.............................................................................140 Figure 6.1: Evolution of Vietnam’s Tariff Schedule, 1995 to 2005, and Bound Rates to 2019 ..........175 Figure 6.2: Distribution of Actual and Bound Tariff Rates, 2005 to 2019 .............................................175 Figure 6.3: Potential and Realized Rates of Import Duty Collection .....................................................177 Figure 6.4: Nonpetroleum Import Revenue Scenarios ..............................................................................180 8 Figure 6.5: Petroleum Revenue Scenarios ....................................................................................................181 Figure 6.6: Distribution of Manufacturing Output and Employment by Average Actual and Bound MFN Tariffs, 2005 and 2019 ..............................................................................................................185 Figure 6.7: Changes in Effective Rates of Protection for Import Competing Production Associated with Implementation of Final Bound MFN Tariff Rates ...........................................................................187 Figure 6.8: Changes in Average Effective Rates of Protection Associated with Implementation of Final Bound MFN Tariff Rates........................................................................................................................188 Figure 6.9: Distribution of Industrial, Mining, and Agricultural Value Added in I/O Industries, by ERPa, 2005 and 2019...................................................................................................................................189 Figure 8.1: Incidence of Taxes by Expenditure Per Capita Deciles in Vietnam, 2006.........................227 Figure 8.2: Incidence of Taxes by Expenditure Per Capita Deciles in Vietnam, 2006.........................228 Figure 8.3: Checking the Robustness of the Incidence Results ...............................................................239 Boxes Box 01: Recommendations Based on Framework of a “Good Tax Systemâ€? for Vietnam .....................21 Box 02: Land Administration Reform and Property Tax ............................................................................22 Box 3.1: How Costly Is It to Comply with Tax Regulations?......................................................................80 Box 3.2: Developing a Compliance Management Strategy .........................................................................82 Box 3.3: Examples of Tax-Related Simplification Proposals as Part of Project 30..................................86 Box 3.4: Computerization of Tax Administration Operations ...................................................................89 Box 3.5: Key Elements of a Modern Taxpayer-Tax Administration Relationship ..................................96 Box 3.6: Service Pledge – Inland Revenue Authority of Singapore...........................................................97 Box 6.1: Exemptions from Import and Export Duties ..............................................................................164 Box 6.2: Other Issues in the Duty Payment System....................................................................................170 Annex Annex 1: Vietnam: Key Economic Indicators..............................................................................................273 Annex 2: State Budget Revenues (share of GDP) .......................................................................................274 Annex 3: Annual Revenue Buoyancy ............................................................................................................275 Annex 4: Features of a Good Tax System ......................................................................................................276 Annex 5: Documents Examined for Analyzing Major Tax Regimes in Vietnam ..................................278 Annex 6: Implications of Joining the WTO for Various Sectors of the Economy ................................280 Annex 7: Tax-Royalty Regime versus Production Sharing Arrangements .............................................282 Annex 8: Suggestions about Decentralization of Charges and Fees ........................................................284 9 Annex 9: Tenets of a Good Tax Administration ..........................................................................................285 Annex 10: Summary of Statistics of Variables Used in Empirical Analysis, 1992–2002 .....................287 Annex 11: Corruption and Revenue Leakage: Cross-country Estimates ..............................................288 Annex 12: Summary of Three Methodologies to Measure Shadow Economy .....................................291 Annex 13: Shadow Economy, Tax Collection, and Leakage .....................................................................292 Foreword Vietnam’s taxation system has undergone a fundamental reform in recent years. Tax policy has been modernized taking into account regional and global trends and the need to facilitate tax compliance of the rapidly growing private sector. The structure and transparency of the tax policy framework has improved in line with the principles of the “socialist-oriented market economy.â€? Cornerstones of this reform process were the creation of the necessary legal conditions for accession to the World Trade Organization, the overhaul of the Personal Income Tax system, moving from a taxation of high-income earners to the introduction of a universal income tax, and the adoption of a Natural Resource Tax Law to promote their efficient use and to ensure social equality in exploiting and using the natural resources of the country. At the same time, steps have also been taken to strengthen the administration of the tax system in parallel with tax policy reform. A fundamental reform to organize the structure of tax administration has been launched, key operational functions have been modernized, and taxpayer services and assistance have strengthened voluntary compliance with the tax system and reduced compliance costs. A backbone of the new phase of tax administration reform is the development and deployment of an Integrated Tax Administration Information System, based on an automation of reengineered business processes. It requires substantial investments in terms of training of tax officials and strengthening the integrity and accountability by undertaking organizational and procedural reforms. Results achieved so far have been encouraging. The performance of the revenue collection system has improved considerably, and net collections almost doubled between 2006 and 2010. It is expected that by 2015 a tax-to-GDP ratio of 23 to 24 percent will be achieved. The share of domestic nonoil revenues in total revenue is expected to increase further, reducing the dependency of the budget on revenues from natural resources. Tax policy reform remains a priority for Vietnam, and there are major reform challenges still to be addressed. The Value-Added Tax (VAT) system should further study the international best practice. There is no VAT threshold in place and two different VAT rates continue to apply. The direct tax system requires further strengthening to address newly emerging challenges, such as cross-border tax evasion, transfer pricing, and e-commerce transactions. The use of the tax system for ensuring better protection of the environment should be developed further. Reviewing the design and impact of the tax policy system in Vietnam to inform the future reform process thus is a critical necessity for the Government. The studies in this publication provide an important contribution to this review process. They were initially prepared for a workshop on tax policy reform organized by the Ministry of Finance in cooperation with the 11 World Bank in Hanoi in May 2008. The studies have subsequently been revised and updated. I am very pleased to introduce this important publication of the World Bank to a broader audience. Dr. Do Hoang Anh Tuan Vice Minister of Finance Vietnam 12 Acknowledgments This report was prepared by a team led by Duc Minh Pham and comprised of Gangadha Prasad Shukla, Michael Engelschalk, and Tuan Minh Le under the guidance of Martin Rama and Deepak Mishra. Viet Dinh Tuan of the World Bank was manager of the project. The individual chapters and the overall outline and content of the report were discussed with government representatives of Vietnam during a workshop held in Hanoi on May 29–30, 2008. All chapters were subsequently updated by the respective authors. Ms.Nguyen Thi Thanh Hoai of the Academy of Finance contributed significantly to the review and updating. The editors would like to thank experts from a large number of government and nongovernment institutions in Vietnam who provided invaluable comments and suggestions during the preparation of the report. Special thanks are due to Mr. Do Hoang Anh Tuan, Vice Minister of Finance; Mr. Vu Van Truong, Deputy Director General, General Department of Taxation; Mme. Le Hong Hai, Deputy Director General, General Department of Taxation; Mme. Nguyen Thi Cuc, Chairwoman of the Tax Consultancy Association; Mr. Nguyen Van Phung, Deputy Director, Tax Policy Department, Ministry of Finance; Mr. Le Van Ai, Academy of Finance; Mr. Pham Dinh Thi, Ministry of Finance; Mr. Quach Duc Phap, Institute of Finance, Ministry of Finance; Mr. Dang Hung Vo, Former Vice Minister of Ministry of Natural Resources and Environment; Mr. Vo Tri Thanh, Central Institute of Economic Management; Ms. Mai Thi Thu Van, Ministry of Finance; Mr. Bui Ba Cuong, General Department of Statistics; Mr. Nguyen Dinh Cung, Central Institute of Economic Management; and Mr. Hoang Huu Phe, Director General, VINACONEX R&D. The excellent editorial work of Diane Stramm is duly acknowledged. The project benefited from secretarial support provided by Ms. Linh Anh Thi Vu and Ms. Chi Kim Tran. Special acknowledgments go to Mr. Vikram Nehru and Ms. Victoria Kwakwa of the World Bank, who provided valuable support and guidance during the preparation process. The project also benefited from comments and suggestions received from the staff of the International Monetary Fund and from numerous World Bank colleagues. Their support is gratefully acknowledged. 13 Abbreviations AFTA ASEAN Free Trade Area BOT build-operate-transfer BT build-transfer BTA Bilateral Trade Agreement CIT Corporate Income Tax CPAs Certified Public Accountants EU European Union FDI foreign direct investment GDP gross domestic product GDT General Department of Taxation GSO General Statistics Office IMF International Monetary Fund IT information technology ITAIS Integrated Tax Administration Information System MFN most favored nation MIS Management Information System MoF Ministry of Finance MPI Ministry of Planning and Investment MTEF Medium Term Expenditure Framework OECD Organisation for Economic Co-operation and Development OOG Office of Government PIT Personal Income Tax PREM Poverty Reduction and Economic Management SCT Special Consumption Tax SEDP Socio-economic Development Plan SMEs small and medium enterprises SOE state-owned enterprise TAMP Tax Administration Modernization Project TIN Taxpayer Identification Number U.K. United Kingdom VAT Value-Added Tax VIES VAT information exchange system WCO World Customs Organization WTO World Trade Organization 15 Executive Summary OVERVIEW OF THE TAX SYSTEM IN VIETNAM By Martin Rama, Deepak Mishra, and Duc Minh Pham In 2010, after two decades of rapid economic growth, Vietnam passed the threshold to become a lower-middle-income economy. Sustained market-oriented reforms combined with intensive efforts to integrate into the world economy are among the key drivers of this achievement. The reform of tax policy and administration has been a vital part of this transition. This is leading to a fundamental change in the composition of taxpayers, from large state-owned enterprises (SOEs) and foreign-invested companies to a myriad of small and medium private enterprises. Economic transition is also leading to an equally important change in the sources of government revenue, away from cross-border trade-related taxes and revenue collection from crude oil toward a greater share of domestic tax revenue, in particular taxation of business profits, labor income, and capital gains on land. However, completing the transition to a market economy will require changes going beyond tax collection and administration procedures, and will involve changes to the tax instruments themselves. At the end of this process, Vietnam should have a set of taxes that is simple and transparent, secures a stable flow of revenues for the government, encourages an efficient allocation of resources, and does not risk constituting a source of inequality or unfairness. An ambitious strategic target has been set in the Socio-Economic Development Plan 2011– 2015 (SEDP). During the next five-year period, the state budget revenue is aimed at securing macroeconomic stability and investments in infrastructure. Average total revenue collection is expected to account for approximately 25.1 to 25.4 percent of gross domestic product (GDP), out of which the revenue from taxes and fees will account for 23 to 24 percent of GDP. Total revenue is expected to reach approximately 3,880 trillion dong, or one and a half times the 2006–10 level. This would represent an average annual growth rate of state budget revenues of approximate 15.6 percent per year. The purpose of the series of studies in this volume is to shed light on the issues Vietnam will be facing in the process of reforming its tax policy and administration. The studies are also expected to lead to concrete policy recommendations contributing to the preparation of key policies and legislative documents to ensure the achievement of the state budget revenue target and other tax administration reform targets in the SEDP 2011–2015. It is expected that the individual studies in this series will become useful inputs into the debate surrounding the issuance of new laws and regulations. It is also hoped that the volume will support the reform momentum in the tax policy area, leading to increased efficiency, transparency, and equity. A Tax System in Transition The first two major tax policy reforms undertaken in Vietnam in 1990 and 1998 were to adjust to the transition to a market economy. A comprehensive legal framework was developed to restructure the dated tax system in the centrally controlled economy. Taxes imposed on the state, nonstate, and agricultural sectors were integrated into major, more standard instruments, notably turnover and profit taxes. Subsequently, these were transformed into a Value-Added 17 TA X R E F O R M I N V I E T N A M Tax (VAT) and a Corporate Income Tax (CIT), respectively. The tax structure has evolved further to include natural resource taxes; Personal Income Tax (PIT); excises; customs duties; and a number of minor taxes, fees, and charges. Since then, the structural change in the Vietnamese tax system is characterized by three main features: (i) reduced dependence on oil revenue and tariff revenue, (ii) a reduced share of the state sector and an increased share of the nonstate sector in total revenue, and (iii) an increased importance of the VAT. Revenue from tariffs is diminishing in importance given a rapid global integration. Vietnam signed a bilateral trade agreement with the United States in 2001, participated in the ASEAN Free Trade Area in 2006, and obtained World Trade Organization (WTO) membership in 2007. During this process, tariff rates have been reduced and nontariff barriers and trading right restrictions have been phasing out. More recently, this decline in tax revenue has been more than compensated by the solid performance of other tax instruments, in particular by a highly “productiveâ€? VAT (figure 01). Import and export duties still accounted for 19 percent of total tax revenue in 2001, when the reform process accelerated. They fell to 11.4 percent in 2006 but recovered to 15.6 percent in 2010 owing to the significant increase in volume of international trade in recent years (figure 02). In addition, increasing global integration is leading to a higher ratio of imports to GDP, so the reduction in tariff rates has been partly offset by the increase in the tax base. Figure 01: Change in Tax Revenue Structure Agricultural tax PIT 2000: 2.2%; 2010: 2000: 2.4%; 2010: 0.01% 5.1% Excise 2000: 6.7%; 2010: CIT 7.8% 2000: 37.2%; 2010: 32.4% Natural Resources Tax 2000: 9.6%; 2010:5.7% 2000 Export & Import Tax 2000: 17.2%; 2010: 15.6% 2010 VAT 2000: 21.9%; 2010: 32.4% Source: Based on data from the MoF. The Vietnamese tax system in the current transition has the capacity to guarantee a sustained flow of revenues for financing the government-increased recurrent expenditure and investments. Despite the decline in trade-related proceeds, total average tax revenue was 22.2 percent of GDP for 2001–09, a favourable comparison with regional neighbours. This record has been achieved despite a consistent underestimation of tax revenue due partly to poor forecasting capacity. Buoyant revenues are partly the result of high prices of oil in international markets. However, they also stem from the operation of an extremely efficient VAT. Tax “productivity,â€? defined as percentage points of GDP in revenue divided by percentage points of the basic tax rate, is a standard performance indicator, in particular for 18 E XECU TIVE SUMM ARY the VAT. Vietnam’s VAT productivity is close to 0.6, which exceeds the average for industrial countries. Some cascading-inducing features of the Vietnamese VAT may contribute to this exceptional performance. But there is more than cascading involved, which explains why the VAT has become one of the workhorses of the Vietnamese tax system. The other workhorse of the system is the CIT. Together with the VAT, this has been a dominant source of revenue, with each of the two instruments contributing slightly less than 6 percent of GDP. Oil revenue still constitutes the major source of revenue, despite the recent reduction of its share in total revenue. Oil revenue as a share of total revenue decreased from 29.2 percent in 2005 (a peak year in the last decade) to 13.4 percent in 2010 (figure 02, panel A). Meanwhile, nonoil revenues increased steadily from 50.9 percent of total revenue in 2000 to 55.1 percent in 2008. The structure of nonoil revenues has changed over time, as well. The importance of SOEs in the economy has been decreasing steadily as Vietnam moves to a market economy, and the trend is bound to continue. In contrast, the non-state sector has increased in importance. The emergence of a vibrant private sector is changing the potential number and composition of taxpayers. At present, some 500,000 enterprises and organizations, and about 1.7 million business households and 9.0 million individuals, are registered to pay. These numbers are expected to increase rapidly. The expectation is that the number of enterprises will increase by close to 50 percent within five years. The number of “smallâ€? taxpayers could reach at least 2.3 million by 2012. Although tax revenues from all economic sectors have increased in absolute terms over the past decade, as a result of their declining importance in the economy the share in total budget revenues from the state-owned sector has decreased steadily while revenues from the private sector and the foreign-invested sector increased (figure 02, panel B). Figure 02: Changes in Revenue Structure Panel A Panel B A collection priority shi ing from the … and focusing more on border to the domestic market, and emerging SMEs reducing dependence on oil revenue … 35.0 25.00 30.0 State-Owned 20.00 25.0 Oil 20.0 15.00 15.0 Foreign 10.00 10.0 Tari Nonstate 5.00 5.0 0.0 0.00 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009/e2010/e 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009/e Oil Revenue Tari Revenue State-Owned Sector Foreign Invested Sector Nonstate Sector Source: Based on data from MOF and GSO. 19 TA X R E F O R M I N V I E T N A M Simplifying the Tax Instrument and Reducing the Compliance Cost Further simplification of existing tax instruments could result in improved economic efficiency. In the case of the VAT, reducing the number of tax rates (10, 5, and zero percent at present) and moving to a simpler structure with only one positive rate of 10 percent in addition to the zero rate would be in the right direction. While the current rate differentiation may generate modest equity gains, it creates distortions and opportunities for misclassification of taxable transactions. A single positive rate lowers the cost of compliance by reducing the requirements for keeping records and invoices and offering a potential for simplifying tax forms. Similarly, a large number of exemptions (26 at present) create administrative and compliance problems for traders who sell both taxed and exempt items. More critically, exemptions risk breaking the VAT chain and may lead to cascading, as in the case of a turnover tax. The current set of exemptions should be scrutinized and, ideally, only a few of them should be retained. Other simplifications may be considered in the case of the VAT. The current practice of having no registration threshold and thus forcing small traders into the VAT system does not follow international good practice. A suitable threshold could be put in place taking into account the tradeoff between the cost of administration and revenue implications. This would help the administration focus on larger taxpayers, while applying a simplified tax regime to traders below the threshold. The VAT refund system should be simplified to lower compliance costs and to alleviate cash flow problems facing businesses. The CIT currently has multiple rates with a standard rate of 25 percent; a higher rate between 32 and 50 percent stipulated for oil, gas, and other rare natural resources; and reduced tax rates of 20 and 10 percent. This makes the administration complicated. In addition, the current CIT regime includes complex depreciation rules involving 53 types of assets and three possible schedules. To implement the current regime, each fixed asset must be classified and maintained in a separate file and also recorded in an assets record book, numbered, and assigned a separate card with relevant statistics. A better approach would be to place all assets in four or five broad categories and then to depreciate the pool using a declining balance method. When an asset is added, the pool becomes larger; when an asset is sold, the pool becomes smaller. The incentive regime embedded in the Vietnamese CIT is also complex and may not achieve its intended objectives. The regime is aimed at encouraging capital accumulation in technologically dynamic sectors and in poorer regions. These are legitimate development objectives, and more can be done to foster them under the current regime. For instance, deductions could be allowed for nonwage benefits for personnel. But the current regime should also be rationalized. At present, there are three major types of incentives: (a) access to preferential or reduced rates, (b) outright tax exemption or tax holidays for a prescribed period of time, and (c) accelerated depreciation. Over 300 rules define eligibility and terms of application. But many of those rules overlap and some are contradictory. In addition, the system is only partially effective, with business surveys revealing that a majority of beneficiary firms would have proceeded with their investments in the same form and location even if the incentives did not exist. The PIT could benefit from simplification, as well. At present, there are seven income brackets, while four brackets (as in the old regime) would probably be sufficient to ensure appropriate progressivity of the tax. International experience suggests that having too many brackets 20 E XECU TIVE SUMM ARY complicates administration and encourages tax evasion while doing little to promote equity. Importantly, these brackets are not indexed to inflation, which may result in an unplanned increase in the already high number of potential taxpayers. Given that business households will be migrating from the CIT to the PIT, the highest tax rate for the latter should be similar to the CIT tax rate. At 35 percent, the current VAT is much higher. Despite its overall strong performance, the Vietnamese tax system is still unbalanced. Revenue from some tax instruments is almost negligible, at less than 1 percent of total revenue. Effort made could help include the large set of minor taxes, charges, and fees but would result in unnecessary complexity. Tax instruments with a strong potential to influence the allocation of resources and reduce waste are missing. And the main instrument being introduced at present, the universal PIT, will not yield high revenue for some time but could distort incentives if it is not phased in carefully. Current efforts to modernize the tax system are focused on the introduction of a broader PIT. But the challenge is that PIT compliance is typically low in developing countries, and an attempt to expand the tax base too rapidly could penalize the formalization of employment, thus hindering modernization and job creation. See box 01 for a discussion of what constitutes a “good tax system.â€? Box 01: Recommendations Based on Framework of a “Good Tax Systemâ€? for Vietnam Economic efficiency. At present, there are several non-neutralities both with respect to tax policies and their implementation. A lack of harmonization among taxes, a multiplicity of rates, and a large number of exemptions and tax incentives give rise to a great deal of inefficiency in resource allocation because taxpayers have incentives to base their business decisions on minimizing their tax burden instead of optimizing output and production. Equity. There are numerous violations of horizontal equity. For instance, taxpayers subject to personal income tax are taxed at different rates depending on the type of income earned. The numerous exemptions also lead to a violation of vertical equity. The same thing is true for indirect taxes with multiple VAT and excise rates and many exemptions. Transparency. Transparency primarily refers to tax laws that are not overly complex and discriminatory and tax regulations and administrative requirements that are easy to access and comply with. From this perspective, Vietnam has made significant strides in achieving transparency, although a great deal of reform is yet to be implemented. For example, in its present form, the VAT has zero plus two other rates and a long list of exemptions. Excise taxes have many different tax rates and exemptions. The corporate income tax has several tax rates and a series of incentives related to the nature of industries and their location that leave effective tax rates different from their published rates. Revenue adequacy and stability. While lack of adequate data precludes the estimation of tax elasticities, volatile buoyancy points to the need for improvement in several areas of the tax system. The tax base is narrow and needs to be diversified and broadened. It is also important to monitor growing sectors of the economy with a view to bringing them into the tax net at the right time. This includes commodity sectors with high income elasticity of demand. International experience shows that a broad-based and simple tax system is more elastic and buoyant. Thus, there is a need to examine the tax laws with a view to simplifying and rationalizing them. Finally, it may be helpful to estimate the extent of current tax expenditures and examine their rationale and justification. 21 TA X R E F O R M I N V I E T N A M In addition, the number of potential taxpayers is growing much faster than current administrative capacity can handle. The mushroom growth of small and medium enterprises (SMEs) suggests that the “missing middleâ€? of the size distribution pyramid could be populated quite rapidly. Furthermore, taxing a much larger number of domestic private businesses as they grow into sizable enterprises is bound to be challenging. Efforts have been made to shift the responsibility for collection of these taxes to local governments, but decentralization creates additional accounting and supervision problems. Taxpayers complain about ambiguities and contradictory provisions in the tax system, which increase the compliance burden and the discretionary powers of tax inspectors. Revenue forecasting capacity remains weak. Projections for the following year are primarily based on the growth rate of revenues during the current year, adjusted after consultations with line ministries and the General Department of Taxation (GDT) field offices. No econometric or micro-simulation modeling tools are used, and the quality of the data is poor. To address these concerns, a third wave of reforms to tax policy and tax administration is underway. At present, natural resources may be placed in four categories for the purpose of taxation. Taxing natural resources has the potential to increase economic efficiency and reduce environmental damage. Minerals are only subject to royalties. Natural forest products are also subject to royalties, with rather high rates up to 35 percent applying. While high rates aim to protect forest resources, they may also encourage the exploitation of high-value natural wealth and create an incentive for evasion, thus negating their purpose. Natural aquatic resources and natural water are subject to a combination of CIT and a low rate of royalty. Most items in this category are taxed at 1 to 5 percent, except for pearls and sea slugs, which are taxed at 6 to 10 percent. A similar combination applies to oil and natural gas, although in this case CIT rates vary (from 6 to 40 percent) depending on the amount and depth of extraction. The combination of CIT and royalty is a sensible model, and it should be generalized to all types of natural resources. The introduction of CIT in the case of minerals, and especially of natural forest products, should allow reducing royalty rates. While the aim of such high rates is to protect natural resources such as forests, in practice they tend to promote the exploitation of high-value natural wealth and foster tax evasion. Royalty rates should be reduced in number and should not be kept at an abnormally high level. This would also help reduce administrative and compliance costs. However, the CIT rate should not be reduced below 20 or 25 percent - not even in the case of the oil and natural gas sectors. Property tax reform is another challenge which currently is being addressed. Box 02 summarizes key issues in the land administration reform in relation to property tax reform. Box 02: Land Administration Reform and Property Tax Taxing land can potentially increase economic efficiency and improve equity. In 2010, the associated revenue was 2 percent of GDP and hardly sustainable given that 87 percent of the actual revenue came from land-use right allocation levies. In addition to broadening the tax base, betterment levies in the form of up-front lump-sum payments from the beneficiaries can make property tax become a useful instrument to improve public services, because it provides sub- national governments with resources to invest in local infrastructure. By taxing income derived 22 E XECU TIVE SUMM ARY from property, this instrument can also lead to significant improvements in the quality of land use. Taxing property can also help contain land speculation, one reason for the real estate bubble, and thus contribute to improved land administration and equity. Property taxation may result in a situation where poor people are unable to pay. Exempting subsistence farms from property taxes and introducing a minimum imposable threshold are obvious solutions to poor people owning land of little value. The problem of poor people with land of significant value (particularly in cities) is that they may be reluctant to move to cheaper land for understandable social reasons, especially in the case of the elderly. A possible way to address this concern is to allow any poor person to defer his or her taxes (plus interest) until death, and then recover the debt from the value of the property. There is another concern that taxing property may be complex and costly compared to the revenues it may potentially generate. A fundamental challenge is property revaluation on a regular basis. Systematic property assessments conducted by area can bring the cost down. The self-assessment of property prices is another practical solution. Avoidance risks can be mitigated through the provision that the self-assessed value is what will be paid as compensation by government if the property is reclaimed for the purpose of infrastructure development. However, a more serious difficulty in Vietnam is the incomplete issuance of land-use right certificates, especially in urban and upland areas. Household survey data from 2006 suggest that only 76 percent of agricultural land parcels, 68 percent of urban land parcels, and 34 percent of forestland parcels have been granted land-use right certificates. In practice, this means that about two-thirds of the total land parcels of Vietnam still lack a certificate. Despite the promulgation of the Land Law in 2003, the main reason for this problem is the shortage of basic infrastructure for an effective operation of land administration, including cadastral mapping, transaction registrations, and record management to the provision of land administration services. The system itself is cumbersome, not transparent, and inefficient, and does not provide the quality services that end users can rely on. As a result, land administration is generally recognized as one of the most severe constraints to business development and transparent governance in Vietnam. Combined with limited local capacity, this calls for a step-by-step approach in introducing this new tax instrument. Source: Based on Vietnam Development Report 2009. Toward an Equitable Tax System Taxes and fees are instruments not only to generate the state budget revenues but also to introduce a wedge between gross and net income, and thus redistribute resources across households. In many cases, such redistribution, if accomplished through a variety of exemptions, thresholds, and differential tax rates would lead to increased complexity of the tax system that would compromise the redistribution impact. In Vietnam, households are directly subject to several taxes and fees. As consumers, they pay the VAT and excise taxes. As residents, they face a range of local fees such as agricultural taxes, education fees, and taxes on household businesses, among others. All these taxes are progressive. In 2006, a household in the richest quintile of the population devoted 14.8 percent of its spending to paying taxes and fees compared to 8.7 percent for a household in the poorest quintile. But not all taxes and fees are equally progressive. The VAT, excise taxes, and taxes on household businesses are, while agricultural fees are not. And not all households pay all taxes and fees. For instance, 23 TA X R E F O R M I N V I E T N A M mostly everybody is confronted with the VAT, whereas a meager 0.1 percent of households pay the PIT. A study using household survey data from 2006 shows that the Vietnamese tax system also does well in terms of equity. Richer groups, whether measured by income per person or by expenditure per person, face a proportionately higher tax burden (figure 03). Compared to results of a similar study for 1998, the tax burden is now heavier, and it is more progressive. The 1998 study found that taxes represented about 8 percent of expenditures for all but the top quintile, where they stood at just over 10 percent. The figures for 2006 show a tax burden of about 8 percent only for the bottom two population deciles, rising steadily to reach a plateau of about 14 percent in the four top deciles. This higher, more progressive, tax structure is due largely to the introduction of the VAT in 1999, when it replaced a complex system of turnover taxes. The VAT is not progressive in every country. But in Vietnam poor people provide for many of their own needs rather than purchasing them, which in turn reduces the VAT they must pay. Over time this will change, as home production becomes less important, making the tax system gradually less progressive. Figure 03: A Progressive Tax Burden 16 14 12 Tax burden 10 8 6 4 2 0 1 2 3 4 5 6 7 8 9 10 Population decile (from poor to rich) Percent of income Percent of expenditure Source: Jonathan Haughton 2008. Building Modern Tax Administration With almost 44,000 staff on its payroll, GDT is one of the largest public sector institutions in Vietnam. It is organized into three levels with the tax administration headquarters in Hanoi, 63 tax offices at the provincial level, and 694 tax offices at the district level. GDT is charged with collecting all major domestic taxes, in particular the VAT, the CIT, excise taxes, and natural resource taxes, which together account for more than two-thirds of total domestic revenue collection. GDT is also responsible for the collection of a relatively large number of minor taxes, such as the tax on the transfer of land-use rights, the land and housing tax, and the small business tax. 24 E XECU TIVE SUMM ARY The tax reform program adopted by the government confronts GDT with a series of new challenges. Shifting collection priorities from the border to the domestic market, reducing dependence on oil revenues, focusing more on private sector enterprises and less on SOEs and foreign-invested companies, adjusting to self-assessments as opposed to inspections, introducing modern taxpayer service processes, and adapting to the electronic formats increasingly used by taxpayers to keep their records are among the major issues to address. The Tax Administration Law that became effective in July 2007 was an important step toward addressing these challenges. The law provides a foundation for ensuring transparency and integrity in all tax administration operations. In addition, it eliminates gray areas potentially fostering noncompliance by taxpayers and abuse by tax administration officials. Among several important innovations, the law introduces a unique Taxpayer Identification Number and stipulates that third-party information can be acquired from various agencies for effective audit and enforcement, while safeguarding its confidentiality. But there are some weaknesses, too. While the law stipulates the general rights and responsibilities of the tax administration agency and other related sub-national organizations, it does not include provisions on the division of responsibilities between various levels of tax administration, or on the division of responsibilities between the tax administration agency and People’s Councils and People’s Committees at various sub-national levels. An assessment of GDT’s operational processes was conducted by reviewing 15 functional aspects of the tax administration performance (figure 04). Overall performance is below average in all main and support functions scored in accordance with an international best practices research scale, with supporting data for the framework gathered from 30 developing and developed countries’ tax administrations. Among them, the debt collection and enforcement function turned out to be the area with the most advanced processes in place. Multiple payment methods (cash and bank transfer) and the use of technology to access taxpayer account data combined with semi-integrated collection systems to monitor taxpayer delinquency are among key factors contributing to this advanced function. In addition, provisions for the counseling and advising of GDT staff on legal issues exist, so that at least basic support is available to help tax administration staff deal with more complex legal matters. However, the legal support is not yet comprehensive and additional processes to perform critical legal tasks, such as creating the capacity to issue private letter rulings to taxpayers, still have to be developed. Existing processes are less satisfactory in most other areas, with the tax audit function being an area of particular concern. Procedures and methodologies for a risk-based audit selection have not yet been developed, and auditor discretion in the selection of cases for a desk or field audit is substantial. In addition, existing processes for the conduct of an audit provide substantial discretion to the audit team, and the overall audit planning and supervision process does not yet include a systematic ongoing monitoring of audit efficiency. Anti-evasion appears to be the weakest area, where no strategic plan for addressing evasion cases exists. 25 TA X R E F O R M I N V I E T N A M Figure 04: Gap in Operating Process Pre-Filing Taxpayer Services Support Functions (Financian and 5 Filing - Return Human Resources Processing, and Tax 4 Accounting Function Management, and IT 3 Operations 2 Mangement-Strategic 1 Audit Function Planning, Perf Mgmt, QA, Communications 0 Legal A airs (Part of Debt Collection & Collection & Enforcement Function Enforcement Function Anti-evasion Appeals (Activities (Activities Conducted Conducted by Internal by Taxpayer Audit Audit Function) Source: Based on Booz Allen Hamilton 2007. A taxpayer baseline survey conducted in 2007 and 2008 provides some information on how GDT is doing so far. On the positive side, 61.3 percent of the taxpayers interviewed considered the services to the public provided by tax offices to be above average. However, only 3.9 percent of them rate these services as very good, suggesting that there is still scope for improvement. The taxpayer baseline survey also reveals that taxpayers struggle to understand major components of the CIT legislation, such as preferential regimes or the determination of deductible expenses. Less than 40 percent of respondents found the VAT refund system easy to understand and implement. And more than 60 percent reported finding out about changes in the laws and procedures less than a month prior to their introduction and enforcement. This highlights the need to improve taxpayer information and communication. A 10-year tax system reform strategy has been drafted to set a target that taxes and fees (excluding revenue from crude oil and export revenue) should constitute 70 to 75 percent of the total state budget revenue by 2015, and 75 to 80 percent by 2020, comparing favorably with the current level of 62.5 percent in 2010. An ambitious program supported by the World Bank to modernize the tax administration is underway to improve governance in tax administration and raise the level of volunteer compliance with the tax system. Through this program, government resources, combined with financial and technical assistance from the donor community, will be mobilized to implement a comprehensive reform plan to improve the business climate, sustain revenue collection, and reduce corruption. The Tax Administration Modernization Project was prepared in 2007. It is considered to be the main driver to achieve the target that, by 2015, Vietnam’s tax agency will rank in the top five in the East Asia region. The project aims at assisting GDT to strengthen its governance in tax administration and to increase the level of voluntary compliance with the tax system by improving the effectiveness, efficiency, transparency, and accountability of the tax administration. These reforms are expected to have a positive impact on the 26 E XECU TIVE SUMM ARY business climate by developing predictable and enforceable tax administration processes and procedures. The sustainability of revenue collection and greater equity will also be supported through implementation of this project. The project’s development objectives, therefore, are complementary to two higher-level objectives of the country’s four reform pillars, namely business-led development and strengthened governance systems. The development and deployment of an Integrated Tax Administration Information System based on reengineered business processes reflecting international best practice is assumed to be the backbone of the five-year plan and 10-year strategy (figure 05). Figure 05: Tax Administration Modernization Objective Key Indicators Project Development Objectives Key Indicators Strengthened Accountability Introduction of Taxpayer Increased Transparency Service Culture Improved Increased Reduced Corruption Reduced Compliance Burden Administration Voluntary Improved Interaction with Capacity Enhancement of Taxation Compliance Stakeholders Improved Interaction with Improved Compliance with Stakeholders Legal Systerm Improved Compliance with Enhanced E ciency Legal System Improved Business Climate Enhanced E ciency Sustain able Growth in Revenue Collection Reduced Corruption Higher Development Objectives Business-led Strengthened Development State Governance Source: Based on the World Bank-funded Tax Administration Modernization Project Document. This comprehensive modernization program is not risk free. A reform program that intends to reduce the opportunity for part of the tax officers to charge discretionary rents might be seriously resisted. Implementing an ambitious multifaceted reform agenda, with interdependence between a complex computerization component and other required activities might lead to delays in meeting the targets of the reform agenda, considering in particular the weak implementation capacity of the administration. Five factors, therefore, are considered critical for the sustainability of the reforms: • Strong government ownership of, and a high level of commitment to, its reform agenda, including the Tax System Strategy for 2011-20 • Mobilization of private sector support and partnership • Sound and quantifiable performance indicators and client standards introduced, so that any slippage in performance will be easily identifiable and tax officials can be held accountable • Development of a sustainable, in-house capacity for continuous improvement following full implementation of the Reform Plan 27 TA X R E F O R M I N V I E T N A M • Application of a modern Management Information System as a key business strategy to radically streamline operations by simplifying procedures, increasing transparency, removing opportunities for the unlawful use of official discretion, and introducing internationally agreed standards. Such changes, once implemented, will be extremely difficult, if not impossible, to reverse. A Summary of the Volume This volume compiles a number of studies prepared under a World Bank technical assistance program that shed light on the issues Vietnam will be facing in the process of reforming its tax administration and tax policies. The studies also contain concrete policy recommendations that can illuminate discussions on the preparation of key laws and regulations in this area. The studies in this volume fall into two groups, one to discus cross-cutting themes and another to deal with the design and impact of specific taxes. This summary highlights and integrates the key findings of these studies. The legal framework for tax administration reform is set in Chapter 1. The chapter provides an overview of the revenue system in the context of Vietnam’s ongoing reforms and integration into the global economy. It summarizes the main features of direct and indirect taxes and identifies issues in their administration in light of international best practices and experience. This review leads to the conclusion that the latest tax reforms have moved Vietnam forward toward a system of tax administration that supports the country’s reform and that is converging with international norms. Continuing improvements are necessary to iron out the remaining kinks, however, but this can only be accomplished as an integral part of a long-term reform program; ad-hoc adjustments will likely create more inconsistencies. The costs to households and enterprises of complying with existing taxes are the subject of Chapter 2. Poor compliance has detrimental effects not only on government revenue but also on the country’s economic development. Therefore, understanding these costs and who bears them is key to the design of an effective and equitable tax system and its administration. The chapter provides a conceptual framework for measuring tax compliance and its associate costs and critiques Vietnam’s existing approaches to measurement. In assessing the cost of compliance for households and businesses, it finds that international experience is a valuable guide to what Vietnam needs to do as part of overall tax administration reform. In particular: • It must improve the General Department of Taxation’s analytical capability in this area • It must reform tax policies • Tax reform should be implemented in parallel with public sector reform • Focusing on a group of large taxpayers should be complemented by improving compliance among SMEs. The rapidly changing tax administration environment in Vietnam creates substantial challenges for compliance management and revenue collection. This is occurring against the backdrop of new types of business transactions, the emergence of a dynamic SME sector, and Vietnam’s integration into the world economy within the WTO framework. The reform of these existing tax structures must take into account these developments. This reform, undertaken under the Tax Administration Modernization Project, is the subject of Chapter 3. 28 E XECU TIVE SUMM ARY By identifying the key private and public stakeholders and the role they play in defining the strategy and supporting the implementation of reforms, the chapter outlines the elements of a comprehensive strategy to improve voluntary tax compliance that reduces compliance costs while ensuring integrity and transparency in tax administration. In assessing the efficacy of these reforms to meet these objectives, the chapter highlights the challenges faced and concludes that organizational changes in the direction of reengineering business processes and strengthening of the tax administration headquarters and the establishment of specialized administrative structures for the internal audit, the appeals, and the compliance monitoring functions need to be supported by enhanced information technology support. Reforms in human resource management with regard to training and capacity building to enable tax administration staff to perform highly specialized tasks in a function-based organizational structure, as well as workforce planning and workload analysis, and the design of appropriate incentives to reward high performance, are also recommended. As with the preparation of government budgets, effective tax administration requires accurate forecasting of tax revenues. As with most developing countries, both statistical data and institutional capacity to make these forecasts need improvement. This is one task of the tax administration reform and the theme of Chapter 4. The chapter provides an overview of the forecasting tools and methods currently in use and lays out the options available, including specialized modeling tools like input-output-based modeling for the VAT. Some of these methods and tools are applied to illustrate forecast of key taxes such as the VAT, the CIT, and the PIT. Findings from these exercises are used to show how they can be used to assist policy makers in making appropriate decisions. These findings are, in the sequence for implementation: • The need to strengthen the tax database that GDT can access • Training of personnel in relevant economic theory and quantitative techniques • Model building • Proper sequencing according to an appropriate time frame of up to 24 months. Once tax revenues are received, how they are allocated becomes important. This importance is enhanced in the context of tax reform. Lack of coordination between tax reform and assignment can lead to processes working at cross purposes. The most important objectives of tax reform—simplicity, low economic distortions or efficiency, and equity in the distribution of the tax burden—are less vital in decentralization reform, where revenue adequacy and local government accountability take center stage. Given Vietnam’s drive toward fiscal decentralization, budget allocation norms have been established for recurrent and capital expenditures. These norms have been translated into rules for tax revenue assignment. In Chapter 5, an overview of tax assignment to and revenues derived from sub-national levels reveals arrangements that are ad hoc. These arrangements are then assessed in terms of their contribution to the accountability of local officials and their compatibility with the tax administration reform strategy. This chapter concludes with proposals to strengthen arrangements for tax revenue sharing. These relate to strengthening revenue adequacy at the sub-national level, improvement on the apportionment of tax revenues between the central and the provincial governments, and the introduction of own revenue sources for local 29 TA X R E F O R M I N V I E T N A M governments. In terms of tax administration, the recommendation is for tax collection to continue to be centralized. Beginning with Chapter 6, the second part of this volume looks at the administration of key taxes. The internationalization of Vietnam’s economy that culminated in WTO accession has implications for a range of tax instruments and hence their administration. Chapter 6 examines these implications by category of tax instruments and over time, up until the final phase of the WTO implementation timeline of 2012. The categories covered include import taxes, export taxes, customs fees, and the adoption of the Customs Valuation Agreement (Article VII of the General Agreement on Tariffs and Trade). Additional areas include domestic consumption, business revenue tax, and even the provision of tax consulting services. Previous work undertaken in 2004 and 2006 for the World Bank and the Asian Development Bank are referenced. Specific issues - the offsetting impact of a larger trade volume against a lower tax rate, and the potential impact on nontrade-related revenue - are also discussed. This chapter has several major findings: • Implementing Vietnam’s tariff binding commitments will reduce the average and dispersion of tariffs, with a consequent improvement in the incentives for investment resource use. • Although Vietnam’s phasing-in period is long, much of the final bound rates will be in force after seven years. • The tariff reduction will likely result in revenue losses, but these can be largely offset by reducing concessions and exemptions and improving administration efficiency. The VAT, the subject of Chapter 7, is a centerpiece of any domestic tax system, especially in the face of global integration. This is equally true in the case of Vietnam, where it will be increasingly relied upon to offset the likely loss of trade-related revenues in the medium term and to remove tax-induced burdens on enterprises that can undermine their competitiveness. The chapter undertakes an analytical review of the VAT regime in Vietnam and its evolution over time. This review underpins the discussion of key issues related to the current structure of the VAT regime and its implications on equity, efficiency, and the treatment of exporters. These implications are compared with those for VAT systems in other countries in the region. Recommendations for reform in both the system and its administration, including improving cooperation between the tax and customs administrations, conclude this chapter. As Vietnam’s economy grows, the taxation of personal income becomes an increasingly important source of tax revenues. However, such taxation evokes considerable debate because in a developing country context, the PIT can easily become a tax on formalization. In addition, it is complex to administer and its potential in terms of revenue is hitherto still limited. A key input to inform this debate is to understand the sources of personal income and types of expenditure and the incidence of tax (personal income and other taxes) on them. This is the theme of Chapter 8, which summarizes the existing tax structure and the structure of household income and expenditure before proceeding to estimate tax incidence. An important overall finding is the progressivity of the tax structure. The chapter also analyzes the likely impact of the new structure under the 2007 Personal Income Tax Law on revenues and incidence. Simulations using two data sets, however, yielded opposite results—one 30 E XECU TIVE SUMM ARY with higher revenue and the other lower. Finally, the chapter looks at the likely impact of a property tax on nonagricultural real estate, which Vietnam does not have, and comes to the conclusion that it can lead to substantial yields and be very progressive. Chapter 9, the final chapter, provides observations on property and land taxes. Despite a modest contribution to tax revenue, land taxes help broaden the tax base, potentially increase economic efficiency, and improve equity. The property tax becomes an increasingly useful instrument to provide local governments with resources to invest in local infrastructure. Taxing property may be complex and costly compared to the revenue it might generate. The chapter suggests a distinction between the “landâ€? and the “capital improvementâ€? components of property taxes, given the different elasticity of their supply. A fundamental challenge is that property should be reevaluated regularly. Ineffective operation of land administration is one of the most severe constraints to business development and transparent governance in Vietnam. 31 The Legal Framework for Tax Administration I By Duc Minh Pham, Tuan Minh Le, and Gangadhar Prasad Shukla 1. Summary Vietnam is currently following a course of economic integration with the world economy. To facilitate this process, the government has undertaken a series of trade and tax policy reforms and has improved its fiscal management. As a result, the country has enjoyed spectacular growth during the past two decades, and most economic indicators have shown a very robust trend. Inflation has been generally under control, the trade deficit mild, and the exchange rate fairly stable. The budget performance has also been encouraging. The growth in revenues has been impressive, while expenditures have been contained, thus keeping the budget deficits within manageable limits. The country has also been taking steps to reform its tax laws and modernize its tax administration. The tax system has undergone three phases of reform in the past two decades, and the third and current phase of reform has been particularly motivated by the increasing integration of Vietnam into the world economy and its effort to join the World Trade Organization (WTO). The existing tax regime in Vietnam is analyzed below under the framework of a “Good Tax System,â€? that is, equity, economic efficiency, stability, and technical efficiency. Value-Added Tax (VAT) The VAT has proved to be the main workhorse for raising tax revenues in Vietnam. There are three VAT rates: zero percent mostly for exports, a 5 percent tax for 15 categories of goods and services, and 10 percent (standard rate) for the remaining items. Twenty-six goods and services are VAT exempt. At present, there is no threshold based on turnover for exempting small producers and traders from the VAT net. The credit method is used in parallel with the subtraction method, which is applied exclusively to small traders that fail to fully observe regulations on accounting, invoices, and documents, and to gold, silver, and gem-trading activities. The present VAT refund rules in Vietnam are complicated and vary with the nature and activity of the business. Vietnam should reform its VAT regime as follows: i. Simplify the rate structure with a zero rate and one positive rate of 10 percent. ii. The list of exempt items needs to be closely scrutinized and only a few of the 26 items should remain exempt. iii. The practice of having a zero threshold of exempting small traders needs to be discontinued and a suitable threshold put in place based on the tradeoff between the cost of administration and revenue implications. A simplified tax regime can be applied to all traders below the threshold. 33 TA X R E F O R M I N V I E T N A M iv. The credit method should be adopted across-the-board, and the subtraction method applied to small traders must be abolished when the exemption threshold is established. v. The VAT refund system needs to be rationalized and simplified to lower collection (administration and compliance) costs and to alleviate cash flow problems for businesses. Excises or Special Consumption Tax (SCT) The tax base consists primarily of goods whose consumption is regarded as socially and environmentally undesirable, and luxury goods. The tax rates on goods are highly dispersed, varying between 10 percent (gasoline; air conditioners of 90,000 BTU or less; and electrically operated passenger cars of between 16 seats and under 24 seats) to 70 percent (votive gilt papers and votive objects) and on services between 15 percent (lottery business) to 40 percent (dance halls). The tax on gasoline, however, appears to be low compared to neighboring countries. The excise regime should be reformed as follows: i. Tax rates are quite dispersed. Some top excise rates should be lowered while some low rates may be increased. ii. In the short run, the government may retain the low excise rate on gasoline, due to the increase in the international price and high inflation. However, over the medium to long term, the rate may be increased to the level comparable with neighboring countries. iii. A cost-benefit analysis of levying an excise tax on some of the entertainment items may be conducted to compare the cost of administration and compliance with the revenues collected. iv. The provision of tax reductions and exemptions for establishments facing financial difficulty should be eliminated. v. Expansion of the tax base under excises may be considered by bringing into the tax net some additional luxury items primarily consumed by higher-income groups. Trade Taxes The tariff rates could be put into four categories: (i) preferential rates applied to goods originating from countries in the most favored nation (MFN) category; (ii) rates applied to countries without any trade agreements (50 percent higher than MFN countries); (iii) special preferential tariff rates for countries with special trade agreements with Vietnam; and (iv) antidumping and antidiscrimination laws. Thus, Vietnam has a multiplicity of tariff rates. The tariff rates are first assigned based on country of origin (five rates) and then on the classification of goods (four categories) resulting in 20 tariff rates. The following suggestions should be considered to reform the import taxes: i. There should be a basic set of tariffs covering preferential, MFN, or ordinary tariffs, supplemented by a set of special preferential tariffs for countries with treaties. For countries not accorded favored treatment, the MFN rate plus a certain percentage should be used. Preferably, the number of categories based on origin should not exceed three. ii. Within the base category, three rates based on classification of goods are suggested: (a) the lowest rate (for example, 5 percent), applicable to inputs; (b) the next-higher rate (say 15 percent), applicable to intermediate goods; and (c) the highest rate (say 25 percent), applicable to pure consumption goods. 34 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N iii. For goods whose imports are to be discouraged, a high tariff rate may be specified. Alternatively, imports of these goods may be licensed to keep the number of rates at three. iv. The government should implement formal WTO agreements, in consultation with the World Customs Organization (WCO), trade facilitation measures such as harmonization of tariffs, advance rulings, specialized appellate authorities, etc. Personal Income Tax (PIT) A Law on Personal Income was enacted by the National Assembly in November 2007 and became effective in January 2009. The law removes the different tax treatment of Vietnamese citizens and foreigners. It also removes the distinction between regular and irregular incomes. It has a broader base and includes income from capital investment (interest, dividends, capital gains, transfer of immovable property, copyrights, etc.). The number of tax-free job-related allowances has decreased. A new deduction for personal income tax purpose was introduced and extended to cover charitable contributions. The taxable period is the calendar year, but the taxable period equals each occasion of earning income in the case of interest on loans, dividends and profits from other types of individual taxpayer’s capital investment; capital gains and immovable property transfer; lottery prizes; royalties; franchising; and technology transfer. It means that an exempt threshold of 10 million dong applies in each case rather than on the basis of the total annual income from these sources. The PIT law has seven tax brackets: 5, 10, 15, 20, 25, 30, and 35 percent. The number of tax brackets is too large compared to international best practices. The tax brackets should preferably be limited to three or four. The problem of a narrow tax base persists. Corporate Income Tax (CIT) The CIT is one of the major sources of revenue in Vietnam and has been generating about 7.3 percent of gross domestic product (GDP) in recent years. The same tax law applies to domestic and foreign companies, so that the playing field is level. The depreciation rules are complicated, with 53 categories of fixed assets. The investor has to select a method from three possible rules of depreciation (straight line, reducing balance, and method of quantity or volumes). Clearly, this is bound to increase both administrative and compliance costs. A better approach would be to pool all assets in four or five broad categories and then depreciate them using a declining balance method. Some payments by companies are not allowed, although these are legitimate business expenses, such as deductions for benefits in kinds and payment of bonuses. Expenditures for some forms of sales promotion are capped. Such legitimate expenses should be allowed. The standard tax rate is 25 percent. A higher rate between 32 and 50 percent is stipulated for oil, gas, and other precious materials. This is in accordance with international norms. Reduced tax rates of 20 and 10 percent are applied as tax incentives. Multiple rates clearly make the administration complicated. The CIT provides incentives in specific sectors and specific locations. At present, incentives are very complex and need to be rationalized. A better approach could be to lower the overall tax rate to 23 or 20 percent and to eliminate the reduced rates of 10 percent and 20 percent and rationalize the incentive structure. 35 TA X R E F O R M I N V I E T N A M Taxation of Natural Resources Taxation of natural resources in Vietnam has gained greater importance because of its exports of oil and gas and the recent rise in oil prices. Metallic and nonmetallic minerals are subject to royalty alone at varying rates ranging from 5 to 30 percent. Natural forest products are subject to high royalty rates—in the range of 1 to 35 percent. Natural aquatic resources are taxed at 1 to 10 percent. It would be better to impose the normal CIT along with a single royalty of about 5 percent on all these natural resources, because higher royalty rates cause “high gradingâ€? and an inefficiency burden. The oil and natural gas sector is subject to a combination of royalty and CIT. The royalty rates vary from 1 to 40 percent and need to be rationalized. The CIT rate is 50 percent, which is in accordance with international norms. For investment-promotion projects, the tax rate is reduced to 32 percent. According to present CIT law, enterprise income tax incentives are not applicable to income from mineral extraction activities (Circular 130 part G). Similar to the taxation in the oil and natural gas sectors, a higher CIT rate should be imposed on minerals and precious metals, as well. Even when the normal CIT rate is reduced to 20 to 23 percent, a higher rate should be kept for these sectors. The government should consider adopting a production sharing arrangement in the oil and gas sectors along with a royalty and income tax regime. Nontax Regimes (Fees and Charges) According to the legislative documents on charges and fees, the charge rates for services must ensure the recovery of capital within a reasonable time and suit the payment cap ability of payers. In practice, however, there seems to be no linkage between the user charge or user fee and the cost of providing the service. The main features of fees and charges in Vietnam should be as follows: i. The cost of collection and compliance should be compared with the gain in revenues. It may be better to abolish fees and charges where the cost of collection and compliance exceed the revenues (for example, stamp duties and license fees). ii. The arrangements for determining rates and charges differ. In some cases, there are few rates (judicial record cards, issue and renewal of passport, permits for electricity activities), but some are quite complex (66 categories of customs charges and fees and about 200 categories of charges and fees for assaying medicines). iii. Most of the user charges and fees are fixed on a specific basis and not on an ad-valorem basis. The revenues are bound to decline in real terms with inflation. Therefore, the rates should be indexed to inflation. iv. Most user charges and fees are centralized. The international trend is toward more decentralization of both service delivery and revenue collection. An exercise should be undertaken to determine which services must be retained by the central government and which can be delegated to the provincial and local levels. Tax Administration A Tax Administration Law became effective on July 1, 2007, and provides the legal basis to ensure transparency and integrity in tax administration. In general, the law is comprehensive 36 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N and high quality, and it stipulates clear provisions for governing taxpayer services, tax appeals, and dispute settlement. Some commendable features of the Tax Administration Law are as follows: i. The law clearly specifies the rights of taxpayers during the process of tax assessment, examination, audit, investigation, and appeals. The law makes it mandatory for tax administration to compensate for damages caused to the taxpayer. ii. The law emphasizes the granting and use of Taxpayer Identification Numbers (TINs), which is the first step toward a modern tax administration. iii. There is symmetry of treatment of tax payment and tax refunds. If taxpayers delay paying taxes, they are liable to pay a penalty. In the case of overpayment of taxes, a refund must be made within a prescribed time. iv. The law stipulates that third-party information can be acquired from various agencies, while confidentiality of the information for tax purpose must be safeguarded. The law has also exposed the following major shortcomings that should be resolved: i. While the law stipulates the general rights and responsibilities of the tax administration agency and other related sub-national organizations, it does not include specific provisions on the division of responsibilities between various levels of tax administration (that is, the General Department of Taxation [GDT], provincial offices, and district offices). ii. There is a lack of a detailed legal basis in the following administrative operations: (a) the law does not specify cooperation between the tax and customs agencies; (b) the law does not include provisions on accounting and bookkeeping; (c) there is no provision governing electronic filing and the recognition of electronic signature; and (d) while the law is clear about the rights of appeal, it does not yet specify levels of appeal. 2. Introduction Vietnam is currently following a course of economic integration with the world economy. To facilitate this process, the government has undertaken a series of trade and tax policy reforms and has improved its fiscal management. As a result, the country has enjoyed spectacular growth during the past two decades, and most economic indicators have shown a very robust trend. Inflation has been generally under control, the trade deficit mild, and the exchange rate fairly stable. The budget performance has also been encouraging. The growth in revenues has been impressive, while expenditures have been contained, thus keeping the budget deficits within manageable limits. Selected economic indicators for 2001–10 are shown in Annex 1. The country has also been taking steps to reform its tax laws and modernize its tax administration. The features of the existing tax regimes of direct and indirect taxes are analyzed in this document within the framework of a “Good Tax Systemâ€? and in the context of international best practices. Suggestions for improving the tax regimes are presented and their linkage with improved tax administration and compliance are explained. The existing Tax Administration Law is also scrutinized within the framework of “Tenets of a Good Tax Administration,â€? and directions for further reform are presented. 37 TA X R E F O R M I N V I E T N A M Trade Liberalization, Economic Integration, and Tax Policy in Vietnam Vietnam has undertaken major steps toward trade liberalization as a prerequisite for economic integration. One important step is tax policy reform, which has an impact both on trade liberalization and economic integration and is, in turn, shaped by them. Trade Liberalization During the past two decades, the trade policy regime has undergone significant changes mainly in three directions: (i) restrictions on trading rights (that is, the right to import and export) have been relaxed; (ii) tariff rates have been reduced; and (iii) nontariff measures have been slackened. This has also been accompanied by changes in tax policy. Over the years, licensing requirements have been significantly reduced for exporting and importing. Currently, all enterprises with business licenses may engage in foreign trade in the goods specified in their business license without an import-export license except for a few restricted categories. The tariff schedule was rationalized following the country’s accession to the ASEAN Free Trade Area (AFTA) in 1992 and in preparation for the country’s accession to the WTO in 2007. Finally, in the past few years, Vietnam has significantly reduced the use of nontariff barriers. Another significant step was taken in 2000, when Vietnam signed the Bilateral Trade Agreement (BTA) with the United States, which envisages phasing out all nontariff barriers over a period of three to seven years. Economic Integration and Tax Policy In this era of globalization and economic integration, tax policy cannot be made without considering its international implications. Three issues become important in this context: (i) attracting foreign investment, (ii) global competition in the market for goods and services; and (iii) transfer pricing through creative accounting with international companies, such as reporting the bulk of their income in low-tax countries and seeking tax deductions in high-tax countries. Implications for Tax Policy and Tax Revenues Trade liberalization generally implies a decline in revenues from trade taxes. The government must, therefore, look for revenues from taxation of income and domestic consumption. 1 There are limits to increasing such taxation, however. Since Vietnamese goods have to compete in the world market, taxes on domestic or imported components cannot be levied. Further, income taxes have to be designed such that they do not hinder investment. It does not necessarily mean tax holidays and reduced tax rates, but it does mean using a reasonably low CIT rate, providing for tax credits to foreign investors, avoiding discriminatory taxes, 1 Property taxes (particularly land and building taxes) can be another potential source of revenue, and the tax is efficient from an economic perspective. The tax is the largest source of state and local taxes in the United States and elsewhere in the developed world. However, the tax is still largely marginal in developing countries, and Vietnam is not expected to collect large revenues from this tax over the short term. Reasons for this may be attributable to meager performance in the developing world: a tax culture and tradition combined to form political and social resistance to the tax; institutional constraints (in particular, the capacity of revenue administration agencies is typically weak, while the valuation and collection of the tax are sometimes completely segregated); and valuation of real property (both land and improvements on land) are inherently complex. 38 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N and keeping the system simple and transparent. This should also generally take care of the “transfer pricingâ€? problem. Overview of the Revenue System and Tax Reform Process The tax system in Vietnam has undergone three phases of reform in the last two decades. The elements of a modern tax system and tax administration were put in place during the first phase of reforms in the early 1990s. Tax laws enacted, revised, or both included the turnover tax, the special consumption (excise) tax, import and export taxes, the profits tax, the natural resource tax, the agricultural land tax, and the tax on the transfer of land-use rights, and the income tax on high-income earners. The second phase of reform, during the late 1990s to the early 2000s, focused on removal of the discriminatory nature of the tax system. The third and current phase of reform has been particularly motivated by the increasing integration of Vietnam with the world economy and its attempt to join the WTO. It includes the reform of virtually all taxes, including taxes for environmental protections. At the same time, GDT is in the process of a major reform supported by the Tax Administration Law. Revenue Performance Tax revenues in 2010 constituted around 86 percent of total government revenues, and about 24.6 percent of GDP. Total revenue collection is currently about 28.6 percent of GDP, which compares favorably with Vietnam’s neighboring countries. The fact that the tax revenues have been consistently high and total revenues have been more than 20 percent of GDP over the past 10 years is a remarkable achievement for a country like Vietnam. The revenues from different taxes and their buoyancy over the past decade are shown in Annex 2. The annex reveals that the CIT and the VAT are the two major sources of revenue, each contributing slightly less than 6 percent of GDP. The trade tax remains another major tax, accounting for approximately 3 percent of GDP. Excise and natural resource taxes have shown an upward trend, and were 2.5 and 1.3 percent of GDP, respectively. The contribution of the Personal Income Tax, although on an upward trend, is still quite low at 1.3 percent of GDP, according to 2010 statistics. Other taxes are almost negligible at less than 1 percent of total tax revenues. This shows that the tax system is currently somewhat skewed and needs to be suitably reformed. Buoyancy of the Tax System The buoyancy of the total tax system has been rather volatile. In 1997 and 1998, it exhibited negative buoyancy, which shows that the revenues fell in real terms during those two years. During 2003 and 2004, tax revenues grew at less than the rate of growth of GDP. But from 2006 to 2010, the tax system exhibited positive buoyancy. The average buoyancy of the tax system for 1997–2009 was 1.38 (table 1.1). This plausible result is yielded from the following two regressions: Ln (total revenue) 0.0958 Year + 15.001 R² = 0.9689 and Ln (GDP) = 0.0693 Year + 16.793 R² = 0.9973. 39 TA X R E F O R M I N V I E T N A M Table 1.1: Average Buoyancy of Tax System, 1997–2009 1997–2009 Buoyancy Total Tax Revenues 1.3859 Corporate Income Tax 2.3823 Individual Income Tax 1.7219 Land and Housing Tax 0.6006 Licenses Tax 0.6890 Tax on Transfer of Properties 1.9948 Tax on Land Use Right Transfer 2.5041 VAT 1.4828 Excises 1.5549 Natural Resource Tax 1.5468 Agricultural Tax -5.6202 Export and Import Taxes 0.9157 Source: Authors’ calculations. Ultimately, the tax system should become elastic and not just buoyant; that is, growth in tax revenues should be at least proportionate to growth in GDP without the need for making frequent discretionary changes in the base or rates. The recent dramatic improvement in buoyancy may cause some concerns that a higher tax burden may choke off investment and innovation. From the revenue efficiency perspective, however, higher buoyancy relates to more revenue stability and better opportunity for revamping the tax regimes to achieve the seemingly opposite aspects of a good tax system: enhanced economic efficiency and targeted equity. 2 Analysis of Specific Tax Regimes under the Framework of a “Good Tax Systemâ€? The following major direct and indirect taxes are examined in order to find the key issues that need attention during the current reform phase: (i) the Value-Added Tax (VAT), (ii) the Excise Tax, (iii) the Trade Taxes, (iv)the Personal Income Tax (PIT), (v) the Corporate Income Tax (CIT), (vi) Natural Resource Taxation, and (vii) fees and charges. The analysis of these tax regimes is conducted under the framework of a “Good Tax System,â€? targeting, in particular, the aspects of efficiency, equity, transparency, revenue adequacy, and stability. The features of a good tax system are briefly described in Annex 4. An overall picture of the tax system in Vietnam is presented, followed by an analysis of specific taxes. Economic Efficiency At present, there are several non-neutralities with respect to both tax policies and their implementation. A lack of harmonization among taxes, a multiplicity of rates, and a large number of exemptions and tax incentives give rise to a great deal of inefficiency in resource allocation, because taxpayers have incentives to base their business decisions on minimizing their tax burden instead of optimizing output and production. 2 For a detailed discussion of the basic concepts of tax buoyancy, tax elasticity, taxable capacity, and tax efforts, see, for example, Le et al. (2008); Musgrave and Musgrave (1989); and Shome (1988). 40 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N Equity Under the current system, there are numerous violations of horizontal equity. For instance, taxpayers under the PIT are taxed at different rates depending on the type of income earned. The numerous exemptions also lead to a violation of vertical equity. 3 The same thing is true for indirect taxes with multiple VAT and excise rates and many exemptions. Transparency Transparency primarily refers to tax laws that are not overly complex and discriminatory and tax regulations and administrative requirements that are easy to comply with. From this perspective, Vietnam has made significant strides in achieving transparency, although a great deal of reform is yet to be implemented. For example, in its present form, the VAT has zero plus two other rates and a long list of exemptions. Excises have many different tax rates and exemptions. The CIT has several tax rates and a series of incentives related to the nature of industries and their location that leave effective tax rates different from their statutory rates. Revenue Adequacy and Stability While lack of adequate data precludes the estimation of tax elasticities, volatile buoyancy points to the need for improvement in several areas of the tax system. The tax base is narrow and needs to be diversified and broadened. It is also important to monitor growing sectors of the economy with a view to bringing them under the tax net at the right time. This includes commodity sectors with high income elasticity of demand. International experience shows that a broad-based and simple tax system is more elastic and buoyant. Thus, there is a need to examine the tax laws with a view to simplifying and rationalizing them. Finally, it might be helpful to estimate the extent of current tax expenditures and examine their rationale and justification. The present tax regime in Vietnam is analyzed under the framework of a “Good Tax Systemâ€? outlined above. For each tax, first the prominent features of the existing tax regime are summarized, followed by regional and international best practices. Then, suggestions for reforming the current tax regime are presented. We begin with indirect taxes followed by direct taxes, natural resource taxation, charges and fees, and tax administration, in that order. The documents examined for the analysis are listed in Annex 5. 3. Value-Added Tax Main Features of the Current VAT Regime The VAT has proved to be the main workhorse for raising tax revenues in Vietnam. There are three VAT rates: zero percent, mostly for exports; 5 percent tax, for 15 goods and services; and 10 percent (the standard rate), for the remaining items. Twenty six goods and services are VAT exempt. At present, there is no threshold based on turnover for exempting small 3 Horizontal equity refers to a tax system that induces individuals of similar income level, regardless of their sources of income, to pay the same amount of taxes as a share of their gross income. Vertical equity refers to a tax system that induces individuals of higher income levels to pay more in taxes as a share of their gross income. 41 TA X R E F O R M I N V I E T N A M producers and traders from the VAT net. The credit method is used in parallel with the subtraction method, which is applied exclusively to small traders who fail to fully observe regulations on accounting, invoices, and documents, and to gold, silver, and gem-trading activities. The present VAT refund rules in Vietnam are complicated and vary with the nature and activity of the business. Regional and International Practices i. Except for China, Indonesia, and the Republic of Korea, most countries have a single VAT rate in addition to a zero rate. The standard rate in most Asian countries is 10 percent, while the rates in China are 13 and 17 percent and in Indonesia are 5, 10, 20, and 35 percent. Most Organization for Economic Co-operation and Development (OECD) countries have a single rate or two rates, and the standard rate varies between 5 and 25 percent. ii. International practice for exemptions is applicable to (a) agricultural products and inputs; (b) passenger transport; (c) cultural and merit goods; (d) service provided by the public sector, health, and education, in particular; (e) financial services; and (f) real estate and construction. iii. The threshold for exempting small traders in ASEAN countries varies from US$14,100 (the Philippines) to US$125,000 (Cambodia). In OECD countries, the threshold varies from US$1,600 (Denmark) to US$341,000 (Spain). iv. All ASEAN and OECD countries use the credit method of VAT computation except Japan, which uses the subtraction method. v. In developed countries, refunds are usually paid within three to four weeks of the end of a taxable period, and virtually no restrictions are placed on the refund payment. Most developing countries and transition economies limit in some ways the entitlement to a refund; for example, the amount of a refund is capped at the level of the output VAT in each tax period and the remaining balance is carried forward to the following tax period(s). Suggestions for Reforming the VAT Regime in Vietnam In light of international experience, the suggested reforms to Vietnam’s VAT regime are evident. Many of these also strengthen tax administration and enhance tax compliance. They are: i. Vietnam should apply a phased approach to simplify the rate structure, with a zero rate and one positive VAT rate. The lower VAT rate of 5 percent may be eliminated, leaving 10 percent as the single VAT rate. While the existing rate differentiation may generate modest equity gains, if any, it creates distortions and opportunities for misclassification of items, thus raising the cost of administration and lowering the VAT productivity. A single positive rate lowers the cost of compliance by reducing the need to keep records and invoices and also simplifies tax forms. ii. The current list of exempt items needs to be scrutinized and only a few of the 26 items should remain exempt. Too many exemptions create administrative and compliance problems for traders who sell both taxed and exempt items. More critically, exemption breaks the VAT chain and may lead to cascading, as in the case of a turnover tax. iii. The current practice of having a zero threshold of exempting small traders does not follow international best practice. A suitable threshold may be put in place based on the 42 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N tradeoff between the cost of administration and expected revenue generated. Instituting a turnover threshold for small traders will help the administration focus on big taxpayers rather than wasting its resources on a large number of taxpayers who contribute very little to tax revenues. A simplified tax regime can be applied to all traders below the threshold. iv. The credit method should be adopted across-the-board and the subtraction method currently applied to small traders should be abolished when an appropriate exemption threshold is established. This will improve tax compliance since it leaves a trail of invoices in all cases that may be audited and verified. v. The current VAT refund system needs to be rationalized and simplified to lower collection (administration and compliance) costs and to alleviate cash flow problems for businesses claiming genuine VAT refunds. 4. Excises or the Special Consumption Tax (SCT) The tax base consists primarily of goods whose consumption is regarded as socially and environmentally undesirable (alcohol, cigarettes, and gasoline); luxury goods (cars and air-conditioners); and some entertainment items more likely to be consumed by the rich (dancing and massage establishments, casinos, jackpot games, golf, and lottery business). Main Features of the Current Excise Tax Regime i. While there is a rationale for including entertainment items in the tax base, taxation of many of these items entails high administrative costs without generating much revenue. Examples are taxes on playing cards and votive objects. ii. The tax rates on goods vary between 10 percent (gasoline, air-conditioners of 90,000 BTU or less, and electrically operated passenger cars between 16 seats and 24 seats) to 70 percent (votive gilt papers and votive objects), and on services between 15 percent (lottery business) to 40 percent (dance halls). Thus, the tax rates are highly variable. The tax on gasoline, however, appears to be low even in comparison to the neighboring countries (for example, Cambodia and Laos, with a combined excise and import duty of 43 percent and 68 percent, respectively). A low import duty of 20 percent and an excise tax of 10 percent in Vietnam undoubtedly create incentives for gasoline smuggling. The rates for cars are differentiated by seating capacity and by fuel: gasoline, biofuel, or electrically operated cars. Liquor with different alcoholic content is currently taxed at two different rates—45 percent and 25 percent. Beer is taxed at 45 percent. Air-conditioners with a capacity of less than 90,000 BTU are subject to 10 percent excises while other air-conditioners are not taxed. A large rate differentiation among types of the same good creates scope for abuse. Similarly, the rate on dance halls, massage parlors, etc., is too high. iii. The provision of tax reductions for establishments facing difficulties caused by natural disasters or unexpected accidents is unnecessary. Equity issues are best addressed by income tax laws. Regional and International Practices In OECD countries, excise taxes are primarily imposed on alcoholic drinks, cigarettes (the so-called “sinâ€? taxes), and motoring-related activities, and to a lesser extent on betting and gambling. The rate of excise tax on gasoline varies greatly across countries (in Canada and 43 TA X R E F O R M I N V I E T N A M the United States it is around 10 U.S. cents per liter, while in France, Germany, and the U.K. it is 60 to 80 U.S. cents). The tax rates on tobacco and alcohol also vary significantly. Denmark and the U.K. tax them heavily, while Italy, Portugal, and Spain tax them lightly. Most countries use specific tax rates to target the alcohol in drinks and discriminate more against dangerous forms of smoking. When it comes to automobiles and motoring fuel, at least four types of taxes are imposed in OECD countries. They include (i) taxes on motor vehicles, (ii) taxes on motor fuels, (iii) license fees, and (iv) users’ fees and charges (driving license fees, road tolls, parking fees). Taxes on betting and gambling are targeted less at revenue generation and more at regulation. Excise taxes in OECD countries also address environmental issues. There is a wide variety of excise taxes in Asian countries. Both specific and ad-valorem rates are used, and they vary widely across items and countries. Taiwan taxes rubber tires, cement, beverages, gasoline as well as diesel and kerosene, fuel oils, electric appliances, and motor vehicles. China taxes tobacco, alcoholic beverages, cosmetics, jewelry, and motor vehicles. In Malaysia, mainly motor vehicles, beverages, and tobacco are taxed. The Philippines has a very long and differentiated excise tax schedule on beverages, tobacco products, fuel, automobiles, and cosmetics. Singapore applies excise taxes on alcohol products, tobacco, petroleum products, and motor vehicles. In Korea, excise taxes apply to a wide range of goods, including alcoholic beverages, luxury items, petroleum products, casinos, and nightclubs. Thailand levies excise taxes mainly on four product groups: petroleum products, beverages, tobacco, and motor vehicles. In Indonesia, excises are applied on 37 groups of goods including luxury items, electrical appliances, alcoholic beverages, and furniture. Suggestions for Reforming the Excise Tax Regime in Vietnam i. Tax rates under the present structure are quite dispersed. Some top excise rates should be lowered, while some low rates may be increased. The general approach should be to look at the impact of all the taxes—VAT, excise, import duty—on the final consumer price while assessing the correct level of excises. This helps improve tax administration and tax compliance and would lower the costs of both. ii. Over the short run, the government may retain the relatively low rate of excise on gasoline, given the recent dramatic increase in the international price and high inflation. However, over the medium to long term, the rate may be increased to a level comparable with the ones in neighboring countries. iii. A cost-benefit analysis of levying an excise tax on selected entertainment items should be conducted to compare the cost of administration and compliance with the revenues collected. iv. The provision of tax reductions and exemptions for establishments facing financial difficulty should be eliminated. This practice is tantamount to an implicit subsidy to establishments that fail for a variety of reasons that do not merit government support. This step would also simplify administration. v. Expansion of the tax base under excise taxes may be considered by bringing into the tax net selected additional luxury items primarily consumed by higher-income groups, such as electrical appliances, jewelry, cosmetics, motorboats, and furniture. 44 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N 5. Trade Taxes Trade taxes have been of special significance in Vietnam, because they used to bring about one-third of total tax revenues. This share has been falling mainly due to tariff reduction following Vietnam’s increased international commitments, and it decreased from about 21 percent in 1997 to 11.2 percent in 2010. With Vietnam joining the WTO, the contribution of trade taxes is likely to decline further. Tariff rates fall into the following three broad categories: i. Preferential rates applied to goods originating from countries in the most favored nation (MFN) category. ii. Common tariff rates applied to goods from other countries without any trade agreements. For those countries, tariff rates in this case were 50 percent higher than for MFN countries. iii. Special preferential tariff rates on goods from countries with special trade agreements with Vietnam. There is a fourth type of import duty under which several supplementary tax laws have been enacted that are in the nature of antidumping and antidiscrimination laws. Main Features of the Current Law on Import and Export Tax i. The law applies to all goods that either pass through the country’s borders or nontariff zones, and to goods purchased or exchanged by residents of Vietnam. ii. Two types of taxes are specified: an “absolute taxâ€? or specific tax, based on units of goods, and an “ad valoremâ€? tax based on the value in the import-export declaration. iii. Import duties depend on two factors: country of origin and classification of goods. There are effectively four types of duties, which vary by country of origin: • Preferential rates for countries that accord Vietnam MFN status. Over 160 countries, including China, European Union member countries, the United States, and several major developing countries accord Vietnam MFN status. • Special preferential rates are further reduced tariff rates commonly applied in the case of ASEAN members and with China and India, for example. • Ordinary rates that are 150 percent of preferential rates and applicable to countries that do not grant Vietnam MFN status and in addition discriminate against Vietnamese goods. • Special types of duties are safeguard measures (antidumping duties) on specified goods deemed to be dumped and causing injury to domestic producers. The duty equals the difference between the goods’ export price and their normal value. iv. The logic of this fourth class of duties (antidiscrimination) is questionable since rates 150 percent higher than MFN rates are already imposed on goods from countries not accorded MFN status. These conform to the letter but not the spirit of WTO rules, create potential for disputes, and hamper trade facilitation. Because the tariff rates are first assigned based on country of origin (five rates) and are then levied based on type of goods (four categories) for each country type, at least 20 different import tariff rates are in place. International experience shows that an import-substituting strategy does not lead to growth or innovation. It may be more efficient to have two or three basic moderate customs rates based on the classification of goods. 45 TA X R E F O R M I N V I E T N A M There is a special provision of exemptions from customs duty on first-time imports into the country. Most of these goods are common consumer items subject to a VAT regime. Therefore, the government should specify the mode of subsequent taxation of these goods when they are disposed by the first-time importer. Another provision in the law prohibits the import of certain goods, particularly used consumer goods including garments and home appliances. Thus, the duty-exempt goods are new goods, and when the duty-free importers dispose of the used goods in the domestic market, they are conferred a rent. The rationale for doing this is unclear. Provisions for the remission of duties and refunds have been specified in detail. These provisions are modern and in conformity with the laws in many developing countries. Regional and International Practices In the context of trade taxes, the international norm is set by WTO provisions. Since Vietnam has joined the WTO, it will go through an extensive review of its trade taxes and regulations that will involve rationalization of tariff rates and will have implications for various sectors in different ways. Some of the sector-specific implications of joining the WTO are summarized in Annex 6. Suggestions for Reforming Trade Taxes in Vietnam i. There should be a basic set of tariffs covering preferential, MFN, and ordinary tariffs, supplemented by a set of special preferential tariffs for countries with treaties. For countries not accorded favored treatment, the MFN rate plus a certain percentage should be used. Preferably, the number of categories based on origin should not exceed three. ii. Within the base category, three rates based on classification of goods are suggested: (a) the lowest rate (for example, 5 percent), applicable to inputs; (b) the next-highest rate (say 15 percent), applicable to intermediate goods; and (c) the highest rate (say 25 percent), applicable to pure consumption goods. iii. For goods whose imports are to be discouraged, a high tariff rate may be specified. Alternatively, imports of these goods may be licensed to keep the number of rates at three. iv. Export taxes should be imposed only on exhaustible resources. For other exports, various export protection schemes that involve a zero tax burden on exports through duty drawback equal to all taxes (federal, provincial, and local) may be considered. v. The inclusion of an “unjust enrichmentâ€? provision in refund law may be considered. In cases where refunds are due to the importer, but the importer is unable to prove that the wrongly charged higher tax paid has not been recovered from the buyer of the goods, the refund is not due to the importer but to the consumer. This excess amount may not be returned to the importer but may be kept in a “Consumer Welfare Fund.â€? vi. The government should implement formal WTO agreements, in consultation with the World Customs Organization (WCO), trade facilitation measures such as harmonization of tariffs, advance rulings, specialized appellate authorities, etc. vii. Vietnam should consider becoming a signatory to the Protocol on Simplification and Harmonization of Customs Procedures (the revised Kyoto Convention). 46 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N The main thrust of these tax policy reforms is to further rationalize tariff rates. This is bound to simplify customs administration and help enhance compliance. 6. Direct Taxes Revenues from the PIT and the CIT combined were about 8.7 percent of GDP in 2010. CIT revenue accounts for about 90 percent of total income tax revenue. 6.1. Personal Income Tax (PIT) Before the PIT law was passed by the National Assembly in November 2007, and before it went into effect on January 1, 2009, the taxation of personal income was regulated by the Ordinance on Income Tax on High Income Earners (Ordinance). An attractive feature of the PIT is that it is revenue elastic. In the case of Vietnam, however, the tax applies to only very high income earners, which leaves the majority of people out of the tax net and thus contributes very little to tax revenues. i. Narrow Tax Base: The Ordinance had covered a narrow tax base and was like an excise tax; that is, it covered very few taxpayers at high tax rates. The high exemption level virtually excluded the majority of Vietnamese from the tax base. 4 While the share of PIT revenue from foreign taxpayers increased from 48.2 percent in 2000 to 62.4 percent in 2003, for Vietnamese taxpayers it declined from 42.4 percent to 30 percent. The number of Vietnamese taxpayers in 2001 was 0.46 percent of the total population, but the number declined to 0.3 percent by 2003 due to the increase in income threshold. ii. The PIT as a Tax on the Process of Formalization of the Economy: The PIT used to be a tax on very high income earners. The question is to what extent this should be changed. While the PIT can serve as an instrument of income redistribution, it might also hinder the process of formalization. Gradual expansion of the formal labor market and increasing wage employment are the key drivers of poverty reduction in Vietnam. In trying to expand the tax base and make it progressive, the PIT may end up being a tax on formalization of the economy, and thus could hinder modernization and job creation. Thus, for sometime in the future, as a matter of policy choice, the PIT may remain a tax on high-income earners only, and its expansion to middle-income groups must be gradual. Any radical change in the PIT structure resembling the PIT in developed countries may do more damage than good. iii. Many Exemptions: Both the Ordinance and the 2009 law provide a long list of payments as allowances and social support payments that are still not taxed. International experience shows that this practice is likely to erode the tax base and become a major tax loophole. iv. Exemption Threshold: The Ordinance regulated an exemption threshold of 5 million dong per month. This threshold was very high—more than eight times the per capita income and four times the average income of employees working in enterprises at the time the Ordinance was effective. According to international practice, the threshold varies 4 The exemption threshold is set at 5 million dong per month for Vietnamese citizens and 8 million dong for foreign residents, or about than US$314 and US$503 at the current exchange rate of 15,900VND/US$, respectively. 47 TA X R E F O R M I N V I E T N A M between 2 to 24 percent of the income of the average production worker. The threshold has been abolished according the 2009 PIT law. v. The Individual as a Taxable Unit: The taxable unit is the individual income earner and this is in conformity with the international PIT trend. The Ordinance did not include provision for personal allowances for spouse, children, medical expenses, etc., but rather allowed for a flat high exemption. This practice facilitated the administration but narrows the tax base considerably. The law promulgated in 2009 has been revised to include this provision. vi. Inequity in the Tax System: The tax system is highly inequitable in Vietnam, and the principle of horizontal equity is violated in many ways for a variety of reasons, including the exemption of different kinds of allowances and savings incomes. vii. Tax Rates and Tax Brackets: Tax rates are highly differentiated, and this needs to be rationalized. According to the 2009 PIT law, there are seven tax rates between 0 percent and 35 percent. The highest tax bracket of 35 percent is too high compared to international best practices and developing countries, where the highest tax rates are converging to 30 percent. This should be seen in the context of tax competition and integration of PIT and CIT rates. With the highest tax rate in a developed country at 40 to 45 percent, it becomes attractive for workers to move to a country where the highest tax bracket is around 30 percent. At present, there is substantial disparity between the CIT rate of 25 percent and the highest PIT rate of 35 percent. Thus, this rate should be lowered to around 30 percent. The 2009 Law on Personal Income Key features of the PIT Law are as follows: i. No distinction between Vietnamese citizens and foreigners living in Vietnam: Personal income taxpayers include residents who earn taxable incomes inside and outside the Vietnamese territory and nonresidents who earn taxable incomes inside the Vietnamese territory. ii. No distinction between regular and irregular income: Although a scheduler tax rate has been adopted for some of these incomes, this still amounts to implicitly maintaining the difference in taxation of different kinds of incomes. iii. Broader coverage of incomes: The law includes several types of incomes, such as income from capital investment including interest on loans, dividends, income from capital gains, transfer of immovable property, copyrights including technology transfer, transfer of industrial rights, franchising, and income from inheritance and donations. Taxation of interest and dividends will help expand the tax base and is a welcome change. iv. Nontaxable income, tax-exempt income, tax reduction: The number of tax-free, job-related allowances has decreased, but some allowances remain. One major change is the inclusion of individual businesses that were subject to CIT in the previous PIT regime. One notable exemption in the act is overseas remittances. In addition, income from agriculture, forestry, feeding and catching seafood, etc., is tax exempt. v. Taxable period: The taxable period is the calendar year for incomes from businesses and employment. However, for both Vietnamese and foreign residents, the taxable period equals each occasion of earning income in the case of income from interest, dividends, and profits, capital gains, immovable property transfer, winning lotteries, royalties, franchising, technology transfer, inheritance, and donations. This means the exemption 48 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N threshold applies on a case-by-case basis rather than on the total annual income from these sources. vi. Social and medical insurance: These contributions are explicitly deductible from business income or from salaries and wages (Article 21), but their payment is not clearly mentioned as income. These contributions might escape taxation at both ends unless the word “pension,â€? taxable under the current PIT law, is meant to cover these items. vii. Family deduction: A deduction for the taxpayer and dependents has been introduced. For the taxpayer, the monthly amount is 4 million dong, and for each qualified dependent it is 1.6 million dong. Thus, the threshold has gone down from 5 million dong to 4 million dong, but the family deduction effectively increases the threshold. This may further decrease the number of taxpayers and lower tax revenues. viii. Charitable contributions: A deduction for charitable contributions has been introduced, which is in accordance with international practice. ix. Tax rates: The law has seven tax brackets: 5, 10, 15, 20, 25, 30, and 35 percent. The highest tax bracket has been reduced from 40 percent to 35 percent, which is a welcome step, but this has been offset by the increase in the number of tax brackets. Presumably, this is to promote vertical equity, but international experience suggests that having too many brackets complicates administration and encourages tax evasion while doing little to promote equity. The number of tax brackets should preferably be limited to three or four. A scheduler tax structure has been adopted for incomes from capital investment (interest and dividends), royalties and franchising, winning prizes, inheritance or donations, and the sale of immovable property.5 x. Threshold for tax liability: A threshold of 10 million dong is applied for income from royalties, franchising, winning prizes, inheritance, etc. Thus, these incomes also receive a special tax break under the new law. There is little justification for this tax break. xi. Narrow tax base: The problem of a narrow tax base persists. However, this may be acceptable in the short run in the interest of formalization of the labor market and the economy. xii. Nonresidents: There are scheduler rates instead of a flat rate. A tax rate of 20 percent is applicable for income from salaries or wages. Tax rates applicable to income from business are 1, 2, and 5 percent. Is the new PIT law more progressive or does it generate more revenue? A microsimulation model for PIT revenue forecasting indicates interesting results about relative productivity and progressivity of the two PIT regimes, current and previous (see Chapter 8: Taxation in Vietnam: Who Pay What). If the Ordinance and the 2009 PIT law were applied exclusively to income from wages and salaries, the new law would yield lower 5 Taxing income from capital investment is a complex issue, and a tradeoff between collection and economic efficiency costs need to be made. In a classical regime, corporate income is taxed at different levels; for example, corporate income is taxed at the corporate level, and then distributed dividends are taxed at the individual level. Such a regime is simple for administration, but is likely to generate unintended economic distortion. Hence, some countries attempt to introduce in their income tax system some form of partial integration (for example, in the United States, dividends are taxed at a substantially lower rate than the highest marginal rate of personal income tax). 49 TA X R E F O R M I N V I E T N A M revenues and look less progressive. This is not surprising in view of the newly introduced personal and family deductions. If, however, other forms of income as provided in the new law are incorporated into the tax base, the system becomes more progressive and yields higher tax revenues. Regional and International PIT Practices Revenues from the PIT contribute about 25 to 30 percent of total tax revenues in developed countries but substantially less in developing countries. There has been a shift from taxation of family to individual income and a gradual move from scheduler to a global system. However, some characteristics of a scheduler system have been retained in the sense that pension funds and capital gains are taxed differently. The number of tax brackets also varies across countries but is in the range of three to five. The threshold at which income taxes first hit the taxpayer also varies considerably, ranging from 2 to 24 percent of the income of the average production worker. The OECD countries typically permit a variety of deductions or allowances including a standard allowance, premiums for retirement and insurance, mortgage interest payments, interest payments on national and post office savings and government bonds, and unemployment and other welfare payments. The Asian countries present a varied picture. In Singapore, tax rates vary from 0 to 21 percent, with a threshold of 20,000 Singapore dollars. In Cambodia, tax rates are between 0 and 20 percent, with a threshold of riel 500,000 a month (or less than US$127). China adopted tax rates between 0 and 45 percent and a threshold of RMB 1,300 (or US$183). In Thailand, tax rates are between 5 and 37 percent while in the Philippines they are between 5 and 25 percent with a zero threshold after allowing some basic deductions. Indonesia has five rates between 5 and 35 percent while rates in Malaysia are between 0 and 28 percent, with a variety of deductions. Suggestions for Further Reforming the PIT in Vietnam The PIT Law, effective January 1, 2009, substantially improved many provisions. There is still scope for further reform, however. Improvements include: i. Some job-related allowances have been eliminated but the number of items under “nontaxable incomeâ€? and “tax-exempt incomeâ€? has, in fact, increased according to the law. This category should be further examined and restricted to truly deserving cases. ii. Social and medical insurance contributions are tax deductible and, therefore, should be clearly mentioned as income so that they do not escape taxation at both ends. iii. Income from agriculture and forestry are now tax exempt. While this simplifies tax administration, a distinction should be made between normal agriculture and forestry and those large farms dealing in cash crops and large-scale plantations. iv. The number of tax brackets is still too large. International experience suggests that too many brackets complicate administration and often encourage evasion while doing little to promote equity. The brackets should preferably be limited to three or four. v. The highest tax bracket of 35 percent is still on the higher side and needs to be reduced. This is particularly important in the context of the tax structure of neighboring countries. Tax rates on income from capital investment, interest, and dividends, in particular, are low (5 percent) and need to be increased to around 15 percent. 50 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N vi. The distinction between regular and irregular incomes still exists, and royalties and lottery prizes are taxed at lower rates and with a threshold of 10 million dong for each transaction. This income should be treated in the same way as regular income. Alternatively, a withholding tax of around 15 percent may be used, and it should become the final tax. vii. While schedular tax rates may be used in some cases (transfer of immovable property, interest and dividends, inheritance), the law applies them to too many items. viii. To account for inflation, the threshold and tax brackets should be adjusted periodically for bracket creep. 6.2. Corporate Income Tax (CIT) The CIT is a major source of revenue in Vietnam and has been generating about 7 percent of GDP in recent years. Main Features of the CIT The same tax law applies to domestic and foreign companies so that the playing field is level. (i) Tax base: a. The CIT covers incorporated businesses. b. The tax base includes capital gains earned from a variety of sources. There may be a tax loophole in that the gains in the sale of land use rights may escape taxation if the entire company is sold and land is a part of the company. A better alternative is to tax land value as such. (ii) Depreciation rules: a. The depreciation rules are complicated, as there are 53 different categories of fixed assets with different recovery periods, and an investor first is required to determine the period applicable and then to select one out of three possible depreciation methods. If the investor is not satisfied with the recovery period specified in the law, Circular 203/TT- BTC dated October 20, 2009, offers the possibility to initiate a consultation process with relevant government agencies to agree on the depreciation rules for fixed assets to follow in the individual case. b. To implement the depreciation regime, each fixed asset must be classified and maintained in a separate file and also recorded in an assets record book, numbered, and assigned a separate card with relevant statistics. Clearly, all this is bound to increase both administrative and compliance costs. c. A better approach would be to place all assets in four or five broad categories and then to depreciate them, using a declining balance method. When an asset is added to the pool, it becomes larger, and when something is sold, it becomes smaller. This also solves the problem of having to keep tabs on old and new assets and determine which asset has been depreciated to what extent at the time of its sale. For instance, if an asset has been expensed in year 1 and sold in year 2 at 90 percent of its original price, the pool and the tax liability for the original buyer are reduced while the new buyer will simply apply his depreciation schedule. If an asset has appreciated greatly, the pool may become 51 TA X R E F O R M I N V I E T N A M negative, indicating that the excessive capital gains would need to be taxed. Assets under accelerated depreciation should be kept in a separate pool. d. Depreciation rules may still be used as instruments of tax incentive for particular types of investment or investments in notified geographic areas. e. Accelerated depreciation up to twice the normal rate is applicable for “high income efficiencyâ€? establishments. (iii) Disallowed deductions and limits on deductions: Some payments by companies are not allowed under the law, although they are legitimate business expenses under internationally acceptable practices. Also there is a cap on the following forms of expenditures: a. Deductions for benefits in kind and payment of bonuses; such payment of bonuses would, of course, be taxable under the PIT paid by the employee. b. There is a cap on expenditures for advertising and other forms of sales promotion, including entertainment, and such costs cannot exceed 10 percent of the total deductible expenditure. For a new enterprise, such expenses cannot exceed 15 percent of deductible expenses for the first three years from the date of establishment. This kind of expenditure is generally legitimate and should not be capped. There has to be clear rules about allowing entertainment as an expense. (iv)Tax rates a. The present tax regime sets the standard tax rate at 25 percent. A higher rate - between 32 and 50 percent - is stipulated for oil, gas, and other precious and rare natural resources. Reduced tax rates of 20 percent and 10 percent as part of the CIT incentive scheme are applied to investment in preferred economic sectors or geographic areas. Multiple rates make the administration complicated. b. The existing regulatory framework does not differentiate tax rates on foreign company branches and domestic enterprises. A higher income tax rate on branches of foreign companies is advocated because branches are taxed only once, and there is no occasion to apply a withholding tax. (v) Tax incentives a. The CIT provides incentives for investments made in specific sectors, specific locations, and if they meet some social criteria (female or handicapped people). b. At present, incentives are very complex, and they need to be greatly rationalized. There are three major types of incentives: (i) preferential or reduced rates (20 percent and 10 percent), (ii) outright tax exemption or a tax holiday for a prescribed period of time, and (iii) accelerated depreciation. There are detailed criteria for each category that are sometimes unclear. Regional and International Practices in Enterprise Taxation In most developed countries, income arising from all sources including business, trade, and nonbusiness income, is included in the base. Most countries also tax capital gains at the full 52 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N CIT rate, but a few exempt capital gains if reinvested in business. Subject to restrictions in some countries to prevent tax avoidance, all expenses incurred wholly and exclusively in earning the income or maintaining the assets are tax deductible. All countries provide a depreciation allowance, but the rules vary. Some standardized method, such as the straight line or declining balance method, is followed. Tax incentives are generally offered in the form of accelerated depreciation at a higher rate than normal. All OECD countries allow loss carry forward ranging from five years to indefinitely, while some allow loss carry backward from one to three years. Tax rates in developed countries also vary considerably (from 19 percent in Canada to 45 percent in Germany). Top tax rates have been declining over the years under the increasing pressure of tax competition. Selected Asian countries apply various tax laws, and all offer different forms of fiscal incentives for certain types of investments or investments in designated geographic locations. The standard rates vary widely. For example, Indonesia has three rates - 10, 15, and 30 percent - while Malaysia has 28 percent for non-petroleum companies and 38 percent for petroleum companies. Korea applies two rates - 15 and 25 percent depending upon income - whereas the Philippines, Singapore, and Thailand have a single rate, which are 32 percent, 26 percent, and 30 percent, respectively. Suggestions for Reforming Corporate Income Taxes in Vietnam Following are suggestions for reforming corporate income taxes in Vietnam: i. The depreciation rules should be streamlined by placing all assets in four or five categories of pools and depreciated using a declining balance method. If an asset appreciates significantly, the pool becomes negative and the capital gains will be taxed. This will lower both administrative and compliance costs. ii. The provision of allowing accelerated depreciation up to twice the normal rate for “high economic efficiencyâ€? establishments is arbitrary and open to subjective interpretation. iii. Wage compensation in the form of in-kind benefits and bonuses should be tax deductible. Also, the genuine expenditures on sales promotion need not be capped. Such steps encourage voluntary compliance by companies and lower their costs of doing business in Vietnam. iv. Further lowering of the tax rate in the range of 20 to 23 percent and applying a higher rate on branches of foreign companies may be considered. This should be accompanied by rationalization of the overly generous incentive scheme. This helps tax administration and lowers compliance costs. v. The criteria for tax incentives are complex and need simplification. Also, outright tax exemption or tax holidays are not cost-effective and should be phased out. 7. Taxation of Natural Resources Taxation of natural resources in Vietnam has gained greater importance because of its export of oil and gas and the recent rise in oil prices. 53 TA X R E F O R M I N V I E T N A M Assessment of Natural Resource Taxation in Vietnam The natural resources in Vietnam may be placed into the following four categories, each with its own tax rates: i. Metallic and nonmetallic minerals: These are subject to royalties alone. The rates on minerals are in the range of 5 to 25 percent, on nonmetallic minerals 3 to 20 percent, and on gems 16 to 30 percent. This broad band of differentiated rates is in place perhaps to compensate for the absence of a CIT. It is better to impose the normal CIT along with a single royalty of about 5 percent because higher rates cause “high gradingâ€? and an inefficiency burden. ii. Natural forest products: These are subject to high royalty rates—in the range of 1 to 35 percent. One objective of such high tax rates could be to protect the forest resources, but high rates tend to promote the exploitation of high-value natural wealth and tax evasion, thus negating their very purpose. A CIT with a lower royalty may be considered for this sector, as well. iii. Natural aquatic resources and natural water: Most items in this category are taxed at 1 to 5 percent except pearls and sea slugs, which are taxed at 6 to 10 percent. These may also be subjected to CIT in combination with one lower rate of royalty. iv. Oil and natural gas: The oil and natural gas sector is subject to a combination of royalties and CITs. The royalty rates vary with the amount and depth of extraction and are in the range of 6 to 40 percent. The royalties on gas are taxed at rates of 1 to 30 percent. A higher royalty ensures early revenue flows to the government but leads to higher inefficiency costs to the economy. The CIT rate for this sector is 50 percent for normal exploration and production. For an investment promotion project, the tax rate is reduced to 32 percent. Currently, there is no separate CIT tax incentive for petroleum exploration. International Practice in Natural Resource Taxation The tax regime commonly applied to this sector comprises a combination of royalties, CIT (around 30 percent), and production sharing arrangements. Production sharing arrangements are most commonly used in the oil and gas sectors. A comparison of normal tax regimes with production sharing agreements, along with a numerical illustration, is presented in Annex 7. Suggestions for Reforming Natural Resource Taxation in Vietnam i. Vietnam should move to a robust combination of royalties and CIT for the taxation of all types of natural resources. A higher CIT rate may be maintained for the oil and natural gas sectors. ii. The rate of royalties in all sectors of natural resources is highly differentiated and for many sectors is quite heavy. These royalty rates should be reduced in number and should not be kept at an abnormally high level. This will help reduce administrative and compliance costs. iii. The royalty rates for oil and natural gas are also highly differentiated, and these rates need to be rationalized. 54 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N iv. Just as for the oil and natural gas sectors, a higher CIT should be imposed on minerals and precious metals. Even when the normal CIT rate is reduced to 20 to 23 percent, the rate for this sector may be kept at a higher level. v. The concessions and exemptions granted to certain categories of investors need to be closely examined and reduced in scope since such incentives are not effective and only result in revenue losses. This also simplifies the administration. vi. The government should consider adopting a production sharing arrangement in the oil and gas sectors. A royalty and tax regime can also be maintained along with the production sharing arrangement for the sake of revenue maximization. These are well tried instruments and would in the long run help in administration and in generating higher revenues. 8. Nontax Regimes (Fees and Charges) The Government of Vietnam has promulgated a large number of decrees, circulars, and decisions to impose different kinds of charges and fees and collect revenues from them. Assessment of Charges and Fees in Vietnam According to the Ordinance on charges and fees, the determination of charge rates is based on two principles: (i) the rates of charges for services provided by the state must ensure the recovery of capital within a reasonable period of time, and (ii) the rates for services provided by organizations and individuals must ensure the recovery of capital within a reasonable period of time and suit the payment capability of payers. In practice, however, there seems to be no linkage between the user charge or fee and the cost of providing a specific service. While the law specifies the rationale for fixing the rates of a user charge or fee, it seems that this important principle has not been applied in practice. The estimation of service delivery costs will also help determine the level of user charges and fees. The main features of charges and fees in Vietnam are as follows: i. Revenue and efficiency versus cost of collection: The cost of collection and compliance should be judged in the context of the expected gain in revenues and the efficiency of imposing a particular charge or fee. It may be better to abolish those charges and fees where the cost of collection and compliance outweigh the revenue and efficiency considerations (for example, stamp duties and license fees interfere with the efficiency but bring little revenue). ii. Multiplicity of rates: There is a wide variety of arrangements in determining the rates of charges. In some cases, there are only a few rates (judicial record cards, issuance and renewal of passport, permits for electricity activities), but some are quite complex (66 categories of customs charges and fees for various activities and about 200 categories of charges and fees for assaying medicines). iii. Indexing for inflation: As expected, most of the user charges and fees are fixed in Vietnamese dong, that is, on a specific basis but not on an ad-valorem basis. The level of the user charges or fees is bound to decline in real terms with inflation over time. Therefore, a system should be prescribed in the law to index the rates to the inflation rate in the economy. 55 TA X R E F O R M I N V I E T N A M iv. Need for decentralization: Many user charges and fees are centralized, but only a few are decentralized to provincial authorities. The international trend is toward more decentralization of both service delivery and revenue collection. A major exercise should be undertaken to determine which services must be retained by the central government and which are to be delegated to the provincial and local levels. Some degree of decentralization is suggested in Annex 8. International Practice in Levying User Charges and Fees In most developed countries, determination of user charges follows the “marginal cost pricing,â€? that is, the rate is linked to the cost of providing the service at the margin. Due to practical problems in applying this principle, however, average cost pricing is applied as the next best option. Also, equity considerations are a major concern in any kind of user charge or user fee pricing. Normally, it is better to leave this aspect for other public policies such as tax and expenditure policies at the national level or progressivity of the property taxes at the local level. Subsidized charges and fees may still be necessary in some cases, and they are justified on the grounds of “basic needâ€? such as supply of drinking water. It is common practice to apply the rule of cross-subsidization if it is possible to charge a higher price from high-income consumers and then subsidize the poor. In most countries, these services and charges are decentralized in the interest of efficiency and efficacy of service delivery. The role of user fees and charges is crucial in the preservation of natural resources and the protection of the environment. When appropriate levels of user fees are imposed on consumption of resources or on the negative externality generated in the process, the costs are internalized. Thus, the role of user fees is recognized internationally to be important in containing the demand and exploitation of valuable resources. Suggestions for Reforming Nontax Regime (Charges and Fees) in Vietnam i. A rational basis for determining the rate of a user charge or fee should be laid down. The best principle is to follow the marginal cost or at least the average cost of providing a particular service. ii. In some cases, where the cost of levying the charges and fees and collecting them outweighs the gain in efficiency and revenues, it may be better not to recover any user fees and charges. On the other hand, in some sectors such as mining, substantially higher fees may be charged for the privilege of prospecting, exploitation, or extraction of a natural resource. This will also help the tax administration and reduce compliance costs. iii. For some services, the rate structure is too complex, and it will improve the system if the rates are rationalized to the extent possible. It will also simplify administration. iv. Currently, all user charges and fee rates are set at the national level, and the laws and rules apply nationwide. Thus, there is a high degree of centralization. It is more efficient to provide these services and collect revenues at the regional and local level. Also, there is bound to be cost variation across regions, so a uniform and single system is not likely to be efficient. It is worthwhile to analyze all the different types of services and to allocate them to the central, provincial, and municipal levels of government, both for legislation and implementation. This will also make the administration more effective. 56 T H E L E G A L F R A M E W O R K F O R TA X A D M I N I S T R AT I O N v. By their very nature, rates for most charges and fees are fixed in dong and not as a percentage of the price. With a given level of inflation, the government or the agency providing the service will begin to lose revenue unless there is a well established system of revising the rates on a regular basis, in keeping with the inflation rate. 9. Tax Administration The adoption of the new Tax Administration Law, which was enacted on November 29, 2006, and which became effective on July 1, 2007, was a milestone in the ongoing tax administration modernization. It provides a legal basis for GDT’s concerted effort to ensure transparency and integrity in all tax administration operations. In addition, it is expected to help eliminate gray areas and uncertainties serving as breeding grounds for noncompliance by taxpayers and abuse by tax administration officials. In general, the law is comprehensive and of high quality. It consists of 14 chapters, covering all key tax administration business functions (tax registration, filing and accounting, assessment, audit and investigation, collection and debt management). The law has clear provisions for governing taxpayer service, tax appeals, and tax law violation settlement. Commendable Features of the Tax Administration Law The following are selected commendable aspects of the Tax Administration Law: i. Clear definition of the rights and obligations of taxpayers, the tax administration agency, and other related organizations. The law specifies the rights of taxpayers during the process of tax assessment, examination, audit and investigation, and appeals. Taxpayers are entitled to good taxpayer service that helps them understand the binding laws and regulations, and to obtain easy access to tax information and compliance guidance. The law mandates the tax administration to compensate a taxpayer for damages caused by the administration. To avoid abuse by tax officials, the law stipulates that a tax audit can be conducted on a periodic basis but cannot exceed once a year. The tax administration agency is legally responsible for the confidentiality of taxpayer information. The law also specifies the purpose and scope of tax administration; the rights and responsibilities of the tax administration agency; and the roles and responsibilities of other stakeholders such as the people’s committees, other government bodies related to tax administration, the media, and tax agents. The law includes a clear provision on international cooperation, which stipulates that, among other thing, the tax administration agency is authorized to negotiate, sign, and organize implementation of international agreements with tax administrations of other countries. ii. Unique Taxpayer Identification Number (TIN). The law emphasizes the granting and use of TINs, which is the first step toward modern tax administration. iii. Symmetric treatment of tax payment and tax refunds. The law deals explicitly with tax payment deadlines and tax refunds. If taxpayers delay paying taxes, they are liable to pay a penalty at a daily rate of 0.05 percent of the amount of late payment. On the other hand, overpayment of taxes is to be refunded or deducted from the tax liability in the subsequent period. The tax refund must be made at the request of the taxpayer within five working days from the date of receipt of the letter of such request. 57 TA X R E F O R M I N V I E T N A M iv. Third-party information for effective audit and enforcement. The law stipulates that third- party information can be acquired from various agencies, but the confidentiality of the information for tax purpose must be safeguarded. However, the new law also has some shortcomings (see below) that should be resolved in the next round of revisions. Areas for Future Improvement Areas for future improvement of the tax law include the following: i. Lack of specifics in institutional arrangements. While the law stipulates the general rights and responsibilities of the tax administration agency and other related subnational organizations, it does not include specific provisions on the division of responsibilities among various levels of tax administration (that is, GDT, provincial offices, district offices), on the division of responsibilities between the tax administration agency and people’s councils, or on people’s committees at various subnational levels. Experience in many developing countries indicates that the issue of dual subordination in tax administration at subnational levels is always one of the stumbling blocks to effective implementation of tax laws nationwide. ii. Lack of a detailed legal basis in some tax administration operations. First, the law does not make explicit the precise form of cooperation between the tax and customs agencies. Second, the law does not include any provisions for accounting and bookkeeping for tax purposes. Third, there is no provision governing electronic filing or, related to it, the recognition of electronic signature for electronic filing. Fourth, while the law is clear about the rights of taxpayers to appeal a decision by the tax authority, it has yet to specify what levels of appeal are available. Vietnam has made commendable strides in strengthening the legal framework for tax administration. Tax regimes are being revamped and a Tax Administration Law has been adopted. Such progress creates a solid legal foundation for the current comprehensive tax administration reforms. The structure of tax collection has been converging with the general trend in more advanced countries, with gradual reduction of trade taxes as a share in total tax intake. The Tax Administration Law establishes the legal basis for self-assessment and supports the strategic direction of reforms aimed to enhance transparency and integrity in the tax system and to create an effective culture of voluntary compliance. It is not surprising, however, that there still exists a number of shortcomings in the current legal setting for the tax administration that call for certain revisions along the reform process. In addition, GDT needs to strengthen the legal foundations for its operations by further prescribing specific rules and regulations for each tax administration function and task. Periodic review of tax laws and legislation is necessary and becomes a routine process in developed countries. Tax reform is a long-term, continuing process. Any shortcuts with ad-hoc changes cannot make a system elastic, but rather risk creating more uncertainties for investors and opportunities for discretionary treatment by tax officials. 58 Tax Compliance and Sources of Revenue Leakage: Conceptual II Framework and Assessment By Tuan Minh Le, Duc Minh Pham, G. P. Shukla, and Gregory Kisunko 1. Summary With respect to taxes, Vietnam faces the same challenge as other developing countries: how to effectively plug revenue leakage and thereby enhance compliance. A growing body of literature indicates that low compliance has the following detrimental impacts on economic growth and development: countries have fewer resources for development, the tax system tends to be more distorted and less equitable, and tax policies are more vulnerable to frequent ad-hoc changes. The typical paradox of the “missingâ€? middle reflects the reality in Vietnam, a country still in transition; that is, a myriad of household or small businesses coexist with large state-owned and foreign-invested companies, while medium-size enterprises are still emerging. This chapter provides a conceptual framework for assessing tax compliance and presents practical approaches to measuring compliance for Vietnam. While tax compliance or noncompliance is a straightforward concept, its measurement— and especially, any measurement of tax potential—is technically difficult. In principle, tax potential can be quantified in three distinct ways: (i) revenue forecasting, using micro data and applied by tax type; (ii) analyzing tax administration data to gauge compliance in key functions, specifically tax filing, accounting, and payment; and (iii) relying on taxpayers’ self-reporting for compliance. However, taxpayers have little incentive to reveal their true estimates, and thus relying on their estimates is not feasible. Given the limited data available for Vietnam, we apply the first two approaches to estimate compliance of selected taxes and of selective tax functions. The results indicate that the key challenge for enhancing tax compliance for Vietnam is twofold: how to sustain the relatively high compliance rate by the formal, large businesses, and how to improve compliance by small or household businesses. Such a challenge is not unique. As Vietnam embarks upon a new wave of tax policy and administration reforms, it can draw on the following lessons from international experience: • Analytical capacity within the General Department of Taxation (GDT) must be improved as an integral component of successful tax administration modernization. Without a systemic approach to develop a quality database and train sufficient staff for revenue forecasting, GDT will face an uphill battle to gather intelligent information on the perceived risks of noncompliance. • Effective reform of tax administration requires successful tax policy reforms. Vietnam is moving in this strategic direction; it is in the process of revamping all its tax policies while concurrently modernizing its tax administration. 59 TA X R E F O R M I N V I E T N A M • Tax reforms need to be implemented in parallel with public sector reforms, in general, and public financial management reform, in particular. Tax administration cannot become an “island of integrityâ€? in an overall public sector facing a high incidence of corruption, and at the same time, corruption or wasteful spending of public expenditures undermines people’s willingness to comply with their tax obligations. • Any focus on a large group of taxpayers as the main stable source of revenues needs to be complemented with improved compliance by small and medium enterprises (SMEs). Voluntary compliance by the hard-to-tax SMEs can only be achieved with lower compliance costs through a simplified tax regime, greater reliance on automatic filing, and a different approach to compliance management that would emphasize service rather than discretionary punishment. 2. Introduction Higher tax compliance with a broader base creates an opportunity for a country to reduce tax rates across-the-board and make the tax regime more stable, not affected as much by sectoral fluctuations in the economy, and less reliant on ad-hoc changes in the tax structure to meet revenue targets. Many developing countries with weak governance and accountability, an expansive shadow economy, and low tax administration capacity face similar challenges of how to plug the revenue leakage and thereby enhance compliance. This challenge is also true for Vietnam, where a myriad of household enterprises and small businesses coexist with large state-owned enterprises and foreign-invested companies while the “middleâ€? part of the distribution of enterprises is only emerging. Compliance by domestic private micro and small businesses is estimated to be very low, with enforcement costs often higher than what is being collected, whereas compliance is relatively high for large, established enterprises. The analysis of determinants of revenue leakage would facilitate the understanding of the specific nature and systematic patterns in noncompliance, which is necessary for developing a new risk-based compliance management philosophy in tax administration. This chapter provides a conceptual framework for assessment of compliance and presents practical approaches to measuring compliance and reasons for noncompliance in the case of Vietnam. The structure of the chapter is as follows. Section 2 establishes a conceptual framework for understanding compliance and noncompliance. Sources of noncompliance and its detrimental impact on the economy are explored. Section 3 presents measurements of compliance and noncompliance. Two different approaches—revenue forecasting modeling and estimation based solely on tax administration data—are applied to measuring tax revenue potential and level of compliance in Vietnam. Section 4 offers a conclusion. 2.1. Tax Compliance and Revenue Leakage: A Conceptual Framework A Conceptual Framework in Assessing Tax Compliance and Revenue Leakage Tax compliance is defined as the ratio between actual tax collection and potential tax collection. The noncompliance rate, or the relative size of revenue leakage, is equal to 1 minus the compliance ratio. The contentious issue is how to precisely define the tax potential 60 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E of an economy, and researchers still lack agreement on what exactly is being measured. For example, the Organisation for Economic Co-operation and Development (OECD) Compliance Measurement – Practice Note (2001) poses the following question: “How is compliance defined—is it compliance according to the tax authority’s or the taxpayer’s interpretation of the law and its application to the facts? Is it from another more neutral perspective?â€? (p. 3). Tax administrations tend to consider the different interpretation of the laws by taxpayers as evidence of noncompliance without regard to the fact that tax laws—as observed in many developing countries—may not be consistent and transparent, or they may have grey areas open to widely different interpretations by taxpayers and tax administrations. While the means to define compliance or, related to it, noncompliance, are still subject to much dispute, the sources of noncompliance or channels for revenue leakage are conceptually straightforward. Sources and determinants of revenue leakage or noncompliance Tax evasion and avoidance are the two fundamental sources for noncompliance or revenue leakage. These reflections of taxpayer behavior are, in turn, determined by multiple interrelated factors, both internal (tax policy, tax administration capacity, efficiency and integrity, and compliance costs) and external (appeals system, accountability in public expenditures, corruption, and the size of the shadow economy). Figure 2.1 depicts a conceptual framework for analyzing the determinants of noncompliance or revenue leakage. Figure 2.1: Determinants of Noncompliance and Revenue Leakage Internal Determinants External Determinants Accountability in Public Tax Policy Expenditures Taxpayer Behaviors Tax Tax Corruption Administration Tax Evasion Avoidance (Capacity, E ciency and Shadow Integrity) Economy Noncompliance Appeals and Compliance Review Cost System Conceptually, tax evasion and avoidance are different, but both lead to revenue leakage, less equitable tax systems, and more distortions to the competitive market framework. Taxpayers are said to engage in tax evasion practice when they use illegal means to evade paying their full tax liability. Tax evaders choose not to report or to underreport their earned incomes, sales revenues, or wealth while they attempt to overreport the amount of deductible expenses or even participate in outright smuggling or money laundering schemes. In tax avoidance, taxpayers abuse loopholes in the existing tax regimes in order to lower their tax liability. 61 TA X R E F O R M I N V I E T N A M Internal factors leading to noncompliance The complexity of tax codes and legislation and other undesirable features of tax policy, including an excessively high rate structure and a nonfunctioning system of penalties for noncompliant behavior, are the driving forces for tax evasion and tax avoidance. The OECD (2001) suggests that tax laws in many countries are imprecise and unclear, which creates a fertile ground for firms’ active engagement in tax planning. This is especially true in developing countries with weak administration and accountability, a lack of a tax culture, and excessively complex tax regimes. Complex tax policies, a poor legal framework for tax administration, weak or corrupt tax administration, and the lack of a proper taxpayer service function tend to impose high costs of compliance on taxpayers and, hence, encourage firms, especially SMEs, to operate in the shadow economy (see, for example, ITD 2007; Johnson et al. 1997; Shneider and Enste 2000). Recent large-scale surveys of tax practitioners in South Africa6 and Tax Compliance Cost Surveys in a number of Former Soviet Union countries, most recently Armenia and Uzbekistan,7 provide empirical evidence that the tax compliance burden on small businesses is significantly higher in relative terms than for larger businesses. Table 2.1 illustrates this finding. While the tax system in South Africa undoubtedly remains burdensome (and hence is still associated with substantially high compliance costs for SMEs), it leaves little space for interpretation of the laws by tax officials; that is, capacity and integrity of tax administration are high and corruption within the ranks is low. In this environment, the most common means of noncompliance for SMEs is tax evasion—businesses stay outside of the system, that is, they remain fully or partially in the shadow economy. Not surprisingly, survey respondents estimated that up to 60 percent of businesses in the lowest turnover bracket prefer to remain in the shadow economy. However, this is a relatively unusual situation; the majority of developing countries have legal and regulatory frameworks that either allow tax officials to interpret the law to varying degrees, or have inefficient and/or corrupt tax administration, or both. SMEs in those countries rarely have resources to access sophisticated tax avoidance schemes, thus they resort to either staying in the shadow economy or to colluding with tax officials in order to reduce their tax burden. 6 At the request of the National Treasury and the South African Revenue Service (SARS), the Foreign Investment Advisory Service (FIAS), a multidonor facility of the World Bank Group, provided assistance with a set of surveys of tax compliance costs for SMEs. The objective of the surveys was to document tax compliance costs for small businesses in South Africa, to identify the most onerous compliance burdens based on these costs as targets for tax administration reform, and to serve as a baseline against which future progress could be measured. The first survey was a large-scale, countrywide, web-based survey of professional tax practitioners whose main job is to help SMEs (as defined by South African tax legislation) comply with the tax requirements imposed by the state. It was complemented by a direct survey of SMEs and a survey of informal firms regarding their perceptions of tax compliance. A report on the Tax Practitioner survey results, entitled “Tax Compliance Burden for Small Businesses: A Survey of Tax Practitioners in South Africa,â€? which focused primarily on different turnover bands, was published by FIAS in September 2007 and is available at: http://www.fias.net/ifcext/fias.nsf/AttachmentsByTitle/FIAS_ tax_practioners_report_SA/$FILE/FIAS_Tax_Practitioners_Report_-_FINAL_29+Aug.pdf. 7 For Uzbekistan, see http://www.ifc.org/ifcext/uzbeksme.nsf/Content/Home. 62 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E Table 2.1: South Africa: Tax and Accounting Costs as Percentage of Turnover (Based on Highest Turnover in Each Turnover Bracket) R1– R300,001– R1,000,001– R6,000,001– R300,000 R1,000,000 R6,000,000 R14,000,000 Cost of Minimal Tax Services 2.2 0.7 0.1 0.1 Cost of Tax and Accounting 11.5 3.2 0.6 0.3 Services Source: FIAS. Apparently, some evidence shows that reasons for noncompliance may be different in South Africa and Vietnam. According to the results of an Enterprise Survey sponsored by the Enterprise Analysis Unit of the World Bank, South African businesses are expected to give a gift to a tax official in 3.13 percent of cases, while in Vietnam it is a much more common practice (33.68 percent).8 However, a number of factors of the gift-giving practice and, related to it, the compliance pattern, (for example, social culture), may be vastly different between the two countries and are not accounted for in such surveys. The practice is not so different if one compares Vietnam with other countries in the region. For example, according to the same survey, almost 39 percent of businesses in China report giving gifts to tax officials (survey 2003), and the figure was around 22 percent for the Philippines in 2009. External factors behind noncompliance The primary factors that are external to the tax system that have an impact on noncompliance or revenue leakage include poor public perception of accountability in public expenditures, corruption, the size of the shadow economy, and the efficiency of the judicial and appeals system. The OECD (2007) traces the historical development of the formation of accountable and effective states that has been closely linked to taxation. The concept of “fiscal social contractâ€? is related to the fact that citizens are willing to pay their taxes (and, hence, induce higher compliance), only if they have, in return, the rights to be represented in the decision-making process about how public finance is raised and used. Tax compliance is strongly linked with accountability and public sector governance. Compliance induces taxpayers to participate more actively in the process of shaping accountability and governance. At the same time, government accountability in the use of public funds contributes to the improvement of taxpayers’ perception of compliance. This clearly indicates that taxation and accountability work both ways. The challenge in a number of developing countries is not necessarily to tax more, but to tax broader groups of individuals and enterprises more consensually on the basis of greater willingness to pay and an increased propensity to mobilize demand for better public services. Other channels for revenue leakage are corruption, the shadow economy, and the interrelation between the two, which tends to deplete the tax base and create further distortion in taxation. Finally, if people perceive that the appeals and review system is efficient and effective and there is a fair degree of equity in the treatment of taxpayers as regards penalties for noncompliance, 8 Data for South Africa and Vietnam are from 2007 and 2009 surveys, respectively. These and other relevant indicators can be found at: http://www.enterprisesurveys.org/Custom/. 63 TA X R E F O R M I N V I E T N A M irrespective of their wealth or status, compliance is likely to be enhanced. Thus, an important external factor that induces compliance behavior is the existence of an independent appeals system capable of processing disputes between tax authority and taxpayer in a fair and timely manner (Gordon 1996). Needless to say, noncompliance or revenue leakage has a detrimental impact on economic growth and development. It imposes additional fiscal constraints on countries and leaves them with fewer resources to meet their development objectives through public expenditures and investment. In addition to revenue loss, the tax system risks becoming further distorted, with a more inequitable domestic revenue collection structure (Tanzi and Davoodi 2000). On the other hand, higher compliance or a reduction in revenue leakage has a direct impact on the enhancement of tax productivity. It allows countries to raise a certain amount of tax revenues without the need to frequently change the tax regime (typically, raise tax rates or introduce new taxes and fees). Relatively stable, broad-based tax regimes with low rates would help reduce uncertainty for businesses (lower risks for doing business and hence lower tax-induced barriers to investment) and make tax administration more efficient and effective. Given the data available, empirical estimates can be made to reveal the detrimental impact of just two specific external factors (corruption and the shadow economy) on revenue leakage and low compliance. These are outlined in the following section. Assessing the Impact of Corruption and the Shadow Economy on Revenue Leakage The level of corruption and the share of the shadow economy in a country are two of the leading factors that have a significant impact on tax compliance. It is not surprising that these are closely linked. Corruption and Tax Collection Tax economists and practitioners have long acknowledged that corruption is related to evasion, wasteful tax exemptions, and weak tax administration. These incidences directly lower government tax take and widen the gap between what tax administrators collect and what actually goes to treasuries. Tanzi and Davoodi (1997) show empirically that high corruption lowers government revenues.9 Using 1980–95 cross-country data, they find that corruption has a statistically significant negative impact on revenue collection, controlling actual per capita GDP (a proxy for the stage of development). We test Tanzi and Davoodi (1997) empirical results by using the same set of independent variables (GDP per capita and corruption) with an updated database for 1992–2002. We also extend Tanzi and Davoodi’s work by examining the impact of corruption on both revenue and tax collection. 9 Applying the ordinary least squares in a fuller empirical test, Bird et al. (2004) also find that institutional qualities, such as corruption or political stability, as measured by the governance quality index, have a statistically significant impact on a country’s tax intake. They use different instrumental variables (including ethnic fractionalization, language, and latitude) and the Hausman Chi-square test to examine but fail to detect the presence of simultaneity of the tax or revenue intake and institutional variables. 64 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E Annex 10 describes the data and summarizes the statistics of the variables. To ensure consistency, the data used are taken from the same sources: data on revenue or tax-to-GDP ratios and real GDP per capita are from World Bank World Development Indicators 2005 (WDI 2005), and data on the corruption index are from the International Country Risk Guide (ICRG). The original ICRG corruption index ranges from 0 (most corrupt) to 6 (least corrupt).10 To make it consistent and easy to compare with the results in Tanzi and Davoodi (1997), we have rescaled the ICRG index by multiplying it by -10/6 so that the index ranges from -10 (least corrupt) to 0 (most corrupt). Our empirical results lend support to Tanzi and Davoodi’s hypothesis that corruption is associated with lower government revenue (table 2.2). The results indicate that controlling for the level of income, an increase in corruption by one standard deviation (2.1) reduces the revenue and tax collection as a share of GDP by 1.64 and 1.39 percentage points, or by 7 and 9 percent, respectively (see annex 10 and table 2.2). Table 2.2: Revenue-Depleting Impact of Corruption Tanzi and Davoodi Le, Pham, Kisunko, and Shukla (2008)b (1997)a Independent Variables Dependent Variable: Dependent Variable: Dependent Variable: Revenue/GDP Revenue/GDP Tax Revenue/GDP (%) (%) (%) Constant 12.9 15.6565 11.0566   (13.7) (15.35) (14.30) Real per capita GDPc 3.73 4.4570 1.4190   (5.34) (9.67) (4.24) Corruption Index -1.71 -0.7789 -0.6635   (-9.28) (-3.71) (-4.27) Adjusted R-squared 0.28 0.3085 0.1347 Observations 1,042 662 662 Sources: World Development Indicators 2005; ICRG. Note: t-stats are shown in brackets. The coefficients on GDP per capita are multiplied by 10,000. a. Tanzi and Davoodi (1997) use data for 1980–95 from the following sources: IMF, Government Finance Statistics; World Tables; Business International; and Political Risk Services (ICRG). b. Data for 1992–2002. c. GDP per capita is in real purchasing power parity 2000 international dollars. Using the results of the regression analysis (table 2.2), we estimate the specific losses in revenue and tax collection due to corruption for those countries that have higher corruption indexes than the mean (the revised corruption index of -5.6, see Annex 11). The corruption impact is negative and significant. Annex 11 provides a detailed account of corruption-induced leakage 10 The ICRG provides alternatives for gauging the quality of the institutional setting of a country, particularly the corruption index and bureaucratic quality. The corruption index ranges from 1 to 6, where a higher number means lower corruption. For detailed presentation of the ICRG methodology, see http://www.prsgroup.com/ICRG_Methodology.aspx. 65 TA X R E F O R M I N V I E T N A M by country illustrated in figure 2.2.11 The revenue and tax leakage is presented for three groups of countries with low, medium, and high corruption. The low and high corrupt groups include the top and bottom 20 percent of the whole country sample, respectively (Annex 11). The tax and revenue loss is small for countries with low corruption (in the magnitude of 0.2 and 0.3 percent of GDP, respectively), but it is significant for countries with high corruption (approximately 2 and 2.2 percent of GDP, respectively). Figure 2.2: Higher Corruption Associated with Greater Revenue and Tax Loss Corruption & Revenue Loss Revenue / Tax Loss (% of GDP) Low Medium High 0.0 0.5 1.0 1.5 2.0 2.5 Corruption Tax Revenue Loss Nontax Revenue Loss Sources: ICRG; WDI 2005; authors’ estimates. Tanzi and Davoodi (2000) further hypothesize that statistically, corruption has a detrimental impact on the equity of a tax system. They find corruption has a larger negative impact on direct taxes than on indirect taxes. The results of their empirical work indicate that corruption exacerbates the imbalances in the tax take from direct and indirect taxes; it reduces the tax collection from a certain group of taxpayers (and normally those well-to-do businesses that have the ability to give bribes to tax officials), and increases relatively the collection from the poorer groups of taxpayers. However, independent of the corruption, by its very nature, the collection of direct taxes is technically more challenging than the collection of indirect taxes, and this is even more so in developing countries. Corruption and the Shadow Economy Economists have attempted to analyze the deep-rooted correlation between corruption and the shadow economy, both of which lead to the erosion of the tax base and, hence, leakage. The shadow economy conceptually includes activities not recorded in government statistics. As defined by Schneider and Klinglmair (2004), the shadow economy consists of all unreported income earnings from the production of legal goods and services, either from monetary or barter transactions, including all economic activities that would be taxable were they reported 11 The revenue and tax losses for a particular country in Annex 11 are estimated as a product of the corresponding corruption coefficients (-0.78 and -0.66, respectively), and the difference between the rescaled corruption index of this country and the mean rescaled corruption index (-5.6, Annex 10). For example, the revenue loss for Congo would be: [-0.78] * [-1.02 – (-5.6)] = -3.3 (Annex 11). 66 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E to tax authorities.12 Choi and Thum (2005) examine the connection between corruption and the shadow economy and their implications for the official economy. Surprisingly, they demonstrate that the shadow economy and official economy may be compliments rather than substitutes. They explain that an open option for an entrepreneur to flee the official economy constrains the ability of corrupt officials to artificially introduce distortions for businesses operating formally in order to extract private gains. The shadow economy thus mitigates government-induced distortions and leads to more efficiency of the official economy. However, the more common belief is that the shadow economy is a substitute for the formal economy. For example, Johnson et al. (1997) in their study of the pattern of the unofficial economy in the Former Soviet Union and Eastern Europe show that bad public policies—including burdensome combinations of taxes and excessive bureaucratic regulations, and corruption— may lead to the expansion of the unofficial sector at the expense of the formal sector and as a consequence result in the leakage of the government’s tax collections, continued deterioration of good public service provision, and poor overall economic performance. Shneider and Enste (2000) and Johnson, Kaufmann, and Zoido-Lobaton (1998) also conclude that corruption and the shadow economy are complimentary; a country with more corruption tends to have a larger shadow economy; this implies that there should be some negative correlation between the official and shadow economies. In any case, since we are dealing with “ghostâ€? activities, we recognize that estimating the size of a shadow economy would be technically challenging. Schneider and Enste (2000) and Schneider and Klinglmair (2010) offer an extensive review of the three most widely used methodologies for estimating the size of the shadow economy: (i) the direct or micro approaches, (ii) the indirect or indicator approaches, and (ii) the model approach (see annex 12 for a summary). Worldwide Comparison and Impact of Shadow Economy on Revenue Leakage Schneider, Buehn and Montenegro (2010) apply various statistical measures to estimate the size of the shadow economies over 2006–07 (Annex 13). They find substantial differences in the size of the shadow economy by country group. While the cross-country average for OECD countries was a modest 18.7 percent, the figure was significant for developing countries (38.7 percent) and only slightly lower for economies in transition (38.1 percent). The size of the shadow economy in Vietnam was relatively small (just below 16 percent).13 Combining the 12 The General Statistical Office (GSO) of Vietnam defines informal or shadow economy activities as including the following: (i) household production in rural areas; (ii) unregistered business activities in urban areas; (iii) not reporting income so as to evade taxes; (iv) domestic service; (v) smuggling; (vi) renting of houses or furniture; (vii) secondary and unreported activities of administrative offices, army offices, prisons, reeducation camps, and orphanages; and (viii) operations of nonprofit institutions, charitable associations, the Red Cross, and the like (Tenev et al., 2003, 14–5). 13 Based on the sample surveys conducted in 1989, 1992, 1994, and 1996, the General Statistics Office (GSO) of Vietnam estimates that the size of the Vietnamese shadow economy is much more significant (more than half the size of the formal sector GDP) (GSO 2001). The sensitivity of the estimate to the applied methodology may account for the different results obtained. To ensure cross-country data consistency for the empirical exercise, we use the comprehensive country sample as estimated by Schneider and Klingmair (2010). 67 TA X R E F O R M I N V I E T N A M shadow economy dataset by Schneider, Buehn and Montenegro (2010) with the data on tax-to-GDP ratios from World Development Indicators (WDI), we find that the correlation between the size of the shadow economy and tax collection is negative and substantial (-0.23) (figure 2.3 and annex 13). Figure 2.3: Negative Correlation between Size of Shadow Economy and Tax-to-GDP Ratio Size of Shadow Economy and Tax Collection 40 35 Tax Revenue (% of GDP) 30 25 Vietnam 20 15 10 5 0 0 10 20 30 40 50 60 70 Size of Shadow Economy (% of GDP) Sources: Schneider and Klinglmair 2010; WDI 2007. Estimating Country-level Impact of the Size of Shadow Economy on Tax Leakage Annex 13 shows our rough estimate of the extent of tax leakage due to the shadow economy. The share of tax collection in GDP weighted by the size of the shadow economy is a proxy for tax leakage. The estimation relies on the implicit assumption that there are no systematic differences in the nature of business in the shadow and official economy. However, shadow businesses are typically either small or engaged in illegal activities and, hence, they are generally associated with higher evasion risks and pose a significant challenge to tax administrations. Thus, the estimates should be regarded as indicative and as the upper bound of the leakage. A summary of the impact of the shadow economy on tax leakage by three groups of countries with low, medium, and high shadow economies is presented in table 2.3. Using the database by Schneider and Klinglmair (2010), we classify countries into three groups by the size of their shadow economy. The low and high groups consist of the 40 percent in the top and bottom quintiles of the data. The results are striking; moving from a country with a small shadow economy (averaged at 17.6 percent of GDP) to a country with a significant share of shadow economy (averaged at 35.7 percent of GDP) is correlated to an estimated substantial loss of tax revenue. The loss sharply increases from approximately 3.5 percent of GDP to 6.1 percent of GDP, respectively. 68 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E Table 2.3: The Bigger the Shadow Economy, the Greater the Tax Leakage Size of Shadow Economy and Revenue Leakage: Summary Statistics Size of Shadow Economy (% of GDP) Tax Revenue Leakage (% of GDP) Low 17.6 3.5 Medium 28 4.9 High 35.7 6.1 Sources: Schneider and Klinglmair 2010; WDI 2007; and authors’ estimate. 2.2. Measurements of Compliance and Noncompliance As noted, the key and most contentious issue in measuring tax compliance is how to gauge the size of potential revenue. Conceptually, the latter can be measured in three ways: • Tax potential is estimated as a result of a revenue forecasting exercise using micro data. The forecasting models are applied to each type of taxes. • Tax potential is as regarded by a tax administration agency. This indicator can be inferred from the compliance level estimated for the key functions of tax administration, specifically, tax filing, accounting, and payment. The estimates rely primarily on the comprehensiveness and quality of the tax administration database. • Tax potential is as regarded by taxpayers. In practice, if surveyed, taxpayers have little incentive to reveal their true estimates, and thus tax potential as regarded by them is unobservable. In self-assessment, they typically submit to the tax administration the returns that already reflect, among other things, their interpretation of gray areas of tax laws to their own benefit. This implies that, realistically, there are two practical approaches to measure tax compliance or its opposite, noncompliance: (i) modeling to forecast potential tax revenues and then comparing them with the actual collection; and (ii) using tax administration data to estimate the gap between the potential and actual figures in key tax functional processes. The empirical modeling is conducted on the basis of data from multiple sources and may provide a fuller scope of tax potential. But it is sensitive to the quality of data available, assumptions made, and specific models applied. It should be particularly noted that in order to assess tax potential using microbased revenue forecasting models, data should ideally come from two sources: (i) the tax administration data set, and (ii) a nontax database, such as industrial and commercial surveys, national accounts, and the census. Tax administration data by themselves are insufficient to estimate tax gaps using microsimulation, unless they are complemented with external data. The second approach that relies on tax administrative data may underestimate the scope of leakage due to an inherent lack of information on the potential tax base or the growth potential of the base. However, it reflects the best pragmatic estimates done by, and at the discretion of, tax administrations. Therefore, in the absence of routine work on revenue forecasting, such administrative data are typically the sole source of information used in the preparation of their compliance enforcement work plans. 69 TA X R E F O R M I N V I E T N A M For designing compliance strategies, advanced tax administrations normally apply both approaches as compliments. Given the limitation of data, we attempt to apply the two approaches to estimate compliance and noncompliance of selected taxes (the VAT and CIT) and of selective tax functions. Modeling of Tax Potential, Tax Compliance, and Noncompliance for the VAT and CIT Estimation of the VAT potential and compliance The estimation of the VAT gap for Vietnam is calculated using the Input-Output (I-O) Model.14 The model is built and simulated on the basis of data from the 2005 Vietnam I-O Tables, the Household-Expenditure Survey, National Accounts Aggregates, and VAT collection data from GDT. The total potential VAT base for the country (Bv) can be expressed as follows: B v = ∑ Biv â‹… α iv + ∑ K j (1) i j Where: Biv = VAT base for commodity i α iv = VAT taxable proportion for commodity i K j = Business inputs purchased by tax-exempt sector j. The potential VAT revenues for the economy (Rv) is equal to the summation of all the adjusted tax bases across commodities and sectors multiplied by the VAT rate, at the full compliance rate. R v = B v â‹… Ï? v ⋅θ v (2) Where: q v = VAT compliance rate (100 percent for estimation of VAT revenue potential) rv = VAT rate. Finally, the tax gap (Gv) is estimated as the difference between the potential VAT revenue (Rv), and actual VAT collection (TTRv). G v = R v − TTR v . (3) In the VAT model under consideration, assuming the full compliance rate, the VAT has a revenue potential of approximately 48,980 billion dong for FY06, whereas the actual collection was 36,469 billion dong. This implies that the VAT compliance rate in Vietnam is 74 percent and that the tax gap is estimated to be 12,511 billion dong, equivalent to 1.3 percent of GDP. 14 Jenkins, Kuo, and Shukla (2000) present in detail the I-O modeling methodology. For a discussion of various methodologies for estimating the VAT base and an application for the case of Romania, see Le (2007). United Nations (1999) provides a framework for compilation and analysis of I-O tables. 70 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E Estimation of the CIT Potential and Compliance The limitation of the data from individual tax returns for the CIT precludes direct application of the microsimulation model. For Vietnam, we rely on the concept of an operating surplus in business operations and follow a national accounting approach to estimate the CIT potential and compliance. The Gross Operating Surplus (GOS), which is obtained from the Input- Output Tables (2005), is used as the basis to approximate the CIT base. From the income approach, GDP is defined as: GDP = ΣVA = W + GOS + TSP Where: ΣVA = the sum of value added across economic sectors W = compensation of employees, including wages, salaries, and other labor costs (employer’s social security contribution plus employees’ social security contributions) GOS = Gross Operating Surplus of enterprises (including profits, rents, interest, and depreciation) TSP = taxes, less subsidies on products. The Estimated Operating Surplus, which is net of depreciation, is used as a proxy for the CIT base.15 From the gross operating surplus (derived on the basis of the 2005 I-O tables and the official GDP data by economic sector), estimates of the operating surplus distribution by state-owned, domestic nonstate, and foreign direct investment are made (table 2.4). Table 2.4: Estimation of Operating Surplus by Economic Sector, 2005 (Billions of dong, current prices) GDP Operating Surplus State-owned 311,241 75,777 Domestic Nonstate (private) 382,804 90,018 Foreign Investment 134,166 31,550 Total 839,211 264,933 Sources: Vietnam Department of Statistics; author’s calculations. Table 2.5 presents the estimate of the CIT compliance and collection gap, with the CIT potential estimated on the basis of the standard CIT rate of 28 percent. The estimation excludes the foreign investment sector due to a complex set of rules effectively applicable to this sector, for example, various statutory rates and special treatments for foreign direct investment in the natural resource sector—mostly in the oil sector—or in different types of business, such as build-operate-transfer (BOT) or build-transfer (BT). 15 With a comprehensive database of the CIT returns, which is part of the objectives of the ongoing tax administration reform process, the GDT will be able to estimate the base more directly using a microsimulation approach, rather than relying on the OS concept. 71 TA X R E F O R M I N V I E T N A M Table 2.5: Estimation of the CIT Tax Gap, 2005 (Billions of dong, current prices) Operating General Potential Actual Tax Tax Surplus Tax CIT CIT Compliance Gap Rate (%) Revenue Revenue (%) A B C=A*B D E=D/C F=C-D State 75,777 28 21,217 15,107 71.2 6,110 Nonstate 90,018 28 25,025 6,633 26.3 18,572 Sources: Vietnam Department of Statistics; Ministry of Finance; author’s calculations. The domestic nonstate sector, consisting mostly of small and medium private businesses, stands out as the most problematic in compliance. Statistics show that the sector accounts for more than 90 percent of total CIT taxpayers but contributes less than 9 percent of total collection.16 Table 2.5 reaffirms this; compliance by this sector is just over 26 percent, while compliance by the state-owned sector is more than 71 percent. The VDR 2008 also reports that the SOEs remain the most important taxpayers, accounting for approximately 54 percent of total CIT. A sensitivity analysis indicates that if the compliance rate for the domestic nonstate sector is improved to the approximate level by the state sector (70 percent), CIT collection can be improved by almost 3 percentage points of GDP. Compliance Estimates on the Basis of Tax Administration Data On the basis of tax administration data, compliance is typically estimated for the following three major stages or key functions of the tax administration: • Filing compliance: Filing compliance can be disaggregated into three indicators. First, filing compliance by registered taxpayers. This indicator measures the ratio between the number of tax returns filed and the number of taxpayers actually registered with a tax administration. Actual returns filed may include both timely filing and late filing. Second, compliance of filing on time, which is defined as the ratio between the number of tax returns filed on time and the number of total tax returns actually filed. Third, composite filing compliance, which is the ratio between tax returns filed on time and the total number of registered taxpayers. The composite filing compliance is estimated as the product of the first two indicators. • Reporting compliance: Reporting compliance refers to truthful reporting of taxable income, sales, and production as relevant to specific direct income or indirect sales taxes. Reporting compliance is defined as the ratio between the income or sales reported and the income or sales as expected to be reported. Reporting compliance can be detected during the process of tax assessment and accounting. • Payment compliance: Payment compliance refers to the payment of taxes in a timely manner. It is defined as the ratio between the amount of tax liability paid on time and the full amount of tax liability owed by taxpayers. 16 The magnitude of the foregone revenues is large, and some other policy-related factors like differences in thresholds, definition of small business taxable income, and exemptions may be contributing to this. 72 TA X C O M P L I A N C E A N D S O U R C E S O F R E V E N U E L E A K A G E At present, the tax administration does not have sufficient information for estimation of reporting or payment compliance.17 However, GDT has managed to maintain the tax filing database, which allows for analysis of filing compliance. Table 2.6 provides a summary of tax filing compliance by major taxes (business income tax, individual income tax, VAT, excise, and the natural resource tax). The estimates indicate that the composite filing compliance varies substantially. In common with other developing countries, individual income tax has a very low level of compliance (28 percent). Compliance management of the VAT, the business income tax, and the excise tax is fairly good; in particular, compliance reaches 84 percent for the VAT, which does not have any threshold. Table 2.6: Tax Filing Compliance by Major Taxes, 2010 Filing Compliance of Compliance of Composite Filing Tax Type Registered Taxpayers Filing on Time Compliance Business income tax 71 68 58 (preliminary figure) Individual income tax 44 36 16 VAT 91 79 72 Excise duty 47 40 19 Sources: GDT and authors’ estimates. 3. Conclusion Noncompliance in the payment of taxes depletes the tax base and leads to revenue leakage. Countries with low compliance levels are likely to require frequent tinkering with their tax regimes to maintain revenues at a certain percentage of GDP. Frequent and ad-hoc changes in tax regimes, in turn, risk making the system less stable and equitable. Enhancing voluntary compliance has become a key institutional and operational objective of worldwide tax reforms. Due to the limited availability of quality microdata on business operations and economic sectors, and the lack of a comprehensive tax administration database, only some preliminary estimates of tax compliance can be made. While the results indicate that Vietnam has attained commendable achievements in compliance with major taxes, a number of issues remain. The challenge is twofold: to sustain the relatively high compliance rate by the formal, large businesses, and to improve the compliance of SMEs. The government is currently embarking on a comprehensive tax administration modernization reform program, with a focus on improved integrity, enhanced taxpayer service, and risk management to ensure high tax compliance and better equity. The following lessons can be learned from other revenue reforms:18 17 As part of the ongoing Tax Administration Modernization Project (TAMP), the GDT is to reengineer tax procedures and automate the administration, which will facilitate the development of a reliable database for complete compliance analysis. 18 These lessons are also featured in the 2007 World Bank Development Committee report on Fiscal Policy for Growth and Development. For the report, see: http://siteresources.worldbank.org/DEVCOMMINT/Documentation/21289619/DC2007- 0004(E)-FiscalPolicy.pdf. Also, for country experience in revenue reforms and lessons learned, see Casanegra de Jantscher and Bird (1992). 73 TA X R E F O R M I N V I E T N A M 1. Analytical capacity within GDT must be improved to support revenue administration reforms. The new philosophy in revenue administration based on voluntary compliance requires effective compliance measures and improved taxpayer service. At present, GDT does not have sufficient in-house analytical capacity to estimate compliance, by tax, or by major economic sector. Without a systematic approach to develop a quality database and to train sufficient staff in revenue forecasting, GDT will face an uphill battle to gather intelligent information on the perceived risks of noncompliance. 2. Effective reform of tax administration requires successful tax policy reforms. Tax policy and tax administration are linked. In countries with poor tax administration capacity—posed as a binding institutional constraint to tax reforms—tax policy reform should take into account the administrative capacity and contribute to reducing opportunities for corruption. Frequent adjustments of tax regimes may help keep the system buoyant, but there are no credible short-term solutions for enhancing compliance that deal with tax policy or tax administration reforms in isolation. Vietnam is currently revamping its entire direct and indirect tax policies while modernizing its tax administration. 3. Tax reforms need to be implemented in parallel with public sector reforms, in general, and public financial management reform, in particular. Corruption in the public sector in general, and in tax administration in particular, is one of the key determinants of low compliance. The recent trend in tax administration reforms shows that projects increasingly focus on good governance, with anticorruption being at the core. However, there is “no island of integrity,â€? and at the same time, corruption or wasteful spending of public expenditures undermines people’s willingness to comply with their tax obligations. Thus, tax reforms should be conducted in parallel with public administration reforms and reforms in budgeting and public expenditures. 4. Any focus on a large group of taxpayers as the main stable source of revenues needs to be complemented with improved compliance by SMEs. There are strong arguments for widespread inclusion of SMEs in the tax net as a direct approach to shrink the shadow economy and enhance the equity of the tax system (ITD 2007). Although the benefits for SMEs of being in the tax net are substantial (for example, easier access to finance and the possibility to have formal contracts with the formal sector), their compliance costs are disproportionably high. While good analytical capacity to project revenues, and reforms to streamline tax laws and reduce opportunities for corruption are necessary, they are not sufficient to change compliance behavior (especially in the SME sector), if taxpayers’ needs and views are not taken into consideration. Voluntary compliance by this segment of taxpayers can only be achieved with lower compliance costs through a simplified tax regime, greater reliance on automatic filing, and a different approach to compliance management that would emphasize service rather than discretionary punishment. As part of the overall reform program, GDT is to focus its resources on establishing a working system of feedback mechanisms and improving capacity to act on the results of feedback assessment. 74 Tax Administration By Duc Minh Pham and Michael Engelschalk III 1. Role and Challenges of Tax Administration in Vietnam With almost 44,000 employees, the Tax Administration of Vietnam (General Department of Taxation, GDT) is one of the largest public sector institutions in the country. It is organized at three levels, with tax administration headquarters in Hanoi, 63 tax offices at the provincial level, and 694 district tax offices. GDT thus maintains a physical presence in all provinces and districts in the country. GDT is charged with collecting all major domestic taxes, in particular the Value-Added Tax (VAT), the Corporate Income Tax (CIT), excise taxes, and natural resource taxes, which account for 76.4 percent of domestic revenue collection.19 GDT is also responsible for the collection of a relatively large number of minor taxes, such as the tax on the transfer of land use rights, the land and housing tax, and the small business tax. Domestic tax revenues, including tax revenues from crude oil, collected by GDT, amounted to 25.1 percent of GDP in 2010. GDT is operating in a rapidly and fundamentally changing environment, which will substantially increase the challenge of ensuring effective tax revenue collection. The greater integration of Vietnam into the international trade system has led to a substantial decrease in the collection of revenues from international trade taxes. A milestone in the trade integration process has been the accession of Vietnam to the World Trade Organization (WTO) in January 2007. Vietnam’s international trade integration has been accompanied by tariff reductions, and the share of total tax revenues from import and export taxes has decreased from 28.9 percent in 1996 to 13.2 percent in 2010. Shifting collection priorities from the border to the domestic market, however, puts a much greater burden on the compliance management system and increases the risk of tax evasion. Taxes from state-owned enterprises (SOEs) still constitutes the major source of tax revenues, although with the move to a market economy, the importance of SOEs in the economy and for revenue collection is decreasing rapidly. According to 2010 estimates, 35.1 percent of CIT revenues and 36.4 percent of VAT revenues from domestic production are paid by SOEs. The increasing role of the private sector in the economy so far has not been reflected in the revenue collection results, and the contribution of the nonstate sector to total revenue collection remains small at 17.4 percent in 2009. The decrease in trade tax revenues has been compensated primarily by the operation of an extremely efficient VAT. The efficiency ratio of the VAT in Vietnam is close to 0.6 and thus exceeds the average VAT efficiency in Organisation for Economic Co-operation and Development (OECD) countries. While some cascading-inducing features of the Vietnamese VAT system have contributed to the exceptional revenue performance at the expense of economic efficiency, the short-to-medium-term challenge is that the planned revision of the 19 Including domestic tax revenues from crude oil. 75 TA X R E F O R M I N V I E T N A M VAT, including the rationalization of the exemption and enhancement of the VAT refund, may result in a revenue loss. 2. The Need to Modernize the Tax Administration System GDT’s current level of efficiency compares favorably with a number of other countries in the region. However, operational costs of tax administration amount to 2 percent of tax collected, and thus are considerably higher than the less than 1.5 percent in most OECD countries. Around half of OECD country tax administrations have operational costs of less than 1 percent of revenue collected. In addition, a number of Asian countries have achieved a higher level of cost-efficiency (for example, Indonesia at 0.58 percent, Malaysia at 1.41 percent, and Singapore at 0.80 percent of revenue collected [OECD 2010]). An extensive network of field offices, insufficient information technology (IT) support for core operational functions, and labor-intensive business processes contribute to comparatively high operational costs. Existing operational processes are not fully suited to effectively addressing compliance management challenges in a market economy. GDT will also need to reform its operations to ensure fairness and transparency in administering the tax system. Major progress has been made in modernizing the organizational structure of tax administration. Starting in 2007, organizational units at all levels (headquarters, regional offices, and district offices) have been gradually reorganized along functional lines. GDT now has 13 major departments, of which only the department for Personal Income Tax (PIT) administration shows features of a tax-type organizational structure. Separate departments have been created to manage operational functions, such as tax return processing and accounting, collection of arrears, and taxpayer services and information. With the establishment of a special department for large taxpayer administration, the organizational basis for developing targeted compliance management strategies for the most important taxpayer segment has been developed. A pilot self-assessment program was begun in 2004 for the VAT and CIT in Ho Chi Minh City and Quang-Ninh Province. It was later rolled out to other provinces and extended to other types of taxes and led to an organizational change in headquarters and provincial offices. The reorganization was aimed at functionalizing business processes in the tax administration in accordance with international practice. A typical provincial tax office now has the following divisions: (i) taxpayer services and information, (ii) tax return processing and accounting, (iii) collection of arrears, (iv) tax examination, (v) tax audit, (vi) Personal Income Tax administration, (vii) general affairs, (viii) internal inspection, (ix) organization and personnel, (x) administration and finance, and (xi) IT. The strategic management capacity of GDT is being upgraded. This requires, in particular, the further development of analytical and planning efficiency, such as building tax gap analysis and revenue forecasting capacity, and improving the framework and tools for monitoring field office performance and supervising operational work in provincial and district offices. Tax administration needs to apply various planning tools in order to design a clear strategy for responding to compliance challenges and provide guidance for operational work. A 76 TA X A D M I N I S T R AT I O N strategic planning process must be introduced and stakeholder involvement in the strategic planning process must be solicited to ensure wide acceptance of the plan. Operational and annual business plans will need to be prepared to ensure consistency in operations and implementation of the strategic plan. Capacity development will also require increasing the staffing level at GDT. While many modern tax administrations assign around 10 percent of their staff to headquarters functions, only 1.1 percent are assigned in Vietnam. This is not sufficient to efficiently perform all relevant planning, monitoring, and analysis tasks. An assessment of GDT’s operational processes was conducted by reviewing 15 functional aspects of tax administration performance (figure 3.1). Overall performance is below average in all main and support functions scored in accordance with an international best practices research scale, with supporting data for the framework gathered from 30 developing and developed country tax administrations. Among them, the debt collection and enforcement function turned out to be the area with the most advanced processes in place. Multiple payment methods (cash and bank transfer), the use of technology to access taxpayer account data combined with semi-integrated collection systems to monitor taxpayer delinquency, are among the key factors contributing to this advanced function. In the legal area, provisions for the counseling and advising of GDT staff on legal issues exist, so that at least basic support is available to help tax administration staff deal with more complex legal matters. However, the legal support is not yet comprehensive, and additional processes to perform critical legal tasks, such as creating the capacity to issue private letter rulings to taxpayers, still have to be developed. Figure 3.1: Staffing Proportion in Headquarters 50,000 40,000 30,000 20,000 10,000 0 Australia Malaysia Sweden Indonesia Vietnam Total tax administration sta ng Headquarters sta ng Sources: GDT 2010; OECD 2010. Existing processes are less satisfactory in most other areas, with the tax audit function being of particular concern. Procedures and methodologies for a risk-based audit selection have not yet been developed, and auditor discretion in the selection of cases for a desk or field audit is substantial. In addition, existing processes for the conduct of an audit provide substantial discretion to the audit team, and the overall audit planning and supervision process does not yet include a systematic and ongoing monitoring of audit efficiency. Anti- 77 TA X R E F O R M I N V I E T N A M evasion appears to be the weakest area, where no strategic plan for addressing evasion cases exists (figure 3.2). Figure 3.2: Operational Gap Pre-Filing Taxpayer Services Support Functions (Financian and 5 Filing - Return Human Resources Processing, and Tax 4 Accounting Function Management, and IT 3 Operations 2 Mangement-Strategic 1 Audit Function Planning, Perf Mgmt, QA, Communications 0 Legal A airs (Part of Debt Collection & Collection & Enforcement Function Enforcement Function Anti-evasion Appeals (Activities (Activities Conducted Conducted by Internal by Taxpayer Audit Audit Function) Source: Based on an internal document of Booz Allen Hamilton 2007. Note: The outer line in the diagram indicates best practice performance. A critical challenge to be addressed rather urgently is the reengineering and streamlining of business processes, which is a precondition for the implementation of an integrated IT system. Appropriate risk assessment will need to be introduced to maximize operational efficiency. In a situation of scarce resources and limited capacity, tax administration operations must concentrate on high-risk taxpayers and operations. Auditing low-risk transactions with low revenue yield generally implies a waste of such scarce resources. Risk analysis capacity, therefore, must be created to determine tax audit and arrears collection priorities. The 673 largest taxpayers in the country contribute more than 70 percent of total domestic tax revenues. With just a few large taxpayers being of crucial importance for overall revenue collection, and considering the special tax avoidance and evasion risks in the large taxpayer segment of the taxpayer population, special capacity needs to be developed to reduce the avoidance and evasion risks. According to international good practice, specialized large taxpayer offices have been established in Vietnam to deal with these taxpayers. Staffing these offices with highly qualified and experienced tax officers and creating the necessary legal framework for dealing with tax avoidance and evasion schemes such as transfer pricing is of fundamental importance for compliance management. An indispensable requirement for operational efficiency is the capacity to collect and process a large quantity of information. Collecting and matching data permits the tax administration to detect unregistered taxpayers and tax evasion. IT support is becoming an extremely important factor and an integral part in all activities of the tax administration. There is considerable demand for IT support from all tax administration business programs. 78 TA X A D M I N I S T R AT I O N Therefore, GDT must establish a long-term ICT strategy to effectively support the future business model. New small and medium enterprises (SMEs) entering the tax net require information and services tailored to their needs in order to reduce tax compliance costs and facilitate business formalization. Compliance costs tend to be largely regressive and, therefore, put a special burden on SMEs. GDT needs to develop its taxpayer service function to be able to respond to the service needs of SMEs. The World Bank’s Doing Business report ranked Vietnam 124 out of the 183 nations surveyed in 2011 in terms paying taxes, a ten-grade improvement over the 2010 ranking. The costs of paying taxes in Vietnam include the following: Payments: This indicator measures the total number of taxes and contributions paid per year. In Vietnam, this figure is more than twice the average number of payments in OECD countries and significantly higher than the average in the East Asia & Pacific region (table 3.1). Time: This indicator reflects the time it takes to prepare, file, and pay (or withhold) the CIT, the VAT, and social security contributions (in hours per year). The time for complying with taxes was 1,055 hours per year, which was unchanged during 2006–10. It has been slightly reduced to 941 hours in 2011, remaining significantly higher than the average of the East Asia & Pacific region and OECD countries (table 3.1). Profit tax: The profit tax indicator measures the amount of taxes on profits paid by the business as a percentage of commercial profits, while the labor tax and contributions indicator reflects the amount of taxes and mandatory contributions on labor paid by the business as a percentage of commercial profits. Vietnam tends to tax profits at a lower rate than the average of OECD and East Asia & Pacific economies, while labor taxes are almost at the same level as OECD countries and are twice as high as in East Asia & Pacific region countries (table 3.1). Table 3.1: Paying Taxes in Vietnam East Asia & OECD Indicator Vietnam Pacific Countries Payments (number per year) 32 24.5 14.2 Time (hours per year) 941 218.2 199.3 Profit tax (%) 12.5 18.3 16.8 Labor tax and contributions (%) 20.3 10.3 23.3 Other taxes (%) 0.3 6.8 3.0 43.0 Bottom of Total tax rate (% profit) 33.1 35.4 Form Source: World Bank, Doing Business 2011. See box 3.1 for a discussion of the costs of complying with tax regulations in Vietnam. 79 TA X R E F O R M I N V I E T N A M Box 3.1: How Costly Is It to Comply with Tax Regulations? Assessments of the cost of complying with taxes can be sensitive to the methodology used. Three recent attempts to measure such cost in Vietnam yielded substantially different results. Investment Climate Assessment. Conducted by the World Bank in 2005, this assessment surveyed a representative sample of firms in manufacturing across 25 provinces in Vietnam. The focus of the survey was not to estimate the tax burden, but rather to assess the broader investment climate. However, one of the questions asked referred to the time spent dealing with tax inspections, including preparation of relevant documentation. According to the results, firms were subject to an average of 2.5 inspections per year, with each lasting 9.5 hours, on average. From these figures, it could be inferred that firms need to spend about 24 hours of work to comply with taxes. But tax inspections, no matter how burdensome, are only part of the overall compliance cost. Therefore, the investment climate assessment produces an underestimate of tax compliance costs. Doing Business. This is an attempt to quantify various dimensions of the overall business environment across countries. One of those dimensions is the costs of complying with taxes. In this respect, the Doing Business report relies on expert assessments for a hypothetical enterprise with 60 employees and a turnover equivalent to 0.7 million dollars per year (in Vietnam’s case). During 2005–07, the time needed for this enterprise to be current in its obligations was 1,055 hours per year. This estimate refers to the CIT, PIT, VAT, and labor taxes. It includes the time to prepare tax returns and the time required for bookkeeping. However, the latter may be incurred anyway, regardless of tax payments. Also, few enterprises pay the PIT in Vietnam, and the estimate assumes that electronic filing is not used. Therefore, the Doing Business report overestimates compliance costs. Central Institute for Economic Management. This evaluation was conducted in 2007 using a “standard cost modelâ€? developed by the Dutch government. The model relies on 360 interviews with SMEs. In the case of Vietnam, the average time to comply with taxes was estimated at 1,959 hours. Most of this cost (1,733 hours) was associated with the VAT, which respondents seem to consider particularly annoying. However, this evaluation mixes the one-time cost of obtaining a Taxpayer Identification Number with the recurrent cost of paying taxes. Also, the methodology makes it difficult to distinguish between activities that are part of doing business, activities specifically related to the payment of taxes, and more subjective “annoyance costs.â€? A closer analysis shows that the estimate for the time spent on inspections is four times higher than in the Investment Climate Assessment. The cost of dealing with most taxes is twice as high as in the Doing Business report, and the cost of dealing with the VAT about 30 times higher. Therefore, the Central Institute for Economic Management study is likely to have vastly overestimated tax compliance costs. To avoid similar discrepancies in the future, a consistent methodology to monitor tax compliance costs and service delivery should be developed. Such methodology should be anchored on representative samples and well-tested questionnaires. Estimates should be produced on a regular basis, focusing on the costs (time and money) faced by different groups of taxpayers. Similar surveys of tax officers would allow for a more complete picture. Key metrics should include not only the cost of complying with taxes, but also indicators related to the efficiency of administration, the integrity of its officials, and the taxpayer satisfaction. Source: World Bank 2008. 80 TA X A D M I N I S T R AT I O N 3. Ensuring Compliance in a Changing Environment Like many developing countries, Vietnam faces the challenge of fighting substantial underground economy activities from the compliance management standpoint. Schneider, Buehn, and Montenegro (2010) calculated that in 2007 in the East Asian region, the shadow economy accounted for an average of 32.3 percent of GDP. Vietnam is among the countries less affected by an informal sector. With an estimated shadow economy of 14.4 percent of GDP, Vietnam is close to the top performers in the region (compared to Japan with 10.3 percent of GDP, China with 11.9 percent, and Singapore with 12.2 percent), and is even below the OECD country average of 17.1 percent of GDP. Although these figures do not give direct evidence of the level of tax evasion and the size of the tax gap, they are encouraging and indicate that a major informal sector operating completely outside the tax net has not yet developed. However, this does not necessarily reflect above-average tax morale of Vietnamese taxpayers, or a particularly efficient GDT compliance management strategy. Rather, it needs to be attributed primarily to the still early stage of development of the country’s market economy. The efficiency of the tax system in Vietnam in this respect is not fundamentally different from that found in other countries at the same level of transition. The total number of registered taxpayers in Vietnam is low for a country with a population of more than 86 million. Major taxpayer segments include around 500,000 enterprises of all forms, 1.7 million business households, and 11 million individuals (table 3.2). Table 3.2: Registered Taxpayer Population: An International Comparison Personal Corporate Value- Registered Number Labor Income Tax Income Tax Added Tax PIT Payers Country of Citizens Force Taxpayers Taxpayers Taxpayers / Labor (million) (million) (million) (million) (million) Force (%) Argentina 40.13 16.4 0.93 0.29 0.92 5.7 France 62.63 28.51 36.4 1.6 4.2 1,247.7 Germany 81.9 41.7 26.8 2.5 5.7 64.3 Japan 127.51 66.17 23.69 3 3.63 35.8 Korea 48.75 24.39 5.23 0.44 5.34 21.4 Mexico 107.55 45.4 23.6 0.9 4.9 52 South Africa 49.32 17.38 5.54 1.83 0.74 31.9 Turkey 71.9 23.81 14 0.6 2.3 58.8 Vietnam 86.02 47.74 9.0 0.5 2.2 22.7 Sources: OECD 2010l. Data for Vietnam: GSO and GDT 2009. This demonstrates that the level of private entrepreneurial activity in the country is still at a rather moderate level. The registration of private businesses, mainly SMEs, is on the rise, however, and the registered enterprises in 2010 - 430,000 - is five times higher than it was in 2006. From a compliance management standpoint, the transition to a market economy and the rapid development of a private sector creates two major challenges. First, ensuring large 81 TA X R E F O R M I N V I E T N A M taxpayer compliance will increasingly require dealing with private business operations instead of SOEs. Simultaneously, the international dimension of business operations is of growing importance. This is not only reflected in the growing level of foreign direct investment in Vietnam, but also in cross-border business transactions of Vietnamese enterprises. With access to high-quality tax advice now available in the country (all major international tax consulting firms operate branch offices in Vietnam), GDT will be confronted with the use of sophisticated tax avoidance schemes, including international transfer pricing operations. Second, an emerging active medium and small segment of the taxpayer population requires the development of new compliance management strategies. It also poses capacity-building challenges for GDT. International experience shows that the development of a market economy will not affect the importance of large taxpayers for tax revenue collection. Even with a booming SME segment of the economy, a small number of major enterprises remain the backbone of the revenue system (figure 3.3). However, these large taxpayers will increasingly consist of private businesses instead of SOEs. Ensuring the compliance of large enterprises, therefore, must remain a top GDT priority. This requires building capacity to monitor large taxpayer operations and reducing the tax compliance burden (box 3.2). Figure 3.3: Taxpayer Segmentations: Population vs. Revenue Contribution Proportion of Proportion of tax revenue registered taxpayers contribution Large taxpayers > 1% 70+ % Medium taxpayers 5–25% 10–25% 0–10% Small taxpayers 70–95% Source: IMF calculations. Box 3.2: Developing a Compliance Management Strategy One of the core tasks of tax administration reform is the development of GDT capacity to manage compliance and address compliance risks. The overall level of tax compliance in a country is determined by a number of factors. These are partly business related and partly reflect the economic and social environments. The educational level of business operators, the general tax morale in the country, and fear of the tax authority can also be important factors in this context. Based on the general impact of external factors on the tax compliance attitude, combined with personal characteristics (for example, level of risk aversion) of the individual taxpayer, the following four levels of taxpayer compliance attitudes can be distinguished: 82 TA X A D M I N I S T R AT I O N Factors in uencing taxpayer behaviour A itude to compliance Compliance strategy Have decided not to comply Use the full force of the law Business Industry Industry Don’t want to comply, Deter by detection Taxpayer Taxpayer but will if we pay a ention Sociologica Economic Try to but don’t Assist to comply Psychological always succeed Psychological Our strategies aim to create Our strategies aim to create Willing to do pressure down pressure down Make it easy the right thing 1. The disengaged: At the top of the pyramid is an attitude of disengagement. It characterizes those who have decided not to comply. People with this attitude either deliberately evade their responsibilities or choose to opt out. 2. Resisters: The attitude of resistance characterizes active confrontation. The system is seen as oppressive, burdensome, and inflexible. This attitude characterizes those who do not want to comply but who will if they can be persuaded that their concerns are being addressed. 3. Triers: Those who are basically willing to comply have a more positive attitude. But they also have difficulty complying and do not always succeed. They may have difficulty understanding or meeting their obligations, but their expectation is that, in any dispute, trust and cooperation will prevail. 4. Supporters: The attitude here is one of willingness to do the right thing. There is a conscious commitment to support the system and accept and effectively manage its demands. There is an acceptance of the legitimacy of the role of tax officers and a belief that they are fundamentally trustworthy. Opinion surveys reveal that compliance management in Vietnam can build on relatively high tax morale and a widespread feeling that tax evasion is unethical. This indicates that the major part of private business operators in Vietnam should fall under the categories of supporters and triers. In this case, one could expect a direct and visible impact of compliance facilitation initiatives on the actual level of voluntary tax compliance. Unclear is the percentage of newly emerging private businesses, which basically resist complying with tax obligations. Noticeable improvements in the compliance behavior of resisters would require a broader set of measures, comprising, in particular, strengthening tax enforcement capacity and developing programs to change their overall attitude toward government. This goes beyond the competence and responsibility of GDT and calls for a governmentwide response to noncompliance. The development of a compliance management strategy, therefore, requires a thorough analysis of compliance attitudes and compliance problems (obstacles). Such information is not yet available in Vietnam. Efforts will, therefore, need to be made to monitor on an ongoing basis the scope of, and reasons for, noncompliance. The tax administration reform program currently does not put much emphasis on such analytical work. Further analytical capacity development and the creation of a responsible organizational unit to carry out such analytical work should be envisaged in the next phase of the reform program. Source: Based on OECD 2004. 83 TA X R E F O R M I N V I E T N A M 4. Dealing with Small and Medium Taxpayers GDT faces completely different challenges in the growing importance of the SME segment of the taxpayer population. In particular, small taxpayers with high mobility and low visibility can be a major burden for a tax administration. Experience in Eastern Europe and the countries of the Former Soviet Union has demonstrated that with the deepening of the privatization process and the emergence of a large number of newly created SMEs, there is the risk that the performance of a tax administration will decrease. For example, there is the risk that with increasing privatization and the rapid multiplication of the number of new taxpayers, collection results will drop. While the voluntary compliance of new taxpayers is low due to a lack of an established tax ethos in the country, enforcing a high level of compliance fails due to the low enforcement capacity of the tax administration. Such developments are not without impact on the revenue system. According to the assumptions of the Vietnam Development Report 2006, failure to ensure small business tax compliance could result in a decline in total tax revenues of 7 to 9 percent within a decade. Compliance management is particularly difficult in the micro and small business segment of the taxpayer population. Data for 2006 show that of the 3.3 million individual business establishments in Vietnam, only 39.4 percent were business registered and only 37.3 percent had a tax registration. Nonincorporated small businesses with less than 10 employees in practice are free to decide whether to keep books for the determination of their profits and turnover, or to be taxed on a presumptive basis. There are no incentives provided for improving record keeping among small unincorporated businesses, and the presumptive tax regime is quite advantageous in practice. Not surprisingly, therefore, empirical evidence indicates that there is substantial abuse of the presumptive regime by larger businesses. In addition, the lack of a VAT registration threshold increases compliance costs even for microenteprises and results in considerable administrative resources spent on collecting taxes from businesses with low revenue potential. For the administration of a large number of small businesses, administrative structures and processes must be designed in a way that reduces administrative costs and resource needs to a minimum in order to guarantee the efficiency of tax administration and ensure the proper allocation of scarce expert resources to taxpayer segments with higher revenue yield. By offering different tax calculation approaches to determine taxable income and value added in different regions, the tax administration is currently trying to align the presumptive tax liability as closely as possible with the standard regime. However, the result is a highly complex calculation of liabilities, which is beyond the capacity of many taxpayers. It is thus not surprising that the majority of business households opt for an assessment of tax liabilities by tax district office officials instead of self-assessment. Moreover, calculation errors are frequent among the small number of households attempting to self-assess liabilities. As a result, the presumptive assessment of PIT and VAT liabilities is time and resource intensive for the tax administration and problematic for many small businesses. In particular, the estimations of turnover levels done by the district audit departments in cooperation with communes and the subsequent calculation of liabilities consume considerable administrative resources and lengthen the assessment process, involving 84 TA X A D M I N I S T R AT I O N multiple steps between business households, local GDT departments, and the tax advisory council. The administrative processes for presumptive taxation result in extensive direct contact between taxpayers and tax administrators, which increases compliance costs and the risk of collusion and corruption. In Hanoi, for example, business households indicated that they received on average of seven visits from tax officials, and 25 percent reported that they had to go through a negotiation process with tax officers to determine the amount payable in the settlement for 2006.20 To reduce administrative costs, unnecessary interactions, and the related compliance burden for small businesses, the administration of the presumptive tax system needs to be fully based on self-assessment of tax liabilities. This, of course, requires a dramatically simplified calculation method for small businesses, which needs to be based on a size-based differentiation of bookkeeping standards and tax treatment. 5. The GDT Reform Initiatives Despite continuing efforts to establish a revenue management system in accordance with a market economy since the launch of Doi moi (economic renovation) process in late 1980s, Vietnam for the first time set up a formal five-year plan for tax administration reform in response to the need to prepare the five-year Socio-economic Development Plan (SEDP) 2006–2010. The main objectives of this plan were to develop a modern, fair, and transparent revenue collection agency, to promote voluntary compliance through a balance between high-quality taxpayer services and effective enforcement, and to enhance revenue collection. The implementation of the reform plan has, to a large extent, aligned the tax administration structure and operations in Vietnam with international good practice. The June 2007 special law on tax administration creates a transparent and comprehensive legal basis for tax administration, including clear definitions of rights and responsibilities of taxpayers and tax inspectors, ensuring sufficient access of the tax administration to information while guaranteeing tax secrecy, and clearly defining enforcement powers. The introduction of self-assessment, initially piloted for the CIT and the VAT in Ho Chi Minh City and Quang-Ninh in 2004 and rolled out nationwide in 2007, led to the restructuring of the tax administration along functional lines, combined with the creation of special administrative structures for large taxpayers. Coupled with piloting reorganization were the following: (i) the establishment of a dedicated taxpayer service function in all tax offices; (ii) the strengthening of the arrears collection function, with newly occurring arrears being reduced to 5.5 percent of total tax collection in 2010, the gradual introduction of e-filing, which in 2010 was rolled out to 19 provinces (the number of businesses filing electronically, however, was still low, with 7,200 enterprises using the option in 2010); and (iii) the reform of the tax payment system with a gradual move to tax payments made by bank transfer. For human resource development, GDT implemented a major staff training and capacity- building program. As part of this program, the percentage of staff with a tertiary or higher degree has been increased from 44.7 percent in 2006 to 53.1 percent in 2010. 20 Tax Compliance Cost Survey, internal document, World Bank, 2007. 85 TA X R E F O R M I N V I E T N A M A tax processes simplification program was launched as part of the governmentwide Master Plan to Simplify Administrative Procedures in the fields of State Governance (“Project 30â€?) (box 3.3). Box 3.3: Examples of Tax-Related Simplification Proposals as Part of Project 30 Regulation now enables businesses to print their own invoices. Businesses no longer need to receive approval from the Tax Authority for self-printed invoices when issuing and using them; rather, they just need to inform the Tax Authority. The purchase of VAT invoices issued by the Ministry of Finance is only applicable to newly established businesses and to small enterprises that do not have printing means. VAT declarations. Small and medium enterprises will declare their VAT every three months and large enterprises will do so monthly. All firms currently declare their VAT on a monthly basis. More banks (including joint stock commercial banks) will be allowed to engage in the tax information system and tax collection. VAT refund procedures. Business risk criteria will be published so that businesses can define what VAT refund category they belong to, making the VAT refund process easier and more transparent. The period for the VAT refund procedure in the category of “getting VAT refund before examination of tax refund applicationsâ€? was shortened from 15 working days to 6. The time for the VAT refund procedure for the category of “examination of the tax refund applications before getting VAT refundâ€? was shortened from 60 working days to 40, according to the Ministry of Finance. Source: OECD 2011. A new Tax System Reform Strategy was approved by the Prime Minister for 2011–20.21 The strategy aims at continuing and deepening the tax administration reform process and addressing on a priority basis remaining challenges and operational weaknesses. Three areas are of particular importance in this respect: • Strengthening the tax audit function: Progress in modernizing the tax audit function has been less impressive than reform results achieved in other areas. Expert staff allocated to the audit function is insufficient, with only 20 percent of total staff working in the audit area, which is far less than the international average of 30 to 35 percent. Audit selection is largely done on a manual basis, with little consideration given to risk analysis. Pilot exercises to develop a risk-based audit selection system have started and will result in the implementation of an automated audit selection system. According to the performance objectives of the reform strategy 2011–2020, at least 90 percent of audit cases willl be selected based on an automated risk management system by 2015. • Ensuring sufficient IT support: The current tax administration IT system does not have sufficient capacity to fully support and link all operational functions and procedures and to process a rapidly increasing quantity of data. Consequently, a special initiative to develop IT capacity for the processiong of PIT returns had to be launched in combination with the introduction of a universal PIT. Realizing the full benefits of operational modernization 21 Decision of the Prime Minister No. 732/QD-TTg, May 17, 2011, approving the Tax System Reform Strategy 2011–2020. 86 TA X A D M I N I S T R AT I O N and ensuring the availability of e-services for taxpayers will require the introduction of a new Integrated Tax Administration Information System (ITAIS). The requirements for such a system have been prepared and the system is expected to be fully operational by 2016. In combination with the ITAIS implementation, business processes will have to be reviewed and reengineered. • Introducing a proactive compliance management approach: GDT has not yet started to develop special targeted programs to address compliance risks and behaviors of individual taxpayer segments. Filing and payment noncompliance, and deficiencies in keeping proper documents and records, will have to be addressed more proactively. This requires further strengthening the compliance analysis capacity and moving to a more taxpayer- segment-oriented approach of compliance management. The 10-year Tax System Reform Strategy set a target that taxes and fees (excluding revenues from crude oil and export revenues) should constitute 70 to 75 percent of total state budget revenues by 2015, and 75 to 80 percent by 2020, compared to the current level of 62.5 percent in 2010. An ambitious program supported by the World Bank to modernize the tax administration is underway to improve governance in tax administration and raise the level of voluntary compliance with the tax system. Through this program, government resources, combined with financial and technical assistance from donors, will be mobilized to implement a comprehensive reform plan to improve the business climate, sustain revenue collection, and reduce corruption. The tax administration modernization program is expected to be one of the key pillars of the business-led development strategy and to provide a significant contribution to improving key public services and state governance in the next five years. By 2015, Vietnam’s Tax Agency is expected to rank among the top five in the East Asia region in key performance areas, including but not limited to enhanced efficiency, increased transparency and accountability, reduced compliance burden, increased taxpayer satisfaction, and improved compliance with the legal system. 6. External Assistance A Tax Administration Modernization Project (TAMP) was prepared in 2007, with technical and financial support from the World Bank. The project aims to assist GDT to strengthen its governance in tax administration and to increase the level of voluntary compliance with the tax system by improving the effectiveness, efficiency, transparency, and accountability of the tax administration. These reforms are expected to have a positive impact on the business climate by developing predictable and enforceable tax administration processes and procedures. The sustainability of revenue collection and greater equity will also be supported through implementation of this project. The project’s development objectives, therefore, are complementary to two higher-level objectives of the country’s four reform pillars, namely business-led development and strengthened governance systems (figure 3.4). 87 TA X R E F O R M I N V I E T N A M Figure 3.4: Tax Administration Modernization Objective and Indicators Key Indicators Project Development Objectives Key Indicators Strengthened Accountability Introduction of Taxpayer Increased Transparency Service Culture Improved Increased Reduced Corruption Reduced Compliance Burden Administration Voluntary Improved Interaction with Capacity Enhancement of Taxation Compliance Stakeholders Improved Interaction with Improved Compliance with Stakeholders Legal Systerm Improved Compliance with Enhanced E ciency Legal System Improved Business Climate Enhanced E ciency Sustain able Growth in Revenue Collection Reduced Corruption Higher Development Objectives Business-led Strengthened Development State Governance Source: Based on World Bank-funded Tax Administration Modernization Project Document. This reform program, technically and financially supported by the World Bank, consists of the following four main components: • An institutional development component, which supports the implementation of the key requirements and instruments for good governance in the tax administration. It will (i) increase transparency and accountability of the administration, (ii) foster the efforts of the GDT to counteract corruption, (iii) improve the legal framework for tax administration operations, and (iv) build capacity to analyze compliance trends. Outreach activities and the promotion of stakeholder participation in the reform process will create external support for the institutional and operational modernization. • An operational modernization component, which will enable the GDT to ensure a high level of voluntary compliance with the tax system. Streamlining and simplification of processes, and the establishment of a service culture responding efficiently to taxpayer service and information needs, will reduce the compliance burden for taxpayers. In parallel, the activities under this component will increase the ability of the tax administration to detect tax evasion and collect the full amount of taxes due. Introducing new functional business processes will also increase transparency and reduce the discretion of tax officials. • IT development, which will primarily support the procurement and implementation of a proven ITAIS. In addition, three other subcomponents would be carried out in parallel with the activities and tasks for ITAIS, including procurement, installation, and testing of IT hardware and system software, e-tax applications, and a pilot data warehouse (box 3.4). ITAIS is closely linked to project Component Two since it facilitates an automation of the full functionality of tax administration, including but not limited to registration and deregistration, returns processing, payment processing, refunds processing, taxpayer accounting, assessments, compliance and delinquency control, collections, audit and inspection, appeals (objections), revenue accounting, and taxpayer services. In addition, 88 TA X A D M I N I S T R AT I O N ITAIS also supports all tax administration functions, including but not limited to case management, taxpayer notices, workflow, document control, data entry and importing, report writing, and system administration. The project management component supports advisory services and the necessary office infrastructure to assist the Project Management Unit in implementing all aspects of the project. It will also support activities aimed at developing and implementing appropriate strategies for managing organizational change at all levels of the tax administration, including communication strategies to explain the rationale and potential impact of proposed changes to all managers, staff, taxpayers, and other stakeholders. Coupled with strong commitments to the reform program and World Bank support, GDT also receives complementary assistance from other members of the donor community. The International Monetary Fund (IMF) provides assistance in developing a self-assessment scheme, improving the law on tax administration, and organizational restructuring of the tax administration. With IMF technical assistance, GDT has piloted self-assessment since 2004 and is currently rolling it out to provinces. The IMF is also supporting the reform of the PIT and the implementation of a strategic planning process. The International Finance Corporation (IFC) is supporting GDT in reforming the tax system for micro and small businesses, simplifying tax filing requirements, and introducing a risk-based approach for tax audit selection. Japan, through the Japan International Cooperation Agency ( JICA), provided a three-year program (2005–08) for the development of a training plan and training materials to support tax administration reform. The program has been extended and is currently in its second phase. A multidonor trust fund (MDTF) has been established to support basic training on tax administration functions, provide consultancy services on taxpayer accounting, conduct surveys on taxpayer service needs, and implement IT applications in self-assessment implementation. In addition, the U.S. Treasury provides technical assistance for capacity enhancement focusing on some tax administration functions, anticorruption strategy, and human resource development and management. This is a comprehensive modernization program, but not a risk-free one. A reform program that aims at reducing the opportunity for part of the tax officers to charge discretionary rents risks being seriously resisted. Implementing an overly ambitious multifaceted reform agenda, with interdependence between a complex computerization component and other required activities, and particularly given the current weak capacity of the system, would lead to delays in responding to timeliness and meeting targets of the reform agenda. Box 3.4: Computerization of Tax Administration Operations Computerization of tax administration operations is the backbone of the reform process. Modern tax administrations have made considerable investment in information technology not only to facilitate the processing of large numbers of tax returns and management of taxpayer accounts, but also to support the functioning of a self-assessment system and implement an automated, risk-based approach to audit selection and arrears management. This requires an integrated tax administration information system, which provides a common taxpayer accounting function across all taxes and is capable of matching data and cross-checking information from the various taxes and administrative functions. On average, IT-related costs amount to more than 11 percent of aggregate administrative costs in OECD countries. In non-OECD countries surveyed by the 89 TA X R E F O R M I N V I E T N A M OECD Forum on Tax Administration, IT-related costs have increased from an average of 6.6 percent of aggregate administrative costs in 2005 to the OECD level of 11.7 percent in 2009. Tax administrations have chosen different approaches for the implementation of computerization programs. While earlier computerization projects in OECD countries adopted a strategy of system development from scratch, more recently a number of comprehensive and proven integrated systems became available. Reuse of such systems requires a certain degree of system customization, but avoidance of the significant risks associated with in-house development, such as unqualified system developers, lack of cooperation between system developers and tax administration specialists during the development phase, or problems with system maintenance and updates. International experience has shown that, particularly for developing countries, acquiring a proven commercial off-the-shelf tax administration information system generally is a quicker and less risky option for tax administration computerization than attempting to develop bespoke solutions from scratch. Vietnam is following this approach and has started the procurement process for an off-the-shelf solution. The organization of a discovery workshop prior to the finalization of the bidding documents, which offered GDT the opportunity to collect additional information on existing commercial off-the-shelf solutions, provided valuable input into the procurement process. Tax administration computerization exercises risk focusing primarily on system procurement and installation. To guarantee the success of computerization it is vital, however, that the exercise is guided by a clear vision of the future business redesign and functional requirements coupled with an agreed-upon IT strategic plan, including expected requirements for IT architecture, tax applications, and data warehousing. The lack of such a comprehensive business reengineering vision and an IT planning process may inhibit the development of a full range of integrated services because component parts will not have been built with a future business redesign to guide them; reduce the take-up rates of e-services, because taxpayers will not get consistent, high- quality experience; and reduce the efficiencies that could be gained from understanding and bringing together future tax administration and taxpayer requirements. Additional risks should be foreseen and mitigated when implementing an integrated tax administration information system. The launch of modern tax administration information that reduces the opportunity for part of the tax officers to abuse their direct contact with taxpayers for charging discretionary rents risks being seriously resisted. The interdependence between a complex computerization component and other required activities and, in particular, given the current weak capacity of the system, would lead to delays in meeting reform targets in a timely manner. Significant organizational changes and new staff responsibilities and competencies required as a result of the business process reengineering and IT application may create unexpected institutional problems. There may be a shortage of IT staff with the required expertise to provide the ongoing required maintenance of the integrated system. Hence, to be successful, every tax modernization project must include a well-designed change management component and a plan for staff retraining and the additional hiring and training of the required IT support staff. Source: Authors based on OECD 2010 and Engelschalk, Weist, and Melhem 2000. 90 TA X A D M I N I S T R AT I O N 7. The Critical Issue of Reform Management and Sustainability Experience around the world demonstrates that the most important ingredient for effective tax administration is clear recognition at the highest levels of government of the importance of the task and the willingness to support good administrative practices—even if political friends are damaged (see Bird and Casanegra de Jantscher 1992). However, even with substantial government support, resistance to change will come from within and outside the organization. The reforms will dramatically change the way the tax administration of Vietnam conducts its management and operations, and the way it relates to taxpayers and other stakeholders. Support for the reform process from staff and external stakeholders is essential for the reforms to be able to implement the changes envisaged and achieve the results expected. Appropriate strategies for managing organizational change at all levels of the tax administration, including communication strategies to explain the rationale and potential impact of proposed changes to all managers, staff, taxpayers, and other stakeholders are essential to reduce resistance to change. GDT is, therefore, adopting an interactive process of project implementation, which, in particular, encourages the participation of managers and staff in the change process, pilots important changes to learn from experience, sequences change activities in a manner that minimizes disruption in normal workflows and smoothes the implementation of new processes and systems, requests stakeholder feedback to make strategic and tactical changes, and offers extensive training in change management. Five factors are considered critical for the sustainability of the reforms: (i) strong government ownership of, and a high level of commitment to, its reform agenda, including the Tax System Reform Strategy for 2011–2020; (ii) mobilization of private sector support and partnership; (iii) sound and quantifiable performance indicators and client standards introduced, so that any slippage in performance will be easily identifiable and tax officials can be held accountable; (iv) development of a sustainable, in-house capacity for continuous improvement following the implementation of the Reform Plan initiatives; and (v) application of a modern Management Information System (MIS) as a key business strategy to radically streamline operations by simplifying procedures, increasing transparency, removing opportunities for the unlawful use of official discretion, and introducing internationally agreed standards. Such changes, once implemented, will be extremely difficult, if not impossible, to reverse. Monitoring stakeholder satisfaction with reform results will be a particularly important tool to measure progress in achieving reform objectives and convince stakeholders of the benefits of the reform process in order to increase the level of support for further reforms. GDT has now embarked on piloting a feedback survey. An important contribution to increasing the sustainability of the reform process is the development of a new approach to monitoring the performance of the GDT. Certainly, revenue collections will remain a core criterion for performance monitoring. This is indispensable given that the prime task of tax administration is to collect all taxes due. But with more emphasis given to the quality of operations, the reduction of compliance costs, improvements in good governance, and cost efficiency, a broader set of performance indicators needs to be developed to permit wider performance monitoring regarding all aspects of tax administration efficiency and effectiveness. GDT is in the process of modernizing its results monitoring framework, and, in accordance with international practice, a list of performance 91 TA X R E F O R M I N V I E T N A M indicators has been developed within the framework of preparation of the TAMP and will be used by GDT management to measure its performance (table 3.3). Table 3.3: A New GDT Results Monitoring Framework Area Performance Indicators Improving • Perception of the integrity of staff as measured by periodic internal governance and and external stakeholder surveys increasing the • Public perception of the quality of tax administration performance level of voluntary as measured by periodic surveys compliance • Tax revenues paid voluntarily/total tax revenues collected • Number of active taxpayers • Compliance with major taxes: CIT, PIT, VAT Institutional • Perception of taxpayers and other stakeholders regarding the level development of honesty of GDT staff • Perception of managers and staff regarding the level of professional skills in selected functional areas Operational • Perception of taxpayers regarding the quality of services provided development by the tax administration • Percentage of written queries answered within 30 days • Operational costs/tax collected • Percentage of tax audits resulting in additional tax assessments • Additional tax collected as a result of tax audits/additional tax assessed through tax audits • Tax arrears recovered during the year/total tax arrears at the beginning of the year • Average time taken to process a VAT refund request • Percentage of taxpayers filing electronically • Average time required to settle administrative disputes Source: Based on World Bank-Financed TAMP document. 8. Can GDT Become a Model for Integrity Enhancement? Compliance behavior is closely linked to the credibility of the revenue collection agencies. Also, the business environment and tax compliance costs depend to a substantial extent on the efficiency and integrity of revenue collection. The Transparency International Corruption Perception Index 2010 lists Vietnam as a country with severe corruption problems (ranked 116 out of 178 countries surveyed). International experience shows that in a corrupt environment, the tax administration generally is one of the public sector institutions particularly affected by breaches of integrity. The Transparency International Global Corruption Barometer 2006 lists tax administration as the third-most-corrupt government agency worldwide (behind the police and the judiciary). The situation in Vietnam is not substantially different from the 92 TA X A D M I N I S T R AT I O N global picture, and recent surveys reveal a particularly high incidence of corruption in the GDT (table 3.4). Table 3.4: Which Are the Most Corrupt Government Agencies in Vietnam? Investment Climate Survey Diagnostic Study on Corruption Traffic police Land administration agency Customs department Customs department Tax department Traffic police Land administration agency Tax department Market controller Regulators in construction Construction permit authorities Construction permit authorities Import/export license authorities Health care Planning and investment agencies Regulators in transportation Economic police Source: World Bank 2006. Given the severe consequences of corruption in revenue administration for business development and for the state budget, it is not uncommon to select tax and customs agencies as pilot institutions for the implementation of anticorruption programs. Such an approach seems feasible as long as it is part of a longer-term and more comprehensive public sector integrity strategy. An isolated attempt to create an “island of integrityâ€? in tax and customs administration would, however, have a high risk of nonsustainability. In Vietnam, counteracting corruption has become a core element of public sector reform. The government is committed to counteracting corruption in the various government agencies, and various steps have been taken to reduce corruption possibilities, including the creation of an anticorruption committee and the introduction of asset declaration obligations for civil servants, as part of the law on corruption passed in 2005. A new National Strategy for Anti- Corruption to 2020 was promulgated by the government in 2009, defining basic and long- term objectives and precise steps, roadmaps, and responsibilities of individual government agencies for the implementation of a comprehensive integrity policy. Breaches of integrity are facilitated in various ways in GDT as a result of organizational and procedural deficiencies. Some of the most important issues are: • Business processes: A lack of clear and uniform business processes combined with a low level of computerization of tax administration operations increases the discretion of tax officials and rent-seeking opportunities. • Taxpayer-tax administration interaction: Direct contact between tax officials and taxpayers is not yet reduced to a minimum; in particular, administrative assessment practices require extensive interaction between taxpayers and tax officers. • Incentive system: A corporate identity and an incentive system rewarding correct behavior and efficiency are lacking. • Staff rotation: Human resources management practices reducing corruption opportunities, such as a regular staff rotation program, do not yet exist. 93 TA X R E F O R M I N V I E T N A M Reducing corruption in tax and customs administration requires addressing both motives and opportunities for corrupt behavior (table 3.5). Table 3.5: Key Issues in Anticorruption Strategy Addressing Motives for Corruption Addressing Opportunities for Corruption • Clear mission and vision statement • Transparent and clear legal framework • Performance-linked compensation • Functional tax administration organization packages • Limited contact with taxpayers • Competitive base pay • Arm’s-length, transparent, and • Transparent and nonarbitrary reward nondiscretionary business processes procedures • Transparent human resource, procurement, • Effective sanctions for corruption and budgetary processes • Stronger accountability to taxpayers • Computerization of tax administration through independent surveys processes • Independent internal and external audits • Citizen review and oversight • Internal anticorruption units Source: Das-Gupta, Engelschalk, and Mayville 1999. Business processes (combined with the computerization of reengineered processes) and human resource management are the two cornerstones of an anticorruption strategy in revenue administration. While business process reengineering can be accomplished within any given institutional environment, a number of countries have adopted an approach to grant greater autonomy to revenue agencies to facilitate the introduction of efficient human resource management techniques and incentive schemes. In the Asian region, the Internal Revenue Authority of Singapore is the most prominent, and worldwide one of the most successful, examples of a semiautonomous revenue agency. But granting greater autonomy is not necessarily the panacea for greater integrity in revenue administration, and the value added of autonomy can be minimal when the existing framework of rules and regulations provides sufficient flexibility for the introduction of an integrity-enhancing human resource management system, which seems to be the case in Vietnam. GDT is receiving full government support to adapt and upgrade its human resource base to meet the requirements of a modernized, function-based tax administration. GDT is also in a position to provide bonuses to high-performing staff, so that salaries can be increased by up to 1.8 times the base salary. A number of steps have already been taken by GDT to implement some of the key components of an anticorruption strategy. Further reforms are envisaged as part of the TAMP. The expansion of self-assessment and the rollout of a functional tax administration structure are core elements of the integrity enhancement strategy. A gap analysis of business processes has been carried out, and the reorganization and computerization of business processes will be supported by the TAMP. Tax laws are being redrafted, and ambiguities and gaps in tax laws will be revised, so that tax administration discretion is reduced. Regular taxpayer feedback surveys will provide a monitoring tool for the success of the anticorruption measures. A dedicated internal audit 94 TA X A D M I N I S T R AT I O N union will be set up within the framework of the TAMP, and the TAMP will also support improving the operation of a hotline in GDT for reporting allegations of corruption. A short-term plan for the implementation of the government’s national anticorruption strategy in the tax administration was prepared by GDT and promulgated in September 2010. The implementation plan focuses, in particular, on strengthening human resource management aspects of counteracting corruption, such as developing codes of conduct for tax officers and improving staff rotation policies; reducing the direct contact between taxpayers and tax officials through, for example, expanding e-filing possibilities and revising procedures for tax audits; and increasing transparency in tax laws and regulations. 9. Stakeholder Involvement in the Reform Process Improving stakeholder relations plays a key role for compliance management and for deepening the success of the tax administration modernization process. In an environment that is characterized more by a cooperative relationship between tax administration and taxpayers and a high level of voluntary compliance instead of an enforcement-based taxation approach, an active stakeholder dialogue and a cooperative relationship with stakeholders need to be developed. Ministry of Finance (MoF): A prime stakeholder for tax administration is the MoF. Tax administrations generally are administrative structures subordinate to the MoF. This is also the case in Vietnam, with GDT being a government agency under the supervision of the MoF. Experience has shown that changing the accountability relationship between the tax administration and the MoF is important for tax administration modernization. Traditionally, accountability mechanisms are exclusively determined by revenue maximization considerations. Revenue collection targets are set by the MoF, and tax administration performance is determined by whether these targets have been met. Such a limited focus on revenue collection is not appropriate for a modern tax administration. Working toward the reform objective of developing a modern, fair, and transparent revenue collection agency implies improvements and high performance in areas that do not or do not immediately translate into higher collection results. It can even be that in the transition stage, actual decreases in revenue collection results occur. In Vietnam, for example, in the first year after the introduction of self-assessment, collections from self-assessed taxpayers fell in absolute terms compared to the administrative assessment results achieved in the previous year. An exclusive focus on collection results does not fully assess the level of improvements and the achievements realized in other crucial reform areas, such as taxpayer services or efficiency of the appeals system. The reform of the accountability mechanism is, therefore, an essential requirement for facilitating the reform process and supporting its sustainability. Key criteria used to measure the performance of a modernized tax administration could include (i) the quality and efficiency of taxpayer services, (ii) the level of voluntary compliance with filing and payment obligations, (iii) increases in the number of registered (active) taxpayers, and (iv) perception regarding the level of integrity in tax administration. Other government agencies: There are a number of government agencies with a substantial interest in tax administration operations and the tax administration reform process. Agencies of particular relevance are the customs directorate, the Ministry of Economy, and agencies 95 TA X R E F O R M I N V I E T N A M responsible for business registration. Cooperation with customs is particularly important to ensure the proper operation of the VAT; the Ministry of Economy has a stake in improving the investment climate and reducing tax-related barriers for business growth and formalization. Cooperation with regulatory agencies responsible for business registration should be sought to streamline and facilitate business registration processes. Involving these agencies in the reform process, therefore, is highly important. Consideration could be given to including some key stakeholders, such as the Ministry of Economy, in the steering group overseeing the reform process. For other agencies with a more narrow interest in tax administration operations, an ongoing dialogue should be established to ensure that the benefits from the tax administration reform process are maximized for all stakeholders and that the interests and constraints of other stakeholders are appropriately considered. Taxpayers: Facilitating compliance for taxpayers is a core reform objective. Voluntary compliance requires establishing a partnership between the tax administration and taxpayers. Relations with taxpayers, therefore, need to be structured in a new way (box 3.5). Box 3.5: Key Elements of a Modern Taxpayer-Tax Administration Relationship Following are the key elements of a modern taxpayer-tax administration relationship: 1. Apply tax laws in a fair, reliable, and transparent manner. 2. Outline and communicate to taxpayers their rights and obligations and the available complaint procedures and redress mechanisms. 3. Consistently deliver quality information and treat enquiries, requests, and appeals from taxpayers in an accurate and timely fashion. 4. Provide an accessible and dependable information service on taxpayers’ rights and obligations with respect to the law. 5. Ensure that compliance costs are kept to the minimum level necessary to achieve compliance with the tax laws. 6. Where appropriate, give taxpayers the opportunity to comment on changes to administrative policies and procedures. 7. Use taxpayer information only to the extent permitted by law. 8. Develop and maintain good working relationships with client groups and the wider community. Source: OECD 2001. GDT has planned several initiatives to improve the relationship with taxpayers. Modernizing taxpayer services is one of the main components of the Tax Administration Reform Plan. Activities include the establishment of taxpayer service centers and the development of new service and information programs. Regular collection of taxpayer feedback will occur through periodic surveys. An area that will require further attention in the future is the participation of taxpayers in the GDT strategic planning process. The development of a GDT reform plan should ultimately lead to a regular strategic planning process with the development of a multiyear strategic plan. The design of such a strategic plan, and its acceptance by stakeholders, will benefit greatly from an extensive dialogue with all major stakeholders, in 96 TA X A D M I N I S T R AT I O N particular, businesses, throughout the planning process. During plan implementation, regular stakeholder forums are a useful way to maintain the dialogue with the taxpayer community. Increasing the involvement of the private sector in the reform process will also have an impact on the accountability arrangements of GDT. The recognition that taxpayers are an important stakeholder of the tax administration logically leads to accountability relations between the tax administration and taxpayers. These are substantially different from the accountability relations between the tax administration and the MoF, since taxpayers’ expectations regarding the performance of tax administrations differ from MoF priorities and, therefore, lead to the development of a new set of performance benchmarks. In modern tax administrations, the establishment of taxpayer charters has frequently been a visible outcome of this process. See box 3.6 for how Singapore handles its accountability to its taxpayers. Box 3.6: Service Pledge – Inland Revenue Authority of Singapore We are committed to providing excellent service. You can expect courtesy, competence, clarity and convenience from us. Courtesy: We will be attentive and polite when we serve you. Competence: We will ensure that you are served by well-trained officers and our tax assessments are accurate. Clarity: We will provide clear and complete information to help you fulfill your tax obligations. Convenience: We will continuously seek improvements to make it simple for you to meet your tax obligations. We will respond to you in a timely manner: We will answer most of your telephone calls within 2 minutes. We will reply to most letters within 3 weeks. We will usually attend to you within 20 minutes when you visit us. We will usually make refunds to you within 30 days. We recognize your desire for excellent service. To help us deliver service to meet your expectations, we need your co-operation to: 1) be timely in filing your return; 2) give us accurate and complete information; 3) pay your tax on time; and 4) comply with tax laws. Source: Internal Revenue Authority of Singapore. 10. Summary of Key Findings and Recommendations The tax administration of Vietnam requires fundamental organizational and procedural reforms to address the challenges of managing compliance in a market economy. GDT prepared an initial reform program which was successfully implemented over the five-year time period 2005–10 and which has been updated with the preparation of the Tax System Reform Strategy 2011–2020. The major challenge for tax administration in Vietnam in the forthcoming years will be a rapidly growing private sector with a large number of SMEs supposed to enter the tax net. Appropriate compliance management strategies will need to be developed to ensure a high 97 TA X R E F O R M I N V I E T N A M level of voluntary SME compliance. Despite the growth in the SME segment of the taxpayer population, the major contribution to tax revenue collection will continue to come from the limited number of large businesses. With the privatization of large business operations and the decreasing role of SOEs for revenue generation, a highly specialized large taxpayer administration must be established. A rather centralized large taxpayer office with industry- specific information gathering and technical expertise would be an appropriate response to ensure proper compliance from the major taxpayers. Little information is currently available on compliance risks and the level of tax evasion in the medium and small segment of the taxpayer population. Creating analytical capacity in GDT to measure compliance behavior and the size of the tax gap should be envisaged. A special analytical unit in GDT headquarters staffed with highly qualified experts could take responsibility for such analytical work. The level of voluntary tax compliance and the efficiency of revenue collection are strongly influenced by the level of integrity in tax administration. Corruption surveys indicate that major efforts are needed to improve integrity in tax administration. This requires the effective implementation in the tax administration of the government’s anticorruption strategy and a commitment to stakeholders that integrity enhancement is a priority for GDT management. The major elements of such an anticorruption strategy will be implemented within the framework of the GDT reform plan and the TAMP. However, the human resources management element of integrity enhancement requires further ongoing attention. With the move to a modern, service-oriented tax administration, the accountability arrangements for GDT need to change. Accountability and results agreements should cover all operational areas and not be limited to monitoring collection results. Modern tax administrations are accountable not only to the MoF, but also to other stakeholders. This requires a transparent monitoring of service commitments to taxpayers. Starting with the GDT five-year reform plan 2005–2010, a regular, ongoing strategic planning process was introduced. The multiyear strategic plan outlining the challenges for tax administration in Vietnam in the forthcoming years, and actions required to address these challenges, should be regularly updated and supplemented with annual and departmental plans. The involvement of all stakeholder groups in the planning process, and in the process of monitoring the implementation of the plan, is of key importance. Conducting regular feedback surveys from taxpayers and tax administration staff is both an important step to deepen the accountability arrangements to taxpayers and a tool to support further development of the GDT strategy. GDT implemented a pilot taxpayer survey in 2008, and a survey on the implementation of the Tax Strategy 2005–2010 was carried out in 2010. Taxpayer surveys should be conducted on a regular basis in the future and should use a consistent methodology to ensure comparability of results. Results from the surveys should be made available to stakeholders. Dedicated administrative structures to deal with taxpayers outside the large taxpayer segment are not yet available. Following the establishment of a special large taxpayer administration, the potential benefits and risks of further taxpayer segmentation should be analyzed. This process is linked to further reforms on the tax policy side, however. With the planned 98 TA X A D M I N I S T R AT I O N introduction of a VAT registration threshold and modernization of the presumptive tax regime for small taxpayers, transforming the district office network into special small business tax offices responsible for taxpayers with a turnover below the VAT registration threshold, should be considered. Development and deployment of an integrated IT system based on reengineered business processes is a key element for securing the success of the tax administration modernization agenda. Strong government ownership of the reform agenda and leadership in its implementation, with prioritized activities monitored by quantifiable performance indicators and client standards, are among the critical factors for achieving sustainability of the modernization program. 99 Forecasting Tax Revenue: Methods And Tools IV By Gangadhar Prasad Shukla, Daniel Alvarez, Tuan Minh Le, and Duc Minh Pham 1. Summary Revenue forecasting is the starting point in the preparation of government budgets and is at the core of public sector budgeting. Well before the budget cycle starts, the revenue forecast should be in place. Revenue forecasts are crucial for avoiding unanticipated deficits that would have to be filled by unplanned borrowing, expenditure cut backs and ad-hoc revenue measures. Revenue forecasting has become all the more important since the advent of the medium term expenditure framework (MTEF) which is based on revenue predictability over the medium term. In addition to forecasting future tax revenues for budget preparation, forecasting techniques may be used for estimating revenue potential and monitoring tax collections, evaluation of economic and structural aspects of fiscal policy and for purposes of tax expenditure analysis. 1.1. Organizational Set Up and Forecasting Process The General Department of Taxation (GDT) has a Revenue Projection Department (RPD), headed by a director, and is in the process of strengthening its database and constructing revenue forecasting models. The officials of this department are recruited from among young university graduates with different academic backgrounds and there is no system of entry-level institutional training. Most of the job expertise is gained through self-study and on-the-job learning. The legal authority and responsibility for revenue estimation is laid down in the Budget Law of 2002 under the chapter “State Budget Estimationâ€?. It lays down the broad guidelines for estimation of revenues and expenditures. The GDT forwards its revenue projections to the Budget Planning Department of the Ministry of Finance (MoF). The Budget Planning Department discusses the forecasts with the Monetary Policy Department within the Ministry of Planning and Investment (MPI). This MPI unit also conducts its own revenue estimates. These two sets of figures are reconciled and an agreed set of revenue forecasts is sent to the Office of the Government (OOG). The OOG forwards the forecasts to the Budget Committee of the National Assembly (NA). The NA debates these forecasts and arrives at a final set of figures which are sent back to the Government. 1.2. Data Collection and Data Integrity The task of collection and maintenance of data is performed by the field offices of GDT. The information is maintained in electronic form. The electronic reporting by the field offices is supplemented through hard copies of taxpayer information and statistical reports submitted to the GDT. At present, there are no prescribed norms and practice for the verification or cross checking on the part of GDT for ensuring the quality and integrity of data. 101 TA X R E F O R M I N V I E T N A M 1.3. Present Status of Revenue Forecasting in GDT The GDT maintains the database and conducts revenue projections for the various taxes, for example, VAT, excise taxes, personal income tax, corporate income tax, and natural resource tax including oil revenues. The customs department is responsible for maintenance of the database and revenue forecasting exercise for the import and export taxes. In the month of May, the preliminary projections for the following year are made. These projections are mainly based on the growth rate of revenues from January to April of that year over the collections of the previous year in the same period. The growth rate of sales revenues of producer companies and distributors are also considered before arriving at the final growth rate figure. In the month of August, discussions are held with other ministries to get their feedback and then the preliminary projections are sent to the field offices. The provincial offices assess whether the projected figures seem to be of the right order and also if there is a gap between the growth rate of revenue collections and the growth rate of sales in which case they try to find out the reason behind this divergence. The revenue forecast is finalized by the month of November and submitted to the National Assembly for approval. This is the model applied to most tax revenues. Thus, the present practice of revenue forecasting is based only on the past years’ trend and therefore is rudimentary and subjective in nature. Currently no macro or micro simulation models are being used. 1.4. Revenue Forecasting Techniques and Modeling Three categories of models are mainly available for most taxes. (a). GDP Based or Macro Model GDP based or macro modeling is a relatively simple model with manageable data requirements and is based on the relationship between the tax base and the tax revenue. The underlying principle is that if there are no changes in tax laws, the increase in tax revenue will be a function of the increase in the tax base over time (GDP, income, consumption, value of imports etc.) and the elasticity of tax revenues with respect to the base. Thus, to apply this model, one has to estimate the elasticity of overall taxes or a particular type of tax using a regression analysis on time series data of tax revenue and tax base. If tax rates and/or tax base also change (discretionary changes), then the increase in revenue in that particular year would depend upon the normal increase in the tax base plus the impact of the discretionary changes. If one were to estimate the elasticity in the year of discretionary changes by the usual regression analysis, one would end up calculating buoyancy which includes the impact of discretionary changes also. Thus to apply the macro model, one has to separate the growth in tax revenues coming purely from an increase in the tax base and the increase coming from any discretionary changes. (b) Tax Specific Microsimulation Models Microsimulation Models serve two purposes: 102 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S i. Impact Analysis: They enable us to analyze the impact of a tax policy change on taxpayers, specific groups of stakeholders as well as the tax revenues. Thus one can have a scenario analysis where alternative scenarios of tax policies are examined to see how they compare and pick up the one that is best suited to the country’s needs. ii. Forecast Revenues: The models also enable a more precise forecast of revenues with or without any changes in tax policy or tax structure. For GDP based forecasting models, the outcome of the tax intervention is hard to model. (c) Monthly Receipts Model This model forecasts monthly receipts rather than actual revenue potential. It is not suitable for impact analysis and its scope is limited to the forecasting of expected receipts each month. However, it is a useful tool for allocating monthly targets of tax collection among the staff of the tax administration and monitoring their performance. In addition, it does take into account the seasonality in revenue receipts. The data requirement is minimal. Only the monthly receipts of the 12 months for the previous year, the monthly receipts for the month in the current year and the growth rate of the GDP as projected for the current year are required. 1.5. Strategy for Revenue Forecasting in Vietnam The strategy for revenue forecasting comprises database construction, capacity building of the officials concerned and actual construction of the revenue forecasting models. (a) Strengthening of Database The first step would be for the GDT to assign the task of database building to a team of officials within the RPD who should systematically go over the data available within the field offices of GDT, department of statistics, central bank, chambers of commerce and industries, research institutions, etc. and should collect and enter the relevant data into electronic formats. (b) Personnel Training The staff training should comprise a robust combination of theory and applications (exercises, case studies, presentations) and should be very much hands-on. The officials need two sets of training in the following forms: (1) tools training that should include training in econometrics package like e-views, STATA, advanced spreadsheet operations in Excel, and (2) substantive subjects which includes Economics of Taxation and revenue forecasting techniques. (c) Model Building The actual task of model building should be undertaken once the staff training and database preparation are complete. The building of the macromodels and the monthly receipts model may be taken up earlier as these are comparatively easy to make. (d) Timeframe The entire process described above should take about 18 to 24 months provided it is done in an organized and systematic manner. Once the models are ready, they should be calibrated 103 TA X R E F O R M I N V I E T N A M to correct any special features of the forecast. It should be noted that while the first cut of the forecasting models should be able to answer “what if â€? questions and yield a decent set of revenue forecasts, the process would be far from complete. It will take at least two more years before the database is complete and clean and the models have been fully refined. 2. Introduction Revenue forecasting is the starting point in the preparation of government budgets and is at the core of public sector budgeting. Well before the budget cycle starts, the revenue forecast should be in place. The forecasts should also be revised during the course of the year if some major parameters affecting tax revenues, for instance the GDP growth rate, undergo change. Both the initial and revised forecasts are crucial for avoiding unanticipated deficits in the course of a budget year that would have to be filled by unplanned borrowing, or expenditure cut backs and emergency ad-hoc revenue measures. Revenue forecasting has become all the more important since the advent of the medium term expenditure framework (MTEF) which is based on revenue predictability. Changing the psychology of the spending departments and agencies from a “needsâ€? to an “availabilityâ€? mentality is the most important feature of the MTEF approach to budgeting and therefore predictability of funds becomes significant to reallocating future budgets. A medium-term perspective is necessary because the time span of an annual budget is too short to adjust expenditure priorities, and uncertainties become larger over the longer term. Since MTEF is a multiyear exercise, revenue forecasts two to three years forward are necessary.22 At present, the General Department of Taxation (GDT) in Vietnam maintains the database and conducts the revenue forecasting exercise for total tax revenues, VAT revenue, excises, personal income tax, corporate income tax and natural resource taxes including oil revenues. The projections for the following year are based primarily on the growth rate of the current year revenues over the previous year’s revenue collections during the same period of time. This growth rate is suitably adjusted after looking at the growth rate of turnover of the relevant industries and sectors of the economy and after due consultations with line ministries and field offices of the tax department at the province and commune levels. Evidently, the present forecasting exercise does not employ any kind of econometric or microsimulation modeling for total tax revenues or for major individual taxes. It is simply the use of some sort of trend analysis. The quality of data is also poor. What is currently available from the GDT and its field offices mainly comprises revenue collections by tax type and by months. Consistent and integrated database on tax base and the underlying economic parameters that affect the performance of the various bases is virtually nonexistent. 22 In developed countries, the time horizon for revenue forecasting may be extended substantially. For example, the U.S. Congressional Budget and Impoundment Control Act of 1974 requires the Congressional Budget Office (CBO) to prepare detailed budget projections in an annual report to the Congress. The report contains projections of federal outlays and revenues for the current year and the next 10 years to be used as a budget baseline against which the effects of any proposed changes in tax and spending laws can be measured (CBO, 2001: Description of CBO’s Models and Methods for Projecting Federal Revenues). 104 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S Before getting into specific tax forecasting models, some general issues are examined concerning the relevance of revenue forecasting, the details of the organizational structure and the specifics of the revenue forecasting exercise in Vietnam compared with international practices. 2.1. Why Revenue Forecasting? Revenue forecasting may be utilized for various purposes. a. Forecasting Future Tax Revenues. This forms the basis of budget preparation. The main task is to assess the revenues available for forthcoming budgets. The budget cycle has the following components: i. Strategic planning—policy framework first ii. Macroeconomic framework iii. Formulate budget program: revenues before expenditures iv. Budget law v. Budget execution vi. Accounting & reporting vii. Audit & monitoring viii. Program evaluation & policy analysis ix. Start over again. This process has to start at the beginning of year zero to prepare budgets for years 1, 2, 3. Month 1 is the first month of year 1 and month 12 is first month of year zero. The following table 4.1 presents a desirable sequence. Table 4.1: Budget Cycle Time Activity 9 months • Macro forecast (medium term: years 1, 2, 3) • Revenue forecast and debt policy (medium term) • Determine revenue gaps: review and revise basic revenue policies based on such fundamentals as the size of government (increase/decrease expenditures), debt burden, and new tax policies (changes in structure, administration, incentives, distribution of burden). 6 months • Revised macro forecast • Revised revenue forecast including tax measure requirements to meet debt policy • Medium term expenditure ceilings: detailed budgets prepared (updated) for years 1, 2, 3 in consultation with the spending agencies 3 months • Refined medium term forecast with emphasis on year 0 and 1 revenues with final target for discretionary revenue measures for year 1. (The final forecast for year 0 changes the basis of the forecast for years 1, 2, 3.) 0 months • Present budget to the legislature Source: Authors’ own study of best practices in sound budget making in developed economies. 105 TA X R E F O R M I N V I E T N A M The budget making process should start with a macroeconomic and revenue forecasting exercise. The macro forecast predicts the economic growth rate which feeds into the revenue forecasting exercise by predicting the growth in tax bases (GDP, consumption etc.) and indicates the level of the desired public sector investment and expenditures. These two forecasts are also essential to determine the amount of debt required to meet the government expenditures as well as the need for deficit financing, if any. Once the revenue envelope is established, the expenditure limits should be decided, keeping in mind government policies and demands from line ministries and spending agencies. It is not uncommon, however, to see the entire process turned upside down in most developing countries and transition economies, where expenditure demands are first collected and then an effort is made to raise commensurate revenue amounts through different sources. This practice has the potential of increasing the debt burden and also putting excessive pressures on the tax administration to raise more and more revenues which may potentially destroy the integrity of the tax system. The suggested timeframe may vary somewhat from country to country but it is important to follow this sequence and timetable as closely as possible to avoid any ad hoc decision making in the budgeting exercise which may prove costly to the economy at a later stage. The main tasks of revenue forecasting comprise: • Computation of elasticity (and buoyancy) of various taxes and the entire tax system; • Forecasting revenues for each type of tax and impact of policy change on major criteria of a good tax system (for example, revenue adequacy, efficiency and equity); and • Evaluation of the effect of inflation and price changes on tax revenues. b. Estimating Revenue Potential and Monitoring Tax Collections. This is one of the most important functions of the Ministry of Finance and the tax administration and should involve the following set of activities: • Establishing a reliable database and an information system within the government; • Estimating the country’s tax capacity; • Identifying key tax handles with growth potential; • Analyzing the mix of tax and nontax revenue instruments and rationalizing them over time; • Exploring the possibility of charging user fees for public sector services wherever it is possible to exclude the nonpayers; and • Examining the level of tax effort and compliance. c. Evaluation of Economic and Structural Aspects of Fiscal Policy. One of the main purposes of having revenue forecasting models is to assess the economic impacts of various fiscal policies, and sometimes even of nontax economic policies. This will imply: • Projection of economic impacts of consumption taxes in terms of efficiency, incidence and equity; 106 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S • Provision of a framework for analyzing policy options for income tax, using a microsimulation model; • Assessment of the impact of changes in one tax regime on revenues from other taxes; • Scrutiny of nontax revenue sources and curtailment of “nuisance taxesâ€?; • Estimation of tax revenue impacts of alternative premises about the economy such as economic growth rate or growth rate of specific economic sectors; and • Measurement of the effect of macroeconomic policy changes on the tax regime and tax revenues such as (a) removal of quantitative restrictions on imports, (b) trade liberalization policies and their effect on import substitution and export promotion, (c) deregulation of certain economic activities, (d) currency fluctuations, (e) outcome of interest rate changes in terms of business profits. d. Tax Expenditure Analysis. The provisions of exemptions and concessions in tax laws tend to make the tax base narrower and the tax system less buoyant. Also, a concession to one party may set a precedent for others. These concessions and exemptions amount to spending from the budget and should, therefore, be spelled out clearly in the budget, for the sake of transparency and public debate. They involves the following: • Quantification of revenue impacts of: (a) exemptions, (b) incentives, (c) credits, (d) rate relief, (e) subsidies, (f) tax deferrals, and (g) tariff exemptions for selected items/ industries/free zones; and • Inclusion of tax expenditure in the budget for the sake of public debate and transparency. 2.2. Present Status of Revenue Forecasting in Vietnam The GDT already has a revenue projection department and is in the process of strengthening its database and constructing revenue forecasting models. 2.3. Organizational Structure The work of revenue forecasting and database maintenance for that purpose is performed by the Revenue Projection Department (RPD) within GDT. The officials of this department are recruited from among young university graduates with different academic backgrounds— economics, econometrics, commerce, finance, accounting, etc. There is no system of entry- level institutional training in taxation or revenue forecasting. Most of the job expertise gained is through self-study and on-the-job learning. The Revenue Projection Department has the following organizational setup. It is headed by a Director who is responsible for making estimates of tax collection as provided for in the Budget Law. The Director is assisted by three vice directors. The main duty of Vice Director I is building and setting revenue plans. Vice Director II is responsible for working on accounting, and tax statistics which are the essential elements of forecasting revenues. The main duties of Vice Director III are forecasting revenues and analyzing the impact of new tax policies on tax collections. There are 24 tax officials working in this department and they can 107 TA X R E F O R M I N V I E T N A M be assigned any task and responsibility by the Director. There is no specific distribution of work among this group of officials. At the provincial level, there is a forecasting unit manned by two to three people who are engaged in the task of revenue projection for each type of tax. Thus, there are 63 field units that work with the central revenue projection department. These units are equipped with computers for data maintenance and revenue forecasting task. Finally, GDT has its field offices at the commune or district level that support the tasks of tax collection and data gathering. 2.4. Legal Framework The legal authority and responsibility for revenue estimation is laid down in the Budget Law of 2002 under the chapter “State Budget Estimationâ€? (articles 37 to 49). It lays down the broad guidelines for estimation of revenues and expenditures at all levels. Enterprises will have their own production and business plans and estimates of the following year’s tax amounts which are used in forecasting of VAT and corporate income tax revenues. All budget spending units shall estimate their revenues and expenditures and their immediate superior managing agency will sum up the estimates. Similarly the customs department shall estimate their revenues and the provincial/ municipal Finance-Pricing Services shall be responsible for budget estimates of units under them. The Ministry of Finance is primarily responsible for synthesizing and estimating the budget revenues and expenditures and for coordination with the Ministry of Planning and Investment and other line ministries. 2.5. Reporting Procedure The GDT forwards its revenue projections to the Budget Planning Department of the Ministry of Finance (MoF). The Budget Planning Department discusses the forecasts with the Monetary Policy Department within the Ministry of Planning and Investment (MPI). This MPI unit also conducts its own revenue estimates. It should be noted that the MPI estimates are for reference only while the MOF has the prime role in the forecasting process. These two sets of figures are reconciled and an agreed set of revenue forecasts is sent to the Economics Department within the Office of the Government (OOG). The OOG forwards the forecasts to the Budget Committee of the National Assembly (NA). The Budget Committee comprises of technical experts who review the forecasts and submit them to the NA. Finally, the NA debates these forecasts and arrives at a final set of figures which are sent back to the Government. As these NA approved figures flow back down to the MoF and GDT, these become the revenue targets for the GDT for that particular year. 2.6. Coordination within MoF The Budget Planning Department in the MOF coordinates with the GDT on revenue forecasting. This Department makes its own ‘rough and ready’ revenue estimates for the following year and uses those figures to finalize the GDT forecast, but it does not have the 108 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S apparatus to conduct a detailed revenue forecasting exercise. That task is performed by the GDT and then the MoF makes only peripheral adjustments. 2.7. Linkage of Revenue Projection with the Budget Process The Budget Planning Department of the MoF is responsible for budget making. As outlined above, this Department is fully involved in the process of finalization of revenue projection figures in consultation with MPI. Thus the revenue projection exercise by the GDT fits in completely with the budget making process of the Government. 2.8. Data Collection, Data Maintenance, and Data Integrity The initial task of collection and maintenance of data is performed by the field offices of GDT situated within each province. The field offices have their own information department which is a blend of data processing and information technology (IT) departments. These two wings of the information department are in the process of being separated to become stand alone departments. The information department of the field office at the provincial level collects data on tax revenues, tax returns and taxpayers. This information is maintained at the field offices in electronic format. The IT Center of GDT has access to this electronic dataset and can utilize it as the database at the national level. The electronic reporting by the field offices is supplemented through hard copies of taxpayer information and statistical reports submitted to the GDT. It may be noted that while the field offices can input data in the electronic database, they cannot alter it. At present, there are no prescribed norms and practice for the verification or cross checking on the part of the GDT for ensuring the quality of data collected through the information departments of the field offices in the provinces. 2.9. Present Status of Database and Revenue Forecasting in GDT The GDT maintains the database and conducts revenue projection for the following set of taxes: i. Total tax revenues ii. VAT revenue iii. Excises (Special Consumption Tax) iv. Personal income tax v. Corporate income tax vi. Natural resource tax including oil revenues. At present some data is available on land related taxes along with the natural resource data but no data is available on property taxes. Also, the Customs Department is responsible for maintenance of the database and revenue forecasting exercise for the Import and Export taxes or Trade taxes (Tariff revenues). It should be noted here that actual returns and data are maintained at the field office levels and what is available with the GDT is simply a summary of the data. Therefore, any scrutiny 109 TA X R E F O R M I N V I E T N A M of actual tax returns and data available for revenue projections has to involve the field offices as well. The present practice of revenue forecasting for different kinds of taxes is outlined below. Value Added Tax: Step I, in the month of May, the preliminary projections for the following year (say 2012) will be made. These projections are mainly based on the growth rate of revenues between January to April that year (2011) over the collections of the previous year (2010) in the same period. The growth rate of sales revenues of producer companies and distributors between the previous year (2010) and this year (2011) are also considered before arriving at the final growth rate figure. Then the same growth rate is applied on that year’s (2011) estimated revenue figures to arrive at the following year’s (2012) projections. The VAT revenue projections are made by sectors – food, beverages, pharmaceuticals, etc. Step II, in the month of August, discussions are held with other ministries to get their feedback and then the preliminary projections are sent to the field offices. The provincial offices assess whether the projected figures seem to be of the right order and also if there is any gap between the growth rate of revenue collections and the growth rate of sales, in which case they try to find out the reason behind this divergence. Step III is the finalization of the revenue forecast by the month of November for submission to and approval by the National Assembly. Retailers do not pay taxes based on their actual sales but rather on estimates made by the tax department about their sales revenue, input costs and tax liability (using subtraction method). Distributors and producer companies submit their returns which are maintained by the field offices. Special Consumption Tax: The revenue projection for Special Consumption Tax is done separately for each category of goods subject to this tax. The revenue projection model is almost the same as that for VAT. Personal Income Tax: The revenue projection method is the same as in the case of other taxes like VAT. The personal income tax data is maintained under two categories: foreigners and Vietnamese nationals. Salaried persons do not submit tax returns but income tax is deducted at source by the employers and they submit a return at the end of the year. Self-employed professionals have to file tax returns and they are expected to file the returns at the field offices of the tax department. Corporate Income Tax: The revenue projection is done in a manner similar to other taxes described above. The only additional point is that the figures for sales and profits are cross checked with those for VAT to ensure that the two are consistent. Natural Resource Taxes: The revenues are projected, enterprise by enterprise, based on their production plans that outline the quantities of production and the expected prices in the future and their actual performance in the previous years. 110 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S 2.10. Data from other Sources The Department of Statistics publishes a series of documents that are relevant to the revenue projection task. In addition to a large amount of statistics on various sectors of the economy, it publishes data on population growth, investment in different sectors—state, nonstate, and foreign invested sector—and national accounts. All three are of relevance to the revenue projection task. 3. Conclusion The following three aspects become clear from the above description. i. The organizational set up for data collection and revenue forecasting and the necessary staff are in place. However, the staff has not been properly trained either in tax analysis or revenue forecasting. All the learning is on-the-job and therefore not methodical; ii. The present practice of revenue forecasting is based only on the past years’ trends and therefore is rudimentary and subjective in nature. Currently no macro or micro simulation models are being used; iii. Since the models are not in place, the necessary database has not been constructed in a scientific way. It should be noted that most developing countries are in a similar situation and are struggling with the problems of a lack of a good database, the absence of revenue forecasting models and a dearth of capacity and skill. 3.1. International Practice in Revenue Forecasting Developed countries such as Canada, Australia, New Zealand, the Netherlands and the US use both macroeconomic and microsimulation models for revenue forecasting and impact analysis of tax policy changes. In fact, many of the present day models were first constructed by the finance ministries or tax administrations in some of these countries. For instance, Canada was a pioneer in developing a microsimulation revenue forecasting model for Personal Income Tax which enables both revenue projection and impact analysis of policy changes. Similarly, Australia has developed an input-output based forecasting model for VAT which again performs the dual task of revenue projection and impact analysis. Since the late 1960s, major OECD countries and the US have developed “microsimulation modelsâ€? for their main taxes, particularly for personal and corporate income taxes. These models are data intensive and are constructed with the help of samples of tax return data. They have dual major objectives: (a) estimating the tax policy impact on revenue collections; (b) forecasting revenues for one or two years into the future. As regards the developing countries, few have fully developed models and most of them are in the process of building suitable models. Some of them are more advanced (for example, Malaysia, South Africa) than others (for example, Ghana, Sri Lanka). In the majority of developing countries, only the simple trend analysis is being used; some apply a preliminary form of GDP based modeling that relies on the concept of tax buoyancy instead of the 111 TA X R E F O R M I N V I E T N A M proper concept of tax elasticity.23 Recently, some countries such as Russia and Bulgaria have established and developed internal dedicated revenue forecasting functions within tax administration agencies as part of comprehensive tax administration modernization programs. The IMF has done a comparative study of forecasting practices in developing countries and transition economies based on a survey of 34 countries (Kyobe and Danninger 2005) from Africa, Asia, Latin America, the Middle East, and the CIS. Some key findings of this study are summarized below. • In most countries forecasting responsibilities are poorly defined and there are few formal rules and regulations for the revenue forecasting exercise. • Estimation techniques are generally rudimentary in nature. About 85 percent of countries use subjective assessment and simple extrapolation techniques as their main methods for revenue forecast. • The process involves multiple executive agencies, outside the ministry of finance, requiring a high degree of coordination. • It is common to have multiple competing forecasts. • Revenue forecasts are often produced late in the budget process, thus disabling the important link between revenue forecasting and budget making. • Public accountability in terms of access to forecast results and participation of nongovernmental agencies in the revenue forecasting exercise is limited. • These characteristics do not differ significantly along regional or per capita income levels. More transparency and rigor are applied in Latin American countries which also have higher income levels. • Higher formality and transparency are found in countries with a higher per capita income and lower corruption levels. 3.2. Revenue Forecasting Techniques and Modeling Two Categories of Models are mainly available for most taxes. A third type of model, referred to as “Monthly Receipts Modelâ€? that forecasts revenue receipts for each kind of tax instead of revenue potential, is also in use (King 1995, 254–57). (a). GDP based or Macro Model GDP based or macro modeling is relatively simple with manageable data requirements. This model is based on the relationship between the tax base and the tax revenue. The underlying principle is that if there are no changes in tax laws (tax rates and the statutory tax base do not change), the increase in tax revenue from year to year will be a function of the increase in the tax base over time (GDP, income, consumption, value of imports, etc.) and the elasticity 23 Both tax buoyancy and tax elasticity are measured as the ratio between the real growth rate of tax collection and the real growth rate of the tax base (for example, GDP). However, the elasticity of a tax system or an individual type of taxes measures the efficiency of the underlying tax structure and administration, excluding the impact of discretionary changes in tax rates and/or base. For macro-based forecasting purpose, one should use the concept of tax elasticity. 112 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S of tax revenues with respect to the base. Thus, to apply this model, one has to estimate the elasticity of overall taxes or a particular type of tax using a regression analysis on time series data of tax revenue and tax base. The next step is to forecast the growth in tax base and finally to forecast tax revenues with the help of tax elasticity and increase in the tax base. The forecast of growth in the tax base (GDP, consumption) can be simply extracted from the real sector (macroeconomic) forecasts. If tax rates and/or the tax base also change (that is, there are discretionary changes in tax rate and/or tax base), then the increase in revenue in that particular year would depend upon the normal increase in the tax base plus the impact of the discretionary changes made. Now one has to estimate the impact of the discretionary changes as well. If we were to try and estimate the elasticity in the year of discretionary changes by the usual regression analysis, we would end up calculating buoyancy which includes both the impact of economic growth and discretionary changes. If T0 is tax revenue this year, then T1, the expected tax revenue next year, can be linked to T0 and other factors as follows: T1 = T0 + Growth in Revenue without Change in Tax Structure + Change Due to Discretionary Measures Thus to apply the macro model, one has to separate the growth in tax revenues coming purely from an increase in the tax base and the increase coming from discretionary changes. While it is called the GDP based model, the tax base need not always be GDP and it could be payroll revenues or corporate profits or consumption or value/volume of imports. (b). Tax Specific Microsimulation Models In addition, there are tax specific models, sometimes referred to as microsimulation models that serve two purposes: Policy Impact Analysis: They enable us to analyze the impact of a tax policy change or change in tax structure on taxpayers, specific groups of stakeholders as well as the tax revenues. Thus one can have a scenario analysis where alternative scenarios of tax policies are examined to see how they compare and pick up the one that is best suited to the country’s needs. Forecast Revenues: The models also enable a more precise forecast of revenues with or without any changes in tax policy or tax structure. For the GDP-based forecasting models as described above, the outcome of the tax intervention is hard to model, especially for the PIT and CIT. Another category of micro simulation models combines tax modeling with economic modeling (for example, with sectors, types of consumer goods, etc.), provided that key economic performance indicators (for example, consumption, wages, salaries, corporate sector income and profits, etc.) and revenues are interdependent. This allows explicit feedback of tax impact on economic fundamentals. For practical reasons, these models are often handled separately. One model focuses on economic forecasting and on simulating policy changes, whereas the other(s) look specifically at tax revenue. Even though the models do not have to be automatically linked, they need to be produced in such a way that they generate consistent macroeconomic and revenue projections (See, for example, CBO 2001, p.5). 113 TA X R E F O R M I N V I E T N A M Clearly these models are more elaborate, and they need more extensive data. It also takes time to build them. But the models can and should be built by users in a transparent way and must not be “black boxesâ€? where the user does not understand what is happening. Once the models are functional, they are user friendly and effective. Two crucial aspects need to be taken into account: i. The time period analyzed: in an MTEF perspective, the time horizon for forecasting should typically be around three years to ensure a proper overview of the macrorisks in terms of budgetary outcomes (deficits) in the near future. ii. Cash versus accrual basis: The essential question is whether one deals with cash receipts or with the tax liabilities due in a given period. Typically, the macromodels forecast cash receipts, whereas the microbased models study liabilities or tax potential which may actually come to fruition some time (say, half a year) later and need to be corrected to arrive at the actual cash collections which are of most interest to the government. (c). Monthly Receipts Model There is a further category of models that forecast monthly receipts rather than actual revenue potential. It is not suitable for impact analysis and its scope is limited to the forecasting of expected receipts each month. However, it is a useful tool for allocating monthly tax collection targets among the staff of the tax administration and monitoring their performance. In addition, it takes into account the seasonality in revenue receipts. The data requirement is minimal. Only the monthly receipts of the 12 months for the previous year, the monthly receipts for the month in the current year and the growth rate of the GDP as projected for the current year are required. The three sets of models are fully explained in annex 1. 3.3. Strategy for Revenue Forecasting in Vietnam The strategy for revenue forecasting comprises database construction, capacity building of the officials concerned and actual construction of the revenue forecasting models. Steps Involved in Building of Revenue Forecasting Models for Vietnam Three main steps are involved in model building in Vietnam: strengthening the database with economic and tax revenue data, training of GDT staff from the Revenue Projection Department (RPD) and the actual creation of the models. The first two steps can be taken simultaneously while the third step is best undertaken once the first two are complete. It should be noted that these models can also be used for forecasting revenues at the provincial or commune levels provided the data is available separately for a specific province or commune. (a). Strengthening of Database The first step would be for the GDT to assign the task of database building to a team of officials within the RPD who should systematically go over the data available within the field offices of GDT, the Department of Statistics (at MPI), the Central Bank, chambers of commerce and industry, research institutions, etc. and should collect and enter the relevant data into electronic formats mentioned in annex 1. 114 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S While the model building exercise can start even when the database is in the process of being created and cleaned, it must be emphasized that the breadth, depth and integrity of the dataset will determine the ultimate quality of the models and their results. So this exercise is important, should be first to start and should be undertaken with due diligence and care. (b). Personnel Training The staff training should comprise a robust combination of theory and applications (exercises, case studies, presentations) and should be very much hands-on. Generally, half the time should be spent on theory, principles and demonstrations with the rest devoted to hands-on and computer based exercises and applications. GDT officials need two sets of training in the following forms: (1) tools training that should include training in econometrics packages like e-views, STATA, advanced spreadsheet operations in Excel, including the writing of simple macros and basic principles of micro- and macroeconomics used in tax policy analysis and revenue forecasting; and (2) substantive subjects which include Economics of Taxation (tax policy analysis) and revenue forecasting techniques. (c). Model Building The actual task of model building may be undertaken once the staff training and database preparation are complete. The building of the macromodels, which requires the estimation of tax elasticities on the basis of a clean dataset of tax and macro data, may start earlier if these data are readily available. The same holds for the monthly receipts model. This process should be completed method by method (first macro, subsequently micro), tax by tax, and in stages. (d). Time Frame The entire process described above should take about 18 to 24 months provided it is done in an organized and systematic manner. Once the models are ready, they should be calibrated to correct any special features of the forecast. It should be noted that while the first cut of the forecasting models should be able to answer “what if â€? questions and yield a decent set of revenue forecasts, the process would be far from complete. It will take at least two more years before the database is complete and clean and the models have been fully refined. But this is part of the normal model building process and not unique to Vietnam; every country has to go (or has gone) through the same route. 3.4. Revenue Forecasting Model and Results for Vietnam For illustration purposes, and given the limited data available, we attempt to apply microsimulation modeling to three major taxes, specifically, the VAT, personal income tax and corporate income tax using data from 2004–07. Value Added Tax To assess the impact on VAT revenue collection from the reforms suggested, a revenue estimation model was constructed. The model is divided in two parts. First, using Vietnam Input-Output Table data, the tax base was estimated by the extent to which goods and 115 TA X R E F O R M I N V I E T N A M services are covered under the VAT, following current tax code rules in terms of base and rate definition (The VAT Law 1997 and subsequent revisions as well as supporting legislation for the implementation of the VAT Law). Then, using alternative VAT rates and the new base with rationalized exemptions, an assessment of the impact on revenue collection was made. Data requirement The main data requirement is an updated Input-Output (I-O) table. If an I-O table is available for previous years, it needs to be updated by augmenting the entries in the ratio of GDP of the year of forecasting and the GDP of the I-O table year. If, however, the I-O table is more than five years old, this ageing of data may not yield optimal results. In addition, household surveys of income and expenditures and details of government consumption are needed. Growth rates for consumption as a share of GDP can be extracted from macroeconomic forecasts or simply by trend or regression analysis on past data. Methodology The tax base is estimated by employing the equivalence of the value added tax to a retail sales tax levied on the final selling price of goods and services. Thus, the main concept in the Input- Output revenue estimating model is to capture the total expenditure (personal, business, and government) that is incurred in the economy. While constructing the VAT base, it is necessary to arrange all the expenditures, including personal and government expenditures, intermediate and capital expenses, under the current tax laws and rules. The starting point for constructing the VAT base is the detailed information available for domestic consumption. Only those domestic expenditures in the final demand categories that are meant for personal and government use, not those for the further production of goods and services for commercial purposes, are considered as final consumption for constructing the VAT base. The amounts of personal and government expenditures are depicted in the final demand matrix of the I-O tables. Furthermore, since trade and distribution margins of retailers, wholesalers and transportation operators are part of the expenditures eventually paid by the final consumers; they should all be included in the VAT base. As final government expenditure is generally subject to VAT, the VAT base is simply calculated by adding domestic personal and government expenditures at retail prices. Wages and salaries paid to public servants are, however, not subject to VAT and hence are excluded from the VAT base. From detailed final consumption expenditure by goods and services, adjustments are made to subtract zero rated and exempt goods and services. When the VAT code zero-rates or exempts certain commodities, the full value of zero-rated goods and services has to be removed from the potential VAT base. Capital expenditures (with the exception of residential construction) are not considered part of the tax base, since they are used as business inputs into the production of final goods. Further adjustments have to be made to the VAT base in order to take into account the effects 116 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S of those cases in which tax-exempt sectors are not allowed to claim taxes paid on their business inputs and there is an element of cascading. The expected VAT revenue by commodity at 100 percent compliance can be calculated by multiplying the tax base with the taxable proportion and tax rates. In the case of Vietnam, separate estimation needs to be made for those commodities subject to a standard rate (10 percent) and those subject to nonstandard rates or a preferable rate (5 percent). In general, VAT revenues (R) at full compliance are estimated as follows.   R = ∑ Ci Ï?i + ∑ K j + ∑ Ï?i {Ei − Si } ∑ β m rm  i j i m Where, • Ci = before-VAT final expenditures (private and nonwage government consumption) of commodities in sector i; • Ï?i = proportion of the final consumption of commodities in sector i, subject to VAT (net of proportion of exempt or zero-rated commodities and services); • Kj = input purchase by exempt sector j; • Ei = inputs purchased by below-threshold small businesses in sector i; • Si = outputs sold by below-threshold small businesses in sector i; • β m = proportion of the final consumption (adjusted by the taxable proportion) of commodities subject to a positive VAT rate rm; and • r1,…,rm = positive statutory rates as stipulated in a country’s VAT code. Finally, the estimated VAT revenues at full compliance are to be adjusted by the existing compliance level to derive the projected VAT revenues. Data Available in Vietnam The 2005 Input-Output Tables are available for Vietnam at time of the calculation. The I-O Tables include detailed private consumption expenditures and immediate inputs – broken down into 112 categories. Household final consumption was obtained from the 2004 Income- Expenditure Household survey while the final Government consumption was only available for three items: science and technology, state management, and other remaining services. Growth factors from 2004 to 2006 were estimated based on the historical performance and future prospects for private consumption, government expenditures and business intermediate and capital expenditures. Model Results Under the VAT features as they stand today, assuming a GDP nominal growth of 16.7 percent for 2007, tax collection for FY 2007 was estimated to be VND 42,577 Billion, or 3.7 percent of GDP. Using the IO Model to forecast VAT collection, the following policy scenarios are also analyzed: 117 TA X R E F O R M I N V I E T N A M i. Base case scenario. Revenue collection for FY2007 is estimated, following the current base and rate structure. ii. Only one positive rate. A unified 10 percent tax rate substitutes the current structure of two rates—5 and 10 percent—and allowing zero percent rate only for export transactions. The current exemption structure is maintained. iii. One-rate structure plus exemption rationalization. Starting from Scenario 2, preferential treatment is allowed only for a few goods and service categories, specifically financial services, primary education, basic healthcare (not higher education and specialized health services), some cultural and merit goods and services, and aid financed activities. The rest of the exempt items were moved to the taxed category. The results are summarized in table 4.2. Table 4.2. VAT Policy Impact Scenarios, 2007 Billion Dong Percentage of GDP I. Base Case 42,577 3.7 II. 10% only positive rate 45,039 4.0 III. 10% as single positive rate and 50,359 4.4 exemptions rationalization Note: Estimated compliance rate of 74 percent for 2006 is applied in the 2007 VAT revenue projection. Option 3 refers to the broadest revenue enhancement scenario. With a unique positive tax rate, and a broad base, it is expected that an increase in the overall compliance rate will take place due to efficiency gains in tax administration and reduction in compliance costs. As can be seen, the rate of compliance is an important parameter in this model. This coefficient was calculated using actual collection for FY 2006 and the revenue potential estimated for that year. Using the dataset from private consumption expenditures from the latest I-O Tables and the Household Income-Expenditure Survey, the current Vietnam VAT structure was modeled for FY 2006 assuming full compliance by taxpayers. As a result, potential collection is estimated to be VND 48,979.6 billion (5 percent of GDP). The compliance rate is obtained from the ratio of actual VAT collection for FY2006—VND 36,469 billion—and the revenue potential for 2006, thus resulting in a compliance rate of 74 percent. This is the figure used in forecasts for FY 2007, assuming no major changes in the tax administration effort during 2006–07. Personal Income Tax The microsimulation model is used as an analytical tool to evaluate the impact of the reform proposals to the PIT structure in Vietnam at the individual level. The model aims to forecast personal income tax revenues and to assess the distributional effect of a given policy proposal on a particular segment of the population.24 24 Recently, a team of consultants applied an econometric model to analyze the revenue and incidence impact of the new draft Law on Personal Income Tax. The study was funded by DANIDA and the Tax Policy Department (TPD) was the coordinator as well as the beneficiary. The model uses two main sources of survey data: the VHLSS-2006 and the 2005 General Department of Taxation survey 118 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S The main advantage of microsimulation modeling for personal income tax lies in its capacity to estimate the distributional effect of a given policy proposal on particular sectors of the population. Since these models are capable of dealing with distributional impacts by identifying potential winners and losers in the society from a given policy proposal, they have proved to be powerful tools in the hands of analysts (Gupta and Harding 2007). Data Requirements The following data are required for microsimulation modeling for personal income tax. • Individual or family-based annual income tax returns; • Household surveys, particularly the Household Income-Expenditure Survey (VHLSS) to cover nonfilers; • Personal income tax regime; and • Basic social and economic data: consumer price index, growth rates of population, GDP, and investment. The process of collecting data is often complex, since the models require large representative samples from the total population. Data is sometimes collected through surveys for specific purposes. Analysts tend to use mostly administrative tax data or the censuses carried out by government agencies. Methodology The following steps are used when constructing a micro simulation model for personal income tax (Harding 1996). (a) Database Construction In practice we generally rely on a relatively small sample size. Typically, the sample size ranges from 0.5 to 5 percent of the total filing population, depending upon the size of the total potential taxpayers and the available resources to collect and analyze the data. Different strata are established on the basis of income sources (for example, employees, self-employed, investment, pensioners, etc.), place of residence (for example, urban, rural, provinces, foreign residence, etc.), and income levels (for example, low, medium, high). (b) Data Cleaning and Data Completion Continuous cross-checking against external data sources is carried out at different stages of the model development process. This includes filling in missing data on nonfilers on the basis of household surveys. of income and spending by high income individuals. While the model is elegant and well elaborated, it does not escape from some typical self-selection and respondent biases inherent in survey data. In this study, we follow the micro-simulation approach, applied by revenue administrations in developed countries, that relies mostly on tax administration data (that is, the individual or household income tax returns). For a discussion of the practicality and objectives of the microsimulation model, see Gupta and Vishnu 2000. In addition, in the case of Vietnam, the microsimulation model approach would enable the GDT to pinpoint the level of compliance by different groups of taxpayers and thereby develop appropriate compliance for specific segments of taxpayers. 119 TA X R E F O R M I N V I E T N A M (c) Sample Weight The sample weight indicates the number of taxpayers, or potential taxpayers, in the population with similar characteristics represented by the sample. After the sample is drawn, or surveyed, the distribution of the sample with certain characteristics must be compared against the whole population. If the sample distribution is skewed (too many or too little sample with certain characteristics), the sample weight must be corrected, or some samples must be dropped. For instance, if we have a sample size of 100 tax returns and 12 of those are in the highest tax bracket, the sample weight for that bracket in the sample is 12. (d) Data aging, Updating and Validation This is done by applying forecasted growth factors such as the growth rates of population, income (or GDP), and investment, as well as the consumer price index. (e) Construction of Typical Taxpayer Tax Calculator Model A typical taxpayer model calculates the tax liability of a typical individual, or family. The model also helps simulate the impact of proposed policy changes on the tax liability of the individual taxpayer. Typically, the structure of a tax calculator model consists of three main components, namely, (1) personal income tax parameters (tax code regulations); (2) taxpayer personal information (historical data); and (3) tax calculator module. (f) Construction of Aggregate Tax Calculator Model and Impact Distribution Analysis The aggregate model is based on the whole database of the personal income tax filers. Similar to the typical taxpayer tax calculator model, the aggregate tax calculator model uses personal income tax parameters and taxpayer personal information as its input. Under the aggregate model, each record from the personal information database will be automatically fed into the typical taxpayer tax calculator model; and the output will be generated based on the predefined taxpayer categories. The macro module is a computer program which is used in the aggregate model to automate the simulation process. The simulation process for each taxpayer personal record in the database is generally done in three steps: (a) read the personal record from the database; (b) calculate the tax liabilities under the current and proposed tax structure using the tax calculator model; and (c) update the tax liabilities in the personal information database and calculate the weighted impact. An impact table is generated after all records in the database are processed. Using the typical taxpayer model, the personal income tax law currently enacted in Vietnam is applied to individuals. This is then compared with the tax liability of the same individual under a proposed tax policy change. Having developed the typical taxpayer model, it is possible to simulate the impact on the tax liability of the proposed policy changes in the tax system, such as changes in the definition of income, modifications in the size and scope of deductibles, tax structure, etc. The model reads individual sets of data obtained from the microsimulation database and performs individual tax calculations based on the laws and regulations. As the database is 120 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S formed by a stratified sample, each observation carries a weight corresponding to the number of tax filers represented by that observation. Calculations of the model have to be adjusted by this weight in order to obtain aggregate revenue estimates. Figures 4.1a and 4.1b provide a conceptual framework for the construction of a typical taxpayer and aggregate model structures as applicable in the case of Vietnam. Note in the figures, the Current Regime refers to the provisions stipulated in the personal income ordinance of 2004 while the Proposed Changes refer to the provisions in the personal income tax law enacted in 2009. Figure 4.1a: Typical Taxpayer Model Structure Typical Taxpayer Model Current Tax Proposed Model Typical Taxpayer Regime Changes Current Tax Proposed Regime Changes Aggregate Model MacroAggregate Model Module Macro Module Personal Information Personal Database Information Database Current Tax Proposed Current Tax Proposed RegimeRegime Changes Changes Impact Distribution Impact Distribution Analysis Analysis Figure 4.1b: Aggregate Model Structure Personal Income Tax Personal Income Tax Parameters Parameters Current Tax Proposed Regime Proposed Current Tax Changes Regime Changes Taxpayer Tax Calculator Taxpayer Personal PersonalInformation Tax Calculator Information Personal Income Tax Liability Personal Income TaxProposed Current Tax Liability Regime Changes Current Tax Proposed Regime Changes 121 TA X R E F O R M I N V I E T N A M Data Available in Vietnam A data set of 3,000 individual annual income tax returns from FY 2004 was available to construct the model, representing approximately 1 percent of the total taxpayer population in Vietnam at that time. There is no evidence of the criteria followed in the sampling exercise, however. The limitation of the tax return data set in terms of size and representation may seriously reduce the scope of the microsimulation analysis. Model Results The model comprises eight tables. Table 4.1 is the Typical Taxpayer Model. Table 4.2 is the Parameters Table containing a set of parameters related to personal income tax law such as income from different sources, deductions, credits and tax brackets provided in the current tax code (the base) and in the tested tax system (the option). Table 4.3 gives the Impact for Typical Taxpayer with Different Income Level, which helps to test the tax system with related policy issues. Essentially, this is a sensitivity table. Table 4.4 is the Sample (Raw) Data coming out of the data sampling process, providing information for all representative taxpayers and indicating the weight assigned to each sampled observation. Table 4.5 is the Growth Factors Table used to age sample data for further use in construction of the model. Table 4.6 is the Aged (Forecasted) Data; while table 4.7 is the Forecasted and Weighted Data. Finally, table 4.8 contains the Tax Calculations and Simulations. This table consists of two parts. The first part represents the calculation of individual tax liabilities of each representative taxpayer while the second part provides the estimates of the aggregate income tax liability of all taxpayers for base and option scenarios and the difference between the two scenarios (impact analysis). The total tax liability for the whole population is the sum of all weighted tax liabilities. Under the old 2004 Ordinance, the tax rate applicable for income below 5 million dong per month (60 million per year) was zero percent, but there was no family allowance. The present PIT Law as approved by the National Assembly in November, 2007 allows for family unit taxation in such a way that taxpayers will receive an allowance of 4 million for self and 1.6 for the dependent per month. In order to assess the progressivity of the PIT law 2009 and to estimate the impact of the new marginal tax rate schedule, combined with family allowances, on a wage earner with one dependent, an estimation of average rates was conducted. For this exercise, a tax calculator was used to estimate the tax liability of a typical taxpayer under both PIT systems (that is, Ordinance 2004 and the PIT law 2009), for different annual income scenarios (30 to 3,000 million dong). The typical taxpayer calculator is constructed for a representative household with one dependent. Under the approved law 2009, the taxable income was estimated by deducting a yearly amount of 67.2 million dong (5.6 million dong per/month) to reach taxable income. Once statutory marginal tax rates were applied under both PIT laws, total impact and average tax rates were estimated. The results are presented in table 4.3. 122 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S Table 4.3: Average Tax Rates for Different Levels of Income Tax Ordinance 2004 and Current Law Enacted in 2009 (Million Dong per year) Tax Payable Tax Payable Average Tax Rate Average Tax Rate Total (Ordinance (PIT Law Impact (Ordinance 2004) (PIT Law 2009) Income 2004) 2009) (%) (%) 30 0.0 0.0 0.0 0 0 90 3.0 1.1 -1.9 3 1 180 12.0 8.3 -3.7 7 5 330 45.0 31.6 -13.4 14 10 540 114.0 76.2 -37.8 21 14 810 222.0 147.8 -74.2 27 18 1,200 378.0 273.5 -104.5 32 23 1,500 498.0 378.5 -119.5 33 25 2,000 698.0 553.5 -144.5 35 28 3,000 1,098.0 903.5 -194.5 37 30 Source: Own calculations, using Vietnam PIT Schedules. Thus the PIT law of 2009 appears to be less progressive than the previous law of 2004 if only wage incomes are taken into the tax base as was done in the law of 2004. Prima facie, it also yields less revenue since the revenue collections for FY 2009 are estimated to be reduced by 21.7 percent in nominal terms. Figure 4.2: PIT Progressivity (wage income) 40 35 37 33 35 30 32 30 Average Tax Rates 25 27 28 21 23 25 20 18 15 14 14 10 10 5 7 5 3 1 0 0 30 90 180 330 540 810 1,200 1,500 2,000 3,000 Annual Income (Million VND) Source: Own calculations using Vietnam PIT Schedules. The progressivity and revenue potential of the PIT law 2009 need to be examined more closely, however, because the tax base in this law is expected to expand by including other forms of income. Progressivity and Revenue Potential of the PIT law 2009 Different from the revenue forecasting exercise done above, now we also include other types of incomes in the tax base as envisaged by the PIT law 2007. This is done by considering other nonwage incomes in order to estimate the full impact of the new tax rate schedule. 123 TA X R E F O R M I N V I E T N A M According to the VHLSS-2004, the average composition of income, the structure of income per capita by source of income, is as follows: Salary or wage (32.7 percent), Agriculture (22.6 percent), Forestry (1 percent), Fishery (3.6 percent), Industry (5.4 percent), Construction (0.4 percent), Trade (9.9 percent), Services (6.8 percent), and others (17.7 percent). The PIT law 2009 exempts income from primary activities while other income sources are taxed. Therefore, besides wages and salaries, which were practically the only source of income taxed by the old PIT law, two new categories of income were introduced into the analysis: (i) business income, which includes industry, construction, trade and services income sources, and (ii) other incomes such as capital investment, royalties, capital gains, immovable property, donations, inheritance, and capital transfer. By using this new income source categorization, a new tax calculator was used to assess the tax liability of a typical taxpayer under both PIT systems, for different annual income scenarios (30 to 3,000 million dong). Considering an average of two dependents per household instead of one, taxable income was estimated by deducting a yearly amount of 86.4 million dong (67.2 million dong in the previous exercise) to reach taxable income. Similar to the previous progressivity exercise, statutory marginal tax rates were applied under both PIT laws. The only difference is that other income was taxed at a 13.33 percent flat rate, which is the average tax rate applicable under the PIT law 2007 to sources of income such as investment, royalties and alienation of immovable property. Applying the PIT Micro simulation model to the data set of 3,000 individual tax returns available for Vietnam, under the income tax law of 2007, the tax revenue collection is estimated to increase by approximately 69 percent in nominal terms, to VND 1,853 billion from VND 1,098.7 billion. Measured by average tax rates, the PIT law 2009 turns out to be more progressive at different levels of income compared to the old system, when different sources of income are included into the tax base. This is the outcome even when the taxpayers are given an allowance for two dependents under the new law. The results are presented in table 4.4 and figure 4.4. Table 4.4: Assessing Progressivity of Old (2004 and New (2009) PIT System (VND million) Total Wages Business Other Taxable income Tax Payable Average Tax Income Income Income Rates Old New Old New Old New System System System System System System 30 14 9 7 14 30 0.0 1.0 0% 3% 90 41 27 22 41 90 0.0 2.9 0% 3% 180 81 55 44 81 180 2.1 8.3 1% 5% 330 149 100 81 149 330 8.9 26.2 3% 8% 540 244 164 132 244 540 24.8 60.9 5% 11% 810 366 246 198 366 810 55.8 115.8 7% 14% 1,200 542 365 293 542 1,200 114.8 210.2 10% 18% 1,500 677 456 367 677 1,500 169.0 292.3 11% 19% 2,000 903 608 489 903 2,000 259.3 440.8 13% 22% 3,000 1,355 912 733 1,355 3,000 440.0 737.9 15% 25% 124 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S Figure 4.3: PIT Progressivity (all sources of income) 30 Average Tax Rates 25 25 20 22 19 15 18 15 14 10 11 11 13 8 10 5 5 7 3 3 3 5 0 0 0 1 30 90 180 330 540 810 1,200 1,500 2,000 3,000 (5) Annual Income (Million VND) Source: Own calculations using Vietnam PIT Schedules. Corporate Income Tax A microsimulation model is applied to project CIT revenues and analyze the revenue impact of policy proposals at an individual company level. Data Requirements (a) Tax returns filed by companies in the last two to three years: Stratified samples should be selected from these to represent different categories of industries and amounts of assets/ investments. If only samples are available instead of the entire dataset, the selected samples should represent different categories of industries and the amount of assets. Ideally, all the top 5 percent companies by assets should be included in the samples. To maintain confidentiality, the company name and identification should be excluded from the file. (b) Basic Data pertaining to the business sector: Such data related to operations, revenues and expenditures of the business sector such as corporate investment (by sector and type of asset), corporate income, profits, tax liability, dividend payments, and tax deductible expenses such as interest payments, costs of goods sold, fiscal depreciation. In addition, if the tax agency publishes any aggregate or disaggregate data on the corporate taxes that would be useful. The sources for these business sector related data could be enterprise censuses or surveys. Methodology Conceptually, a microsimulation for CIT is similar to the one applicable to PIT. As part of the model, a CIT calculator is to be constructed in such a manner that it is capable of computing business income tax liability in Vietnam for a typical business. The existing tax law is applied to a representative business to be compared with the tax liability of the same business under a proposed tax policy change. Having developed the typical taxpayer model, it is possible to simulate the impact on the tax liability of the proposed policy changes in the tax system, such as the change on the CIT tax rate and the rationalization of incentives (Citro and Hanushek 1991). The model reads individual sets of data from the microsimulation database and performs individual tax calculations based on laws and regulations. As the database is formed by a stratified sample, each observation carries a weight corresponding to the number of tax 125 TA X R E F O R M I N V I E T N A M filers represented by that observation. Calculations of the model have to be adjusted by this weight in order to obtain aggregate revenue estimates. The proposed CIT Model for Vietnam developed in Microsoft Excel has four major components: 1) Table of Parameters; 2) Corporate Tax Calculator Model; 3) Tax Returns Database; and 4) Macro Modules. The Table of Parameters contains information about the existing statutory corporate tax rates and the proposed changes to those rates. The Corporate Tax Calculator Model uses the data (that is, tax rates) stored in the Table of Parameters and calculates the tax liability of an individual corporation. Tax Returns Database stores the tax returns data from national samples of selected corporations. Finally, the Macro Module automates the process of simulating the calculation of tax liability for the corporations in the database. The following steps are involved in forecasting corporate income tax revenues: 1. Sampling of tax returns. Stratified sampling is suggested for corporate income tax forecasting. First, different strata of companies are established on the basis of some critical categories such as size of assets, income, industry, or region. Second, samples of firms will be selected from these strata. Different weights are assigned to each stratum, and will be used to determine the proportion of firms in each stratum to be randomly drawn for the ultimate sample. Also, note that samples from different filing periods may also be selected to account for noncalendar filing and filing extensions; 2. Data cleaning. This is to ensure consistency and reliability of data selected for the simulations. 3. Data completion process. Filling in missing data relies on either data from corporate financial statements, and/or imputation of data; 4. Construct a tax calculator to simulate corporate income tax. The corporate income tax base is derived from gross income from different sources minus itemized deductions including costs of goods sold, depreciation allowance, interest payments, overhead expenses, and any net operating loss from prior years. Tax payable is estimated as the product of the tax base and corporate income tax rate. Tax liability should also be adjusted for any tax credits that are allowed. From this base tax calculator, the impact of any proposed changes in the corporate income tax code on a representative firm and on government tax revenue can then be simulated; Data Available for Vietnam The GDT provided a data set for microsimulation modeling purposes, including a sample of 91 enterprise income tax returns filed for the 2005 period, which represents approximately 0.15 percent of the total CIT taxpayer population in Vietnam. There is no indication of the criteria followed in the sampling exercise. The limitation of the tax return data set in terms of size and representation may seriously reduce the scope of the microsimulation analysis. Nevertheless, a revenue forecasting model was prepared using the data available to assess the impact of the reforms proposed for the CIT structure in Vietnam. The model is designed in such a way that it could readily produce the estimations to fully assess the revenue impact of the reform proposed above, once additional data is considered. 126 F O R E C A S T I N G TA X R E V E N U E : M E T H O D S A N D T O O L S Model Results According to the database available, three types of corporations are considered: state owned, foreign invested, and nonstate owned. Also, the different CIT tax rates shown in the data sample are considered. For the sake of simulation, a flat rate of 28 percent across corporations was considered instead of the structure with differentiated rates. In any event, this table provides flexibility as changes to the current and proposed tax rates can be easily done without modifying the macro codes. The formulae in the Tax Calculator Model simulate the tax calculation procedures as presented in the 91 tax return samples. The Calculator Model presents the calculation of individual corporation tax liability under the existing law and the proposed changes in the law, as well as the changes (impacts) on the corporations’ tax liability due to the discretionary changes. The Tax Return Database stores the tax returns data from national samples of selected corporations. The Macro Module automates the process of simulating the calculation of tax liability for the corporations in the database. The Module reads individual corporation records in the database, stores them in the Tax Calculator Model and saves the calculated tax liability in the designated location in the Tax Returns Database. This process continues until the last record in the database is covered. The microsimulation performed for the 91 corporations tax returns data shows the revenue impact from the 28 percent flat rate proposed. Using the 91 CIT individual tax return database available, a policy analysis scenario was constructed using a 28 percent CIT flat rate across the different corporation types. After performing the tax calculator model, it is expected that the government would collect VND 105.8 billion, instead of the VND 71.2 billion estimated, which represents an additional CIT collection of VND 34.6 billion, or an increase of 48.6 percent in nominal terms. 3.5. Summary of Model Results The summary of results of impact analysis and revenue forecasting for VAT, personal income tax and corporate income tax are presented in table 4.5. Even with a considerable data constraint, it is clear that the models may be applied by the GDT to analyze the implications of tax policy changes and forecast revenues. Similar analysis could not be done for excises because of the unavailability of data. However, as the GDT strengthens its database and builds the forecasting models, it may have credible results within 18 to 24 months and the results could be utilized for the budget making exercise. Table 4.5: 2007 Revenue Forecasting Summary Base Case Scenario Policy Scenariosa Nominal Increase over 2006 Nominal Increase VND Billion revenue (Percentage) VND Billion (Percentage) VAT 42,577 16.7 50,359 38.1 CITb 71.2 15.8 105.8 48.6 PIT 1,098.7 21.3 1,853 69.0 Note: a. Policy scenarios constructed under the following assumptions: VAT: 10 percent single rate and exemption rationalization; CIT: Microsimulation using sample of 91 returns with a 28 percent flat rate across all corporations; and PIT: Microsimulation using 3,000 returns and applying the PIT Law 2007. b. The revenue figures are for the sample size of 91 corporations or 0.15 percent of the total companies in Vietnam. If prorated, base case total revenue would be VND 46,466 and 70,000 billion dong. 127 TA X R E F O R M I N V I E T N A M 3.6. Suggestions for Improving Revenue Forecasting a. The present status of revenue forecasting in Vietnam is rudimentary in nature. This, however, is not peculiar to Vietnam as a majority of developing countries and transition economies are in a similar situation. The GDT has already been taking steps to augment the database, computerize the department and build necessary models. What is needed is a concerted effort to complete the process. b. The starting point would be the enhancing of the database. This should be done after fully understanding the data requirements for various kinds of econometric and microsimulation models. A team of officers could be assigned exclusively this task. c. While the task of strengthening the database is underway, the personnel in the revenue projection department (RPD) of GDT should undergo necessary training in economic and econometric tools, tax policy analysis and revenue forecasting models. d. At the end of these two exercises, which should take about 9 to 12 months, GDT may begin to build specific macro and microsimulation models and calibrate them. The first round models are seldom precise and it takes time and experience to fine tune them. Later on, some models may be constructed even for nontax revenues like user charges. e. Currently there is no effort to specialize within the RPD and any official can be asked to work on a specific issue. However, in the interest of developing expertise over the long run, sections within the RPD may be assigned specific taxes to work on, for instance, direct versus indirect taxes, natural resource taxation. f. Data from other sources, for instance, GDP growth rate from macroeconomic models, Input-Output tables from the Ministry of Planning, income-expenditure household surveys, and business surveys are complementary to forecasting of revenues. The Ministry of Finance may like to coordinate the activities of these other agencies so that the revenue forecasting process is smooth and the results are reliable. g. Parallel to building technical capacity of staff working on revenue forecasting, it is essential that institutional reforms are in place to create a whole new set of the right incentives. The sole revenue-based criteria in performance evaluation of the GDT must be replaced with multiple indicators measuring both efficiency and effectiveness of its functions. In particular, the indicators that measure quality of key tax processes (registration, accounting, auditing, debt management and taxpayer service) should be established and valued to be as important as the collection indicator. A balanced mix of performance criteria would help mitigate a potential detrimental risk that revenue agencies—and their operational units—have incentive to lower their revenue forecasts so as to lower their assigned collection targets. The ongoing Tax Administration Modernization Project follows this strategic direction: the institutional development component is designed to enhance the HRM policies with a new comprehensive set of measurable performance indicators. 128 Tax Reform and Revenue Assignments in Vietnam V By Jorge Martinez-Vazquez and Minh Van Nguyen 1. Introduction Tax reform and decentralization reform have some similar objectives but also their own separately independent ones, and unless both policy processes are harmonized, decentralization and tax policy reforms can work at cross purposes. The most important objectives pursued by tax reform include those of simplicity, low economic distortions or efficiency, and equity or fairness in the distribution of tax burdens. In the case of decentralization reform, those objectives are also important but they necessarily take a second seat to the objectives of revenue adequacy for subnational budgets and accountability of local government officials. The attainment of these decentralization objectives requires of significant own taxes and fees assigned to subnational governments. However, pursuing a decentralized system with more tax sources assigned to subnational governments will likely mean that the entire tax system for the country will be less simple, and a bit more distortionary, and even perhaps less fair. But, it would be wrong to reach the conclusion that we should not have fiscal decentralization on the revenue side of subnational budgets. Instead, what is needed is the coordination of the tax reform effort with the desired level of revenue decentralization for the country. This coordination needs to reach a compromise between the decentralization objectives of accountability and revenue autonomy for subnational governments and those of simplicity and efficiency for the entire tax system at the national level. Even though subnational taxes may be less progressive than desired, it needs to be kept in mind that what matters from the perspective of distributional equity is the overall level of progressivity for national and subnational taxes combined. Subnational taxes are better guided by the application of the benefit principle, that is, a correspondence between the taxes paid and the value of the services received. The application of the ability to pay principle and the objective of income redistribution are much more effectively pursued at the national level. After all, potentially less progressive taxes at the subnational level can be compensated with more progressive taxes at the national level. In sum, it is important to revisit the issue of revenue assignment in decentralization design at the time of major tax reform. Because Vietnam is involved in a major tax reform effort, it is justified to take an in-depth view of its revenue assignment system. This is the major motivation of this paper. As part of the drive toward decentralization in Vietnam, budget allocation norms have been established for recurrent and capital expenditures. Rules for tax revenue assignment are the natural complement to those norms. However, the current criteria for revenue assignments, in particular which taxes are retained and the sharing arrangement for each province, fundamentally are still ad hoc. The time of fundamental tax reform is a good opportunity to put the question of revenue assignments on a more sound footing in Vietnam. In this paper 129 TA X R E F O R M I N V I E T N A M we review the current system and analyze these assignments against public finance principles and best international practices. The paper also suggests several avenues for reform.25 2. Review of the Existing Revenue Assignments According to the 2002 State Budget Law there are three types of taxes in Vietnam. First, taxes assigned 100 percent at the central level; second, taxes assigned 100 percent at the provincial level; and third, taxes shared between the central and provincial governments.26 The tax revenues assigned 100 percent to the central government include export and import taxes, VAT and excises on imports; taxes and other revenues from the petroleum industry; and corporate income tax on enterprises with uniform accounting. The tax revenues assigned 100 percent to the subnational level include land and housing taxes, natural resource taxes excluding those on petroleum activities, license tax, tax on transfer of land use rights, fees on land use, land rent, revenues from the leasing and sale of dwellings publicly owned, registration fees and most other fees and charges. The taxes shared between the central and provincial governments include all VAT receipts with the exception of VAT on import goods; corporate income tax with the exception of receipts from enterprises under the whole-unit accounting system; personal income tax; special consumption tax on domestic goods and services; and gasoline and oil fees. “Sharing ratesâ€? are determined which establish the proportion of total revenues from shared taxes which a province may retain. The “sharing ratesâ€? between the center and the provinces are not stated in the Budget Law. It is left for the Standing Committee of the National Assembly to fix the sharing rates for each province for a period of three to five years. A peculiarity of tax sharing in Vietnam vis-à-vis most other countries is that tax sharing rates are uniform for all shared taxes for each province but they differ by province.27 The tax sharing rates are calculated as part of the budget process at the start of each stability period of at least three years by using a formula. The formula estimates the gap between expenditure needs (estimated on the basis of norms) and revenue capacity.28 For the larger group of relatively poor provinces the tax sharing rate is 100 percent. 25 Sections of this report build on chapter 6 of the World Bank Vietnam 2004 PER-IFA, and on Martinez-Vazquez (forthcoming). 26 The Budget Law leaves provincial governments with considerable discretion in the design of revenue assignments to districts and communes within their borders. However, article 34 of the 2002 Law provides some general principles and minimum standards the provinces need to follow in designing the revenue assignments for their local governments. 27 By far the most common approach internationally calls for sharing rates that differ by tax but that are the same for all subnational governments. 28 Expected revenues for the provincial government are determined by the tax administration on the bases of the actual revenue collections of the previous years, taking into account any tax policy changes applicable in that year and the expected economic growth during the year. The minimum expenditure needs of the provincial government are derived on the basis of the prevailing system of expenditure norms, and cover all current expenditures (salaries, operation and maintenance, and so on) and (some minimum amount of) capital expenditures. The norms are adjusted for different regions depending on geography and remoteness. 130 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M In addition to the shared revenues, as discussed below, the poorest provinces get an additional equalization transfer. Relatively wealthier provinces get sharing rates between 100 and zero percent. Table 5.1 shows the tax sharing rates for the provinces for the 2004–06, 2007–10 and 2011–15 “stability periods.â€? For the 2007–10 stability period, out of the 64 provinces, there are 11 which are considered “surplusâ€? provinces with sharing rates below 100 percent. This represents a reduction by 4 provinces in the number considered as surplus units by comparison to the 2004–06 stability period. Otherwise the changes in sharing rates between those two stability periods were relatively small and generally downward. More recently, the MOF and the provincial governments made a compromise to extend the stability period from three to five years. The quid pro quo of the agreement was more stability for the provinces and the ability of the center to get higher shares from the surplus provinces. The experience has been that are the richer, more economically dynamic provinces that tend to benefit from the length of the stability period. From the last column in table 5.1, it is apparent that most of the share rates have been reduced. An exception is Hanoi, which because it has been merged with Hatay, which was a poorer jurisdiction, it has been allowed to retain a higher revenue share. Table 5.1: Tax Sharing Rates for the Provinces 2004–06, 2007–10, and 2011–15 Stability Periods Sharing Rate Sharing Rate Sharing Rate Province   2004–06 2007–10 2011–15 1 Ha Giang 100 100 100 2 Tuyen Quang 100 100 100 3 Cao Bang 100 100 100 4 Lang Son 100 100 100 5 Lao Cai 100 100 100 6 Yen Bai 100 100 100 7 Thai Nguyen 100 100 100 8 Bac Kan 100 100 100 9 Phu Tho 100 100 100 10 Bac Giang 100 100 100 11 Hoa Binh 100 100 100 12 Son La 100 100 100 13 Lai Chau 100 100 100 14 Dien Bien*   100 100 15 Ha Noi 32 31 42 16 Quang Ninh 98 76 70 17 Hai Phong 95 90 88 18 Vinh Phuc 86 67 60 19 Hai Duong 100 100 100 20 Hung Yen 100 100 100 21 Bac Ninh 100 100 93 22 Ha Tay** 100 100 23 Ha Nam 100 100 100 24 Nam Dinh 100 100 100 131 TA X R E F O R M I N V I E T N A M 25 Ninh Binh 100 100 100 26 Thai Binh 100 100 100 27 Thanh Hoa 100 100 100 28 Nghe An 100 100 100 29 Ha Tinh 100 100 100 30 Quang Binh 100 100 100 31 Quang Tri 100 100 100 32 Thua Thien Hue 100 100 100 33 Da Nang 95 90 85 34 Khanh Hoa 52 53 77 35 Quang Nam 100 100 100 36 Quang Ngai 100 100 61 37 Binh Dinh 100 100 100 38 Phu Yen 100 100 100 39 Ninh Thuan 100 100 100 40 Binh Thuan 100 100 100 41 Dak Lak 100 100 100 42 Dac Nong *   100 100 43 Gia Lai 100 100 100 44 Kon Tum 100 100 100 45 Lam Dong 100 100 100 46 Ho Chi Minh City 29 26 23 47 Dong Nai 49 45 51 48 Binh Duong 44 40 40 49 Binh Phuoc*   100 100 50 Tay Ninh 99 100 100 51 Ba Ria – Vung Tau 42 46 44 52 Binh Phuoc 100 100 53 Long An 99 100 54 Tien Giang 99 100 55 Vinh Long 99 100 56 Can Tho 95 96 91 57 Ben Tre 100 100 58 Tra Vinh 100 100 59 Soc Trang 100 100 60 An Giang 100 100 61 Dong Thap 100 100 62 Kien Giang 100 100 63 Bac Lieu 100 100 64 Ca Mau 100 100 *Denotes newly created provinces. **Merged with Hanoi. Source: General Department of Statistics (GSO) 132 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M Currently, Vietnam complements revenue assignments to the provinces with two types of transfers: the “balancing transferâ€? and specific grants. The balancing transfer is an equalization grant designed to increase the financing viability of poor provinces. These grants are unconditional and calculated using a formula. In addition, these grants are fixed in nominal terms during each stability period of three-to-five years.29 For specific conditional grants, there are several types. An important type is those transfers for the implementation of “National Target Programs,â€? such as those for various national health programs. A second type is conditional transfers that are specifically designed for a particular province. The third type is “matchingâ€? grants designed to implement important projects for socioeconomic development. There are also emergency grants for help with natural disasters and so on.30 In Vietnam, all tax collections are centralized. The General Taxation Department collects all domestic taxes with offices that extend through the provinces and the districts, and the Customs Department collects all taxes falling on imports and export tax. Only minor fees and charges are collected by financial agencies and service providers. 29 The formula used for the determination of equalization transfers to the poor provinces is also the same formula used to derive the customized tax sharing rate for the rich provinces discussed above. For the most part, provinces implement with their lower-tier governments a system of balancing or equalization transfers that is similar to that of the central government with the provinces. However, between the provincial and lower levels, the provinces set their own norms and so effectively set their own transfers. 30 Capital transfers in Vietnam are part of the equalization transfer and treated as a general transfer. The capita transfer is now being determined by a formula that takes into account poverty, economic development level, and other factors for each province. In the recent past there was no formula for determining the amount of capital transfer; instead, the capital transfers were calculated by using a grossing up factor for the expenditure needs determined for the calculation of the equalization grant. 133 134 Table 5.2: Provincial Revenues by Revenue Sources and Summary Statistics, 1997–2010 (in Million VND)   1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010e TOTAL REVENUE                   Total (Million VND) 30,360,918 31,875,486 40,122,103 47,508,691 58,842,075 65,822,571 85,850,251 105,039,065 124,193,506 146296,591 174810,837 235073,044 218711,704 273689,143 TA X R E F O R M I N V I E T N A M Maximum (Million VND) 3,795,677 3,652,867 3,876,337 4,482,944 6,173,867 7,593,201 10,163,321 12,346,213 13,024,259 14439,759 15065,902 26684,144 23474,893 32830,801 Minimum (Million VND) 190,768 218,584 258,408 317,738 393,632 422,105 515,109 628,279 823,278 991,184 1076,351 1154,006 1411,484 1641,467 Average (Million VND) 497,720 522,549 657,739 778,831 964,624 1,079,059 1,407,381 1,641,235 1,940,524 2285,884 2731,419 3673,016 3471,614 4344,272 Standard Deviation 511,995 483,147 530,261 601,639 807,217 1,014,225 1,350,802 1,651,534 1,793,452 2086,635 2205,545 3693,735 3622,196 5065,666 Coefficient of Variation 1.03 0.92 0.81 0.77 0.84 0.94 0.96 1.01 0.92 0,91 0,81 1,01 1,04 1,17 TAXES 100% ASSIGNED TO SNG Total (Million VND) 10,663,832 10,188,870 10,897,540 11,438,666 12,855,610 15,809,860 22,817,136 33,720,219 35,757,399 41827,070 43819,223 60477,542 52029,040 64455,744 Maximum (Million VND) 2,292,169 2,351,102 2,310,872 2,407,533 3,380,660 4,638,815 5,216,889 6,172,473 5,883,673 6389,724 4778,388 12033,857 9965,650 14031,610 Minimum (Million VND) 11,167 10,967 13,806 17,020 18,967 25,365 30,628 16,579 38,470 37,672 53,960 39,583 55,413 64,183 Average (Million VND) 174,817 167,031 178,648 187,519 210,748 259,178 374,051 526,878 558,709 653,548 684,675 959,961 825,858 1023,107 Standard Deviation 303,698 304,245 299,857 321,548 433,124 597,627 718,074 877,410 907,762 1008,224 937,854 1722,516 1454,326 2050,044 Coefficient of Variation 1.74 1.82 1.68 1.71 2.06 2.31 1.92 1.67 1.62 1,54 1,37 1,79 1,76 2,00 SHARED TAXES Total (Million VND) 9,739,476 9,397,695 8,713,205 11,147,368 13,000,372 14,735,167 19,927,159 31,770,783 39,446,837 45463,862 55429,802 73488,958 82627,357 104928,015 Maximum (Million VND) 1,310,961 1,159,989 1,144,284 1,729,470 2,012,719 2,197,020 3,906,979 5,946,359 6,886,196 7778,749 10369,631 12752,406 14377,220 19542,484 Minimum (Million VND) 8,219 8,094 9,677 8,826 10,631 11,941 13,646 16,721 29,275 3,265 5,710 9,349 8,150 8,955 Average (Million VND) 159,664 154,061 142,839 182,744 213,121 241,560 326,675 496,418 616,357 710,373 866,091 1166,491 1311,545 1665,524 Standard Deviation 224,206 202,066 198,463 289,192 331,695 369,871 572,976 847,728 992,612 1148,415 1460,475 1901,030 2444,747 3343,462 Coefficient of Variation 1.40 1.31 1.39 1.58 1.56 1.53 1.75 1.71 1.61 1,62 1,69 1,63 1,86 2,01 TRANSFER Total (Million VND) 9,957,610 12,288,921 20,511,358 24,922,657 32,986,093 35,277,544 43,105,956 39,548,063 48,989,270 59394,292 75645,100 95152,440 84055,307 104305,384 Maximum (Million VND) 461,193 567,523 863,980 1,060,089 1,362,348 1,798,383 1,913,027 1,971,745 2,625,883 4621,036 4074,251 10307,310 4815,793 6163,917 Minimum (Million VND) 8,796 29,626 153,418 61,008 242,473 215,737 264,586 61,724 44,342 141,446 58,131 87,863 56,133 56,000 Average (Million VND) 163,240 201,458 336,252 408,568 540,756 578,320 706,655 617,938 765,457 928,036 1181,955 1510,356 1334,211 1655,641 Standard Deviation 98,271 111,644 163,460 174,002 215,481 274,265 318,065 362,248 489,151 681,430 724,636 1406,175 871,686 1082,136 Coefficient of Variation 0.60 0.55 0.49 0.43 0.40 0.47 0.45 0.59 0.64 0,73 0,61 0,93 0,65 0,65 Source: Ministry of Finance and the author TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M 135 TA X R E F O R M I N V I E T N A M The evolution of provincial revenues for the period 1997 to 2010 is shown in table 5.2 in millions of VND. This table also shows the evolution of the three main components of: “own revenuesâ€? or taxes assigned 100 percent to subnational governments, shared taxes, and transfers. Total provincial revenues have increased steadily over the period quadrupling in size. As illustrated in figure 5.1, a significant part of the growth in total revenues is explained by the faster growth in transfers. However, with the beginning of the stability period in 2004, revenues from shared taxes and, to a lesser extent, own taxes jumped quite significantly. The jump in revenues from shared taxes was due to a large extent to the two additional taxes (special consumption taxes and gasoline and oil fees) that were added to the list of sharable taxes starting that year. In absolute size of collections, provinces can differ considerably; for example in the distribution of revenues by province the difference between the maximum and the minimum was twentyfold in 2010. Note also that even though this difference and the coefficient variation decreased somewhat over time up to 2006–07, the tendency has been again significantly upwards after that (figure 5.2). Figure 5.1: Total Provincial Revenue, by Revenue Sources (Million VND) 300000000.0 250000000.0 200000000.0 150000000.0 100000000.0 50000000.0 0 1997 19981999 2000 200120022003 2004 2005 2006 20072008 2009 2010 TOTAL REVENUE SNG TAXES SHARE TAXES T NSFER Source: Ministry of Finance and the author Figure 5.2: Coefficient of Variation of Provincial Revenue, by Revenue Sources (Million VND) 2.5 2 1.5 1 0.5 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 TOTAL REVENUE SNG TAXES SHARE TAXES T NSFER Source: Ministry of Finance and the author 136 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M Table 5.3 shows the evolution of revenues (total and of the three main components) in millions of VND per 1000 population, while the trends for the four categories of revenues are illustrated in figure 5.3. Normalized by population, fiscal disparities have increased over time in Vietnam up to recent times. For example, the differences between the richest and the poorest province in revenues per capita jumped form fourfold in 1997 to fivefold in 2005 while the coefficient of variation also increased by 25 percent. However, in more recent years (up to 2010) these trends have stabilized or decreased slightly. As illustrated in figure 5.4, fiscal disparities in revenues per capita are created by the distribution of revenues from shared taxes and also from own taxes. The distribution of overall revenues shows smaller disparities than revenues from shared and own taxes because of the compensating or equalizing effects of transfers. However, considerable differences across provinces in their abilities to raise revenues remain in the system even after transfers. Table 5.3: Provincial Revenue per 1,000 Population and Summary Statistics, by Revenue Sources, 1997–2010 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010e TOTAL REVENUE Maximum 861 924 1,386 1,535 1,962 2,141 3,405 4,509 4,552 4277 7,556 6,932 5,695 7,083 (Million VND) Minimum 228 240 286 329 392 446 576 540 814 760 1,026 1,088 1,394 1,647 (Million VND) Average (Million 408 423 524 612 748 826 1,061 1,284 1,494 1859 2,258 2,866 2,698 3,281 VND) Standard 155 168 220 254 335 398 543 669 720 751 1,036 1,319 1,032 1,311 Deviation Coefficient of 0.38 0.40 0.42 0.41 0.45 0.48 0.51 0.52 0.48 0.40 0.46 0.46 0.38 0.40 Variation TAXES 100% ASSIGNED TO SNG Maximum 472 474 456 461 629 972 2,238 2,981 2,918 2,389 5,638 4,615 2,586 3,112 (Million VND) Minimum 33 31 35 31 36 37 50 52 92 54 73 106 59 74 (Million VND) Average (Million 143 135 142 147 163 198 282 412 430 432 499 585 509 594 VND) Standard 89 83 83 94 110 170 306 406 395 367 724 645 438 549 Deviation Coefficient of 0.62 0.61 0.58 0.64 0.67 0.85 1.08 0.99 0.92 0.85 1.45 1.10 0.86 0.92 Variation SHARED TAXES Maximum 503 410 466 676 972 1,408 1,795 2,086 2,246 2,456 2,759 5,770 5,227 5,661 (Million VND) Minimum 23 20 22 25 27 29 37 52 95 11 16 25 22 24 (Million VND) Average (Million 131 125 114 144 165 185 246 388 475 485 567 774 804 987 VND) Standard 91 82 83 127 155 204 263 351 416 492 524 902 828 1,001 Deviation Coefficient of 0.69 0.66 0.73 0.88 0.94 1.10 1.07 0.90 0.88 1.02 0.92 1.16 1.03 1.01 Variation 137 TA X R E F O R M I N V I E T N A M TRANSFER Maximum 646 736 1,050 1,307 1,617 1,793 2,007 2,136 3,083 3,197 3,862 5,410 5,256 6,605 (Million VND) Minimum 5 16 86 52 145 138 187 39 37 42 42 56 36 37 (Million VND) Average (Million 134 163 268 321 419 442 533 483 589 944 1,192 1,507 1,385 1,700 VND) Standard 148 165 210 252 307 337 404 504 693 755 898 1,281 1,137 1,393 Deviation Coefficient of 1.10 1.01 0.79 0.78 0.73 0.76 0.76 1.04 1.18 0.80 0.75 0.85 0.82 0.82 Variation Source: Ministry of Finance and the author Figure 5.3: Average of Provincial Revenue per 1,000 Population, by Revenue Sources (Million VND) 3500 3000 2500 2000 1500 1000 500 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 TOTAL REVENUE SNG TAXES SHARE TAXES T NSFER Source: Ministry of Finance and the author Figure 5.4: Coefficient of Variation of Provincial Revenue per 1,000 Population, by Revenue Sources (Million VND) 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 TOTAL REVENUE SNG TAXES SHARE TAXES T NSFER Source: Ministry of Finance and the author 138 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M In terms of consolidated State Budget revenues (central and subnational governments combined), the subnational governments’ share for 2002 was 26 percent in 2003 to 47 percent in 2005.31 While the subnational share oscillated mildly up and down 25 percent during 1997–2003 (table 5.4), it jumped considerably in 2004 and 2005. As is illustrated in figure 5.5, the introduction of the new stability period in 2004, when, as we mentioned above, the two additional taxes of special consumption taxes and gasoline and oil fees were added to the list of sharable taxes starting in 2004, ushered in a heightened participation of subnational governments in total fiscal resources. However, in the following years, the subnational governments’ share in consolidated State Budget revenues fell rapidly to stabilize around 30 percent during 2008–10. Table 5.4: Outcomes of State Budget Revenue Decentralization in Vietnam (in Billion VND)   1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010e Revenues 65,352 70,612 78,489 90,749 103,773 121,716 175,908 170,375 202,683 241,117 341,057 430,549 454,766 559,170 Of Which: Local budget 19,264 20,280 19,571 22,269 25,463 30,545 45,521 74,688 96,113 87,291 99,249 133,966 134,656 169,384 revenue Share in total state budget 29.5 28.7 24.9 24.5 24.5 25.1 25.9 43.8 47.4 36.2 29.1 31.1 29,6 30.3 revenue Growth rate (previous year=100%) State budget 4.8 8 11.2 15.6 14.4 17.3 44.5 -3.1 19.0 19.0 41.4 26.2 5.6 22.9 revenue Local budget 8.5 5.3 -3.5 13.8 14.3 20 49.0 64.1 28.7 9.1 13.7 34.9 0.5 25.8 revenue Note: “Local budget revenueâ€? excludes revenue from transfers. Source: Ministry of Finance and the author Figure 5.5: Share of Local Budget Revenue in Vietnam 50 45 40 55 30 25 20 15 10 5 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 31 Of course, this figure excludes central government transfers to provincial governments. 139 TA X R E F O R M I N V I E T N A M Figure 5.6: Share of Local Budget Revenue in Vietnam 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Source: Ministry of Finance and the author 3. What is to be Accomplished by Revenue Assignments? The big expectation behind decentralization reform is that it can increase the overall efficiency of the public sector, strengthen democratic representative institutions, and ultimately lead to improved social and economic welfare for countries that decide to adopt it. But a critical assumption for the realization of these benefits is that decentralized governments will generally be accountable and responsive to citizens’ needs and preferences. At the same time, revenue assignments tend to play a critical role in the development of accountable subnational governments. There is general agreement among experts in decentralization that the increased accountability associated with decentralization can only be assured when subnational governments have an adequate level of autonomy and discretion in raising their own tax revenues. Thus, if effective fiscal decentralization requires a meaningful degree of revenue autonomy at the regional and local levels of government, the question is, which taxes should be allocated at these levels? This is known in the fiscal decentralization literature as the “tax assignment problem.â€?32 At this juncture we need to remember that revenue assignment is just one element in the design of the entire system of government decentralization and that if revenue autonomy is to work effectively to increase accountability it has to be within the context of other well- designed institutions in a decentralized system, including other fiscal issues (the assignments of expenditure functions, transfers, etc.); and political decentralization, with democratically elected officials, and administrative decentralization—in particular in what pertains to civil service issues. All are important in assuring good outcomes from decentralization. Besides increasing accountability, the other fundamental function to be fulfilled by revenue assignments is to provide adequate levels of financing so that subnational governments can deliver on the functions that have been assigned to them. 32 See, for example, Martinez-Vazquez, McLure, and Vaillancourt (2006). 140 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M However, this revenue adequacy requirement does not offer a concrete guide for the reform of revenue assignments; in fact, the same adequate levels of financing can be obtained from many different tax assignment combinations or via intergovernmental transfers. But in the case that all subnational government financing is from revenue sharing or other forms of transfers from higher-level governments, there is a danger that subnational governments will become spending agents of the center becoming less efficient,33 and that imposing a hard budget constraint on subnational governments will become more difficult. 34 Besides enhancing accountability and providing revenue adequacy, there are several other benefits from revenue assignments with tax autonomy. Subnational tax autonomy is the best way, if not the only way, to address in a permanent way the difficult problem of vertical imbalances, or mismatch of expenditure needs and revenue sources at different government levels. Adequate revenue autonomy is also a key indicator of subnational governments’ borrowing capacity and creditworthiness. There is also some evidence that countries with more revenue autonomy at the subnational level are countries that over time have shown greater macroeconomic stability.35 On the other hand, greater tax autonomy in revenue assignments can pose some additional challenges. Depending on the geographical distribution of economic activity and tax bases, greater subnational tax autonomy can lead to larger horizontal fiscal disparities across subnational governments. Richer jurisdictions can have the ability to finance their expenditure needs with relatively little effort while poorer communities may have to exert much greater tax effort with their residents to provide for similar expenditure needs. However, this is not a fatal consequence of subnational tax autonomy since these horizontal fiscal disparities can be well addressed through the proper design of equalization grants. With the realization that tax autonomy is paramount, we need to ask how much tax autonomy is needed. Is it the case that subnational governments should finance themselves entirely from autonomous tax sources? In reality, full own-financing by all subnational governments is generally not feasible or even desirable. Instead, the generally accepted rule is that subnational governments need to raise their own funds at the margin and operate with hard budget constraints, which means that revenue sharing and grants should represent only inframarginal 33 A number of recent studies (for example, Ter-Minassian 1997; Ebel and Yilmaz 2002) suggest that outcomes of decentralized spending depend on the form of financing used for these expenditures, with a crucial aspect being the extent of control that local governments can exercise over the sources of their revenues. 34 A hard budget constraint implies that those local governments given autonomy will be asked to balance their budgets without recourse to any end-of-year assistance from the central government and a clear understanding that there will be “no-bailoutâ€? at year-end or in case of debt default. See Rodden et. al. (2003). 35 Traditionally, it has been thought that greater subnational revenue autonomy may compromise the ability of the center to implement stabilization policies; in reality, the reverse seems to happen. It could be that greater subnational revenue autonomy leads to more conservative budget policies and lower deficits at all levels of government. See Martinez-Vazquez and McNab (2006). 141 TA X R E F O R M I N V I E T N A M funding.36 Only the richest subnational governments should be close to financing their full expenditure needs from their own revenue sources.37 Despite the well accepted consensus on the importance of subnational tax autonomy, in the international practice generally we tend to observe low levels of tax autonomy. One reason for this state of affairs, already mentioned above, is that central governments are reluctant to devolve taxing powers for fear of competing with local governments for the same taxing base and for fear of “losing control.â€? Ironically, many subnational governments are happy not to get tax autonomy because quite simply they do not want to take on the responsibility of making politically unpopular taxing decisions to meet their budget needs. Using intergovernmental transfers as opposed to revenue autonomy appears to be a much easier path for all concerned. To some degree, insufficient revenue autonomy can also be the result of the lack of administrative capacity in some subnational governments. When low capacity is combined with the desire to provide all subnational governments (regardless of size and capacity) with the same autonomous taxing powers, low levels of tax autonomy can follow. This situation (differing administrative capacities) poses a dilemma in decentralization design. A uniform intergovernmental fiscal system under which all subnational governments must operate has appeal. If all subnational governments have the same expenditure responsibilities and revenue raising powers, management of the system and evaluation of its success are made easier. Uniform treatment of all subnational governments also seems generally fairer. On the other hand, a more effective route for effective decentralization may be the adoption of an asymmetric tax assignment providing more tax autonomy to larger subnational governments with more capacity, according to transparent objective criteria, and let the smaller ones “grow into this roleâ€? over time.38 Although decentralized systems in some developed countries have high levels of tax autonomy, in reality it is quite rare, especially among developing countries to find significant taxing powers devolved to subnational governments at the onset of decentralization. Often, there is considerable reluctance from central government to let go part of its authority and control over taxes, which in turn is justified because of the need to facilitate attainment of proper capacity at the subnational level. However, these stumbling blocks generally linger for many years after the introduction of a decentralization program. With the passage of time a culture of financial dependency takes hold with subnational governments becoming accustomed to almost exclusively relying on central transfers for their financing needs. 36 For a full development of this argument, see McLure (1998). 37 Several things need to be done to get this right. First, there is a need to devise a sensible way to measure the expenditure needs of subnational governments and to keep these measurements reasonably updated. Next, there is a need to apply the golden rule for revenue assignment: tax assignments should be sufficient to fund the expenditure needs (net of conditional grants) of the wealthiest subnational governments. Sometimes, however, it may be advisable to break this rule somewhat and to have even the wealthiest subnational governments partly financed by central transfers. This may be because of vertical externalities in the use of tax bases, economies of scale in the administration of some taxes, the need to maintain the integrity (harmonized nature) of some taxes, and other considerations in tax administration, which are discussed below. 38 See Bird and Ebel (2007) on the possibilities and problems with asymmetric fiscal federalism. 142 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M 4. Desirable and Less Desirable Forms of Tax Autonomy Given that tax autonomy is required in the design of revenue assignments for subnational governments, we need to address two questions: (i) what type of revenue autonomy is desirable? (ii) what kind of tax instruments should be used to provide this autonomy? With respect to the form of tax autonomy, four dimensions have been identified.39 The first is which level of government has the right to choose the taxes that this given level can impose. There are good reasons, for example, to eliminate any ability of subnational governments to introduce internal tariffs. Provided that these general restrictions are to be in place, there is a choice between two approaches: an open list of taxes to be determined by the subnational governments within general limits and restrictions, or instead a closed list of allowable taxes determined at the national level from which subnational governments can choose. Overall, a closed list of subnational taxes is preferable because it avoids the introduction of nuisance taxes in some cases, or higher and inefficient distortionary taxes which can easily impede local economic development and growth. 40 Whether an open list or closed list approach is adopted, an additional decision needs to be made as to whether the base of specific taxes should be used exclusively by one level of government or whether these bases can be used simultaneously by several levels of government. The latter approach has the potential disadvantage of introducing vertical tax externalities due to the fact that one level will not typically take into account the impact its policies may have on the tax base and revenues of the other level of government.41 But several corrective policies can be implemented to correct for this type of externality such as providing intergovernmental grants. In practice, when an open list approach is chosen it is generally the case that the cohabitation of bases is allowed. In contrast, it is often the case that a closed list approach is used to eliminate the possibility of the cohabitation of tax bases.42 In practice, the choice between exclusive or shared tax bases comes down to weighing the advantages and disadvantages associated with each choice. As we just discussed, a disadvantage of cohabitation is vertical externalities. The most important disadvantage of using the exclusive basis is that typically subnational governments will be shut out of any opportunity 39 See Bird (2000); Bird and Ebel 2007; Boadway (1997); McLure (1998 and 2000a); and Musgrave (1983). 40 The international experience shows that providing subnational governments with freedom to select their own taxes (the open list approach) can easily backfire when subnational governments introduce highly inefficient (distortionary) forms of taxation. A recent example is provided by Indonesia, which adopted an open-list approach in the 2001 decentralization reform. See Alm, Martinez-Vazquez, and Indrawati (2004). In the international experience, where subnational governments are given more constitutional discretion, as in the case of some federal systems, open lists with some general restrictions are common. Closed lists are used more frequently in unitary systems of government. 41 See, for example, the discussion in Boadway, Marchand, and Vigneault (1998). 42 Sometimes the country Constitution, even in the case of some federal countries, is used to clearly delineate what taxes can be used at different levels of government; for example, this is the case in India, Pakistan or Switzerland. 143 TA X R E F O R M I N V I E T N A M to use significant (either in size or elastic over time) tax bases, thus drastically reducing any meaningful possibility of subnational tax autonomy. The imposition of exclusive tax bases can also lead to cumbersome tax structures. All things considered, a closed list allowing for the cohabitation of tax bases by different levels of government and using intergovernmental transfers to correct for vertical externalities is the preferred approach. The second dimension of tax autonomy relates to which level of government can legislate over the structure of the tax bases and which level has discretion to set the tax rates. Of the two types of autonomy for structuring subnational taxes, autonomy to define the tax base is generally less desirable than autonomy to set tax rates.43 Variations in the definition of the tax base, either through special exclusions from tax, deductions from the tax base, and credits against the tax liabilities can more easily lead to complexity and lack of harmonization across jurisdictions. The most important unwanted consequence of the lack of harmonization and complexity is the higher tax administration cost for all the jurisdictions involved and higher compliance costs for taxpayers who have tax obligations in several jurisdictions. Autonomy in the form of setting tax rates generally tends to be more desirable because it is simpler to deal with across jurisdictions for both tax administrators and taxpayers. Focusing on autonomy over tax rate setting has the additional advantage of generating political accountability because tax rates are the most visible element of a tax. Tax rate setting autonomy also may be preferred because it has a more direct impact on revenues and spending ability of subnational jurisdictions and because households and businesses have an easier time comparing their taxes to the services they receive. The third and last dimension of revenue autonomy refers to which level of government is put in charge of administering the various taxes. Centralized tax administration even of subnational taxes is likely to be more efficient because of economies of scale. However, administration by subnational governments of their own taxes is likely to enhance accountability at the subnational level if taxpayers are more aware of subnational taxes under this arrangement. This efficiency-accountability tradeoff is likely to differ for different taxes. For example, the efficiency gains from the centralized administration of subnational piggyback personal income taxes may dominate any increase in accountability generated by decentralized administration of those taxes. In contrast, there may be no significant efficiency gains in the centralized administration of subnational property taxes by comparison to the losses in local accountability implied by the centralization of the administration of these taxes. The administration of subnational taxes or even shared taxes by the central administration can present a problem with low incentives even for shared taxes when the central administration’s 43 The ability to change either base or rate opens up the possibility of fiscal competition among subnational governments. See, for example, Wilson (1999). Interjurisdictional fiscal competition can have both good aspects, such as offering choices to taxpayers and keeping public officials more accountable, and also bad aspects, such as a “run to the bottomâ€? with subnational governments competing with ever lower taxes to the point that they may lack sufficient revenues to provide basic services. The ability to change tax bases or tax rates may also give rise to “horizontalâ€? fiscal externalities, whereby the policies of one jurisdiction (for example, raising tax rates) can have an effect on the tax bases of other jurisdictions (raising their tax bases related to mobile taxpayers). Intergovernmental grants and other policies can be implemented by the central government to correct horizontal fiscal externalities 144 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M share in the revenues is relatively small. What this means is that when cost advantages make it desirable to centralize the administration, there is a need for setting incentive compatible arrangements between levels of government for the collection of taxes.44 Finally, we need to mention the issue of whether certain forms of tax revenue sharing on a derivation basis can contribute to the revenue autonomy of subnational governments. The more generally accepted view is that tax sharing is not a form of revenue assignment because subnational governments do not have a direct role in the structure and administration of the tax; in this view, revenue sharing should be considered just another form of transfer. In the minority view, shared taxes may be considered a form of tax assignment when the shared rates are stable over a period of several years and especially when the subnational authorities can influence the level of administration and affect the size of the tax bases. For these reasons, it is customary in many transitional countries, especially those in the former Soviet Union, to consider shared taxes as part of the own revenues of subnational governments.45 5. Significant Policy Issues with Revenue Assignments in Vietnam Despite the progress made on several fronts of decentralization policy in Vietnam, the current system still lacks a sufficiently well developed revenue assignment. In this section we identify the main drawbacks of the current system. The most significant issue with current revenue assignments in Vietnam is the lack of any significant revenue autonomy, in any of the forms we reviewed in the previous section. In particular, subnational governments in Vietnam have no ability to increase revenues at the margin. The only form of revenue autonomy for district and commune governments that currently exists is the ability to introduce certain fees, such as for waste collection and school tuition. Provincial governments have no discretion or autonomy in any dimension. Most financing for subnational governments in Vietnam is in the form of revenue sharing and other types of transfers. Although not strictly an issue in revenue assignments, it is also worth mentioning that some of the current mechanisms used for tax sharing are problematic. At present, revenues are split between the central government and the provincial jurisdiction where revenues are actually collected (the so-called “derivation principleâ€?). This creates problems of fairness, especially for the VAT and the enterprise income tax. The problem with the VAT is that it tends to be credited and debited in the jurisdiction where the headquarters of the enterprise is located and not necessarily where the tax is actually incurred. This helps large official cities (Hanoi or HCM City) and a few more industrially developed provinces. The use of the derivation principle for sharing VAT revenues also has led to some forms of market protectionism, whereby some provincial governments have been identified as creating artificial barriers to internal domestic trade. Another issue that should be mentioned is the elimination of explicit revenue assignments at the district and commune levels in the 2002 State Budget Law, which leaves it up to the provincial governments to define the revenue assignments for the local governments. This 44 See Martinez-Vazquez and Timofeev (2005). 45 See, for example, Martinez-Vazquez and Boex (2001), and Martinez-Vazquez, McLure, and Vaillancourt (2006). 145 TA X R E F O R M I N V I E T N A M arrangement provides provincial governments with budget flexibility and the ability to adapt to the particular circumstances in the province. It may allow for higher levels of equalization than would be possible were there a concretely specified revenue assignment. However, the lack of concrete revenue assignments also imposes costs on lower-tier governments since it can significantly limit revenue certainty and predictability.46 The consequences of the changes in revenues assignments for local governments introduced in 2002 have not been studied in any depth before 2007 although the casual evidence hinted at the existence of responsible behavior by most provincial governments but also to the presence of some abuses. A study of the Finance Academy (2007) reveals some issues related to real practice of expenditure and revenue assignments by provinces after the 2002 revised Budget Law came into effect. In the case of expenditure assignment the Finance Academy study found that some provinces actually changed their expenditure assignments between stability periods. For example, Ha Tinh province during the stability period 2004–06 assigned all education expenditures at the provincial level but in the stability period 2007–10 all education expenditures have been assigned to the district level. The study also highlighted that the shifting in expenditure assignment has taken place more often at the commune level. In the case of revenue assignments, the Finance Academy study also found that provinces tend to assign smaller taxes to communes while keeping for themselves larger revenue. The study also found instances of shifting in assignments during the stability period 2004–06; for example, that behavior was observed for Bac Can province. Of course, some of these changes may also be an indication of the drawback of using the policy of “stability periods,â€? especially at a time when Vietnam is experiencing very rapid economic growth and frequent policy changes in other areas (such as, raising minimum salary every year, higher expenditure demands due to inflation, and so on). Districts and communes with different economic conditions end up with significant disparities among them. Because of the limited revenue autonomy that provinces currently enjoy, it is really hard for poor provinces to mobilize additional revenues to meet their needs in between stability periods.47 On the other hand, “surplus provincesâ€? are generally in a better position to meet their expenditure needs because they are able to keep larger amount of revenue collected during economic expansions (as the amount of revenue they have to surrender to the center has been fixed at the beginning of the stability period for the whole duration of this period). Of course, these are the provinces that will experience a reduction in their tax sharing rates in the next round of stability periods. The findings in the Finance Academy (2007) study would seem to point out a direction of reform to introduce more stability and transparency in the assignment of expenditure responsibilities and revenue sources at the subprovincial level. However, this will not be a cure for all problems. For example, the 2007 CFAA (Country Financial Accountability Assessment) found that in the case of some provinces, such as Ha Nam province, have experienced difficulties in redistributing resources from high-revenue communes with real estate booms to poorer ones 46 This negative impact on certainty and predictability is partially offset by the practice of “stability periods.â€? 47 Recently, the government of Vietnam abolished inflation and economic growth correction allowances during the stability periods. However, there seems to be agreement within the Ministry of Finance that these allowances should be re-introduced in the next round of revisions to the Budget Law. 146 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M because the current Budget Law mandates specific minimum rates the center has to assign to the communes for the land and housing tax. As pointed out above, in Vietnam the administration and collection of all practically all taxes is centralized. A potential problem with centralized tax collections is the lack of incentives that central government bureaucrats (within the tax administration) may have to mobilize and collect subnational revenues. This problem is mitigated in Vietnam because there is de facto dual subordination of tax administrators to the central administration and the subnational authorities; this means that provincial and district officials can have a recognizable influence on the decisions and activities of tax administrators. For example, it is not unusual for the provincial authorities to provide bonuses to the tax administrators with a better performance in collections. Provincial authorities also usually provide office facilities, housing and other amenities to tax inspectors and other officials of the central government tax administration agency. The de facto dual subordination and the role played by subnational incentives for tax collection has been formalized in practice by letting subnational authorities retain a share of the collections that are above the targeted amount for tax collection in the central government plans.48 On the other hand, in some other countries, notably China (before the 1994 reform that created separate tax administrations for central and provincial taxes) but also other transitional countries like Russia and other former Soviet Republics), de facto dual subordination of central tax administrators was a source of problems as the subnational authorities “abusedâ€? their position to push provincial interests over those of the central government. This means that the practice of de facto dual subordination of central government tax administrators to subnational officials is unlikely to be a good solution in the longer run in order to address the issue of incentives, especially if after the reform of revenue assignments in Vietnam subnational governments are provided with significant tax autonomy. 6. Guidelines for Reforming Tax Assignments in Vietnam 6.1. What kinds of Tax Instruments are best suited for Subnational Governments? The criteria that should guide the assignment of revenue sources across different government levels in a country reflect the dual role of taxes. First, taxes simply provide the means to finance the provision of public goods and services, but taxes are also used as an instrument to achieve government policy objectives, such as the redistribution of income through progressive taxation. The classic starting point for the principles of tax assignments is Musgrave’s (1959) seminal work, where he argued that government policy should pursue three fundamental economic objectives: assuring macroeconomic stability; achieving a more equitable distribution of income; and assuring a more efficient allocation of resources when there is market failure. Although there continues to be some controversy on this, the general consensus among public finance economists is to agree with Musgrave that policy decisions on economic stabilization and income distribution are best assigned to the central government, while some of those related to allocation efficiency (how to best use the resources available to provide goods and services) may be assigned to local governments. 48 The “surplusâ€? collections retained at the subnational level include only 100 percent assigned subnational and shared taxes. The retention rates for surplus revenues have varied by province, from 100 percent for Hanoi, to 79 percent for and Ho Chi Minh City and up to 30 percent for all other provinces. 147 TA X R E F O R M I N V I E T N A M Beyond the guidance provided by Musgrave’s governmental roles, there are some characteristics of taxes that are commonly acknowledged as desirable regardless of whether these taxes are to be assigned at the central or subnational levels. These include: i. Revenue buoyancy, meaning that overall, revenues should change roughly in proportion to the economic base; ii. Equity, meaning that good revenue sources are “fairâ€? or equitable in the sense of horizontal equity under which taxpayers in similar circumstances should be treated similarly and vertical equity under which taxpayers with different incomes should pay according to their ‘ability to pay’; iii. Efficiency, meaning that the tax should have relatively low administration and compliance costs and create a minimum of distortion in the economy; and iv. Political acceptance, meaning that taxes need to be sensitive to the historical and institutional framework in a country. There are in addition, several other features that are desirable for taxes to be assigned at the subnational level.49 First, the benefit principle which relates revenue sources to the benefits being provided should be implemented to the largest extent possible. Fees paid for water or sanitation services are clear applications of the benefit principle. Second, subnational revenue sources should have a tax base that is relatively evenly distributed across jurisdictions. This helps to minimize fiscal disparities among subnational governments and reduces the burden put on equalization grants to allow a more uniform quantity and quality of services. Third, subnational tax sources should have immobile bases to minimize the likelihood and effects of tax competition among jurisdictions in a “race to the bottom.â€? 50 Fourth, subnational taxes should be geographically neutral in the sense that they do not interfere with domestic or international commerce, they do not distort the location of economic activity across the national territory, and they are not exported (so that the taxes levied by a subnational government are not borne primarily by residents in other jurisdictions). Fifth, subnational taxes should be implemented without undue costs of compliance and administration due the existence of multiple jurisdictions. Sixth, subnational taxes should exhibit generally stable tax bases over the economic business cycle; this means that revenue sources that are highly sensitive to general economic conditions (for example, profit taxes) should be assigned to the central government, which has greater ability to deal with cyclical fluctuations in revenues through borrowing and other means. Seventh, subnational taxes should be highly visible so that tax burdens are clearly perceived by local residents.51 And eight, subnational tax assignments need to be stable over time; a typical problem of transitional countries has been unstable assignments, with the assignments not being established in permanent laws but instead decided in annual budgets. Ad hoc assignments decided on an annual basis may also result in a lack of uniformity, unnecessary complexity, and perverse incentives toward revenue mobilization. 49 See, for example McLure (1998). 50 However, as we mentioned above, not all forms of tax competition are undesirable; moderate levels of tax competition can provide for more accountability among politicians and bureaucrats. 51 Of course, subnational governments are likely to think quite differently about this and they would likely prefer less visible taxes. 148 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M 6.2. Selecting Tax Instruments for Assignment at the subnational level Because there are no taxes that comply with all desirable criteria listed above, generally there is a need to reach a compromise between the different criteria. In practice naturally, there are disagreements on what should be in the minimum list of requirements for tax assignment at the subnational level. One such minimum list would include revenue autonomy at the margin, stable assignments over time, sufficient revenues for the wealthiest subnational government unit, and for taxes to be based as much as possible on the benefit principle and on less mobile tax bases. But, even though we cannot select one single best assignment, it is clear that the criteria allow us to select among better and worse tax assignments. 6.2.1. Better choices of subnational taxes Fees and user charges: The most straightforward way to raise revenue in accordance with the benefit principle is by charging user fees to cover the cost of providing specific local government services. Besides generating revenue for local governments, user charges are able to function as a pricing mechanism, thereby ensuring that locally provided goods are only used by local residents as long as their benefits exceed the cost to the user. One feature of this source of subnational revenues is that revenues raised from user fees and other nontax revenue sources are generally not available for general-purpose funding of local services or infrastructure. One general argument that is sometimes made against the reliance on user fees at the local government level is that user fees are potentially regressive. However, as we have already commented, one needs to be careful not to overstate the importance of the redistributional role of subnational governments. As noted earlier, equity and distributional issues are much better addressed at other levels in the overall fiscal system of the country. To the extent that the main purpose of “realâ€? licenses and user fees is to recover the administrative costs of issuing the licenses or the cost of providing the public services, it is important to price the service right. Requiring subnational governments to set the fee levels below the actual cost of provision imposes an unfunded mandate and it can easily lead to poor provision of services. While user fees provide important efficiency benefits, it is important to balance the cost of collecting and administering user fees with the amount of revenues collected; certain types of user fees involving many small transactions may be too costly to collect. The international practice of bundling the collection of a variety of fees into a single payment can make good sense. For example, it is possible to collect refuse collection fees or street lighting fees as a surcharge on property taxes. Property Taxes and Betterment Levies: There is ample consensus in the public finance literature that the real estate property tax is one of the best avenues to provide subnational governments with tax autonomy. Something else makes the property tax particular in the revenue assignments problem. The general rule is to 149 TA X R E F O R M I N V I E T N A M assign the property tax to local governments as opposed to intermediate level (provincial or regional) governments.52 Several features make property taxes especially attractive as a subnational tax. Most important, the property tax is a visible tax and thus conducive to political accountability; in addition the tax, for the most part, falls on an unmovable base. The more homogeneous are the distribution of property values and the distribution of the population in a jurisdiction, the closer the property tax comes to being a benefit tax.53 However, depending on how the property tax is structured, it can easily move away from the benefit link; for example, when the property tax burden falls just on a few classes of property, such as nonresidential property, there is little to be said about the property tax as a benefit tax. Other advantages of property taxes are their revenue potential and stability. Note also that from a vertical equity viewpoint the property tax can be progressive, especially in developing countries, and therefore can increase the overall vertical equity of the tax system. However, in practice, the property tax can be made regressive by exemption policies that target wealthier households.54 The property tax also has the desirable feature that much of the tax burden is quite likely borne by residents in the jurisdiction where the services financed by property taxes are provided. The property tax also imposes a relatively low compliance cost on taxpayers because taxpayer intervention in terms of the determination of tax liability is minimal, except in the case of appeals. Typically the property tax poses no significant problem of tax base competition with the central government, basically because this is not a tax that central governments tend to covet. 55 Finally, a part of property tax might be thought of as a charge for land that can lead to significant improvements in the quality of land use. The main drawback of the property tax is that, perhaps due to its visibility, it is likely unpopular with taxpayers and, as a result, also with public officials. Other drawbacks include the fact that it can lead to liquidity problems for homeowners with valuable real estate assets but low incomes.56 In addition, the property tax administration requires costly revaluation of property on a regular basis, and it is difficult to enforce, because the confiscation of property may be 52 However, despite the wide agreement on the advantages of the property tax as a subnational tax, subnational governments in developing and transitional countries make relatively little use of the property tax. On average, transitional and developing countries raise property tax revenues that are equivalent to only about 0.6 percent of GDP. See Bahl and Martinez-Vazquez (2007) for an investigation of this puzzle. 53 The balance between the services received by property owners and the property taxes they pay on their real estate typically can be capitalized into property values. That is, property taxes do not have to reduce the market value of dwellings if the general perception is that the quality of services provided by the local government is good. 54 See Bahl and Linn (1992) and Sennoga, Sjoquist, and Wallace (2007). 55 Of course, low interest may also reflect the perception that the property tax is complex and has low revenue potential vis-à-vis its associated political costs, although there are exceptions (for example, China, Indonesia, and Jamaica.). 56 Being “house rich and income poorâ€? can be a problem for elderly people. Some countries use special exemption schemes (“homestead exemptionsâ€? or “circuit breakersâ€?) to increase equity in the implementation of property taxes. 150 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M considered too extreme because of the political fallout. Finally, the property tax lacks revenue elasticity, meaning that the tax typically exhibits little automatic revenue growth. In practice there are several forms of the property tax. For example, some countries separate the taxation of land and improvements, or structures, and a few others tax only land values or rents. Although a tax on land tends to be more efficient, it also has less revenue potential and it is generally more difficult to administer properly, for example in terms of valuation or assessment of properties. There is another type of property tax in the form of “betterment leviesâ€? or lump-sum payments exacted up front by subnational governments from land and housing developers and also from homeowners as a charge for public service improvements, such as road paving, drain infrastructure, sidewalks, street lights etc, which all have a visible benefit on property values. Betterment levies can be useful in providing subnational governments with liquidity to invest in needed infrastructure; they also have the advantage of being more directly contractual than property taxes and therefore reinforcing the benefit principle feature in subnational government financing. There are different modalities for the administration of the tax, including centralized or central oversight over cadastres and reevaluation processes, which can make this type of tax even feasible in developing countries. Note that tax autonomy is largely preserved as long as subnational authorities are given some discretion over rate setting.57 Vehicle and Transportation Taxes: These are generally attractive taxes at the subnational level because of a strong link between the ownership of vehicles on the one hand, and the use of local services and infrastructure (particularly roads) on the other hand. In addition, subnational taxes and charges on vehicles can counteract the negative externalities associated with local traffic congestion and air pollution in the local area. Vehicle and transportation taxes also tend to have elastic revenues. It is perhaps for this reason that the central governments in some developing countries, wrongly, tend to keep them fully centralized.58 An annual tax on the registration of motor vehicles is often assigned at the subnational level. However, excise taxes levied one time on the purchase of new motor vehicles and taxes on gasoline and other motor fuels are more frequently assigned at the central level. It is also quite common that the revenues from taxes on motor fuels are earmarked to a national road fund. In some cases, the revenues in the road fund are shared between the central and subnational governments according to a formula with the funds dedicated to road maintenance and construction. Natural Resource Taxes (when resources are evenly distributed): There is at least a partial link between taxes on natural resource extraction and the benefit principle at the local level. Natural resource taxes can be justified at the local level to the extent that extraction activities use local infrastructure (for example, roads needed to transport heavy machinery and mined resources), place stress on other local infrastructure (temporary worker camps, hospital facilities required to treat injuries incurred by those working in this industry, 57 For international experience with the property tax see Bird and Slack (2004) and Bahl and Martinez-Vazquez (2007). 58 But there are some transportation taxes, such as in the case of air travel, which is rightly allocated at the central level, since air traffic control and other similar services, should be centrally provided. 151 TA X R E F O R M I N V I E T N A M and so on), and –depending on the type of extraction—may pollute the environment or cause other negative externalities increasing health costs of local residents. There has been growing interest in the fiscal decentralization literature in the pros and cons of the assignment of natural resource revenues to subnational governments.59 Notwithstanding the arguments for some form of local taxation of natural resources, there are two major arguments against local taxation of natural resources. First, in the case of geographically concentrated natural resources, local taxation could cause extensive horizontal fiscal imbalances (for example, the recent cases of Indonesia, Nigeria, and Russia). These fiscal disparities can lead to inefficient population migration and location of business. Second, given the high volatility of world commodity prices, local taxation of natural resources would not constitute a stable source of revenue. Therefore, some balance must be reached, especially in the case of the uneven geographical distribution of resources, between first, centralized taxation of natural resources to address disparities and avoid or correct for negative economic externalities, and second, sharing some of the revenues with subnational governments to compensate for the environmental damage of the extraction process and so on. Local Business Taxes: Certain forms of business taxes or business license fees are justified at the subnational level as an indirect but administratively easier way to tax income of business owners (especially nonwage incomes), and as a benefit tax for the services and infrastructure provided by subnational governments. Where it is not feasible to recoup costs of local government services through user charges, some form of broad-based levy on general business activity is warranted. To avoid economic distortions, ideally the tax would apply equally to labor (payroll) and capital (assets) used by businesses.60 Where this form of business taxation at the subnational level is not feasible it may be possible to use business permit charges which may vary by type, size, or location of the business. Excise Taxes: Subject to the area size, cross-border trade and smuggling limitations, excise taxes have good potential at the subnational level, especially when they are designed as piggyback taxes on the national taxes. An advantage is that excises tend to be more politically acceptable, can be easily administered in coordination with national wholesalers as withholding agents, and allow for rates differentiated by province.61 Moreover, the benefit principle accords well with 59 See, for example, McLure (1996) and Bahl and Tumennasan (2004). 60 An example of this is a business value tax (BVT) as discussed in Bird (2003) and calculated by adding payroll, interest, rents, and net profits on the basis of annual accounts. An example of a BVT is Italy’s regional business tax (known as the IRAP) prior to the elimination of payroll from the tax base in 2003. The IRAP is discussed in Keen (2003). 61 For example, some OECD countries allow subnational government surcharges on excises. In the Netherlands, provinces impose a surcharge on the motor vehicle tax levied by the central government. Provinces are free to set the rate of the surcharge, subject to a ceiling imposed by the central government. 152 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M the assignment of (destination-based) excises on alcohol and tobacco to the subnational level (to the extent that the latter is responsible for health care) and on vehicles and fuel (to the extent of subnational government involvement in road construction and maintenance). The ability to charge differential rates across subnational jurisdictions is of course limited by the possibility of cross-border trade and smuggling. The extent to which excise piggyback surtaxes can be used at the local level depends also on the technology of product distribution and points of sales. An interesting aspect of excise taxation at the subnational level is the taxation of public utility services. There is significant revenue potential in some of these services, as in the case of electricity and phone services. Besides revenue potential and administrative ease, subnational excises on public utility services are attractive because of the benefit principle; for example, excises on electric consumption and phone services should be in most cases good proxies for the demand of local public services by both households and enterprises. Compared to other commodities, taxation of public utilities would be associated with relatively low distortions, as most utilities show relatively low price-elasticity of demand. Also, the demand for public utilities has been shown to be income elastic, which would bring two additional benefits to this form of subnational taxes in the form of revenue buoyancy and a fairer distribution of tax burdens. Flat-rate Piggyback Income Taxes Progressive income taxes are best assigned at the central government level, because, as discussed above, income redistribution should be an objective pursued by the central government. Another reason for this assignment is that progressive income taxes tend to act as automatic economic stabilizers and macroeconomic stabilization should also be primarily a function of the central government. Nevertheless, there are several possibilities for the taxation of personal income by subnational governments. The most commonly used form of subnational income taxation internationally is a flat-rate income tax as a surtax or “piggybackâ€? tax on the central government personal income tax. This type of tax is almost always collected by the central government administration and shared on a derivation basis.62 To enhance revenue autonomy, local governments may be allowed discretion in setting the flat rate between minimum and maximum rates, which are centrally legislated.63 A flat rate local piggyback income tax easily satisfies the benefit principle and, being quite visible, it promotes political responsibility and accountability at the subnational level. This is also an elastic tax with revenues increasing commensurate with income, so that as the demand for local services increases with income, so do tax revenues. 62 Generally speaking a local income tax should be levied at the place of residence because it is there where most taxpayers consume subnational government services. However, because of administrative convenience, subnational piggyback taxes are often withheld at source at the place of work by employees. However, it is quite feasible to distribute the funds according to where workers reside. 63 Other forms of tax autonomy that are less desirable but nevertheless practiced include the ability to modify the base of the tax by providing more or less deductions and exemptions. 153 TA X R E F O R M I N V I E T N A M 6.2.2. Worse Choices of Subnational Taxes As we have seen, the principles of tax assignment do not provide a deterministic list of taxes to be assigned at the subnational level; however, those principles are helpful in identifying better choices and what may be poorer choices. As just discussed above, progressive personal income taxes are not a good choice for tax assignment at the subnational level; ultimately, it would seem to make little sense to have income redistribution only within the boundaries of subnational jurisdictions, since richer taxpayers tend to live in richer jurisdictions. In addition, the mobility of high income taxpayers and businesses could easily lead to distortion in the location of economic activity. Another tax that is ill-equipped for application at the subnational level is the corporate income tax or profit tax. This is a tax more appropriately assigned to the central government level because of its link to macroeconomic stabilization and, to the extent that corporations are owned by wealthy individuals, this tax also affects income redistribution. Perhaps even more relevant, is that even when levied by the central government, the corporate income tax hardly meets sounds principles. There are no well accepted reasons that explain why incorporated businesses benefit more from public services than do unincorporated ones, or that the benefits received vary with profits. The main justification for a corporate income tax is to prevent avoidance of individual income tax through incorporation; the corporate income tax also works well as a withholding tax on foreign shareholders, who otherwise only may have to pay tax in their countries of residence. Clearly, it is administratively easier to tax profits at source rather than as individual income after distribution among shareholders. At a more practical level, the assignment of profit taxes at the central level is justified by the difficulty of apportioning well the profits of enterprises across subnational jurisdictions where they operate. The few countries that have a corporate income tax at the subnational level attempt to apportion the nationwide profits of enterprises among subnational jurisdictions by using a formula. These apportionment formulas generally use a weighted index with combinations of three factors: payroll, assets, and sales. Despite these formulas, the allocation of profits (and therefore the apportionment of tax revenues) across jurisdictions tends to be quite arbitrary because of the imprecise link between the location of those factors and the generation of profits. In a few other countries where no apportionment formulas are used, the typical norm is to share the revenues between the central and subnational governments on a derivation basis, that is, according to the jurisdiction where the taxes have been actually collected. This practice leads to an even more arbitrary distribution of revenues, since the shared revenues stay in the very few jurisdictions where companies are registered or have their headquarters. This means that the capital of the country and a few other large cities where enterprises have their headquarters tend to benefit unfairly vis-à- vis jurisdictions where the enterprises have factories and other forms of economic activity that use local resources and public services. As we saw above, this is a current problem in Vietnam’s decentralization system. Another tax that traditionally has been thought a poor choice for assignment to the subnational level is the VAT. The main difficulty lies in the fact that the debiting and crediting of the 154 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M VAT is likely to take place in different subnational jurisdictions, which generally will imply an arbitrary apportionment of VAT revenues across those jurisdictions.64 This also makes it problematic to share VAT revenues with subnational governments on a derivation basis. The problem here is, as in the case of the profit tax, is that the tax revenues get arbitrarily allocated to the subnational government where the headquarters of the company are located and not to where the economic activity actually takes place. However, in order to address that arbitrariness, there are countries that use a formula to share VAT revenues with subnational units.65 7. Recommendations for Reform Realizing the full benefits of fiscal decentralization in Vietnam will require the reform of revenue assignments at the provincial and local levels of government. In this paper we have argued that besides adequate revenues to fund the public expenditure needs of subnational governments, what we most want from revenue assignments is to increase the accountability of subnational government officials to their residents. This will be fundamentally achieved by granting subnational governments a significant level of tax autonomy. Before we go into our recommendations for how best to achieve the goals of enhanced revenues and autonomy of subnational governments it will be important to outline the general contours of the reform. We must realize that the full financing of subnational governments from autonomous tax sources will not be feasible. Instead the goals should be that all subnational governments will have a capability to increase their own funds at the margin by exercising greater tax effort. Funds from revenue sharing and grants should represent only inframarginal funding. Operationally, this will translate into the rule that own revenue sources should fund the expenditure needs (net of conditional grants) of the wealthiest subnational governments, and the revenue needs of the relatively poorer subnational governments should be supplemented with equalization grants. Many elements of this formula are already in place in Vietnam. What Vietnam’s decentralization system lacks is tax autonomy and granting a significant degree of tax autonomy should be the main thrust of the reform in revenue assignments. However, we have seen that not all forms of tax autonomy are equally desirable. All things considered, the best way to provide subnational governments with tax autonomy is to have a closed list of taxes for which subnational governments can set tax rates within some minimum and maximum values that are legislated by the National Assembly. Subnational governments would not be allowed to change any other aspect of the taxes, such as the definition of the tax base. 64 Revenue sharing on a derivation basis for the VAT also means that, as in the case for the sharing of corporate income taxes, the tax tends to be paid according to the place of registration or the location of the headquarters of business firms. 65 For example, the VAT can be shared on the basis of population (as in Belarus and Germany), or on the basis of the regional shares in aggregate consumption (as in Canada’s Maritime Provinces, Japan, or Spain). In the case of Canada’s Harmonized Sales Tax for the Maritime Provinces, all three provinces have a uniform rate that piggybacks on the federal VAT. 155 TA X R E F O R M I N V I E T N A M Good choices for enhancing revenue autonomy at the subnational level in Vietnam include: • Selectively making better use of some fees and charges for excludable services under the benefit principle at the provincial and local; however, some exclusions may be used for poor households or in areas with high incidence of poor households and where equalization grants should be providing additional needed revenues. • The assignment of the real estate property tax to local governments with discretion to set the tax rate between maximum and minimum values set in the national law. While the assessment of property values could be left to the national tax administration, other administrative functions (such as collections) could be decentralized asymmetrically depending on the administrative capacity of the local government. Selected local governments could also be allowed to charge betterment levies for the construction of new housing and the urbanization of additional land for construction.66 • The assignment of a motor vehicle tax with variable rates set in the national legislation to the provincial governments. This would be an annual tax associated with the renewal of the annual license to operate the motor vehicles. • The assignment of a piggyback flat rate personal income tax to provincial governments. This tax would allow provincial governments to introduce a flat rate set between a minimum and a maximum rate (for example, 1 percent and 4 percent) legislated in national law. The tax would be charged on the same tax base as the national income tax and paid on a residence basis. • The Government should also explore the feasibility of using piggyback taxes or surcharges on the special consumption (excise) taxes on alcohol and tobacco. This would allow provincial governments to add a surcharge to the national excise taxes and they would be withheld by producers or distributors according to the destination (province) of the merchandise shipments/sales to retailers. A second thrust of the reform in revenue assignments would be the improvement of the apportionment of shared taxes between central government and provincial governments. In the case of the VAT, tax revenues would not be shared any longer on a derivation basis but rather according to a formula. This formula could be based on either an equal per capita basis, or in proportion to estimates of shares in final consumption for each province. In the case of the profit tax, tax revenues would not be shared any longer on a derivation basis but rather on the basis of a formula. This formula would be based on the geographical distribution of the enterprise payroll, and to the extent possible also on the geographical distribution of the enterprise assets and sales. The sharing of personal income tax revenues with subnational governments would be according to the place of residence of workers, where most local services get consumed, as opposed to the place of works as is now the case. A third thrust of the reform in revenue assignments would be to introduce an explicit list of tax and other revenue sources for local governments (districts and communes). The current flexible unwritten approach to revenue assignments for districts and communes has 66 Betterment levies, like one-time or multiyear charges, are linked to infrastructure improvements directly benefiting certain developers or homeowners, such as improvements in street lighting, sidewalks and so on. 156 TA X R E F O R M A N D R E V E N U E A S S I G N M E N T S I N V I E T N A M advantages but its costs in terms of uncertainty and unpredictability for local governments outweigh those advantages. In terms of tax administration, we recommend that, for the time being, tax collections for subnational taxes remain centralized as there are advantages of economies of scale and information in the centralized administration of most of the taxes that have been suggested to provide subnational governments with more tax autonomy and enhanced tax accountability. However, at some point in the near future there should be an examination of how to introduce “incentive compatibleâ€? contracts between the central tax administration and the subnational governments for the collection of subnational taxes; this can be in the form of benchmark payments linked to the performance of collections. In the longer term, for certain taxes, such as the real estate property tax, some of the administration functions including registration and collections could be devolved to subnational tax administrations. This devolution would be phased out cautiously and in an asymmetric way taking into account the administrative capacities of subnational governments. 157 Integration and Government Revenues: Assessing the VI Implications of Vietnam’s Accession to the WTO By Robert Warner 1. Summary This paper presents an analysis of some of the implications for future tax policy of Vietnam’s tariff reduction and binding commitments, entered into as one of the agreements associated with accession to the World Trade Organization. The main observations arising from the analysis are: • Implementing Vietnam’s tariff binding commitments will reduce the average and dispersion of tariffs compared to the current rates, with a consequent improvement in the incentives for investment resource utilization. But compared to some other countries, Vietnam does not seem to have taken the opportunity of accession to lock in a tariff regime more suitable for an open trading economy. Some bound rates are higher than current applied rates, and over 50 percent of bindings are at rates of 20 percent or higher. • Vietnam’s phasing in period is quite long (12 years in total before all bindings are at their final rates. However, over 99 percent of final bound rates will be in force after 7 years. • The tariff cuts that will follow from implementation of bindings look set to reduce tariff revenues. But a large part of this revenue loss could be made up by reducing concessions and exemptions and improving the efficiency of customs administration. Shortfalls could also be made up by adoption of a different approach to taxing petroleum products. More importantly, however, any assessment of the revenue effects needs to be carried out in a way that takes account of the broader range of changes affecting the revenue and expenditure system: growth should expand the base for some revenue sources, for example, and reductions in subsidies will reduce demand for revenue. Some revenue functions could be shifted from the tariff on the special consumption tax, and improving the structure and administration of the VAT could generate significant yields. • The pattern of bindings seems to reflect a fairly typical trade negotiator’s approach to tariff setting: most of the big cuts implied by the bindings seem to be in areas where the tariff is largely redundant — where Vietnam is a successful exporter or where import competition is limited. • The cuts will reduce effective protection across the traded goods element of the economy: but pockets of high protection will be allowed to persist if Vietnam maintains rates at bound levels. 159 TA X R E F O R M I N V I E T N A M The findings also suggest that it would be useful to look harder at ways of analyzing the impact of changes in the structure and administration of other taxes and revenue instruments. The analysis of the revenue impacts of tariff bindings presented in this paper is fairly rudimentary, and has been constrained by lack of access to data that should inform a thorough assessment of the revenue effects of large policy changes. It would be useful to try to develop a more robust analytical framework for revenue analysis. 2. Introduction Vietnam’s accession to the WTO will impact on revenues in a number of ways: there will be first round effects as a result of implementation of the various commitments, and later effects as the response of producers, traders, investors and consumers to the changes that implementing the commitments will bring. The first round effects will include: • an impact on the rate structure of some taxes (eg customs duties, some of which will have to fall as a result of tariff bindings); • a consequent impact on the valuation base for other taxes on imports whose tariff rates change (for example, the valuation base for VAT on imports will change as import duties change); • changes in the way in which the value for duty of imports will be determined (application of the GATT Valuation agreement); • other changes in administration of trade taxes and in charges levied on imports; • elimination of some concessions on import duties resulting from implementation of the TRIMs agreement. The second round effects will include the following. • A change in the trajectory of imports over time as traders respond to price reductions consequent on tariff changes and other charges. This might in turn have some impact on the exchange rate, affecting the value for duty. • A change in the transaction base for all other taxes if the economy grows more rapidly as a result of implementation of the agreement. This will depend to some degree on which parts of the economy respond most vigorously. By no means all transactions and incomes are taxed under the current system of direct and indirect taxes: small informal enterprises may escape some parts of the tax net, and significant elements of consumption are exempted from the VAT. Nontax revenues may also be affected: land fees and charges, for example may be affected by changes in investment, and changes in the zoning of land and its valuation of land taxes. This paper presents an assessment of some of these effects: primarily the immediate impact on tariff revenues of the implementation of tariff bindings. It updates some modeling presented in a paper prepared for the ADB (CIE 2007), and draws on earlier analysis carried out for the World Bank during preparation of its support for Customs reform (CIE 2004). 160 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S 3. Vietnam’s system for Taxing and Managing Trade The Government of Vietnam raises revenues from taxes, fees and charges. Because trade is such an important part of the Vietnamese economy, imports are linked to most of the government’s revenue sources: they play an important role in the production and income of most enterprises that are subject to taxes on income or that collect value added taxes on behalf of the tax department. There are, however, three sources of revenue that are directly dependent on imports: tariffs, VAT on imports and the special consumption tax. As table 6.1 shows, taxes on external trade, including export and import tariffs, import surcharges and VAT collected on imports has accounted for up to 25 percent of total collections in recent years (In 2010, the percentage was 23 percent.). However, there has been a steady decline in the share of revenues contributed by export and import tariffs: from 18 percent in 2002 to 11.2 percent in 2010. Taxes raised on Imports Import Tariffs Vietnam’s import tariff has 3 schedules: normal, preferential and CEPT. There are now also preferential rates applying to goods from countries with which Vietnam has a preferential trade agreement, bilaterally or as a member of ASEAN. The preferential schedule broadly corresponds to an MFN schedule, since the normal schedule applies to imports from countries with which Vietnam does not have a trade agreement. Table 6.2 summarizes the MFN tariff schedule for 2005 through to 2009. The data for 2007 and beyond includes the effects of changes in applied rates necessitated by implementation of the tariff binding commitments made by Vietnam upon accession into the WTO. As the table shows, the simple average of tariff rates has fallen from 18.4 percent to 11.3 percent. Table 6.1: Composition of Revenues 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 In percent of total revenue Tax revenue 73.1 74.1 77.5 72.2 76.6 79.6 82.1 84.7 84.8 86.0 Corporate 24.8 24.1 31.0 29.8 33.0 38.2 35.1 29.9 19.4 25.7 income tax Individual 2.0 1.9 2.0 1.8 1.9 2.0 2.2 3 3.1 4.7 income tax Capital user 1.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 chargesa Land and 0.3 0.2 0.3 0.2 0.2 0.2 0.2 0.3 0.2 0.3 housing tax Licence Tax 0.4 0.3 0.5 0.4 0.3 0.3 0.3 0.4 0.3 0.2 Tax on transfer 1.2 1.1 1.2 1.4 1.3 1.3 1.3 1.7 2.1 2.3 of properties Tax on land use 0.3 0.2 0.3 0.3 0.5 0.5 0.4 0.7 0.1 — rights Value added tax 18.6 21.3 21.6 20.3 21.1 20.7 28.0 29.2 32.1 28.4 (VAT) 161 TA X R E F O R M I N V I E T N A M Of which VAT 5.2 8.0 9.9 .. .. .. .. on importb Special 6.0 6.0 5.8 6.7 7.2 6.5 6.1 7.5 9.2 8.7 Consumption Tax (Excises) Slaughter tax 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Agricultural Tax 0.8 0.7 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 Import and 16.8 18.0 14.6 11.3 10.9 10.0 8.4 11.7 14.2 11.2 export duties Other taxes on 0.1 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 trade Other taxes 0.2 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Nontax revenue 25.0 24.0 20.7 26.2 22.3 19.0 16.8 13.5 17.4 17.6 Grants 1.9 1.9 2.0 1.5 1.1 1.4 1.1 2.2 1.8 1.0 Total Revenue 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (inc. grants) Memo items In percent of GDP Total Revenue 21.6 22.7 24.9 26.7 25.9 27.1 24.9 28.1 27.4 28.6 (inc. grants) Import and 3.6 4.1 3.6 3.0 2.8 2.7 2.1 3.3 3.9 3.2 export duties Source: CIE 2004a; IMF 2006, 2007; Hoai personal communication. Note: a. Abolished in late 2001. b. Not reported after 2003. Table 6.2: Summary of Vietnam’s Preferential or MFN Tariff Structure, 2005–09 Tariff 2005 2006 2007 2008 2009 bracket Tariff Tariff Tariff Tariff Tariff % % % % % % lines lines lines lines lines 0 3,087 28.9 3,268 29.6 3,270 29.1 2,761 30.3 2,774 30.3 >0–5 1,920 18.0 2,074 18.8 2,207 19.7 1,918 21.1 1,909 20.9 >5–10 1,048 9.8 1,187 10.8 1,210 10.8 965 10.6 950 10.4 >10–15 177 1.7 290 2.6 860 7.7 768 8.4 836 9.1 >15–20 855 8.0 859 7.8 1,123 10.0 808 8.9 823 9.0 >20–30 1,285 12.0 1,189 10.8 1,281 11.4 1,108 12.2 1,165 12.7 >30–40 985 9.2 1,046 9.5 646 5.8 581 6.4 505 5.5 >40–50 1,001 9.4 778 7.1 130 1.2 63 0.7 55 0.6 >50–100 319 3.0 288 2.6 437 3.9 122 1.3 124 1.4 >100 12 0.1 50 0.5 61 0.5 5 0.1 5 0.1 Total 10,689 100.0 11,029 100.0 11,225 100.0 9,099 100.0 9,146 100.0 Range 0–150 0–150 0–150 0–140 0–140 Mean 18.4 17.3 14.3 11.6 11.3 CVa 120.7 127.8 131.8 118.8 120.8 Source: IMF 2007; WTO 2011. Note: a. Coefficient of Variation (standard deviation as a percentage of the mean) The CEPT schedule applies to imports from ASEAN countries of goods on Vietnam’s inclusion list. For those goods not on the inclusion list (for example sensitive agricultural products which have a longer time period to be brought under the CEPT umbrella), the MFN rate 162 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S applies. Table 6.3 summarises the CEPT schedule from 2005 to 2013. From 2006 onwards, all products except those on the sensitive list are to be at rates between 0 and 5 percent. Note that for goods from ASEAN countries to be eligible for the CEPT rate, they must meet certain rules of origin requirements. Some imports (for example from entrepôt ports like Singapore) will be subject to the MFN rate, because they do not meet the requirements under the rules of origin framework specified for the agreement. There does not seem to be any readily available information on the proportion of imports originating in ASEAN countries to which the CEPT rates are applied. Table 6.3: Summary of Vietnam’s CEPT tariff schedule Tariff Bracket 2005 2006 2007 2008 2009 2010 2011 2012 2013 0 31.9 53.0 52.1 56.1 56.5 57.0 57.0 57.0 57.1 >0–5 50.8 46.5 45.7 42.8 42.4 41.9 42.0 42.1 42.4 >5–10 13.2 0.0 0.1 0.0 0.2 0.1 0.1 0.3 0.0 >10–15 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 >15–20 3.7 0.2 0.1 0.1 0.0 0.1 0.3 0.0 0.0 >20 0.2 0.3 2.0 0.9 0.9 0.8 0.6 0.6 0.6 Mean 4.5 2.5 2.5 2.8 2.7 2.7 2.6 2.5 2.4 CVa 108.4 144.1 133.7 255.8 256.2 258.7 240.5 230.2 205.8 Source: ASEAN Secretariat 2011. Note: a. Coefficient of Variation (Standard Deviation as a Percentage of the Mean). In recent years, the share of Vietnam’s imports coming from ASEAN has been declining, from about 30 percent in 1998 to around 25 percent in 2006 and to 19.3 in 2010 (table 6.4.) Thus at least 80 percent of Vietnam’s imports are now probably subject to MFN rates. (Vietnam has preferential trade agreements with other partners, but in some cases, such as the agreement with the United States, the preferential rates have largely been matched by the post WTO accession MFN rate.) Table 6.4: Imports by Region of Origin: ASEAN and the Rest of the World, 1995–2005 ASEAN Rest of World Total US$m % US$m % US$m % 1995 2270.1 27.8 5885.3 72.2 8155.4 100.0 1996 2905.5 26.1 8238.1 73.9 11143.6 100.0 1997 3220.5 27.8 8371.8 72.2 11592.3 100.0 1998 3344.4 29.1 8155.2 70.9 11499.6 100.0 1999 3290.9 28.0 8451.2 72.0 11742.1 100.0 2000 4449.0 28.5 11187.5 71.5 15636.5 100.0 2001 4172.3 25.7 12045.7 74.3 16218.0 100.0 2002 4769.2 24.2 14976.4 75.8 19745.6 100.0 2003 5949.3 23.6 19306.5 76.4 25255.8 100.0 2004 7768.5 24.3 24200.3 75.7 31968.8 100.0 2005 9326.3 25.4 27434.8 74.6 36761.1 100.0 2006 12546.6 27.9 32344.5 72.1 44891.1 100.0 163 TA X R E F O R M I N V I E T N A M 2007 15908.2 25.3 46856.5 74.7 62764.7 100.0 2008 19 501 24.1 61212.8 75.9 80713.8 100.0 2009 13815 19.7 56152.8 80.3 69967.8 100.0 2010 16412 19.3 68389 80.7 84801 100.0 Source: GDC; GSO 2008; MOIT. There is an extensive set of exemptions from import duties provided for in current legislation. Box 6.1 presents a summary of these exemptions from the most recent IMF summary of the tax system and from information available from the Ministry of Finance. Box 6.1: Exemptions from Import and Export Duties Duty exempt imports and exports include: • Temporary imports for re-export or temporary exports for re-import. • Goods which are moveable assets. • Exports and imports of foreign organizations and individuals enjoying diplomatic immunities in Vietnam at levels stipulated by the Government consistent with international treaties of which the Socialist Republic of Vietnam is a member. • Goods imported for processing for a foreign party which are exported or goods exported overseas for processing for a Vietnamese party which are reimported pursuant to a processing contract. • Goods which are imported or exported within the quantity of duty-free baggage of individuals upon exit from or entry into Vietnam. • Selected goods which are imported in order to form fixed assets of a project which is an encouraged investment or of a project which is funded by Official Development Aid (ODA) as stipulated in Decree 87/2010/ND-CP. • Agricultural species imported for investment projects as regulated. • First time import of certain equipment and goods stipulated by Decree 87/2010 /ND- CP to form fixed assets of encouraged investment or ODA funded investment in hotels, offices, apartments for lease, residential housing, commercial centers, technical services, supermarkets, golf courses, tourist resorts, sporting resorts, entertainment areas, medical diagnosis and treatment establishments, and entities providing training, cultural, financial, banking, insurance, auditing, and consultancy services. • Selected goods imported in order to support petroleum operations. • Shipbuilding enterprises are exempted from export duty for ship products and import duty for imported machines, equipment to form fixed assets; materials, semi-products that cannot produced domestically. • Imported materials that cannot produced domestically for software development. • Goods which are imported for direct use in scientific research and development of technology, including machinery, equipment, accessories, materials and means of transportation which cannot yet be produced domestically, and technology which is not available domestically; and scientific books and data. • Raw materials, materials and component parts imported for production of projects on the list of sectors where investment is especially encouraged or on the list of regions with especially 164 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S • difficult socioeconomic conditions (Decree 87/2010 /ND-CP referred) or production of components, mechanic, electric and electronic items shall be exempted from import duty for a duration of five years from the commencement of production. • Goods which are manufactured, processed, recycled or assembled in nontariff zones without using raw materials or component parts which are imported from abroad, upon import thereof into the domestic market; in the case where raw materials and component parts imported from abroad are used, upon import of goods into the domestic market, import duty must be paid on that part of the goods which is imported raw materials or component parts which form a constituent part of such goods. • Goods which are imported for selling in duty free shop. Source: IMF 2007; MOF 2010. VAT Vietnam’s VAT, which replaced a multirate turnover tax in 1999, was restructured in 2004 and again in 2009, with the number of rates reduced from 4 to 3 (a previous top rate of 20 percent was eliminated). Table 6.5 summarizes the structure of the VAT, as it applies to goods. A range of commodities are exempted from the tax, including unprocessed products of agriculture and fisheries. In addition, a set of commodities are exempted if imported for a particular use (for example, machinery, equipment and means of transportation that are imported). Further, imports funded by nonhumanitarian or nonrefundable aid are also exempt from the tax. The VAT is levied on the duty paid price of imports (using the value that is used for determination of the duty payable), so that changes in import duties impact on the amount of VAT collected on imports, as do changes in the basis for determining value for duty. An important feature of the VAT is that the system provides for either a credit or a subtraction system to be used in for determining tax payments. The subtraction system cannot readily be applied with a multiple rate tax, and the coexistence of the two systems can turn the VAT into a variable turnover tax. (Most enterprises, however, use the credit method.) Table 6.5: Application of VAT to Goods since 2009 Treatment / Commodity Exempted Unprocessed or semiprocessed products of cultivation, husbandry, aquaculture or seafood • Products which are animal breeds and plant varieties • Salt products • Irrigation and drainage, soil ploughing and harrowing; dredging of intrafield canals and ditches for agricultural production • Certain specialised machinery, equipment or means of transportation that cannot be produced locally for direct use in scientific research and technological development activities; for prospecting, exploring and developing oil and gas fields • Leased airplanes, drilling rigs and vessels • Transfers of land use right, state-owned dwellings sold by the state to current tenants • Credit services and investment fund services, life and some other forms of insurance 165 TA X R E F O R M I N V I E T N A M • Technology transfer and software • Special-purpose weapons and military equipment for security and defense purposes • Goods transferred out of border gate or transited via the Vietnamese territory; goods temporarily imported for re-export • Medical and veterinary services, teaching and job-training activities, selected public services such as sanitation and water drainage, public transportation in inner cities • Newspapers, magazines, certain books • Imported goods that are: funded by humanitarian or nonrefundable aid; gifts to state and other bodies; • Some exported unprocessed minerals • Gold imported in bars • Appliances for the disabled • Goods and services of business individuals with a monthly average income level lower than the common minimum salary level applicable to domestic organizations and enterprises. Zero rated • Exports of goods and services to a foreign country or an EPZ, or products deemed to have been exported, except for unprocessed minerals Subject to rate of 5 percent Clean water • Fertilizer, pesticides and growth stimulants for plants and animals • Medical equipment and instruments, medicines and pharmaceuticals • Various agricultural products, including fresh foodstuffs, sugar and by-products, jute, sedge, bamboo and thatch products, animal feeds) • Children’ toys and book, except some kind of book • Teaching aids • technical and scientific services; some kind of services for agricultural production such as services of digging, embanking and dredging canals, ditches, ponds and lakes for agricultural production; • Preliminarily processed rubber latex; preliminarily processed turpentine; nets, main ropes and fibers for making fishing nets • Special-purpose machinery and equipment for agricultural production;and Cultural, exhibition, physical training and sports activities; art performances; film production; film import, distribution and screening. Subject to a rate of 10 percent • All goods not exempted or subject to a rate of 0 or 5 percent. Source: IMF 2007; Shukla 2007; VAT law. The Special Consumption and Excise Tax The special consumption tax is effectively an excise tax which in principal applies equally to imports and locally produced goods, and in 2009 the tax was replaced by an excise tax. Goods directly exported by production or processing enterprises are not subject to the tax, nor are imports funded by humanitarian or nonrefundable aid, or goods imported by units of the armed forces or donated to political and social organizations. The base for the tax when levied 166 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S on imports is the value for duty plus the import duty paid. (The tax is also levied on certain services, including dancing halls, massage parlors, casinos, lotteries and betting activities, and golf course memberships and services). Table 6.6 summarizes the current coverage and rate structure of the tax, and table 6.7 summarizes the excise rate structure. Table 6.6: Special Consumption Tax on Goods (as at August 2007) Goods Tax rate (%) Cigarettes 55 Spirits • 400 proof or more 65 • Between 200 and 40º proof 30 • Below 200 proof, including wines brewed from fruit 20 • Medicinal spirits 15 Beer • Bottled, canned and fresh beer 75 • Draught beer 30 Automobiles • 5 seats or less 50 • 6 to 15 seats 30 • 16 to under 24 seats 15 Petrol, naphtha, reformate compounds and other components to mix with petrol 10 Air conditioners (under 90 000BTU capacity) 10 Playing cards 40 Votive paper 70 Source: IMF 2007. Table 6.7: Excise Tariff since April 2009 No. Goods or Services Tax Rate (%) I Goods 1 Cigarettes, cigars, and other tobacco preparations 65 2 Liquor a/ Of 20o proof or higher From January 1, 2010, through December 31, 2012 45 From January 1, 2013 50 b/ Of under 20o proof 25 3 Beer From January 1, 2010, through December 31, 2012 45 From January 1, 2013 50 4 Under-24 seat cars a/ Passenger cars of 9 seats or fewer, except those specified at Points 4e, 4f and 4g of this Article Of a cylinder capacity of 2,000 cm3 or less 45 167 TA X R E F O R M I N V I E T N A M Of a cylinder capacity of between over 2,000 cm3 and 50 3,000 cm3 Of a cylinder capacity of over 3,000 cm3 60 b/ Passenger cars of between 10 seats and under 16 30 seats, except those specified at Points 4e, 4f and 4g of this Article c/ Passenger cars of between 16 seats and under 24 15 seats, except those specified at Points 4e, 4f and 4g of this Article d/ Cars for both passenger and cargo transportation, 15 except those specified at Points 4e, 4f, and 4g of this Article e/ Cars running on gasoline in combination with 70% of the tax rate for cars of the electricity or biofuel, with gasoline accounting for not same kind as specified at Points 4a, more than 70% of the used fuel 4b, 4c and 4d of this Article f/ Cars running on biofuel 50% of the tax rate for cars of the same type as specified at Points 4a, 4b, 4c, and 4d of this Article g/ Electrically operated cars Passenger cars of 9 seats or fewer 25 Passenger cars of between 10 seats and under 16 seats 15 Passenger cars of between 16 seats and under 24 seats 10 Cars for both passenger and cargo transportation 10 5 Two- and three-wheeled motorcycles of a cylinder 20 capacity of over 125 cm3 6 Aircraft 30 7 Yachts 30 8 Gasoline of all kinds, naphtha, reformade 10 components and other components for mixing gasoline 9 Air conditioners of 90,000 BTU or less 10 10 Playing cards 40 11 Votive gilt papers and votive objects 70 II Services 1 Dance halls 40 2 Massage parlors and karaoke bars 30 3 Casinos and prize-winning video games 30 4 Betting 30 5 Golf business 20 6 Lottery business 15 Import Tax Administration The revenue collected in any year from taxes on imports is only partly determined by the statutory rate of tax: other factors that are important are policies that affect: 168 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S • The eligibility of consignments for tax exemptions or concessionary treatment; • The process for determining the base to which the rate of tax is applied; and • The terms and conditions for payment of assessed. As indicated above, certain classes of goods are exempted from the VAT and the special consumption tax. There are also a wide range of circumstances under which imports are exempted from import duty. These exemptions not only affect duty collections, but also impact on collections of these other taxes. The nature of these exemptions and concessions are discussed below. Valuation for duty also has an important impact on collections—this has been discussed in some detail in an earlier CIE report (CIE 2004a). Finally, the terms for payment of duty impact on the cash flow to the budget and the costs of financing budget outlays. Vietnam has a rather unusual set of arrangements for paying import taxes. These are also discussed below. Exemptions and Concessions As indicated earlier, Vietnam’s law provide for some standard exemptions from duty, such as: • Temporary imports and goods in transit; • Duty free luggage of passengers; and • Imports by foreign organizations enjoying immunities and privileges consistent with international conventions observed by Vietnam. However, as box 2.5 shows there are a number of additional circumstances where imports may be exempt from duty, and the laws on investment provide for a range of incentives for approved projects. As a result of these provisions, a significant range of capital goods, raw materials and intermediate inputs can be exempted from duty depending on the ultimate user of the goods, the resources used to fund the imports, the location where the goods are to be used, and the sector or nature of the production process in which they will be used. Thus the same goods could be subject to considerably different treatment according to end use and other characteristics unrelated to the characteristics of the goods themselves. Such an extensive and complex set of exemptions erodes the revenue base and adds considerably to the challenge of administration and compliance. It can also significantly alter the protective impact of duties, which to a considerable degree undermines the logic of having a highly non-uniform tariff structure. The fact that some exemptions only apply if the goods in question are not produced locally adds an additional dimension of complexity and scope for discretion in administration, and further skews the protective impact of the duty regime, There is also a set of duty suspension, exemption and rebate systems for enterprises involved in export production. Enterprises in export processing zones are not liable to pay duties on capital goods and raw materials. 169 TA X R E F O R M I N V I E T N A M Terms and Conditions for Payment Importers currently face four terms of payment of import duty (and in turn for VAT and special consumption tax on imports). Amendments made to the Law on Import and Export duties in 2005 and to the Decree 87/2010, to Circular 194/2010 stipulate the following terms. • With respect to goods being supplies and materials imported for production of goods for export, the time limit for payment of duties for taxpayers that with good tax law observance records shall be 275 days (nine months) from the date taxpayers register their customs declarations. In special cases, the time limit for payment of duties may be extended according to the cycle of production or storage of the supplies and materials by the enterprise in accordance with the regulations of the Government. • With respect to goods temporarily exported for re-import or goods temporarily imported for re-export, the time limit for payment of duties for taxpayers with good tax law observance records shall be 15 days after the date of expiry of the duration from temporary export to re-import or from temporary import to re-export in accordance with the regulations of competent body. • With respect to goods being machinery, equipment, materials, fuel, supplies and means of transportation imported for production, the time limit for payment of duties shall be 30 days from the date taxpayers register their customs declarations. • With respect to imported consumer goods, duty must be paid prior to receipt of goods. In cases where the amount of duty payable is guaranteed by credit institutions or other organizations authorized to conduct a number of banking operations in accordance with the provisions of the Law on Credit Institutions, the time limit for payment of duties shall be 30 days from the date taxpayers register their customs declarations. The fact that Customs releases goods to importers before duties are paid creates some particular problems for ensuring integrity of the payment process, problems which are exacerbated by the current state of interaction between Customs and the Treasury, to which duty payments are made. It also means that in effect the Treasury is financing part of the working capital of most enterprises. Box 6.2: Other Issues in the Duty Payment System Under present arrangements, importers receive an assessment of duty and VAT payable from Customs, and in most cases have 30 days in which to make payment. Payment is made (typically by cash, cheque or bank transfer—there is currently no legal basis for e-payment and notification) into a designated account of the Treasury in the Ministry of Finance. This account does not have to be at the port of entry of the goods. When the Treasury is notified of the payment, it sends a paper notification to the customs house at the port of entry. (Although the General Department of Customs now is a part of the Ministry of Finance, its computer systems are not currently integrated into the Treasury/MOF system, so it is not possible to send notifications.) Given this system, it is not uncommon for notification to reach the port of entry some time after the period for payment has expired. In theory, Customs would refrain from processing additional shipments without evidence of payment, but will make accommodation for established traders. 170 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S One problem is that notification of payment only goes to the port of entry. Customs Headquarters does not receive notification of actual payments made, so is not able to implement effective management control over the payment system and effect reconciliation of payments against payment notices. Presumably, this also means that traders who have not paid accounts may be able to bring shipments through other ports of entry: customs’ information systems would appear to be unable to inform all customs houses of delinquency in payments. This system is under revision: under new arrangements Customs will issue the tax notice electronically to the Treasury, which will in turn notify Customs electronically when payment is made. The current payment and invoicing system, and lack of communication between computer systems also create difficulties with respect accounting for trader’s VAT credits on tax paid on imports. Nontariff Barriers The other aspect of Vietnam’s system for managing imports that has long had an impact on revenue collections are the policies and processes for managing trade, including regulation of entry into importing of particular commodities, prohibitions and quantitative restrictions on imports, and regulation of transactions. • Import prohibitions: Most prohibitions appear to have been driven by concerns of health, security and safety. However, imports of cigarettes and a wide range of second hand goods were prohibited until Vietnam’s accession to the WTO. • Import licensing by the Ministry of Industry and Trade, typically used to manage volumes of imports, to protect local producers, but also to ‘manage’ demand and supply of key commodities has been largely dismantled as a result of implementing WTO commitments. • Specialised management by line agencies. A significant range of goods is subject to management by line agencies. This management, which is largely aimed at quality and safety issues can take the form of inspection of shipments, appointment of authorised importers, import licensing, registration, or certification as to compliance with standards. This form of management has in the past been used to protect state enterprises belonging to the managing agency from competition, either from imports or from other traders. The Government introduced a set of tariff rate quotas in 2003 for raw milk, condensed milk, eggs, maize, raw tobacco, salt and cotton. These quotas, which allow a certain volume of imports at a lower tariff than applies to “over-quotaâ€? imports frequently act as a straight quantitative limit on imports. They are permitted, however, under the WTO Agreement on Agriculture, and Vietnam has retained them in the postaccession period. Trade in certain commodities (such as petroleum, aircraft, newspapers, cigarettes and videotapes) remains restricted to state enterprises. As a consequence of WTO accession, foreign firms and individuals are now allowed the same trading rights as domestic enterprises. 171 TA X R E F O R M I N V I E T N A M WTO Accession and Tariff Bindings On 11 January 2007, Vietnam formally acceded to the World Trade Organization (WTO), 12 years after submitting its application for membership. As part of the accession process, Vietnam negotiated three schedules of commitments that are attached to its Protocol of Accession, relating to: • Tariffs (to be reduced and bound or subject to ceiling bindings) and other measures affecting trade in goods; • Market access, domestic support and export subsidies affecting agricultural trade (with bindings); and • Commitments on trade in services, consistent with the General Agreement on Trade in Services. This chapter assesses some aspects of the schedule of tariff bindings. Vietnam’s Tariff Bindings Vietnam has agreed to bind all of its tariff rates, with a 12-year transition period for all bindings to reach their final levels. As table 6.8 shows, the bindings for around 60 percent of the tariff lines were set at their final level upon accession in 2007, and the bulk of the transition will be completed by 2014. As the table shows, the average bound rate declines from 17.5 percent to 13.5 percent over the period. The degree of dispersion amongst the rates, as measured by the standard deviation also falls (from 22.0 to 14.6) although the coefficient of variation does not fall very much. These aggregate measures imply that a tariff regime that set tariffs at the bound rates would probably become less distorting over time. However, there would be increasing complexity in the tariff, as the number of rates in the schedule would increase from 37 to 48. (Of course, there is no obvious reason to suppose that the Government would set rates at the bound levels. However, it is almost inevitable that there will be pressure from some quarters to raise rates to the highest level permitted by the bindings.) Table 6.8: Vietnam’s Tariff Binding Transition Proportion of Tariff Lines Average Maximum Number of Standard CVa Reaching Final Rate Rate Rates Deviation Bound Rate % % % No No % 2007 59.5 17.5 200 37 22.0 114.4 2008 1.3 17.5 200 39 21.5 114.6 2009 3.9 17.3 200 39 20.0 115.8 2010 11.1 16.6 200 44 20.0 118.9 2011 1.1 16.4 200 45 20.0 119.5 2012 16.2 14.8 200 49 19.7 124.6 2013 0.4 14.7 200 49 19.6 124.4 172 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S 2014 6.4 13.6 135 46 18.5 108.4 2015 0.0 13.6 135 46 18.3 108.3 2016 0.0 13.6 135 46 14.7 108.3 2017 0.1 13.5 135 47 14.7 107.9 2018 0.0 13.5 135 47 14.7 107.9 2019 0.1 13.5 135 48 14.6 106.8 Source: Author’s calculations based on WTO 2007. Note: a. Coefficient of variation. b. The maximum rate from 2007 to 2013 is actually a complex rate levied on used motor vehicles, typically set at the minimum of 200 percent or 150 percent plus a specific US dollar amount, which depends on the vehicle engine. Comparison with Tariff Commitments of other recently acceding countries Table 6.9 compares Vietnam’s the tariff binding commitments with those of other countries that have acceded to the WTO since its formation in 1995. While other countries tariff policies do not provide any very meaningful basis for Vietnam’s tariff setting, one thing that does seem to emerge from the comparison is that Vietnam has not taken advantage of the opportunity that WTO accession provides to lock in a lower and less dispersed tariff. The majority of recently acceding countries have bound 90 percent or more of their nonagricultural tariffs at rates of less than 20 percent (compared to 71 percent for Vietnam): and Vietnam has bound a much lower share of its agricultural tariffs at rates below 20 percent (48.5 percent, compared to over 80 percent for most other acceding countries). Vietnam’s transition period is also longer than for most other recently acceding countries (the transition period for China, Chinese Taipei Cambodia and Nepal was ten years, compared to 12 years for Vietnam). However, over 99 percent of Vietnam’s tariff lines reach their final bound rate in 7 years. Table 6.9 compares Vietnam transition with those for other recently acceding countries. Table 6.9: Tariff Bindings for Countries Acceding to the WTO Nonagricultural tariffs Agricultural tariffs Year Lines Lines Lines Lines Tariff of Simple below below Simple below below Country lines Max. Max. acces- averageb 20 30 per- averageb 20 30 per- bounda sion percentc centd percentc centd No. % % % % % % % % Albania 2000 100.0 6.0 20.0 100.0 100.0 10.6 20.0 100.0 100.0 Bulgaria 1996 na 12.6 40.0 89.1 97.5 34.9 98.0 34.9 47.6 Cambodia 2003 100.0 18.4 42.5 74.7 83.9 29.3 60.0 35.8 50.0 China 2001 100.0 8.9 50.0 96.5 99.5 15.0 65.0 84.2 95.7 Chinese 2002 100.0 4.8 90.0 98.8 99.6 17.5 400.0 83.2 94.4 Taipei Croatia 2000 100.0 5.0 25.0 100.0 100.0 10.4 55.0 90.6 95.0 Ecuador 1996 100.0 20.1 40.0 57.6 99.7 25.8 86.0 37.4 87.3 173 TA X R E F O R M I N V I E T N A M Estonia 1999 100.0 6.6 30.0 99.3 100.0 17.7 59.0 68.0 85.5 Georgia 2000 100.0 5.8 20.0 100.0 100.0 12.1 30.0 84.5 96.5 Jordan 2000 100.0 15.0 30.0 77.2 100.0 25.0 200.0 39.8 95.0 Kyrgyz 1998 100.0 6.7 20.0 100.0 11.7 30.0 100.0 Republic 100.0 99.9 Latvia 1999 100.0 9.3 55.0 98.3 98.7 33.6 55.0 38.1 41.7 Lithuania 2001 100.0 8.2 30.0 99.2 100.0 15.6 100.0 80.1 91.4 Moldova 2001 100.0 5.7 20.0 100.0 100.0 12.4 25.0 99.9 100.0 Mongolia 1997 na 20.0 30.0 87.6 100.0 18.4 75.0 88.8 90.8 Nepal 2003 100.0 23.7 60.0 47.3 95.9 42.3 200.0 7.0 35.6 Oman 2000 100.0 11.0 20.0 100.0 100.0 30.5 200.0 83.4 87.6 Panama 1997 na 11.5 81.0 84.3 99.5 26.1 260.0 84.3 99.5 Vietnam 2007 100.0 12.5 100.0 70.9 92.8 21.0 135.0 48.5 68.3 Source: WTO (2001); Vietnam Law Data; author’s calculations. Note: a. Proportion of tariff lines bound individually. b. Of lines bound individually. c. Proportion of individually bound tariff lines bound with rates below 20 percent. d. Proportion of individually bound tariff lines bound with rates below 30 percent. n.a. = Not available. The Impact of Bindings on Applied Tariffs While comparisons of Vietnam’s tariff binding commitments with those of other countries tell us a little bit about how the country’s negotiators approached the accession process, they tell us nothing about the economic implications of the commitments. What is important is how the commitments could shape the evolution of the structure of applied tariff rates, and in turn what that implies for levels of protection and revenues. A starting point for addressing this question is to compare the bindings with the applied rates in place immediately prior to accession. When this analysis was carried out, the most recent year for which an electronic version of the tariff is available was 2005 (provided by the Ministry of Finance). An unofficial version of the tariff schedule for 2006 is available on the website of the Ho Chi Minh City’s Investment and Trade Promotion Centre (ITPC 2007). (More recently, electronic summaries of Vietnam’s tariff schedules have been available on the WTP website – this data was used for tables in section 1 of this chapter.) A comparison of the 2006 information with the 2005 schedule confirms while some changes were made to tariffs between the two years, so it is not unreasonable to use the 2005 schedule as a starting point for assessing Vietnam’s tariff commitments. (The one area where changes have occurred is in the tariffs on petroleum products (items in the four-digit tariff heading 2710): rates here are changed irregularly during the year as the Ministry of Finance tries to influence the domestic price of fuel.) 174 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S Figure 6.1: Evolution of Vietnam’s Tariff Schedule, 1995 to 2005, and Bound Rates to 2019 25 20 Standard devn. Average 18 20 16 14 15 12 Per cent No 10 10 8 6 5 4 2 0 0 1995 1996 1997 1998 2001 2003 2005 2007 2010 2012 2014 2019 Source: CIE 2007. Figure 6.1 summarizes the evolution of basic indicators of Vietnam’s MFN tariff rates since 1995, and compares them with indicators for the bound rates up to 2019. As the chart shows, the bindings seem to continue a process of reducing the average and dispersion of rates that began after 2003. This process followed a period of escalation in the tariff structure that began after the 1997–98 Asian financial crisis, and accompanied a process of phasing out some the key nontariff barriers. Figure 6.2 summarizes the shifts in the distribution of rates that implementation of the bound rates would bring in its train. It shows that the bindings could lead to a large reduction in the number of zero-rated tariff lines, a significant increase in the number of tariff lines with rates between zero and 5 percent, and a consolidation of rates in the range of 5 to 20 percent. Figure 6.2: Distribution of Actual and Bound Tariff Rates, 2005 to 2019 35 30 25 20 15 Per cent 10 5 0 2005 2007 2010 2012 2014 2019 0 >0-5 >5-10 >10-15 >15-20 >20-25 >25-30 >30-35 >35-40 >40-45 >45-50 >50-75 >75-100 >100 Source: CIE 2007. (In practice, as table 2.2 shows, during 2008–09, Vietnam’s applied tariff has been adjusted in ways that have produced a lower average applied rate than the average bound rate, and a lower dispersion of rates as measured by the standard deviation and coefficient of variation.) 175 TA X R E F O R M I N V I E T N A M 4. Analyzing the Revenue Impact of Tariff Bindings An important concern for the government during the course of negotiating tariff commitments for WTO accession has been the impact on revenues. As table 2.1 showed, import and export duties have accounted for between 10 and 18 percent of revenues in the last few years. (VAT on imports, which is levied on the duty paid value of imports and is therefore affected by tariff changes, has accounted for between 8 and 10 percent of revenues, and some of the Special Consumption tax, which accounts for around 6 percent of revenues, is levied on imports). So since the contribution to revenue of taxes on imports remains significant, it is reasonable to try to work out what the effects of tariff changes consequent upon the application of tariff bindings might be. Other aspects of WTO accession may also affect tariff revenues, such as application of the Customs Valuation Agreement and elimination of quantitative restrictions on imports. These issues are discussed in an earlier CIE paper (CIE 2004a): this chapter focuses solely on the direct effects of binding commitments on tariff revenues. Relationship between Tariff Rates and Revenues. Changes in tariff rates have two direct in principle effects on revenues: they alter the statutory rate of duty to be levied on imports, and alter the base for calculation of VAT on imports. In practice, of course, the effects are shaped by the extent to which the statutory rates are actually applied. As figure 4.1 shows, there has for some time been a large difference between the potential rate of duty collection (the import weighted average duty rate) and the realised rate of collection (duties expressed as a percentage of the value of imports). On average, over the period 1994 to 2003, actual collections have been 51 percent of potential revenues. However, as the chart shows, the relationship between potential and realised duties does not seem to be stable, so it is difficult to predict actual collections from statutory duty rates without more knowledge of what is driving the difference. The Vietnamese import taxation system has had an extensive set of exemptions that can affect the duty levied on shipments of goods that may have non-zero statutory rates. This may be a key reason for the outcomes shown in figure 6.3. Unfortunately, when the data in the chart was compiled, there seemed to be no consistent information on the incidence of either of exemptions or of assessed or collected rates of duty by tariff item or groups of tariff items (see CIE 2004a). This means that it is not really possible to explore the relationship between the composition of imports, statutory duty rates and actual collections. This in turn means that it is not possible to construct scenarios of how tariff reductions consequent on WTO accession would affect revenues. All that can be done at this stage is to explore how these reductions would affect notional revenues assuming that statutory rates are applied in full. 176 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S Figure 6.3: Potential and Realized Rates of Import Duty Collection Potential rate Realised rate 1994 1995 1997 1999 2000 2001 2002 2003 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 Rate Source: CIE 2004a. a. Potential rate of duty is the weighted average duty rate; realized rate of duty is the ratio of actual duty collections to the value of imports. It is worth noting that this absence of a clear relationship between statutory tariff rates and actual collections is not unique to Vietnam. Pritchett and Sethi examined this relationship for a number of countries implementing tariff changes in the late 1980s and 1990s, and found that: • The collected rate of duty for any given item in the tariff code is only weakly related to the official rate for that item; • The variation of collected rates around the official rate increases with the level of the official rate; • Collected rates, on average, increase much less rapidly than official rates • When the official rate is high, the rate at which collected rates increase as official rates increase falls. From this evidence they argue that in reforming the system of tariffs and tariff revenue collection, the change in official rates, especially at the high levels, is likely not to be the most important element of reform affecting revenue. They also argue that assuming that tariff revenues fall one-for-one with rates overstates the impact of rate reductions on revenues, because it assumes constant collection performance. The General Department of Customs collects data on the value of imports that enter under different types of processing regime. As table 6.10 shows, over 60 percent of the value of imports in 2002 and 2003 were classed as imports of goods destined for domestic consumption, as opposed to further processing or temporarily admitted. Customs sources have suggested that a large proportion of imports for further processing of cleared by foreign invested enterprises are exempt from duties under the various provisions determining exemptions and concessional treatment. 177 TA X R E F O R M I N V I E T N A M Table 6.10: Value of Imports by Customs Processing Treatment 2002 2003 Customs processing treatment $US mil. % $US mil. % Consumer goods 11908.3 60.8 14508.6 63.7 Imports by foreign invested enterprises 2767.3 14.1 3097.4 13.6 Temporary admission 374.2 1.9 609.5 2.7 Imports for further processing 4533.2 23.1 4551.6 20.0 Total 19582.9 100.0 22767.2 100.0 Source: CIE 2004a. How would Tariff Reductions and Bindings Affect Revenue Collections Implied by Statutory Duty Rates? Analysing the effects that tariff binding commitments may have on revenue involves: • conjecturing how Vietnam’s tariff bindings may affect applied duty rates; • estimating how these changes in duty rates would affect notionally collectible revenues for a given level and structure of imports; • adjusting these estimates for the impact of concessions and exemptions and other factors affecting collection efficiency; and • factoring in how the volume of imports of different commodities might change in response to the price effects of duty changes. The scope for implementing these steps is seriously constrained by data problems. The main constraints are due to problems of aggregation and of tracking of revenue assessments. These problems can be summarised as follows. The aggregation problem Tariff rates are specified at the eight-digit level of the HS tariff nomenclature. However, comprehensive import data is only reported at the six-digit level of the HS (and this only for 2003), so some way is required of creating a representative tariff rate to apply to the import data. Whatever method is used will lead to inaccuracies in the estimates of implied revenues. The problem of revenue assessments The other main problem is—as discussed above—that no data seems to be available on actual duty assessments at any level of the HS. Nor is information readily available on the incidence of concessions and exemptions, or on the proportion of imports from ASEAN countries that is eligible for CEPT treatment. This means that it is not possible to develop any systematic way of assessing the impact of these exemptions and concessions on revenue collections, and how it might vary across different commodities and over time. Projecting Revenue Impacts These issues can be explored further with a simple model to project how complementary policy changes can offset the immediate revenue impacts of tariff reductions. The model 178 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S superimposes estimates of estimated tariff changes associated with the bindings on a scenario of projected growth in imports. The model takes no account of how imports may respond to price changes caused by tariff reductions. But it does allow consideration of the impact of rationalising the current system of exemptions and concessions and so increasing the efficiency of duty collections (as measured by the difference between notional and actual revenues raised on imports.) The model can be extended to take account of the effect on revenues of duties collected on imports under the CEPT — but for now the focus is on imports subject to MFN rates. The Growth Scenario The analysis has been carried out assuming that Vietnam’s GDP will continue to grow at and average 7.5 percent per annum over the period 2007 to 2019. A set of consistent growth rates for imports of consumer goods, petroleum and production inputs has been constructed, by assuming a relationship between GDP and imports. Table 6.11 shows that over the period 1995 to 2006, a 1 percent change in GDP has been associated on average with a 2.7 percent increase imports. For the purpose of this analysis, it has been assumed that the elasticity of imports to GDP will average 1.6 over the projection period, with real imports of consumer goods and petroleum products projected to grow at 9 percent per annum and real imports of production inputs (raw materials, intermediates and capital goods) projected to grow at 13.5 percent per annum. (Because the key data set, the disaggregated information on imports, is for 2003, the analysis generates estimates in 2003 prices.) Table 6.11: Real Imports, GDP, Growth, and Elasticities Real imports Real GDP Import growth GDP growth Elasticity 1995=100 1995=100 % % 1995 100.0 100.0 .. .. .. 1996 130.3 109.3 30.3 9.3 3.2 1997 138.7 118.3 6.4 8.2 0.8 1998 159.4 125.1 15.0 5.8 2.6 1999 189.9 131.0 19.1 4.8 4.0 2000 248.5 139.9 30.9 6.8 4.5 2001 272.5 149.6 9.7 6.9 1.4 2002 344.6 160.2 26.5 7.1 3.7 2003 432.7 171.9 25.6 7.3 3.5 2004 506.1 185.3 16.9 7.8 2.2 2005 545.1 201.0 7.7 8.4 0.9 2006 646.8 217.4 18.6 8.2 2.3 Trend 18.8 7.0 2.7 Source: Author calculations. Growth and Collection Efficiency Figure 6.4 shows the results of three scenarios explored with the model, assuming real GDP growth of 7.5 percent per annum over the scenario period 2007 to 2019, with consistent rates 179 TA X R E F O R M I N V I E T N A M of growth of imports of petroleum products (9 percent), consumer goods (9 percent) and intermediate products (13.5 percent). The scenarios are: • Base scenario: No change in applied MFN tariffs from 2006 rates, and no change in the incidence of concessions and exemptions; • Scenario 2: Application of the tariff bindings, leaving applied rates that are below bindings unchanged; • Scenario 3: Application of the tariff all bindings, including increasing rates where the bound rate is higher than the current applied rate; and • Scenario 4: As for the second scenario, but with exemptions and concessions progressively reduced so that the difference between notional and actual revenue collections narrows progressively over time. (Figure 6.4 shows the projected trajectory of collection efficiency for imports of petroleum products, intermediate and consumer goods. It is assumed that over the period the gap between implied and actual collections on imports of intermediate goods falls to 30 percent—reflecting a judgment that imports entering into the production of exports will remain eligible for duty drawback and exemption, and some concessions and exemptions will remain. For consumer goods, the gap is assumed to fall to 5 percent.) Figure 6.4: Nonpetroleum Import Revenue Scenarios 6000 Base scenario 5000 Scenario 2 Scenario 3 US$ billion 4000 Scenario 4 3000 x 2000 1000 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: Author’s calculations. As the chart shows, under scenario 1, nonpetroleum tariff revenues are projected to grow steadily from US$1.6 billion to US$5.5 billion (in 2003 prices) between 2006 and 2019. Under the second scenario, these revenues increase, but are affected by the tariff reductions, and reach US$3.1 billion in 2019. If rates were set to the bound rate in each year, revenues would grow to US$3.7 billion. If action were taken to reduce the incidence of concessions and exemption according to the trajectory in table 6.12, revenues would dip slightly below those of the first scenario until 2010, and catch up with those revenues in 2018. 180 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S Table 6.12: Trajectory of Collection Efficiency Petroleum Intermediate Consumer Year products (%) goods (%) goods (%) 2006 100 22 75 2007 100 26 78 2008 100 30 81 2009 100 33 84 2010 100 37 87 2011 100 41 90 2012 100 44 93 2013 100 48 95 2014 100 52 95 2015 100 55 95 2016 100 59 95 2017 100 63 95 2018 100 67 95 2019 100 70 96 One critical element that shapes overall tariff revenues is the taxation of imports of petroleum products. As indicated in the previous chapter, tariff rates on these products are frequently adjusted to balance revenue and pricing objectives. Figure 6.5 shows what could happen to petroleum revenues raised on imports from sources subject to MFN rates, and prices remained unchanged from 2003 levels. If 2006 rates continued to apply, real revenues would grow to around US$630 million by 2019. If the upper bound rates were applied, revenues would be some US$1.3 billion higher in 2019. This would come close to completely compensating for any reduction in nonpetroleum revenues under the various scenarios. Figure 6.5: Petroleum Revenue Scenarios 2500 2000 With 2006 rates All bound rates Binding reductions US$ billion 1500 1000 500 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: Author’s calculations. This analysis illustrates a number of things: • The bindings could have quite a large impact on nonpetroleum tariff revenues, all other things being equal. 181 TA X R E F O R M I N V I E T N A M • Policy changes with respect to the extensive and complex concessions and exemptions can readily compensate for the effect of rate reductions, and could allow, from a revenue perspective, more radical reductions in tariff rates. • There is considerable scope to make up for revenue shortfalls by increasing taxation of petroleum products. It is also worth noting that the foregoing analysis takes no account of what tariff reductions would do to the base for other taxes on imports (primarily the value added tax) and for the base for income taxes. Reduction in the price of duty paid price of imports could reduce the value basis for the VAT, but might also lead to higher volumes of imports. Similarly, growth and increased formalisation of the economy would increase the base for other taxes, including income and property taxes. However, this may also require further attention being paid to the structure and administration of these taxes, especially the VAT. (Recent analysis of the VAT suggests that there are still some serious problems with the VAT, in terms of the base, rate structure and calculation methods (CIE 2004a). On balance, it is unlikely that the Government would suffer a revenue shortfall in the medium to longer term, although there may be some issues to deal with in the short term. 5. Industry and Tariff Protection in Vietnam One other important consequence of Vietnam’s tariff biding commitments is the effect they may have on the structure of protection for local producers. Much of the rationale for Vietnam’s complex and dispersed structure of tariff rates appears to have been to provide protection for certain sectors of the economy. This motive may also explain some features of the country’s tariff binding commitments. So it may be useful to examine how these commitments may impact on industry assistance. The starting point for analysis of the impact of tariff binding scenarios is the current structure of applied MFN tariffs, and how this bears on the products of different industries. We can use information on the commodity composition of the output of industries to calculate average tariff rates for each industry. This gives us some idea of the nominal rate of protection that the tariff regime makes available to producers in the industry: that is, the extent to which tariffs allow producers to charge prices that exceed the duty free price of competing imports. The next step is to estimate effective rates of protection for industries, and get an indication of how the combined effect of tariffs on inputs as well as on imports of products that compete with the output of these industries. This gives us an idea of how much the tariff regime may allow returns to—and costs of—land, labor, and capital to be higher (or lower) than they would in the absence of these tariffs. Given these starting points, and the framework for calculating the various measures, we can then see how alternative tariff reduction and binding scenarios may change protection to industries. Tariff Protection by Economic Activity Table 6.13 provides information on the average tariff protection accorded to commodities classified to the broad economic activities (two-digit divisions) of the International Standard Industrial Classification). The average MFN tariff for each activity has been calculated using 182 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S a set of standard concordances linking HS tariff lines to the commodities classified as the products of each ISIC industry. This is not a good indication of the actual level of protection accorded these industries in Vietnam. (To do this would require weighting tariff lines by the value of production of commodities that compete with imports to which the tariff would apply. Unfortunately, there is not as yet data on production of commodities based on an international commodity classification that can be brought into concordance with the tariff nomenclature system.) However, it is reasonable to surmise that big variations in the calculated tariff averages for these economic activities signifies variation in the protection available to import-competing production under these activities. Table 6.13: Vietnam’s Tariffs and Tariff Bindings by Economic Activity Average Average Change Average 2007 2019 Required ISIC Industrya 2005 Bound Bound by 2019 Tariff Tariff Tariff Binding % % % % 1 Agriculture, hunting and related service 12.7 15.5 13.0 0.3 activities 2 Forestry, logging and related service activities 2.0 1.9 1.6 -0.4 5 Fishing, aquaculture and service activities 24.6 24.6 14.9 -9.7 incidental to fishing 10 Mining of coal and lignite; extraction of peat 4.4 4.8 4.8 0.3 11 Extraction of crude petroleum and natural 7.0 10.0 9.2 2.2 gas; service activities incidental to oil and gas extraction, excluding surveying 12 Mining of uranium and thorium ores 0.0 0.0 0.0 0.0 13 Mining of metal ores 0.0 0.0 0.0 0.0 14 Other mining and quarrying. 5.0 8.1 8.1 3.0 15 Manufacture of food products and beverages 31.6 30.8 24.3 -7.3 16 Manufacture of tobacco products 50.6 114.7 108.2 57.6 17 Manufacture of textiles 33.2 10.9 10.8 -22.4 18 Manufacture of wearing apparel; dressing 47.6 20.8 19.9 -27.7 and dyeing of fur 19 Tanning and dressing of leather; manufacture 26.5 24.4 17.8 -8.7 of luggage, handbags, saddlery, harness and footwear 20 Manufacture of wood and of products 11.4 10.9 7.7 -3.6 of wood and cork, except furniture; manufacture of articles of straw and plaiting materials 21 Manufacture of paper and paper products 20.7 19.2 14.0 -6.7 22 Publishing, printing and reproduction of 19.6 18.1 11.5 -8.1 recorded media 23 Manufacture of coke, refined petroleum 10.6 32.1 32.1 21.6 products and nuclear fuel 183 TA X R E F O R M I N V I E T N A M 24 Manufacture of chemicals and chemical products 5.0 9.3 6.0 1.0 25 Manufacture of rubber and plastics products 17.3 18.8 12.9 -4.4 26 Manufacture of other nonmetallic mineral products 21.8 23.1 19.7 -2.1 27 Manufacture of basic metals 5.3 12.7 8.9 3.6 28 Manufacture of fabricated metal products, except machinery and equipment 17.0 18.3 15.7 -1.3 29 Manufacture of machinery and equipment n.e.c. 7.1 9.1 7.3 0.2 30 Manufacture of office, accounting and computing machinery 5.5 7.4 1.4 -4.1 31 Manufacture of electrical machinery and apparatus n.e.c. 12.6 13.6 10.6 -1.9 32 Manufacture of radio, television and communication equipment and apparatus 14.0 14.1 8.5 -5.5 33 Manufacture of medical, precision and optical instruments, watches and clocks 7.3 6.5 4.6 -2.7 34 Manufacture of motor vehicles, trailers and semitrailers 48.2 51.9 41.2 -7.0 35 Manufacture of other transport equipment 35.3 36.4 27.7 -7.6 36 Manufacture of furniture; manufacturing n.e.c. 26.5 23.3 18.5 -8.0 Total 18.3 17.9 14.0 -4.3 Note: a. Divisions of the International Standard Industial Classificiation. The table suggests a number of observations about the current tariff structure and possible impact of Vietnam’s binding commitments. • There is considerable variation of average tariffs across economic activities. Import competing production in some activities (such as manufacture of tobacco products [ISIC 16], transport equipment [ISIC 34 and 35]) would seem to be much more heavily protected than others (such as agriculture (ISIC 1) and manufacture of chemicals, basic metals office and other machinery and precision instruments (ISIC 24,27,30,33). • Average tariffs seem surprisingly high for some industries where Vietnam is successfully exporting a considerable volume of production (fishing (ISIC3) and ISIC 18, manufacture of apparel). Activities that are efficient enough to compete on international markets would normally be competitive with imports on the domestic market without any protection • Some of the larger reductions in tariffs that application of tariff bindings implies are for activities where Vietnam is a successful exporter. This suggests that, perhaps not surprisingly given the way that tariff bindings are negotiated, some of the reductions will remove ‘water’ in the tariff rather than reduce levels of protection actually used by some industries. • It also appears that the bindings provide room for Vietnam to significantly increase tariffs for some activities where large investment is under way — such as petroleum refining 184 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S (ISIC 23), or to allow big increases in protection of local manufacturing — such as tobacco products (ISIC16). Tariffs on Products of Manufacturing Table 6.13 does not give a good sense of how much economic activity is sheltered by tariffs. The 2002 enterprise survey conducted by the General Statistics Office (GSO) provides information on the level of activity at the four-digit level of the ISIC for the manufacturing sector. Calculating average tariffs for each of these 124 four-digit industries allows us to examine the quantum of manufacturing output, employment and value added protected by different tariff levels. Figure 6.6 shows the shares of manufacturing employment and gross output in industries with average tariffs falling into various ranges. It shows that while the final bound rates will lead to the bulk of the sector’s employment operating with lower average tariffs, a significant share of the sector’s output will continue to be protected by high tariffs of more than 50 percent. Figure 6.6: Distribution of Manufacturing Output and Employment by Average Actual and Bound MFN Tariffs, 2005 and 2019 Employment Gross output 35.0 30.0 30.0 25.0 25.0 2005 Value 20.0 2005 20.0 2019 Value 2019 15.0 15.0 10.0 10.0 5.0 5.0 0.0 0.0 0 >0-5 >5-10 >10-15 >15-20 >20-25 >25-30 >30-35 >35-40 >40-45 >45-50 >50-75 >75 0 >0-5 >5-10 >10-15 >15-20 >20-25 >25-30>30-35>35-40 >40-45 >45-50 >50-75 >75 Source: Data provided by GSO and author calculations. Note: The figure was prepared by calculating simple average actual or bound tariff rates for four-digit industries, and ranking the shares of gross output and employment in these industries by these tariff levels. While these data suggest that a high proportion of manufacturing employment may be dependent on high levels of protection, some important caveats have to be made. First, as table 6.14 shows industries falling in the two-digit activity covering manufacture of textiles (ISIC 17) and manufacture of wearing apparel; dressing and dyeing of fur (ISIC 18)) together account for over 23 percent of manufacturing employment. While tariffs on imports of the products of these industries are high, much of their output is exported with no or limited assistance. This means that much of the employment may not benefit from the high tariffs. Second, many industries involved the manufacture of food products and beverages (ISIC 15, which accounts for 16 percent of manufacturing employment and has an average MFN tariff of 35.7 percent—see appendix table A.1) may not face much competition from imported products. This is because local production is very competitive, or involves goods that do not readily enter into international trade. There may well be ‘water in the tariff â€? for parts of this industry and producers may not ‘need’ the full extent of the available tariff protection. (This is not true of all production in this sector: sugar production, for example is not competitive and may well be relying heavily on the average tariff of around 70 percent that applies to this industry.) 185 TA X R E F O R M I N V I E T N A M The structure of manufacturing activities varies across the regions of the country, and so the dependence of employment and production on tariff protection also varies. Table 5.2 shows the average tariff protection available to the manufacturing given the structure of industry in each region. It shows that the manufacturing sector in the Red River Delta, Mekong River Delta and South Central Coast regions has a larger proportion of activities with high tariff protection than in the North East and Central Highland regions. (This could suggest that industry in the latter regions might be less negatively affected by tariff reductions than the counterparts in the more developed parts of the country.) The table also shows that the average tariff on manufacturing weighted by gross output in four-digit industries is quite a lot higher (29.5 percent) than the simple average of rates on manufacturing tariff lines. Effective Rates of Protection (ERPs) To get a fuller understanding of how activities will be affected by changes in tariff protection associated with implementation of the tariff bindings, it is necessary to take account of how tariffs on inputs affect costs of production as well as how tariffs on imports that compete with outputs allow producers to charge higher prices. To this end, the effective rate of protection concept is useful in showing how returns to factors of production can be affected by the combined effect of these two influences on costs and prices. Calculating ERPs requires information on the raw materials and intermediate products used in production processes. Such information is compiled in Vietnam in the Input-Output table prepared by GSO. This table uses a different basis for organising data on products and inputs than the enterprise surveys discussed earlier in this chapter, working directly with information on commodities rather than enterprises. Figure 6.7 shows how the ERPs for import competing activities in broad groups of agricultural, mining and manufacturing activities would change upon implementation of the tariff immediate and final bindings. Protection levels fall for all groups except for precision equipment, where protection becomes slightly less negative. Table 6.14: Regional Distribution of Manufacturing Activities and Average MFN and Bound Tariffsa Gross Average Average Change Average Number of Value Value Employ- 2007 2019 Required 2005 Region Enterprises of Added ment Bound Bound by 2019 Tariff Output Tariff Tariff Binding (%) (%) (%) (%) (%) (%) (%) (%) Red River 26.5 21.6 21.2 22.8 34.9 34.7 26.7 -8.2 Delta North-East 4.5 3.7 3.6 5.3 19.3 20.0 15.8 -3.5 North-West 0.5 0.1 0.1 0.4 26.8 29.0 24.0 -2.8 North 3.9 2.7 3.2 2.9 26.1 32.4 26.8 0.7 Central Coast 186 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S South 5.7 4.5 5.3 6.9 31.3 39.5 32.1 0.8 Central Coast Central 1.9 0.5 0.6 1.5 19.7 19.1 15.7 -4.0 Highland South East 38.9 56.9 58.4 52.6 28.2 27.6 21.7 -6.6 Mekong 18.3 9.9 7.7 7.6 30.1 31.0 23.6 -6.4 River Delta All regions 100.0 100.0 100.0 100.0 29.5 29.8 23.3 -6.3 Source: Data supplied by GSO, author’s calculations. Note: The table shows, for example, that the Red River Delta accounts for 26.5 percent of all manufacturing sector enterprises in the country, and that the 2005 weighted average tariff applied to the products of industries in the region is 34.9 percent. a. Average MFN or bound tariff for each region calculated by weighting the average tariff for each four-digit industry by the share of industrial gross output in the region accounted for by that industry. Figure 6.7: Changes in Effective Rates of Protection for Import Competing Production Associated with Implementation of Final Bound MFN Tariff Rates Agriculture Mining & quarrying Food beverages & tobacco Building materials Paper and wood products Chemicals Precision equipment 2005 Household appliances Transport equipment Non-electrical machinery 2007 Electrical machinery Metal products 2019 Textiles, clothing & footwear Other -25 0 25 50 75 100 125 150 175 200 225 Per cent Source: Author calculations. Note: The chart indicates the average effective rates of protection for import competing production in broad industry classes (aggregates of industries identified in the Vietnamese Input-Output industry classification) using the 2005 applied MFN tariff rates and the 2007 and 2019 bound rates. Figure 6.8 provides an attempt to take account of the fact that there is a fair degree of export-oriented production in some of these industry groups, for which the tariff provides no protection. The figure suggests that application of the bindings does little to change the impost that tariffs place on export-oriented activities. 187 TA X R E F O R M I N V I E T N A M Figure 6.8: Changes in Average Effective Rates of Protection Associated with Implementation of Final Bound MFN Tariff Rates Agriculture Mining & quarrying Food beverages & tobacco Building materials Paper and wood products Chemicals Precision equipment 2005 Household appliances Transport equipment Non-electrical machinery 2007 Electrical machinery Metal products 2019 Textiles, clothing & footwear Other -25 0 25 50 75 100 125 150 175 200 225 Per cent Note: The chart indicates the average effective rates of protection for exporting and import competing production in broad industry classes (aggregates of industries identified in the Vietnamese Input-Output industry classification) using the 2005 applied MFN tariff rates and the 2007 and 2019 bound rates. As table 6.16 shows, implementation of the bound rates would reduce the average ERP across all traded goods producing activities identified in the Input-Output table—from 47.1 to 34.5 percent for import substituting production, and from 32.9 to 24.2 percent across importing competing and exporting activities. Table 6.15: Summary Measures of Changes in ERPs Import Substitution Average ERP 2005 2007 2019 2005 2007 2019 Mean 47.1 43.9 34.5 32.9 31.4 24.4 Standard Deviation 72.0 68.4 55.7 67.6 65.4 51.7 CV 152.8 155.9 161.5 205.2 207.9 211.6 Source: Author calculations. The table suggests that the dispersion in effective rates would also fall. But as chart 6.9 shows, the distribution of value added in these industries across different levels of effective protection suggests that the occurrence of pockets of quite high and very low protection will persist. 188 I N T E G R AT I O N A N D G O V E R N M E N T R E V E N U E S Figure 6.9: Distribution of Industrial, Mining, and Agricultural Value Added in I/O Industries, by ERPa, 2005 and 2019 50 120 Distribution Cumulative distribution 45 40 100 35 80 30 Percent 25 2005 60 Percent 2005 20 15 2019 40 2019 10 20 5 0 0 0 -1 0 0 5 10 15 20 25 30 35 40 45 50 60 70 80 0 >1 0 50 0 -1 0 0 5 10 15 20 25 30 35 40 45 50 60 70 80 0 >1 0 50 10 15 <-2 10 15 <-2 Per cent Per cent Source: Date: CIE calcuations. Note: a. ERP for import competing activities only. 189 Value-Added Tax By Gangadhar Prasad Shukla, Tuan Minh Le, and Duc Minh Pham VII 1. Introduction Value Added Tax (VAT) has been an important source of revenue in Vietnam. It presently contributes about 30 percent of the total tax revenues. Except for a few years in the past, it has been generally a buoyant tax, the revenues growing at a rate greater than the growth rate of the GDP. Prior to the introduction of a VAT, Vietnam had adopted a turnover tax in October 1990 in its first round of tax reform. In an effort to mitigate the shortcomings of this tax, several rounds of amendments have been carried out since then. Finally, the turnover tax was replaced by a VAT in 1999. Subsequently, several amendments have been made with some major changes adopted in 2003, 2006, and the last VAT Law revision conducted in 2008. Significance of VAT for Vietnam The significance of VAT as a source of revenue can be judged from the fact that today more than 135 countries in the world have a VAT, which raises about 27 percent of total tax revenues or 5 percent of GDP in these countries. As is clear from the table below (table 7.1), VAT is a major source of revenues wherever it has been employed. Although there are few systematic studies about the revenue impact of introducing a VAT, there is some evidence to show that the presence of VAT is associated with higher tax revenues and increased general government revenues (Nellor 1987; Ebrill et al. 2001). VAT as an instrument of raising revenues has gained additional importance in view of the decline in trade tax revenues due to regional trade associations and the adoption of trade liberalization norms of the World Trade Organization (WTO). In the era of globalization where tariff barriers are being dismantled and the role of import tariffs as a major source of revenues has greatly diminished, VAT has become one of the main workhorses for raising revenues. Vietnam is in a similar situation after its accession to the ASEAN and the WTO as part of its strategy to integrate with the world economy. The country has to look to domestic consumption taxes as a means of making up for the loss of revenues from rationalization and phasing out of trade taxes. In addition, VAT offers a neat way of taxing international trade transactions without the negative aspects of import tariffs. As VAT has already become a major source of revenues for Vietnam, its design and pace of reform become critically important for the success of the overall tax reforms. 191 TA X R E F O R M I N V I E T N A M Table 7.1: VAT Rates and Revenues by Regions Central North Sub- Asia & Europe Africa & Small Saharan EU America Average Pacific & Middle Islands Africa Russia East Average 16.4 10.9 19.8 18.6 16.3 13.9 16.1 16.0 Standard Rate (%) VAT revenue 3.9 3.3 7.0 6.4 5.7 4.9 4.7 5.1 as % GDP VAT revenue 28.4 21.7 20.7 27.8 28.1 33.0 18.0 26.5 as % of tax revenues Source: IMF estimates (Ebrill et al. 2001); International Bureau of Fiscal Documentation (IBFD) 2005; Worldwide Summaries (PricewaterhouseCoopers) and Consumption Tax Trends (OECD 2006). Overall Assessment of VAT Design and Revenue Performance from International Perspective Tables 7.2 and 7.3 present a comparative picture of VAT regimes in selected OECD and some Asean countries. The picture emerging from table 7.2 is very diverse. The standard VAT rates in some countries like Denmark, Poland, Hungary, and the U.K. are as high as 25, 22, 20 and 17.5 percent respectively. Other countries like Australia and New Zealand have moderate rates of 10 and 12.5 percent respectively. On the other hand, countries like Canada and Japan have comparatively low tax rates at 6 and 5 percent respectively. It may be noted, however, that the rate in Canada is low because the country has a dual VAT and provinces apply their own General Sales Tax and VAT. Similarly the threshold varies from a low of US$8,184 in Denmark to a high of US$86,300 in Japan. The contribution of VAT revenue also varies considerably from as low as 2.6 percent of GDP in Japan to as high as 9.7 percent in Denmark. Table 7.2: VAT in Selected OECD Countries VAT Rates Threshold VAT Revenue (percent) (US$) Date VAT Rates Standard Other General (percent (percent First Rate Rates Thresholds of tax of GDP) Introduced Basic revenue) Australia 2000 10 37,159 ... ... Canada 1991 7 6 25,000 6.9 2.6 Czech 1993 5,19 19 5 42,826 18.7 7.0 Republic Denmark 1967 25 25 8,184 19.2 9.7 Germany 1968 7,16 16 7 21,342 18.3 6.9 Hungary 1988 5,14,15, 20 20 5,14,15 19,000 22.0 8.1 Poland 1993 3,7,22 22 3,7 12,000 22.1 3.0 192 VA L U E -A D D E D TA X Japan 1989 5 5 86,300 9.4 2.6 New 1986 12.5 12.5 25,955 25.9 4.2 Zealand Slovakia 1993 19 19 48,857 21.5 7.2 Spain 1986 4,7,16 16 4,7 341,500 17.6 6.2 Switzerland 1995 2.4,3.6,7.6 7.6 2.4, 3.6 37,800 11.0 3.9 Unnited 1973 5,17.5 17.5 5 106,700 18.7 6.8 Kingdom Source: Ebrill et al. 2001; International Bureau of Fiscal Documentation (IBFD) 2005; Corporate Taxes 2004–05; World Summaries (PricewaterhouseCoopers) 2006; and Consumption Tax trends (OECD 2006). It is clear from table 7.3 that the standard tax rates in Asian countries are more moderate at around 10 percent. While the Philippines has a low threshold of US$10,600, Singapore has a comparatively high threshold of US$620,100. The contribution of VAT to revenues is around 3 to 4 percent of GDP in most of these countries. While the VAT system and its performance in Vietnam may look very different from most of the OECD countries, it compares well with the Asian countries. Table 7.3: VAT in Asian Countries Threshold VAT Rates (percent) VAT Revenue (US$) Date General (percent Standard (percent VAT First Rates Other Rates Thresholds of tax Rate of GDP) Introduced Basic revenue) Cambodia 1999 10 10 61,000 (services) 35.5 2.9 125,000 (goods) China 1994 13,17 17 13 224,000 27.9 3.9 (wholesale) 3,000 & 1,200 (goods & services) Indonesia 1985 10,15,20,25,35 10 15,20,25,35 109,200 19.3 2.7 Philippines 1988 12 12 10,600 22.1 3.0 Singapore 1994 5 5 620,100 8.4 1.2 Taiwan 1986 5 5 ... ... Province of China Vietnam 1999 5,10 10 5 No threshold 24.5 4.0 Thailand 1992 7 7 30,900 20.8 2.8 Mongolia 1998 15 15 8,300 29.9 5.3 Korea 1977 10 10 12,500 19.2 4.6 Source: IMF World Economic Outlook; IBFD; Ernst and Young; VAT and Sales Tax Worldwide various years. In order to assess the VAT efficiency, two efficiency ratios are generally used: production efficiency ratio and consumption efficiency or C-efficiency ratio. The production efficiency 193 TA X R E F O R M I N V I E T N A M ratio of VAT is measured by the ratio of VAT revenues to GDP divided by the standard VAT rate expressed as a percentage. A low ratio indicates erosion of the tax base due to poor tax policy such as zero-rating and reduced rates or poor enforcement of the law. Consumption efficiency or C-efficiency ratio is a similar measure where the share of VAT revenue as a percentage of consumption divided by the standard rate is estimated. If a uniform consumption tax is imposed on all consumption, its efficiency will be 100 percent. Zero-rating of some items will reduce the ratio below 100 and taxation of investment goods will raise it to more than 100 percent. Alternatively, a high C-efficiency could be an indicator of multiple breaks in the VAT chain due to excessive exemptions resulting in taxation of both the final consumption and some of the constituent intermediate goods. Both these ratios show the same thing but one is normalized with reference to an income type VAT while the other to a consumption type VAT. Clearly the three factors that would determine the revenue productivity of VAT in a country are: (a) The rules that prescribe the rates, the base, the threshold and other issues related to tax structure; (b) The scale of taxable consumption in the economy that is, the final expenditure made by the consumers on taxable items; and (c) The effectiveness of the tax administration and the level of compliance. Since VAT on tradable goods are collected by the customs department, the third factor related to tax administration would also depend to some extent on the share of trade in the economy that is, the trade/GDP ratio, because it is easy to collect VAT on the imports. Table 7.4 shows an estimate of productivity and consumption efficiencies across the geographical regions. The estimates (for the year 2001) indicate that both production efficiency and consumption efficiency ratios for Asia are comparable to the ones in the EU, Central Europe, North Africa and the Middle East, and Latin America and they are significantly higher than in Sub-Saharan Africa. Table 7.4: Production and Consumption Efficiency Ratios for VAT by Region Sub- Asia & EU Central North America Small Saharan Pacific Europe & Africa & Islands Africa Russia Middle East Production 27 35 37 38 36 37 48 Efficiency ratio (%) Consumption 38 58 57 64 62 57 83 Efficiency ratio (%) Source: IMF estimates (Ebrill et al. 2001). A comparative picture of production and consumption efficiencies for Vietnam and some of its neighboring countries is presented in table 7.5. Both the productivity efficiency and consumption efficiency in Vietnam compare well with its neighboring countries. In fact, these are significantly higher than the regional averages presented in table 7.4. In addition to many other factors, this could also be the result of a series of cascading-induced exemptions in the VAT law. 194 VA L U E -A D D E D TA X Table 7.5: Productivity and Consumption Efficiency in Vietnam and Neighboring Countries VAT Collection VAT VAT Household VAT General (% of Productivity Consumption Consumption Collection VAT Rate Household Rate Efficiency (% of GDP) (% of GDP) (%)   Consumption) ( C) / (E) ( D) / (E) (A) (B) (C) (D) (E) (F) (G) Vietnam 65 5.6 8.615 10 0.56 0.86 Thailand 57 3.5 6.140 10 0.35 0.61 Philippines 72 4.5 6.250 10 0.45 0.63 Malaysia a 43 2.2 5.116 10 0.22 0.51 Indonesia 65 3.9 6.000 10 0.39 0.60 Sources: World Bank (WDI 2006); and author’s estimates 2007. Note: a. Sales tax To ensure that Vietnam has a robust and clean VAT law and yields sufficiently high productivity and consumption efficiencies, the logical sequence to proceed would be to first fix the tax structure and then focus on the tax administration and tax compliance issues. Eliminating or at least minimizing the exemptions would be necessary to give a realistic picture of the performance of the VAT. Once these issues are tackled successfully, the expected revenues can be ascertained by forecasting VAT revenue with the help of forecasting models in order to see the level of compliance in the economy and then try to improve it. Some typical features of the VAT in Vietnam such as tax structure, exemptions, computation method for tax liability, refund and equity are examined below, bearing in mind international practice of VAT regimes. Overall Structure of VAT The VAT law introduced in 1999 was consumption-based, following the destination principle, and applied to both goods and services. Exports were zero-rated while 26 categories of goods and services were exempt. A special feature of this tax was the exemption of goods and services subject to Special Consumption Tax (SCT) or excises. A rate of 5 percent applied to 13 categories of items, a rate of 20 percent applied to 5 groups of goods and services including gold, silver, gems, hotel tourism and catering services, lotteries of all kinds, shipping agents and brokerage services, while the rate of 10 percent applied to all remaining goods and services. A number of articles of the original VAT have been subsequently amended. The main characteristics of the current VAT system (applied since January 2009) are discussed below. The present VAT law mandates three tax rates. A tax rate of zero (0) percent is for the export of goods and services, and international transportation A tax rate of five (5) percent is applied to a select group of about 15 categories of goods and services. This group includes clean water; fertilizers and pesticides; medical equipment and medicines; teaching and learning aids; children’s toys and books; unprocessed products of cultivation, husbandry, , forestry and fresh foodstuffs; sugar and by-products; products made from jute and bamboo; semiprocessed cotton; feed for cattle and poultry; technical and scientific services; special 195 TA X R E F O R M I N V I E T N A M purpose machinery and equipment for agricultural production and newsprint; cultural exhibits and sports activities. Finally, a tax rate of ten (10) percent is applied to the remaining items under VAT. A major change compared to the earlier VAT law has been that VAT now applies to all goods and services subject to the Special Consumption Tax (SCT). It is an important change and ensures that Vietnam law conforms to international practice. Twenty five categories of goods and services are VAT exempt under the new law. Rate Structure As pointed out above, Vietnam uses two VAT rates in addition to a zero rate. Currently, of all the countries that have adopted VAT, about 54 percent of countries have a single rate in addition to a zero rate, while about 23 percent of countries have two rates, and a small number have more than two rates.67 The issue of multiple VAT rates may be analyzed on the dual principles of efficiency and equity. Efficiency gains could be realized by following the inverse elasticity rule according to which goods/services with low elasticity of demand should be taxed at a higher rate and goods/services with high demand elasticity should be taxed at a lower rate. This would justify higher VAT rates for the so-called “sinâ€? goods like alcohol and tobacco products in countries applying a differentiated rate structure.68 It could also be argued that some other goods such as gasoline and some luxury items have low elasticity of demand so far as the rich are concerned. A higher level of taxation for all such goods, however, can be easily ensured by imposing excises on those items in addition to the normal VAT. Another aspect of the efficiency point could be made on the grounds that higher rates would result in higher tax revenues. There is, however, no strong evidence to support this hypothesis. On the other hand, it is a well known fact among tax administrators that higher tax rates are more likely to encourage tax evasion and lower tax revenues. The rationale for multiple rates is rooted primarily in an equity concern: the poor should not be taxed heavily and therefore some goods and services that are predominantly consumed by them should be lightly taxed. The list of goods and services with a 5 percent VAT rate in Vietnam mostly includes items related to agriculture and animal husbandry, education, medicine, science and technology. Studies of tax systems where rate differentiation has been implemented, however, tend to show that in the majority of such cases, the tax benefits largely go to the rich who are likely to spend larger sums on those items compared to the poor. This is also true of food items that are often taxed lightly on the grounds that the poor spend a greater share of their earnings on 67 The data about tax rates, exemption levels, etc. are drawn from a 2001 study of VAT practices around the world conducted by the IMF. 68 In the case of Vietnam, such “sinâ€? goods as alcohol and tobacco are now subject to a standard VAT of 10 percent and an excise (or special consumption tax). This conforms to international best practice in VAT design. 196 VA L U E -A D D E D TA X food in comparison to the rich. Even if it is accepted that distributional gains may be realized by rate differentiation in VAT, such gains are likely to be quite modest.69 Also, there are very few items that can be differentiated clearly on the basis of consumption by different income groups. Therefore, for equity or distributional considerations, income taxes and expenditure policies are preferred policy instruments compared to the use of different rates of indirect taxation on different commodities. Even though there may be some marginal benefits of rate differentiation in VAT, several costs imposed by this kind of policy are well known and would generally outweigh such benefits, if any. Some of these costs are as follows (Ebrill et al. 2001): a. A single positive rate lowers the cost of compliance by simplifying the requirements of keeping records and invoices. It also simplifies tax forms which goes a long way in promoting self assessment. b. Rate differentiation creates opportunities for misclassification of items, thereby raising the cost of administration. On the other hand, a single positive rate removes the confusion that might arise in the treatment of border line cases and thus lowers the cost of administration. c. The use of a single rate helps limit the numbers of refunds. If there is a low tax rate along with a zero rate, some taxpayers particularly exporters and importers of capital goods could be in a sustained credit position vis-à-vis the tax department. This would result in additional refund claims.70 In the case of Vietnam, a lower VAT rate on several input items is sometimes justified on the grounds that it will reduce the cash flow burden of the VAT taxpayers. The VAT payers, however, typically enjoy a cash flow benefit from collecting higher VAT on their outputs so this cannot be a serious problem. Thus, there is hardly any justification for maintaining multiple rates on a long term basis. Vietnam should therefore consider moving to a rate structure with zero and one positive VAT rate. This implies that the 5 percent rate should be eliminated and only the 10 percent rate be applied to all goods and services subject to VAT. The introduction of a single VAT rate should be part of an overall tax policy reform rather than implemented in isolation. For example, it may have to be synchronized with the reform in excises where some luxury items are brought in the tax net so that the elimination of the lower VAT rate is not seen as a step unfavorable to the poor. Future Outlook Out of the 115 countries for which data is presently available, the VAT rate is less than 10 percent in 25 countries, more than 10 percent and below 15 percent in 24 countries, more than 69 For evidence to support this argument, see Sah (1983, 89–101). 70 Sijbren Cnossen (1994) estimates that for a simple VAT regime in New Zealand (with a single positive rate), the administration cost of the VAT is about US$50 per registrant per year, but the cost quadruples for the case of the U.K., where the VAT is structured with two positive rates and multiple zero rating (cited in Ebrill et al. 2001). 197 TA X R E F O R M I N V I E T N A M 15 percent but below 20 percent in 46 countries and more than 20 percent in 15 countries. As shown in table 7.5 above, many of the low rate countries are in Asia – Singapore, Indonesia and Thailand for instance. On the other hand, the Philippines, China and Mongolia have rates varying between 12 and 17 percent. Also, with the gradual reduction of trade taxes under the provisions of WTO (for example, tax revenues from crude oil tend to be shrinking), the question of recouping the loss of revenue may arise for Vietnam in the future. The revenue from the corporate income tax is currently at the level of just under 6 percent of GDP and it will be unrealistic to expect a phenomenal increase in this. Personal income tax cannot be expected to yield substantial amount of revenues, at least in the short term, because of the cautious approach adopted by the Government toward this tax, which is justified to some extent given the present stage of the growth of the economy and the labor market. The scope of additional revenues from property and land related taxes is limited not only in Vietnam but generally in most countries. Thus VAT has to play a dominant role in any increase of revenues in the future. At some later stage in the future, the Government may therefore consider enhancing the VAT rate in parallel with improvement of compliance management. 2. Exemptions The list of exempt goods and services includes 25 categories of items. Compared to the international best practices, this is obviously a long list of exemptions and needs to be examined carefully. But first, a brief note on some aspects of exemptions is presented below. Why give exemptions under a VAT? First of all, what is the rationale for exempting any goods and services? One possible reason could be if output were hard to identify and measure or if it were administratively hard to tax; for instance, financial services and small traders. In the case of financial services, it is not easy to measure their added value. In the second case, the administrative cost of taxing traders below a certain threshold is comparatively higher than the revenue gains from taxing them. Therefore, there is some justification for exempting these sectors. Another reason for exemption could be that it is a second best option, which proves economically and administratively superior to zero-rating or reduced rates for the goods and services that require tax concession. Revenue Implications of Exemption Second, what would be the implication of exemptions for tax revenues? Effectively, exemptions break the VAT chain and may result in an increase of VAT revenues if there is a cascading of taxes. If, however, exemptions are applied at the last stage, revenues will fall. So the overall impact of VAT exemption is unclear and there is no clear cut evidence that it is a revenue enhancing measure. In any case, cascading is not a desirable feature and one reason for moving away from turnover taxes has been that it causes cascading. 198 VA L U E -A D D E D TA X Anomalies created by exemption There are several anomalies created by exemption. i. Exemption distorts the input choices because of an element of tax that remains in the chain of production. This aberration does not remain confined to a single sector but would pervade all those sectors where the products made from the exempt input are further used as production inputs. ii. By introducing an element of cascading, exemption will create an incentive for vertical integration of units giving rise to artificial groupings of activities and organizational forms that are not necessary from a commercial or economic viewpoint. iii. It would create a major problem if the final product is exported because zero rating will not be able to eliminate the cascading created by exemption at an intermediate stage. Similarly in the particular case of Vietnam, it creates an incentive to import inputs, equipment, or materials that domestically have an exemption in their production chain because the import price does not have an element of tax in it since exports from other country would be zero-rated. iv. Exemption also creates administrative and compliance problems for traders who sell both taxed and exempt items. Their input tax payments must be allocated between the two kinds of sales. Some Goods and Services commonly exempted under VAT Looking at the international practice of exemptions, the following broad categories of goods and services are generally included in the exempt category. i. Agricultural products and key agricultural inputs ii. Passenger transport iii. Cultural and other merit goods iv. Aid financed activities v. Services provided by the public sector, health and education in particular vi. Financial services vii. Real estate and construction The rationale behind exempting the first two categories of goods and services is some sort of distributional or equity consideration in the sense that by exempting these sectors, low income consumers and agricultural farmers will benefit. This concern is, however, partially addressed since small farmers and transporters will normally fall in the category of small exempt traders below the prescribed threshold limit. The third category reflects the sensitivity of taxing certain goods and services in a particular country, while the fourth group might arise because of conditions imposed by the donors. The extent to which exemption may have to be granted to these two categories of goods and services would be country specific and somewhat item specific. Goods and services supplied by the public sector in competition with the private sector should be fully taxed except where these are purely noncommercial services given free of 199 TA X R E F O R M I N V I E T N A M charge. Defense, other not-for-profit public services, some categories of health and education services fall in this category. It is hard to tax a service that is given away free except by taxation of inputs. In fact, the public sector may be regarded as the final consumer of noncommercial services that it produces and gives away free of charge. This is the practice currently adopted in the EU countries as well. As regards health and education, the standard advice and practice is to exempt basic services— primary education, basic healthcare—and tax specialized services at the normal tax rate. One relevant question is: does it make a difference whether such public sector services that are provided free of cost are taxed or exempt? Since the final price is zero, taxation or exemption apparently does not make any difference. As regards tax revenue on inputs, that is also ineffective because that will increase the budgetary costs to the government for providing those services. The only difference is that exemption would break the chain with its accompanying implications. Some countries (Canada, some EU countries) have therefore adopted the practice of taxing these services and then rebating the VAT to public bodies engaged in these kinds of activities. This virtually converts the exemption to zero-rating. The problem of assessing the sales value for such services is often solved by looking at the user fee collection, if there is a user fee, or by the amount of budgetary subsidy provided to the agency in question. In any case, the overall revenue implication for the government would be zero. Financial services such as banking and insurance are typically VAT exempt because it is not easy to value their output and also because these are internationally mobile and putting them under VAT may drive them out of the country’s boundary while they can still do business from off shore locations. In that case, the country will not only lose VAT revenues but also personal income tax and corporate income tax revenues. Taxation of financial services is further discussed in a subsequent section. Taxing of Real Estate and Construction The ideal treatment of real estate and construction would be to tax the services that flow from them. Credit may, of course, be given if services are used as business input. The leasing of real estate for commercial purposes can be easily subject to VAT but this would be difficult in the case of owner occupied houses. To avoid distorting choices between renting and house ownership, the commercial leasing of residential property is also generally exempt. One way to tax residential property under VAT would be to tax at the time of house purchase. Taxing the subsequent resale does not make a lot of sense as the tax revenues paid by the buyer and the refund claimed by the seller would cancel out. The construction activities should be taxed in the normal way and credit given for construction as a business activity. This practice would ensure that those who construct houses for their own occupation, and are therefore VAT exempt on the final value of the house, at least pay VAT on their inputs. Those who buy houses from contractors would, of course, pay VAT on their purchases while the contractor can claim the refund on inputs. 200 VA L U E -A D D E D TA X Some Guiding Principles for Vietnam on Exemptions The above discussion provides some useful guidelines for reexamining the existing list of exemptions in Vietnam. Currently, the list of exempt categories of goods and services is long and broadly includes: (a) products of cultivation, husbandry, animal breeds and plant varieties; (b) financial/credit services and insurance (including engagement in stock exchange); (c) state owned dwelling houses sold to tenants ; (d) healthcare and animal health services; (e) renovation, and construction of cultural, physical and sports infrastructure ; (f) radio and TV broadcasting funded through state budget, newspapers and magazines publications; (g) public services of sanitation, water drainage, and irrigation in agricultural production; (h) public transportation by bus and tramcar; (i) weapons and military equipment for defense; (j) humanitarian aid and nonrefundable aid; (k) transfer of technology; (l) post, telecommunication and internet services; (m) machinery, equipment and supplies which cannot be produced domestically and need to be imported for direct use in scientific research and technological development activities; machinery, equipment, spare parts, special- purpose means of transport and supplies which cannot be produced domestically and need to be imported for prospecting, exploring and developing oil and gas ; (n) goods and services of business individuals with monthly incomes below the common minimum salary level applicable to domestic organizations and enterprises; (o) salt products; (p) transfer of land use rights; (q) teaching and vocational training, (r) goods transited through Vietnam territory; (s) Export of unprocessed mined resources; (t) Gold imported in the form of bars or ingots etc. Clearly there is no justification to exempt goods and services of business individuals based on income considerations measured by monthly average income level in comparison to the minimum wage level (item 25 of the exemption list). Goods and services provided by the public sector free of cost on a noncommercial basis are good candidates for exemption. This includes primary education and basic health. Cultural and other merit goods, aid financing activities, financial services may be considered for exemption because of their unusual situation. Residential rentals are another good candidate for tax exemption while new residential properties can be taxed. Other items in the list in section 5 of the act need close scrutiny and should be brought under the VAT net unless there are some strong reasons to keep them tax exempt. A cautionary note on taxing health and education services When bringing any good or service under the tax net from an exempt list, one has to make sure the sale price of the output is the commercial price and not the below market or subsidized price. Otherwise, the tax department may end up paying more tax refund on the input than collecting tax on the output. This kind of problem may arise in cases such as secondary education where the output price may not be market determined and has some hidden element of subsidy. Many countries (Botswana, South Africa) ended up losing revenues by not exercising caution. If there is an element of subsidy in sales price, it should be grossed up before applying VAT to avoid this problem. 201 TA X R E F O R M I N V I E T N A M Tax Exempt Threshold for Small Traders An important issue for simplifying tax administration is the choice of threshold for taxation under VAT. A high threshold creates the incentive for businesses to split or for some businesses on the borderline to misrepresent sales in order to remain below the threshold and therefore outside the VAT net. As against this, small traders will volunteer for VAT if they are supplying to big businesses who would naturally prefer VAT payments to be shown on their invoices for claiming input credit. On the other hand, keeping too low a threshold increases the burden on the tax administration. Present practice in Vietnam Currently, in Vietnam there is no threshold based on turnover for exempting small producers and traders from the VAT net. This means that everyone is subject to VAT. To deal with the problem of small businesses that cannot maintain records, a provision has been made to apply the subtraction method of VAT calculation for them. For those businesses that fail to fully observe regulations on accounting, invoices and documents and those trading in gold, silver and gems, the subtraction method or tax based directly on value added is applied. The value added is the selling price of goods or services minus the purchase price of such goods and services. This is estimated based on sales invoices or declared sales turnover and estimated purchase value or deemed markup as a proxy for input costs. The only exemption provided is based on the income for those small businesses whose monthly income is below the common minimum salary level in domestic organizations and enterprises. This practice of keeping the zero threshold and applying the subtraction method on small traders is somewhat unique to Vietnam. Clearly it increases the burden on the tax administration without a commensurate yield in revenues. A common characteristic of VAT in most countries is that the bulk of revenues—80 to 90 percent—come from a relatively small proportion—10 to 20 percent—of VAT payers. According to an estimate by the Ministry of Finance team (Watanabe and Thi Ninh 2006), about 95 percent of traders in Vietnam are using the subtraction method and the revenue collection from them amounts to about 6.2 percent of total VAT revenues. Thus the situation in Vietnam is more skewed toward a very small number of VAT payers paying almost the entire tax revenues while the tax administration has to spend substantial resources to collect a very small amount of revenue from a large number of small VAT traders. This, therefore, calls for the introduction of a tax exempt threshold in order to lower the burden on the tax department and also on the small VAT payers. This will clearly break the VAT chain and create some degree of cascading. Thus it may place those small VAT payers who transact with bigger taxable enterprises at a disadvantage. To mitigate this adverse impact, the option for voluntary registration should be available to the exempt VAT payers as well. Desirable threshold for Vietnam What would be the desirable threshold for Vietnam? Universally there is a great variation in the threshold level across countries. For instance, the threshold in Netherlands is as 202 VA L U E -A D D E D TA X low as US$1,643 while the threshold in Chile is US$477,633. The exemption levels vary considerably among Vietnam’s neighbors as well. Cambodia has a threshold of US$61,259 for services and US$122,520 for goods, while Indonesia has a threshold of US$109,165, Philippines US$10,612 and Thailand US$30,964.71 The level of threshold for Vietnam should be chosen so that it eliminates the bulk of those small VAT payers who do not make substantial contributions to revenues while imposing a large administrative burden. Most countries start with a high threshold and gradually lower it. In the case of Vietnam, the threshold has been virtually zero and this approach cannot be applied. It is, however, possible to examine, with the help of econometric modeling, several alternative thresholds in terms of their revenue implications and the resulting number of taxpayers. The one which is found to be most appropriate and cost effective should be selected. Taxing the “below threshold tradersâ€? The taxpayers below the threshold would be VAT exempt but some sort of taxation should be applied to them. Many countries have opted for a turnover tax of 2 to 3 percent which is then tax deductible by the next registered buyer. A similar tax may be applied in Vietnam. The question is who would administer this turnover tax? The VAT department is generally not suitable for this task because they are accustomed to dealing with self-compliance. In the case of Vietnam, however, the GDT presently administers this group of very small traders. So a division may be carved out within the GDT to look after this group exclusively when the majority of them move to the exempt category below the threshold. Another alternative is to give this task to the local government that may already be dealing with this group in the context of local taxation or charges and fees for doing business and therefore has better knowledge of them.72 It should be noted that there will still be a category of very small businesses that cannot be subjected to even this turnover tax because the revenues from them may not even be adequate to cover the costs of administering them. VAT computation method Both the credit method, called the value added tax credit method, and the subtraction method, referred to as the calculation of tax based on added value method, have been adopted for calculating the VAT liability in Vietnam. As explained above, the subtraction is applied to those individuals and small traders doing business who fail to fully implement regulations on 71 Source: IMF database 2006. 72 From July 1, 2007, when the new Tax Administration Law was in effect, the GDT would be reorganized by functional structure. However, to support smooth implementation of the new Personal Income Tax Law (starting from January 1, 2009), in the interim, the Department of PIT remains. There is a view that the PIT Department be also responsible for administering micro and small businesses with turnover below the prospective VAT threshold. But over the mid- to long term, the dilemma cannot be resolved—administering small businesses would overburden the already limited resources of the PIT Department, and the Department may not be equipped with appropriate strategies and specialized resources to deal with compliance by the special group of small taxpayers. 203 TA X R E F O R M I N V I E T N A M accounting, invoices, and vouchers to serve as the bases for tax calculation under this method and to businesses involved in gold, silver and gems trading activities.73 Once a suitable threshold for exempting small traders is chosen and implemented, the need for this dual computing system may be abolished and only the credit or tax deduction method should remain applicable. 3. Taxation of Special Sectors of the Economy There are some sectors of the economy: agriculture, real estate, services, including tourism, for instance, that have some special characteristics and therefore deserve separate treatment under VAT. Agriculture Theoretically, the agriculture sector should be treated like any other sector of the economy and should be put under VAT. In practice, however, most countries, developed or developing, give special treatment to agriculture. In most countries, agriculture is exempt under VAT, while a few transition economies tax it with the provision of exempting small traders, which includes small farmers. Many countries also exempt agricultural inputs such as fertilizers, pesticides, seeds, etc. In a nutshell, around the world agriculture is taxed lightly under VAT. Some special features of agriculture are behind this universal practice. First, is the administrative problem of taxing a sector which is largely informal and scattered and therefore difficult and costly to administrate. Second, the distributional aspect is quite relevant in this case. Any tax burden will be shared between the consumers and producers and will be generally regressive in either case, particularly for basic foods. Third, taxation of agriculture is a politically sensitive matter whether in developed or developing economies. While bringing agriculture within the purview of VAT may be the desirable objective in the long run, exemption of this sector seems to be the compromise option in the near future. This excludes farmers from any administrative burden and at the same time ensures that the sector is taxed lightly through input taxation. This approach, however, creates the problem of carrying the tax into the final food prices as well as into the prices of food exports. To alleviate this burden of input taxation, both zero-rating and exemption of agricultural inputs have been suggested as remedies. Thus the preferred strategy for developing countries for this sector may be to exempt it within a high threshold so that large and rich farmers are still under VAT. Agricultural inputs may be either zero-rated or exempt. It is worth emphasizing the need to unambiguously define the limited number of inputs used exclusively for agriculture (for example, fertilizer, seeds) in order to avoid further compliance and efficiency costs. Finally, small farmers should be given the option of voluntary registration so that they are not disadvantaged. 73 Many developing countries have adopted levying of a turnover tax in the range of 2 to 3 percent, called a “unifiedâ€? tax, for unincorporated business in lieu of both the VAT and income taxes. Vietnam, however, has gone beyond that stage and its current dual system can easily move to a pure credit invoice method. 204 VA L U E -A D D E D TA X The present practice in Vietnam generally fits into the strategy presented above and this may be allowed to continue for some time to come. The first two categories of items in the list of exempt goods/services pertain to the agriculture sector and related activities. However, the large farmers may be considered for inclusion in the VAT system. Also, small farmers may be given the option of voluntary registration under VAT. Financial Sectors Financial services are treated differently under VAT for a variety of reasons. First, it is difficult to identify on conventional invoice computation the value added on each service provided by financial institutions. There is seldom a market price for these services that can be used for computing the VAT liability. If normal interest rates were taxed, that would be taxing a combination of inflation and return on savings. Thus, instead of taxing consumption it would become taxation of income. Second, this taxation would be mostly applicable to households while for businesses it would be deductible as credit. As a result, the tax base tends to exclude the business community and essentially remains narrow. Third, putting them under VAT would increase the cost of borrowing to the private sector which becomes a politically sensitive matter. Finally, both capital and financial services are internationally mobile and they can easily move offshore and operate from there. Keeping these special features in mind, the preferred approach has been to exclude financial services from VAT. Under this arrangement, at least they keep paying taxes on all their inputs: buildings, computers, office equipments, stationery and these are not credited back. That is how most European VATs treat financial services. Vietnam has adopted a similar approach and has exempted insurance, credit provision services, capital transfer, derivative financial services and forward and future contracts. This seems to be a reasonable approach. Tourism Typically all services should be taxed under VAT except for a few exempt services such as healthcare, education, social and financial services. Therefore, tourism should be under normal VAT. It is true that as of now tourism is a budding industry in Vietnam and the industry may be seeking tax concessions from the Government. Often the labor content as a proportion of the final price in tourism is high and, as a result, these services are provided by small scale establishments. Therefore, the best approach may be to keep the tourism industry under the VAT net but exempt the small tourism operators and travel agents under the exemption clause for small traders if their turnover is below the prescribed threshold. Real Estate Taxation of real estate and the construction sector has been dealt with in section 6.4 above in the context of the exempt items. 205 TA X R E F O R M I N V I E T N A M 4. Zero-rating and VAT Refunds Zero-rating of exports is essential in order that all the taxes are taken out of the production chain of an exportable item and it remains internationally competitive. A destination based credit method VAT easily meets this requirement. However, for this to be effective, prompt refund of input VAT is necessary. In principle, refunds should be paid without any delay otherwise it becomes a penalty against investors. Precise data about the amount of refund paid are not readily available for developing countries but it is possible to have a sense of the dimension of this problem by looking at some of the developed countries. For instance, according to the refund data collected by the IMF for the year 1995–96, the U.K. collected over US$68 billion dollars and returned US$26 billion in refunds. Thus refunds were about 60 percent of net collections which is unusually high. In the case of France, the refund is normally about 40 percent, in Sweden 80 percent while for Korea it is about 50 percent. In most developed countries, VAT refunds exceed 40 percent of the gross VAT collections while in most developing countries of Africa, Asia, and Latin America, it amounts to less than 20 percent of the VAT revenues. Thus the quantity of refund is substantial and ideally the excess input tax should be refunded fully after each tax period to avoid any adverse impact on the trade and business. Different countries have, however, adopted different practices for processing and paying refunds. Some key features and international practices of refunding excess credit are examined below. Some Important Features of VAT Refunds A key feature of the credit method VAT is that some businesses would pay more VAT on their inputs than they collect on their sales and would therefore claim excess credit refunds. It is therefore important to have a refund system in place if the VAT regime has to be implemented successfully. At the same time, a large amount of VAT revenue is lost as a result of VAT refund abuse. While countries have found it generally difficult to estimate the extent of revenue loss, it is believed to be substantial. Tax authorities in the U.K. estimated the amount of VAT losses in 2002–03 to be around 16 percent of the net VAT receipts. VAT abuse comes in different forms, ranging from traders omitting sales from records to falsification of invoices. Sometimes, businesses are launched without any legitimate activity for the sole purpose of stealing money through the refund system. The following are some notable aspects of VAT refunds around the globe. • Most of the VAT refund claims are made by exporters both in terms of numbers and amounts. Typically, a small number of large exporters account for the bulk of VAT excess credit. • The other businesses claiming refunds are comprised of those traders that have made large purchases of capital goods relative to current sales. Also, in countries with a dual rate structure, refund claims arise when outputs are taxed at the lower rate while inputs are taxed at a higher VAT rate. • The VAT registration process should contain sufficient checks and controls to prevent registration of fictitious traders who want to register with the sole intent of stealing from 206 VA L U E -A D D E D TA X the VAT system. The registration staff should carefully assess and verify initial information presented for registration. • The level of threshold for compulsory registration is relevant not only for effective administration of VAT but also for reducing refund abuse. If the threshold is too low, the tax administration resources become over stretched in managing the VAT registered businesses and monitoring refund claims. Vietnam, with zero threshold, faces this problem. So an appropriate level of threshold is important for effective monitoring and management of VAT refund. • An effective risk-based audit system plays an important role in reducing VAT refund fraud. Exchange of VAT and income tax information and VAT and customs information are important in this regard. Where direct and indirect taxes are administrated together in a unified tax administration, exchange of VAT and income tax information is routine. Where these taxes are administered separately, however, exchange of information is generally limited. In some countries, this kind of exchange is legally prohibited. As a substantive share of VAT revenues are collected through the customs agency, the cooperation between VAT and customs also becomes important. International Practice in VAT Refund and Policy Implications for Vietnam In most developed countries, refunds are usually paid within three to four weeks of the end of the taxable period. Also, there are virtually no restrictions on the refund payment. In developing countries and transition economies, however, the situation is different and this turns out to be one of the weak links in the VAT administration. Many of these countries limit the entitlement to refund. The IMF conducted a survey of 36 such countries and come up with the following findings74: • Ninety percent of the countries in the survey reported that their tax authorities are bound by law to make refunds within a prescribed timeframe, generally 30 days. It is also true that the statutory deadlines are often not met. Around 40 percent of the countries go further and provide for interest payments on late refunds. In eight countries, refunds are focused on exporters only. Six countries give refunds to exporters immediately and the rest give refunds with some time lag, both to exporters and other businesses. • Most countries (60 percent of the surveyed countries) now ask their taxpayers, particularly nonexporters, to typically carry forward the refund for a specific period – generally six months. While this practice enables the tax administration to verify genuineness of refund claims, these measures create hardship to investors and are not welcomed by firms who are making large investments and production and sales do not begin immediately thereafter. Many VAT systems allow for refunds to be offset against other tax liabilities, for instance, the income tax liability. • Four of the surveyed countries make refunds in excess of a floor credit only. One country adjusts only once every year at the end of the year. One country makes no refunds at all even though it is provided in the law and one other country pays refunds in an ad hoc manner. There is one country that pays refunds only after an audit. Two countries in Africa have enacted legislation to deny all outstanding claims as of a specific date. This, of course, has serious implications for the long-term integrity of the tax system and the 74 See Harrison and Krelove (2005) for details. 207 TA X R E F O R M I N V I E T N A M credibility of tax administration. Thus clearly there is a great deal of variation among countries. Also, the actual refund practices are found to be quite different from what is laid down in the law. • Two-thirds of the surveyed countries undertake some sort of risk assessment in processing VAT refunds. Sometimes it is quite sophisticated (UK) while in many countries it is often rudimentary and may amount to verifying each refund claim. A quarter of surveyed countries have a statutory requirement to verify every refund claim. • It is also found that a lack of an appropriate audit program in many countries contributes to the existence of a poor refund system. A third of the surveyed countries report that they do not have a VAT audit program. These countries end up carrying out extended prepayment checking of claims rather than relying on a sound postpayment audit. This naturally delays the entire process. It is important for the tax administration to have a sense of the level of refunds payable so that financial resources are in place in a timely fashion and also as a warning signal against attempted fraud. • Exchange of information about VAT registered traders and the value of supplies across countries has become important for international transactions. For instance, EU member states exchange information via a VAT information exchange system (VIES) about VAT traders and sales of goods and services. • There are two methods of budgeting for VAT refunds: (a) making payments from gross VAT revenues, (b) paying from budget expenditure appropriations. Irrespective of the method used, it is necessary to have a forecasting and monitoring system to estimate VAT collections and level of refunds. Actual collections and refunds should be tracked against forecasts and variances should be explained and reconciled. • Some countries have adopted innovative practices to reduce refund fraud. For instance, Kenya provides for mandatory certification of VAT claims by certified public accountants (CPAs) for claims exceeding specified amounts. Sanctions are also imposed on CPAs who knowingly certify false claims. A number of countries give preferential treatment to taxpayers with sound compliance histories (“goldâ€? status in Pakistan). Some countries (Bulgaria) have introduced a VAT bank account system where a VAT registered business must open a VAT account for depositing VAT payments and spending money on VAT purchases. This is meant to reduce cash transactions and thus reduce VAT fraud. • The following lessons with significant policy implications for Vietnam’s VAT system can be learned from analysis of international experience in VAT refunds. • The number of taxpayers should be kept at a level that can be realistically managed by choosing the right threshold and verifying the VAT registration applications to prevent fictitious traders entering the system. • Suitable forecasting and monitoring systems should be established to anticipate refund levels and make provisions for refund payments. • Refunds should be processed in a reasonable period of time and interest paid on late payments. Excess VAT credit should be offset against VAT and other tax arrears. • Exporters should be paid refunds promptly. In a functioning risk management, the tax department should make a clear distinction between claimants with a history of compliance and those who are new. The Tax department should maintain historical profiles for each refund claimant. 208 VA L U E -A D D E D TA X • Verification of VAT refund claims should be a component of a wider audit program. A prerefund audit should be implemented for high risk claims while a normal post audit should be conducted for lesser risk cases. • Given the substantial size of VAT refunds, one part of an audit department may simply focus on refund audits, that is, verifying the facts of refund claims. • VAT refunds also create opportunities for corrupt practices by tax and customs officials and this has to be carefully monitored. • Some countries, especially among transition economies, have tended to delay refunds and use the resources to overcome temporary revenue problems. While this kind of interest free loan from the business sector may appear attractive in the short run, it is bound to create a serious backlash and adversely affect the investment climate in the country and therefore such practices should be avoided. VAT Refunds in Vietnam The present VAT law (article 13) in Vietnam is somewhat complicated and varies with the nature and activity of a particular business. This is despite the fact that the 2008 VAT law rationalizes the refund system considerably. Below are some conditions for getting refund: i. Business establishments which pay value-added tax, according to the tax credit method are entitled to VAT refund if, for three or more consecutive months, they have some input value-added tax amount not yet fully credited. ii. Business establishments having registered to pay VAT according to the tax credit method are entitled to tax refund if they have new investment projects and some amount of value- added tax on purchased goods or services used for investment not yet fully credited and the remaining tax amount of 200 million Vietnam dong or more. iii. Business establishments which export goods or services in a month are entitled to value- added tax refund on a monthly basis if they have a noncredited input Value-Added Tax amount of 200 million Vietnam dong or more. iv. Business establishments which pay VAT according to the tax credit method are entitled to value-added tax refund if upon ownership transformation, enterprise transformation, merger, consolidation, separation, split, dissolution, bankruptcy or operation termination, they have an overpaid value-added tax amount or have some input value-added tax amount not yet fully credited. v. Business establishments possessing value-added tax refund decisions issued by competent agencies as provided for by law, and cases eligible for value-added tax refund under treaties to which the Socialist Republic of Vietnam is a contracting party. As mentioned above, excess input tax should preferably be refunded fully after each tax period, quarterly in the case of Vietnam. This is, however, not feasible in most developing countries with limited administrative capacity. The carry forward system is mainly advocated as a safeguard against fraud; but that can only really be detected through a quick and effective audit system. Thus one main area of weakness in refunds could be the lack of effective and prompt auditing capacity. Therefore the long term solution is strengthening administration and especially enhancing the audit capacity in Vietnam. 209 TA X R E F O R M I N V I E T N A M Also, in Vietnam some other complications related to VAT refunds need to be remedied. There is no justification for distinguishing between the VAT refund on capital goods versus the VAT refund on other inputs. A carry forward period of three months may be applied to every business and a threshold for obtaining the current period refunds may also be fixed but it should not be excessive and should remain uniform across different businesses. This would render the system simpler and more transparent. All new businesses must be audited for a couple of years before being given refunds to make sure they are genuine units and not a scam. Otherwise, a new unit may come up and claim a lot of refunds based on false invoices and then disappear. Older, more established units are less likely to commit fraud. To deal with the concerns of investors about late refunds, some countries have adopted the practice of allowing deferment of VAT payments on specific capital imports until VAT input deductions are claimed. This approach requires effective customs control and coordination between customs and domestic VAT administration. Sometimes, zero-rating of capital equipment items has been tried in a few countries but this may result in a net revenue loss to the extent that capital equipment is used by unregistered businesses as well. Importance of Customs in VAT Collection An important feature of VAT collection brought out by empirical studies of countries in Asia, Africa, Latin America and Central Europe is that the revenue collected on imports normally accounts for a large proportion of total VAT revenues in most countries. A study by the IMF of this ratio in more than twenty developing and transition economies shows that the ratio varies between 40 percent (Peru) to 70 percent (Bulgaria and Haiti). This clearly brings out the importance of customs collection in VAT revenue collection. There is another way in which customs plays an important role in VAT collection. As the tax policy analysis for VAT would show, under the credit method if the final sales are VAT exempt, the chain terminates at the stage of the previous input purchase. In that case, there is some loss of revenues at the stage of final sales but total revenues are not lost because the VAT collected on the input is not refunded. In many developing and transition economies, this is what happens on the sale by the informal sector of the economy. The informal sector is responsible for a substantial share of final sales to consumers in these countries and by its very nature this sector often does not pay VAT. But if the customs department is effective at the border, the imported inputs used by the informal sector would be collected. Since the informal sector is not registered under VAT, the taxes on inputs paid at the border are not refunded to them even though the tax on value added at their stage is lost. Thus customs effectiveness is an important factor in the effectiveness of total VAT collections in the countries where an informal sector is prominent. In addition, exchange of information between the customs department and the VAT collection agency would help the latter in crosschecking whether the traders who are importing part of their inputs are paying the right level of tax on their production. Thus clearly the extent of cooperation between the customs and VAT administrations is crucial for the successful 210 VA L U E -A D D E D TA X enforcement of VAT. In fact, close cooperation among the different parts of tax administration is helpful for enforcement of all taxes – income tax, VAT and custom tariffs. However, this kind of cooperation in developing and transition economies is more of an exception than a rule. Clearly, a separate VAT department independent of the domestic tax department that administers income taxes is bound to make this kind of cooperation harder to achieve. In the case of Vietnam, since VAT is administered by the GDT, which is responsible for all domestic taxes, cooperation and coordination between GDC and GDT would achieve the objective of enhancing VAT enforcement. The question is what kind of arrangement between the two agencies would ensure this cooperation? Does it have to be some sort of formal arrangement or would an informal arrangement do? In most countries, the problem arises due to two reasons. First, the confidentiality guaranteed to the taxpayer by the law may be a hurdle in sharing of information. In most countries, however, the law does not preclude sharing of information within the tax administration bodies and the ministry of finance. For instance, in Vietnam article 6 of the Tax Administration Law mandates that the taxpayer information should be kept secret by the tax administration bodies. But this cannot be interpreted as stopping the GDC and GDT from information sharing between them or with the Ministry of Finance (MoF). Second and more importantly, it is administrative apathy and compartmentalization that is the main problem. Vietnam may not be an exception to this problem and historically there may be a general lack of coordination among GDC, GDT and the MoF (tax policy unit). This may, however, be remedied through an administrative arrangement where clear instructions are laid down for periodical sharing of information on prescribed formats. If it is deemed necessary, the task of coordination and cooperation among the various tax agencies may be supervised by a senior officer or deputy minister nominated by the Minister of Finance. 5. Policy and Administrative Recommendations Rate structure a. Vietnam should move to a rate structure with zero and one positive VAT rate instead of zero and two positive rates. The lower VAT rate of 5 percent could be eliminated. Exemptions a. The current list of exempt items needs to be closely scrutinized and only a few of the 25 items should remain exempt. For instance, financial services, primary education, basic healthcare (not higher education and specialized health services), some cultural and merit goods and services, aid financed activities, and residential rentals are good candidates for tax exemption (while new residential construction can be taxed). The rest of the items may be moved to the taxed category. b. When moving an item from the exempt list to the taxed list, care should be taken to ensure that the sales price of that good or service is not below market or subsidized price otherwise the VAT refund on inputs may exceed the VAT on the output and the government may end up losing revenues. In such cases, it is better to leave it in the exempt category. 211 TA X R E F O R M I N V I E T N A M An input-output VAT base modeling is applied to assess the revenue impact of rate unification and/or rationalization of exemption. The main source of data is Vietnam’s I-O Tables 2005, supplemented by the results of the 2004 household expenditures survey and other macroeconomic data (for example, economic growth rate, growth rates of various components of GDP, in particular, private and government consumption and investment). Details of modeling methodology and projected results are presented in Chapter 4. The results are summarized as follows: a. Unification of the rate structure (at the standard rate of 10 percent) with existing exemptions would generate expected revenue gains of about 5.8 percent. b. Unification of the rate structure combined with rationalization of exemptions would expect to generate revenue gains of about 18 percent. In both reform scenarios, it is assumed that the existing compliance rate remains unchanged. The estimates do not take into account the savings in reduced compliance and administration costs due to the simplification of the VAT regime. Small traders a. The current practice of having a zero threshold of exempting small traders should be abolished and a suitable threshold adopted. The final choice should be based on the tradeoff between the cost of administration and revenue implications. b. For the small traders below the threshold some sort of taxation should be envisaged. Applying a turnover tax of 2 to 3 percent which is then tax deductible by the next registered buyer may be considered. It should be administered by a special division within the GDT that is familiar with the problems of dealing with small traders. Alternatively this responsibility may be handed over to the local authorities. Method of computing VAT liability a. With the adoption of the credit method across the board, the subtraction method of VAT calculation for some traders who are unable to observe accounting regulations may be abolished. Exports and VAT refund a. The current VAT refund system needs to be rationalized. There are different regulations for different kinds of businesses and this makes the system complex and opaque. While the long-term solution of this problem is linked to improved administrative capacity and practices, some short-term measures are required. While deferment on imported capital goods or zero-rating their sales may offer some relief, these measures have their own problems. At least, some uniform system of waiting and threshold should be mandated to mitigate this problem. For instance, a carry forward period of three months may be applied to every business and a threshold for obtaining the current period refunds may also be fixed but it should not be excessive and should remain uniform across different businesses. b. All new businesses should be audited before paying refund to make sure these are genuine units. 212 VA L U E -A D D E D TA X Administrative cooperation between GDT and GDC a. Cooperation and coordination between the GDT and GDC is at the core of enhanced effectiveness of VAT administration and collections. This arrangement may be better finalized either through the provisions of the Tax Administration Law or through administrative orders. Timeline for Developing a Modern VAT Regime in Vietnam The time needed to develop a green field VAT satisfactorily depends upon several factors (see Tait 1991) including: a. Complexity of the VAT law b. Experience with operating indirect taxes c. Experience with operating credit mechanisms d. Existence of a modern information technology (IT) system used as an operational tool e. Experience with implementing large scale tax changes f. Availability of high quality trained staff g. Political commitment. With these elements in place, the introduction of a green field VAT can usually take between two to three years. The various steps involved would include: a. Preparatory and development steps b. Legislation and regulations c. Organizational changes d. Operational design e. Publicity and advertising f. Manuals and forms g. Training development and training delivery h. Computerization of headquarters and the field offices. In Vietnam, VAT has already been introduced, and the GDT has gained extensive experience with its operations including credit mechanism and refunds. There is commitment for tax reform at the highest order. In fact, an ambitious reform plan in which all the taxes are to be reviewed and suitably reformed one by one has been launched and the process is likely to be completed in the near future. Thus in Vietnam, it is the case of reforming the existing VAT law and administration in the country. Clearly this can be done in fifteen to eighteen months from start to finish. The Government is in the process of formulating the laws and taking them through the National Assembly. This perhaps may be the most time consuming part of the whole process. The important thing at this point is that the preparation of the manuals should also be completed as soon as possible as they are vital for an efficient VAT administration and consistency throughout the department. 213 Taxation In Vietnam: Who Pays What? VIII By Jonathan Haughton 1. Introduction This paper has two purposes. The first is to document the incidence of taxation on households in Vietnam, using recent data, from 2006. The most recent publicly available study uses data from 1998 (Haughton, Quan, and Bao 2006), and so does not take into account the rapid changes in the tax system and in household spending and income patterns since then. Information on who bears the burden of different taxes is helpful to policy makers when they consider changes in the structure of taxes, and so will be useful as Vietnam reshapes and modernizes its tax system over the coming few years. The second, and more important, purpose of this study is to trace the implications of tax incidence for the design of the personal income tax in particular, and the tax system in general. In December 2007, the National Assembly approved a major reform of personal income taxation, to come into effect in January 2009. Here we evaluate the revenue and incidence effects of this change, and situate the discussion in the wider context of tax reform. The next section summarizes the current structure of taxes in Vietnam. Direct taxes fall on income, and indirect taxes are imposed on expenditures, and so an understanding of tax incidence calls for an appreciation of how the different components of income and expenditure vary as one goes from poor to rich households, or from region to region; this is addressed in section 3. In section 4 we then present detailed information, both graphically and in tabular format, on the incidence of several important taxes, and compare the relative progressivity of each. We provide an analysis of the revenue and incidence effects of proposed changes in the personal income tax in section 5; in the concluding section we assess the implications of our findings for the direction of tax reform. 2. Current Tax Structure A list of the main taxes, and the revenue that they have yielded annually since 1997, are shown in table 8.1. There is a clear discontinuity in total revenue mobilization, from about 20 percent of GDP up to 2000, to about 25 percent of GDP in the years since 2003. Among the major taxes, the value-added tax (VAT) has steadily increased in importance, and now collects the equivalent of 7 percent of GDP, a remarkable performance for a tax whose main rate is just 10 percent. Also noteworthy is the relative decline in revenue from taxes on trade, from over 4 percent of GDP in the late 1990s to about 2 percent of GDP now. This has occurred despite a very rapid increase in imports relative to GDP, and reflects a sharp fall in the average effective import tariff—from 5.7 percent in 2003 to an estimated 1.8 percent in 2008—due in part to cuts in rates on imports from ASEAN countries. 215 TA X R E F O R M I N V I E T N A M The third major source of tax revenue is income tax. The revenue collected by the corporate income tax peaked at over 10 percent of GDP in 2006; much of this revenue comes from the state-owned industrial sector, whose share of GDP is now shrinking (EIU 2008). The personal income tax yields just over 2 percent of government revenue, a share that has changed little over the past decade. In table 8.1, the numbers for 2007 are preliminary and the figures for 2008 are estimates. However, if they can be believed, Vietnam will face some difficulty in maintaining its current level of revenue mobilization (relative to GDP), with anticipated relative declines in revenue from corporate income tax, and taxes on trade. The decline of oil production in the middle of the decade did not help – tax revenue attributable to oil (as a fraction of GDP) peaked in 2005, although some recovery has occurred since. Table 8.1: Revenue from Taxes and Fees, 1997–2008 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 (e) trillions of VND, current prices Total revenues and grants 65.4 73.0 78.5 90.7 103.9 121.7 153.0 191.3 217.1 264.2 287.9 323.0 of which: Taxes Corporate income tax 11.6 13.1 14.5 22.2 25.8 29.3 47.4 57.0 71.7 100.8 98.9 108.1 Individual income tax 1.5 1.8 1.9 1.8 2.1 2.3 3.0 3.5 4.2 5.2 6.9 8.1 Capital user charge 1.5 1.7 1.5 1.6 1.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Land and housing tax 0.3 0.3 0.3 0.4 0.3 0.3 0.4 0.4 0.5 0.6 0.6 0.7 Licence tax 0.4 0.3 0.4 0.4 0.4 0.4 0.8 0.7 0.7 0.8 0.8 0.9 Tax on the transfer of properties 1.0 1.0 1.0 0.9 1.2 1.3 1.8 2.6 2.8 3.4 4.5 5.2 Tax on land use right transfer 0.3 0.4 0.3 0.2 0.3 0.3 0.4 0.6 1.0 1.3 1.7 2.0 Value added tax 11.8 11.8 17.2 17.2 19.3 25.9 33.1 38.8 45.7 54.8 78.6 96.7 Special cons. tax for domestic 4.6 5.6 4.5 5.3 6.2 7.3 8.9 12.8 15.7 17.1 17.0 19.9 Natural resouces tax 3.4 3.3 4.6 7.5 8.4 8.5 9.7 17.4 21.9 20.2 19.1 19.6 Agricultural tax 1.7 2.0 2.0 1.8 0.8 0.8 0.2 0.1 0.1 0.1 0.1 0.1 Imp - Exp. tax, special cons. tax on imports 13.5 14.9 14.4 13.4 17.5 21.9 22.4 21.6 23.6 26.3 25.0 26.2 Other taxes 1.5 2.9 2.2 0.3 0.3 0.4 0.2 0.0 0.0 0.0 0.0 0.0 Fees, charges and non-tax From discrepancy of import prices Fees and charges (include gasoline fee) 7.1 7.0 7.3 12.3 13.6 15.8 16.6 23.4 22.1 24.5 9.0 9.9 Rental of land 0.5 0.5 0.6 0.6 0.6 0.5 0.5 1.0 1.0 1.6 1.8 1.7 Others 2.1 4.3 3.4 2.8 3.4 4.4 4.6 8.5 3.8 3.9 2.7 2.9 Capital revenues (revenues from sale of State - owned houses, land use right assignment) 17.7 17.6 Grants 2.6 2.1 2.4 2.0 2.0 2.3 3.0 2.9 2.3 3.6 3.4 3.6 Memo items: GDP, in current prices 313.6 361.0 399.9 441.6 481.3 535.8 613.4 715.3 839.2 973.8 1143.6 1358.7 GDP, in 1994 prices 231.3 244.6 256.3 273.7 292.5 313.2 336.2 362.4 393.0 425.1 461.0 502.5 Imports, in current prices 142.5 159.7 164.7 226.7 238.8 301.7 391.7 503.2 583.0 718.0 977.5 1459.9 as percentages of GDP Total revenue and grants 20.9 20.2 19.6 20.5 21.6 22.7 24.9 26.7 25.9 27.1 25.2 23.8 of which: Corporate income tax 3.70 3.63 3.63 5.03 5.36 5.47 7.73 7.97 8.54 10.35 8.64 7.96 Individual income tax 0.48 0.50 0.48 0.41 0.44 0.43 0.49 0.49 0.50 0.53 0.60 0.60 Value-added tax 3.76 3.27 4.30 3.89 4.01 4.83 5.40 5.42 5.45 5.63 6.87 7.11 Excise taxes 1.47 1.55 1.13 1.20 1.29 1.36 1.45 1.79 1.87 1.76 1.49 1.46 Natural resources tax 1.08 0.91 1.15 1.70 1.75 1.59 1.58 2.43 2.61 2.07 1.67 1.44 Taxes on trade 4.30 4.13 3.60 3.03 3.64 4.09 3.65 3.02 2.81 2.70 2.19 1.93 Fees, charges, and non-tax revenues 3.09 3.27 2.83 3.55 3.66 3.86 3.54 4.60 3.21 3.08 2.73 2.36 Grants 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 All other taxes 2.96 2.96 2.52 1.72 1.45 1.08 1.10 1.02 0.88 1.00 0.98 0.91 Memo item: Revenues from oil n.a. 4.0 4.6 6.5 7.4 6.8 7.0 8.0 8.7 8.2 6.3 4.8 Non-oil revenues n.a. 16.2 15.0 14.1 14.2 15.9 17.9 18.8 17.1 18.9 18.9 18.9 Memo item: Oil revenue as % of total revenue 20.0 23.6 31.4 34.4 29.8 28.2 29.7 33.8 30.3 24.9 20.3 Import duties as % of total import value 9.5 9.3 8.7 5.9 7.3 7.3 5.7 4.3 4.0 3.7 2.6 1.8 Sources. IMF (2003, 2006, 2007) for figures to 2006; Ministry of Finance web site for 2007 and 2008(e) figures; EIU (2008) for growth rate for 2007. Notes. Figures for 2007 are revised estimates. Excise taxes are those levied on domestically produced goods; excise collected on imports is included in trade taxes. GDP figures for 2008 assume 11% inflation and 7% growth in real GDP. Estimates of GDP for 2007 assume 8.3% inflation between 2006 and 2007. Revenue figures for 2007 are revised estimates. 216 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Taxes Affecting Households In measuring the incidence of taxation on a household, it is necessary to trace the effect of each tax on the household’s income or expenditure. This can be done for some, but not all, taxes. For instance, there is no obvious way to allocate the burden of a tax on natural resources (such as a royalty on oil) to individual members of the population. However, given reasonable assumptions about tax shifting, it is possible to trace the burden of direct taxes (that are levied on income) and indirect taxes (that are levied on spending). Table 8.2 summarizes the essential features of these taxes; for a fuller treatment, see Shukla (2006) or IMF (2007). The personal income tax is a direct tax that is in effect mainly a tax on wages and salaries. It is generally remitted to the General Taxation Department by employers, and three fifths of the revenue comes from taxing approximately 50,000 foreign employees. The remaining tax is collected from about 150,000 Vietnamese citizens, most of who are working for foreign-invested companies. The very small number of taxpayers—just 0.2 percent of the population—is due in large part to the high threshold: one only begins to pay the tax with an income of at least VND 60 million annually, which is about six times higher than average per capita income. The rates are graduated, and reach a maximum of 40 percent on annual income above VND480 million (US$30,000). The tax brackets are wider for foreigners working in Vietnam. We assume that the burden of this tax falls on the employee. Although most of the revenue from the corporation income tax comes from taxing the profits of incorporated businesses, this tax is also applied to the profits of approximately 100,000 household enterprises. It is levied at a flat rate of 28 percent. In measuring incidence, we assume that the burden of this tax is on the household (or business) that has to pay the tax. A number of taxes and fees are collected directly from households. A business license tax is levied on all registered enterprises (which may or may not be formally incorporated), and starts at a flat rate of VND1 million per year. The agricultural land tax has in effect been suspended, but the land and housing tax is collected in urban areas, although it in practice it is levied on the value of land but not housing. The property transfer tax is collected at a rate of 2 percent of the value of land (if agricultural), and 4 percent on other land, with higher rates for transfers involving businesses. In each case we suppose that the tax falls on the initial owner. Many households also pay a number of local fees and “contributions.â€? Some of these are, in principle at least, paid in return for a service—for irrigation, to maintain dikes, for veterinary services, for plant protection, for schooling, and for health care. In practice, many of these fees are only loosely related to the services with which they are supposed to be associated, and so have most of the characteristics of a tax. In measuring incidence, we assume that rural fees are like taxes, and that the burden falls entirely on the payer. We are not able to measure all of the contributions: for instance, Vietnamese households are required to provide ten days of labor for the public good annually (or an equivalent in cash); the time is spent mending roads and dikes and otherwise maintaining the local infrastructure, but unfortunately data on the extent of these contributions are not available for 2006. 217 TA X R E F O R M I N V I E T N A M The most important direct tax is the value-added tax (VAT), which is levied at a rate of 10 percent on most items; however, some goods face a lower rate of 5 percent, others (such as exports) pay 0 percent, and some important items are exempt (such as some basic unprocessed foods). We assume that the VAT is passed on to consumers in the form of higher prices; thus, if we know how much households consume of each good or service, we can apply the appropriate VAT rate and infer how much tax the household is effectively paying. In addition to the VAT, there are excise (“special consumptionâ€?) taxes on a number of important goods, including cigarettes (now taxed at 65 percent), beer (75 percent on canned beer, 40 percent on draught beer), automobiles (50 percent on sedan cars, for instance), gasoline, air conditioners, dance halls, and betting. Here again we assume that the effective incidence of the tax is on the consumer, even when the tax is paid formally—the statutory incidence—by the vendor. Table 8.2: Summary of Main Taxes in Vietnam Tax Base Rates Personal income Levied on wages, salaries and benefits, Tax at 10% begins at an income level of tax but not including most allowances, VND60m (US$3,750) annually, rising interest, dividends. to a maximum of 40%. Wider brackets Separate, lower, tax is levied on for foreigners. Note: A new personal irregular income such as lottery income tax structure will come into winnings. effect in January 2009. Enterprise Taxable income is defined as total 28%. Reductions for investments income tax revenue less deductible expenses in favored sectors (e.g., scientific (depreciation, cost of goods sold, research) and areas (e.g., mountainous research and development costs, areas). Refund of 50 to 100% of tax on interest). Losses may be carried reinvested profit. forward up to 5 years. Applies also to household enterprises. Social security Applied to salaries 15% paid by employer plus 5% paid by insurance employee. VAT Applies to most goods, uses credit 0% for exports, 5% for “essentialsâ€? and method. Exemptions include 10% standard rate. agricultural production, salt, some imported equipment, credit, business services, education. Special sales tax Base is sales price divided by (1+tax Cigarettes: 65%. Beer: 40 to 75%. (excise) rate). Not levied on goods that are Liquor: 20 to 65% depending on directly exported, or on goods brought proof. Autos: 15 to 50%, with highest to Vietnam by aid agencies. rates for sedans. Gasoline: usually 15% but varies inversely with world price. Natural Royalties are levied on sales value. 2 to 5% on metallic minerals, 1 to 3% resources on coal, 6 to 25% on oil, 0 to 10% of natural gas, 5 to 40% on natural forest products. 218 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Import duties Levied on cif price, average tariff is Rates vary from 0% to 60%, with most about 8%. Some exemptions for aid, in the 1%, 3%, 5%, 10%, 15% brackets. goods in transit, education, research, for export processing, and certain machinery & equipment. Export duties Levied on a few items only. Oil: 4%. Wood: 5 to 20%. Cashews: 4%. Property taxes Agriculture and land use tax is being Computed as 50 to 650 kg. paddy/ha phased out; exempts barren land, but paid in cash equivalent. reclaimed land, and households “in difficulty.â€? Land and housing tax. As for agriculture and land use tax, but higher rates. Land use right transfer tax. 2% for agricultural and fishing land, 4% otherwise. Registration fees Applied to boats, cars, motorbikes, and 0.5% to 2%. guns. Sources: Haughton, Quan and Bao 2006; IMF 2007; Shukla 2006. In this study we do not attempt to track the incidence on households of import duties, export taxes, resource taxes, property transfer taxes, land taxes, or taxes levied on incorporated businesses. In most cases the reason is pragmatic: adequate data are not available. We are thus able to track the incidence on households of slightly over half of all tax revenue. But before presenting these results, it is helpful to set out in more detail the components of income and expenditure—and hence, the tax bases of the various taxes under consideration. 3. The Structure of Household Income and Expenditure In order to measure the burden of different taxes on households, it is essential to have relatively detailed household-level data both on income and expenditure. So, for instance, if one were considering replacing a wage tax with a higher cigarette excise tax, it would then be straightforward to compute the effect on the real income of every household in the data sample, and hence trace the effect on the distribution of expenditure (or income). The data that we use for this study come mainly from the Vietnam Household Living Standards Survey of 2006 (VHLSS-2006). In the course of the survey, 9,189 households were interviewed; of these, 4,298 were randomly (in clusters) from the 2004 round of the VHLSS (in effect creating a data panel), while the remainder were chosen randomly using a process of stratified cluster sampling. Sampling weights have been created for the survey, and have to be used in all computations. The actual surveying was done between May and November, with 93 percent of all households interviewed in June, September, or October. Given the length of the questionnaire, every household was visited twice; the information on income and spending is based on the recall of respondents. The survey was administered by the Social and Environmental Statistics Department of the General Statistics Office, which 219 TA X R E F O R M I N V I E T N A M has considerable experience in collecting good quality survey data. The response rate to the survey is believed to be over 90 percent, even in urban areas, but detailed information on this aspect of the survey was not collected. Table 8.3 provides a breakdown of the sources of income by income per capita deciles. The procedure that we followed was first to create a measure of income based on the survey data. Most of the sources of income are straightforward; however, the methods used to measure income from durable goods housing, merit a few further comments. Although there are a few renters, most Vietnamese households own their homes. This should be seen as an asset that generates income; the homeowner is paying rent to himself, and the imputed value of the services of the house—which we took to be 2 percent of the reported value—is included both as expenditure and income. The inclusion of imputed rent in household income has been recommended by the United Nations since 1977 (Yates 1994); the need to include it as a component of household expenditure is also widely recognized (Deaton and Zaidi 2002). Similar reasoning applies to durable goods. A motorbike, for instance, is used up (“consumedâ€?) over a number of years, and so rather than include the full purchase price as part of expenditure in the year in which it was bought, it is more appropriate to spread the value of the consumption over the expected useful life of the asset. Thus, for each major asset we compute average depreciation rates from the available survey data, adjust for inflation, and assume a 3 percent real interest rate, to arrive at the rental equivalent of durable goods. It might be argued that this rental equivalent should also be counted as part of the household’s income—it is paying the “rentalâ€? of the motorbike to itself—but this is not quite so: the income is the rental receipt less the depreciation of the asset, so only the interest component of the rental should be counted in income. By analogy, if the money had been put into an interest-bearing bank account, we would count that interest in income. Our results are not particularly sensitive to the choice of parameters here; if housing services are imputed using a rate of 3 percent instead of 2 percent, and if durable goods use an interest rate of 5 percent rather than 3 percent, then total expenditure would be about 8 percent higher and measured income would be 10 percent higher, and both would be distributed somewhat more unequally. Since prices vary by region and by month of the year, we adjusted income using price deflators supplied by the General Statistics Office (GSO). Unless otherwise indicated, all numbers related to the survey are expressed in average prices for 2006; they thus can be expected to differ slightly from the GSO figures, which are expressed in the prices of January 2006. The price-adjusted measures of household income were then divided by household size to obtain income per capita, and households were then sorted from poorest to richest into ten deciles, each of which contains an equal number of individuals. It is not surprising that as one household moves from low- to high-income household, the share of income from agriculture, livestock, and forestry falls, from 53 percent in the bottom decile to 10 percent in the top decile. This pattern has a simple corollary: taxes on the agricultural sector will tend to fall more heavily on poorer households. 220 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Table 8.3: Sources of Household Income for Vietnam Households, by Income Per Capita Decile, 2006 Income per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All percentage of total income from each source of income Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 of which Wages and salaries 23.8 27.4 27.6 28.0 29.3 28.5 30.8 30.3 30.6 27.4 28.4 Agriculture 36.0 31.3 28.0 23.6 20.5 19.3 15.3 14.4 10.1 7.2 20.6 Livestock 9.1 9.0 8.0 7.9 7.1 6.0 5.3 4.3 3.2 2.2 6.2 Forestry 7.5 3.1 2.3 1.5 1.2 0.8 0.6 0.4 0.2 0.4 1.8 Aquaculture 2.0 3.2 3.2 4.0 4.1 3.0 3.2 2.4 2.9 3.2 3.1 Household enterprises 3.6 7.7 11.3 14.4 15.7 18.9 20.3 20.6 20.3 21.4 15.4 Durable goods 0.8 0.9 1.2 1.1 1.2 1.2 1.4 1.4 1.4 1.5 1.2 Housing 5.0 5.5 6.0 6.5 7.1 7.8 8.6 10.1 13.0 15.8 8.5 Subsidies, ed & hlth 1.8 1.3 1.2 1.3 1.0 1.1 0.9 0.9 0.9 0.9 1.1 Other income 10.5 10.6 11.3 11.7 12.7 13.2 13.4 15.0 17.3 18.4 13.7 of which: foreign remittances 0.2 0.4 0.7 0.8 1.0 1.0 1.6 2.0 3.9 5.8 1.7 domestic remittances 6.4 6.7 6.3 6.3 6.8 7.0 6.2 6.0 5.4 5.8 6.3 thousands of VND per year Memo items: Income/capita/yr 2,370 3,511 4,387 5,279 6,268 7,508 9,076 11,118 14,660 29,343 9,351 Income/household/yr 13,417 18,234 22,193 25,981 29,885 36,739 43,756 50,816 65,663 120,728 42,737 Expenditure/capita/yr 2,394 3,254 3,839 4,599 5,319 6,178 7,470 8,852 11,603 20,670 7,417 Expenditure/household/yr 13,221 16,519 18,997 22,202 24,870 30,012 35,683 40,030 51,643 83,409 33,656 Notes: Agriculture includes agricultural services; forestry includes hunting; "other income" includes rent, remittances, interest, and transfer payments such as pensions. The average exchange rate in 2006 was VND15,994/US$. Notes: Agriculture includes agricultural services; forestry includes hunting; “other incomeâ€? includes rent, remittances, interest, and transfer payments such as pensions. The average exchange rate in 2006 was VND15,994/US$. More unexpectedly, the importance of wages and salaries in total income does not vary systematically with per capita income. This is in contrast with income from household businesses, which constitutes just 4 percent of the income of those in the poorest decile but a fifth of the income for those in the top four deciles. A flat tax on household business income, such as the corporate income tax, is therefore likely to be progressive, in the sense of representing a proportionately greater tax burden on rich households than on poor. A breakdown of household expenditure is shown in table 8.4. In this case, households are sorted by real expenditure per capita—that is, expenditure deflated for differences in the prices of main commodities across regions and over months of the year. Engel’s Law, which states that as real incomes rise, the share devoted to buying food falls, is clearly evident: the poorest households devote over three fifths of their spending to food, compared with about a quarter for those in the top quintile. The other striking differences in the consumption patterns of the poor and the rich is that more affluent households spend relatively more heavily on housing, utilities, and durable goods; thus, flat taxes on these groups of goods and services will be inherently progressive in nature. 221 TA X R E F O R M I N V I E T N A M Table 8.4: Breakdown of Household Expenditure by Vietnamese Households, by Expenditure Per Capita Deciles, 2006 Expenditure per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % of total expenditure devoted to each expenditure heading Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 of which: Food 60.8 55.2 51.7 47.7 46.4 43.0 40.4 37.8 33.4 26.9 44.3 Non-food items 21.6 22.7 23.2 24.1 24.4 25.2 25.9 26.5 25.8 26.2 24.6 Durable goods 3.9 5.6 6.8 8.0 8.3 9.5 9.6 9.9 11.7 12.3 8.6 Housing 5.2 6.0 7.0 7.9 8.6 9.1 10.6 12.2 16.3 22.7 10.6 Utilities 1.8 2.1 2.3 2.3 2.3 2.4 2.5 2.5 2.9 3.0 2.4 Health 3.3 4.0 4.6 4.9 5.1 5.4 5.5 5.8 5.0 4.4 4.8 Education 3.4 4.5 4.4 5.1 4.9 5.3 5.4 5.2 5.0 4.4 4.8 Also: Home production 34.1 24.7 19.7 16.6 13.5 10.1 7.4 5.1 2.8 1.4 13.5 Purchases 65.9 75.3 80.3 83.4 86.5 89.9 92.6 94.9 97.2 98.6 86.5 thousands of VND per year Memo items: Income/capita/yr 2,996 3,949 5,015 6,179 7,058 8,413 10,086 12,133 15,933 27,939 9,351 Income/household/yr 17,957 20,445 25,109 30,793 33,917 39,361 48,935 55,507 71,528 111,603 42,737 Expenditure/capita/yr 2,000 2,938 3,640 4,324 5,077 5,981 7,149 8,811 11,593 22,681 7,417 Expenditure/household/yr 11,839 15,233 18,208 21,410 24,397 28,054 34,336 40,437 52,104 90,624 33,656 Notes: Non-food items include daily non-food purchases, annual purchases, and "other household expenditure". Durable goods refer to the rental value of durable goods. Housing measures actual rent paid, or imputed rent (in the case of owners), plus home repairs. Utilities comprise spending on water, electricity, and garbage collection. Average exchange rate in 2006 was VND15,994/US$. For completeness, a breakdown of income and expenditure by region is shown in table 8.5, along with a comparison between urban and rural areas. Region 1 is the Red River Delta, centered on Hanoi, while Region 7 is the Southeast, centered on Ho Chi Minh City. These two relatively affluent areas stand out: there is relatively less spending on food, and more on housing; and income is more likely to come from household business and less likely to originate in agriculture. The urban-rural split shows a similar pattern. The Central Highlands (region 6) is noteworthy in that income is heavily dependent on agriculture, but unlike the other poor hilly areas (regions 2 and 3 in the northeast and northwest), households in region 6 do not rely much on home-produced food. This reflects the importance of cash cropping, particularly coffee, in the Central Highlands. 222 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Table 8.5: Breakdown of Household Expenditure and Income, by Region and Urban Rural, 2006 Regions 1 2 3 4 5 6 7 8 Urban Rural All % of total expenditure devoted to each expenditure heading Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 of which: Food 40.5 49.3 54.9 46.5 44.0 47.0 37.8 47.0 36.5 47.2 44.3 Non-food items 22.0 25.1 23.7 23.0 24.3 25.0 25.2 27.5 23.7 24.9 24.6 Durable goods 9.6 7.9 6.9 8.6 9.6 8.3 8.7 7.6 10.1 8.0 8.6 Housing 14.4 7.5 7.1 8.2 9.1 7.5 16.1 7.4 17.3 8.1 10.6 Utilities 2.7 1.9 1.3 2.2 2.4 1.8 3.1 2.3 3.3 2.1 2.4 Health 5.7 4.4 2.9 6.6 5.9 5.3 4.5 2.8 5.0 4.7 4.8 Education 5.0 3.9 3.1 5.0 4.6 5.1 4.3 5.5 4.2 5.0 4.8 Also: Home production 13.3 26.4 36.0 18.9 10.3 14.1 3.5 8.5 2.7 17.5 13.5 Purchases 86.7 73.6 64.0 81.1 89.7 85.9 96.5 91.5 97.3 82.5 86.5 percentage of total income from each source of income Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 of which Wages and salaries 29.0 23.0 17.8 21.9 34.0 23.1 36.9 29.1 36.1 25.5 28.4 Agriculture 16.4 25.6 39.3 23.3 15.3 41.4 10.6 21.3 5.0 26.2 20.6 Livestock 6.8 10.7 9.2 9.3 5.6 5.6 3.0 3.4 1.5 7.9 6.2 Forestry 0.0 5.9 8.4 3.0 1.8 2.0 0.3 0.6 0.2 2.4 1.8 Aquaculture 1.6 1.9 2.0 3.3 3.1 0.6 1.5 7.5 1.2 3.8 3.1 Household enterprises 16.4 10.7 6.7 12.8 17.9 10.8 19.7 17.1 23.9 12.3 15.4 Durable goods 1.3 1.1 1.1 1.3 1.4 1.1 1.2 1.1 1.4 1.1 1.2 Housing 12.0 5.9 5.7 6.6 7.5 5.9 13.1 5.7 14.5 6.4 8.5 Subsidies, ed & hlth 1.2 1.1 1.2 1.3 1.4 1.4 0.9 0.9 1.0 1.2 1.1 Other income 15.2 14.0 8.7 17.1 12.0 8.0 13.0 13.4 15.3 13.1 13.7 of which: foreign remittances 1.3 1.2 0.3 1.9 1.7 0.8 2.9 2.1 2.4 1.5 1.7 domestic remittances 6.2 5.3 4.0 7.3 6.1 4.4 6.2 7.4 5.6 6.6 6.3 thousands of VND per year Memo items: Income/capita/yr 9,951 7,916 5,678 6,729 8,396 8,113 13,882 9,008 14,377 7,518 9,351 Expenditure/capita/yr 8,241 5,772 4,393 5,053 6,766 5,955 11,852 6,698 12,394 5,602 7,417 Notes: Non-food items include daily non-food purchases, annual purchases, and "other household expenditure". Durable goods refer to the rental value of durable goods (for expenditure) or the interest on durable goods (for income). Housing measures actual rent paid, or imputed rent (in the case of owners), plus home repairs. Utilities comprise spending on water, electricity, and garbage collection. Average exchange rate in 2006 was VND15,994/US$. Agriculture includes agricultural Note: The regions are: 1: Red River Delta, 2: Northeast, 3: Northwest, 4: North Central Coast, 5: South Central Coast, 6: Central Highlands, 7: Southeast, 8: Mekong River Delta. A technical point is worth noting at this point. The measures of expenditure used in this study were built using the detailed data of the VHLSS-2006. This was necessary in order to be able to calculate the incidence of taxes, where one often has to apply different tax rates to detailed expenditure (or income) data. But there are some differences between our measures, and the summary measures produced by the GSO, as set out in table 8.6. Our measures of nonfood spending, the services of durable goods, and housing are all substantially larger than the GSO numbers, and the difference is too large to be due to the increase in consumer prices between January 2006 (the GSO price base period) and the average for 2006 (our base price period). These discrepancies are probably mainly due to differences in the assumptions used to impute the values of housing and durable goods services; clearly even the esoteric assumptions about interest rates and depreciation play a role here. However, the net effect of these differences on tax incidence is relatively modest. 223 TA X R E F O R M I N V I E T N A M Table 8.6: Spending Comparison (in VND ’000 per Household per Year, 2006) Current Study GSO Data (prices of 2006) (prices of January 2006) Food 11,563 11,341 Nonfood 7,972 6,427 Durables: services 3,240 1,638 Housing 4,735 3,410 House repair 107 0 Water 118 118 Electricity 693 693 Garbage 23 23 Health 1,488 1,478 Education 1,479 1,486 Insurance 17 0 Income tax 26 0 Total 31,460 26,610 4. Tax Incidence The essential results on tax incidence are shown in tables 8.7 and 8.8, and the associated figures 8.1 and 8.2. As explained above, we have assumed that the ultimate weight of all these taxes falls on households. For the value-added tax, excise taxes, and personal income tax, the amounts reported in tables 8.7 and 8.8 are obtained by multiplying the tax rates by the relevant bases. Thus these are measures of the hypothetical amount of tax paid; the actual amount of tax collected may be higher (if households underreport income or spending) or lower (if there is tax evasion). On the other hand, the amounts paid in fees and in household enterprise taxes are the actual values reported by households in the VHLSS-2006. The difference between table 8.7 and table 8.8 is that the former ranks households by real expenditure per capita, while table 8.8 sorts households by real income per capita. It turns out that in the case of Vietnam it does not make a lot of difference which one chooses, a reflection of the relatively close correlation between household income and expenditure in the country; this is in contrast with what has been observed in Latin America or the United States, where the pattern of tax incidence can look very different depending on whether households are ranked by income or expenditure (see Haughton, Quan, and Bao 2006 for an example). The most important finding is that the tax system is progressive overall: more affluent households, whether measured by income or by expenditure, generally face a proportionately higher tax burden than poorer households. This may also be seen at a glance in the top left panels of figures 8.1 and 8.2; as one moves from left to right, equivalent to moving from poor to rich, the proportion of spending (or income) going to pay taxes generally rises. This pattern differs somewhat from that found by Haughton, Quan, and Bao (2006) for 1998 in two respects: the tax burden is now heavier, and it is more progressive. The earlier study 224 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? found that taxes represented about 8 percent of spending for all but the top quintile, where it stood at just over 10 percent. The numbers for 2006 show a tax burden of about 8 percent only for the bottom two deciles, rising steadily to reach a plateau of about 14 percent in the four top deciles. The higher, more progressive, tax structure is likely due in large part to the introduction of the VAT in 1999, when it replaced a complex system of turnover taxes. Of the taxes considered in tables 8.7 and 8.8, by far the largest is the VAT. This tax is generally progressive, representing 4.5 percent of the expenditure of households in the bottom expenditure decile, compared to 7.0 percent for those in the top decile. The main reason appears to be that richer households are more likely to buy the goods and services they need— incurring VAT in the process—in contrast to poor households; in the poorest decile, as much as a third of household spending is on home production, which is untaxed, compared to little more than 1 percent for those in the top decile. On the other hand, the VAT rate drops off in the highest decile, due mainly to the relatively high savings rate for this group, which means that a substantial fraction of income escapes all taxes on consumption spending. Although VAT is not levied on unprocessed agricultural products, it is applied to some inputs such as fertilizers, and to processed products; we were able to take such details into account by using the disaggregated expenditure data from the VHLSS survey. The excise tax is also progressive, and this is also somewhat surprising, as Haughton, Quan, and Bao (2006) found this tax to be approximately proportional in 1998. The main explanation seems to be the rise in spending on transportation, most notably motorbikes and the gasoline required to run them, both of which are subject to excise taxes and are disproportionately owned by richer households. For instance, although 59 percent of households owned at least one motorbike in 2006, the proportion varied from 21 percent in the poorest (expenditure) decile to 93 percent in the top decile. Households in the top tenth of the expenditure distribution owned an average of 1.7 motorbikes, compared to 0.22 per household in the poorest tenth. Not all taxes are progressive. Agricultural fees, while modest overall, are relatively burdensome for the lower-middle part of the expenditure distribution. Educational fees exceed 1 percent of spending for all groups, but come close to 2 percent of spending for those in the middle- upper expenditure brackets. 225 TA X R E F O R M I N V I E T N A M Table 8.7: Tax Paid by Vietnamese Households in 2006, by Expenditure Per Capita Deciles Table xxx. Tax paid by Vietnamese households in 2006, by expenditure per capita quintile Expenditure per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % hh paying thousands of VND p.a. Household expenditure per capita 2,000 2,938 3,640 4,324 5,077 5,981 7,149 8,811 11,593 22,681 7,417 Total tax paid 156 252 372 458 557 709 984 1,151 1,679 3,146 946 100.0 of which: value-added tax 90 144 212 263 319 420 525 662 1,007 1,577 522 100.0 excise taxes 13 23 30 38 47 60 73 93 141 333 85 97.3 educational fees 22 37 58 73 88 112 130 168 191 264 114 62.1 agricultural fees 12 22 31 37 33 34 28 28 23 12 26 40.2 other fees 16 22 29 34 41 48 65 83 133 247 72 99.0 taxes on household enterprises 3 3 12 12 30 35 163 116 184 651 121 15.4 personal income tax - - - - - - - - 1 61 6 0.2 as percentage of expenditure Total tax paid 7.8 8.6 10.2 10.6 11.0 11.9 13.8 13.1 14.5 13.9 13.4 of which: value-added tax 4.5 4.9 5.8 6.1 6.3 7.0 7.3 7.5 8.7 7.0 7.0 excise taxes 0.6 0.8 0.8 0.9 0.9 1.0 1.0 1.1 1.2 1.5 1.1 educational fees 1.1 1.3 1.6 1.7 1.7 1.9 1.8 1.9 1.6 1.2 1.5 agricultural fees 0.6 0.8 0.9 0.9 0.7 0.6 0.4 0.3 0.2 0.1 0.4 other fees 0.8 0.8 0.8 0.8 0.8 0.8 0.9 0.9 1.1 1.1 1.0 taxes on household enterprises 0.1 0.1 0.3 0.3 0.6 0.6 2.3 1.3 1.6 2.9 1.6 personal income tax - - - - - - - - 0.0 0.3 0.1 Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Results are preliminary, and based on the VHLSS survey of 2006. All figures are in prices of 2006. Table 8.8: Tax Paid by Vietnamese Households in 2006, by Income Per Capita Deciles Income per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % hh paying Income per capita 2,370 3,511 4,387 5,279 6,268 7,508 9,076 11,118 14,660 29,343 9,351 Total tax paid 174 290 372 460 558 684 888 1,098 1,528 3,409 946 100.0 of which: value-added tax 96 153 201 257 309 376 494 627 886 1,818 522 100.0 excise taxes 15 25 33 40 48 57 83 100 130 319 85 97.3 educational fees 30 54 68 84 99 113 130 162 185 217 114 62.1 agricultural fees 14 26 29 29 29 31 29 26 22 26 26 40.2 other fees 18 25 29 33 44 54 69 90 137 220 72 99.0 taxes on household enterprises 1 6 11 17 28 54 83 93 167 749 121 15.4 personal income tax - - - - - - - - - 62 6 0.2 as percentage of income Total tax paid 7.4 8.3 8.5 8.7 8.9 9.1 9.8 9.9 10.4 11.6 10.1 of which: value-added tax 4.0 4.4 4.6 4.9 4.9 5.0 5.4 5.6 6.0 6.2 5.6 excise taxes 0.7 0.7 0.8 0.8 0.8 0.8 0.9 0.9 0.9 1.1 0.9 educational fees 1.3 1.5 1.6 1.6 1.6 1.5 1.4 1.5 1.3 0.7 1.2 agricultural fees 0.6 0.7 0.7 0.6 0.5 0.4 0.3 0.2 0.2 0.1 0.3 other fees 0.8 0.7 0.7 0.6 0.7 0.7 0.8 0.8 0.9 0.7 0.8 taxes on household enterprises 0.0 0.2 0.3 0.3 0.4 0.7 0.9 0.8 1.1 2.6 1.3 personal income tax - - - - - - - - - 0.2 0.1 Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Results are preliminary, and based on the VHLSS survey of 2006. All figures are in prices of 2006. The two direct taxes are highly progressive, as the bottom panels of figures 8.1 and 8.2 shows. The tax on business income mainly hits the top four deciles; and the personal income tax is borne almost entirely by households in the top decile. A word of caution is in order here; 226 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? despite a sample of 9,189 households, only 14 people reported paying personal income tax. This represents just 0.15 percent of the sample. Figure 8.1: Incidence of Taxes by Expenditure Per Capita Deciles in Vietnam, 2006 Tax as a % of expenditure, by decile, 2006 VAT as a % of expenditure, by decile, 2006 Taxes paid as % of expenditure Taxes paid as % of expenditure 16.0 10.0 14.0 9.0 12.0 8.0 7.0 10.0 6.0 8.0 5.0 6.0 4.0 4.0 3.0 2.0 2.0 1.0 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Expenditure per capita decile (poor = 1, rich = 10) Expenditure per capita decile (poor = 1, rich = 10) Overall VAT Excise Tax as a % of expenditure, by decile, 2006 Educational Fees as a % of expenditure, by decile, 2006 Taxes paid as % of expenditure 1.6 2.5 1.4 Taxes paid as % of expenditure 2.0 1.2 1.0 1.5 0.8 0.6 1.0 0.4 0.5 0.2 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Expenditure per capita decile (poor = 1, rich = 10) Expenditure per capita decile (poor = 1, rich = 10) Excises Educational fees Agricultural Fees as a % of expenditure, by decile, 2006 Other Fees as a % of expenditure, by decile, 2006 Taxes paid as % of expenditure Taxes paid as % of expenditure 1.0 1.4 0.9 1.2 0.8 0.7 1.0 0.6 0.8 0.5 0.4 0.6 0.3 0.4 0.2 0.2 0.1 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Expenditure per capita decile (poor = 1, rich = 10) Expenditure per capita decile (poor = 1, rich = 10) Agricultural fees Other fees Tax on business income as a % of expenditure, by decile, 2006 Personal Income Tax as a % of expenditure, by decile, 2006 Taxes paid as % of expenditure 3.5 Taxes paid as % of expenditure 0.3 3.0 0.3 2.5 0.2 2.0 1.5 0.2 1.0 0.1 0.5 0.1 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Expenditure per capita decile (poor = 1, rich = 10) Expenditure per capita decile (poor = 1, rich = 10) Business income Personal Income Tax 227 TA X R E F O R M I N V I E T N A M Figure 8.2: Incidence of Taxes by Expenditure Per Capita Deciles in Vietnam, 2006 Tax as a % of income, by decile, 2006 VAT as a % of income, by decile, 2006 Taxes paid as % of income 14.0 7.0 Taxes paid as % of income 12.0 6.0 10.0 5.0 8.0 4.0 6.0 3.0 4.0 2.0 2.0 1.0 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Income per capita decile (poor = 1, rich = 10) Income per capita decile (poor = 1, rich = 10) Overall VAT Excise tax as a % of income, by decile, 2006 Educational Fees as a % of income, by decile, 2006 Taxes paid as % of income Taxes paid as % of income 1.2 1.8 1.0 1.6 1.4 0.8 1.2 1.0 0.6 0.8 0.4 0.6 0.2 0.4 0.2 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Income per capita decile (poor = 1, rich = 10) Income per capita decile (poor = 1, rich = 10) Excises Educational fees Agricultural fees as a % of income, by decile, 2006 Other fees as a % of income, by decile, 2006 0.8 1.0 0.7 0.9 Taxes paid as % of income Taxes paid as % of income 0.6 0.8 0.7 0.5 0.6 0.4 0.5 0.3 0.4 0.2 0.3 0.2 0.1 0.1 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Income per capita decile (poor = 1, rich = 10) Income per capita decile (poor = 1, rich = 10) Agricultural fees Other fees Tax on business income as a % of income, by decile, 2006 Personal income tax as a % of income, by decile, 2006 3.0 0.3 Taxes paid as % of income Taxes paid as % of income 2.5 0.2 2.0 0.2 1.5 1.0 0.1 0.5 0.1 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Income per capita decile (poor = 1, rich = 10) Income per capita decile (poor = 1, rich = 10) Business income Personal Income Tax In table 8.9 we summarize the amount of tax paid by region: on average, tax collected comes to 12.8 percent of expenditure; the rate is lower than this in the four northern regions, and higher (or close to) this rate in the four southern regions. In part this reflects the greater affluence of the southern part of the country, although it is noteworthy that the highest tax burden, measured relative to expenditure, is neither the area around Hanoi (Region 1) nor Ho Chi Minh City (Region 7), but rather the South-Central Coast (although the difference may reflect sampling variation—two large household enterprises boosted the numbers for this region substantially). 228 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Table 8.9: Tax Paid by Vietnamese Households in 2006, by Region and Urban/ Rural Residence Tax paid Expenditure Tax as % of Memo: VAT as Income # of households per capita expenditure % of expend. per capita sampled '000 VND '000 VND Region 1 923 8,241 11.2 6.8 9,951 1,944 2 592 5,772 10.3 5.3 7,916 1,317 3 508 4,393 11.6 6.3 5,678 429 4 584 5,053 11.6 5.6 6,729 1,014 5 1,261 6,766 18.6 10.0 8,396 852 6 814 5,955 13.7 7.3 8,113 582 7 1,506 11,852 12.7 6.5 13,882 1,188 8 944 6,698 14.1 8.2 9,008 1,863 Urban 1,685 12,394 13.6 6.7 14,377 2,307 Rural 676 5,602 12.1 7.3 7,518 6,882 Overall 946 7,417 12.8 7.0 9,351 9,189 Note: The regions are: 1: Red River Delta, 2: Northeast, 3: Northwest, 4: North Central Coast, 5: South Central Coast, 6: Central Highlands, 7: Southeast, 8: Mekong River Delta. It is helpful to summarize the progressivity of taxes, and this can be done with the help of table 8.10. The first two rows of that table measure the inequality of per capita expenditure and income respective, using the Gini coefficient. This is obtained by sorting individuals from poorest to richest along a horizontal axis, and graphing their cumulative expenditure (or income) on the vertical axis, to create the Lorenz curve. Denote by A the area between this curve and a line of equality, and by B the area under the Lorenz curve. Then the Gini coefficient is given by A/(A+B); it varies between 0 (perfect equality) and 1 (complete inequality); For expenditure, the Gini coefficient is 0.390, while for income per capita it is 0.403. It is common for income inequality to be greater than expenditure inequality. These numbers reflect moderate inequality. Official calculations show an expenditure (per capita) Gini coefficient of 0.358; the discrepancy appears to be due to differences in the assumptions made when constructing expenditure (see table 8.7), most notably in housing, durable goods, and nonfood spending. The bottom part of table 8.10 reports concentration ratios (or “quasi-Gini coefficientsâ€?) for the various taxes. These are constructed in a manner similar to the Gini coefficient: households are still sorted by per capita expenditure on the horizontal axis, but now the vertical axis graphs the cumulative proportion of tax paid, forming a “quasi-Lorenzâ€? curve. Unlike the Gini coefficient, a concentration ratio could take on a negative number. A high concentration ratio, meaning a value close to one, indicates that tax payments are highly unequal—indeed, they are especially concentrated among higher-income individuals. Such a tax is therefore progressive. Kakwani has suggested that the difference between the concentration ratio and the Gini coefficient serve as a measure of the progressivity of a tax: if tax payments are more unequal than per capita expenditure, then the tax is progressive and the Kakwani measure is positive. 229 TA X R E F O R M I N V I E T N A M The Kakwani measure for all the taxes we have considered here is 0.071, which denotes a moderate degree of progressivity. In table 8.10 we display the remaining taxes in order, starting with the most progressive and ending with the most regressive. These numbers confirm the patterns seen in figures 8.1 and 8.2: the income taxes are highly progressive, excises and VAT are somewhat progressive, and educational and (especially) agricultural fees are regressive. Table 8.10: Measures of Tax Progressivity Gini coefficient (GX) Level, VND m p.a. Expenditure per capita 0.390 7,417 Income per capita 0.408 9,351 Memo: Using higher costs for durables and housing: Expenditure per capita 0.406 8,029 Income per capita 0.412 10,275 Concentration Ratio Kakwani Measure Measures relative to expenditure (CT,X) (CT,X – GX) All allocable taxes 0.461 0.071 Personal income tax 0.986 0.597 Business income tax 0.686 0.296 Excise taxes 0.504 0.114 Fees: other 0.459 0.069 VAT 0.443 0.054 Fees: educational 0.354 -0.036 Fees: agricultural -0.018 -0.408 Property tax 0.632 0.242 Property tax: own residence 0.619 0.229 Property tax: other than residence 0.751 0.361 Property tax with VND100m threshold 0.776 0.386 Memo: all allocable taxes, using higher costs 0.462 0.056 for durables and housing (see text) Memo: all allocable taxes, excluding fees 0.495 0.105 Source: Based on Vietnam Household Living Standards Survey 2006. It is not easy to reconcile the tax revenue estimated on the basis of actual or imputed data from the VHLSS-2006, and the data provided by the national budget, but an attempt is made in table 8.11. The first column of numbers shows the estimated tax payments per capita, for the major tax categories. When multiplied by the population, which was 84 million in 2006, one gets the estimated levels of revenue shown in the middle column. These may be compared with the budgetary data in the final column. The VAT numbers based on the VHLSS-2006 data are understated. One plausible reason is that substantial amounts of sales, which are subject to VAT, are made to outsiders, including tourists and foreigners resident in Vietnam—groups that are not covered by the VHLSS. 230 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Another possibility is that households understate their actual spending, which is a common feature of expenditure surveys. Similar reasoning applies to the understatement of excise tax revenues, much of which consists of “sin taxesâ€? on goods such as alcohol and tobacco, notorious for being underreported in household survey data. Based on the VHLSS-2006 numbers, the personal income tax shows a very modest amount of total revenue—just VND500 billion—compared to a budget total of VND5.2 trillion. We argue below that this is just about plausible, and reflects the fact that most of the revenues from this tax come from collections on the wages and salaries of foreigners working in Vietnam. It is also quite possible that the VHLSS-2006 did not adequately capture information on the incomes of the very top earners in Vietnam. Taxes collected from household enterprises contribute an estimated one tenth of total revenue from the corporate income tax, which is plausible given that the latter mainly collects revenue from larger corporations. Perhaps the most interesting entry is the last one, which projects fees to yield a total of VND 17.8 trillion, compared to VND 3.9 trillion in the budget. The explanation is that most of the revenue from fees and contributions goes to local bodies—schools, municipalities, health clinics, and the like—rather than into the central budget. These levies—and they are numerous—lack transparency, and little is known about the extent to which they are true user fees, or how they are determined. Table 8.11: Reconciling Survey with Budget Data Tax revenue, based on VHLSS-2006 Based on budgeted Per capita, Nationally, national receipts, VND VND ‘000 VND trn trn VAT 0.522 43.9 54.8 Excise taxes 0.085 7.1 17.1 Personal income tax 0.006 0.5 5.2 Business profits tax 0.121 10.2 100.8 Fees 0.212 17.8 3.9 Source: VHLSS-2006 data from table 8.7; budgetary information from table 8.1. Revising the Personal Income Tax In December 2007, after a period of vigorous debate, the National Assembly approved a new Personal Income Tax Law, came into effect on January 1, 2009. Compared with the previous personal income tax, the new version is levied on a broader base—it is designed to include most forms of income, and not just labor compensation—it applies a single rate structure to foreigners and Vietnamese, and it has seven rather than four rates, the highest of which is 35 percent instead of the previous 40 percent. Before applying the tax rates, the taxpayer may deduct a personal allowance of VND 48 million, plus a further VND 19.2 million for each bona fide dependent. These details are summarized in table 8.12. One important feature of the new arrangement is that household business income, which previously was taxed at a flat rate of 28 percent, is now taxed under the personal income tax. 231 TA X R E F O R M I N V I E T N A M Table 8.12: Structure of Current and Previous Personal Income Taxes Previous Tax Current/New Tax Coverage Wages, salaries, bonuses, housing Wages, salaries, bonuses, housing allowances allowances, business income, interest, dividends, capital gains, prizes Rates & Base For Vietnamese Income p.a., VND m Tax rate, % Taxable income p.a., Tax rate, % VND m 0–60 0 0 – 60 5 60–180 10 60 – 120 10 180–300 20 120 – 216 15 300–480 30 216 – 384 20 >480 40 384 – 624 25 For foreigners 624 – 960 30 0–96 0 >960 35 96–240 10 240–600 20 Personal deduction: VND 48 m 600–960 30 Deduction/dependent: VND 19.2 m > 960 40 Sources: IMF 2007; Shukla 2006; Law on Personal Income Tax /2007/QH12. We would like to know what effect the new income tax would have on revenue and on incidence, and we approach this problem in two ways. First, we use the data from the VHLSS-2006: we strip out the business income tax and current personal income tax, and then apply the rules of the new tax. The results are shown in table 8.13, where we see that the estimated revenue would fall from VND 127,000 per person to just VND 21,000 per person. The new tax would fall almost entirely on those in the top decile, and so would be highly progressive. However, this cannot be the final word on the matter, if only because two thirds of personal income tax revenue is collected from foreigners, and the effect on them of the new tax structure is not included here. Some further insight into the problem may be gained by turning to data from another source. In late 2005, the General Department of Taxation undertook a survey of “high income individuals.â€? The sample of 15,500, representing about 5 percent of its tax rolls, included 3,200 foreign taxpayers, 7,200 Vietnamese paying the personal income tax, and 5,100 Vietnamese paying business income tax. There were 11,532 responses to the survey, representing a 74 percent response rate. The survey used a relatively short questionnaire that asked about income and expenditure, but not about taxes paid. An effort was made to assure respondents that their answers to the questionnaire would not be used for tax collection purposes, but it is not clear how effective this assurance was. Response rates were relatively modest for foreigners, and in Ho Chi Minh City, but were high elsewhere; the results reported below represent weighted averages that adjust for the differential response rates. 232 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Table 8.13: The Distributional Effects of Revising the Personal Income Tax Law Expenditure per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % hh paying thousands of VND p.a. Household expenditure per capita 2,000 2,938 3,640 4,324 5,077 5,981 7,149 8,811 11,593 22,681 7,417 Total tax paid 156 252 372 458 557 709 984 1,151 1,679 3,146 946 100.0 of which: taxes on household enterprises 3 3 12 12 30 35 163 116 184 651 121 15.4 personal income tax - - - - - - - - 1 61 6 0.2 Total new tax paid on income 8 1 13 190 21 1.0 as percentage of expenditure Total tax paid 7.8 8.6 10.2 10.6 11.0 11.9 13.8 13.1 14.5 13.9 13.4 of which: taxes on household enterprises 0.1 0.1 0.3 0.3 0.6 0.6 2.3 1.3 1.6 2.9 1.6 personal income tax - - - - - - - - 0.0 0.3 0.1 Total new tax paid on income - - - - - - 0.1 0.0 0.1 0.8 0.3 Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Based on the VHLSS survey of 2006. All figures are in prices of 2006. Income per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % hh paying Income per capita 2,370 3,511 4,387 5,279 6,268 7,508 9,076 11,118 14,660 29,343 9,351 Total tax paid 174 290 372 460 558 684 888 1,098 1,528 3,409 946 100.0 of which: taxes on household enterprises 1 6 11 17 28 54 83 93 167 749 121 15.4 personal income tax - - - - - - - - - 62 6 0.2 Total new tax paid on income 212 21 1.0 as percentage of expenditure Total tax paid 7.4 8.3 8.5 8.7 8.9 9.1 9.8 9.9 10.4 11.6 10.1 of which: taxes on household enterprises 0.0 0.2 0.3 0.3 0.4 0.7 0.9 0.8 1.1 2.6 1.3 personal income tax - - - - - - - - - 0.2 0.1 Total new tax paid on income - - - - - - - - - 0.7 0.2 Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Based on the VHLSS survey of 2006. All figures are in prices of 2006. The key data are shown in table 8.14, which divides respondents into three groups— Vietnamese and foreigners who pay personal income tax, and Vietnamese who pay business income tax (here labeled “household enterprisesâ€?). The reported annual incomes of those surveyed are relatively high, especially for foreigners, where the total came to VND 632 million (about US$40,000) in 2005. For foreigners, the most important source of income by far was wages and salaries, while other household members added relatively little to total household income. In contrast, the Vietnamese taxpayers relied on income from a greater variety of sources, with large contributions from business income and, in the case of those who paid personal income tax, other family members. The reported levels of spending in table 8.14 are implausibly low; even so, the low proportion of income that expatriates spend inside Vietnam is quite striking, and plausible enough. The Vietnamese taxpayers appear to spend heavily on assets rather than consumption. This is 233 TA X R E F O R M I N V I E T N A M consistent with the idea—which is at the heart of the “permanent income hypothesisâ€?—that many of those who are caught in the tax net in a given year have temporarily high incomes; in order to smooth their consumption patterns, they spend a high proportion of the transitory income on acquiring assets that will allow them to maintain higher levels of consumption into the future. At the bottom of table 8.14 we estimate the amount of revenue that should, in theory, have been collected from these taxpayers in 2005. The personal income tax revenue should have been VND 15.5 trn (= VND 14.849 trn + VND 0.648 trn), when it actually came to VND 4.2 trn. This is evidence that evasion is widespread under the current income tax. The new income tax arrangements came into effect in 2009, at which point incomes and prices (but not the tax brackets) were substantially higher than they were in 2008, or than they were in 2005 at the time of this survey. So in order to simulate the effects of the new income tax regime, we first project income out to 2009 for every taxpayer in the survey, and then apply in turn the current tax rules, and the proposed tax rules. Note that the personal income tax would be very elastic; although nominal GDP rose by an estimated 83 percent between 2005 and 2009, personal income tax from Vietnamese taxpayers was expected to rise by 253 percent over the same period even under current rules, making this tax very elastic (see table 8.14). The most important results are displayed near the bottom of table 8.14, and shows that the theoretical amount of tax revenue collected from these high-income tax payers would rise by 8 percent with the change in the income tax regime. This is an underestimate, because it is computed on the basis of those who paid these taxes in 2005; by 2009 the number of households caught in the tax net would be higher, reflecting growth both in incomes and the total working population. Although total revenue would rise, the amount paid by foreigners would fall by 15 percent while the tax paid by Vietnamese would rise sharply—by 178 percent for personal income taxpayers, and by 37 percent for those who pay household business tax. These higher tax payments result mainly from the provision in the new law that would tax global income; when wages and salaries, capital income, and business income are added together, many households find themselves pushed quickly into relatively high tax brackets. These proportionate changes assume that the degree of tax evasion does not change between 2005 and 2009, despite major alterations in the tax code. 234 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? Table 8.14: Income and Estimated Tax for High-Income Households, 2005 Vietnamese Foreigners Household Full sample % foreign enterprises thousands of VND per year Taxpayer income 195,003 631,835 244,158 302,514 of which: Wages and salaries 620,939 43,753 177,172 Enterprise 498 810 327 507 Business 37,474 1,548 191,095 80,459 Finance and capital 79,732 8,190 7,022 40,869 Transfers 98 5 227 121 Other 5,938 342 1,734 3,386 Memo items: Total household income 370,619 671,127 266,619 399,276 Household current spending 79,273 116,266 39,741 74,014 Household asset purchases 255,995 38,100 59,914 145,971 Personal income tax/taxpayer Estimate, 2005 4,504 239,751 74,315 76,649 Estimate, 2009, old rules 15,907 527,604 136,219 162,466 Memo: % growth 253% 120% 83% 112% Estimate, 2009, new rules 44,283 447,019 186,635 175,300 Personal income tax revenue (VND trn) Estimate, 2005 628 14,849 7,336 22,812 65% Memo: Actual PIT revenue, 2005 4,200 Estimate, 2009, old rules 2,217 32,677 13,446 48,340 68% Estimate, 2009, new rules 6,171 27,686 18,423 52,280 53% Percentage change 178 (15) 37 8 If the simulation based on the VHLSS-2006 data is to be believed, the new income tax regime was expected to reduce tax revenues sharply, by about 5 percent overall. But when we use the information on high-income households, we conclude that the tax change would increase government revenue. An attempt at reconciling these two contradictory conclusions is summarized in table 8.15. Fewer people would pay personal income tax (table 8.13), and the amount paid by high-income taxpayers would fall slightly, essentially because foreigners would pay less (table 14), for a net loss of revenue, had the new system been in place in 2006, of VND 0.5 trillion dong, as shown in the top row in table 8.15. It is more difficult to reconcile the numbers for tax on household enterprise profits. Over half of these profits are collected from households in the top decile, and fewer of these are expected to have to pay the tax under the new integrated income tax regime (table 8.13); business tax revenue from households in the lower deciles will almost evaporate (table 8.13), as they are subject to the more generous personal income tax regime.75 On the other hand, high-income households who pay business income tax would generally have to pay more, as the base on which they have to pay tax is broadened (table 8.14). The net effect of all these changes would be a reduction in tax revenue of VND8.2 billion, or about 3 percent of total tax revenue in 2006. 75 The drop in revenue from taxes on household enterprise income is probably overstated, because some of the taxes reported by households as business income tax may in fact represent other business taxes, which would remain in effect. Unfortunately, we are unable to quantify this effect. 235 TA X R E F O R M I N V I E T N A M Table 8.15: Estimating the Net Effect on Tax Revenue of Personal Income Tax Changes, 2006 Tax Change Tax revenue Unexplained Change Net revenue, based on national tax change based on budgeted revenue in tax VHLSS, national revenue nationally receipts VND trn % VND trn VND trn % VND trn (1) (2) (3) (4) (5) (6) Personal 0.5 -73 5.2 4.7 -3 -0.5 income tax Business 10.2 100.8 90.6 profits tax Of which: Top decile 5.5 -73 2.3 +37 -3.2 Other deciles 4.7 -96 -4.5 Net revenue -8.2 effect Notes: Total revenue in 2006 was VND 264.2 trillion; the net revenue effect represents a 3.1 percent reduction in revenue. (1) from table 8.11, with breakdown by decile based on table 8.13. (2) based on changes in table 8.13. (3) and (4) from table 8.11, with top decile figure based on table 8.14. (5) based on table 8.14. 5. Property Tax Vietnam does not currently have a property tax. Calls for the introduction of such a tax are becoming more common; proponents argue that such a tax would be both efficient (especially the part that falls on land), and equitable. In this section we examine the effects of introducing a property tax on nonagricultural real estate owned by households, levied at a rate of 1 percent of the property value. Our procedure is straightforward: we assume that there would be no evasion, and that the property values reported by households are not systematically over-or understated. Then we apply the tax rate to the values of real estate reported in the VLHSS of 2006, assume that the tax is borne by owners, and derive the incidence numbers shown in tables 8.16 and 8.17. We also compare these results with the case of a property tax that is levied at 1 percent but would exempt the first VND100 million (US$6,250) worth of household property from the tax. The first point to note is that a 1 percent tax on nonagricultural household-owned real estate is a heavy tax. It would represent, on average, 7.8 percent of household expenditure, or 6.2 percent of household income; that is about three-fifths of the total amount currently collected directly and indirectly from households. It also represents a tax of perhaps 20 percent on the income stream from property, assuming the latter yields a return of 5 percent. A tax of this size would meet strong resistance, and so is best thought of as an upper bound to the 236 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? (politically) feasible rate. Moreover, this burden supposes that revenue collection is efficient and that there are no evasions which are rather optimistic assumptions. Table 8.16: The Distributional Effects of Introducing a Tax on Nonagricultural Real Estate Expenditure per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % hh paying thousands of VND p.a. Household expenditure per capita 2,000 2,938 3,640 4,324 5,077 5,981 7,149 8,811 11,593 22,681 7,417 Total tax paid 50 82 119 168 210 277 384 545 1,028 2,944 580 97.9 of which: taxes on residence 49 80 113 157 200 257 357 500 922 2,587 522 97.5 taxes on other property 1 2 5 10 10 20 26 46 106 358 58 8.9 Total tax paid with VND100m threshhold per household 1 4 12 44 62 109 197 335 798 2,677 424 35.8 as percentage of expenditure Total tax paid 2.5 2.8 3.3 3.9 4.1 4.6 5.4 6.2 8.9 13.0 7.8 of which: taxes on residence 2.4 2.7 3.1 3.6 3.9 4.3 5.0 5.7 8.0 11.4 7.0 taxes on other property 0.1 0.1 0.1 0.2 0.2 0.3 0.4 0.5 0.9 1.6 0.8 Total tax paid with VND100m threshhold per household 0.1 0.1 0.3 1.0 1.2 1.8 2.8 3.8 6.9 11.8 5.7 Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Based on the VHLSS survey of 2006. All figures are in prices of 2006. Income per capita deciles 1 (poor) 2 3 4 5 6 7 8 9 10 (rich) All % hh paying Income per capita 2,370 3,511 4,387 5,279 6,268 7,508 9,076 11,118 14,660 29,343 9,351 Total tax paid 56 96 132 178 230 315 430 626 1,038 2,702 580 97.9 of which: taxes on residence 54 92 126 166 219 298 404 562 955 2,343 522 97.5 taxes on other property 2 5 6 12 10 17 26 64 83 359 58 8.9 Total tax paid with VND100m threshhold per household 2 9 23 51 81 149 247 419 813 2,443 424 35.8 as percentage of expenditure Total tax paid 2.4 2.7 3.0 3.4 3.7 4.2 4.7 5.6 7.1 9.2 6.2 of which: taxes on residence 2.3 2.6 2.9 3.2 3.5 4.0 4.5 5.1 6.5 8.0 5.6 taxes on other property 0.1 0.1 0.1 0.2 0.2 0.2 0.3 0.6 0.6 1.2 0.6 Total tax paid with VND100m threshhold per household 0.1 0.3 0.5 1.0 1.3 2.0 2.7 3.8 5.5 8.3 4.5 Notes. Revenue from value-added tax, excise taxes, and personal income tax, are estimated based on spending and income patterns. Based on the VHLSS survey of 2006. All figures are in prices of 2006. A property tax would be highly progressive, representing a burden of less than 3 percent of income or expenditure for the poorest fifth of the population, but over 7 percent for the richest fifth. Nine-tenths of the revenue would come from taxing principal residences, with the remainder coming from property that households rent out to others. A property tax that is only levied above a threshold would be even more progressive, collecting just 0.1 percent of spending from the poorest decile and 11.8 percent from the top decile. The burden of the tax would be four times as high (as a percentage of household spending) 237 TA X R E F O R M I N V I E T N A M Table 8.17: Tax Paid by Vietnamese Households, by Region and Urban/Rural Residence, Based on a 1 Percent Property Tax Tax paid Expenditure Tax as % of Tax paid with Tax as % of # of households per capita expenditure threshhold expenditure sampled '000 VND '000 VND Region 1 791 8,241 9.6 599.6 7.3 1,944 2 305 5,772 5.3 172.6 3.0 1,317 3 244 4,393 5.5 140.8 3.2 429 4 248 5,053 4.9 119.5 2.4 1,014 5 387 6,766 5.7 238.1 3.5 852 6 312 5,955 5.2 190.4 3.2 582 7 1,422 11,852 12.0 1,224.0 10.3 1,188 8 277 6,698 4.1 135.7 2.0 1,863 Urban 1,425 12,394 11.5 1,211.6 9.8 2,307 Rural 272 5,602 4.9 136.2 2.4 6,882 Overall 580 7,417 7.8 423.6 5.7 9,189 Note: The regions are: 1: Red River Delta, 2: Northeast, 3: Northwest, 4: North Central Coast, 5: South Central Coast, 6: Central Highlands, 7: Southeast, 8: Mekong River Delta. This analysis has a number of limitations, due largely to data imperfections. Because it is based on household survey data, it does not include the value of property held by corporations or the government. It assumes that property values in 2006 were realistic, correctly reported by respondents, and would not fall if a property tax were introduced. This last assumption is not particularly compelling; with the introduction of a property tax we would expect the value of land to fall by the capitalized value of the stream of expected property tax payments. In response to a 1 percent property tax, property values could easily fall by 10 percent. Nonetheless, the results appear robust enough to have confidence in the conclusion that such a tax would be progressive, and could have very substantial yield. 6. Checking for Robustness It is useful to ask how robust our results are to some of the more important, or controversial, underlying assumptions. In figure 8.3, the top left panel shows the distribution of taxes relative to expenditure in the baseline case (left-hand bars), and under the assumption that fees are really payments for services and not taxes in disguise (right-hand bars). The exclusion of fees reduces the measured tax burden of course, but it has a particularly marked effect for poorer households; the concentration coefficient for taxes net of fees is 0.495, compared to 0.461 when fees are viewed as taxes. In short, when fees are excluded, the tax system looks more progressive. The top right panel of figure 8.3 looks at the distribution of taxes in the baseline case (left), and then on the assumption that housing services should be valued at 3 percent (rather than 2 percent) of the capital value, and the real interest rate used in valuing the services of durable goods should be 5 percent (rather than 3 percent). These alternative assumptions increase measured expenditure by about 8 percent, and so reduce the observed overall burden of taxation. The 238 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? net effect is a modest reduction in measured tax progressivity: the Kakwani measure falls from 0.071 in the baseline case to 0.056 with the new assumptions (see table 8.10). Figure 8.3: Checking the Robustness of the Incidence Results 16.0 16.0 14.0 Using higher cost of 14.0 Fees as payments for services durables & housing Taxes paid as % of expenditure Taxes paid as % of expenditure 12.0 12.0 Fees as taxes Baseline 10.0 10.0 8.0 8.0 6.0 6.0 4.0 4.0 2.0 2.0 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Expenditure per capita decile (poor = 1, rich = 10) Expenditure per capita decile (poor = 1, rich = 10) Distribution of taxes for baseline (le ) and Distribution of taxes for baseline (le ) and if if fees are considered as payments for measure of expenditure uses higher costs for services (right) durables and housing (right) 16.0 20.0 Double tax on alcohol & 18.0 14.0 tobacco 16.0 Taxes paid as % of expenditure Taxes paid as % of expenditure 12.0 Uniform VAT at 10% 14.0 Baseline Baseline 10.0 12.0 8.0 10.0 8.0 6.0 6.0 4.0 4.0 2.0 2.0 10 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 Expenditure per capita decile (poor = 1, rich = 10) Expenditure per capita decile (poor = 1, rich = 10) Distribution of taxes for baseline (le ) Distribution of taxes for baseline (le ) and if taxes on alcohol and tobacco and if VAT is levied at a uniform are doubled (right) 10 percent (right) One can also explore the effects of other tax changes; the lower left panel of figure 8.3 shows the effect of doubling excise taxes on alcohol and tobacco, and leads one to conclude that the effects on progressivity would be very modest – although the change may be understated here, since spending on alcohol and tobacco are typically substantially underreported in household surveys. Finally, in the lower right panel of figure 8.3 we show what would happen if the VAT were levied at a single rate of 10 percent (while leaving zero-rated and exempt goods unchanged). This is a rough and ready estimate that does not account for VAT that is partly collected on intermediate goods only, but is nonetheless a reasonable first pass at the problem. The overall burden of taxation would rise, but progressivity would essentially remain unchanged – the Kakwani measure would only change from 0.071 to 0.072 (which, if anything, represents greater progressivity, relative to expenditure). At least in the context of current-day Vietnam, a higher VAT rate would not disproportionately burden the poor. 7. An Assessment of Incidence and Its Implications Fairness, whether measured by horizontal or vertical equity, is only one dimension of a good tax system. Taxes must also raise revenue, and do so efficiently, with minimal distortions and low administrative and compliance costs. Even so, the measures of tax incidence that underpin discussions of fairness are important, because they are central to the ethical and political debates about tax policy. This paper provides some of the information required for those discussions. 239 TA X R E F O R M I N V I E T N A M The key finding is that the Vietnamese tax system is relatively progressive, in that it imposes a higher relative burden on the rich than on the poor. However, it must be remembered that only about half of all taxation can be attributed to households; for now, we have not yet allocated to households the taxes that are levied on business corporations—a controversial issue—or the tariffs collected on imports (for which more data would be needed). Nor do we have information on the extent of in-kind contributions make in lieu of taxes that are required of all citizens. And the model we use is a mechanical one; ultimately it would be desirable to build a full computable general equilibrium (CGE) model to track the effects of tax changes on revenue, income distribution, and output (see for instance Shoven and Whalley 1984). The VAT and excise taxes are somewhat progressive. This is not the case in every country, but occurs in Vietnam mainly because poor people provide for many of their own needs rather than purchasing them, which in turn reduces the VAT they have to pay. Over time this will change, as home production becomes unimportant, and gradually this tax will become less progressive. As Vietnam considers its options for raising more tax revenue, changes in rates and base of the VAT are well worth considering: it is administered effectively, raises a lot of revenue, could be changed at minimal expense and, as we have shown, does not unduly burden poorer households. On the other hand, if expanded too much, the VAT too could discourage the development of the formal sector, especially in retailing. In an effort to help the poor, it would be tempting to reduce or even abolish agricultural and educational fees, which are regressive. There is a recent precedent: the unpopular agricultural tax was essentially abolished, which provided substantially more relief to poor than to rich households. However, even if the size of local fees is only loosely related to the services they finance, one should be cautious about reducing important sources of local revenue, especially if the result is diminished local services—poorly maintained irrigation systems, weaker schools, or understocked health clinics. As a general proposition, it would be more desirable to ensure that fees are set so they correspond more closely with the services they pay for. Tax exemptions could be considered for the very poorest households; there is already an infrastructure for determining who is poor in setting educational scholarships. The personal income tax has received more attention over the past year or two than its feeble contribution to state revenue would appear to justify. This still begs the question of what the future direction of public policy should be: should the personal income tax be expanded robustly, or should it remain an unimportant tax. Those who argue for expanding the personal income tax emphasize its revenue potential and its strong progressivity; opponents worry that it will hinder the development of the formal sector, and it is expensive to administer relative to the revenue it yields. The new personal income tax has one major advantage over the current version: it would tax fewer households, essentially because many of the households who currently pay the business income tax— levied at a flat 28 percent once profits exceed the minimum wage—would not have to pay tax when the new personal income tax comes into effect. This should make it more attractive, initially at least, for people to create formal (rather than informal) institutions, such as registered businesses. Our best, although admittedly rough, estimate is that the new income tax, had it been in place in 2006, would have reduced government revenue by 240 TA X AT I O N I N V I E T N A M : W H O PAY S W H AT ? about 3 percent; almost all of this effect would be through the reduction in revenue from the household business income tax. If the personal income tax continues to raise little revenue—and in the Asian countries surveyed by Alm and Wallace (2004), it raised just a sixth of all tax revenue—should it be abolished entirely? Certainly, a case for this can be made, principally on the grounds that expenditure taxation is more efficient that direct taxation (see for instance Tuerck et al. 2006) because it creates fewer distortions in the decisions to work and save. On the other hand, the personal income tax, as currently constituted and as proposed, is the most effective mechanism for ensuring that expatriate workers pay a reasonable amount of tax to Vietnam; that alone would be reason enough to keep the tax. But if this is the main rationale for it, then the tax should continue to apply just to those in the highest earnings brackets. By 2009 the threshold for paying personal income tax will be about four times the level of average income, and the number of people subject to the tax will begin to rise very rapidly. This would probably imply substantial extra collection costs for limited additional revenue, which suggests that at a minimum the tax brackets should be adjusted for inflation over time. It would probably also be desirable to match the top income tax rate with the corporation income tax rate of 28 percent (to ensure that people do not rearrange their business activities simply to avoid taxes), and simplify the structure of the tax by reducing the number of brackets. The implication is that the personal income tax should continue to play a limited role over the coming decade; if more tax revenue is needed, it will be necessary to look elsewhere—to the VAT, a property tax, and perhaps excise taxes. A property tax looks particularly attractive as an option: it would be highly progressive, and would yield substantial revenue. However, the administrative costs of such a tax can be high, especially at first (establishing ownership, creating maps, and the like); and it can create serious burdens on the cash flows of households and businesses, especially in a time of rapidly inflating house prices. Finally, it is worth noting that the discussion in this paper has focused entirely on the revenue side of government. It would be desirable to measure the incidence of government expenditures as well as revenue, as proposed by Demery (2000) and done by Haughton (2005) in the context of Peru. The reason is that even a regressive tax might be desirable if the ensuring spending were returned even more heavily to the poor. 241 Property and Land Taxes By Bruno Moser and Nicolaus Tideman IX 1. Property and Land Taxes: Legal Background and Revenue The basis for holding land in private possession is expressed in the constitution and is further regulated in the land law 2003.76 The drafters of this law made an effort to define, foresee, and regulate every variety of land use (and the amount of money that should be paid to the state for the privilege of each type of use).77 Nevertheless, total revenue from land taxes (apart from royalties from natural resources) was less than one million dollars in fiscal year 2006. However, provinces and municipalities derive significant income from allocating (selling) land use rights. For example, in recent years Ho Chi Minh City has derived more than 40 percent of its annual budget from this source. This practice is not sustainable. A summary of laws related to land taxes is provided below. 1.1. Summary of laws related to land taxes In Vietnam, several taxes are applied to property and land use: the House and Land Tax, the Agricultural Land Use Tax, the Land Use Levy, the Land and Water Surface Rental Charges, and Registration Fees. Until 2009 there was also a Land Use Right Transfer Tax. 1.2. The House and Land Tax This tax is levied on houses and urban land but it has been applied only to land until now. Thus houses are presently not subject to taxation. Thus the base for this tax consists of all residential land and construction land. The tax base is not the market value but rather the plot area, and variable rates are applied according to various categories of land and its location. Some of the main features are given below. The land tax shall not be collected if the land is used for public purposes, social welfare or charity, with no commercial or residential purposes, or if it is used for public worship. When land is leased, the organizations and individuals that lease the land shall not have to pay land tax; the liability will be with the land lessons. Tax base and tax rate The tax calculation is based on land acreage, land grade and the tax rate per acre. 76 VN-Constitution Article 17: The land, forests, rivers and lakes, water supplies, wealth lying underground or coming from the sea, the continental shelf and the air […] come under ownership by the entire people. VN Land Law 2003 Article 1: This Law makes provisions for powers and responsibilities of the State as a representative of the entire people’s ownership of land and exclusive manager of land; the management and use of land and rights and obligations of land users. 77 A summary of land tax related laws was provided by the Ministry of Finance. The economic effects of the various land related taxes and fees are discussed in section [[II and III]]. 243 TA X R E F O R M I N V I E T N A M The tax rate per acre is prescribed for each type of street in each type of urban center. Urban land has been classified into 5 grades, and the rates have been linked to the tax rate for agricultural land use. For instance, for grade-I urban centers (highest grade), the land tax rates have been fixed between 9 and 32 times that on the use of agricultural land of the highest grade in the region, and for grade-V urban centers (lowest grade), the land tax rates are fixed at 5 to 13 times that on the use of agricultural land of the highest grade in the region. For townships, the land tax rates are between 3 and 13 times that on the use of agricultural land of the highest grade in the region. For residential land and construction land in suburban areas, areas adjacent to traffic hubs or along main traffic routes, the tax rates are prescribed as follows. For land in population quarters and construction land in the suburban areas of grade-I urban centers, the land tax rates is 2.5 times that on the use of agricultural land of the highest grade in the communes. For suburban areas of grade-II, -III, -IV, and -V urban centers, the land tax rates is 2 times that on the use of agricultural land of the highest grade in the villages. Finally, for residential land and construction land in rural, delta, midland and mountainous areas, the land tax rates is equal to the average agricultural land use tax rates in the communes. Tax exemption and reduction There is a temporary exemption from land taxes in the following cases: • Land for construction of offices of State administrative and nonbusiness agencies, social organizations, cultural works and land for defense/security purposes. • Land in forests and high mountains, and areas of ethnic minorities. • Residential land of the families of war invalids and martyrs. • Residential land of the disabled who have lost their working capacity, juveniles, and lonely and helpless elderly persons who are incapable of paying tax. 1.3. Agriculture Land-Use Tax Individuals and organizations using or empowered to use farmland for agricultural production are required to pay this tax. It is the land used for agricultural production that is subject to this tax: cultivated land, aquacultural land, and forest-planted land. Tax calculation The farmland-use tax is to be calculated on the basis of the area of land, land categories, and tax rates determined on the basis of kilograms of rice per unit of land. The land category is dependent on the nature of the land, its location, climate, irrigation conditions and average productivity obtained in normal conditions. Tax rates are determined on the basis of kilograms of rice per hectare in case of cultivated and aquacultural land; for perennial land the rates are slightly higher, and for perennial trees on cultivated land rates are related to the rate applied to cultivated land. Wood trees and one- harvest perennial trees are subject to a tax rate of 4 percent of the output value. The price of rice is determined by the people’s committee. 244 P R O P E RT Y A N D L A N D TA X E S Tax reduction and exemption Bare-hill land and mountainous land are exempt from taxation. Those households who move to new economic zones to reclaim land for agricultural production are also tax exempt. In the event of devastated crops due to natural disasters and wars, taxpayers are entitled to tax reduction or exemption. Currently, a wider exemption and tax reduction is applied. Exemption has been extended to all farmers until the year 2010. Thus presently there is no revenue collection from this tax. 1.4. Land Use Right Transfer Tax Until 2009, organizations, family households and individuals were subject to this tax when transferring their land use rights. The base of this tax was the value of the land area in question. Exemptions The law did not apply if the State allotted or leased out land, if the transfer took place among family members or when land use rights were donated to the People’s Committees or to political organizations. Tax calculation The land use right transfer tax was based on the land area, land prices and the transfer tax rates. The land prices were the prices set according to the Government’s price bracket for different land categories. In case of land use rights that were transferred by auction, it was the auction price. If land was transferred by agreement between two parties, the agreed price was the basis. The tax rate was two percent for land used for agricultural production, forestry, and aquaculture and four percent for residential land and land for project construction. Repeal The land use right transfer tax was repealed in 2009, on the ground that it was sufficient to collect personal and corporate income taxes on land price gains. 1.5. Land and Water Surface Rental Charges There is a provision for collection of rental charges in the case of land and water surface leased by the State. Those who only rent land do not pay agricultural land use tax or land and housing tax. The level of rent is much lower than for the land use tax. The annual land rental unit price is calculated at 0.5 percent of the land price, while water surface rental price bracket varies between 10 and 250 million dong/km2/year. Land and water surface rental charges are reduced to 50 percent for cooperatives and to 40 percent if the taxpayer is facing a reduction in output caused by natural disaster or fire. 1.6. Land Use Levies Land use levies are collected when the State assigns land with provision for the collection of land use levies; the land use purpose changes; and when industrial parks, hi-tech parks, economic zones are built. 245 TA X R E F O R M I N V I E T N A M Households and individuals that are assigned residential land; economic organizations that are assigned land for the purpose of building dwelling houses for sale or lease or for the purpose of production and business or for agricultural, forestry, aquaculture are liable to pay this tax. Persons who lease land from the State and those using land for construction of infrastructure for common use in industrial parks are exempt from the land use levy. The land use levy depends upon land area, land prices and land use terms. The price of land is determined by the People’s Committees or the auction price as the case may be. The land use levy is not collected in cases of investment promotion, construction of public works for commercial or social purposes in the fields of education, healthcare, culture, physical training and sports. 1.7. Registration Fee This fee applies to the registration of houses and land. It also applies to registration of vehicles and weapons. Value of properties for registration fee calculation The value of a property for calculation of the registration fee is the actual price of the transfer of land use rights or the domestic market price under normal conditions. The value of houses for calculation of the registration fee is the actual value of a house transferred on the market at the time of registration. Where such value cannot be ascertained or it is lower than the market price, the price is set by the provincial People’s Committee. Registration fee rates The registration fee rate is calculated as 0.5 percent of the value of the houses and land. The maximum registration fee for a property per registration however is 500 million dong. Table 9.1: Revenue from Land-Related Taxes (million VND) No. Types of Taxes 2002 2004 2004 2005 1 Agricultural land use tax 772 151 130 132 2 Land-use right transfer 327 408 640 984 3 Land-use collection 3,244 8,149 14,202 14,176 4 Land and houses tax 336 359 438 515 5 Land rental 459 513 846 799 No. Types of Taxes 2006 2007 2008 2009 2010 1 Agricultural land use tax 111 113 97 67 56 2 Land-use right transfer 1,251 2,328 3,017 260 — 3 Land-use collection 15,416 28,677 31,598 37,695 41,691 4 Land and houses tax 594 711 902 1,203 1,394 5 Land rental 1,281 2,180 2,268 2,626 2,903 246 P R O P E RT Y A N D L A N D TA X E S According to the State Budget Law, land taxes have been found to be an excellent source of revenue for local governments in a large number of countries, provinces, and municipalities. 2. The Economics of Property Taxes Reason for Separate Analyses of Taxes on Land and on Improvements For economic analysis it is essential to distinguish between the “landâ€? and the “capital improvementâ€? components of property taxes. This distinction tends to be lost in everyday language when we talk about “propertyâ€? and “real estateâ€?. From an economic perspective, a property tax is a combination of two taxes, one tax on the value of land and the other on the value of improvements. The same tax rate is generally applied to both components of the tax base, but it is possible to have different tax rates on land and improvements, and there are some places that use different rates. The consequences of taxes on land and improvements are different, both in terms of efficiency and equity,. In terms of efficiency, the principal reason for the difference between taxes on land and on improvements is the elasticity of their supply. Land value derives from nature, good governance, the provision of public infrastructure, and the growth of the community. Any value that comes from efforts of the possessor of land is classified as an improvement. Since the value of land is independent of what the possessor of land does, the elasticity of the supply of land is zero, by definition. Improvements to land are due to people, generally depreciate over time, and generally require continual maintenance and protection. Therefore the elasticity of the supply of improvement is quite different from 0 (zero). Sometimes, as in mining, what is provided by nature can be depleted. Because depletion has a complex impact on issues of efficient taxation, the taxation of depletable natural assets is not included in this analysis. From the perspective of equity in a sovereign, sustainable commonwealth, land is the heritage of all people of all generations. Both the Constitution of Vietnam and the Land Law of 2003 stipulate clearly that all things provided by nature are owned by all of the people of Vietnam.78 78 VN-Constitution Article 17: The land, forests, rivers and lakes, water supplies, wealth lying underground or coming from the sea, the continental shelf and the air […] come under ownership by the entire people. VN-Constitution Article 18: The State manages all the land in accordance with the plan and the law, and guarantees that its use shall conform to the set objectives and yield effective results […] organizations and individuals are responsible for the protection, enrichment, rational exploitation and economical use of the land; they may transfer the right to use the land entrusted to them by the State, as determined by law. VN Land Law 2003 Article 1: This Law makes provisions for powers and responsibilities of the State as a representative of the entire people’s ownership of land and exclusive manager of land; the management and use of land and rights and obligations of land users. VN Land Law 2003 Article 5-1: Land is under the ownership of the entire people, and the State is the representative of the owner. 247 TA X R E F O R M I N V I E T N A M Thus, rather than considering land to be “private propertyâ€?, it might be better to consider it as subject to “private possessionâ€? or “secure tenure.â€? There are also important differences between taxes on land and taxes on improvements in terms of administration. The proper taxation of improvements requires continual on-site evaluations that are somewhat subjective and open to corruption. Land values on the other hand, because they are generally smooth functions of location, are much more objective, transparent, and easy to assess. For all of these reasons, an economic analysis of property taxes should be divided into an analysis of a tax on natural opportunities (urban sites, agricultural land, fisheries, electromagnetic spectrum, landing slots, etc.) and a tax on humanly produced capital improvements. 3. Land Tax The efficiency of land taxes Taxing land values is one of the most efficient ways of raising substantial government revenue. When properly administered, a tax on land, unlike most other sources of public revenue, introduces no inefficiency, or “excess burdenâ€? into an economy. Therefore, whenever a nation or a commune replaces a revenue source that has an excess burden with a tax on land, the overall efficiency of the economy improves.79 There are two requirements for a tax on land to have no excess burden. First, the tax must be independent of how productively the land is used. Second, the tax must not be greater than the value of using the land. If these two conditions are met, then all land that can be used productively will continue to be used when land is taxed, and those using land will continue to have an incentive to use it as productively as they can. A tax that is independent of how efficiently land is used can collect up to the full rental value of land without creating any excess burden. Furthermore, enforcement and compliance costs of land taxes are minimal. Enforcement costs involve the maintenance of a cadastral system, frequent and transparent assessments, sending out bills, and dealing effectively with delinquencies (of which there will always be a few). The compliance costs of a land tax are comparable to the costs of paying bills such as electricity. For urban and commercial land, a monthly payment would probably be most convenient. This could even be done by an automated debit from the possessor’s bank account. For agricultural land, it might be best for payments to be due shortly after the harvest. For southern provinces this could mean up to three payments per year, while in mountainous regions in the center and the north one single annual payment would be appropriate. 79 The excess burden of a tax that has an excess burden is roughly proportional to the square of the tax rate. It is also roughly proportional to the reciprocal of the sum of the reciprocals of the elasticities of supply and demand. 248 P R O P E RT Y A N D L A N D TA X E S While there are other ways of raising public revenue without excess burden,80 the economically most important way of raising public revenue without excess burden is by taxing land.81 Land taxes and superefficiency For several reasons, a tax on land is not merely efficient. Rather, it offers “superefficiency,â€? meaning that a tax on land does not simply provide public revenues without excess burden, it actually helps to offset other circumstances that generate economic inefficiency, thereby producing an economy with greater overall efficiency than an economy without the tax. The impact of land taxes on land speculation The higher the rate of taxation of land, the less rewarding land speculation is. As the rate of taxation of land approaches 100 percent of the rental value of land, profitable land speculation becomes virtually impossible, because the selling price of land falls lower and lower while the holding cost rises higher and higher. Therefore, taxes on land tend to take land out of the hands of speculators and put it into the hands of those who plan to use it. With land speculators out of the market, the selling price of land falls further, lowering the holding costs for the general public and thus making land even more available to those who wish and offer to use it productively. To elaborate: Land speculation is attractive when people have different beliefs about how fast land prices will rise. Those who expect a rapid rise buy land and wait. This reduces efficiency for two reasons. First, those who speculate in land are generally not the persons who are most talented in managing the land while waiting for the price to rise. Second, when speculation is attractive, land is most valuable to those who expect the most rapid price rise. Such extreme beliefs are generally wrong. (They lead to the phenomenon known as the “winner’s curse,â€? where the fact that a person was the highest bidder is an indication that he paid too much). Land owners who believe that land will rise rapidly in value think it is unproductive to improve land currently, because that would make it unavailable for even greater improvements later. Thus the pursuit of speculative profit produces a bias against improving land. By reducing or eliminating the profit from land speculation, taxing land eliminates this bias against improving land, thereby generating more intensive, productive, and sustainable land uses. A higher land tax, in addition to reducing land speculation, tends to reduce premature conversion of agricultural land to development uses. Because speculation is reduced or eliminated by a high land tax, the most valuable sites are developed rapidly when land is taxed heavily (up to 100 percent of the market rental value), leaving little or no market for the development of those that are not ripe for development. 80 Other ways of raising public revenue without excess burden are taxes that compensate for the harm caused by pollution, congestion, and global warming, and prices (no greater than marginal cost) for public services. 81 From an economic perspective, taxes on land include taxes on other scarce natural opportunities such as fishing and the electromagnetic spectrum (used for mobile phones, radio, and TV). 249 TA X R E F O R M I N V I E T N A M The impact of land taxes on access to land for entrepreneurs Taxing land improves economic efficiency by putting land in the hands of people who get greater returns on their assets. If lending markets were perfect, all people would face the same interest rate. But in fact, some people face high interest rates while others enjoy low rates. Some people enjoy yet lower rates because instead of borrowing, they are looking for ways of investing their money.82 When the tax on land is increased, potential users of land reduce their purchase bids for the land by the present discounted value of the future land taxes.[Gaffney] For people enjoying low interest rates, this present discounted value of the future taxes is greater than for people facing high interest rates. Thus an increase in land taxes decreases the value of land to people facing high interest to a lesser degree than it decreases the value of land to people enjoying low interest rates. Therefore increasing the tax on land shifts the possession of land toward persons who face high interest rates. Because these persons face high interest rates, they require greater returns on the investment. Thus a tax on land also shifts land to uses with greater productivity. The impact of land taxes on access to loans for entrepreneurs The selling price of land is the present value of future rent minus taxes. As taxes approach the full rental value of land, the selling prices approaches zero. This has three important implications: i. When a government plans to tax land at a level corresponding to a significant proportion of the rental value of land, it is advisable to make the tax base the annual rental value of the land instead of the selling price of the land, and ii. As the tax rate approaches 100 percent of the rental value of land, the acquisition of land becomes more and more feasible for those with little or no access to credit. Therefore the only money needed to open a business will be what is needed for actual capital investments. iii. With land of little value as collateral for loans, banks will be unable to simply lend to whoever is able to offer land as collateral. Instead, they will need to engage in careful evaluation of and support for proposed business plans. This should improve the efficiency with which the nation’s savings are invested. The impact of land taxes on rent-seeking Rent-seeking is the use of scarce resources to arrange to receive benefits for which one does not pay. One form of rent-seeking is persuading the government to spend money, raised from taxes on others, on projects that benefit themselves. While rent-seeking can be personally beneficial, the resources that individuals spend on rent-seeking efforts are a waste from the perspective of society as a whole. And the projects that are undertaken as a result of rent- seeking efforts are often wasteful as well. Taxes on land reduce the potential profit from rent- seeking, thereby reducing the amount of rent-seeking activity, and therefore improving social 82 The laundering of illegitimate funds is one of the most extreme examples of “enjoyingâ€? low interest rates. 250 P R O P E RT Y A N D L A N D TA X E S efficiency. To the extent that governments refrain from spending on wasteful projects because of reduced rent-seeking, there is a further improvement in social efficiency. The impact of land taxes on foreign direct investment (FDI) One of the major problems FDIs face when considering Vietnamese investments is the high cost of land, if it is available to them at all. Just as a significant land tax removes the motive for land speculation, it also provides substantial government revenue, permitting the revenue- neutral reduction or elimination of inefficient taxes on income and commerce. With such a policy, a nation can expect to experience a sharp increase in FDI, giving an important boost to its socioeconomic development. With more rapid development, not only will there be more demand for labor in general, resulting in higher wages, but there will also be greater demand for skilled labor. If Vietnam implements low taxes on commerce, it is possible that many international companies will choose Vietnam as their headquarters, both to take advantage of Vietnam’s high educational standards and to lower their global tax liabilities. 4. Equity Land as the common heritage of all persons of all generations As already mentioned a land tax gives expression to the constitutional declaration that land belongs to all. Since an equal physical division of a nation’s land is nearly impossible and would be economically disastrous, an equal financial division, through a tax on land, is an appropriate expression of equal rights to land. Collecting the annual rent of land for public purposes also reflects Adam Smith’s first cannon of taxation, “The subjects of every state ought to contribute toward the support of the government … in proportion to the revenue which they respectively enjoy under the protection of the state.â€? Financing from direct beneficiaries of local services and infrastructure In a market economy, participants are generally expected to pay for the goods and services that the economy provides for them. This principle can be applied to local services and infrastructure that provide benefits to some persons and not others. When an executive body at the lowest political level builds a new bridge across a river, the biggest beneficiaries are those who have possession of the land on the other side of the valley, which had previously been difficult to reach. If they are now required to pay a few bags of rice a year for the bridge, they are still better off than without the bridge. It is fair for the community to place such an obligation on them. Of course, a rice paddy near the bridge benefits more than the one that is a half-day foot hike away. Similarly, land near the bus and truck stop, or at the end of the paved road, is likely to receive the greatest benefit from a public investment. Payments according to the increase in rental value generated by infrastructure like the bridge maintain equality between those who do and those who do not receive special benefits from such investments. If the bridge is a 251 TA X R E F O R M I N V I E T N A M worthwhile investment, the total increase in land rent that it generates will be greater than the cost of the bridge, so that if the bridge is financed by the increase in land rent it generates there will be no need for the bridge to be subsidized by people in unaffected regions.83 The same principle applies at the next level of government. Thus if a province constructs a large bridge that benefits several communes, land values will rise in those communes. The province can pay for the bridge by requiring the communes to pay in proportion to the increases in the rental value of land that they receive. The province would therefore not need to charge citizens who do not benefit, and would not need a subsidy from the national government either. A similar analysis applies to the national government. If the national government provides bridges or other items of infrastructure to individual regions (access to national highways or high-speed trains, for example), then fairness requires the regions that benefit to compensate the national government according to the benefits that they enjoy.84 A practice of having substantial taxes on land would provide the funds with which individual provinces and communes could pay for national infrastructure in proportion to the benefits that they receive. In sum, capturing the annual rental value of land allows individuals and communities to pay for the benefits that they receive from government infrastructure projects, in the way that fairness requires. 5. Tax on Improvements Unlike a tax on land, the component of a property tax that falls on improvements85 can have a significant excess burden. Therefore, when there is a substantial tax on land, it is important to tax improvements at a much lower rate, if they are to be taxed at all. While taxes on improvements are generally inefficient, there are at least two reasons why a small tax on improvements might be included in a package of efficient sources of public revenue. First, the more improvements to land a nation has, the more it will need to spend on police, courts, national defense and diplomacy to protect its property. It is reasonable to assign some portion of these costs to improvements and therefore to levy a tax on improvements to collect that much revenue. If a tax on improvements simply compensates for the cost of protecting additional improvement, then there is no excess burden of the tax. The second reason that one might wish to levy a tax on improvements is to erode the wealth of the wealthy. The efficiency rationale for eroding the wealth of the wealthy is that it is difficult to maintain the cohesion of a society if inequality in the distribution of wealth is too extreme. 83 The extension of the Jubilee line in London’s rapid transit system has dramatically increased neighboring land values. The increases around just two of the six new stations would have generated enough revenue to pay the total construction cost of £3.5 billion. It is estimated that the total increase in land value along the extension was three times as much as the construction cost. 84 Similarly, negative externalities of national projects, such as fenced in high ways with no nearby exit or even crossing, should be compensated according to the losses in land value. 85 Improvements are not only buildings but also private investments in irrigation ditches, private roads and bridges, continual individual efforts to maintain rice paddies, etc. 252 P R O P E RT Y A N D L A N D TA X E S Thus a tax on the value of property for this purpose would have a significant exemption, so that people would not be penalized for providing for their basic needs and retirements. If a society has a consensus that an extreme concentration of wealth is costly, then a tax that compensates for this cost has no excess burden. If a tax on improvements exceeds the marginal cost of protecting improvements and the cost of more extreme inequality in the society, then there is an excess burden to the tax, because the more improvements are taxed, the fewer will be built and the smaller will be expenditures on maintaining improvements. The excess burden of a tax on improvements will be roughly proportional to the square of the amount by which the tax on improvements exceeds the sum of the marginal costs of protecting improvements and the marginal costs of inequality. 6. Tax on Real Estate Transfers86 Taxes on real estate transfers are questionable in terms of efficiency, equity, and sustainability. They also invite corruption and numerous evasion schemes. While the portion of a real estate transfer tax that falls on improvement punishes investment and therefore causes inefficiency, the portion on land at least captures some of the unearned gains, though with a long and variable delay. The tax on real estate transfers in Vietnam takes account of this efficiency difference by applying a lower tax rates to the transfer of improvements than to the transfer of land. However, this encourages investors to evade the tax by having improperly large portions of their assessments assigned to buildings. The most important flaw in tax on real estate transfers is that the tax is collected only in the event of a recorded transfer. This leads to inefficient subleasing schemes. These diminish the security of tenure, diminish transparency, and add to the difficulty of public revenue forecasting and budgeting. Furthermore, even a moderate transfer tax of a few percentage points will thwart some transactions. An annual tax on land values has none of these difficulties. It encourages improvements and captures even subtle land value increases due to demographic changes, public services, and altered tax policies at times when they occur, while providing a continuous stream of public income that is both foreseeable and transparent. One informal variation on transfer taxes is a tax on changes in land use. Land conversions are often promoted by local and provincial governments to acquire substantial, easy, one-time revenue.87 86 In November 2007 the National Assembly has voted for the introduction of a 25 percent on capital gains on real estate transfers. This is on the top of the 2 percent transfer tax already in place and brings it almost in line with the current income tax on rent which is at 28 percent. 87 This tempting practice is not sustainable: it deprives future generations of such benefits, it often converts land to “higherâ€? uses prematurely (thereby disrupting unnecessarily well established communities and livelihoods), and it invites for corruption. Furthermore, the taxes, fees, and bribes that must be paid for land use conversions can easily bankrupt poor households, as they lack the resources to pay upfront the unexpected charges that are out of line with the current use and rental revenue, reflecting actual local market demand. 253 TA X R E F O R M I N V I E T N A M 7. Policy Implications In Vietnam more than 86 million persons compete with thousands of state owned enterprises and private companies for the use of land.  In addition to this private demand for land, there is communal demand for land, for such uses as parks, streets, waterways, and natural reserves.    One of the central aims of land legislation in Vietnam has been the creation of an efficient land market, designed to promote equitable outcomes and sustainable development.  However, even though many legislative measures have been taken to eliminate inefficiencies and inequalities, these measures have not managed to control price inflation, land speculation, encroachment onto public lands, persistent resistance to formalization, and corruption. And the SEMLA reports argue that the Land Law and associated decrees have reached such a level of complexity that they are difficult for even the most sophisticated citizen to understand. In this section we show how taxing (collecting the rent of) land according to its full market rental value could help resolve, in a efficient and equitable manner, a number policy questions that have been discussed repeatedly in the five year socio-economic development plans (SOEDs). Advantages of Formalization of the Land Market88 Land deals between two private parties do not necessarily need to be registered formally. People sometimes consider it sufficient to transfer land use rights informally, by a handshake or a simple rental agreement. However, if there is no legal formalization, there remains a risk that investments by the person who does not have legal tenure might be taken without compensation. Thus it is valuable, for both efficiency and equity, to have a simple and inexpensive process by which land use rights can be legally transferred. Keeping track of land title information does not request a large bureaucracy and can be financed by fees equal to marginal costs of the service or by general public revenues. The public has an interest in a transparent and efficient system of land titles, to avoid the social costs of idle land and to promote the efficient and sustainable use of land. Managing land title registration and tax collection, once these are well established, is like managing the front desk and back office of any hotel: Room allocation, proper registration, and full rent collection. The fee for land registration, as an adjunct to a system of land value taxation, should be no more than the cost of a clerk’s time. Any reduction in transfer and registration costs will be reflected in increased land prices. Hence, by taxing land values those value increases would be recovered for the public good. 8. Land Recovery Land use rights are allocated to land users by the state, and the state has power to recover these rights in prescribed circumstances.89 Article 38 of the Land Law of 2003 lists the 88 Land markets remain largely informal. Recent figures from the ADB suggest that only about 50 percent of transactions of agricultural land are formally registered. In urban areas estimates vary, suggesting that between 60 percent and 85 percent of land is transacted informally (CIEM 2006; Gillespie 2002). 89 This is equivalent to “eminent domain,â€? a nearly universal practice around the globe. 254 P R O P E RT Y A N D L A N D TA X E S reasons for which the state is authorized to recover land use rights. The first clause is quite comprehensive, allowing recovery when “the state needs to use the land for the purpose of national defense, security, national benefit, public benefits or economic development.â€? Notice must be given by the state prior to recovery: 90 days for agricultural land and 180 days for nonagricultural land. In most cases, households from whom land is recovered are eligible to receive compensation from the state, either in form of similar land elsewhere or in cash. Some are eligible for additional support to help redress the effects of disruption on their livelihoods.90 While there have been problems in the land recovery process, which have caused complaints, resistance, conflict, and significant delays in land clearance and public works, the main cause of public outcry has been the compensation scheme. Official prices do not match market prices for land. This makes it almost impossible for “compensatedâ€? title holders to gain the use of land comparable to what has been recovered from them. Further fuel is added to sensed injustice when land is recovered, not for basic infrastructure needs but for economic development projects and industrial zones. In these cases the difference between the compensation and the value in the new use value is drastically widened. Under a regime of full land value taxation this issue disappears because public collection of the rental value of land drives the selling price of land toward zero, no matter what the designated current or future use is. Land recovery under land value taxation would cause only the loss of tenure, and not a financial loss. Since the selling price of land would be close to zero, no compensation would be needed for the loss of land. Under land value taxation, compensation would be limited to the financial and psychological costs of dislocation and the loss of capital improvements attached to the site. It might be suggested that a compensation scheme should be compassionate. Under a regime of public collection of the full rental value of land, compassionate compensation is more feasible since no funds are needed to compensate for land, which belongs to the entire people in the first place. 9. Settling Land Disputes  One of the ways in which land value taxation promotes equity is in land disputes.91  If two stakeholders claim title to the same piece of land and records do not clearly specify which one of them is the proper title holder, the question can often be resolved by consulting tax records. If one of them has been paying the tax on the property and the other has not, it is reasonable to presume that the true owner is the one who has been paying the tax. The one who has not been paying the tax should be obliged to explain how it is that he is the owner despite the fact that the other has been paying the tax. 90 Decree 197/2004/ND-CP on compensation, support, and resettlement. 91 Gillespie (2001) reports that resolution of land and housing disputes can be time consuming, expensive and uncertain, with over 70 percent of unresolved civil law cases concerning housing disputes. The district courts of HCMC hear about 600 cases a year with one-third going to appeal at higher courts and taking about an average of up to three years for settlement. 255 TA X R E F O R M I N V I E T N A M Land value taxation also reduces the stakes in land disputes by reducing the value of land. (It is also possible to have disputes over the value of improvements, and land taxation does not reduce the magnitude of these disputes.) 10. Land Fragmentation and Consolidation In many communes, the land holdings of households are fragmented into several plots per household, to give each household the same amount of land of each type. This is a great hindrance to agricultural productivity, since it limits economies of scale.92 Communes with highly productive land have been pressing for reallocations that consolidate holdings, using available land-quality data to achieve fair redistributions among households. If reliable data were collected for tax purposes, households would find it easier to recognize which trades would be fair while raising everyone’s productivity. Similarly in urban settings, inefficiently small plot sizes and the difficulty of consolidating adjacent lots impede economical construction, resulting in the loss of much valuable land to dead spaces between buildings and inefficiently small buildings. Taxing land at its full rental value would encourage title holders to synchronize their development plans or pass their land use rights to someone who can build in a cost-effective way. Major redevelopment and infrastructure project are often delayed because of the bargaining power of current lease holders. If they were obliged to pay the full rental value of land they would find it in their interest to clear the way as soon as they were offered relocation compensation or capital shares in the in new developments. 11. Internalizing Externalities Many human activities generate benefit or harm to those who use land that is close to the activity in question. Economists call these benefits and harms “positive externalitiesâ€? and “negative externalitiesâ€? respectively. When all of the rent of land is collected publicly, any externalities that would affect the wealth of those who have title to land will affect the public treasury instead, through their effects on the amount of rent that can be collected publicly. This means that it is not necessary to consider a special scheme for compensating the holders of land for such negative externalities (or charging them if they are the beneficiaries of positive externalities) because the payments that they are required to make for the use of land adjust with the externalities. Efficiency will still require that those who cause externalities be charged93 or 92 Due to different historical land reform efforts, the north is twice as fragmented as the south. Average plot size is also much smaller and plots are much more widely dispersed than in the south, making land consolidation among households very difficult. Furthermore, in an effort to expedite land distribution following the 1993 Land Law only single land use right’s certificates have been issued for all plots of land. Many households are thus forced to reregister their land on separate certificates at additional costs, resulting in public reluctance toward the efforts. 93 For example, farmland close to unsealed landfill sites in Da Nang has become polluted and unproductive resulting in protests and compensation claims by the local population. Tu Liem district of Hanoi provides another example. There, a common complaint amongst farmers is that they are compensated only for the land taken for construction. Left over, adjacent land in small, often awkward-shaped plot corners which no longer have irrigation or other agricultural infrastructure is not compensated, despite its loss of productive potential. 256 P R O P E RT Y A N D L A N D TA X E S credited, to motivate them to undertake efficient amounts of externality-generating activities, but a general compensation scheme for those who have title to surrounding parcels will not be needed, because the government will be the party that benefits or loses from externalities. The biggest externality landlords normally enjoy is through the public provision of infrastructure and public services. Building bridges, roads, schools etc. tends to enrich the landlords, while tenants face higher rents. Taxing land values will assist the public to recover the costs associated with the service provision and thus assist in funding further projects. 12. Land Taxes and the Poor When policy makers consider taxing land, they sometimes worry that poor people may be unable to pay their land taxes. This concern should be divided into two parts: Poor people with land of little value and poor people with valuable land. The solution to the problem of poor people with land of little value is to give every citizen with an identity card showing the right to an exemption from some amount of land taxes, say VND 1,000,000 per year.94 Every family member over the age of 16 should be allowed to contribute his or her tax exemption to the family tax bill. Any family with land of little value will then have no tax bill. Furthermore, allowing citizens who have no land to sell their tax credits to those who do have land will permit all citizens to benefit equally from the land tax credit. A subsistence farmer with land of inconsequential value should not be required to pay a land tax. However, a farmer who enjoys the use of a fertile, sunny, well irrigated patch of land with greater value than what others in the village can enjoy should be required to compensate the other farmers in the village, for reasons of efficiency and equity. Similarly, a community of subsistence farmers on land of little value should not be required to pay any land taxes. But a community of farmers with very high land value per capita should be expected to compensate the rest of the nation for their superior access to the nation’s heritage of land. The problem of poor people with land of significant value (particularly in cities) is difficult because it draws on our feelings of compassion while frustrating us with its inefficiency. These persons may be reluctant to move while having no funds with which to pay their land taxes.95 If we let them stay where they are, they impose on the community a significant loss in terms of the opportunity to grow and to obtain the additional land tax revenue that comes with growth. We have two recommendations. 94 Depending on the revenue estimates made and the time a land tax has been in place and well established it is fair to assume that total revenue exceeds government needs. The excess revenue could be used to establish a citizen’s dividend and/or finance a system of social security for all. 95 However, many originally registered land holders are “ghost-owners.â€? The elderly person holding the land title might have been moved by the family, the old hold demolished and a new one constructed at its place, rented out at full market value to whomever is in need of real estate. Furthermore, current land use fees (for example, within the city of Hanoi) are as low as VND 40,000 per month, not even enough to fill the gas tank of a motorbike. Migrant workers and students, on the other hand, pay VND 200,000 per month and more, for as little as a few square meters of housing. 257 TA X R E F O R M I N V I E T N A M First, allow any poor person to defer his or her land taxes until death, if there is enough value in the property to pay the taxes. At death, sell the property and pay the taxes, with interest. Second, when it is apparent that there will not be enough value in the property to pay the taxes at death, make explicit, case-by-case decisions about whether to grant tax waivers on a compassionate basis. Because cities need to be redeveloped and because all redevelopment could be blocked if no poor person was required to pay their land taxes, it may be reasonable to put pressure on people to give up their valuable land. One possible compromise is to help people with valuable land exchange their land use rights for high-rise condominiums in their neighborhoods, so they can live in familiar surroundings, in apartments equipped with modern conveniences, without occupying a significant amount of valuable land. 13. Application and Exemptions Except for cases of hardship that should be dealt with on a one-by-one basis, there should be no exemptions to the land tax owed, to keep the system simple, transparent and fair. This no-exemptions rule should include all ministries, government agencies, schools, hospitals, museums, SOEs, etc.96 Efficiency, budget transparency, and accountability are the reasons for this rule. Having all public entities pay for the land they use at market prices gives them, the public, and policy makers the opportunity to consider whether the location and size of current public activities represents good land use, taking account of the changes that occur over time in technology, demography, and public needs. A government body that uses space above its essential needs or at an inappropriate location causes a loss of productive opportunities and thus diminishes the wealth of the nation. Furthermore, having public bodies pay for their land use gives local communities an opportunity to receive compensation for their provision of land to higher levels of government. 14. Equitization of State Owned Enterprises (SOEs)97 Current ambitions to equitize SOEs should be considered from the perspective of what happened in the privatization process to the economies and social fabric in the former East Block or in reference to the doings of the “Deutsche Treuhandâ€?. As SOEs are massive land holders, all of their land is getting “privatizedâ€? to those with insider knowledge and deep 96 As of 2002, SOEs held up to 95 percent of land in Hanoi under lease to organizations, leaving only a relatively small share available for private use. Furthermore, data from the 2006-PCI found that of the firms currently with a lease, 13.7 percent lease informally from an SOE. It should be no question that the SOEs should be liable for all implicit (collected from third parties or not) rent and pass it on to the state as part of their total rent payment for the privilege of monopolized control of a location. Furthermore, leasing out land by SOEs is currently illegal (Land Law 2003, Article 109/2). However, in economic terms leasing out land (a simple sublease) would not cause any market distortions were they forced to pay the market rent just as everyone else does. Thus, this could be another law that could be nullified in the future. 97 The Vietnamese government ordered privatization of State Owned Enterprises in 1992. More than 3,500 SOEs have been privatized by the end of 2005. Another 1,505 are planned to be privatized by 2010 (among them all the national banks). 258 P R O P E RT Y A N D L A N D TA X E S pockets. The companies will be sold at below-market prices, since it is impossible to predict future land prices. Thus, many companies will be bought up simply for the purpose of speculating on their land. Factories, even profitable ones, will be closed and the land held until the public has made major infrastructure investments. Taxing land values would prevent the damage from speculative purchase of SOEs, since it takes the profit out of land speculation. Taking land speculation out of the equitization equation would also provide greater opportunities for groups with less deep pockets (such as the workers) to bid on SOEs, since less capital would be required. Taxing the land of SOEs, whether privatized or not, would ensure that the use of this land, this part of the Vietnamese homeland, would provide a steady stream of income for future generations. 15. Taxes on Rental Income from Real Estate Rental income from real estate is generally subject to income taxation. Even though this captures a part of land rent for the public treasury, it is an inefficient way to do so. First, there is evasion. One simple form is misstated invoices. Furthermore, even with the best accounting, the net income of real estate is difficult to specify, because of uncertainty in such items as depreciation. Between laxity in the tax code and eager accountants and lawyers, there are generally many legal ways for real estate investors to reduce their taxes, at a substantial cost in social efficiency and equity. Subjecting real estate income to income taxes also places a burden on those who use their land productively, compared to those who let their land sit idle, speculating. Whatever taxes are to be paid by those who have title to land are best levied on the rental value of land, without regard to the current income from land. This ensures that those who have the use of land have the best incentive to use the land productively. 16. Protecting against Sudden Increases in the Rental Value of Land Taxing land values will generally lead to smoother changes in land rent, since speculative elements, corruption, and rent seeking are minimized. However, major infrastructure projects, nearby private investments, global economic developments, and changes in the legal codes could lead to unanticipated, rapid changes in land rents. This could lead to losses in the value of improvements that investors have placed on land. To protect themselves against such losses, investors might wish to take out insurance against change in land rents. It is reasonable to expect that insurance companies would offer such insurance if land was fully taxed. The purchase of such insurance would help to give investors confidence that they would be able to recoup capital investments in accordance with their initial business plans. 17. Pilot Projects If a national effort of making land the major source of public revenue appears to be too ambitious, pilot projects could be designed. However, each tax- or duty-free enterprise zone 259 TA X R E F O R M I N V I E T N A M throughout the nation represents such a “pilot projectâ€? already, based on the reasoning that lower taxes and duties will spur economic growth.98 An interesting project is emerging on the eastern banks of Hanoi’s Red River, a new multi- billion-dollar city center along the eastern bank of Hanoi’s Red River has been agreed upon by Vietnam and Korea. The large and modern commercial and residential area would require the resettlement of 170,000 people and is subject to approval from the national assembly of the communist country. The urban development would by 2020 run along a 40-kilometer stretch of waterfront, across the Red River from the ancient centre of Hanoi. The 4,200-hectare urban project would require about US$7 billion in investment for new high-rise buildings, residential blocks, parks and dykes to be built along the flood-prone Red River. By applying the principles of land value taxation this mammoth project could move forward more quickly, more smoothly, and in a more equitable way than under the current tax, investment, and land laws. Turning the 4,200 hectare area into on single tax-free zone today, charging annual market based land use fees, would put all stake holders on a level playing field and assure transparency in the process. However, as the project is only supposed to be finished by 2020, Vietnam would lose the opportunity to benefit on a large scale from the growth and sustained prosperity that would stem from a switch to land value taxes envisioned within the current five year socioeconomic development plan. 18. Practical Application of Taxing Scarce Space: London’s Congestion Charge London’s great success with its congestion charge, facilitating traffic flow and funding increased public transportation services, provides a good example of how charging for the use of a scarce resource (in this case city streets) can increase efficiency and equity. Travel times have been greatly reduced while convenience has been improved, leading to a more efficient city. Unfortunately, London has minimal taxes on land. With greater taxes on land, the rent increases due to more efficient transportation would generate a substantial increase in the city’s revenue and distorting taxes could be taken off commerce. Hanoi and Ho Chi Minh City could get their increasingly congested traffic under control by introducing a charge of say, US$20 per car per day, for use of city streets (buses and trucks US$100). From such a charge approximately US$5 million per day would flow into Hanoi city’s treasury. Even if we were to allow for the most expensive administration costs imaginable, $4 million per day would remain for upgrading the public transportation system. This would help to limit motorbike traffic significantly, since people would have the option of enjoying an efficient and comfortable public transportation system. Whole streets and city sections could then be limited to public transportation or foot traffic only. Furthermore, it would only take 50 days of revenue to fund the whole of the French project of a tramway or subway system. Having a system of capturing the increased land rents for public benefit would help turn Hanoi and Ho Chi Minh City into sustainable, first class world cities of the 21st century. 98 Note the strong relationship of a given tax regime and rent: where taxes are lower, rent is commonly higher and vice versa. Thus, without the full public collection of land rent, the lower tax rates will be absorbed by the landlords. 260 P R O P E RT Y A N D L A N D TA X E S 19. Administrative Issues of Land Value Taxation Assessing the Value of Land The cornerstone of a system of property or land value taxation is its assessments. They need to be accurate, up-to-date, and transparent. The use of today’s computer technology, with geographic information systems (GIS) and World Wide Web permits assessments that are updated as frequently as one wish. The full transparency of such a system provides a level playing field to all market participants and relevant government agencies. There are three basic ways to achieve good assessments: • Self-assessment • Professional assessment, and • Public auctions A system of self-assessment is suitable for assessing unimproved land, and for assessing the value of the combination of land and improvements, but not (at least directly) for assessing the value of land that is improved. In a system of self-assessment99 each possessor of land use rights declares the value of his property to the government, with the condition that any citizen has the right to purchase the property at the declared price. An excessively high price is discouraged by the resulting high tax payment, while an excessively low price is discouraged by the likelihood that that would result in an obligation to sell the property. If the land is unimproved, the assessed value can be a rental value. The self-assessment process will reveal the rental value of the land. If the land is improved, the self-assessment must refer to a selling price, namely the selling price of the combination of the land and the improvement. When there is only partial land value taxation, so that land has a significant selling price, it is possible to use a system of self assessment to determine land prices indirectly. From the self- assessments of the combination of land and improvements one can make a statistical estimate of the selling price that land would have if it had not been improved. As one moves toward 100 percent land value taxation, the selling price of land goes to zero and the usefulness of self assessment in determining the value of land fades away. Professional assessments can be outsourced to a professional real estate or appraisal company or provided by a government agency. It is important that an appeal process be provided when professional assessments are used, to hear lease holders’ arguments as to why their assessments are excessive, as well as to gather additional market information in order to improve the accuracy of future assessments.100 99 As strange as the idea sounds at first hand, anytime someone leases an office space or rents an apartment, a basic self assessment takes place: Can I afford the location? What is the value to me? Am I willing and able to make regular payments? For a discussion of the theory of self-assessment, see Plassmann and Tideman (forthcoming). 100 The International Association of Assessing Officers IAAO acknowledges an error-bandwidth of +/- 20 percent as generally satisfactory. However, this applies to real estate appraisals, including the hard-to-assess improvements. If it were only land values to assess, the accuracy would be greatly increased. 261 TA X R E F O R M I N V I E T N A M Public “rental auctionsâ€? can be used to determine which potential user of the land would place the greatest value on the opportunity to use the land. In such auction, each potential user submits a single, sealed bid, representing how much he is willing to pay for the use of the land for the first year. The land goes to the highest bidder, but the price that the highest bidder is required to pay—and the price that becomes the initial assessed value—is the second- highest bid. As a consolation prize, the second highest bidder receives 1 percent of the first year’s rent. The results of such public rental auctions provide data for a land value map. Other points on the map are derived by interpolation. There are two reasons for requiring a payment of only the second highest bid. First, the second-highest bid represents the opportunity cost of the land, what the land would have been worth to someone else if the highest bidder had not wanted it. It is fair that the user of land be charged no more than this. Second, if the price that must be paid is the second-highest bid, then a bidder will not be concerned about the possibility that he will be the highest bidder and the second-highest bid will be much lower, implying that he could have obtained the land for much less than he bid. When bidders have this concern, they have an incentive to bid less than the land is actually worth to them, and it will be possible that the land will not go to the person who can make the best use of it. It can be shown that when the amount that must be paid is the second-highest bid, no bidder can do better with a false, strategic bid than he does by reporting honestly what the land is worth to him (see Vickrey 1961, 837). The payment of 1 percent of the first year’s rent to the second-highest bidder provides an incentive to knowledgeable people to participate in the process and reveal the potential value of the available land, thereby providing important information about land values. The highest bidder would be entitled to continue using the land for as long as he was willing to pay the (regularly revised) rental value of his land in an unimproved state. In future years, the rental value would be interpolated from the results of public rental auctions of nearby sites. Because people at the local level have relevant local knowledge, assessments should be done on local level. Because the locality may be tempted to try to prove that its land has little value, to get greater support from the province, provincial government should oversee the assessment process and adjust assessments across local jurisdictions. Similarly, the central government should oversee the assessment activities of provinces to achieve comparability across provinces. 20. The Optimal Duration of Land Use Rights Collecting the annually adjusted rental value of the land makes it practical and reasonable to have open-ended land use rights. As long as the tax is paid, the lease holder enjoys the privilege of using the site, forever, if he wishes to. Thus the current land law with various durations of leases for various uses could be simplified. There is an important efficiency reasons for land use rights of indefinite duration. With such rights the leaseholder does not have an incentive to slow down capital investments and maintenance efforts toward the end of the lease. Rather, he has an incentive to keep his 262 P R O P E RT Y A N D L A N D TA X E S improvements at the most efficient level, so as to continually achieve the greatest return on the use of the site. 21. Revenue Potential of Land Rent As in many other places, data on the amount of rent in the economy is not collected in Vietnam. Therefore, until substantial resources are spent on gathering the necessary data, only a rough estimate of the revenue potential of rent can be made. One inference that can be made from economic theory is that when money is spent on worthwhile infrastructure (roads, bridges, utilities, rapid transit systems, etc.) the resulting increase in the rental value of the surrounding land will be great enough to pay the cost of the infrastructure. This is the “Henry George Theoremâ€? (see Arnott and Stigliltz 1979). Thus the rent of land should be more than enough to finance all worthwhile infrastructure expenses. The total amount of rent in the Vietnamese economy should be much more than the cost of worthwhile infrastructure. This is because land has value even before infrastructure is added and because there are a number of components of rent other than land. The most important other source of rent is oil. But there is also coal, spectrum rights, fishing rights, water rights, and perhaps others as well. To get as rough idea of the magnitude of land rent in Hanoi and Ho Chi Minh City, one can employ a spatial model of rent. Suppose that: • The selling price declines exponentially from a peak of $P per square kilometer in the center of a city. • One kilometer from the center of the city, the value of land is Pe-f. • The fraction A of land is available for private use, the rest being streets, parks, etc. • The radius of the city is r. • The interest rate in the economy is I. • The rate of growth of the economy is G. ï‚· The annual tax rate on the sale value of land isTT The annual tax rate on the sale value of land ï‚· • The annual tax rate on the sale value of land isisT Then the total selling price ofprivate land inthe city is • Then the total selling price ofofprivateland ininthecity isis Then the total selling price private land the city rr ï?² 1 ï?² 1 rr 2ï?°AP 2ï?°AP S   2ï?°ï?²APefï?²ï?²dï?²  2ï?°AP[(   22))efï?²ï?²]]  S  2ï?°ï?²APe f dï?²  2ï?°AP[(   00 ff ff e f 00 ff   22 [[1 ((1 fr ))efr ]] 1  1  fr e fr For city withaasize greater than about 5 km, theeffect of the radius of the city is inconsequential and For aaFor awith with a size greater than aboutkm,thethe effect of the radiusofthe city is inconsequential and city city size greater than about 5 5 km, effect of the radius of the is inconsequential can be treatedas treated so that theformula becomes can be treated be infinity,assothat theso that thebecomes becomes and can as infinity, infinity, formula formula 2ï?°AP 2ï?°AP S S ff22 To translate from from thevalueof land to therental value of land, land,employs the fact that the rental To translate the sale sale of land land rental value of land, one one employs the fact that To translate from the sale value value ofto the to the rental value of one employs the fact that the rental of rental proportion of proportion of of land will be approximately + T – G, so I + price the land, as aa proportion of the selling price of land will be of land will be approximately that the price of land, asprice of land, as athe selling pricethe selling price approximately II + T – G, so that the T – G,valueof the private landvalue ofcityis private land in the the private land in the city is so that the total rental in the city is total rental value of total rental 2ï?°AP 2ï?°AP R R ((II  T  G )) T G ff22 263 Some plausible suggestions for the parameters in the formula are: Some plausible suggestions for the parameters in the formula are: To translate from the sale value of land to the rental value of land, one employs the fact that the rental Tprice ofFland, as N proportionM A X R E O R M I a V I E T N A of the selling price of land will be approximately I + T – G, so that the total rental value of the private land in the city is 2ï?°AP R ( I  T  G) f2 Some plausible suggestions for the the formula the Some plausible suggestions for the parameters inparameters inare: formula are: A A ½ – fraction of a cityathat is private land land ½ – fraction of city that is private P $17 billion/sq. km. (Ho Chi Minh City); $8 billion/sq. km (Hanoi) – peak price P $17 billion/sq. km. (Ho1Chi Minh City); $8 billion/sq. km (Hanoi) – peak price of land (divide of land (divide by million to get price per square meter) by 1 million to get price per square meter) f 0.7 – exponential rate of decline of land prices (factor of ½ per kilometer) f I 0.7 – 15 percent – rate of decline of land prices (factor of ½ per kilometer) exponential interest rate in the economy T 0.1 percent – current annual rate of taxation on sale price of land I 15 percent – interest rate in the economy T G 0.1 percent – current annual rate of taxation on sale price of land 8 percent – rate of growth of the economy G 8 these parameter values, of estimated total Withpercent – rate of growth thethe economy rental value of land is US$7.7 billion per year these Chi Minh City and US$3.6 billion per year for Hanoi. The introduction of land value With for Hoparameter values, the estimated total rental value of land is US$7.7 billion per year for Ho taxation would generate one force that for Hanoi. these numbers up and another taxation would Chi Minh City and US$3.6 billion per year would push The introduction of land value that would push them down. The greater efficiency that comes with removing taxes from commerce generate one force that would push these numbers up and another that would push them down. The would push that comes with unprofitability from commerce would push is taxed would greater efficiencyland rents up. Theremoving taxes of land speculation when landland rents up. The mean that a land speculation when land is taxed would mean that substantial land rents unprofitability of substantial amount of land would become available forause, pushingamount of land down. Allowing for the greater influence rents down. Allowing for the greater influence of would become available for use, pushing landof the second factor, we estimate the total land rent the second factor, we estimate the total land rent in Ho Chi Minh City and Hanoi to be $9 billion per year in Ho Chi Minh City and Hanoi to be $9 billion per year (allowing for a 20 percent negative (allowing for a 20 percent negative error margin). We estimate the total land rent for the rest of error margin). We to be US$1 total land year, for a rest of US$10 billion per year in land Vietnam, conservatively,estimate the billion per rent for thetotal of Vietnam, conservatively, to be rent US$1 billion the whole country.per year, for a total of US$10 billion per year in land rent the whole country. The total amount of rent in the country would include not only land rent, but also the rent The total amount of rent in the country would include not only land rent, but also the rent from oil from coal extraction, spectrum rental, fishing rights, and other sources of rent. Oil of rent. extraction,oil extraction, coal extraction, spectrum rental, fishing rights, and other sourcesrevenue was Oil revenue 2006 (plus billion billion (plus US$3.8 billion in corporate income taxes from US$8.0 billion inwas US$8.0 US$3.8in 2006 in corporate income taxes from the oil and gas industry); the oil higher industry); should now be much prices should now be that rent is great enough with today’sand gas oil prices itwith today’s higher oilgreater.itThus it is likely much greater. Thus it is a stroke rent is great enough that, with a taxes of commerce and could scrap in revenue- that, withlikely thatof a pen, Vietnam could scrap all stroke on a pen, Vietnam replace themall taxes on commerce a free market based, efficient and equitable system of market based, efficient neutral fashion withand replace them in revenue-neutral fashion with a freetaxing land values, and thus and sustainable growth taxing land values, and thus unleash a equitable system ofpotential without precedent. unleash a sustainable growth potential without precedent. 22. Sharing the Revenue from Rent Collection among Levels of Government 231 There are two philosophies that can be employed in designing a system for sharing the revenue from rent collection among levels of government. One can design a simple system or one can design a system that is intended to reward jurisdictions according to the efficiency with which they manage their affairs. A Simple System One simple system would use a rule of revenue sharing in thirds (1/3). In terms of transparency, local knowledge, and accountability land rent is best collected at local level, so that is the level that would collect rent. The locality would keep 1/3, which would be shared 264 P R O P E RT Y A N D L A N D TA X E S between communes and other subprovincial governments and pass on 2/3 to the province (=33 percent of total for the local community). The province would keep one third of what it received from the local governments and pass on two thirds to the national government (= 22 percent of the total for the province and 44 percent of the total for the national government). Each level of government would pay its expenses from its share of rent and distribute the excess revenue back to the population as an equal dividend among its citizens. In this way the poorest farmer in the most distant village in the most distant province would receives the same national dividend as the richest rich person in central Hanoi. To the one it might be a year’s worth of labor; to the other it might not pay the maid. In addition to the revenue from a traditional land tax, the central government would distribute revenue from other rent generating privileges such as mineral royalties and licenses from such things as the electromagnetic spectrum, hydropower licenses, airport landing slots, and fisheries. (Where the extraction of mineral deposits, such as coal, means having one’s province turned upside-down, polluted, and degraded, additional compensation for the local people who endure those costs is appropriate.) Therefore, the national divided might reasonably be expected to contribute the biggest share in a citizen’s dividend, further emphasizing the equal rights of all citizens to the benefits of the nation’s natural wealth. Good governance includes an understanding that land is “under ownership by the entire people.â€?101 A table according to the plan outlaid above would look like this: Three-Level Government Rent Local Province National Share Share Sharea Initial Share 100% 67% 15% Share to keep and dispose of according to principles of good governance 33%  22% 44% a. Not including revenues from natural resources. The only way to sabotage the egalitarian properties of a land tax would be by bad assessment. Therefore assessments should be public and frequent. Tax delinquency should be dealt with by levying penalties for late payment and by taking property and selling it at auction when payment is late enough to raise reasonable doubt as to whether the title holder will pay.102 101 Vietnamese Constitution, articles 17 and 18; and Land Law 2003, articles 1 and 5. 102 Making land tax delinquency public makes use of the shame principle to reduce the need to use force. However, there are cases when it is unreasonable to dislocate a person, even though he or she is unable to support the community in the same way as others. Such persons should be dealt with on a case-by-case, compassionate basis. Tax deferral until death provides a compromise that is often satisfactory. If no compassionate relief has been granted, a payment delinquency of 90 days should raise a red flag. After 180 days the title should be auctioned by a “Vickreyâ€? (second price) auction. Any improvement would be sold in the auction. The proceeds of the auction would be used to pay back taxes and administrative costs. Any remainder should be reimbursed to the delinquent former possessor. 265 TA X R E F O R M I N V I E T N A M A System with Efficient Incentives One limitation of the rule of thirds is that it does not give localities and provinces an incentive to make worthwhile investments in infrastructure. If a locality spends VND 2 billion on an infrastructure improvement that raises rent by VND 3 billion, then the locality is allowed to keep only VND 1 billion, making the investment worthwhile for the nation but not for the locality. To motivate localities to make all worthwhile local infrastructure investments, each locality must be allowed to keep all of the increase in rent that is caused by its investments. This does not mean that the locality must be allowed to keep all of the rent of land. Rather, it means that size of the payment of a portion of rent from the locality to the province must be independent of the investments by the locality. Similarly, the size of the payment of a portion of rent from the province to the national government must be independent of investments made by the province. To achieve both fairness and efficiency among provinces and localities, there should be a procedure by which the national government assesses the “unimproved valueâ€? of each province and each province assesses the unimproved value of each locality in the province. The unimproved value of a province or locality would be an assessor’s appraisal of what the rental value of the province or locality would be as a place for potential settlement and improvement if there had been no investments in the place. A province’s rental payment to the national government would be this unimproved value minus enough money to pay a citizen’s dividend to every citizen in the province. The province would be allowed to keep all of the increase in rental value that came from investments in infrastructure, and would therefore have the funds to finance them. The province would also have the money for a citizen’s divided for each of its citizens. If a province did not have enough money for a full citizen’s dividend for all of its citizens when it paid nothing to the national government, then the national government would make a payment to the province from its revenue from rich provinces. A locality’s payment to its province would be the unimproved rental value of the locality, minus enough money to pay a citizen’s dividend to every citizen in the locality. Thus the locality would have the money to pay for worthwhile infrastructure investments and to pay a dividend to every citizen. If a locality did not have enough money for a full citizen’s dividend for all of its citizens when it paid nothing to the province, then the province would make a payment to the locality from the province’s other revenues. 23. Should Land be Rented or Sold? Since many countries use systems of “private ownershipâ€? of land rather than regarding land as the common heritage of the citizens and using a system of use rights of parts of the common heritage, it is appropriate to discuss the differences between private ownership of land and rights to use parts of the common heritage. Rental and sale of land can be made virtually equivalent, if the rental is for a long enough period of time at a fixed rent. Thus economic theory does not provide a basis for asserting 266 P R O P E RT Y A N D L A N D TA X E S that one or the other is more appropriate. But each designation has a different set of political and psychological consequences. Ownership suggests a greater degree of individual control and responsibility. These qualities make ownership attractive. On the other hand, if those who use land are called its owners, there will be a greater tendency to believe that it would be unjust to increase the amounts that owners pay for the use of land (taxes) simply because the rental value of land has increased. But if the taxes are to not increase as rental values increased, then the economy will become burdened with other taxes, those who own land will gradually become a privileged class, and future generations will lose the opportunity to have access to land on the same terms as the current generation. Calling the users of land renters emphasizes the idea that the land of the nation is the common heritage of all generations, and makes it easy to understand why the users should pay more when rental values rise. But to make rental work effectively, those who have land use rights must have options of indefinite duration to continue using the land they rent, provided that they are willing to pay up-to-date rental values. It will also be necessary to have security bonds against land abandonment, so that people will have an incentive to maintain the value of the land they use. Since neither “rentalâ€? nor “saleâ€? has connotations that fit the best way for a nation to manage its land, a new term is needed to describe appropriate land relations. The term in English that seems most appropriate to us is “private possessionâ€? of land. Whatever term is selected, it will be important for the public to acquire an understanding that a person with land use rights: • Has the right and responsibility to decide how land is to be used • Is entitled to continue using the land for as long as he or she wishes, upon continuing payment of the up-to-date rental value of similar unimproved land • Is entitled to transfer the right to use the land to anyone else on whatever terms are mutually agreed. 24. Dealing with Land that Has Been Sold by the State Some title holders have bought land from the government and therefore have already paid at least some of the land rent the land can command today. Out of fairness, and to avoid double taxation, a credit in the amount of the payment already made should be given to those who have bought land use rights from the government. To avoid major revenue shortfalls and to soften the transition it might be suggested that the credits could only be applied to a maximum of 20 percent of the new annual, market based tax that is due. The remainder of the credit would be carried forward with interest to future years. If land use rights were bought from private parties, the state would not have a responsibility for the possibility that the current possessor had paid too much to the previous possessor, but only for any original purchase from the state. With a record of the amount paid to the state, a corresponding credit would be granted to the current possessor. 267 TA X R E F O R M I N V I E T N A M 25. Securing the Natural Value of Land by Bonds There is a possibility that land leases, whether they are short term or long term, might induce title holders to deplete or pollute the soil or abandon the site altogether as their improvements come to the ends of their economic lives.103 To secure the public against inheriting such burdens, title holders might be required to make security deposit, in addition to their rent payments, similarly to the deposits required of those who rent apartments. To elaborate, a person who was paying rent equal to the full rental value of unimproved land, and who owned a building that had come to the end of its economic life, would have an economic incentive to leave the building on the site and disappear, rather than pay the cost of demolishing the building. The natural way to deal with this potential problem is to require every user of land to place in the hands of the government a specified value of government bonds, as a guarantee that the land would be left in a condition no worse than unimproved land. In the event of abandonment, the value of the bonds would be used to restore the land to an unimproved condition. To avoid the need for an individual determination of the required size of each bond, there could be a general rule that the required bond would be such as would pay interest equal to a specified fraction – say 20 percent—of the rent of the land. Ownership of the bond that guaranteed clearance of the land would run with the land. Once such a bond had been provided, the amount that the user of the land would be required to pay each year would be the remaining 80 percent of the rental value of the land. If the rental value of the land were to rise or fall, the percentage of the rent that would be paid by the security bond would be different from 80 percent. The amount of rent due would always be the full rental value of the land, minus the amount of interest provided by the security bond. When the land was transferred to a new user, the amount of the security bond would be adjusted so that the interest from the bond paid 20 percent of the current rent of the land. To provide the users of land with a source of bonds, it would be sensible for the government to issue bonds like British “Consols.â€? These bonds have no maturity date, but simply pay a specified amount of interest indefinitely. For activities, such as the operation of power plants and chemical factories, that held the potential of causing costs greater than the regular bond, additional bonds could be required. 103 This is not only a problem where land is held in common by the state. In fact, it is probably much worse in the West where many industrial sites have been abandoned. For example, there are about 100,000 contaminated land sites in the U.K. (1993). A detailed survey by the (then) Department of Environment identified 39,600 hectares of land “so damaged by industrial or other development that it is incapable of beneficial use without treatment.â€? Half of this land is in urban areas. Total contaminated land, according to the Royal Commission on Environmental Pollution, could cover up to 200,000 hectares. The cost of dealing with this toxic legacy could be as high as 10 billion. Many cities in the United States are also obstructed to rejuvenate derelict sections (Detroit, Philadelphia, the Bronx, etc.), as the land does not automatically come back into the possession of the state when tax payments default. Crumbling buildings and brownfields thus sit afoul for many years, further impacting negatively the area, depriving the community of rapidly gaining access for clean-up and reconstruction. 268 P R O P E RT Y A N D L A N D TA X E S For agricultural uses of land, the equivalent of leaving remnants of use on the land is to deplete the soil or to use pesticides and chemical fertilizers to such a degree that the land requires extra effort before it yields as much as land in the condition provided by nature. Bonds required of agricultural users of land would ensure that users were not motivated to deplete or spoil the soil and then abandon the land. A user of land would have the option, at any time, of selling the improvements on the land and the security bond, for whatever price someone else was willing to pay. If the user wished to cease using the land without finding another user, that too would be permitted. The user would simply be required to give some specified advance notice and leave the land in a condition no worse than unimproved land. Upon relinquishing the land in adequate condition, the user would be entitled to the return of his security bond. 26. Implementing Land Value Taxation Public recovery of the full rent of land is the most reliable way to increase and sustain the wealth of a nation. Implementing an efficient and equitable system of public finance based on collection of the full rent of land is an expression of good governance. Implementation requires visionary leadership and public awareness, education, and support, resulting in a sustained effort to overcome the predictable antipathy of government officials and vested interests.104 An impressive simplification of current laws is foreseeable. Unlike other nations, Vietnam has great advantages arising from its tradition of commitment to respect for human rights and alleviation of poverty. Furthermore, Vietnam has the opportunity to learn from other nations that have undergone transformation from centrally planned economies to market economies.105 The nation’s leadership has repeatedly committed itself to reform. Its SEDP clearly embraces a sustainable, prosperous future with no room for corruption, speculation, or special, uncompensated privileges. Raising the issue of land taxation in public discourse is the first step to its introduction. Such discussion can be expected to have beneficial effects in the land market, as speculators get driven off by the prospect of high taxes on land while real investors are attracted by the prospect 104 Sir Winston Churchill, in an address to the House of Commons on May 4, 1909, said: “Nothing is more amusing than to watch the efforts of land monopolists to claim that other forms of property and increment are similar in all respects to land and the unearned increment on land.â€? [[http://www.grundskyld.dk/2-churchill-uk.html]] In the 21st century, global tax consultants and a myriad of other representatives of privilege have joined hands with land monopolists to shift the source of public revenue from land to labor and capital. 105 Unfortunately, most former communist and socialist countries have simply copied the public finance practices of western nations. While their market liberalizations have freed entrepreneurial forces and stimulated commerce, thereby greatly increasing the productivity of their economies, these nations have also privatized land and other natural resources, creating privileges without proper public compensation. Passing up the chance to get public revenue from land, they have focused taxes on labor and capital, limiting the benefits that they can get from their market reforms. 269 TA X R E F O R M I N V I E T N A M of low taxes on commerce. The steps that must be taken to introduce land value taxation are listed below. Once the political leadership has announced its intention to introduce land value taxation, the remaining points can be addressed simultaneously. • Bring the land value tax option to the table.106 • Declare the political will to introduce land value taxation, display leadership for its introduction, and immediately enact supporting decrees. • Start recovering land rent on the basis of current available assessment records. • Complete cadastral records, including updated assessments and cadastral maps, and the issuance of LURCs. Ensure the data access to all stakeholders on equal terms, taking advantage of the opportunities the web has to offer, including the use of Geographical Information Systems (GIS).107 • Simplify, alter, or nullify obstructing and obsolete laws, decrees, and regulations.108 106 UN Habitat Action Agenda Section B. 56(h) to which all UN member states agreed: Consider the adoption of innovative instruments that capture gains in land value and recover public investments. The Vancouver Action Plan – the 1976 founding document for UN Habitat – states: “Social justice, urban renewal and development, the provision of decent dwellings and healthy conditions for the people can only be achieved if land is used in the interests of society as a whole…. Excessive profits resulting from the increase in land value due to development and change in use are one of the principal causes of the concentration of wealth in private hands. Taxation should not be seen only as a source of revenue for the community but also a powerful tool to encourage development of desirable locations, to exercise a controlling effect on the land market and to redistribute to the public at large the benefits of the unearned increase in land values… The unearned increment resulting from the rise in land values resulting from change in use of land, from public investment or decision or due to the general growth of community must be subject to appropriate recapture by public bodies (the community).â€? 107 Taking advantage of numerous past and ongoing government- and ODA-funded programs in the area of land administration and forming the foundation for the effective and equitable operation of the land/property taxation system, despites of numerous efforts, by 2006 only about 63 percent of parcels of eligible agricultural land, 49 percent of parcels in eligible forest land and 70 percent of eligible urban households have been granted with LURCs. Cadastral records including maps—the basis for Vietnam’s torrent-based land administration is incomplete and not updated. 108 In addition to the nullification of the conflicts between the Land and Housing Laws which the recent Session of the National Assembly has paved the way for, the current land law could be simplified to half of its length. (For example, the differentiation of land classes and lease terms are no longer important, other than for general land use planning efforts. Lease durations would not need to be differentiated as leases would be permanent—provided the annually adjusted, market value reflecting land tax is being paid. Thus, for example, Decree 84/07, which states that foreign investors can now lease land for 70 years, with extensions available at no extra cost could be scraped. Similarly, many other laws and decrees referring, for example, to the current tax and investment laws could be eliminated, as land rent will substitute for current revenue and foreign direct investment would transcend onto the nation naturally). However, it is generally easier to introduce a law than to repeal one, as vested interests have gotten accustomed to their privileges. Thus, even greater leadership is required to induce drastic simplification of current law. 270 P R O P E RT Y A N D L A N D TA X E S • Encourage provincial and local government to shift to land value taxes to finance their duties under decentralization. Let them start collecting land value taxes while lowering other taxes and tariffs in a revenue-neutral fashion.109 • Turn land registration, public assessment, and tax payment into a One-Stop-Shop (OSS) experience. Establish and operationalize a system for the resolution of land-related complaints and disputes, especially related to assessment.110 • Absorb feedback so that every step can be adjusted and fine-tuned, based on that feedback. • Declare the whole nation a tax-free zone, except for socially harmful things such as pollution, congestion, and other privileges. • Deal with cases of personal hardship on a compassionate basis. Where applicable, offer payment deferment. • Develop a citizen’s dividend system, financed by surplus land rent. • Start reaping the benefits of public rent recovery combined with low or no taxes on income and commerce and witness the nation prosperity in a efficient, equitable, and sustainable fashion. Of course there will always remain some open challenges and questions, but unless the initial steps are taken they will not even have a chance to arise. Instead, the possibility of real reform will remain buried in every day’s emergency plans of standard political procedures to deal with the symptoms of a deeply flawed, inefficient and inequitable system of funding public revenue, of a continuous fight of corruption and poverty, and of endless efforts of economic micromanagement. The biggest flaw in any open market economy, committed to the principles of good governance, remains the failure to recover the full economic rent for the benefit of the public. 109 The central government could encourage provinces and local jurisdictions to shift to a revenue system based on land value, by permitting them to declare their constituencies free of all income taxes as long as they provided the central government the same amount of tax revenue (adjusted for GDP growth) as in the year before the shift, using the revenues they now retrieve in a efficient and equitable fashion from the full market value land rental collection This push-and-pull approach would encourage individual provinces to move quickly, without the central government having to appear overly authoritarian. By inviting provinces to embrace the potential for the socioeconomic benefits of taxing land values, the central government could induce provinces to implement and benefit from it faster than any compromise-seeking central rule could possibly achieve. In addition, localities or provinces that would make use of it would serve as pilot programs for the ultimate national implementation, inducing others to copy their policies and delivering valuable information for fine-tuning the system as its benefits become more and more evident. 110 This is critical as feedback to the assessed value will improve overall market accuracy and thus limit future complains. Furthermore, the implementation of collecting the annual land value goes in parallel steps with the improvement of the land administration system. Some complaints and disputes are inevitable. However, as the “winningâ€? landowner then owes the annual rent of the site—nullifying any future speculative gains—and limiting the interest in possession to the economic feasibility of the projected return on the use, a noticeable decrease in title claims can be expected. This would improve the situation with the number of complaints to the government currently encounters, whereof reportedly 80 percent are related to land. 271 TA X R E F O R M I N V I E T N A M As demonstrated, the implementation of full rent recovery supports many current policy goals, all complementing each other. Thus, land value taxation cannot be analyzed based on static economic observations but has to be seen in the full light of its dynamic compilation, rising simultaneously in an upward spiral economic efficiency and equity. Thus, requesting a definite answer to all possible and impossible questions without daring to embrace the principles and to take the initial step is like asking on how to reach personal enlightenment without heartfelt commitment. For any nation, province, or locality that is committed to reform, liberty, and pervasive sustainable prosperity, raising revenue by taxing the private use of the common heritage in land is the economic imperative. “Social Reform is not to be secured by noise and shouting; by complaints and denunciation; by the formation of parties, or the making of revolutions; but by the awakening of thought and the progress of ideas. Until there be correct thought, there cannot be right action; and when there is correct thought, right action will follow.â€?111 111 Henry George, Social Problems, 1883, Chapter 22. This time might have arrived now, since in the United States the debate around the constitutionality of (income) taxation is currently taking the form of a popular movement similar to the times of the Boston Tea Party (for more information google and YouTube, for example, Aaron Russo and Ron Paul). 272 Annex Annex 1: Vietnam: Key Economic Indicators   2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Output, Employment and Prices Real GDP (% change, 6.9 7.1 7.3 7.8 8.4 8.2 8.5 6.2 5.3 6.8 y-o-y) Industrial production 14.6 14.5 15.5 16.0 17.1 16.8 16.7 13.9 7.6 14.0 (% change, y-o-y) Unemployment (%, 6.3 6.0 5.8 5.6 5.3 4.8 4.6 4.7 4.6 4.4 in urban areas) CPI (% change, 0.8 4.0 3.2 7.8 8.8 6.7 12.6 19.9 6.5 11.8 y-o-y) Foreign Trade, BOP and External Debt Trade balance -1.1 -3.0 -5.1 -5.5 -2.4 -2.8 -10.4 -12.8 -8.3 -7.1 (billions US$) Exports of goods 15.0 16.7 20.2 26.5 32.4 39.8 48.6 62.7 57.1 72.2 (billions US$) (% change y-o-y) 4.0 11.2 20.8 31.3 22.5 22.7 21.9 29.1 -8.9 26.4 Imports of goods 16.2 19.7 25.3 32.0 34.9 42.6 58.9 75.5 65.4 79.3 (billions US$) (% change y-o-y) 3.2 22.1 28.0 26.5 15.7 22.1 38.3 28.1 -13.3 21.2 Current account 0.6 -0.7 -1.9 -1.6 -5.6 -0.2 -7.0 -10.8 -6.1 -4.0 balance (billions US$) (% GDP) 2.1 -1.9 -4.9 -3.5 -1.1 -0.3 -9.8 -11.9 -6.6 -3.8 FDI inflows (billions 1.3 2.0 1.9 1.9 1.9 2.3 6.6 9.3 6.9 7.1 US$) External debt 12.3 12.7 13.8 15.6 17.2 19.1 23.0 30.2 38.8 44.5 (billions US$) (% GDP) 37.8 35.0 33.7 33.5 32.7 31.3 32.4 33.4 41.6 42.5 Debt service ratio (% 6.3 10.0 8.5 6.0 5.6 3.0 3.1 2.9 4.2 4.1 exports of g&s) Sources: Vietnam Government Statistics Office; State Bank of Vietnam; IMF; World Bank staff estimates.World Bank staff estimates 273 TA X R E F O R M I N V I E T N A M Annex 2: State Budget Revenues (share of GDP)           Rev. Rev. Est.     2001 2002 2003 2004 2005 2006 2007 2008 2008 2009 A Total revenues and 21.6 22.7 25.8 27.8 28.4 27.1 27.6 28.1 28.1 26.7 grants I Current revenues 21.0 22.1 23.8 25.2 26.1 25.0 24.7 25.4 25.4 24.0 I.1 Taxes 19.1 19.8 20.9 21.7 22.8 23.7 23.2 24.2 24.2 22.3 1 Corporate income tax 6.9 6.9 7.7 8.0 9.0 10.3 9.0 9.1 9.1 6.8 2 Individual income tax 0.4 0.4 0.5 0.5 0.5 0.5 0.6 0.9 0.9 0.9 3 Land and housing tax 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 4 License tax 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 5 Tax on the transfer of 0.2 0.2 0.3 0.4 0.3 0.3 0.5 0.5 0.5 0.6 properties 6 Tax on land use right 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.0 transfer 7 Value added tax 4.0 4.8 5.4 5.4 5.5 5.6 6.1 6.1 6.1 6.4 8 Special consumption 1.3 1.4 1.4 1.8 1.9 1.8 1.5 1.5 1.5 1.8 tax 9 Natural resources tax 1.7 1.6 1.6 2.4 2.5 2.1 1.7 1.8 1.8 1.1 10 Agricultural tax 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11 Export & import tax 3.6 4.1 3.6 3.0 2.8 2.7 3.4 4.0 4.0 4.6 12 Other taxes 0.03 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 I.2 Fees, charges, and 1.9 2.3 2.9 3.4 3.3 1.4 1.5 1.2 1.2 1.7 nontax 13 Revenue from 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 discrepancy of import prices 14 Fees and charges 1.1 1.1 1.1 1.1 1.0 0.8 0.7 0.8 0.8 1.0 15 Rental of land 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.2 16 Others 0.7 1.0 1.8 2.2 2.2 0.4 0.5 0.3 0.3 0.5 II Capital revenues 0.2 0.2 1.5 2.2 1.8 1.7 2.5 2.2 2.2 2.3 III Grants 0.4 0.4 0.5 0.4 0.5 0.4 0.4 0.5 0.5 0.4 Source: MoF 2010. 274 ANNEX Annex 3: Annual Revenue Buoyancy   1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Total Tax 0.22 1.34 1.92 1.98 1.58 1.75 1.60 1.68 1.46 0.77 1.68 -0.51 Revenues Corporate 0.51 1.83 20.92 1.93 0.91 2.77 1.43 2.81 2.85 -0.67 1.20 -4.04 Income Tax Individual 1.82 -0.23 -0.63 1.53 1.30 2.45 1.32 1.34 1.69 3.84 6.81 0.84 Income Tax Land and -2.64 1.12 0.53 -1.90 -0.24 0.05 1.65 1.03 0.87 1.30 0.61 9.83 Housing Tax Licenses Tax -2.62 0.59 1.18 0.39 -0.25 10.50 -2.85 0.22 0.26 0.29 -0.67 1.48 Transfer of -0.61 0.87 -1.93 3.90 1.03 3.72 4.21 -0.15 1.46 6.70 1.04 4.27 Properties Land Use 2.22 1.42 -5.79 5.76 0.79 2.27 5.81 5.19 2.20 8.60 0.95 3.12 Right Transfer Value Added -0.60 -0.46 -0.58 1.65 3.97 2.65 1.07 1.10 1.36 2.12 0.90 2.12 Tax Excise 1.42 0.21 1.95 2.49 1.71 1.89 4.31 1.66 0.23 -0.71 0.17 5.03 Natural -2.00 6.46 8.48 1.53 -0.28 0.91 8.46 1.55 -1.29 -1.08 1.53 -6.29 Resources Tax Agricultural 1.02 -0.84 -1.83 -8.72 -1.15 -10.77 -2.65 -0.82 -1.77 -1.55 -4.62 -6.55 Tax Import- 0.15 -1.55 -1.41 4.23 2.85 -0.54 -1.39 0.10 0.45 4.14 4.43 3.95 Export Tax Source: Own calculation based on MoF data, 2010. 275 TA X R E F O R M I N V I E T N A M Annex 4: Features of a Good Tax System Principles of Taxation: There are two broad principles of taxation. i. Ability to Pay Principle: Each taxpayer is asked to contribute according to his/her ability to pay. The advantage is that the system may be made progressive and it partly takes care of the redistributive function of government. This is the principle behind the income taxes. ii. Benefit Principle: Those who benefit from a good/ service should pay the price. For goods/ services that are rival in consumption or exclusion of nonpayers is feasible, a user charge/ fee can be collected. The financing of such goods and services can be done either directly (toll taxes, social security taxes) or indirectly (taxes on gasoline or automobiles used to finance highways). Equity: A tax system should be “fair.â€? Horizontal Equity means “people, whose circumstances are the same, should pay equal taxes (tax equals equally).â€? This principle demands broad coverage of the tax base and is the source of much of the detail and complexity of a tax law that is required to define a comprehensive tax base and deal with all the exceptional circumstances/ cases. Vertical equity demands that higher tax burden be placed on individuals with a greater capacity to pay in order to “improveâ€? the income distribution in society. Societywide social judgments depend upon and arise out of the political process. Efficiency: It refers to the negative aspects of taxation. For instance, a tax distorts market prices of goods and services, labor (wages), and savings (interest rate) thus introducing distortions in the market. It changes the behavior of consumers and producers, interferes with their choice of consumption, work and savings, and imposes an efficiency cost on the economy. A tax is efficient if efficiency loss is small. The efficiency cost of a tax depends on the price elasticity of demand and supply of the good/service taxed and the tax rate. Taxes such as an income tax or a franchise fee do not change consumer behavior and are least distortional. A high income tax, however, will result in lower investment. An ideal neutral tax would be a poll (head) tax, but it is regressive; again a general sales tax (value added tax) would be nondistortional but regressive. Tax systems should be designed to collect a given amount of revenue with the minimum of efficiency cost to the economy. Revenue Adequacy: It is an important consideration from the point of view of the government. Tax revenues depend on tax rates, tax base or volume of sales and elasticity of demand and supply. Stability: It is a prerequisite for creating a suitable investment climate. Stability of tax revenues is necessary to maintain continuity of fiscal policies and a stable tax system which is crucial for investors. A tax system must be elastic, that is, tax revenues should increase with growth of GNP. To ensure this, appropriate tax rates and tax handles should be used by including the growing sectors of economy in the tax base. Simplicity and transparency: A simple tax system is easily understood by taxpayers and reduces the cost of compliance and administration. Thus, the lack of complexity and discriminatory 276 ANNEX taxes makes the system transparent and easy to comply with. However, over simplification is not desirable otherwise some sectors of economy, specially the self-employed, will remain out of the tax net. Low Collection Cost: High cost of collection reduces net tax revenues available to the government. But some costs may have to be legitimately incurred in bringing the hard-to-tax sectors within the purview of taxation. Expenditures are also justified for taxpayer education as it ultimately reduces cost of compliance. Neutrality: A tax incentive must not distort investment decisions across sectors by favoring particular sectors and in the process promoting low return investments over higher return investments. 277 TA X R E F O R M I N V I E T N A M Annex 5: Documents Examined for Analyzing Major Tax Regimes in Vietnam VAT i. Law on Value Added Tax, 2008 ( No 13/2008/QH 12) ii. Decree No 123/2008/ND-CP of December 8, 2008 detailing and guiding the implementation of a number of articles of the Law on Value-Added Tax iii. Circular No. 129/2008/TT-BTC of December 26, 2008, guiding the implementation of a number of articles of the Value-Added Tax Law and guiding the implementation of the Government’s Decree No. 123/2008/ND-CP of December 8, 2008, detailing and guiding the implementation of a number of articles of the Value-Added Tax Law Corporate Income Tax i. Law on Enterprise income Tax (No. 14/2008/QH12) ii. Decree No. 124/2008/ND-CP of December 11, 2008, detailing the implementation of a number of articles of the Law on Enterprise Income Tax; iii. Circular No.130/2008/TT-BTC of December 26, 2008, guiding the implementation of a number of articles of Enterprise Income Tax Law No. 14/2008/QH12, and guiding the implementation of the Government’s Decree No. 124/2008/ND-CP of December 11, 2008, which details the implementation of a number of articles of the Law on Enterprise Income Tax iv. Circular No.18/2011/TT-BTC of February 10, 2011, Amending and Supplementing Circular No.130/2008/TT-BTC of December 26, 2008, guiding the implementation of the Government’s Decree No. 124/2008/ND-CP of December 11, 2008, which details the implementation of a number of articles of the Law on Enterprise Income Tax Personal Income Tax i. Law on Personal income tax (No. 4/QH 12/2007) ii. Degree No 100/2008/ND-CP of September 8,2008, detailing a number of articles of the Law on personal income tax iii. Circular No 84/TT-BTC of September 30, 2008, guiding a number of articles of the Law on personal income tax and guiding the Government’s degree No 100/2008/ND-CP of September 8,2008, detailing a number of articles of the Law on personal income tax iv. Circular No 62/2009/TT-BTC of March 27, 2009, amending and supplementing the Ministry of Finance’s circular No 84/TT-BTC of September 30, 2008, guiding a number of articles of the Law on personal income tax and guiding the Government’s degree No 100/2008/ND-CP of September 8,2008, detailing a number of articles of the Law on personal income tax Special consumption tax (excises) i. The law on Excise tax (No. 27/2008/QH12) 278 ANNEX ii. Decree No. 26/2009/ND-CP of March 16, 2009, detailing a number of articles of the Law on Excise Tax; iii. Circular No 64/2009/TT-BTC of March 27, 2009 guiding the government’s decree No 26/2009 ND-CP of March 16,2009, which detail a number of articles of the Law on Excise Tax Trade taxes i. Law No. 45/2005/QH11 of June 14, 2005 (Law on Import and Export Tax) ii. Decree No./87/2010/ND-CP of August 13,2010 detailing a number of articles of the Law on Import and Export Tax iii. Circular No 194/2010/TT-BTC of December 02, 2010 guiding the customs procedure, customs monitoring , import and export tax and tax administration on export and import goods Natural Resource tax i. The law on Natural Resource tax ( No.45/2009/QH 12) ii. Decree No./50/2010/ND-CP of May 14, 2010 detailing a number of articles of The law on Natural Resource tax iii. Circular No 105/2010/TT-BTC of July 23, 2010 guiding the government’s decree No 50/2010 ND-CP of May 14, 2010, which detail a number of articles of the law on Natural Resource tax iv. Petroleum Law, (Amended and supplemented in 2000) v. Decree No. 48/2000-CP of December 17, 1996 of the government detailing the implementation of the petroleum law vi. Petroleum Law, (Amended and supplemented in 2008) vii. Circular No.32/2009/TT-BTC of February 19, 2009 guiding the implementation of tax provisions applicable to organizations and individuals conducting oil and gas prospection, exploration and exploitation activities under the Law on Petroleum viii. Decree No. 115/ 2009 of the government amended and supplemented some articles of Decree 48/2000 detailing the implementation of the petroleum law Nontax Regime (Fees and Charges) i. Circular No. 63/2002/TT-BTC of July 24, 2002 ii. Circular 71/2003/TT-BTC of July 30, 2003 iii. Circular 45/2006 of May 25, 2006 Tax Administration i. Tax Administration Law 279 TA X R E F O R M I N V I E T N A M Annex 6: Implications of Joining the WTO for Various Sectors of the Economy Agriculture The WTO’s agriculture agreement includes specific commitments by members to improve market access and reduce trade-distorting subsidies in agriculture from 2005. For developing countries, however, the relaxed commitments essentially imply that net exporting countries do not take actions that harm the food security of vulnerable countries. Sanitary and Phytosanitary Standards The WTO defines these measures as “Sanitary and Phytosanitary measures or regulations— implemented by governments to protect human, animal and plant life and health, and to help ensure that food is safe for consumption.â€? This is an agreement on how governments can apply food safety, quarantine and animal and plant health standards. The standards are based on specifications of international organizations such as FAO. Countries may set their own standards, and are free to adopt standards that are higher than those specified by international organizations, subject to notifications requirements. Textiles Due to the end of the Multi-fiber Arrangement (an agreement that restricted how much of textiles and clothing individual exporting countries could sell to individual importing countries) in 2005, trade in textiles is now part of the overall WTO framework. Technical Barriers to Trade Technical regulations and industrial standards are important, but they vary from country to country. Having too many different standards makes life difficult for producers and exporters. If the standards are set arbitrarily, they could be used as an excuse for protectionism. Standards can become obstacles to trade. The Technical Barriers to Trade Agreement (TBT) tries to ensure that regulations, standards, testing and certification procedures do not create unnecessary obstacles. “Developing country members are not expected to use international standards as the basis for their own technical regulations, standards, or testing methods, if the international methods are not appropriate to the country’s individual development situation.â€? [Article 12.4] Subsidies The agreement on subsidies applies to agricultural as well as industrial goods and regulates the use and the action countries can take to counter the effect of subsidies. Under this agreement, a country can use the WTO dispute settlement mechanism to seek the withdrawal of a subsidy. Member countries have also been given the freedom to launch investigations and charge countervailing duties on subsidized imports if they are found to hurt domestic industry. 280 ANNEX Dumping While countries are allowed to impose antidumping duties, such action is based on the member being able to prove dumping, calculate its impact, and show injury to domestic industry. Customs Valuation This basically amounts to accepting the transaction value as the basis for valuation, with the GATT valuation code as backup. Rules of Origin These rules are the basis for assessing duty, imposing antidumping and countervailing duties, implementing safeguards and origin marking. All members are obliged to frame well-publicized rules of origin that specify how the country of origin of imported goods is determined, with the ultimate objective of having a harmonized set of rules in place for all countries. Preshipment Inspection Some countries prefer to insect imported goods before they are shipped from the exporting country to the importing country to ensure quality control. This practice, however, is becoming increasingly rare. Recognizing the practice among developing country members, the agreement calls on developed countries to provide developing country members technical assistance. Import Licensing This agreement requires members to have a set of simple, transparent and predictable procedures. The agreement also requires governments to publicize the basis for granting such licenses. Investment This applies to trade in goods, and contains prohibitions on applying certain trade-distorting measures, subject to relaxations for developing countries for protecting infant industries or balance of payment considerations. Safeguards Members may restrict temporarily certain imports if a domestic industry is threatened with injury due to a surge in imports, subject to an agreed definition of “surgeâ€?. The focus is on transparency and established, predictable rules, as in the case of dumping Information Technology The agreement provides for members to eliminate duties on IT products, developing signatory countries were to do so by 2005. State Trading Enterprises Article XVII of GATT requires that state trading enterprises (STEs) function on the basis of nondiscrimination, and requires members to notify STEs to the WTO. Source: website www.taxindiaonline.com. 281 TA X R E F O R M I N V I E T N A M Annex 7: Tax-Royalty Regime versus Production Sharing Arrangements The fundamental difference between a tax-royalty regime (concessionary system) and production sharing arrangement (contractual system) (PSC) is the approach toward the issue of ownership. The tax-royalty regime or the concessionary system allows individual ownership of resources. When a government enters into an agreement with an investor for natural resource exploitation, there is a transfer of title in favor of the investor. In return, the investor pays taxes and royalty. Under the production sharing agreement or the contractual system, the government retains the ownership of the natural resource. The mineral or oil companies have the right to receive a share of production or revenues from the sale according to the terms of the contract. The basic cash flow under a PSC with a royalty would look like the following.112 Gross Revenues = Total oil and gas revenues Net Revenues = Gross revenues - Royalty Cost recovery or ‘cost oil’ = = Operating costs + Capital costs (depreciation, depletion and amortization) including intangibles + Unrecovered costs carried forward with interest. Profit oil = Net revenues - Cost recovery Contractor profit oil = Profit oil Ð¥ Contractor percentage share Government profit oil = Profit oil Ð¥ Government percentage share Taxable income = Gross revenues - Royalties - Operating costs - Capital costs (depreciation, depletion and amortization) including intangibles + Investment credit (if allowed) – Government profit oil Net Cash flow = Gross revenues - royalties - Operating costs - Capital expenditure – Government profit oil - Taxes PSC vs. Normal Tax/Royalty Regime: An Illustration A comparison of pure royalty/tax regime and a PSC is illustrated with the help of the following numerical example. Royalty/Tax System Production Sharing Arrangement One barrel of Oil (US$50) One barrel of Oil (US$50) Contractor Share: 10 percent royalty Contractor Share : 20 percent royalty Government Share: $5 Royalties: $10 Net revenue $40 Net revenue $45 112 These examples are based on flow diagrams from Daniel Johnston (1994). 282 ANNEX Deductions $22 Cost recovery 40 percent (limit): $20 Taxable income $18 Provincial Taxes 10 percent $1.8 Amount available for profit sharing $45 - $20 = $25 Net Income $16.20 Profit sharing 40 percent/60 percent (taxable) Federal Income Tax (40 percent) $6.48 $10 $15 40 percent income tax $4.0 Net income $9.72 Net income $6 Total Gov. revenue $18.28 or 37 percent Total Gov. revenue $24.0 or 48 percent Total contractor’s take $31.72 or 63 percent Total contractor’s take $26 or 52 percent 283 TA X R E F O R M I N V I E T N A M Annex 8: Suggestions about Decentralization of Charges and Fees Centralized charges and fees: Charges • Charges in the field of information and communication • Charges in the fields of security, social order and safety • Charges in the judicial field • Charges in the fields of finance, banking and customs. Fees: • State management fees related to citizens’ rights and obligations • State management fees for national sovereignty. Charges that should be levied at the provincial and/or municipal level: • Charges in the fields of agriculture, forestry and aquatic resources • Charges in the fields of industry and construction • Charges in the fields of commerce and investment • Charges in the fields of communications and transport • Charges in the cultural and social fields • Charges in the fields of science, technology and environment. Fees at the provincial or municipal level: • State management fees related to production and business. Charges at purely local level: • Charges in the field of education and training • Charges in the healthcare field. Fees at local level: • State management fees related to the right to own or use property • State management fees in other fields. Within some broad category, a few charges/fees should be at one level of government while some others should be assigned to another level. For instance, in the category of “charges in the fields of security, social order and safetyâ€?, the following two can be at the central level. • Charge for technical inspection of machinery, equipment, supplies and substances subject to stringent safety requirements • Charges for security, social order and safety. The “charges for verification of papers or documentsâ€? could be delegated to the provincial government while the “vehicle watch chargeâ€? could be at the local level. There are several examples of this kind and it only shows that the entire list needs close scrutiny and proper assignment done rather than just looking at the broad categories. 284 ANNEX Annex 9: Tenets of a Good Tax Administration Objective The principal objective of a modern tax administration should be to collect the proper amount of tax in an efficient and equitable manner, with a high degree of integrity that respects the rights of taxpayers. One aspect of this objective is promoting voluntary compliance, which is the most cost-effective way of collecting taxes. Voluntary compliance involves the accurate self-assessment by taxpayers of tax owed, the filing of accurate tax returns, and the payment of tax liability on time. Encouraging voluntary compliance requires a concerted effort by the tax administration (and other government agencies) to reduce the burden the tax system imposes on taxpayers or to make it as easy as possible for taxpayers to comply with the tax laws. It requires laws and regulations that are clear, consistent, and fair, tax forms and instructions that are understandable and helpful, and taxpayer education and assistance programs to help taxpayers comply with their tax obligations. Expectation of sanctions and fairness The other aspect of a good tax administration is establishing a credible expectation on the part of taxpayers that the tax administration will detect noncompliance, and that it will impose in an impartial manner sanctions that are commensurate with the infraction. In particular, it requires a tax administration organized by function (rather than by tax), not only to take advantage of the efficiencies of specialization, but also to reduce the opportunity for corruption by involving more than a single tax official in the relationship between the tax office and the taxpayer. It requires uniform procedures that are integrated with computer and communications technology and flexible enough to respond to changes in the technology over time. In addition, it requires computer systems designed to automate certain tasks, such as corresponding with taxpayers, verifying and matching information, and selecting tax returns for audit. Organizational Mission The tax administration should develop a written mission statement outlining its goals for the future. The mission should be to collect the proper amount of tax in an efficient and equitable manner, with a high degree of integrity while respecting the rights of taxpayers. The mission should be well publicized and understood by each functionary of the department. Functional Structure The tax administration should be preferably reorganized along functional lines rather than according to tax types in order to take advantage of the efficiencies of specialization and to reduce the opportunities for corruption. Tax Law The tax administration can be only as good as the laws that it implements. So a comprehensive review of the country’s tax laws, regulations, and administrative authorities with an eye toward enacting changes that will improve their transparency, consistency, efficiency, fairness, and administration is the starting point. 285 TA X R E F O R M I N V I E T N A M Human Resources The tax administration should develop a comprehensive human resource program to support the functionally oriented organization. Such an organization requires people with the specific skills necessary to promote voluntary compliance through the provision of services to taxpayers and the effective detection of noncompliance. Taxpayer Education and Assistance A comprehensive taxpayer education and assistance program is an essential element of a modern tax administration system. Such a program promotes voluntary compliance by making taxpayers aware of their rights and responsibilities under the law and by making it as easy as possible for taxpayers to comply with the law. Audit The tax administration should develop a comprehensive audit program that maximizes voluntary compliance for a given amount of audit resources by establishing a credible expectation on the part of taxpayers that noncompliance will be detected and appropriate penalties and interest will be assessed. Appeals In order to build a relationship of mutual trust, the tax administration should develop an administrative appeals process that is fair, expeditious, and transparent. There should be a single avenue of appeal, the first level of which is administrative, and the second level of which is judicial. Collection The tax administration should establish a collection function to collect taxes that are due but not paid and to identify those who have stopped filing returns (“stop-filersâ€?) and those who have never filed returns (“nonfilersâ€?). Processing and Information Systems The tax administration should develop or acquire and modify existing computer processing and information systems to streamline certain tax administration activities. Autonomous Revenue Authority In due course, there should be a move toward establishing a semiautonomous revenue authority as a practical means of achieving the goals of revenue administration reform. This gives the administration adequate autonomy and administrative tools to perform its duties efficiently, fairly and effectively. 286 ANNEX Annex 10: Summary of Statistics of Variables Used in Empirical Analysis, 1992–2002 Number of Standard Variables Mean Minimum Maximum Observations Deviation GDP/Capita 662 13,572.5251 15,786.8781 300 84487 (real)a 662 -4.505952381 1.888415049 -10 0 Corruptionb Revenue/GDP 662 27.5669643 11.8921307 0.3 78.5 (%)c Tax Revenue/ 662 18.07704918 6.648668875 7.3 37.2 GDP (%) Source: World Development Indicator 2005; International Country Risk Guide (ICRG), available at [http://www. countrydata.com/]. a. GDP per capita, PPP (constant 2000 international US$). b. The corruption index from ICRG ranges from 0 (most corrupt) to 6 (least corrupt). To make it consistent and easier to compare with Tanzi, Davoodi (1997), we have rescaled the ICRG index by multiplying it by -10/6 so that the index ranges from -10 (least corrupt) to 0 (most corrupt). c. Revenue, excluding grants (percent GDP). 287 TA X R E F O R M I N V I E T N A M Annex 11: Corruption and Revenue Leakage: Cross-country Estimates ICRG Rescaled ICRG Revenue Loss Tax Loss Country Corruption Corruption (%GDP)c (% GDP)d Indexa Indexb Gambia, the 3.13 -5.21 -0.07 -0.06 Brazil 3.08 -5.13 -0.14 -0.12 Botswana 3.05 -5.08 -0.18 -0.15 Zambia 3.03 -5.05 -0.20 -0.17 Cote d’Ivoire 3.01 -5.01 -0.23 -0.20 Morocco 3.00 -5.00 -0.24 -0.20 Uruguay 3.00 -5.00 -0.24 -0.20 Peru 2.98 -4.96 -0.27 -0.23 Guatemala 2.97 -4.95 -0.28 -0.24 Trinidad and Tobago 2.96 -4.94 -0.29 -0.25 Senegal 2.95 -4.92 -0.30 -0.25 Oman 2.93 -4.89 -0.33 -0.28 Ecuador 2.93 -4.88 -0.33 -0.28 Malawi 2.91 -4.84 -0.36 -0.31 Tanzania 2.90 -4.84 -0.37 -0.31 Tunisia 2.90 -4.84 -0.37 -0.31 Suriname 2.89 -4.82 -0.38 -0.32 Lithuania 2.88 -4.81 -0.39 -0.33 Venezuela 2.86 -4.77 -0.42 -0.35 Albania 2.86 -4.76 -0.43 -0.36 Yemen, rep. 2.85 -4.75 -0.44 -0.37 Turkey 2.83 -4.71 -0.47 -0.40 Philippines 2.80 -4.66 -0.50 -0.43 Latvia 2.77 -4.61 -0.54 -0.46 Argentina 2.75 -4.59 -0.56 -0.48 Mexico 2.74 -4.57 -0.57 -0.49 Bolivia 2.71 -4.52 -0.61 -0.52 India 2.70 -4.51 -0.62 -0.53 Papua New Guinea 2.70 -4.51 -0.62 -0.53 Croatia 2.70 -4.50 -0.63 -0.54 Kuwait 2.69 -4.48 -0.64 -0.55 Ghana 2.68 -4.47 -0.65 -0.56 Jamaica 2.67 -4.46 -0.66 -0.56 Guyana 2.61 -4.36 -0.74 -0.63 Cuba 2.59 -4.31 -0.78 -0.66 Egypt, Arab Rep. 2.59 -4.31 -0.78 -0.66 288 ANNEX ICRG Rescaled ICRG Revenue Loss Tax Loss Country Corruption Corruption (%GDP)c (% GDP)d Indexa Indexb China 2.55 -4.26 -0.82 -0.70 Kenya 2.53 -4.22 -0.85 -0.72 Thailand 2.53 -4.22 -0.85 -0.72 Burkina Faso 2.50 -4.17 -0.89 -0.76 Cameroon 2.50 -4.17 -0.89 -0.76 Algeria 2.50 -4.16 -0.89 -0.76 Angola 2.48 -4.14 -0.91 -0.77 Zimbabwe 2.45 -4.09 -0.95 -0.81 Kazakhstan 2.42 -4.03 -1.00 -0.85 Vietnam 2.42 -4.03 -1.00 -0.85 Uganda 2.41 -4.02 -1.01 -0.86 Colombia 2.37 -3.95 -1.06 -0.90 Mali 2.36 -3.94 -1.07 -0.91 Korea, Dem. Rep. 2.26 -3.76 -1.20 -1.02 Ukraine 2.23 -3.72 -1.23 -1.05 Serbia and Montenegro 2.21 -3.69 -1.26 -1.08 Pakistan 2.17 -3.61 -1.32 -1.13 Russia 2.16 -3.60 -1.33 -1.13 Indonesia 2.12 -3.54 -1.38 -1.18 Sierra Leone 2.10 -3.50 -1.41 -1.20 Honduras 2.03 -3.38 -1.50 -1.28 Azerbaijan 2.00 -3.33 -1.54 -1.31 Guinea-Bissau 2.00 -3.33 -1.54 -1.31 Panama 2.00 -3.33 -1.54 -1.31 Paraguay 2.00 -3.33 -1.54 -1.31 Qatar 2.00 -3.33 -1.54 -1.31 Saudi Arabia 2.00 -3.33 -1.54 -1.31 United Arab Emirates 2.00 -3.33 -1.54 -1.31 Togo 1.99 -3.32 -1.55 -1.32 Ethiopia 1.98 -3.30 -1.57 -1.33 Armenia 1.89 -3.15 -1.68 -1.43 Moldova 1.88 -3.14 -1.69 -1.44 Lebanon 1.85 -3.08 -1.73 -1.48 Bangladesh 1.83 -3.06 -1.75 -1.49 Haiti 1.72 -2.87 -1.90 -1.62 Niger 1.60 -2.66 -2.06 -1.75 Nigeria 1.59 -2.65 -2.07 -1.76 Myanmar 1.44 -2.40 -2.27 -1.93 Sudan 1.44 -2.40 -2.27 -1.93 289 TA X R E F O R M I N V I E T N A M ICRG Rescaled ICRG Revenue Loss Tax Loss Country Corruption Corruption (%GDP)c (% GDP)d Indexa Indexb Liberia 1.26 -2.10 -2.50 -2.13 Somalia 1.05 -1.76 -2.77 -2.36 Gabon 1.00 -1.67 -2.84 -2.42 Iraq 1.00 -1.67 -2.84 -2.42 Congo, Dem. Rep. 0.61 -1.02 -3.34 -2.84 Source: ICRG; WDI 2007; and authors’ estimates. a. ICRG Corruption Index is a monthly index scaled from 0 (Most corrupt) to 6 (Least corrupt). The ICRG indexes in this column are annualized for the period of 1995–2005. b. Rescaled Corruption Index is derived by multiplying ICRG Corruption Index with (-10/6). Therefore, the number is ranged between -10 (Least corrupt) and 0 (Most corrupt). The rescaling follows Tanzi and Davoodi (1997). c. Revenue loss is a product of difference of country’s corruption from the mean and the regression’s corruption coefficient, b*(Xi - Xbar), when total country’s revenue is a dependent variable, Y. d. Tax Revenue loss is a product of difference of country’s corruption from the mean and the regression’s corruption coefficient, b*(Xi - Xbar), when country’s tax revenue is a dependent variable, Y. 290 ANNEX Annex 12: Summary of Three Methodologies to Measure Shadow Economy Direct (or micro) Approaches to Measure the Size of the Shadow Economy. The size of the shadow economy can be estimated based on surveys or as a by-product of tax auditing and other compliance methods. The survey-based methodology relies on the sampling and survey of voluntary respondents. The shadow economy can also be measured by tax auditing and other compliance methods: Estimating the shadow economy based on the discrepancy between income declared for tax purposes and that measured by selective checks such as fiscal auditing and tax compliance data (tax audit data). Indirect (or indicator) Approaches to Measure the Size of the Shadow Economy. There are five different approaches, relying on macroeconomic and other indicators to trace the development of the shadow economy over time. a. The Approach Based on the Discrepancy Between National Expenditure and Income Statistics. This approach is anchored on the comparison between an independent estimate of the expenditures and the national income shown in the national accounts; the discrepancy between the two measurements approximates the size of the black economy. b. The Approach Based on the Discrepancy between Official and Actual Labor Force. c. The Transactions Approach. This approach relates money demand, and transaction volume, reflected in Fisherian quantity equation: MV=pT,113 to the size of nominal GNP (official and unofficial). Shadow economy is estimated as the difference between the total nominal GNP and official GNP. d. The Currency Demand Approach. Tanzi (1980, 1983) econometrically estimates a currency demand function for the US for the period 1929–1980, controlling for conventional determinants of money demand (for example, growth of income, payment habits, interest rates) and using a weighted average tax rate (weighted of direct and indirect taxes) as proxy for changes in the size of the shadow economy. His underlying assumption is that shadow economy transactions are cash-based, and the tax burden is the only cause of the shadow economy. The development of the size of the shadow economy is gauged by the difference in the currency demand when the tax burden is at low and high levels. e. The Physical Input (Electricity Consumption Method). The underlying assumptions are that electricity consumption is the single best physical indicator of overall (official and unofficial) economic activity, and that the growth in consumption of electricity moves in lockstep with the growth of GDP. The Model Approach to Measure the Size of the Shadow Economy. This econometric model (The dynamic multiple indicators multiple causes (DYMIMIC) model) incorporates multiple causes (for example, tax burden, regulation burden, and tax morality) and multiple effects of the shadow economy (for example, increased monetary transactions; increased participation of workers in the hidden sector; development of the production market (higher shadow economy tends to depress the official growth rate of the economy)). 113 In Fisher equation, M is denoted for money demand, V is for velocity, p is for prices, and T is for the number of transactions. 291 TA X R E F O R M I N V I E T N A M Annex 13: Shadow Economy, Tax Collection, and Leakagea Shadow Economy Tax Revenues Tax Leakage Country (2007) (%GDP) (2007) (% of GDP) Algeria 31.2 37.2 11.61 Austria 9.5 20.1 1.91 Bangladesh 34.1 8 2.73 Belarus 43.3 23.7 10.26 Belgium 21.3 25.1 5.35 Bulgaria 32.7 23.1 7.55 Canada 15.3 13.7 2.10 China 11.9 9.9 1.18 Colombia 33.5 13.6 4.56 Costa Rica 24 18.5 4.44 Cote d’Ivoire 47 15.5 7.29 Croatia 30.4 20.2 6.14 Czech Republic 17 15.2 2.58 Denmark 16.9 35.8 6.05 Dominican Republic 30.5 15.9 4.85 Finland 17 21.7 3.69 France 14.7 21.8 3.20 Georgia 62.1 13.7 8.51 Germany 15.3 11.8 1.81 Greece 26.5 20.4 5.41 Guatemala 47.9 12.1 5.80 Hungary 23.7 21.4 5.07 India 20.7 11.9 2.46 Indonesia 17.9 12.4 2.22 Iran, Islamic Rep. 17.3 7.3 1.26 Israel 20.7 27.9 5.78 Italy 26.8 23 6.16 Jamaica 32.5 26.9 8.74 Jordan 17.2 23.8 4.09 Kazakhstan 38.4 12.3 4.72 Kenya 29.5 17.8 5.25 Korea, Rep. 25.6 16.6 4.25 Kyrgyz Republic 38.8 16.4 6.36 Latvia 27.2 15.6 4.24 Lebanon 32 14.8 4.74 Lithuania 29.7 17.7 5.26 Madagascar 38.5 11.4 4.39 Malaysia 29.6 14.8 4.38 292 ANNEX Shadow Economy Tax Revenues Tax Leakage Country (2007) (%GDP) (2007) (% of GDP) Mexico 28.8 17.9 5.17 Moldova 44.3 (2006) 16.2 (2006) 7.18 (2006) Nepal 36 9.8 3.53 Netherlands 13 23.5 3.06 New Zealand 12 30.8 3.70 Nicaragua 43.1 18.3 7.89 Norway 18 29.2 5.26 Pakistan 33.6 9.8 3.29 Panama 60.0 (2006) 14.5 (2006) 8.68 (2006) Peru 53.7 15.6 8.38 Philippines 38.3 14 5.36 Poland 26 18.3 4.76 Portugal 23 21.7 4.99 Romania 30.2 11.8 3.56 Senegal 41.7 16.5 6.88 Singapore 12.2 13.1 1.60 Slovenia 24.7 19.6 4.84 South Africa 25.2 28.9 7.28 Spain 22.2 13.9 3.09 Sri Lanka 42.2 14.2 5.99 Sweden 17.9 22.4 4.01 Switzerland 8.1 10 0.81 Tunisia 35.4 20.8 7.36 Uganda 40.3 12.4 5.00 Ukraine 46.8 16.5 7.72 United Kingdom 12.2 27.8 3.39 United States 8.4 11.8 0.99 Uruguay 46.1 18 8.30 Venezuela, RB 30.9 17.7 5.48 Vietnam 14.4 19.6 2.82 Sources: Schneider and Klinglmair 2010; WDI 2008. a. The correlation between shadow economy and tax-to-GDP ratios is -0.4 (author’s estimate). 293 References Academy of Finance. 2007. On the Real Power of Sub-national Governments in the Current Inter- governmental Fiscal Arrangement in Vietnam. Hanoi. Alm, James, and Sally Wallace. 2004. Can Developing Countries Impose an Individual Income Tax? Atlanta, Georgia State University. Alm, James, Jorge Martinez-Vazquez, and Sri Mulyani Indrawati. 2004. Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia: The ‘Big Bang’ Program and its Economic Consequences. London: Edward Elgar Publishing. Arnott, Richard J., and Joesph Stigliltz. 1979. Aggregate Land Rents, Expenditure on Public Goods, and Optimal City Size. The Quarterly Journal of Economics 93 (4) (November): 471– 500. ASEAN Secretariat. 2011. http://www.aseansec.org/19585.htm. Bahl, Roy, and Bayar Tumennasan. 2004. How Should Revenues from Natural Resources Be Shared in Indonesia?â€? In Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia – The “Big Bangâ€? Program and its Economic Consequences, ed. James Alm, Jorge Martinez-Vazquez, and Sri Mulyani Indrawati. London: Edward Elgar Publishing. Bahl, Roy, and Jorge Martinez-Vazquez. 2007. Property Taxes in Developing Countries: Where Will We Be in 2015? In Roy Bahl, Jorge Martinez-Vazquez, and Joan Youngman (eds.), Making the Property Tax Work in the Developing World. Atlanta: Lincoln Institute. Bahl, Roy, and Linn, Johannes F. 1992. Urban Public Finance in Developing Countries. Washington: Oxford University Press. Washington DC. Bird, Richard M. 2000. “Rethinking Subnational Taxes: A New look at Tax Assignment.â€? Tax Notes International 8: 2069–96. Bird, Richard M. 2003. “A New Look At Local Business Taxes.â€? Tax Notes International 30: 695. Bird, Richard M., and Enid Slack. 2004. International Handbook of Land and Property Taxation. London: Edward Elgar Publishing. Bird, Richard M., and Milka Casanegra de Jantscher, eds. 1992. Improving Tax Administration in Developing Countries. Washington, DC: International Monetary Fund. Bird, Richard M., Jorge Martinez-Vazquez, and Benno Torgler. 2004. Societal Institutions and Tax Efforts in Developing Countries. Working Paper 04-06. Andrew Young School of Policy Studies, Georgia State University, Atlanta. Boadway, Robin. 1997. Tax Assignment in the Canadian Federal System. Reshaping Fiscal Federalism in Australia, ed. Neil A. Warren. Sidney: Australian Tax Foundation, pp. 61–90. 295 TA X R E F O R M I N V I E T N A M Boadway, Robin, Maurice Marchand, and Marianne Vigneault. 1998. The Consequences of Overlapping Tax Bases for Redistribution and Public Spending in a Federation. Journal of Public Economics 68 ( June): 453–78. Bourguignon, Francois, and Luiz Pereira da Silva, eds. 2003. The Impact of Economic Policies on Poverty and Income Distribution. Washington DC: World Bank and Oxford University Press. Booz Allen Hamilton. 2007. Business Process Re-engineering Baseline Findings. Internal Document. Centre for International Economics (CIE). 2004a. Revenue impacts of WTO accession: Vietnam’s customs modernisation. Canberra: Centre for International Economics. ——. 2004b. “WTO Accession and Industry: Assessing the Implications of Tariff Commitments. Canberra: Centre for International Economics. ——. 2007. “Vietnam’s WTO Accession: Assessing the Implications of Tariff Binding Commitments.â€? Report prepared for the Asian Development Bank, Centre for International Economics, Canberra. Casanegra de Jantscher, Milka, and Richard Bird. 1992. Improving Tax Administration in Developing Countries. Washington DC: International Monetary Fund. Choi, Jay Pil, and Marcel Thum. 2005. Corruption and the Shadow Economy. International Economic Review 46 (3) (August). Citro, C. F., and E. A. Hanushek, eds. 1991. The Uses of Microsimulation Modeling. Vol. 1, Review and Recommendations. Washington, DC: National Academy Press. Das-Gupta, Arindam, Engelschalk, Michael, Mayville, William. 1999. An Anti-Corruption Strategy for Revenue Administration. World Bank PREM Note 33, World Bank, Washington, DC. De Jantscher, M. Casanegra. 1990. Administering a VAT. In Value Added Taxation in Developing Countries, M. Gillis, C. S. Shoup, and G. P. Sicat, eds. World Bank: Washington, DC. Deaton, Angus, and Salman Zaidi. 2002. Guidelines for Constructing Consumption Aggregates for Welfare Analysis. Living Standards Measurement Surveys Working Paper 135, World Bank, Washington DC. Demery, Lionel. 2000. Benefit Incidence: A Practitioner’s Guide. Washington, DC: World Bank. Ebel, Robert D., and Serdar Yilmaz. 2002. On the Measurement and Impact of Fiscal Decentralization. World Bank Policy Research Working Paper No. 2809, World Bank, Washington, DC, March. Ebrill, Liam, Michael Keen, Jean-Paul Bodin, and Victoria Summers. 2001. The Modern VAT. International Monetary Fund, Washington, DC. EIU (Economist Intelligence Unit). 2008. Vietnam: Country Report – April 2008. London. 296 REFERENCES Engelschalk, Michael, Dana Weist, and Samia Melhem. 2000. Computerizing Tax and Customs Administrations. World Bank PREM Note 44, World Bank, Washington, DC. General Statistics Office. 2008. http://www.gso.gov.vn/default_en.aspx?tabid=472&idmid=3&Item ID=6540. Gordon, Richard K. 1996. Law of Administration and Procedure. In Tax Law Design and Drafting, Vol. 1, ed. V. Thuronyi. International Monetary Fund, Washington, DC. GSO (General Statistical Office of Vietnam). 2001. Some Issues in Compliance of Main Accounts by International Sectors in Vietnam. Country Report for Interregional Workshop on the 1993 System of National Accounts, May 7–11, Manila, the Philippines. Gupta A., and A. Harding, eds. 2007. Modeling our Future, Population Ageing, Social Security and Taxation. International Symposia in Economic Theory and Econometrics. Amsterdam: Elsevier. Gupta, A., and K. Vishnu. 2000. Micro-simulation in Government Policy and Forecasting. Amsterdam: North-Holland: Elsevier Science Ltd. Harrison, Graham, and Russel Krelove. 2005. VAT Refund: A Review of Country Experience. IMF Working Paper, International Monetary Fund, Washington, DC, November. Haughton, Jonathan. 2004. An Assessment of the Tax System of Lebanon. For office of Prime Minister, Beirut. ——. 2005. An Assessment of Tax and Expenditure Incidence in Peru. For the Inter-American Development Bank. Haughton, Jonathan, Nguyen The Quan, and Nguyen Hoang Bao. 2006. Tax Incidence in Vietnam. Asian Economic Journal 20 (2): 217–39. IFC (International Finance Corporation). 2007. Tax Compliance Cost Survey. Washington, DC. IMF (International Monetary Fund). 2006. Vietnam: Statistical Appendix, IMF Country Report No. 06/423. International Monetary Fund, Washington, DC. ——. 2007. Vietnam: Statistical Appendix, IMF Country Report No. 07/386. International Monetary Fund, Washington, DC. Investment and Trade Promotion Centre (ITPC). 2007. Vietnam’s Custom Tariff Schedule 2006. http://www.itpc.hochiminhcity.gov.vn/en/trade_guide/vietnam_customs_tariff_ schedule_2006 ITD. 2007. Taxation of Small and Medium Enterprises. Background paper prepared for ITD Global Conference, Buenos Aires, Argentina, October. http://www.itdweb.org/smeconference/documents/itd%20global%20conference%20-%20 background%20paper.pdf>. Jenkins, Glenn P., Chun-Yan Kuo, and G. P. Shukla. 2000. Tax Analysis and Revenue Forecasting – Issues and Techniques. Cambridge, Massachusetts: Harvard Institute for International Development, Harvard University. 297 TA X R E F O R M I N V I E T N A M Johnson, S., D. Kaufmann, and A. Shleifer. 1997. The Unofficial Economy in Transition. Brookings Papers on Economic Activity, Brookings Institution, Washington, DC. Johnson, Simon, Daniel Kaufmann, and Pablo Zoido-Lobaton. 1998. Regulatory Discretion and the Unofficial Economy. American Economic Review 88 (2). Keen, Michael. 2003. Tax Reform in Italy. Tax Notes International 17 (February): 665–82. Kiefer, Donald W. 2001. Distributional Tax Progressivity Indexes. National Tax Journal 37: 497–513. King, John. 1995. Alternative Methods of Revenue Forecasting and Estimating. In Tax Policy Handbook, ed. P. Shome. Washington, DC: International Monetary Fund. Kyobe, A., and S. Danninger. 2005. Revenue Forecasting – How is it Done? Results from a Survey of Low-Income Countries. International Monetary Fund Working Paper, International Monetary Fund, Washington, DC. Le, Minh Tuan. 2007. Estimating the VAT Base: Methodology and Application. Tax Notes International 46 (2) (April). Le, Minh Tuan Le et al. 2008. Expanding Taxable Capacity and Reaching Revenue Potential: Cross-Country Analysis. World Bank Working Paper No. 4559, World Bank, Washington, DC, March. Linn, Johannes F. 1983. Cities in the Developing World: Policies for Their Equitable and Efficient Growth. Oxford: Oxford University Press. Lopez Laborda, J., Jorge Martinez-Vazquez, and C. Monasterio. Forthcoming. The Practice of Fiscal Federalism in Spain. In The Practice of Fiscal Federalism: Comparative Perspectives, ed. A. Shah. The Forum of Federations. Atlanta, USA. Martinez-Vazquez, Jorge. Forthcoming. Revenue Assignments in the Practice of Fiscal Decentralization. In Fiscal Federalism and Political Decentralization, ed. Nuria Bosch and Jose Maria Duran. Atlanta: Edward Elgar, U.K. Martinez-Vazquez, Jorge, and Jameson Boex. 2001. Russia’s Transition to a New Federalism, with, World Bank Institute Learning Resources Series. Washington, DC: World Bank. Martinez-Vazquez, Jorge, and Robert McNab. 2006. Fiscal Decentralization, Macroeconomic Stability, and Economic Growth. Hacienda Pública Española-Revista de Economía Pública, no. 179. Martinez-Vazquez, Jorge, Andrey Timofeev, and Jameson Boex. 2006. Reforming Regional- Local Finance in Russia. World Bank Institute Learning Resources Series. Washington, DC: World Bank. Martinez-Vazquez, Jorge, Charles E. McLure, Jr., and Francois Vaillancourt. 2006. Revenues and Expenditures in an Intergovernmental Framework. In Perspectives on Fiscal Federalism, ed. R. Bird and F. Vaillancourt. Washington, DC: World Bank. 298 REFERENCES McLure, Charles E., Jr. 1996. The Sharing of Taxes on Natural Resources and the Future of the Russian Federation. In Russia and the Challenge of Fiscal Federalism, ed. Christine I. Wallich. Washington, DC: World Bank. ——. 1998. The Tax Assignment Problem: Ends, Means, and Constraints. Public Budgeting and Financial Management 9 (4): 652–83. ——. 2000. Tax Assignment and Subnational Fiscal Autonomy. Bulletin for International Fiscal Documentation (December): 626–35. Musgrave, Richard A. 1959. The Theory of Public Finance. New York: McGraw-Hill. ——. 1983. Who Should Tax, Where, and What? In Tax Assignment in Federal Countries, ed. Charles E. McLure, Jr. Canberra: Center for Research on Federal Financial Relations, pp. 2–19. Nellor, David. 1987. The Effect of Value Added Tax on the Tax Ratio. IMF Working Paper 87/47. International Monetary Fund, Washington, DC. Norregard, John. 1997. Tax Assignment. In Fiscal Federalism in Thoery and Practice, ed. Teresa Ter-Minassian. Washington, DC : International Monetary Fund, pp. 49–72. OECD (Organisation for Economic Co-operation and Development). 2001. Compliance Measurement – Practice Note. Prepared by the OECD Committee of Fiscal Affairs Forum on Strategic Management. OECD, Paris. http://www.oecd.org/dataoecd/36/1/1908448.pdf. ——. 2001. Principles of Good Tax Administration. OECD, Paris. ——. 2004. Compliance Risk Management: Managing and Improving Tax Compliance. OECD, Paris. ——. 2007. Taxation and Governance. Policy Paper (draft of October 31). OECD, Paris. ——. 2010. Tax Administration in OECD and Selected Non-OECD Countries: Comparative Information Series. OECD, Paris. ——. 2011. Cutting Red Tape – Administrative Simplification in Viet Nam, 2011. OECD, Paris. Plassmann, Florenz, and T. Nicolaus Tideman. Forthcoming. Accurate Valuation in the Absence of Markets. Public Finance Review. Available as a working paper at: http://econ.binghamton. edu/wp01/WP0101.pdf. Sah, Raaj K. 1983. How much Redistribution is Possible through Commodity Taxes? Journal of Public Economics 20 (1): 89–101. Sanford, C. T. 1995. Tax Compliance Costs: Measurement and Policy. Bath: Fiscal Publications. Schneider, Friedrich, and Dominik Enste. 2000. Shadow Economies: Size, Causes, and Consequences. Journal of Economic Literature 38 (March): 77–114. Schneider, Friedrich, and Robert Klinglmair. 2004. Shadow Economies around the World: What Do We Know? IZA, Bonn. 299 TA X R E F O R M I N V I E T N A M Schneider, Friedrich, Andreas Buehn, and Claudio E. Montenegro. 2010. New Estimates for the Shadow Economy all over the World. International Economic Journal 24 (4): 443–61. Sennoga, Edward, David Sjoquist, and Wallace. 2007. Incidence and Economic Impacts of Property Taxes in Developing and Transitional Countries. In Making the Property Tax Work in the Developing World, ed. Roy Bahl, Jorge Martinez-Vazquez, and Joan Youngman. Georgia State University, Atlanta, Georgia. Shome, P. 1988. On the Elasticity of Developing Country Tax Systems. Economic and Political Weekly XXIII (34) (August 20). Shome, Parthasarathi, ed. 1995. Tax Policy Handbook. Washington, DC: Fiscal Affairs Department, International Monetary Fund. Shoven J. B., and J. Whalley. 1984. Applied General Equilibrium Models of Taxation and International Trade: An Introduction and Survey. Journal of Economic Literature 22: 1007–51. Shukla, G. P. 2006. Tax Policy Analysis and Legal Framework: Taxation Reform and Modernization Program, Vietnam. Duke Center for International Development, Duke University, Durham, North Carolina. ——. 2007. Tax Policy Analysis and Legal Framework: Taxation Reform and Modernization Program, Vietnam. Duke Center for International Development, Duke University, Durham, North Carolina. Tait, Alan A. 1988. Value Added Tax: International Practice and Problems. International Monetary Fund, Washington, DC. ——. 1991. “Value Added Tax: Administrative and Policy Issues.â€? International Monetary Fund, Washington, DC. Tanzi, Vito, and Hamid Davoodi. 1997. Corruption, Public Investment, and Growth. Working Paper/97/139, International Monetary Fund, Washington, DC. Tanzi, Vito, and Hamid Davoodi. 2000. Corruption, Growth & Public Finances. Working Paper/00/182, International Monetary Fund, Washington, DC. Tenev, Stoyan, Amanda Carlier, Omar Chaundry, and Quynh-Trang Nguyen. 2003. Informality and the Playing Field in Vietnam’s Business Sector. International Finance Corporation, World Bank and Mekong Private Sector Development Facility. Ter-Minassian, Teresa. 1997. Fiscal Federalism in Theory and Practice. Washington, DC: International Monetary Fund. Tuerck, David, Jonathan Haughton, Keshab Bhattarai, Phuong Viet Ngo, and Alfonso Sanchez- Penalver. 2006. The Economic Effects of the FairTax: Results from the Beacon Hill Institute CGE Model. Boston MA: Beacon Hill Institute, Suffolk University. United Nations. 1999. Handbook of Input-Output Table Compilation and Analysis. New York: United Nations. 300 REFERENCES Vickrey, W. 1961. Counterspeculation, Auctions and Competitive Sealed Tenders. Journal of Finance 16 (May): 837. Vietnam Law Data. 2004. Vietnam’s Import-Export Tariff, September 2003. Hanoi. Watanabe, Satoshi, and Duong Thi Ninh. 2006. Indirect Tax Reform. In The Final Report of the Joint Research Program in the Vietnamese Tax System. Tax Policy Department, Ministry of Finance, Government of Vietnam. Wilson, John Douglas. 1999. Theories of Tax Competition. National Tax Journal 52 (2): 269– 304. World Bank. 2004. Public Expenditure Review and Integrated Fiduciary Assessment. World Bank, Vietnam. ——. 2006. Vietnam Development Report: Business. World Bank, Washington, DC. ——. 2007. Project Appraisal Document, Tax Administration Modernization Project. World Bank, Washington, DC. ——. 2008. Vietnam Development Report: Social Protection. World Bank, Washington, DC. ——. 2010. Doing Business 2011 – Making a Difference for Entrepreneurs. Washington, DC: World Bank. WTO (World Trade Organization). 2001. Technical Note on the Accession Process. WT/ ACC/10, December 21, Geneva ——. 2004. www.wto.org/english/thewto_e/acc_e/completeacc_e.htm#pan. ——. 2006. Report of the Working Party on the Accession of Vietnam. WT/ACC/VNM/48, 27, October. ——. 2011. Tariff Download Facility. http://tariffdata.wto.org/default.aspx. Yates, Judith. 1994. Imputed Rent and Income Distribution. Review of Income and Wealth 40 (1): 43–6. 301 cứÄ? gnồH nảb tấux àhN iá»›ig ếTh gnàh nâgN :gnud iá»™n mệihn hcárt uịhC mc82x5,02 ổhk ,)tệiv gnếit 003 + hna gnếit 003( nốuc 006 nI Ä?H/50-40/20/BXC/1102-337 :nảb tấux hcạoh ếk ýk gnăđ ốS cứÄ? gnồH nảb tấux àhN iá»›ig ếTh gnàh nâgN :gnud iá»™n mệihn hcárt uịhC mc82x5,02 ổhk ,)tệiv gnếit 003 + hna gnếit 003( nốuc 006 nI Ä?H/50-40/20/BXC/1102-337 :nảb tấux hcạoh ếk ýk gnăđ ốS cứÄ? gnồH nảb tấux àhN iá»›ig ếTh gnàh nâgN :gnud iá»™n mệihn hcárt uịhC mc82x5,02 ổhk ,)tệiv gnếit 003 + hna gnếit 003( nốuc 006 nI Ä?H/50-40/20/BXC/1102-337 :nảb tấux hcạoh ếk ýk gnăđ ốS