ENJEUX MODERNES D’UNE VILLE MILLÉNAIRE REGIONAL INVESTMENT IN VIET NAM CHALLENGES AND OPPORTUNITIES THE WORLD BANK D E C E N T R A L I Z AT I O N & S PAT I A L I N T E G R AT I O N A D V I S O R Y P R O G R A M JANUARY 2024 1 2 REGIONAL INVESTMENT IN VIET NAM THE WORLD BANK D E C E N T R A L I Z AT I O N & S PAT I A L I N T E G R AT I O N A D V I S O R Y P R O G R A M CONTENT 01 Motivation and Key Messages 6 02 An Overview of Public Investment Trends in Vietnam 10 03 Ine ciencies in the PIM and IGF systems Allocative ine ciencies related to strategic infrastructure 14 14 Allocative problems related to under-investments in environmental protection and adaptation to climate risks 17 Ine ciency in the implementation of public investment 18 04 Systemic Problems in Subnational PIM and IGF Systems 20 Lack of a conducive intergovernmental scal framework 21 Lack of an enabling legal framework for vertical and horizontal investment coordination 24 Lack of an e ective incentive and enforcement mechanism 25 Lack of an e ective planning-budgeting-investment framework 26 05 Conclusions Next Steps Recommendations and 30 Investment assignments and plans linking closely with integrated masterplans 31 Adoption of e ective instruments for vertical coordination across levels of government 32 Adoption of e ective instruments for horizontal coordination among sub-national governments 32 Annex 1. OECD principles for e ective public investment across levels of government and selected country application examples 33 Annex 2. Quantifying the ine ciencies of a PIM system 35 1 MOTIVATION AND KEY MESSAGES 1. Vietnam, in seeking to reach upper middle-income status by 2030 and high-income status by 2045, aspires to become a modern and industrialized nation with a higher quality of life for its citizens. To achieve these ambitious development goals, the country needs to ensure that gross capital investments from all sectors account for 32 to 35 percent of GDP on average from 2021 until 2030, including government investment at an average of 7.3 percent of GDP coordination mechanisms. Within the unitary set of rules and norms applied annually to support infrastructure across the country, the share of subnational governments (SNGs) in total development during this period.1 government expenditures increased from 26 percent in 1992 to close to 60 in More importantly, Vietnam needs 2020. SNGs account for as much as 80 percent of infrastructure investment, better investments and there is significantly higher than that of peers in similar level of development and significant scope to increase the income. As analyzed in further details below, in the absence of effective allocative, technical, and operational coordination institutions, there has been a proliferation of uncoordinated efficiency of public investment. investments—executed by both the central government and 63 provinces some of which are very small in size—resulting in dispersed development against the 2. During its development process, agglomeration trends of economic activities while not collectively addressing Vietnam has embarked on rapid both national and regional opportunities and challenges. spending decentralization policy, while lacking the prerequisite 3. A viable public investment strategy will also require effective strategic effective planning and investment coordination in investment planning and execution both vertically and horizontally across levels of government. Vietnam has pursued several 1 Socio-Economic Development Strategy 2021– approaches to regional coordination, however, as discussed in detail below, 2030 (approved by Central Party Committee none of them has produced expected results due to the lack of effective in 2021) and National Spatial Development financing arrangements. The Planning Law (2019) aims at a paradigm shift in Masterplan 2021–2030, vision 2050 (approved by the National Assembly in January 2023). master-planning approach towards integrated spatial development at national 6 and regional levels. This means moving from provincial to an area-based and inter-provincial and transboundary one; and from a sectoral to a multisectoral perspective. A national and six regional masterplans are being prepared, and pilot regional coordination institutions were established in the south. However, the prerequisite institutional set up in terms of both resources and mechanisms to realize the regional investment needs are still missing.2 4. This policy note3 is intended to inform and motivate the discussions on reform options to strengthen the linkages of the public investment at the national and subnational levels. It also aims to prompt a sense of urgency in addressing inefficiencies in the public investment management system. These inefficiencies have risen from allocative duplications, implementation challenges, negative externalities, and suboptimal private sector mobilization and financing. It also focuses on the weaknesses in the public investment management (PIM) and intergovernmental fiscal (IGF) systems to understand the reasons behind the failure to undertake regional investments. The weaknesses in the PIM system prevent the most impactful investments from being prioritized and adequately executed. In terms of IGF system, there is an excessive degree of decentralization of PIM power to provincial government without the necessary prerequisite central level control and monitoring systems. The weaknesses in both PIM and IGF systems leave room for race-to-bottom type of a competition among provinces. Some of the key findings of the note include: • Despite growing needs, Vietnam’s public investment level has been on a declining trend from 8 percent in 2011 to 6 percent of GDP in 2022. While every effort should be made to reverse this trend, given fiscal constraints it is unlikely that Vietnam will be able to raise public investment significantly. The priority focus, therefore, must be on improving the efficiency of investment. • Investment budget suffers from chronic under-execution with only 77 percent of the allocation being effectively implemented during 2017-22 period. Long delays in implementing and completing major projects lead to significant cost increases. A review of a sample of selected large-scale transport projects found an average delay of 5 years and an average cost overrun of double the original cost at the design and budget allocation stage. 2 Regional investments for the purpose of this paper refer to investments that may encompass any one of the specific economic regions or link two or more provinces. 3 Prepared as part of a World Bank’s analytical and advisory program that aims to inform the stakeholders’ decision-making on the potential adjustments to the coefficients, modalities, and incentives for the country’s strategic allocation of fiscal resources corresponding to the national and regional spatial development masterplans. 7 7 • National infrastructure spending—critical to addressing emerging needs for backbone infrastructure that tends to cut across provincial borders—has been squeezed disproportionately, absorbing a declining share of the shrinking capital budget. Central government now accounts for a mere 20 percent of total government investment (compared to 40 percent 7 years ago). At the same time, there is well-recognized over-investment by provinces in low value projects and stranded assets, with potential negative impact on the environment (such as low-occupancy industrial parks and provincial ports). • The investment budget decline is compounded by persistent allocative, technical, and operational inefficiencies. Between 2011 and 2019, it took more than six VND in investment to generate one VND in output. As a result, Vietnam achieved much less growth per dollar of investment than China, Malaysia, Republic of Korea, Singapore, and Thailand when they were at comparable levels of per capita income and development. The IMF (2018) estimated that the output efficiency of public investment for Vietnam could be 23 percent higher if the efficiency of public investment management reached the global frontier. 5. Going forward, it is imperative that the Government inter alia needs to: (i) rebalance the infrastructure investment budget from provincial level to central level, (ii) address legal loopholes particularly in the State Budget Law to enable vertical and horizontal collaboration across levels of government, (iii) institute a matching grant mechanism to facilitate cross-provincial infrastructure investment coordination, and (iv) establish robust monitoring and oversight mechanisms for the use of capital budget resources at the central and provincial levels. 6. After this brief introduction the next section provides a background on the trends in the volume and efficiency levels of public investment in Vietnam, against the backdrop of its development ambition. Section 3 analyzes the financial and opportunity costs and investment inefficiencies that have arisen due to the lack of coordination. Section 4 reviews the key determinants that have hindered more coordinated efforts. Section 5 provides initial recommendations on the needed adjustments to the current frameworks—legal, institutional, incentives, and enforcement arrangements—to enhance vertical and horizontal coordination of regional investments, including those that are feasible within the existing legal framework and those would require revisions to applicable laws in the longer run. 8 9 2 AN OVERVIEW OF PUBLIC INVESTMENT TRENDS IN VIETNAM 7. Vietnam’s rapid developmental progress since 1986 helped to reach lower middle-income status in 2010 and reduce extreme poverty in the country. Underpinning the remarkable growth has been significant levels of investments in infrastructure to achieve universal access to energy, improved power generation capacity, widespread provision of water services, and better logistics and transportation networks, all of which have enabled Vietnam to become an export economy. perceived infrastructure quality had improved in line with increasing capital 8. However, to realize its stock, but Vietnam still trailed emerging and developing Asian countries in development ambition, Vietnam 2019. It ranked Vietnam 77th out of 141 countries worldwide, behind regional will need to sustain the level of peers such as China, India, Indonesia, Malaysia, and Thailand—countries that public investment in addition to Vietnam is competing with for FDI (Table 1). The Global Quality Infrastructure enhancing improving the efficiency Index 2021 ranked Vietnam 51st out of 184 economies, below Indonesia (28th), and effectiveness of PIM system. Malaysia (29th), and Thailand (33rd). One example of infrastructure quality is Vietnam’s infrastructure quality lags the expressway density, which is one of the lowest in the region, while road that of many Asian countries and transport costs are the highest regionally.4 The infrastructure investment gap could impact its attractiveness as an will constrain Vietnam’s ability to attract and retain FDI, including those looking FDI destination and potential growth to relocate from China. in the long term. According to the World Economic Forum Survey, 4 In Vietnam, road transport represented about 79 percent of overall transport budget with the highest budget-to-volume ratio. Road transport remains the most important mode in terms of volumes, accounting for more than 90 percent of passengers and 70 percent of freight volumes during 2010–2020, also the costliest form of domestic freight transport. 10 Table 1: Quality of Infrastructure Overall Roads Railroads Seaports Air transport Utility infrastructure infrastructure Score Rank Score Rank Score Rank Score Rank Score Rank Score Rank Vietnam 65.9 77 3.4 103 3.6 54 3.8 83 4.0 103 79.6 87 China 77.9 36 4.6 45 4.5 24 4.5 52 4.6 66 86.9 65 India 68.1 70 4.5 48 4.4 30 4.5 49 4.9 59 69.8 103 Indonesia 67.7 72 4.2 60 4.7 19 4.3 61 4.9 56 79.4 89 Malaysia 78.0 35 5.3 19 5.1 13 5.2 19 5.5 25 89.7 51 Thailand 67.8 71 4.4 55 2.8 75 4.1 73 5.0 48 78.9 90 Source: World Economic Forum – The Global Competitiveness Index (2019 report) Note: Over infrastructure and Utility infrastructure Scores 1-100 Best; Sub-sector Transport infrastructure Score 1-7 Best; Rank: out of 141 countries. 9. To sustain the level of economic growth to achieve middle-income status by 2030 and high-income status by 2045, it is estimated that the Government needs to invest in infrastructure at least 7 to 7.3 percent of GDP during these periods. That is to stimulate growth to the goal of maintaining at least 7 percent annual GDP growth rate during 2021-30 and between 6.5 to 7.5 percent during 2031-2050.5 The country also needs to ensure that gross capital investments from all sectors account for 32 to 35 percent of GDP on average from 2021 until 2030, including government investment at an average of 7.3 percent of GDP annually to support infrastructure development during this period.6 This is in line with global experience that fast-growing countries are characterized by high public investment, defined as 7 percent of GDP or more.7 5 National Spatial Development Masterplan 2021-30, vision 2050 approved in National Assembly Resolution 81/2023/QH15 dated January 2023. 6 Socio-Economic Development Strategy (SEDS) 2021–2030. However, the government intends to reduce the share of public investment in gross investment to 20-23 percent during 2021-30 from over 30 percent. 7 The Growth Report: Strategies for Sustained Growth and Inclusive Development (Commission on Growth and Development, 2008). The report reviewed 3 economies that have grown at an average rate of 7 percent a year or more for 25 years or longer since 1950; and at that pace of expansion, they almost doubled in size every decade. 11 11 10. Despite critical growing needs, Vietnam’s public investment level has declined in the past decade. From 2011 to 2022, the share of public investment in total government expenditures and GDP fell from 27 to 20 percent, and 8 to 6 percent, respectively.8 Yet the economy remains relatively capital scarce with a public capital stock per capita and per worker well below upper middle-income countries (UMICs) and high-income countries (HICs) (Figures 1 and 2). Meanwhile, global experience suggests that capital accumulation was the most important driver for fast-growing countries when they were at Vietnam’s level of development (Figure 3). Figure 1. Public capital stock per capita and Figure 2. Public capital stock per worker at infrastructure index different income level Source: IMF (2021) Source: IMF (2021), World Bank WDIs Figure 3. Capital stock accumulation as a major driver for fast-growing economies Note: L=labor; H=human capital; K=capital; TFP=total factor productivity Source: Bank staff calculations based on Total Economy Database (https://www.conference-board.org/data/economydatabase/total-economy- database-productivity). 8 Ministry of Finance and General Statistical Office. Excludes investment by state-owned enterprises. 12 11. Sustaining the levels of public investment will be critical to secure productive and social infrastructure as public goods which will facilitate greater private sector investment. Vietnam has ample fiscal space—with Public and Publicly Guaranteed Debt (PPG) at 35.7 percent of GDP in 20229 compared to the 60 percent debt threshold set out by the National Assembly—and can spend more on public investment to boost sustainable growth. 12. While higher investment is necessary, more importantly, Vietnam needs better investments and there is scope to increase the efficiency of public investment. Between 2011 and 2019, it took more than six VND in investment to generate one VND in output.10 As a result, Vietnam achieved much less growth per dollar of investment than China, Malaysia, Republic of Korea, Singapore, and Thailand11 when they were at comparable levels of per capita income and development. The IMF (2018) estimated that the output efficiency of public investment for Vietnam could be 23 percent higher if the efficiency of public investment management reached the global frontier.12 Increased public spending efficiency could have a major impact on aggregate productivity growth and GDP levels. Policy research suggests that a one percentage point rise in public investment through efficiency improvements lifts growth rates by 0.1–0.2 percentage points over the next few years.13 In addition, completed projects need to be adequately maintained—the budget only finances 35-45 percent of the minimum needs in road infrastructure maintenance.14 The failure to maintain assets reduces the usable life of an asset and therefore its contribution to economic growth. 9 Ministry of Finance. 10 Incremental capital output ratio (ICOR) data from Vietnam General Statistics Office (GSO): The 2011–2015 ICOR was 6.25, 2016–2019 ICOR was 6.13 and 2016– 2020 ICOR was 7.04 as 2020 was an exceptional year due to the COVID-19 pandemic. 11 Various sources including World Development Indicators (2022); “Sector Assessment (Summary): Industry and Trade.” (Asian Development Bank (ADB, 2016); and “China’s Productivity Slowdown and Future Growth Potential.” (World Bank Policy Research Working Paper 9298, 2020). 12 Vietnam: Public Investment Management Assessment (PIMA) (IMF, November 2018). 13 Assessing the Effect of Public Capital on Growth (World Bank Policy Research Working Paper 8604/2018). 14 World Bank report “Vibrant Vietnam – Forging the Foundation of a High-Income Economy” (2019). Efficient Infrastructure chapter by Madhu Raghunath, Vivien Foster and Aditi Raina. 13 3 INEFFICIENCIES IN THE PIM AND IGF SYSTEMS 13. Unraveling the reasons for inefficient investment requires an understanding of the decentralized/deconcentrated nature of public investment in Vietnam. It seems the PIM system fails to select and implement projects that are identified as national or regional priorities. In addition, there is well-recognized anecdotal evidence for over-investment in certain types of infrastructure projects (such as industrial parks and provincial ports). This puts at risk the achievement of the development to a decentralization system that does not adequately align national goals and goals of the government that have provincial incentives, and (iii) an inefficient and ineffective public investment been articulated in national plans, management system that fails to promote national priority investment projects. such as the Social and Economic Unless these institutional issues are addressed it would be very difficult to Development Plan for 2021-25 (SEDP). “coordinate” between provinces and the central government to optimize capital investment budget. 14. The high reliance of provincial governments on public investment places a significant burden Allocative inefficiencies related to strategic on effective coordination of infrastructure public investment management institutions The institutional 15. In the context of a highly decentralized public expenditure system with underpinnings of the PIM and IGF 63 competing provinces, the PIM system encourages bottom-up project systems yield the following inefficient identification driven by provincial interests. The current planning system outcomes: (i) allocative inefficiencies fails to coordinate bottom-up proposals with top-down strategic prioritization. due to an excessive degree of The lack of effective vertical and horizontal coordination mechanisms within decentralization of public investment the intergovernmental fiscal architecture limits the ability of the central to provincial governments, (ii) government to promote priority projects and allocate budgetary resources for distortions in allocative decisions due inter-provincial investments. 14 16. The “race-to-the-bottom” type competition among SNGs in investment mobilization has resulted in suboptimal private sector financing and lack of complementarity with public investments. Amidst the lack of a national strategy on FDI mobilization, many SNGs employed their own strategies to mobilize private investments or “to lay the red carpets” for investors to come and invest in their jurisdiction. In almost of the cases provincial strategies include fiscal incentives such as land rent and subsidies. While producing some immediate benefits to the investment recipient SNG, this strategy comes at the expense of other provinces and a total loss for the country because of distorted competition environment, natural resource degradation, environmental damages, increasing tax expenditures, which overall weaken national competitiveness. 17. In the case of establishing economic zones, for example, there have been overlapping investments to attract private sector, reflecting poor planning and coordination, both by ministries and provinces. Over 25 years, Vietnam has established 4 Key Economic Regions (KERs) and 6 Socio-Economic Regions (see Figure 4), 19 Coastal Economic Zones (CEZs), 26 Border Economic Zones (BEZs), 369 Industrial Parks (IPs), 49 key tourism areas, and 859 urban centers. The current approach to establishing economic zones suggests no strategic thinking and most of the investments have proven to be uneconomic and wasteful (see Box 1). The national average occupancy rate at IPs is only 52 percent. Long An province, for example, has zero occupancy at 13 of its 38 IPs while Can Tho Province featured one of its 5 IPs with occupancy of less than 3 percent after 15 years of establishment. Recent research has found that industrial agglomeration through value chains by private sector has been far more effective in promoting private investment than economic and industrial zones pursued by the government.15 18. Vietnam has more than forty-seven seaports of all sizes across provinces, but 95 percent of cargo goes through the three ports invested and operated by the central Ministry of Transport.16 The remaining forty- four ports account for only 5 percent of cargo volume, reflecting uneconomic investments by provinces. While there are 47 seaports nationwide (including 2 special seaports, 12 grade-I seaports, and 20 grade-II seaports), there is a lack of modern facilities that can receive ships of high DWT capacities. 15 Vietnam: Connecting value chains for trade competitiveness (2019). 16 Namely: Hai Phong city in the north and Ho Chi Minh city and Ba Ria-Vung Tau province in the south. 15 15 Figure 4. Six socio-economic regions and four key economic zones in Vietnam The boundaries, colors, denominations and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Box 1. Economic density, growth and investment efficiency, and structural transformation in KERs17 The economic density in the KERs, measured by value of GDP per square kilometer, is still quite low, especially in the Central and Mekong Delta KERs (VND3.93 and 8.17 billion respectively). The growth rates of the economic density in the KERs were even diminishing from 2.33 times in 2000-05 to 2.27 times in 2005-20, while the national average economic density growth rate for the same period increased from 1.90 to 1.96 times. The labor productivity in the KERs (one measure of the quality of economic growth) is low. The labor productivity in the Central and Mekong Delta KERs was even lower than the national average (USD5,081/employee in 2019). The highest KER labor productivity level in Vietnam (of the Southern KER) was equal to only 76 percent and 57 percent in Thailand and Malaysia respectively. The incremental output ratio (ICOR coefficient) is high, reflecting low level of investment efficiency in most of the KERs. Except the Southern KER (ICOR 2.8 in 2018), ICOR of the remaining KERs was around 6, close to the national average which includes many remote, difficult provinces, and significantly higher than those in many developing countries18. The structural transformation growth rate was declining in 2006-20 from 2000-05 period. This was resulted from the declining growth rates of industrial and service production in the Northern and Southern KERs recently. Source: Staff estimates based on publicly available reports and data on GSO, MPI and other websites. 19. There has also been overinvestment in many small airports throughout the country. Currently, there are 22 airports nationwide, only 8 of which are considered international (rated 4E), most of the remaining can only accept smaller aircrafts such as A320/A321 (rated from 3C to 4D).19 There is a concentration of a high number of airports in the central coastal region— among 14 provinces in the region, there are 9 airports with an average distance of around 100 km between two airports (the shortest distance between airports in Quang Tri and Hue is 88km), which is lower than international norm of three hours by car. Only 6 of the 22 airports have experienced growth of passenger volume while the rest have only reached about 40 percent of projected volume. Most of the airports are loss-making and unable to even cover O&M costs. 17 After 15 years of establishment of the Northern, Central and Southern KERs and 9 years of the Mekong Delta KER. 18 According to the Study on ICOR in Developing Countries (IMF Institute, 2015), the ICOR in 44 developing countries varied from 1.2 to 6.56 and averaged at about 2.5. 19 According with ICAO classification. 16 20. Competition to develop hydro-power generation has also resulted in excessive and poorly coordinated provincial investments with negative externalities imposed on downstream provinces. For the last two decades, there was a boom in construction of small-scale hydropower plants in the central coastal region and central highland provinces. In just three provinces of the Central Highlands (Gia Lai, Dak Nong and Kon Tum), the construction of 256 small and medium scale hydropower plants were approved and are being operated without fully consulting the affected stakeholders. For instance, the Dak Mi 4 hydropower plant (of capacity of 210 MW) was developed with partial consultation with the affected district authorities along Vu Gia - Thu Bon River in Quang Nam province such as Nam Giang, Dai Loc, Hoi An and the downstream Da Nang city. These localities were only notified about the areas of land loss and removal of ethnic minority village when the construction started. Ten years after its operation, Da Nang city has sent multiple complaints to Quang Nam province and to the Prime Minister on the problems created by the change of the water flows which exacerbate floods during heavy rains upstream and cause drought in downstream areas in dry season.20 Allocative problems related to under-investments in environmental protection and adaptation to climate risks 21. Vietnam is among the 10 countries most affected by climate change and natural disasters globally.21 Each year, an average of US$852 million worth of economic activity and 316,000 jobs are at risk from riverine and coastal flooding in the agriculture, aquaculture, tourism, and industry sectors. This is particularly acute for many of the growth engines of the country and respective regions—like Ho Chi Minh City, Da Nang, and Can Tho cities—which are on the frontline of climate-related and extreme weather events. For instance, the World Bank estimates that Da Nang city would incur around VND900 billion—about 4.5 percent of its annual spending and 0.8 percent of GRDP—in asset loss annually due to extreme weather events during 2021–2030. Compounding the threats of sea level rises are risks due to land subsidence in populous regions of the country. The lives and livelihoods of more than 17 million citizens in the Mekong Delta could be at risk due to land subsidence, sea level rise, erosion and salination.22 22. Vietnam’s public investments and assets are increasingly exposed to growing climate risks and physical threats related to extreme weather events. In 2022, the cost of direct damage to energy and transport infrastructure from typhoons and floods was estimated at US$475 million. Firms had estimated expenses of US$280 million due to a lack of reliable and resilient infrastructure. Despite such risks, the average investment in disaster protection infrastructure is only an estimated 0.05 percent of GDP. The forecasted cost to improve coastal resilience will climb to US$4 billion by 2035, of which US$2 billion alone will cover the cost of constructing and heightening sea dikes.23 23. Vietnam’s total incremental financing needs to address the green transition and climate change-related challenges could reach an estimated US$701 billion over 2022–2040, or approximately 6.8 percent of GDP per year.24 Public investment—with its critical roles as the trailblazer and the catalyst for private investment—would account for more than one-third of the financing needs, or about 2.4 percent of GDP. The pathway to improve climate resilience alone would account for about two-thirds of this amount as substantial financing will be required to protect the country’s assets and infrastructure as well as vulnerable people. However, climate change adaptation is not fully accounted for in the overall fiscal strategy nor in prioritization of projects for budget funding in Vietnam.25 20 Small and medium scale hydropower plant development master planning in the Central region, tuoitre.vn (2017). 21 Various sources: including climateknowlegeportal.worldbank.org; climatelinks.org; and climaterisk.undp.org. 22 Vietnam Country Climate Development Report (CCDR) (World Bank, 2022). 23 Resilient Shores: Vietnam’s Coastal Development between Opportunity and Disaster Risk (World Bank, 2020). 24 Vietnam Country Climate Development Report (CCDR) (World Bank, 2022). 25 This report also acknowledges the importance of greening public investment but does not discuss related issues in detail. The issues include among other things: the lack of effective green taxonomy (to scale up sustainable investment and apply eligibility criteria for green bonds/credits), green public procurement being in its infancy (as existing legal framework needs to be operationalized and aligned better with international good practices), and green financing (particularly green bonds) not being implemented despite the availability of initial legal enablers. 17 Inefficiency in the implementation of public investment 24. The above-discussed problems are compounded by growing inefficiency in project implementation. Both central and provincial governments are unable to disburse money appropriated to support project implementation. The large difference between budgeted and actual spending—about 23 percent during 2017-2226 (Figure 5) which was significantly larger than in many peer countries such as Malaysia, Cambodia, Singapore, and Thailand—indicates chronic problems in project implementation. One of the main causes that appears to bedevil all road construction projects is difficulties in land acquisition as well as resettlement of those displaced by such land acquisition that cause extensive delays in beginning construction, longer than planned time to complete projects, and cost overruns.27 The consequential capital carried over from one year to the next stood high, at about 18 to 31 percent of total capital expenditures, significantly larger than in many peer countries, and way above international good practices which suggest the level to remain below 5 percent.28 Figure 5: Persistent execution gaps during 2017-2022 Actual/Budget Ratio (%) 100 80 60 40 20 0 2017 2018 2019 2020 2021 2022 General Government Central Government Provincial Governments Source: MPI (mpi.gov.vn) and MoF (mof.gov.vn) websites and World Bank staff estimates. 25. Inefficiency in implementation of public investment is manifested by significant time and cost overrun.29 A review of selected projects,30 especially the large-scale ones (such as urban railway projects), presents some evidence of undesirable outcomes of project management cycle. Average delay time of the projects reviewed is 5 years, and the average cost overrun is 2 times of original cost at design and budget allocation stage.31 The Cat Linh-Ha Dong urban railway project in Hanoi, for example, started in October 2011 and was initially projected to completed in November 2013. It has been delayed for more than seven years and its cost has escalated to VND 18,002 billion (US$868 million) from VND 8,770 (US$553 million) by the end of 2020.32 26 During 2017-22, execution rates map closely at both provincial and central government levels. On average the rates stand at 78 for the provincial level and at 76 for central government, leaving the national average at 77 percent. 27 It was noted that central efforts to limit pending bills to contractors led to provinces asking contractors to continue work but delay issuing invoices for payment which would suggest that financial accounts do not accurately reflect actual project implementation. 28 Public Expenditure and Financial Accountability (PEFA) Framework, Performance Indicator 1. 29 The authors were not able to obtain the consolidated data by major spenders at either ministerial or provincial level, therefore, it is impossible to present a comprehensive picture. 30 Sample of 14 transport projects which accounts for 13 percent of transport investment budget during 2011-2020, sourced from MOT. 31 In the same review, it is worth noting that time overrun tends to vary more widely in the case of smaller projects (those in group B or C) at an average of 3-4 years. 32 Lan, Do Mai and Hoang Oanh. 2021. Vietnam’s Tentative Approach to Regional Infrastructure. 18 26. Procurement practices also contribute to high costs of public investment. Despite a Public Procurement Law that meets international standards, in practice there is limited use of competitive bidding with direct procurement of contracts resulting in a consequent increase in investment costs.33 There is historical evidence of some projects being completed below cost, as indicated by a JICA assessment of construction of about 100 km of National Highway No. 10 from Hai Phong to Ninh Binh together with 40 km of bypass roads and a number of bridges. The assessment concluded that while the project took longer than planned (79 months rather than 54) the actual cost of the project was only 76 percent of the estimated cost because of very competitive bidding for the project. However, it should be noted that this ODA project was implemented between 1998- 2008 and therefore does not represent current conditions.34 Nevertheless it provides a historical comparator to current delays and cost overruns in implementation. 27. In addition, operations and maintenance (O&M) of existing infrastructure is not adequately funded. The budget only finances about 35 to 45 percent of the minimum needs in road infrastructure maintenance.35 There is clearly a preference to start new projects rather than to ensure that existing infrastructure is adequately maintained, and the life of asset is extended. The dual budgeting system with MOF overseeing the recurrent budget while MPI manages the capital budget also accounts for the underfunding of operations and maintenance. 28. There lacks a comprehensive legal framework for asset accounting and reporting and up-to-date and integrated asset registries. These have hindered effective decision-making on new infrastructure investment, O&M, and revenue policy for all infrastructure (at national and sub-national levels). While there is a clear definition of infrastructure assets in the Public Asset Management Law, there is currently only management and accounting guidance for roads and water/irrigation infrastructure assets, but even this is not consistently applied, which impedes accuracy of asset value reporting. 33 See Vibrant Vietnam, p 66-68, World Bank (2021). 34 Ex-Post Evaluation of Japanese ODA Project - National Highway No. 10 Improvement Project. 35 World Bank report “Vibrant Vietnam – Forging the Foundation of a High-Income Economy” (2021). Efficient Infrastructure chapter by Madhu Raghunath, Vivien Foster and Aditi Raina. 19 4 SYSTEMIC PROBLEMS IN SUBNATIONAL PIM AND IGF SYSTEMS 29. Achievement of Vietnam’s high growth targets set for the 2021- 2045 period will depend, in part, on the level, composition, and efficiency of PIM system. The government has identified regional investment as critical to the country’s development in the Social and Economic Development Plan for 2021-25 (SEDP). It proposes “a focus on implementing 3 strategic breakthroughs, especially in key industries and sectors, large and important national, intra-regional and inter-regional projects creating priority.36 There are also broader concerns about the composition and efficiency spillover effects and increasing of public investment which need to be addressed for effective growth. growth momentum.” Investments that span multiple provinces are 30. Weaknesses in the PIM and IGF systems prevent the realization of regional critical to improving inter-provincial investment projects. All public investment decisions (including regional connectivity, addressing the adverse investments) in Vietnam are made and implemented under laws and regulations impacts of environmental pollution that, inter alia, define the public investment management system and the and climate change, as well as intergovernmental fiscal system. Significant weaknesses in the two systems developing the complementarity and current practices offer, we believe, a compelling explanation of the failure between public and private to implement regional investment priorities. Those weaknesses also explain investments. Yet, the evidence shows the broader problems of declining efficiency of all public investment. Unless that such investments have been they are corrected, these institutional weaknesses will make achievement of neglected by both the national and the goals of the Socio-Economic Development Strategy (SEDS) and Plan (SEDP) provincial governments despite difficult. being identified as a national 36 In a recent statement regarding the southern economic region the HCMC chairman noted that “current mechanisms have not encouraged localities in the region to be proactive and strengthen regional linkages” while the party secretary of neighboring Binh Duong said: “localities in the region have been lacking cohesion and only focusing on their own development for years”. 20 31. The weaknesses in the PIM and IGF systems can be discussed under four headings. These are the lack of (a) a conducive intergovernmental fiscal transfer framework, (b) an enabling legal framework for vertical and horizontal investment coordination between the levels of government, (c) an effective incentive and enforcement mechanisms at the regional level, and (d) an effective planning-budgeting-investment framework. Lack of a conducive intergovernmental fiscal framework 32. Public investment efficiency is also affected by Vietnam’s fragmented intergovernmental system and increasingly decentralized capital spending. Since the early 1990s, the Vietnamese Government has embarked on a policy of rapid spending decentralization. Within the unitary set of rules and norms applied across the country, the share of subnational governments’ (SNGs) in total government expenditures increased from 26 percent in 1992 to almost 60 percent in 2022.37 This spending is significantly higher than the international average (24 percent) and the lower middle-income economies’ average (20 percent). 33. The existing fiscal decentralization framework provides the central government very limited and declining share of fiscal resources to address what would generally be considered as national infrastructure needs. This framework leaves the central government only limited resources for public investment—a mere 20 percent of total government investment in 2022, compared to 40 percent seven years ago. Meanwhile, it has channeled between 78-85 percent (2017-2022) of public investment via subnational governments (Figure 6 and Box 2), more than double the averages in unitary countries (34 percent) and overall (39 percent) (Figure 4). This gives each level considerable leeway in spending decisions which are determined by respective subnational authorities. 37 Source: MoF data and WB staff calculations. 21 21 Figure 6. Increasing level of provincial spending in total public investments 81.8% 84.2% 84.8% 81.1% 78.2% 77.9% 65.6% 66.2% 67.4% 58.0% 58.6% 58.9% 42.0% 41.4% 41.1% 34.4% 33.8% 32.6% 21.8% 18.2% 22.1% 18.9% 15.8% 15.2% 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Central Share in Total Investment Provincial Share in Total Investment Source: MoF (mof.gov.vn) and World Bank staff estimates Figure 7. Vietnam high level of decentralization compared to development peers Source: Subnational Governments around the World: Structure and Finance (OECD/UCLG, 2019) and World Observatory on Subnational Government Finance and Investment (OECD/UCLG, 2022 Synthesis Report). Box 2: Transport infrastructure constraints In the past twenty years, transport spending—roughly 23 percent of general government capital spending and 3.5 percent of GDP— has focused more on rural transport infrastructure (75 percent) which is managed by provinces. This equity focus, which was a policy choice, came at the expense of implementing nationally strategic investments. As a result, the expressway density in Vietnam is one of the lowest in the East Asia and Pacific (EAP) region, and the road transport costs are highest when compared to others in the region. Many segments along the North-South expressway axis have not been constructed. The railway infrastructure has also been neglected. Multimodal connectivity is limited, particularly between roadways, seaways, in-land waterways, railway; connection between railways and seaports is almost non-existent. Transport connection at the gateways of major cities, as well as connection to economic zones, industrial parks, seaports, and border gates remain limited. 34. The degree of decentralization also raises a question regarding whether the assignment of investment responsibilities is consistent with principles of public economics. Because Vietnam is so much of an outlier relative to other countries it is natural to ask if the nature of public goods (externalities, spillovers, public goods) was fully considered in assigning 22 responsibilities for specific public investments across levels of government. The difficulty in initiating regional investments and in responding to cross-border pollution externalities suggests that there are significant mismatches in assignment of investment responsibilities. 35. In Vietnam, the State Budget Law (SBL) assigns responsibilities for expenditures in general terms—the same expenditure responsibilities are assigned to both central and provincial authorities. The lack of clarity in spending responsibilities was flagged in a 2015 review of fiscal decentralization.38 The expenditure assignments were described as “broad and vague” in a recent draft provincial PEFA report.39 Sectoral legislation attempts to provide some clarity in terms of assigning exclusive and shared responsibilities. In general, however, having so many shared responsibilities between center and provinces causes ambiguity over functional assignments. Most projects appear to involve funding by the center and the province which reflects the lack of conceptual clarity in roles. The lack of exclusive assignment of responsibilities leads to provincial governments opportunistically negotiating on responsibilities they prefer and leaving the rest to the fiscally constrained central government. As a result, there are resource gaps and incentive misalignments that lead to underinvestment in public goods, particularly those with inter-provincial characteristics. 36. Furthermore, the existing fiscal decentralization policy was operationalized through extensive use of unconditional transfers. The end result is fragmentation of large projects. The North-South national highway, for example, was divided up into more than 20 segments, several of them were financed and executed by the central government (MOT), while others were assigned to provincial governments. As a result, several segments of the project are yet to be completed, leaving such an important backbone transport infrastructure incomplete. Also, the current unconditional transfer system overlooks the opportunity for using conditional transfers for strategic priorities, such as climate change. 37. It is not clear if there has been a serious effort to estimate the costs of key national infrastructure that would be the sole responsibility of the central government. Such an estimate needs to be juxtaposed against available fiscal resources of the central government. It seems highly likely that there is a significant gap between the likely costs of key national infrastructure and the capacity of the central government to fund such investment over the medium to long term. If so, this points to a structural problem of the fiscal system that planning per se cannot resolve. The impact of this constraint on Vietnam’s hopes for sustained growth and development may be substantial. 38. With provincial governments holding the lion’s share of public investment resources, the central government has to be highly selective in the selection of public investment projects. However, anecdotal evidence suggests that rather than being selective and strategic in its investment decisions, the central government is attempting to leverage co-financing from richer provinces to undertake transport infrastructure projects that would largely serve provincial interests. In effect, this seems like an attempt to stitch together a patchwork of national infrastructure investment plan by negotiating co-financing terms with a few provinces at a time. The example of Ring Road 3 (RR3) in HCMC offers an illustration of this approach. A project that was approved in 2011 was stalled for 10 years because of limited central government resources as well as the difficulty of getting provinces to agree on their investment contribution. Finally, HCMC as a rich province with an interest in the project was able to lead an agreement that included contributions from three neighboring provinces as well as the central government. The project has been approved by the National Assembly for inclusion in the MTIP with projected completion in 2027. 39. The risk is that such an ad-hoc, patchwork approach may fail to develop a strong national infrastructure that complements regional and provincial networks. Additionally, by not following a principled approach to defining the role of the central government, such case-by-case arrangements create a complex and non-level playing field that will handicap the poorer provinces who will be left with lower quality and disconnected infrastructure. 38 World Bank. 2015. Making the Whole Greater than the Sum of the Parts: A Review of Fiscal Decentralization in Vietnam. 39 PEFA. Subnational Public Expenditure and Financial Accountability Assessment of Can Tho City, 2022. 23 Lack of an enabling legal framework for vertical and horizontal investment coordination 40. State Budget Law (SBL, 2015) and Public Investment Law (PIL, 2019) do not include institutional mechanism for vertical and horizontal coordination. On vertical coordination, the laws generally assign projects of national or inter-provincial scope to ministries. However, the nature of inter-provincial investments and respective responsibilities are not clearly defined. State Budget Law also provides a general description of responsibilities in infrastructure services without adequate consideration of the resources and incentive mechanisms for their provision. The end result is underinvestment in infrastructure networks, particularly those with inter-provincial characteristics. National infrastructure is also underinvested because of the limited resources of the central government. 41. On horizontal coordination, the SBL is not conducive to coordinating across spatial or sectoral jurisdictions. On the contrary, the SBL (article 9) explicitly bans the transfer of investment funds from one province to another even in cases where the benefits of cross-provincial projects are evident. Although the intention of the legal framework is to maintain fiscal discipline, such a strict rule creates a financing gap for regional projects. With 63 provinces the need for cooperation in provision of linkage infrastructure and management of environmental externalities is self-evident, yet there is no effective institutional arrangement to facilitate such collaboration.   42. No straightforward co-financing mechanisms have been made available transparently to promote either vertical (central-local) or horizonal (local-local) coordination for regional investments. For domestic resources, the only mechanism for such vertical linkage is through the National Targeted Programs. From 2011 to 2021, the government has reduced the size of targeted (or conditional) transfer schemes in terms of expenditure (from 9.1 to 2.8 percent of total expenditure) and in terms of their number (from 16 NTPs and 45 smaller targeted programs to three NTPs and several on-granting schemes).40 There is ongoing debate on the adequacy of size and institutional structure for NTPs to better link resourcing to targeted results. For external financing, governed by the Public Debt Management Law (PDML, 2017), the co-financing arrangement was introduced in 2017.41 Provinces have been categorized into five groups, depending on their fiscal capacity measured by the level of dependency on central budget transfers, with respective co-financing ratios. However, the programmatic mechanism that connects the central budget and multiple provincial budgets (the so-called “umbrella projects”) was removed in recent years due to concerns over its complexity.42 43. Within this context, when regional projects need to be financed and undertaken by provinces, the case is to be made to the National Assembly, which in turn needs to issue a resolution to make special case to adjust Medium Term Investment Program (MTIP). The recent approval by the National Assembly of Ring Roads. 3 and 443 demonstrates the importance of such vertical coordination but only on ad-hoc basis for nationally important projects for now. However, such a process raises four main issues: • First, the case by case “negotiation” process between central and local government leads to the arbitrary treatment of similar cases. The Southeast Ring Road No. 3 and the Northern Ring Road No. 4 were both approved by 2011 as nationally important projects thereby under central government responsibility; however, the latter was prioritized with JICA funding and fully on-granted to the beneficiary provinces while the former was pending for 11 years and is now being treated as a shared central and local responsibility/co-financing project. Another relevant example is the Can Tho - Soc Trang Road—158 km connecting 4 provinces which was included in the National Master Plan and mentioned in the SEDP 2021-25 as one of the critical regional infrastructures for sustained equitable development—this project has stalled because 3 of the 4 40 Source: World Bank staff calculations based on MoF and MPI websites. 41 Through Decree 52 (2017), which was subsequently replaced by Decree 97 (2019) and supplemented by Decree 79 (2021) on ODA and concessional lending. 42 This mechanism was stipulated in Decree 38 (2013) then removed in Decree 56 (2020) and Decree 114 (2021) on the management and utilization of ODA and concessional financing. 43 Ring Road No. 3 of about 76 km connects 4 SNGs, namely Ho Chi Minh City, Dong Nai province, Binh Duong province, and Long An province in the Southeast region; and Ring Road No. 4 of about 113 km connects 3 SNGs, namely Ha Noi city, Hung Yen province and Bac Ninh province in the Capital / Northern region. 24 provinces are not willing to contribute to the cost. It is not clear whether the center will play a role in resolving this impasse. • Second, the co-financing ratios were also defined on an ad-hoc basis following negotiation between central government and the concerned provinces. In the case of the Southeast RR3, the equal shares of 50:50 were assigned to three out of four SNGs while 25 percent on-lending was assigned to Long An province. The applicable co-financing ratios for this project (and 8 sub-projects) are different from those applied to the provinces when it comes to ODA/CL (as governed in Decree 97). • Third, there has been no on-lending mechanism for domestic sources of financing from central government to SNGs. In the case of RR3 and RR4, the 7 concerned provinces proposed that central government issue an umbrella domestic bond for the financing shortfall, then arrange on-lending of respective portions to SNGs. This approach would help in significantly reducing the overall financing and transactional costs of SNG financing. However, this option was dismissed given the lack of applicable law, and the SNGs were eventually requested to issue bonds on their own. • Fourth, the case also shows how cumbersome the process is to make amendments in the MTIP. As the PIL provides, for a project to be financed, it needs be first included in MTIP (PIL, article 51) and then in AIP (PIL, article 53).44 The MTIP serves as the public investment planning tool, but it suffers from multiple shortcomings. The rigidity of the tool combined with strict project financing rules under the SBL leaves no legal flexibility to accommodate economically and socially beneficial linkage projects. Ring roads 3 and 4 experienced long delays in clarifying the financing mechanism. First, the National Assembly had to approve the amendment to the respective provincial MTIPs which triggered prolonged administrative procedures for negotiations for joint financing between provinces and MOT. As such, regional linkage projects face a difficult hurdle to be included in the MTIP. Other intergovernmental PIM mechanics (e.g., such as land resettlement, procurement, revenue mobilization and O&M responsibilities) are also unclear. Lack of an effective incentive and enforcement mechanism 44. Vietnam has a large number of provinces of relatively small size, and the system encourages competition rather than cross-border cooperation. Expenditure decentralization has encouraged provinces and cities to focus on development priorities within their respective administrative boundaries. In addition, the party central committee places the highest priority on economic growth. Thus, the vision and interests of provincial leaders are limited in both space and time. The political incentive structure creates race to bottom type of a competition among the provinces on the basis of GDP growth and investment/FDI numbers. As a result, provinces invest in inefficient public investment projects in the form of seaports and airports, industrial parks, tourism areas and export processing zones in order to attract foreign and domestic private investments. Without some national strategic guidance such competition among provinces can easily lead to uneconomic and wasteful overinvestments, as has indeed happened over the past decade in Vietnam. 45. The coordination mechanisms at the regional level are ineffective in the absence of financial authority to solve collective action problems. Regional coordination efforts have been made by the government since early 2000s but have fallen short in creating the incentives for regional linkage investments. From 2002 to 2020, there have been several efforts to establish steering committees for selected regions such as the Southwest (2002), Central Highland (2002), and Northwest (2004) Regions; and most recently, the regional coordination council for the Mekong Delta region (2020). However, none of these were effective as the committees could only recommend specific remedies to the Prime Minister but had no fiscal authority nor funding for regional investment programs (see Box 3). These experiences suggest that the problems cannot be addressed by just focusing on coordination councils without addressing the real financial constraints and disincentives. 44 MTIP at both provincial and ministerial levels are instituted by legislation. According to PIL (Articles 55, 60, and 62) the preparation and approval of MTIP are to start 1.5 year earlier in the planning period. At this stage, ministries and provinces review and select potential qualified investment proposals and add in their drafting MTIP. Then in the first half of the year preceding the planning period, MPI review and refine the draft of national MTIP and submit to the Government. The Government in turn submits the MTIP draft to National Assembly at the end-year session (October) for review and approval. The National Assembly approves MTIPs in the first session (May-June) of the first year of planning period. Any project ideas or proposals that come later than the due timing of MTIP would need for the MTIP to be amended with their inclusion. 25 Box 3. Challenges of coordination and investment in the Mekong Delta Region Even in regions, such as the Mekong Delta Region, that have received attention from the higher echelons of the Government and support from development partners, progress in promoting horizontal and vertical coordination has been slow and piecemeal. In the Mekong Delta Region, after years of hard work, the progress in strengthening regional coordination institutions yielded initial results, with the establishment of the Regional Coordination Council /RCC (2020), and regional data center (2022) and the Integrated Regional Master Plan (2022, approved by the Prime Minister through Decision 287/2022/QD-TTg). As per Prime Minister’s Decision 825 (2020), the main responsibility of the Mekong Delta RCC is to support the development and implementation of interprovincial investments that promote climate smart and sustainable economic transformation of the region. It includes representation from all provinces and key sectors. A reality check from the 2-year experience with the Mekong Delta RCC has shown that of the four important functions i.e., planning, implementation, monitoring and evaluation: the RCC neither has sufficient authority nor resources to intervene in the planning process; similarly, the RCC cannot intervene in implementation because this function depends on the executive authority of the government, both local and central. On the funding front, in 2017, four provinces of the Long Xuyen Quadrangular (a sub-region of the Mekong Delta region) wanted to contribute their own provincial funds to create a regional fund hosted by An Giang province for the operation of the interprovincial coordination program, but it was not possible because of the lack of a legal basis for establishing such a fund. The Prime Minister issued Decision 593 in 2016 trying to pilot regional linkages in the Delta and MPI subsequently submitted a proposal package of VND 4,000 trillion (about USD 180 million) for regional investments. However, the package could not be appropriated and disbursed as it could not be supplemented to the MTIP 2016-20. In 2019, at the strong advocacy of the World Bank and several other Development Partners, the Government issued Directive 23 with the promise of providing an additional targeted transfers in the amount of USD 2 billion to the region (on top of the balancing transfers) for the 2021-25 period. It took almost another three years for the Government to discuss list of priority investment projects for the package and the respective expenditure assignments between the central and provincial budgets given the ambiguity in relevant laws. Finally, a list of 19 projects, both cross-provincial and within-provincial ones, has been selected at the bottom-up proposals of the provinces and ministries. A co-financing ratio of 90:10 between central and provincial responsibilities was decided on an ad-hoc basis for this case (through Decree 79 (2021), different from the applicable on-lending ratio of external financing as regulated in Decree 97 (2018) and not well grounded by adequate analysis on the fiscal capacity of concerned provinces. Lack of an effective planning-budgeting-investment framework 46. In the context of high levels of decentralization, the planning and budgeting systems face challenges in coordinating top-down strategic prioritization with bottom-up project identification by 63 provinces. Despite what appears on paper (in the various laws and regulations) to be a rigorous, multilayered, and complex system of planning, budgeting, and investment, the system fails to discipline and strategically orient project proposals in the Medium-Term Investment Plan (MTIP) that would support the priorities of the Social and Economic Development Plan (SEDP). The Spatial Planning Law (2019) aims to achieve a paradigm shift in the master planning approach towards integrated spatial development at national and regional levels. The National Spatial Development Master Plan (NSDP) to 2030, vision 2050 places an emphasis on supporting four dynamic zones and three economic corridors, with linkages within and across zones and six socio-economic regions. More recent efforts towards integrated spatial development support a coherent regional investment approach, but face implementation challenges. The plethora of planning tools—including national, regional, sectoral, and provincial 10-year masterplans—are tenuously linked to their respective five-year MTIPs. Investments that span multiple provinces are critical to enhancing inter-provincial connectivity, addressing the adverse impacts of environmental pollution and climate change, and achieving synergies between public and private investments.45 45 A 2018 PIM Assessment by the World Bank noted that the MTIPs are generated through a bottom-up process and in the absence of a clearly defined top-down fiscal constraint these proposals are unrealistic and ambitious without the necessary strategic distillation. Although the PIL creates a role for the MPI and MOF to query the strategic value of projects under Group A, in fact neither ministry plays such a strategic challenge role, confining themselves to issues of whether the proposal has adequate funding to ensure its implementation. These problems are also present in even stronger form at the provincial and lower levels. 26 Figure 8. Vietnam National Spatial Development Masterplan to 2030, Vision 2050 The boundaries, colors, denominations and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. 27 47. Institutional fragmentation of planning, financing and implementation of infrastructure expenditures greatly complicates coordination challenges across levels of government. The process of reconciling bottom-up project proposals (the bulk of which emanate from provincial governments) with the SEDS lacks a rigorous project appraisal, challenge, and prioritization process. Both the Law on Public Investment and the Law on Construction (2020) refer to the Pre-Feasibility Study (PFS) and Feasibility Study (FS) as part of the appraisal process leading up to financing decision.46 However, neither the PFS or FS are required to provide a quantitative assessment of financial, economic, or social benefits and costs of proposed projects. There are no unified guidelines for project appraisal. This renders any independent review (formally required by the level of public investment with the establishment of appraisal committees) to become just a formality. 48. There is an equally serious problem of omission of national and regionally important projects that affects the quality of the MTIP. Given that most investments are undertaken at the subnational level, it is evident that there is no incentive for provinces to propose projects that would reflect national or regional priorities. In fact, the PIL prohibits using provincial resources for other than internal provincial projects so there is effectively a legal barrier to provinces proposing to undertake or collaborate on regional infrastructure projects. Thus, the bottom-up process of project proposals is unlikely to generate regional projects. 49. Finally, two other institutional disincentives for cross-investments beyond the legally defined boundary confinement at either provincial or sectorial levels. First, the missing legal requirements for or the practice of periodic review of portfolio of investments under construction. Second, effective physical monitoring of project implementation and cost control mechanisms do not exist—the outcomes of the lacked legal provisions that link such PIM function with onward financing. As such, provinces or sectors would be discouraged from entering into regional investments that would likely bind them into extra multi-agency scrutiny. In addition, these two sets of issues in project execution reenforce each other, attributing to the ever increasing carried over capital budget and cost overrun. 46 Article 52, Law on Construction emphasize that it be compulsory for detailed PFS for national important projects and group-A projects using public investment funds, PPP projects, and projects of which investment policies are approved by the NA or the PM; for other project groups, the PFS is to be decided by the investment decision makers. 28 29 5 CONCLUSIONS RECOMMENDATIONS AND NEXT STEPS 50. The analysis in this note suggests that the poor performance of the public investment system, both at the central and the provincial level, contributes to very substantial costs and risks. Because of fragmented management, critical public infrastructure is either not put in place or inordinately delayed, regional infrastructure bottlenecks are not funded or addressed, or coordinated climate-resilient infrastructure is not being built. At the same time, there is substantial evidence of wasteful investments undertaken by competing provinces 51. The below recommendations promote a structured approach to strengthen without a strategic purpose. Projects both vertical and horizontal coordination for public investments that of regional or national importance require linkages between national and regional investments as well as do not get funded; or otherwise, they among regions, on the basis of the understanding of the above-discussed must undergo the lengthy ad-hoc inefficiencies and their determinants and structured around the OECD’s practices ‘negotiation’ among key stakeholders with customization to Vietnam (see Annex 1). At this stage, recommendations and finally the clearance by the are kept relatively general as the entry points for deeper-dive studies and Prime Minister and approval by the dialogue in the future. Subject to further discussions within the Bank and National Assembly. Such allocative externally with relevant counterparts, the recommendations will be elaborated and implementation inefficiencies and sequenced in follow-on notes. pose a major risk to the growth and development prospects of Vietnam.47 47 This note also suggests a deep an in-depth Public Investment Efficiency Review to quantify inefficiencies across the public investment management (PIM) and intergovernmental fiscal (IGF) systems (see Annex 2). 30 Investment assignments and plans linking closely with integrated masterplans (i) Review expenditure responsibilities to inform the next amendments of the State Budget, Public Investment, and Sectoral Laws—defining the investment responsibilities to various levels of government, i.e., the nature, size or spatial characteristics of investments that should be national, regional, or provincial.  It should also identify more efficient matching of financing capability and investment responsibilities. (ii) Ensure that MTIPs are strongly linked, prioritized, and sequenced in a results-oriented manner to improved national and regional spatial development masterplans, within credible forecast of budget affordability. (iii) Undertake an independent, realistic costing of the national and regional investment priorities in the MTIP 2021-25 to assess the financing shortfall relative to government revenue and borrowing capacity and potential room for matching financing and to inform their prioritization at MTIP mid-term review in 2023. Mandate the consistency across MTIPs of consecutive 5-year periods with the longer-term masterplans. (iv) Encourage the production of data at the relevant scale and ensure the interoperability of and accessibility to a multisectoral database system to underpin evidence-based decision making and facilitate multi-sectoral solutions for a region. Accordingly, pre-feasibility and feasibility studies of prioritized interprovincial investments should employ the spatial and multisectoral data and analytics from the database and/or system. 31 31 Adoption of effective instruments for vertical coordination across levels of government (i) Institutionalize tools to ensure coherence of investment across levels of government—such as co-financing arrangements, contracts between levels of government, or various forms of regular inter-governmental dialogue—building on the recent exceptional resolutions of the National Assembly for the Northern regional Ring-road #4 and the Southeast regional Ring- road #3. (ii) Consider adjustments in revenue and unconditional transfers arrangements to increase the share of conditional transfers, while establishing ways for conditional grants (such as climate resilient regional investments) or matching grants (to incentivize regional investments) to better help achieve strategic national or regional goals. (iii) Prepare a regional Public Investment Program (mechanism already provided for in the Public Investment Law) for a selected pilot/demonstration region. The regional PIP could act as a “quasi regional budget” mechanism—that aligns with regional masterplan and links central and local budgets (at the same time, addresses past allocation/execution issues). (iv) Review the functions of the Regional Coordination Council with possible strengthening on setting regional priorities and ensuring effective monitoring and evaluation to enforce implementation of regional masterplan and investments. (v) Amend the State Budget Law to institutionalize a co-financing mechanism that integrate on on-lending mechanism (of external financing) and co-financing mechanism (of domestic financing, including on-lending finance from national bond issue etc.) into a holistic framework based on national and subnational fiscal capacity assessments and well-defined formula, to be appliable from 2026 onwards. (vi) Provide fiscal mechanisms and incentives to enable green and climate actions across levels of government, such as ecological . fiscal transfers and payments for ecological services​​​ Adoption of effective instruments for horizontal coordination among sub- national governments (i) Amend the State Budget Law to enabling province-province partnership with clearly defined costs and benefits sharing, and investment and O&M mechanisms. (ii) Provide and institutionalize incentives the modes of collaboration—such as specific public investment partnerships, and joint contracts and authorities—to increase efficiency through economies of scale to match public investment with the relevant geographical area and to enhance policy synergies among subnational governments. 32 Annex 1. OECD principles for effective public investment across levels of government and selected country application examples All countries are confronted by challenges for the multi-level governance of public investment, whatever the institutional context (in federal countries, or highly centralized countries) since mutual dependency across levels of government for public investment holds true in all countries. The OECD Council has therefore issued since 2014 the “Recommendation on Effective Public Investment Across Levels of Government” aiming to help governments at all levels assess the strengths and weaknesses of their public investment capacity, using a whole-of-government approach, and set priorities for improvement. The Recommendation is organized around three re-enforcing pillars: (a) Co-ordinate public investment across levels of government and policies (i.e., investment using an integrated strategy tailored to different regions, adoption of effective instruments for coordinating across national and sub-national levels of government, and horizontal coordination among sub-national governments to increase efficiency through economies of scale and to enhance policy synergies); (b) Strengthen capacities for public investment and promote policy learning at all levels of government (i.e., bolstering conditions for effective investment and to promote continuous improvement from the strategic selection of investment to its execution and monitoring); and (c) Ensure proper framework conditions for public investment at all levels of government (i.e., good practices in fiscal decentralization, public financial management, public procurement, and regulatory quality at all levels of government). The following boxes provide specific examples from other countries that are well aligned with the above-analyzed challenges in Vietnam and OECD principles. Box 4 suggests good country practices in using integrated strategy tailored to difference places to steer regional investments. Box 5, and Boxes 6 and 7, present country examples in vertical and horizontal investment coordination mechanisms respectively. Box 4. Using integrated strategy tailored to different places A national strategy for public infrastructure investments is critical the long-term development objectives of a country. It will help in the prioritization of capital investment projects by all levels of government. A place-based approach to regional development maximizes the growth potential of each region. In Canada, Regional Development Agencies develop place-based regional growth strategies that aim to achieve development in all localities and by involving stakeholders from levels of government (federal/provincial/local). In the Czech Republic, the National Investment Plan consolidates local investment plans of regions and cities. In the Netherlands, Multi-year Programme for Infrastructure, Spatial Planning and Transport is a consultation committee to discuss projects for all level of governments. In Poland, the Coordinating Committee for Development Policy, a permanent body, discusses regional development issues and carries out analysis for regional investment projects. Box 5. Coordination mechanisms across national and subnational governments Coordination across national and subnational governments are important in addressing the multidimensional challenges of climate change, urbanization, demographic challenges etc. It helps aligning policy objectives, taking concrete actions, and establishing partnerships at all levels of government. In Austria, the Ministry of Science, Research and Economy setup a policy platform for national and regional governments to exchange information. In Australia, an independent statutory authority, Infrastructure Australia, works with states, territories, and local governments as well as the private sector to conduct rigorous cost benefit analysis to identify infrastructure investment priorities. In addition, it identifies policy and regulatory reforms for coordinated delivery of critical investments. In Sweden, the National Forum for regional Growth and Attractiveness brings together national and subnational governments for political and strategic dialogue. 33 Box 6. Horizontal coordination among subnational governments Joint action by multiple subnational governments makes it possible to reach relevant scale in investments and internalize positive/negative spillovers of each investment decision. In France, inter-municipal structures with taxation authority facilitate horizontal coordination. Municipalities are allowed to get together to establish a public establishment for inter-municipal cooperation which is governed by the delegates of constituent municipal councils. In Slovenia, laws provide financial incentives for joint municipal administrations. Similarly, in Spain, joint municipal investment projects get priority for regional funds. Box 7. Payment for Ecosystem Services (EPS) Schemes in China In China, payment for ecosystem services (PES) schemes are used widely in different forms to address climate change externalities. PES is viewed as a public institutional arrangement for internalizing the externalities of ecosystem services. China experienced heavy floods in 1998 in major river basins as a result of extensive logging. The initiation of PES schemes was the attempt by government to fund sustainable ecosystem restoration initiatives. The National Forest Law of China was revised to establish a national fund to renumerate ecological services. Subsequently, several ecological conservation projects and programs rolled out: Grain for Green, National Forest Protection Program, Beijing, and Tianjin Sandstorm Source Control Project, Ecological and Environmental Protection Project in Three-river Source, Karst Rocky Desertification Control Project and returning Grazing Land to Grassland Project. 34 Annex 2: Quantifying the inefficiencies of a PIM system Countries spend significant amount of budgetary resources for public investment, including in infrastructure development.48 Improvements in the efficiency of public spending could have a major impact on aggregate productivity growth and gross domestic product levels. Recent policy research suggests that improving the efficiency of public investment, rather than increasing the level of public investment, offers a significantly positive yield in terms of boosting growth. A one percentage point increase in public investment through efficiency improvements increases growth rates by 0.1-0.2 percentage points over the next few years.49 However, the measurement and quantification of the efficiency/inefficiency of a PIM system is a challenging and elusive task in part because public investment, by its very nature, impacts the productivity of many sectors, both directly by enlarging economic activity, as well as indirectly by spurring private and foreign direct investment. There are two possible approaches for such an exercise: (i) macroeconomic measure, such as the ICOR, or, alternatively, (ii) a microeconomic measure based on portfolio review methodology. At the macroeconomic level, recent analysis shows that the productivity of investment in Vietnam as measured by the ICOR has been declining in part because of the misallocation of capital, both by the public sector and the private sector.50 The decline in both the level and efficiency of public investment in Vietnam in recent years is seen to be an important factor that threaten achievement of planned rates of GDP growth. Theoretical measure of efficiency/inefficiency is the level of output achieved relative to the inputs utilized. Two kinds of efficiencies can be estimated with regard to investment - allocative efficiency (whether the best investments are undertaken) and technical efficiency (how well those investments perform). Technical efficiency/inefficiency can be characterized as using less/more inputs to obtain a given level of output. In case of technical inefficiency, the production process uses more inputs than required optimally (or relative to a comparator) to obtain a given level of output. Allocative inefficiency refers to sub-optimal output given input prices and their marginal productivities. In a public investment context allocative inefficiency refers to the failure to invest (or adequately invest) in projects with significant externalities that markets would not provide. The estimated output and productivity foregone would provide a measure of such allocative inefficiency. For the estimation of technical and allocative efficiency/inefficiency, one can employ a macroeconomic measure, such as the ICOR, or, alternatively, a microeconomic measure based on portfolio review methodology. The advantage of the ICOR approach is that it is relatively easy to measure as the data requirements are straightforward and GDP and aggregate investment data is accessible. As noted, the ICOR for Vietnam since 2008 clearly suggests a trend of declining investment efficiency. It does not however provide a breakdown of inefficiency in terms of technical and allocative shares, nor does it distinguish between public and private investment. At the microeconomic level, the approach could be to collect data/information about a sample of projects to estimate economical and social benefits (quantified as the net benefits, in terms of economic NPV, over the discounted total investments). This is a data- intensive exercise that presupposes that there is high quality ex-ante cost benefit analysis and ex-post analysis of public investment projects. Thí micro approach is a much more laborious task and requires information about projects in the public investment portfolio at the national and subnational levels. Systemic review of the demand and cost analysis can be done to decipher: (1) whether the projects were actually economically and socially beneficial (quantified as the net benefits, in terms of economic NPV, over the discounted total investments). The analysis would reveal the financial and economic costs to doing such investment; and, with additional analysis relative to a comparator investment could reveal, (2) the extent of net loss from sub-optimal selection of projects as well 48 OECD countries spend about 20 percent of GDP in gross fixed capital formation (GFCF) of which public investment is about 3 percent of GDP. In Vietnam total GFCF is about 27 percent of GDP with public investment of about -- percent of GDP. 49 Devadas and Pennings (2018), “Assessing the Effect of Public Capital on Growth”, World Bank Policy Research Working Paper 8604. 50 Vibrant Vietnam (World Bank 2020), pp.26-53. 35 as due to the time and cost overruns (in terms of $$$ negative NPVs, either at design or execution stage or both). In practice, this is a data-intensive exercise that presupposes that there is high quality ex-ante cost benefit analysis and ex-post analysis of public investment projects, that there is the scope to do meta-analysis of such reviews. Finally, it requires that the collaboration of the MPI and sector ministries can be ensured, and that data sharing would be possible. In case of limited data availability, the estimation exercise can focus on sectors that consume a big chuck of public investment budget such as energy and transportation. However, it is imperative to have detailed information about the investment portfolio in those selected sectors at the national and subnational levels. At this time, it does not appear that systematic social cost benefit analysis is being undertaken of government funded investments either at the central or the provincial level. SCBA appears to be limited, in most cases, to donor-funded projects which are also often subjected to ex-post evaluation. Because the set of donor-funded projects is limited and does not cover all sectors there is an inherent bias in drawing efficiency conclusions from this approach. In light of the reality of data availability in Vietnam, a portfolio-based approach to measuring aggregate inefficiency seems a very difficult if not infeasible. As such a macroeconomic approach that disaggregates public and private investment and estimates ICORs for each may be considered as a second-best alternative. This could be complemented by selected examples of projects in major sectors where evidence of technical or allocative inefficiency can be found, together with institutional analysis about investment management. As a next step to this analysis, it is proposed that an in-depth Public Investment Efficiency Review be undertaken to try and quantify inefficiencies across the public investment management (PIM) and intergovernmental fiscal (IGF) systems. This approach would include a deep-dive portfolio review in key sectors, such as transport and water management, with the spatial perspective. 36 REFERENCE 1. Total Economy Database (https://www.conference-board.org/data/economydatabase/total-economy-database-productivity). 2. World Bank. 2017. “Vietnam Public Expenditure Review: Fiscal Policies towards Sustainability, Efficiency and Equity” Washington DC. 3. Yin Yin Lam, Kaushik Sriram, and Navdha Khera. 2019. “Strengthening Vietnam’s Trucking Sector Towards Lower Logistics Costs and Greenhouse Gas Emissions. Available at World Bank Document 4. Dat Nguyen. 2018. “Vietnam Warned It is Planning One Airport Too Many” Available at Vietnam warned it is planning one airport too many - VnExpress International 5. 2030 Water Resources Group. 2017. Vietnam: Hydro-Economic Framework for Assessing Water Sector Challenges. Available at ENG_Vietnam-Hydro-Economic-Analysis_August-2017.pdf (2030wrg.org) 6. Nguyễn Văn Huân: “Theoritical study on regional coordination” (2007); “Industrial policies for developing growth poles for regional linkages” (2008); “Regional coordination: From theory to practice (2012); “Intra-regional linakages for sustainable Highlands development” (State Research Thesis- 2014). 7. Trương Bá Thanh: Regional Coordination in the Central region and the Highlands: From theory to practice (2009). 8. Trần Thị Thu Hương: “Research on factors linking localities in the region in Vietnam: the case of the Mekong Delta region” (PhD thesis, 2017). 9. Nguyễn Quốc Toản: “Regional linkages in economic development in the Southern Central region” (2020). 10. Ngô Thắng Lợi, Trần Thị Hạnh, Vũ Cương, Trần Văn Thanh: “Inter-provincial coordination for regional development in Vietnam” (2014). 37