MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 1 Mor , B tt r or Diff r nt? Inv stin in P r u ’s Ro ds LATIN AMERICA AND CARIBBEAN REGION MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS © 2019 International Bank for Reconstruction and Development/ The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and con- clusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522- 2625; e-mail: pubrights@worldbank.org. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 3 Mor , B tt r or Diff r nt? Inv stin in P r u ’s Ro ds LATIN AMERICA AND CARIBBEAN REGION 4 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 5 TABLE OF CONTENTS Acknowledgments......................................................................................................................................................................8 Abbreviations / Glossary...........................................................................................................................................................9 Executive Summary ................................................................................................................................................................ 10 Chapter 1. Introduction.......................................................................................................................................................... 16 Chapter 2. Background and Road Network Features........................................................................................................ 19 A. Paraguay and Its Transport Sector .......................................................................................................................................... 20 B. Road Network Features ............................................................................................................................................................. 22 C. Spatial Efficiency: Public Access to the Road Network. ........................................................................................................... 26 D. Institutional Roles and Responsibilities................................................................................................................................... 29 Chapter 3. Road Sector Expenditures, Needs and Funding Sources................................................................................ 31 A. Budget Overview ........................................................................................................................................................................ 32 B. Rehabilitation and Maintenance of the Road Network – Needs versus Expenditures....................................................... 36 D. Sources of Revenues for Roads................................................................................................................................................. 46 Chapter 4. Government Road Sector Practices and Management (Funding, Procurement, Budget Execution)...... 52 A. Overview of Practices.................................................................................................................................................................. 53 B. Management Shortcomings: Weak Planning and Low Expenditure Efficiency . ................................................................ 53 C. Procurement and Value for Money in Public Expenditures .................................................................................................. 55 Chapter 5. Findings and Recommendations....................................................................................................................... 57 A. Better Budget Management and Governance Practices . ..................................................................................................... 58 B. Addressing Data and Capacity Issues . .................................................................................................................................... 59 C. Striving for Universal Access in Rural Areas . .......................................................................................................................... 60 D. Sustainable Sources of Revenue for the Sector...................................................................................................................... 60 REFERENCES .................................................................................................................................................................................... 62 Appendix A. Rural Road Network Distribution by Climate Zone...................................................................................... 64 Appendix B. Total Costs of Rural Road Maintenance......................................................................................................... 65 Appendix C. Multi-criteria Analysis to Identify Rural Road Priority Areas...................................................................... 66 6 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS FIGURES ........................................................................................................................................... 20 Figure 1: Population Density in Paraguay (2010). Figure 2: Quality of transport relative to neighbors (ranking)...................................................................................................................... 21 Figure 3: Population and Road Distribution.................................................................................................................................................... 23 Figure 4: Road Density by Surface in Paraguay and Comparators.............................................................................................................. 23 Figure 5: Road Density by Population in Paraguay and Comparators........................................................................................................ 23 ............................................................................................................................... 24 Figure 6: Road Network Composition and Growth (km). Figure 7: Share of Paved Roads in the Road Network (primary, secondary, and rural networks)........................................................... 24 Figure 8: Population and Road Distribution.................................................................................................................................................... 26 ........................................................................................................................................................... 26 Figure 9: Road Density by Department. Figure 10: MOPC Organizational Structure for the Road Sector.................................................................................................................. 29 Figure 11: MOPC Total and Roads Budget, Approved................................................................................................................................... 32 ............................................................................... 33 Figure 12: Public Sector Investment in Roads, Select Countries and LAC (% of GDP). Figure 13: Approved, Modified, and Executed Budget for the Road Sector, 2008–17............................................................................... 33 Figure 14: Approved, Modified, and Executed Budgets for Capital Expenditures on the National and Departmental Network........ 34 Figure 15: Approved, Modified, and Executed Budgets for Current Expenditures on the National and Departmental Network....... 34 .................................................................................................................................... 34 Figure 16: Road Length and 2018 Budget Allocation. Figure 17: Regional Distribution of Roads and Budget................................................................................................................................. 35 Figure 18: Road Length by Road Condition (Paved roads)............................................................................................................................ 35 Figure 19: Regional Shares of Force Account Expenditures for Maintenance............................................................................................ 35 Figure 20: Rural Road Capital Needs and Expenditures (actual).................................................................................................................. 40 Figure 21: Rural Roads Maintenance by Force Account................................................................................................................................ 40 Figure 22: Estimated Rural Road Capital Needs............................................................................................................................................. 41 Figure 23: Estimated Rural Road Maintenance Needs per Year................................................................................................................... 41 Figure 24: Rural Road Access (0 to 1)............................................................................................................................................................... 42 Figure 25: Financial Needs for Rural Road Development.............................................................................................................................. 42 Figure 27: Poverty and Road Accessibility....................................................................................................................................................... 43 Figure 26: Poverty Rates by Department......................................................................................................................................................... 43 Figure 28: Distribution of Crop Production Areas.......................................................................................................................................... 44 Figure 29: Spatial Distribution of Cattle........................................................................................................................................................... 44 Figure 30: Health Access and Road Accessibility............................................................................................................................................ 45 Figure 31: People’s Access to Health Care, by Length of Trip and Share of Departmental Population.................................................. 45 Figure 32: Health Access and Road Accessibility ........................................................................................................................................... 45 Figure 33: Funding Sources for the Road Sector............................................................................................................................................ 46 ........................................................................................ 47 Figure 34: Capital (Committed) Budget for the Road Sector by Type of Source. Figure 35: Source 20 Capital Road Sector Executed Budget by Subtype of Source................................................................................... 47 ................................................................................................................................ 55 Figure 36: Number of Bids Invited, by month, 2015-18. Figure 37: Number of Bid Invitations Closed, by month, 2015-18............................................................................................................... 55 Figure 38: Average number of Clarifications and Addendas, 2015-18........................................................................................................ 55 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 7 TABLES Table 1: Paraguay: Quality of Transport Relative to Neighbors................................................................................................................... 21 Table 2: Paraguay’s Road Network (in kilometers, 2012 and 2018)............................................................................................................. 22 Table 3: Paved Road Network Condition (2012 and 2018)............................................................................................................................ 25 Table 4: Projected Expenditure for the Paved and Unpaved Network (US$ million)................................................................................. 36 Table 5: Total Expenditures on the Paved (National/Urban) Road Network (US$ million)........................................................................ 37 Table 6: Estimated Value of the Road Asset Base........................................................................................................................................... 38 Table 7: Importance to Population of Rural Earth Road Network (km)....................................................................................................... 39 ..................................................................................................................................... 39 Table 8: Rural Roads Network by Surface Type (km). Table 9: Total Expenditures on the Rural (unpaved) Road Network (US$ million)..................................................................................... 41 BOXES .................................................................................................................................................................... 27 Box 1: The Impact of Rural Roads. ..................................................................... 37 Box 2: How Useful Is the Rigid Pavement Law (Given Some Unintended Consequences)?. Box 3: Road Funds as an Interim Step............................................................................................................................................................. 48 8 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS ACKNOWLEDGMENTS This report was developed by Kavita Sethi on the basis of a series of background papers, targeted interviews, and secondary data provided by the Government of Paraguay. The background papers and substantial contributions include those on road network evaluation (paved and unpaved roads) by Rodrigo Archondo-Callao and Atsushi Limi, transport and budget data by Pablo Iribarren Santos and Alex Giron Gordillo and spatial information and analysis by Li Qu and Zhen Liu. The Report relies heavily on roads data provided by the Ministry of Public Works and Communication and budget data from the BOOST Initiative and the Ministry of Finance. The team is grateful for the comments and guidance provided by Jesko S. Hentschel and Juan Gaviria, as well as for comments and helpful suggestions from Carole Megevand, Celia Ortega Sotes, Ruth Gonzalez llamas, Monica Tambucho and Lincoln Flores A special thanks is due to the following reviewers for their comments that have helped to improve the document: Vivien Foster, Daniel Benitez and Cecilia Briceno Garmendia. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 9 ABBREVIATIONS / GLOSSARY DCV Rural Roads Directorate, Dirección de Caminos Vecinales DV Road Directorate, Dirección de Vialidad Fonacide Fondo Nacional de Inversión Pública y Desarrollo MH Ministry of Finance, Ministerio de Hacienda MOPC Ministry of Public Works, Ministerio de Obras Públicas y Comunicaciones PYG Guaraní paraguayo RONET Road Network Evaluation Tool Rural Access is the share of rural population living within 2 kilometers (about 20 – 30 minutes walking time) Indicator (RAI) of an all-season road All-season road is a road usable year-round by motorized transport except for occasional interruptions of brief duration due to inclement weather. Road density by is the number of kms of roads per thousand population in a defined geographic or political population zone, region or country. Road density is the ratio of the length of the country’s road network to the country’s total land area. Road den- by surface sity may be specified with respect to the entire network which includes all roads in the country or with respect to a subset of the road network such as paved roads or highways. Sustainable as defined in RONET is the share of the road network in very good, good or fair condition and Network which requires periodic maintenance works, while roads in poor or very poor condition require rehabilitation works. Nominal or current values are derived from market prices without any adjustments for inflation. Constant or real prices adjust for the effects of inflation. Tolls are direct charges for public services and are retained by the collecting entity. User charges (or quasi prices) are indirect charges for infrastructure services that are often levied as fees on proxy transactions. The choice of proxy varies with the type of infrastructure. Earmarking refers to the pre-commitment of taxes to support or fully fund prespecified expenditure items. These revenues may be channelled through the general treasury or may be paid directly to a dedicated fund. First-generation were established in the 1960s and 1970s as extrabudgetary arrangements through which an road funds earmarked stream of tax revenues was put at the disposal of a road department or agency. Second-generation moved away from using earmarked tax revenues in the 1990s. Instead, they are funded by levies road funds or surcharges designated as “user charges” and identified separately from general taxation. 10 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS EXECUTIVE SUMMARY MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 11 CONTEXT AND FINDINGS budget allocations for the road sector. In real terms, road sector spending in 2017 was double the expenditure 1. Paraguay is a middle-income, landlocked country in 2008 while the share of roads in the overall government with a population of about 7 million. The nation is high- budget went from 4.0 percent in 2008 to 5.2 in 2017. Road ly dependent on its transport and logistics infrastruc- sector spending as a share of GDP mirrors those of Eastern ture to connect to regional markets and international European countries’ in 2009, as these countries attempted seaports. Currently, economic growth is reliant on trade of to rapidly improve their infrastructure to Western European natural resources and agriculture while the country seeks to standards. As a share of GDP, the road sector has received 2 diversify its economic base. However, there is a significant percent or more since 2012, which is well above the regional geographic, economic, and demographic divide that the average, and surpassed only by Bolivia. OECD average ex- roads sector needs to bridge. The Eastern Region concen- penditures meanwhile are considerably lower at around 1 trates 96 percent of the population and hosts most of the percent and reflect a more mature stage of infrastructure soy production while the Western Region is a relatively de- development.1 serted area with just 290,000 inhabitants in 2017 (4 percent of the total Paraguayan population) and 43 percent of the 5. The higher road sector budget has been directed to national livestock production. expansion and preservation of the paved roads since 2012 and has, on the whole, allocated sufficient funds 2. The existing length of the network, paved and un- for their maintenance as estimated by this Public paved, is considered sufficient to cater to the vast ma- Expenditure Review. Despite this there has been a slight jority of the population, notwithstanding spatial differ- decline in the overall quality of the network due to over and ences. The national, average road density by surface area of under funding of roads geographically and by functional 0.20 km per sq km is on par with Brazil with a density of 0.19 classification. The quality of the paved national roads has km per sq km, and much higher than Argentina and Bolivia improved while that of paved secondary roads has deteri- with densities of 0.09 and 0.08 km per sq km respectively. orated. The analysis of this PER suggests that to meet the The current paved road network, at the macro-level, serves targets of the “paving program”, the sector selected low about 70 percent of the total population. At the departmen- traffic volume roads and “paved” these with thin seals laid tal level though significant differences are observed in road over existing cobblestones. These “paved” roads had high- density values which range from 7 to 70 kilometers (km) of er roughness values and poor quality from the very start. road per 100 km2 of land area. The selection process for the paving program appears to have applied socio-economic rather than purely economic 3. In road quality and connectivity, on the other hand, criteria and this allowed inclusion of low traffic roads in the Paraguay trails its neighbors and other middle-income program. Both the selection of roads and choice of technol- countries. According to the Global Competitiveness Indices, ogy of the program were distorted and yielded an overall Paraguay is in the bottom third of indexed countries for decrease in the quality of paved roads. The direct cost of road connectivity and quality. The country’s lowest score such technical inefficiency is estimated at about US$160 relates to the quality of its road infrastructure, where it lags million for the period 2012-2017. behind neighboring countries Argentina, Brazil, and Bolivia, and where it places among the four poorest performers in 6. Budget execution is hampered by Institutional inef- the Latin America and the Caribbean (LAC) region. Its per- ficiencies and the road sector budget is underspent to formance is also below the upper-middle-income-country the tune of US$200-400 million annually or 25-35 percent average. These competitiveness ratings reflect a need to of the annual budget and implies a maintenance deficit understand the underlying issues driving quality and poor which leads to higher rehabilitation costs in the future, connectivity and find solutions to improve road infrastruc- at a ratio of 1:3.5. There is considerable volatility in the an- ture overall. nual approved and modified budgets, in 2014, for example, the modified budget of US$818 million was 130 percent 4. The government has attempted to address per- higher than the approved budget. Such fluctuations have an ceptions on quality and connectivity through higher adverse impact on the planning, execution and absorption 1 Source: OECD (2013) 12 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS of funds when contracts are lumpy and multi-year as in the 9. Budget allocations appear to favor the eastern, case of road works. Poor management of the budget pro- southern and central regions2, considering the distri- cess and works planning are attributed to the institutional bution and quality of the road network. The central and “policy environment” which determines what is allocated eastern regions account for about 48 percent of the total and what is actually spent. Assuming that about half the paved road network but received about 65 percent of the amount left on the table is applicable to maintenance activ- total budget allocated to road projects in 2018. By contrast, ities and given that additional road user costs in Paraguay the rural Chaco region has 14 percent of the paved road are US$3.5 for every dollar of deferred maintenance, a con- network but received only 3.2 percent of the total road-sec- servative estimate of the cost of institutional inefficiency for tor budget. Thus, in broad terms there is overinvestment in the road sector in Paraguay is somewhere between US$350- regions where road conditions are relatively good, i.e. par- 700 million annually. ticularly the central and eastern regions, leaving roads in other areas in worse condition with inadequate investment. 7. The road sector also suffers from a complex, bureau- The actual financial needs of a department must obviously cratic system with inadequate oversight and accounta- be determined by taking account of its adopted engineering bility. Policy formulation and road sector management are designs, current road conditions, procurement costs and both the responsibility of MOPC, responsibilities which are other factors. normally separated institutionally to avoid conflict of inter- est issues. The Ministry of Finance (MOF) should, in princi- 10. While preservation of the paved network is the key ple, assess the efficiency of sector expenditures but there is to continued, reliable access by some 70 percent of the an almost complete absence of data, mechanisms and tools population, there is an enormous, unmet demand for to track performance and hold MOPC accountable for the rural road development to address social equity and ru- quality of spending. Technical staff have very little auton- ral development issues. In rural regions, a few paved roads omy, and ministerial and vice-ministerial approvals are re- are spread across a large area and this restricts access to all quired at each step of the procurement and implementation weather-roads to about 42 percent of the rural inhabitants. processes. The budget and project execution systems are Additional access is provided by earth roads with little or no process-heavy, complex and time consuming – there is low maintenance. While the annual modified budget would suf- visibility on the location of bottlenecks and the overall result fice to address in full the maintenance needs of rural roads, is low execution rates for the sector. the budgetary volatility and institutional issues mentioned earlier lead to underfunding of maintenance by about 50 8. One of the main contributions of this PER is the de- percent. Improving earth roads to provide all-weather ac- velopment of physical, financial, and spatial databases cess requires a judicious review of the technologies avail- and profiles for the road sector that have hitherto been able to improve such roads. The traffic threshold at which unavailable. There is a major lack of consistent, good qual- paving an unsealed road is economically justified is mainly ity data on both physical and financial aspects of roads. To a function of the unit cost of the paving works, the traffic set and monitor programs, technicians and policy makers composition and its expected growth, and the condition of require quality and timely disaggregated data. Reliable the unsealed road. In a wet environment, an unsealed road planning, prioritizing, and budgeting of sector expenditures deteriorates rapidly due to rainfall. In a dry environment, an require regular surveys of traffic and road conditions, which unsealed road can remain in passable condition for much are not being conducted in the Paraguayan context. There longer periods and is a more economical option. appear to be no mechanisms in place to relate expenditures to physical outcomes or decision tools to support review of budget proposals or evaluate the efficiency of sector expenditures. 2 As   of 2018, MOPC groups departments into five regions for budgeting purposes. The departmental composition of regions is shown below. Chaco Norte Este Centro Sur Presidente Hayes Concepción Caaguazú Cordillera Itapúa Alto Paraguay San Pedro Caazapá Guairá Misiones Boquerón Amambay Alto Paraná Paraguarí Ñeembucú Canindeyú Central MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 13 11. A move toward sustained, universal access, espe- • Striving for universal accessibility in rural areas, and cially in remote areas, therefore remains a challenge • Sustainable sources of revenue for the sector. that is being addressed but requires further effort. The country does not have a rural roads policy or a strategy to 15. Recommendations on how to improve public man- achieve the SDG goal of universal accessibility at present. agement and governance practices address a core set The roads sub-sector has been receiving less than 50 per- of administrative, organizational, and budgetary issues: cent of its maintenance needs, as RONET model estimates show. Advancing to a universal access goal, that is upgrad- • Experience with road administrations around the world ing roads to a gravel standard, would require an expend- indicates the importance of separating the policy and iture of about US$5 billion. A 100 percent access rate for implementation functions of roads administration. the rural population requires the construction of additional However, the Paraguayan Ministry of Public Works roads to dispersed and remote habitats at an additional, (Ministerio de Obras Públicas y Comunicaciones) is estimated cost of about US$2 billion. An estimated US$300 charged with both functions, with no real separation million annually would be needed for maintaining these between these two aspects of road administration. Nor new and rehabilitated roads. are there mechanisms or tools in place to monitor and evaluate sector performance and encourage efficient 12. A sustainable long-term revenue source for the expenditures. While sector accountability can be en- road sector is another issue. Road sector development hanced by requiring independent annual audits of all and maintenance is financed through a complex mecha- capital and recurrent expenditures and sharing these nism involving multiple domestic and external financing results in an annual sector report in the short term, clear sources. The road sector receives funds sourced from taxes separation of responsibilities, possibly through an inde- and other government revenues; international and domes- pendent road entity, should be a medium term goal. tic credit; and grant funds. While the share of multi-lateral financial institutions (MFIs) has increased over the last dec- • The execution process of public investments in Paraguay ade, sovereign bonds are nowadays the dominant source for road construction, improvement, and maintenance of capital. There are in turn only a few tolled roads and con- is still highly complex and bureaucratic. While discipline cessions that rely on user charges, both of which provide a is important to protect the integrity of expenditures, the more sustainable source of revenue. The bulk of road sector investment cycle needs to be simplified and optimized. expenditures are channeled through the Ministry of Public Further, the Ministry of Finance should develop both the Works (MOPC) and departmental and municipal govern- tools and technical capacity to review the proposed road ments receive some limited budgets for infrastructure. network budget and planning, and subsequently con- firm that expenditures are aligned to these. The latter 13. This Road Sector Public Expenditure Report (PER) as- can be achieved through ad hoc financial and physical sesses current practices in the budgeting and expenditure of audits of road sector expenditures by the MOF. These public resources in the road sector in Paraguay. The Report changes would strengthen the alignment of annual identifies gaps and potential efficiency gains in the function- budgets with the MTEF and help reduce the current un- al and spatial allocation of resources, spending execution, derspending by MOPC by anywhere between 25 to 35 and resource management efficiency. It also considers the percent of the annual budget. adequacy of resources and opportunities for sustainable funding from user charges. Further, the PER looks at ways • Project preparation processes and cycles are opaque to rebalance and afford the investment and maintenance and show considerable bunching over the calendar needs of rural roads which are a priority to meet the access year, instead of a manageable, well spaced sequence of requirements of approximately 1.8 million beneficiaries. preparing and implementing projects. Increasing num- bers of complaints from bidders and large numbers 14. The outcome of this PER is a set of recommenda- of addenda (that seek to clarify contract terms) signal tions that address four basic themes: weak document preparation linked to low technical and procurement capacity. These issues can be addressed • Better public management and governance practices, by requiring MOPC to start preparing specifications and • Addressing data and capacity issues, procurement documents as soon as a project enters the 14 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS budget process, tracking and oversight of the project commensurate with sector responsibilities and planned preparation cycle, and finally, emphasis on developing growth. Budget allocations for capital and recurrent and retaining staff skills through regular training. expenditures have increased substantially over the past seven years, challenging MOPC’s capacity to manage • In terms of budget allocation, the Government of the annual budget and deliver high-quality works. A Paraguay’s emphasis on the operation and mainte- functional review can help align staffing needs and skills nance of the paved network seems to have led to im- with the current and future needs of the sector. provements in road conditions over the past five years and should continue to be the focus of MOPC efforts. 17. Striving for universal accessibility in rural areas: a At the same time, any funding allocations to upgrading rural roads policy and strategy are necessary to address or reconstructing unpaved, secondary and rural roads universal accessibility goals, ensure appropriate technolo- need to take into account the underinvestment ob- gies, such as gravel roads, for all-weather access, and lessen served in some departments, especially in the poorer the cost and time frame for achieving these. rural North and Chaco regions with the lowest levels of road access. It is recommended that standard, asset • The rural road network is a vital public resource for management tools are applied in the development of the country’s more remote areas, despite their lack of the paved roads budget while a transparent, multi-crite- economic weight and their dispersed population. A ria framework is the basis for planning and prioritizing rural roads strategy would be helpful for determining expenditure on unpaved roads. Failure to address the investment priorities based on transparent and objec- discrepancies in rural access in these regions, combined tive socioeconomic criteria that are appropriate in a with their remoteness and difficulties in access to public rural context. Thus, it would help improve the efficiency services and markets, will inevitably perpetuate inequal- of project selection and public expenditures as well as ity. A robust data and information system are needed to address inequities in the rural population’s access to inform the budget allocation process. economic, educational and health facilities. 16. Addressing data and capacity issues will integrate • Such a strategy should include an objective prioritiza- evidence-based decision-making with a better-skilled tion framework which is important for systematically staff. allocating limited resources to high-priority roads in a timely manner, while avoiding unnecessary political in- • A systematic and up-to-date survey of the quality of volvement. The strategy should also address the issue the paved network, as well as the traffic using it, is of the resources required to advance the goal of univer- necessary for the meaningful application of asset man- sal access. agement models and to determine network spending requirements. • The allocation of the roads budget is disproportionately skewed towards paved roads and a gradual rebalanc- • While maintenance expenditure on the paved network ing of funds towards the rural sector is necessary. It is appears broadly adequate, there is a concern that the estimated that about $62 million are needed annually amount, spatial allocation, and type of capital works for maintenance as against the average $28.0 million funded may be inappropriate, especially if expenditure (2013-17) currently expended on rural roads. This main- requirements are based on guesswork regarding traffic tenance gap can be easily covered through, for exam- and road conditions. Conducting a comprehensive sur- ple, a higher budget execution ratio. The rural access vey of the paved road network, developing mechanisms indicator is currently at 42 percent – leaving roughly 60 for regular updates of the roads database, and making percent of rural residents without reliable access in the these data publicly available alongside budget data, are long term has implications for the rural economy, pov- perhaps the best guarantees of improving value for erty reduction and well-being. A stepped approach to money in the roads sector. achieving universal access is recommended. • A functional review of road sector staff and profiles is 18. Sustainable sources of revenue for the sector: at pres- needed to develop both a staffing and training plan ent, Paraguay’s road funding mechanism pools capital from MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 15 various sources, relying heavily on donor resources and, 19. The recommendations presented here are realistic more recently, on domestic borrowing. Road user charges and, if implemented together, can bring improvements provide a more sustainable source of funding since they can quickly and efficiently. These steps can advance govern- be aligned with the maintenance expenditures of the net- ance, road sector management skills, better budgeting and work and channeled through a dedicated fund, if one exists. expenditures, evidence-based planning for both primary and secondary and rural road networks and achieve the • Road funds, managed by independent road boards, are road sector needs of both rural and urban Paraguay. often recommended in situations of insufficient or un- certain budgetary allocations to help avoid deterioration of road assets. However, dedicated funds also reduce Disclaimer – This PER report includes a first-ever database fiscal flexibility and may be just as poorly managed as of physical and financial data for the roads sector, vetted budgetary allocations. Gwilliam et al suggest that there for coherence and reliability. However, the report needs to be are, in general, two long-term institutional options for read with the caution that the analysis relies on various sourc- reconciling fiscal prudence with asset maintenance: a es of incomplete and imperfect network and budget data. The road agency operating on commercial principles, or a data were obtained either directly from the MOPC, reports and reformed and well-functioning budget process with ap- databases developed by consultants to the MOPC, and budget propriate checks and balances. data from BOOST and the MOF. Reconciling data on the size, surface type, and condition of the road network, as well as its • Paraguay’s paved network has been receiving ade- budget, has been one of the biggest challenges in the prepara- quate maintenance funds, notwithstanding uncertainty tion of this study. There is substantial variation in the numbers in the budget allocation and transfer process. In this across different reports, and it is often impossible to reconcile situation, careful consideration of the sector’s long- the analysis within a given report or across reports. As such, the term institutional preferences, together with any need present review applies best judgement in finalizing the numbers for a provisional road fund, in the form of an active presented, and these may not always match road and budget Sistema de Infraestructura Vial del Paraguay (SIVIPAR), data cited elsewhere. is recommended. This consideration should include an assessment of the revenue sources, the structures of user charges, the affordability of proposed user charg- es (specifically the affordability of tolls and fuel taxes) for lower income populations, and the adequacy of the build-up of revenues toward a sustainable financial mod- el. Appropriate checks and balances in the system to pre- vent any potential misuse of funds are also necessary. • The existing system of tolls and concessions to fund the preservation needs of the core network should be expanded. A review of options to increase toll revenues in line with road usage and damage can more fairly dis- tribute the burden of responsibility, with heavier traffic such as semi-trucks covering a great portion of mainte- nance than light traffic, such as cars and pickup trucks. Concessions are an important source of private funding and management for the road sector, but to expand their use, Paraguay needs to strengthen private sector confidence in the rules and regulations governing con- cession contracts. 16 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS CHAPTER 1. INTRODUCTION MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 17 20. This Public Expenditure Review (PER) of Paraguay’s – regulation, financing, oversight, or outright provision (ei- road sector aims to: ther via government resources or through contracts with private companies), and public-private partnerships (such as • Review the size, composition, and spatial allocation of toll roads or bridges) when government funds are not suffi- the flow of funds to the sector; cient. Roads are a mixed public and private good: rural and • Review the structures of governance and the function- uncongested interurban roads are typically nonrival goods ing of public institutions; since adding another driver does not reduce the value of • Assess the quality of public spending in the sector, in- the road to someone else. Roads are excludable but not cluding distributional issues and public financial man- always profitable in view of the high upfront costs of roads agement, and try to link sector spending to outcomes; construction. Roads therefore constitute public goods with and, a strong rationale for public provision. Access to interurban • Provide recommendations to improve the efficiency, roads can be managed by tolling (to prevent traffic conges- equity, and sustainability of sector expenditures. tion), and roads can, in principle, be built and operated by the private sector. In Paraguay, some interurban roads are 21. The PER analysis is conducted within an expenditure tolled, and there is the possibility to increase the share of choice framework applied to the level and type of ex- such roads in total, although tolls may risk excluding poorer penditures. This analysis of the efficiency of public expendi- users. The majority of roads, however, will continue to be tures on roads is organized principally around allocational public goods. issues, both spatial and functional, and the nature of the   institutional “policy environment” that determines what is 23. This PER is structured as follows: allocated, what is actually spent, and what the outcomes are. • Chapter 2 provides some background on the Paraguayan • The functional analysis considers homogenous group- economy, the country’s road network (primary, second- ings of roads – primary and secondary roads and rural ary, and tertiary) features and analysis, and an overview roads – and addresses the quality of the roads for ad- of the government institutions responsible for the vancing economic and social requirements. network; • Chapter 3 describes how Paraguay budgets and man- • The spatial dimension is explored at the departmental ages its road sector, what the funding sources are, and (provincial) level to determine if road expenditure allo- how efficiently the expenditures are being spent; cated to the spatial network is likely to yield the highest • Chapter 4 assesses Paraguay’s goals for its road sector, economic and social welfare returns (in terms of effi- the effectiveness of its budget execution and sustaina- cient economic activity, optimized traffic concentration, bility of its funding, and its sector monitoring practices; and connecting the country’s population with social and resources and markets). • Chapter 5 concludes with the main findings and recommendations. • Technical efficiency and sustainability are considered for each functional group by evaluating: capital ex- 24. Budget and physical data and information for roads penditures on road rehabilitation and upgrade; current was obtained from government documents and verified or recurrent expenditures that include wages and sal- through interviews and consultations with sector actors. aries, operations and maintenance, and interest pay- Budget data was obtained from BOOST (the World Bank pub- ments, among other things; and funding to meet these lic expenditure database), the Department of Finance in the expenses. Ministry of Public Works and Communications, and online data published by the Ministry of Finance. Budget lines were 22. The economic rationale for public spending in the individually examined and assigned to the correct capital or roads sector rests on the need to overcome market fail- recurrent expenditure category for each functional group. ures and reduce disparities in access to road infrastruc- Expenditure estimates were obtained from the Planning ture. Government intervention in general is required when Department, Dirección de Vialidad, and the Department there is an underlying market failure to be addressed. The for Rural Roads within the Paraguayan Ministry of Public type of failure indicates the form of government intervention Works (MOPC). The Road Network Evaluation Tools (RONET) 18 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS methodology was used to assess road quality. Partial data on the physical characteristics of rural roads was collected from the Planning Department in MOPC and verified in tech- nical workshops. Information for the institutional analysis is for the most part from secondary sources but verified to the extent possible in face-to-face interviews with sector repre- sentatives. Data from various sources was cross-checked to avoid double counting. A full review of procurement data is outside the scope of this PER. However, available informa- tion was reviewed to understand the impact of the budget cycle on procurement processes and contract prices. 25. Open-source data were used extensively for spatial analysis and to supplement information on the road network. For the spatial analysis, data on population distri- bution and economic and social centers were downloaded from World Pop, a geospatial demographic website, while rainfall data was obtained from WorldClim.org. All spatial data were processed in ARCGIS3. Poverty and economic data from Anuario Estadistico del Paraguay, 2016, were used to complement expenditure data. The initial findings of the PER were presented and discussed with Government stakeholders at the technical and policy level. 26. One of the main contributions of this PER is the de- velopment of physical, financial, and spatial databases and profiles for the road sector that have hitherto been unavailable. The team worked with partial and sometimes unreliable data and advises caution when interpreting some of the results. Data scarcity, however, poses problems well beyond the scope of this review. To set and monitor pro- grams, policy makers require quality and timely disaggre- gated data. Reliable planning, prioritizing, and budgeting of sector expenditures require regular surveys of traffic and road conditions, which are not being conducted in the Paraguayan context. 3 A Geographic Information System (GIS) for compiling geographic data; creating maps and analyzing mapped information. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 19 CHAPTER 2. BACKGROUND AND ROAD NETWORK FEATURES 20 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS A. PARAGUAY AND ITS TRANSPORT SECTOR 27. Paraguay is a middle-income, landlocked country macroeconomic stability. Looking ahead, to meet the aspi- with a population of 7 million. Currently, economic growth rations of its population, the country needs to sustain this is reliant on trade of natural resources while the country economic growth and make it more inclusive. For this to seeks to diversify its economic base. Figure 1 shows that the happen, Paraguay will have to address structural challenges population is concentrated in the “Oriental” (eastern) half of related to (i) concentration of resources and production, (ii) the country and only 3 percent of the total population resides extent of the informal economy, and (iii) gaps in the delivery in the Gran Chaco (or “Occidental”), a semi-arid area to the of high-quality public services to citizens. west, which accounts for 60 percent of the land territory and includes the rural departments of Boquerón, Alto Paraguay, 29. Continued economic growth, combined with urban and Presidente Hayes. Around 61 percent of the Paraguayan population growth, will place increasing demands on population is urban, with a just under half living in the cap- Paraguay’s transport, and specifically its road infra- ital, Asunción. The urban population is expected to grow at structure. As a landlocked country, Paraguay is highly about 1.7 percent per annum over the next five years. dependent on its transport and logistics infrastructure to connect to regional markets and international seaports. Figure 1: Population Density in Paraguay (2010) The road sector is the dominant mode of transport in the country, concentrating nearly 90 percent of freight volume movements. Higher costs of road transport not only ham- per the profitability of exports but also that of producers and service providers for the domestic market. In the ur- ban centers, underinvestment in road and transport infra- structure limits access to jobs and services, especially for the poor living in peripheral areas. In both rural and urban areas, poor transport infrastructure adversely affects the well-being of the population in a variety of ways, e.g. high- er transport costs impede access to health and education facilities and impact the profitability of agricultural exports from small farms. Moreover, 38 percent of the population lives in rural areas where weak infrastructure links between regions can worsen the impact of shocks such as weather and climate-related events. 30. Paraguay lags behind its neighboring countries Source: SEDAC 2019 in road quality and connectivity. In terms of the qual- ity of transport infrastructure, as measured by the Global Competitiveness Indices, Paraguay needs improvements to 28. The economy has seen solid gross domestic product catch up with neighbors and the region as a whole (see Table (GDP) growth, strong poverty reduction, and macroeco- 1). Both road connectivity and low-quality performance nomic stability in recent years. GDP grew at 4.7 percent place Paraguay in the bottom third of the countries includ- per year on average in the 2004–2016 period, while poverty ed in the competitiveness indices. The country performs rates fell by 50 percent4 compared with 2003. Paraguay has worst in the quality of its road infrastructure, where it lags thus done better than the regional average on both counts. its neighboring countries, Argentina and Brazil, and even These results are due largely to the exploitation of natural Bolivia, and places it among the four poorest performers in resources for trade and a demographic dividend delivered Latin America and the Caribbean (LAC). Its performance is by an expanding working-age population, supported by also below the upper-middle-income-country average. 4 Measured at 5.5 US$ 2011 PPP (Paraguay Systematic Country Diagnostic, June 27, 2018). MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 21 Table 1: Paraguay: Quality of Transport Relative to Neighbors Paraguay Argentina Brazil Bolivia UMIC Score Rank Score Rank Score Rank Score Rank Score Rank Quality of Transport 33.8 110 47.2 72 43.5 84 29.3 126 47.26 73 infrastructure Road Connectivity index 47.3 99 92.6 11 63.7 74 39.8 114 60.3 72 Road Quality index 2.5 129 3.4 93 3.0 112 3.4 98 3.7 80 Efficiency of seaport 3.4 95 3.7 81 3.1 105 1.9 135 47.6 74 services Source: The Global Competitiveness Report, WEF (2018). Note: Quality, efficiency indices: 1-7 (best) or score 1-100 (best), Rank: 1-135 (worst). 31. The unreliability of the transport infrastructure is Figure 2: Quality of transport relative to neighbors (ranking) also seen as a major constraint by private businesses. The World Bank’s 2017 Enterprise Survey shows that 27 per- 0 20 40 60 80 100 120 140 160 cent of Paraguay’s businesses perceive transport infrastruc- ture as a “major constraint” to doing business, compared Quality of Transport to about 24 percent of firms in the LAC region which see infrastructure transport as a major obstacle, and about 19 percent world- wide. Geographically, 30 percent of the firms in Asuncion Road Connectivity index consider transport a major obstacle and 22 percent of those are located in the central region. Road quality index 32. The business community’s adverse rating of Paraguay’s road quality persists despite quality im- Efficiency of provement being a Government policy priority since seaport service 2012. Rapid economic growth between 2004 and 2017 was accompanied by an ambition to improve the quality of the Paraguay Argentina Brazil Bolivia UMIC country’s overall infrastructure to enhance the country’s competitiveness, connectivity with neighboring countries, Source: The Global Competitiveness Report, WEF (2018). and inclusiveness. These objectives are enshrined in the Note: The higher the rank, the lower the quality: 1-135 (worst) Plan Nacional de Desarrollo Paraguay 20305, 2014, and the Transport Sector Master Plans of 2012 and 2018. 5 Plan Nacional de Desarrollo (PND, 2014). 22 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS B. ROAD NETWORK FEATURES 33. The total length of Paraguay’s classified road net- 34. The 2012 Transport Master Plan (PMT) of the work was estimated at 80,127 kilometers6 in 2018. The Ministry of Public Works proposed to address quality is- country’s functional classification of the road network is as sues in the road sector primarily by paving more roads follows: and addressing the maintenance needs of the paved network. The Master Plan also proposed improving net- • Primary or national roads connect the main cities and work connectivity through investment in about 12,000 km provide Paraguay’s links to the outside world of roads (national, departmental, and rural) and mentioned • Secondary or departmental roads ensure the inter- the need for institutional reforms to improve sector perfor- connection of departmental (provincial) capitals as well mance. While the Plan proposed a paving program, it did as the main cities; and not clarify a prioritization framework. The rural subsector in • Tertiary or rural roads provide interconnection within particular received little attention, and no specific strategy departments and links to the outside world for villages for investing in, or managing, rural roads was proposed. and agricultural production centers. 35. The network length in 2018 comprised 4.7 percent Roads fall under the purview of the Direccion de Vialidad national roads, 21.2 percent departmental roads, and (DV) and the Dirección de Caminos Vecinales, both depart- 74.1 percent rural roads. Less than 9 percent of the net- ments within the Ministry of Public Works. There is a signif- work was paved, with the following breakdown: 91 percent icant additional number of tertiary roads that fall outside of of national roads paved, 23 percent of departmental roads the “official” network, and which are managed in principle by paved, and less than one half percent of rural roads paved. departmental and lower-level authorities. These are at pres- National roads have the highest network utilization (vehicle ent estimated by MOPC to number about 20,000 km. MOPC km/year) at 61 percent, while the utilization of departmental is in the process of redefining the classification criteria for roads is lower at 38 percent. In other words, the road net- national, departmental, and rural road networks. Once the work in 2018 carried around 6,477 million vehicle-kms per new classification is finalized, it is expected to inform and year, of which 61 percent travelled on primary roads that impact the planning of maintenance and new investments. represent 4.7 percent of the network length. The average network traffic is estimated at 3,146 vehicles per day, while the average for national roads only is 5,307 vehicles per day. Table 2: Paraguay’s Road Network (in kilometers, 2012 and 2018) 2012 2018 Responsible authority Paved Unpaved Paved Unpaved Primary (national) 3018 No data (ND) 3190 304 Vialidad Secondary (departmental) 2437 ND 3552 11959 Vialidad Tertiary (rural) ND 244 60878 Vecinales Total 5455 ND 6986 73141 Source: MOPC, Plan Maestro de Transporte, 2018. Paraguay Road Network Strategic Evaluation, RONET, Callao, R. Drafts March 2019. Note: Paved roads are asphalt-concrete and surface treated roads; cobblestone and gravel roads are not counted as paved roads. 6 Draft Report, Rural Roads Maintenance Needs Evaluation, Archondo-Callao, March 2019. First Draft, GIS Database. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 23 Figure 3: Population and Road Distribution Figure 4: Road Density by Surface in Paraguay and Comparators Road Density by Surface (km/sq. km) 0 0.1 0.2 0.3 0.4 0.5 Argentina 0.09 Bolivia 0.08 Brazil 0.19 Paraguay 0.20 Uruguay 0.44 LAC average 0.22 Source: WB staff calculations; data from GIS Roads Database Figure 5: Road Density by Population in Paraguay and Comparators Source: WorldPop and MOPC. Road Density by Population (km/ 1000 population) 36. According to the latest road statistics, the density of Paraguay’s road network by surface area is more or 0 5 10 15 20 25 less in line with neighboring countries. Its road density by surface area of 0.20 km per sq km is on par with Brazil with a density of 0.19 km per sq km, and much higher Argentina 5.50 than Argentina and Bolivia with densities of 0.09 and 0.08 km per sq km respectively (see Figure 4). Road density by Bolivia 8.20 department ranges between 6 and 68 km per 100 km sq. Road density by population is relatively high at 11.76 kms Brazil 7.47 per 1000 inhabitants and ahead of Argentina, Brazil, and landlocked Bolivia which have population densities of 5.5 Paraguay 11.76 kms, 7.5 kms, and 8.2 kms per 1000 inhabitants, respec- tively (see Figure 5). Paraguay’s national averages are, how- Uruguay 22.57 ever, skewed by the thinly populated Chaco region which accounts for 3 percent of the population but 61 percent of LAC average 7.48 total land area. If the Chaco region is excluded, the national average road density increases to 0.38 kms per square km of land area, while population coverage decreases to 9 km Source: World Bank compilation of different data sources, including IRF WRS 2018, 2017, 2016; CIA World Factbook; Infralatam; Dulac, World Bank Open per 1000 inhabitants. Data. 24 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 37. Different types of road surfacing are employed on and 2018, an additional 2,715 km of roads were paved, or an the road network, depending mainly on traffic levels average of 271.5 km per year. Meanwhile, about 76 percent but also on the economic and social significance of the of the entire 2018 network, and 97 percent of rural roads, road. Road paving using flexible (surface treatment) or rigid are earth roads. Paraguay still lags behind the region in the (asphalt concrete) methods are largely employed on the na- share of paved roads and, at under 9 percent, the country is tional (primary) roads and the major departmental (second- behind Brazil, Argentina, and Uruguay (see Figure 7). ary) roads. Fewer trafficked roads are being built with stone, cobblestone, gravel or earth. The different properties of sur- 39. The International Roughness Index (IRI) measures face types result in variations in maintenance requirements road quality and is widely employed to analyze changes and expected life, as well as “ride” quality. A breakdown of in pavement ride quality, over time. On average, during the different surface types is shown in Figure 6. the period 2012–18, the quality of the paved network in Paraguay suffered some deterioration despite an increase 38. The most important development in recent years in the share of the national budget being directed to the has been the increase in the length (kms) of paved roads sector. Road condition surveys indicate that in 2012, roads, especially since 2013, which has resulted in a 63 percent of the national paved road network was still significant change in road reclassification. Paved roads in “good” condition, whereas by 2018 this had fallen by 8 accounted for roughly 20 percent of the network reported percentage points to 55 percent. The network as a whole for 2013. In 2018, this ratio fell to below 9 percent because showed a decline in the share of sustainable7 roads of 20 several unclassified rural roads and paths became formally percentage points, from 85 to 65 percent (see Table 3). “classified” and are now managed by MOPC. Between 2008 Figure 7: Share of Paved Roads in the Road Network (primary, Figure 6: Road Network Composition and Growth (km) secondary, and rural networks) Paraguay: Road Network Composition and Growth, 2008-18 Share of Paved Roads in the Road Network 70,000 0 2% 4% 6% 8% 10% 12% 14% 16% Length of Road Network (km) 60,000 50,000 Argentina 14.61% 40,000 61,122 Bolivia 10.80% 30,000 20,000 23,310 24,195 Brazil 13.70% 14,893 10,000 4860 5474 28% h 7003 Paraguay 8.70% 4068 3889 4023 0 1524 2482 2008 2010 2013 2018 Uruguay 10.00% Stone, Cobblestone and Gravel Paved Earth Source: MOPC. Note that for 2018 only, an estimate of earth roads outside Source: Dirección Nacional Vialidad Sisvial (GIS) (for Argentina) / WB Staff the administration of the MOPC has been added. calculations. 7 In RONET, roads in very good, good or fair condition constitute the sustainable network and require periodic maintenance works, while roads in poor or very poor condition require rehabilitation works. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 25 Table 3: Paved Road Network Condition (2012 and 2018) 2012 2018 Classification Paved kms Road Roughness Sustainable Paved kms Road Roughness Sustainable Average IRI Network Average IRI Network National 3018 3.0 86% 3395 3.0 77% Departmental 2437 3.7 85% 4031 4.9 57% Rural 229 5.8 51% Total 5455 3.4 85% 7655 4.1 65% Source: Archondo-Callao, Paraguay Road Network Strategic Evaluation, March 2019. Draft. 40. While there was some change in the performance of paved roads). This 1,823 km increase consists of largely the paved national roads, there was a marked decline bituminous surface treated roads, and thus these roads in the condition of paved roads in the departmental advanced from unpaved to paved status. The deterioration network between 2012 and 2018. From 2012 to 2018, the in overall condition appears to be attributable to the length average roughness of the national network continued at 3.0 of paved roads added to the network given that the 2012 IRI while the share of the network in sustainable condition roads maintained their condition status in 2018. The data decreased from 86 percent to 77 percent (see Table 3). In does not indicate if these “new” roads deteriorated quickly contrast, the average roughness of the departmental net- due to lack of adequate maintenance, or if they started at a work increased from 3.7 IRI to 4.9 IRI and the share of roads high level of roughness due to poor construction standards in sustainable condition decreased from 85 percent to 57 or moved to poor condition from a combination of both percent over the same period. these factors. Surface treatment level of “paving” may be adequate at current levels of traffic where about a third of 41. The fall in the share of sustainable departmental the paved departmental roads carry less than 1,000 vehicles roads appears to be related to the increase in sur- per day, on average, with many carrying fewer than 400 ve- face-treated (“paved”) roads. Between 2012 and 2018, the hicles per day (definition of a low-volume road)8. But these length of paved, departmental roads increased from 2,437 roads will deteriorate quickly as traffic levels increase and km to 4,260 km (this number includes 229 kms of rural, quicken the declining condition of these roads. 8 Transportation Research Circular E-C167: The Promise of Rural Roads, Mike Long, Chair, TRB Low Volume Roads Committee. 26 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS C. SPATIAL EFFICIENCY: PUBLIC ACCESS TO THE ROAD NETWORK 42. Despite spatial differences, the existing road net- 43. Given the current population and road distribution, work is sufficient to cater for the vast majority of the an estimated 69 percent of the country’s population population. It is estimated that the current road network lives within 2 km of a paved road. This assumes that all can serve about 90 percent of the total population in terms paved roads are in good condition and thus transitable (we of current population and road distribution (see Figures 8 know this to be optimistic since the paved road network and 9). Coverage exceeds 80 percent in most departments, contains roads that are in poor condition). However, the except the Chaco region, where coverage remains about 50 paved road network of Paraguay provides good overall ac- to 65 percent, meaning that around half of the departmen- cess to 69 percent of Paraguayans. This also indicates that tal population lives more than 2 km away from the current population groups are highly concentrated around paved road network. In other words, even if the entire road net- roads. From a regional standpoint, the central region of the work in Chaco were rehabilitated, those people would still country has the highest rate of access, followed by the east- be disconnected. While access to all-season roads has a ern and southern regions. significant impact on household wellbeing, roads bring little benefits in areas without markets to sustain nonagricultural jobs (Rozenberg and Fay, 2019). This may be the case for Figure 9: Road Density by Department Chaco with its small, scattered population. Paving rural roads here would not only be expensive but be also unlikely Road Density (km/ 100 km2) to guarantee connectivity with the rest of the network. 0 10 20 30 40 50 60 70 80 Asuncion Figure 8: Population and Road Distribution Guaira Centro Cordillera Central Paraguari Alto Parana Canindeyu Este Caazapa Caaguazu Itapua Sur Misiones Neembucu San Pedro Norte Amambay Concepcion Presidente Hayes Chaco Boqueron Alto Paraguay Total Source: WorldPop data combined with MOPC roads data. Source: WorldPop data combined with MOPC roads data. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 27 44. In Paraguay, rural accessibility is estimated at about 45. Access to the road network in rural areas differs 42 percent, meaning that about 60 percent of the rural significantly across subregions because of the geograph- population, or about 1.8 million rural Paraguayans, are ically skewed distribution of paved roads and differences in left disconnected from a reliable (paved) road network. population and road density. Although systematic data is This number is calculated on the assumption that paved not available, it is presumed that most of the 60,000 kms of roads are all-season (although some of them are in fair or rural roads are in very poor condition. While the majority of poor condition). In terms of comparison, Paraguay’s Rural the population has access to the paved road network, mov- Access Indicator (RAI) of 42.4 percent falls in the same cate- ing toward universal access, as envisioned in the Sustainable gory as Sub-Saharan African countries, all of which are below Development Goals, especially in remote areas, remains 51 percent (Rozenberg and Fay, 2019). Accessibility is particu- a challenge. Improving earth roads to provide all-weather larly limited in the Chaco region, where both population den- access requires a judicious review of the technologies avail- sity and road density are low. In Boquerón and Presidente able to improve such roads. The traffic threshold at which Hayes (departments of the Chaco region), the access rates paving an unsealed road is economically justified is mainly are estimated at 10.6 percent and 11.7 percent, respectively. a function of the unit cost of the paving works, the traffic By contrast, accessibility is higher in the central, southern, composition and its expected growth, and the condition of and eastern regions, where about half the rural residents the unsealed road. In a wet environment, an unsealed road have good access to the paved road network. deteriorates rapidly due to rainfall. In a dry environment, an unsealed road can remain in passable condition for much longer periods and is a more economical option. Box 1: The Impact of Rural Roads Higher Accessibility Means Lower Poverty. “Lack of access has an impact at the most basic level of living: if there is poor access to health services, people will remain unhealthy; children will die; and any epidemic will be likely to have catastrophic results; if there is poor access to clean water, again health will suffer; if there is poor access to basic information the household will be unaware of ideas and tech- nology that might help them to improve their level of living; and if there is poor access to education, children will probably share in the future the limitations that confront their parents today. If access to markets is difficult, farmers are unlikely to diversify their production to include cash crops, or even to grow net surpluses of subsistence staples. And without such new ideas and opportunities poverty remains an endemic feature of rural life” (Edmonds 1998). In a similar vein, the World Bank’s 2008 Transport Business Strategy notes that isolation caused by poor transport is often the main impediment to economic progress and poverty reduction. 28 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS Data from India, China, Thailand, Uganda and Tanzania, analysed by the International Food Policy Research Institute (IFPRI) shows that public spending on rural roads (especially in economically lagging areas) contributes significantly to lifting people out of poverty. In India, for example, for every 1 million rupees spent on rural roads, 124 poor people could be lifted above the poverty line and every rupee invested in rural roads generates more than 5 rupees in returns from agricultural production, the second-largest production growth effect after research and development. Similar results were seen in Thailand and Uganda where the number of poor people declined by 126 and 33, respectively, per million bhat and million shillings made in road investments (Fan et al. 2007). The research concluded that good road networks lower the costs of everything: rural electrification, irrigation, fertilizers, education and health services, agricultural extension services, financial services, output markets, and a host of other goods and services, all of which produce dynamic linkages to new opportunities for migration, investment, and trade. Cook et al. (2005) likewise show that improvements in rural roads and transport services lead to increased farm incomes and off-farm employment opportunities, improved access to health care, education services and common property resources, as well as improved personal security and community participation. In the case of Thailand, it was seen that as the length of paved roads increases, income levels of rural households rise by 3.5%. This research stressed the importance of time savings for the poor, with road improvements expanding the radius of off-farm employment opportunities accessible to the poor. Based on empirical evidence from Burkina Faso, Cameroon, and Uganda, Raballand et al. (2010) postulate a rural transport trap based on inadequate rural road infrastructure in low-income countries. Poor road infrastructure leads to insufficient access to services, and in turn to limited economic opportunities and low value-added activities, which decrease the public investment possibilities in road infrastructure (see graphic). This circular trap could be trans- formed into a virtuous circle by providing connectivity to roads to meet the needs of the rural population. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 29 D. INSTITUTIONAL ROLES AND RESPONSIBILITIES 46. The Ministry of Finance (MOF) oversees the formu- 48. The MOPC is responsible for policy development, lation and management of the budgetary process, the and the management and implementation of all pub- debt and fiscal policy, and national economic policy. lic works and assets related to transport, energy, MOF programs, formulates, executes, controls and eval- mines, and national monuments. Of the four vice-minis- uates the General Budget, which includes the budget of tries within MOPC, the transport sector accounts for two: the Ministry of Public Works (MOPC) and its departments. roads fall under the Vice-Minister for Public Works and MOPC’s overall budget includes the budget for the trans- Communications, while water, air, and urban transport port sector, including the budget for roads. rest with the Vice-Minister for Transport. Figure 10 shows the organization chart for the MOPC. Unlike in some other 47. Management of the road sector is primarily the regional countries, there is neither a separate ministry for responsibility of the MOPC, while departmental and transport nor a decentralized road agency in charge of the municipal governments play a limited role. The legal execution of capital and maintenance road works. Instead, framework for the sector is provided by Law 5552/16 which MOPC manages most of the construction and maintenance focuses on the classification of the road network and other works centrally, while using its departmental units for rural decrees related to safety, signaling, and other aspects of roads maintenance. The construction, rehabilitation, and road transport. The private sector’s role is limited to invest- emergency maintenance of national and departmental ment and management of toll roads. Despite the approval roads is contracted out. Maintenance of rural roads is by of a public-private partnership (PPP) law in 2013 (Law No. force account, although some share of maintenance is also 5016/2013) “ Promoción de la Inversión en Infraestructura contracted out. Pública y Ampliación y Mejoramiento de los Biens y Servicios a cargo del Estado,” private investments in the sector are stagnant and the country has not yet concluded a PPP oper- ation in the road sector. Figure 10: MOPC Organizational Structure for the Road Sector Minister Strategic Projects ANTSV Directorate National Road Safety Agency Viceministry of Viceministry of Viceministry of Viceministry of Public Works and Management and Transport Mines and Energy Communications Finances Rural Roads Roads Directorate Public Credit Finances Directorate (Dirección (Dirección de Vialidad) Directorate Directorate de Caminos Vialidad) Roads Planning UEO MOPC – BID Human Resources Economic Planning Directorate (IDB Projects) Directorate Directorate (Dirección de Vialidad) Source: MOPC. 30 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 49. Within MOPC, all paved roads are the responsibili- carry most of Paraguay’s economic activities, and more than ty of the Directorate of Roads (Dirección de Vialidad, 60,000 km of unpaved and widely dispersed rural roads. To DV) while all other roads, irrespective of network type, ensure general mobility of people and goods in the country, are administered by the Directorate of Rural Roads the priority for Paraguay should be to maintain the paved (Dirección de Caminos Vecinales, DCV). Development of road network which comprises most of the national roads the road sector works program and the budget are under- and a share of the departmental roads, and provides access taken by the Dirección de Planificación which is also charged for about 70 percent of its population. As the paved network with overall sector planning and policy development. Local is the main source of reliable access for the majority of the governments are responsible for rural and departmental population, it is critical that this network is appropriately roads under the government’s decentralization agenda. maintained to provide continued access. But, in practice, weak technical capacity and small budgets appear to have limited the role of subnational governments 52. Since 2012 the government has focused on the main- in the sector. tenance of paved roads and expansion of the paved net- work. The impact on quality has been uneven, with national 50. Departmental units of the Directorate of Rural roads showing improvements while departmental roads Roads carry out maintenance on roads within their indicate a deterioration in quality, with an overall decline in jurisdiction. Units receive funds for gas and spare parts the asset value of the network. Some of the decline in qual- for maintenance equipment from DCV. The allocation of ity may reflect a poor choice of technology in meeting the the funds among departments is based on historical data targets of the paving program, and probably poor oversight from recent years and the amount of the road-maintenance of the civil works. The cost of technical inefficiency alone is equipment available in each department. Each district then estimated at about US$160 million for the five-year period develops a monthly maintenance plan taking into consid- 2012-2017. eration the claims of municipalities, agricultural producers, and the general population. Local authorities may contrib- 53. The rural road network, although less critical from ute to the prioritization of activities in this monthly mainte- an economic standpoint, is important for social equity nance plan. and rural development. The spatial distribution of paved roads shows significant disparity across departments which is not corrected by the presence of earth roads. As a result, CONCLUSIONS access to all-season roads is 70 percent universally, but only 42 percent for the rural populations. The Transport Master 51. Paraguay’s primary, secondary, and rural road net- Plan does not currently include a rural roads policy and works are key to its continued economic growth and to strategy. It is important to develop the same and commit to the achievement of the country’s development goals. a time frame to meet the SDG goal of universal access. This The government has to manage two very different networks document should also provide guidance on the prioritiza- to move forward: approximately 7,000 km of paved and tion and planning of rural road investments. 12,000 km of unpaved primary and secondary networks that MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 31 CHAPTER 3. ROAD SECTOR EXPENDITURES, NEEDS AND FUNDING SOURCES 32 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS A. BUDGET OVERVIEW 54. The Government of Paraguay has devoted consider- 55. Increasing resources to the sector have been ac- able public resources to the road sector over the last companied by considerable volatility in the annual ap- decade, notwithstanding fluctuations in MOPC’s total proved and modified budgets. Since 2012, year-to-year and road sector budget (see Figure 11). The MOPC’s share budgets have seen large fluctuations which tend to under- in the overall budget increased from 4.8 to 6.8 percent be- mine effective planning, project selection, and implementa- tween 2008 and 2017, reflecting the government’s focus on tion. The reliability of the budget process has been shifting the quality and quantity of infrastructure, especially road in- as deviations have fluctuated between 6 and 130 percent frastructure. During this time, MOPC’s approved budget av- (see Figure 13): the 2014 modified budget of US$818 mil- eraged about US$560 million per year, of which about 70% lion was 130 percent higher than the approved budget. In was allocated, on average, to road sector maintenance and the face of such large fluctuations, planning has been con- development (Figure 11). In real terms, road sector spending servative, and budgets consistently have been underspent in 2017 was double that in 2008. As a share of GDP, the road by 25 to 35 percent. The absence of multi-year budgeting, sector has received 2 percent or more since 2012, which combined with insufficient institutional capacity, has led to is well above the regional average, and surpassed only by underspent budgets year upon year. The lumpy multi-year Bolivia (Figure 12) which has been spending an increasing nature of road contracts does not allow large sums of mon- share of GDP (higher than 2 percent), since 2008 on roads. ey to be absorbed without advance planning. Paraguay’s road sector spending as a share of GDP mirrors Eastern European countries’ as they attempted to rapidly improve their infrastructure to Western European stand- ards, around 2009. OECD average expenditures meanwhile are considerably lower at around 1 percent and reflect a more mature stage of infrastructure development.9 Figure 11: MOPC Total and Roads Budget, Approved 1,000 100% 84% 82% 75% 77% Constant 2017 USD 800 75% 80% 65% 68% 66% 600 54% 60% 52% 400 40% 200 20% 0 0% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 MOPC Roads Budget MOPC Total Budget %-share of Roads in Total MOPC Approved Budget Source: BOOST data (2008-2017) Note: BOOST data is at a “Program” level 9 Source: OECD (2013) MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 33 Figure 12: Public Sector Investment in Roads, Select Countries and LAC (% of GDP) 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 Paraguay Argentina Bolivia Uruguay LAC Average (Selected Countries) Source: : Infralatam, (2019). Note: LAC Average includes Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Mexico, Paraguay, Peru, and Uruguay. Figure 13: Approved, Modified, and Executed Budget for the Road Sector, 2008–17 900 800 700 Million USD (2017 Constant) 600 500 400 300 200 100 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Executed Budget Initial Budget Modified Budget Source: : BOOST data (2008-2017) 34 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 56. Execution ratios for capital and current budgets 57. From a geographic point of view, allocation of the are low and fluctuating. Between 2015 and 2018 (Figures roads budget favors the eastern, southern and central 14 and 15), execution ratios for the modified budgets for regions, taking into account the distribution and quali- capital and current spending oscillated between 65 and 75 ty of the road network (see Figures 16–18). While the cen- percent. The correlation between the modified budget and tral and eastern regions account for about 48 percent of the executed capital expenditures is low at 0.50. Execution of total paved road network they received about 65 percent current expenditure, in contrast, shows a surprisingly high of the total budget allocated to road projects in 2018. By correlation with the modified budget of 0.93, and improving contrast, the Chaco region has 14 percent of the paved road execution rates over time, notwithstanding the decline in network and received only 3.2 percent of the total road-sec- absolute amounts allocated for maintenance expenditures. tor budget. The mismatch becomes even more noticeable when roads in good condition are excluded: due perhaps to the small share of budget, about 25 percent of paved roads Figure 14: Approved, Modified, and Executed Budgets for Capital in poor or very poor condition are included. In short, the Expenditures on the National and Departmental Network trend appears to be one of overinvestment in regions where road conditions are relatively good (i.e. the central and 160 500 eastern regions), while roads elsewhere in less favorable 140 400 120 condition Million USD receive inadequate investment10. Of course, the 100 300 actual financial needs of a department must be determined 80 by considering adopted engineering designs, current road 200 60 conditions, 100 procurement costs and other factors, but the 40 20 broad0 trends are clear. 0 2015 2016 2017 2018 2015 2016 2017 2018 Initial Modified Executed Approved Modified Executed Budget Budget Budget Budget Budget Budget Figure 16: Road Length and 2018 Budget Allocation 2500 250 Source: MOPC Department of Finance data, WB staff calculations. 2000 200 Figure 15: Approved, Modified, and Executed Budgets for Current Expenditures on the National and Departmental Network 1500 150 Length of paved roads (km) investment (USD millions) Budget allocates for road 500 1000 100 400 500 50 Million USD 300 200 0 0 100 Centro Este Sur Norte Chaco 0 Paved roads (km) Budget, 2018 (USD mil) 2015 2016 2017 2018 018 Initial Modified Executed Source: Roads and Expenditure data from MOPC, BOOST. ed Budget Budget Budget t Source: MOPC Department of Finance data, WB staff calculations. 10 As of 2018, MOPC groups departments into five regions for budgeting purposes. The departmental composition of regions is shown below. Chaco Norte Este Centro Sur Presidente Hayes Concepción Caaguazú Cordillera Itapúa Alto Paraguay San Pedro Caazapá Guairá Misiones Boquerón Amambay Alto Paraná Paraguarí Ñeembucú Canindeyú Central MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 35 Figure 17: Regional Distribution of Roads and Budget Figure 18: Road Length by Road Condition (Paved roads) 100% 3.2 Length of paved roads (km) 14.7 10.2 90% 0 100 200 300 400 500 600 700 800 900 25.5 80% 19.2 21.5 Asuncion Length of Road Network (km) 70% 9.2 Chaco Central 60% Norte Centro 17.4 19.8 Cordillera 50% 27.4 Sur Paraguari Este 40% 14.1 23.5 Guaira Centro 30% Alto Parana 20% 31.4 37.7 Caaguazu 25.2 Este 10% Canindeyu 0 Paved roads Poor condition Budget, 2018 Caazapa (km) only (km) (USD mil) Itapua Source: Roads and Expenditure data from MOPC, BOOST. Sur Misiones Neembucu 58. Although the relatively meager expenditures on the San Pedro Norte maintenance of unpaved (mainly rural) roads are more Concepcion evenly allocated across regions than the overall budget- Amambay ed expenditures (Figure 19), the correlation between the force account expenditures and the length of earth roads Presidente Hayes Chaco was an abysmal 0.38 in 2018, indicating an unexplained Boqueron regional variation in the allocation of maintenance expendi- Alto Paraguay tures. These maintenance expenditures include monetary allocations and expenditures on lubricants, fuels, and spare Good or fair Poor parts provided to MOPC departmental units responsible for Source: Roads and Expenditure data from MOPC, BOOST. in-house maintenance activities. Figure 19: Regional Shares of Force Account Expenditures for Maintenance 2018 Proxy for Rural Maintenance Cost of Diesel + Lubricant + Repairs of Rural Roads Directorate 14.60 23.75 26.16 12.15 23.34 Chaco Norte Centro Este Sur Source: 2018 MOPC Budget 36 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS B. REHABILITATION AND MAINTENANCE OF THE ROAD NETWORK – NEEDS VERSUS EXPENDITURES B1. PAVED NETWORK – NATIONAL AND SECONDARY ROADS Table 4: Projected Expenditure for the Paved and Unpaved Network (US$ million) Classification 2018 2019 2020 2021 2022 2023-2030 Paved 80 220 250 430 300 100 (pa)11 Unpaved12 36 40 44 48 53 7113 Total 116 260 294 478 353 Source: Plan Maestro, Update 2018 Note: Projected expenditures decline after 2022 when all road rehabilitation programs are expected to be completed, including those under the CREMA for Route N.9. 59. The Transport Masterplan of 2012 included a Road (3,018 kms of national roads and 2,437 kms of departmental Investment Plan (Plan Inversión Vial), which projected roads), and (ii) Case 2: the 2018 network of 7426 kms (3,395 the rehabilitation and maintenance expenditures nec- km national, 4,031 km departmental, and 229 kms of rural essary to prevent the deterioration of the existing road roads). The estimates for the 2012 network were compared network. Based on this objective, and assuming annual with both the originally planned15 and actual expenditures traffic growth of 3.4 percent, the Plan estimated average an- between 2013-17, while estimates for the 2018 network nual expenditures of US$343 million between 2013-17, and were compared with the (updated) planned expenditures US$380 million between 2018-22. Although the estimates shown in Table 5.16 The analysis utilized data on road net- assumed that there would be no upgrading, 2200 kms of work conditions, traffic levels and composition in 2012 and roads were in fact paved between 2013-17. The 2012 Plan, 2018 provided by MOPC. updated in 2018 for the period 2018-30, included a 3000km paving program targeted at departmental roads. For rural 61. For Case 1, in the optimal scenario, defined as expend- roads, the aim is to improve about 500 kms per year to an iture where the present value of total transport costs (the all-season standard. The updated projections for total ex- sum of road agency plus road user costs) is minimized, penditures on both paved and unpaved roads are shown RONET estimated annual expenditures of US$85 million in Table 8. (maintenance plus rehabilitation) between 2013-17,17 and US$81 million subsequently. The annual expenses de- 60. As part of the PER, different scenarios of expendi- crease over time given that the share of total rehabilitation tures towards the preservation of paved roads were expenditures levels out after the first five years, dropping estimated using the Road Network Evaluation Tools from 24 to 8 percent as the quality of the network improves. (RONET)14 methodology developed by the World Bank. On the other hand, under a “do the minimum” scenario with Expenditure requirements for a 20-year period were de- rehabilitation undertaken only when roads deteriorate to a veloped for (i) Case 1: the 2012 road network of 5455 kms “very poor” condition, the annual expenditure requirements 11 This is an average. In 2029 and 2030 the figure is expected to be higher as periodic maintenance becomes due on roads where rehabilitation was completed by 2022. 12 This includes all unpaved roads. 13 This is an average. It is assumed that the allocation will need to increase by about US$4 million per annum to keep pace with the rising maintenance needs of the upgraded rural network. 14 http://www.ssatp.org/en/page/road-network-evaluation-tools-ronet. 15 Elaboracion de un Plan Inversión Vial, 2012. 16 See Archondo Callao, April 2019. 17 At an average of US$15,562 per km per year MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 37 Table 5: Total Expenditures on the Paved (National/Urban) Road Network (US$ million) 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2008- 2013- 2017 2017 Investment 84.9 139.7 141.9 138.8 162.0 260.2 289.5 265.1 221.0 242.3 194.5 255.6 Maintenance & 24.9 36.5 70.2 82.3 95.7 63.3 79.6 90.8 92.4 75.1 71.1 80.3 Other Total 109.8 176.2 212.1 221.1 257.7 323.5 369.1 355.9 313.4 317.4 265.6 335.9 Investment share 77.3% 79.3% 66.9% 62.8% 62.9% 80.4% 78.4% 74.5% 70.5% 76.3% 72.9% 76.0% Maintenance & 22.7% 20.7% 33.1% 37.2% 37.1% 19.6% 21.6% 25.5% 29.5% 23.7% 27.1% 24.0% Other share Source: BOOST database. Executed Budget. Urban/National (assuming being paved), Road Sector only data. Note: Mix Categories comprising of both investment and maintenance activities have been converted into the Investment and Maintenance & Other categories, assuming 65% to be Investment and 35% to be Maintenance & Other. Other refers to unidentified expenditures. are US$42 million and US$47 million respectively. These ex- reevaluated based on the updated road length, condition penditures, however, are insufficient to arrest deterioration and traffic data for 2018. Under the optimal scenario, annual of the quality of the network as measured by the IRI (from maintenance between 2018-22 is now estimated at US$132 3.4 IRI in 2012 to 4.3 IRI in 2017). As expected, the level of million (an average of US$17,198 per km per year) and at service declines as the roads preservation budget declines. US$116 million subsequently. The respective proportions of By contrast, in the optimal scenario, roads in good and fair expenditure on rehabilitation are 21 percent and 18 percent, condition improve from 66 to 85 percent in 2017, keeping due to the need to rehabilitate the new roads added to the the network with an average roughness of 3.4 IRI over time. network prior to 2018. Under a “do the minimum” scenario, the annual expenditure requirements are US$50 million and 62. The RONET estimates for preservation of the paved US$94 million respectively, and the average network rough- network, in the optimal scenario, are about one quarter of ness increases from 4.1 IRI in 2018 to 5.3 IRI in 202318. the expenditures estimated in the Master Plan. Executed maintenance expenses averaged US$80 million per year 64. RONET estimates for maintenance and rehabilita- between 2013 and 2017 and are within five percent of the tion under Case 2 are about 50 percent less for the first estimates obtained from RONET. It is concluded that mainte- five years than the Masterplan Update, and about 30 nance expenditures from 2013 to 2017 were sufficient to cov- percent higher for the later years as the share of reha- er the preservation needs of the paved network (see Table 6). bilitation spending increases over the years. The US$85 million budgeted for maintenance in 2018 falls far short of 63. Case 2: Evaluation of the 2018 network. The 2018 amount indicated for the optimal scenario under RONET. update of the paved network shows an increase of about 40 Given that the level of service is a choice variable it may percent between 2013 and 2017 on the network length. In not be necessary to spend the optimal amount, regardless this period the length of the national roads was unchanged, of the fact that lower expenditure on maintenance implies but the size of the departmental network increased. In terms more rapid decline in road conditions and a shorter time ho- of condition, the national profile improved while that of de- rizon before sections of the network require rehabilitation. partmental roads worsened, on average, from 2013 to 2017. RONET calculations indicate that every dollar of underfund- The network IRI does not show much change. Maintenance ed maintenance implies a cost of US$3.5 to the economy. and rehabilitation expenditures for 2018 and beyond were 18 Archondo-Callao, R. March 2019. 38 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS Box 2: How Useful Is the Rigid Pavement Law (Given Some Unintended Consequences)? In 2017, Congress approved Law 5841/2017, which requires the MOPC to allocate a percentage of the funding for roads to “rigid pavements.” An increasing share of the annual budget, 15 percent for the first year, 20 percent for the second year, and 30 percent for the third year, is mandated for expenditure on paved roads. While the objective of the law is to foster the local cement industry, a requirement for purchase of cement would tend to distort the technology selection in the road sector. In its current form, the law simply encourages more paving irrespective of the need for such roads. Applying a thin layer of asphalt on existing cobbled roads may well be the result of this Law. The continued implementation of this law could lead to highly inflated prices as demand for cement increases beyond production capacity, with adverse consequences for the public exchequer. 65. At a global level, the executed budget for the main- 66. The ratio of the actual asset value to replacement tenance of paved roads between 2013-17 appears to value is a measure of the quality of the paved network, have been more than adequate (see Table 6). However, and both actual and replacement asset values were calcu- the decline over time in the sustainable network of both lated for the years 2012 and 2018. The decline in this ratio primary and secondary networks (more pronounced in the from 89 to 85 percent over the period 2012-18 points to a case of the latter), raises questions about the planning and net depreciation of the asset base, that is a decline in qual- quality of maintenance works. The data on road conditions ity over this period. Since the executed expenditures were shows that the primary network from 2012 to 2018 retained around the estimated optimal expenditures, the decrease in an average roughness of around 3.0 IRI although the share quality suggests that there is room to improve the efficien- of roads in very good condition increased from 22 to 43 cy of maintenance expenditures. By this measure, US$168 percent. Primary roads would therefore appear to have million could have been saved between 2013-17. The dif- received optimal-level maintenance, and possibly higher ference in the actual road asset value and the value which than was necessary. The length of paved, secondary roads would have maintained the ratio of actual to maximum increased substantially from 2012 to 201819 (a 65 percent value at 89 percent simply by benefiting from technical increase) due to paving programs. On the secondary net- improvements in the planning and execution of civil works. work, the “new” roads appear to be included in the ‘poor” condition category due to a common approach consisting of applying a thin seal of paving to gravel roads. Table 6: Estimated Value of the Road Asset Base 2012 2018 Paved Roads (in km) 5455 7655 Maximum Value (US$ mn) 3255 4510 Actual Value (US$mn) 2888 3846 Actual/Maximum (%) 89% 85% Actual Asset Value/GDP (%) 10.5% 14% Source: MOPC and World Bank estimates. 19 In 2012 RONET evaluated 2,437 km of secondary roads, while in 2018 the secondary roads increased to 4,031 km. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 39 B2. UNPAVED NETWORK – RURAL ROADS 67. In general, the benefits of rural roads maintenance Table 7: Importance to Population of Rural Earth Road Network (km) are related to their social importance rather than to traffic levels, given that most rural roads are low-vol- Department Earth Earth Earth Total ume. For this reason and since there are no traffic-related (High) (Medium) (Low) data for the rural network, rural roads were subdivided Alto Paraguay 23 2,124 3,608 3,631 based on the population served by them. The population Alto Parana 509 1,005 6,346 9,213 data was obtained from GIS databases, with the network subdivided into three ranges of beneficiary population Amambay 31 825 1,927 (Table 8). Population classes were used to allocate all rural Asuncion 689 roads to three levels of importance: (i) low: serving a pop- Boqueron 1,808 3,476 4,166 ulation of less than 250 persons (61 percent); (ii) medium: serving a population of between 250 and 1,500 persons (32 Caaguazu 518 1,137 803 3,426 percent); and (iii) high: population higher than 1,500 per- Caazapa 191 1,553 1,259 2,598 sons (7 percent). This was then used to assign the roads to Canindeyu 111 54 3,555 5,360 different maintenance regimes. Central 64 987 52 338 68. To estimate the maintenance requirements for the Concepcion 97 685 1,483 2,684 rural network, in addition to traffic levels or, in this case, Cordillera 361 859 606 1,807 according to the population served, we need to know the road surface types and climatic conditions. Table 9 Guaira 271 3,123 911 2,123 shows the distribution of rural roads per surface type. There Itapua 745 638 5,354 9,517 are 256 km of asphalt concrete roads, 873 km of stone pave- Misiones 91 365 695 1,433 ment roads, 756 km of gravel and stone pavement roads, and 59,227 km of earth roads. The great majority of the Ñeembucu 25 1,036 894 1,286 rural roads are earth roads, totaling 97 percent of the rural Paraguari 278 566 978 2,382 roads in the inventory. The estimation of the maintenance Presidente Hayes 18 2,102 3,308 3,897 needs of rural roads in Paraguay focuses on earth roads, on which there is relatively little information regarding their San Pedro 560 18,733 2,447 5,332 width, condition, and traffic. Total 3,894 36,601 61,122 69. As for climate, we use rainfall as a proxy, dividing Source: MOPC and World Bank estimates. rural roads into areas with three different rainfall levels: (i) a semi-arid climate with rainfall between 30 and 70 millime- ters (mm) per month (15 percent); (ii) a semi-humid climate Table 8: Rural Roads Network by Surface Type (km) with rainfall between 70 and 130 mm per month (42 per- cent); and a humid climate with rainfall higher than 130 mm Asphalt Concrete 256 per month (45 percent). (See Appendix A). Stone 873 Gravel/Stone 765 Earth 59,227 Total 61,122 Source: 2018 GIS Roads Database. 40 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 70. Based on this information, unpaved roads were US$47.4 million for grading and US$8.20 million for annu- then assigned to categories based on the level of rain- al periodic maintenance. In sum, an average of US$61.7 fall and population served, and their maintenance re- million is needed for the maintenance of rural roads every quirements calculated. This enabled us to estimate the year (see Appendix B). Figure 20 summarizes the actual ex- minimum requirements for: (i) annual routine maintenance penditures versus estimated needs, by Department. Further (e.g., cleaning, drainage maintenance, and spot repairs; details are broken down in Figures 21-23. San Pedro depart- (ii) annual grading of gravel and earth roads; and (iii) an- ment received the highest allocation, followed by Caaguazu, nual periodic maintenance of paved, stone and gravel, and Neembucu, Itapua, and the Central departments. Whereas cobblestone roads. The resulting annual needs for routine the allocation for San Pedro is appropriate, this is not the maintenance are US$6.0 million for routine maintenance, case for the Chaco region, which has an even greater need for rural road maintenance since there are hardly any paved roads in that region. Figure 20: Rural Road Capital Needs and Expenditures (actual) Financial needs for rural road development (USD million) Figure 21: Rural Roads Maintenance by Force Account 0 200 400 600 800 1000 1200 Asuncion Guaira Centro Cordillera Central Paraguari Alto Parana Canindeyu Este Caazapa Caaguazu Itapua Sur Misiones Neembucu San Pedro Norte Concepcion Source: WORLDPOP Database and MOPC GIS data on roads. Amambay Presidente Hayes Chaco Boqueron Alto Paraguay 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Budget spending on Diesel + Lubricant + Repairs (USD million) Investment Maintenance Spending Source: MOPC data MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 41 Figure 22: Estimated Rural Road Capital Needs Figure 23: Estimated Rural Road Maintenance Needs per Year Source: WORLDPOP Database and MOPC GIS data on roads. Source: WORLDPOP Database and MOPC GIS data on roads. 71. Table 9 below presents the expenditures on the share of investments. The average expenditures for mainte- Rural Road Network from 2008 to 2017. The expenditures nance and “other” from 2013 to 2017 were US$28.3 million are shown as (i) Investments including all the improvement per year. This total represents, at best, less than half the and rehabilitation of rural roads (including bridges and required maintenance, and reveals a substantial financing culverts); (ii) maintenance and other expenditures. The “oth- gap for maintaining Paraguay’s rural roads network. er” expenditures category might also include an unknown Table 9: Total Expenditures on the Rural (unpaved) Road Network (US$ million) 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2008- 2013- 2017 2017 Investment 30.2 40.5 19.7 10.4 28.5 39.8 80.4 82.9 99.2 117.2 54.9 83.9 Maintenance & 0.5 0.7 2.2 6.9 14.8 19.0 20.8 26.6 32.2 42.7 16.6 28.3 Other Total 30.8 41.2 21.8 17.3 43.3 58.9 101.3 109.5 131.4 159.9 71.5 112.2 Investment share 98.3% 98.2% 90.1% 60.0% 65.8% 67.7% 79.4% 75.7% 75.5% 73.3% 78.4% 74.3% Maintenance & 1.7% 1.8% 9.9% 40.0% 34.2% 32.3% 20.6% 24.3% 24.5% 26.7% 21.6% 25.7% Other share Source: BOOST database. Executed Budget. Rural roads (unpaved). Note: Mix Categories comprising of both investment and maintenance activities have been converted into the Investment and Maintenance & Other categories, assuming 65% to be Investment and 35% to be Maintenance & Other. Other refers to unidentified expenditures. 42 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 72. The above calculations for maintenance require- roads may need to be reconstructed and formalized. This is ments are premised on preserving the rural network estimated to cost a further US$2 billion. Once reconstructed and would not impact the rural access indicator of 42 and rehabilitated, these roads would need maintenance. percent. If, however, the government chooses to go beyond Assuming that rural roads are reconstructed to gravel 42 percent, and perhaps target universal accessibility, signif- standards, about US$300 million per year would be needed icant resources would be required, especially if initial recon- for periodic and routine maintenance. struction or improvement investment is envisaged. Capital investment would be needed to (i) rehabilitate the existing 74. The geographic heterogeneity regarding network rural road network to a maintainable level, and (ii) construct size and population distribution is reflected in the finan- some parts of it as needed. While access to all-season roads cial expenditures needed for universal access in all the presumes paved roads, given the limited traffic levels, reha- regions. The southern region, where the road network is bilitating from earth to a gravel standard is considered more already dense and well-developed, needs to focus resources appropriate in this case. We estimate that a move towards on rehabilitating the existing road network (Figures 24 and universal access with gravel roads would require about 25). The Chaco region, on the other hand, should consider US$5 billion20, assuming key rural roads are improved. rationalizing its network by using GIS tools to know the lo- cation of hamlets, roads, and facilities prior to committing 73. Even after the current network is rehabilitated and resources for rehabilitating the existing network and reclas- upgraded, about 600,000 people would be left without sifying unofficial roads. A pragmatic approach to connectiv- access to an all-season road. As noted in the previous ity should be considered, taking into account smaller roads chapter, the roads network provides access to about 90 designed for bikes or motorcycles. In the central region, percent of the total population. To move toward universal rural road development calls for few financial needs. accessibility, about 26,000 kms of currently unclassified Figure 24: Rural Road Access (0 to 1) Figure 25: Financial Needs for Rural Road Development Source: World Bank calculations. Data from Anuario Estadistico del Paraguay, 2016, and MOPC GIS Data. 20 Assuming 12,500 kms are upgraded over 20 years at a cost of 400k per km, on average. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 43 C. PRIORITIZATION OF RURAL ROAD INVESTMENTS: AN EXAMPLE 75. While the current rural road network is preserved, Figure 26: Poverty Rates by Department some priority roads may need to be rehabilitated and 0% 10% 20% 30% 40% 50% 60% improved. To prioritize rural road investments, a variety of economic and social issues need to be considered. Balancing Asuncion 13% economic efficiency and social equity, which may not always Concepcion 47% be compatible, is a difficult task. There is a substantial ge- San Pedro 47% ographical variation in Paraguay between the Chaco and Cordillera 26% Oriental (eastern) regions. While in Chaco access rates are Guaira 38% low and need improvement, the majority of economic activi- Caaguazu 48% ties and population are concentrated in the Oriental region. Caazapa 55% Itapua 32% 76. Poverty and equity are of concern in terms of road connectivity. The Chaco region has a particularly high pov- Misiones 38% erty rate, while the Central region around the nation’s cap- Paraguari 38% ital is generally much richer than others and concentrates Alto Parana 27% the country’s main economic activities. Unsurprisingly, Central 16% there is a clear correlation (although not causality) between Neembucu 34% road access and poverty incidence: poor people tend to lack Amambay 20% access to the road network (see Figures 26 and 27). While Canindeyu 35% poverty and equity cannot be entirely blamed on the lack Presidente Hayes 16% of road connectivity (many other constraints exist in those areas) good road accessibility is nevertheless a necessary Extreme poverty Non-extreme poverty condition for strong growth. Source: Anuario Estadístico del Paraguay, 2016, and MOPC GIS Data. Figure 27: Poverty and Road Accessibility 60 50 Caazapa Caaguazu Alto Paraguay Concepcion 40 San Pedro Paraguari Poverty rate Canindeyu Itapua Guaira 30 Cordillera Presidente Hayes Misiones 20 Neembucu Central Boqueron Alto Parana Amambay 10 Asuncion 0 20 40 60 80 100 Access to paved road (%) Source: Anuario Estadístico del Paraguay, 2016, and MOPC GIS Data. 44 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 77. From the investment efficiency point of view, the population has access within one hour. Allowing even 4 priority should be to focus on underlying economic hours traveling time, only around 30 percent of the popula- growth potential. Paraguay has various good economic tion would be able to access a health facility, reflecting the opportunities, such as in hydropower and agriculture, in- lack of road network connectivity (see Figures 31 and 32). cluding livestock-raising. While crop production is focused To improve people’s access to healthcare services etc., road on the eastern and southern regions of the country, the cat- accessibility needs to be significantly improved, especially in tle-raising areas extend to the poorer north and Chaco (see the rural areas of the north. Figures 28 and 29). In order to connect farmers and animal breeders in such areas to the national and regional markets, 79. Against this background, a possible prioritization road connectivity and accessibility are essential. framework (“multi-criteria analysis”), could include poverty, road access, economic potential, and social 78. As for access to healthcare and social services, an accessibility. More resources need to be allocated to the estimated 58 percent of Paraguay’s total population has places where poverty incidence is higher, road accessibility access to a health facility within 1 hour (according to is lower, economic potential is higher, and access to social available open data (Figure 30)).21 As in the case of farm- facilities is more limited. For simplicity, these five criteria are ing, there is substantial regional heterogeneity, with the equally weighted here, with different departments prior- Chaco lagging behind in terms of access to health facilities. itized for different reasons. In the Alto Paraguay department, a mere 4.6 percent of the Figure 28: Distribution of Crop Production Areas Figure 29: Spatial Distribution of Cattle Source: IFPRI SPAM. Source: Gridded Livestock of the World. 21 The analysis relies on road network data from the Government and open source data, such as OpenStreetMap. Although the latter is not comprehensive or official, the data coverage is considerable. According to the assessment, 935 hospitals, health clinics, and pharmacies are located. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 45 Figure 30: Health Access and Road Accessibility Figure 31: People’s Access to Health Care, by Length of Trip and Share of Departmental Population 0% 20% 40% 60% 80% 100% Alto Paraguay Alto Parana Amambay < 1 hour Asuncion Boqueron 1 - 2 hours Caaguazu Caazapa 2 - 4 hours Canindeyu Central > 4 hours Concepcion Cordillera Guaira Itapua Misiones Neembucu Paraguari Presidente Hayes San Pedro Source: IFPRI. 80. A sustainable rural road development program Figure 32: Health Access and Road Accessibility needs to be established to meet the existing and sub- Share of population with <1 hour access to health facilities stantial demand for rural road development, while bal- ancing a variety of economic and social development 100% Central objectives. The prioritization framework should be dis- 90% Asuncion Alto Parana cussed, customized, and agreed with all stakeholders before 80% its formal implementation. The framework will also need to 70% Itapua be consistent with the Government’s development strate- 60% Paraguari gy. Roads for investment could also be selected according 50% 40% Neembucu to the concept of fixing priorities based on socioeconomic Caaguazu Cordillera criteria, but including more detailed data on the roads situ- 30% Caazapa San Pedro Presidente Guaira ation. (see Appendix B). The prioritization framework can be 20% Boqueron Hayes Misiones 10% Canindeyu updated and revised with newly available data and through 0% Alto Paraguay Amambay Concepcion program monitoring and evaluation processes. The formula 0 20 40 60 80 100 should however be applied systematically in order to avoid a situation where developed rural roads are fragmented Access to paved road (%) and disconnected, thereby defeating the aim of enhancing Source: IFPRI network accessibility. Rural roads should then be developed to achieve the agreed objectives for the network. 46 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS D. SOURCES OF REVENUES FOR ROADS 81. Road sector development and maintenance is fi- 83. Departmental and municipal governments receive nanced through a complex mechanism involving multi- some limited budgets for infrastructure although the ple domestic and external financing sources, channeled bulk of road sector expenditures are channeled through through the budget. The resources for the road sector the MOPC. These budgets are all merged and do not allow a consist of central government funds (the central budget), disaggregation of the expenditures that correspond to road transfers from the National Fund for Public investment investments. As for rural road maintenance, municipalities and Development (FONACIDE) to MOPC, and resources contribute mainly by collaborating with MOPC in the main- from FONPLATA (a multilateral fund set up by Argentina, tenance planning rather than by providing funding. Some Bolivia, Brazil, Paraguay and Uruguay), FOCEM (Fondo para municipalities have partnership agreements with MOPC to la Convergencia Estructural de MERCOSUR) and multilateral lead maintenance activities with funding provided mainly by or bilateral agencies such as IBRD, IDB, JICA, CAF and the MOPC. Republic of China. The central budget is administered ex- clusively by Paraguay’s central government, while funding 84. Development partners played a major role in from other sources is also channeled through the central Paraguay’s road sector investment up to 2012, when government but administered separately. sovereign bonds became the main source for develop- ment investment. Source 20, composed of multilateral 82. Road works receive funds from three main sources and bilateral funds, and sovereign debt, has been a key categorized as Sources 10, 20, and 30. Source 10 includes supporter of road investment projects since 2008 (Figure funding from taxes and other government revenues; 34). Capital investments in the sector have increased sub- Source 20, international and domestic credit; and stantially since 2012, but while the share of multi-lateral Source 30 from grant funds (Figure 33). While Source 10 financial institutions (MFIs) has increased in absolute terms, supports mainly administrative expenses and wage costs, sovereign bonds (part of Source 20, Figure 35) are the domi- it also includes some expenditures for maintenance in the nant source of capital financing in the roads sector. The cost road sector. The other two categories support capital and of the capital roads program investment should therefore current expenditures. A given project may receive funding include the cost of servicing the bonds. Among multilater- from one or more sources. The MOPC does not summarize al donor partners, the Inter-American Development Bank allocations across resources, by capital, current and admin- takes the lead in supporting the sector, with Spain as the istrative expenses, nor does it compile this breakdown by main bilateral lender. type of road network. Figure 33: Funding Sources for the Road Sector SOURCE 10 SOURCE 20 SOURCE 30 TREASURY TREASURY INSTITUTIONAL RESOURCES (GENUINOS) BONDS (TOLLS) FONACIDE SOVEREIGN BONDS GRANTS YACYRETA IFIs FOCEM LAW SIVIPAR MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 47 Figure 34: Capital (Committed) Budget for the Road Sector by Type of Source 600 500 Constant 2017 Million USD 400 300 200 100 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source 10: Treasury Source 20: Public Credit Source 30: Institutional Source: BOOST database Figure 35: Source 20 Capital Road Sector Executed Budget by Subtype of Source 500 450 400 Constant 2017 Million USD 350 300 250 200 150 100 50 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Opep (Petroleum Exporters) IDB BIRF FONPLATA CAF Spanish Government Jica Jbic Sovereign bonds Bonds Source: BOOST database 48 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 85. MOPC expenditures through the budget are fund- 87. Reliance on external funding sources is not a viable ed by a variety of domestic and external resources. Toll long-term strategy for the upkeep of the road network. revenue is also raised on traffic using part of the primary User-based charges offer a more sustainable, long-term and secondary network. External sources of funding are solution. The country already has an infrastructure fund, predominantly from multilateral official sources, amounting the Sistema de Infraestructura Vial del Paraguay (SIVIPAR), to about 85 to 90 percent of total funding. Of this amount, created in 2003. This was designed to have the following around one-half to two-thirds has typically been provided by revenue sources: (i) 10 percent of the Itaipú and Yacyretá the International Bank for Reconstruction and Development tolls; (ii) 50 percent of the “patente fiscal annual” tax; (iii) 10 and the Inter-American Development Bank. On the bilater- percent of the fuel tax; and (iv) 40 percent of the vehicle im- al side, the principal sources of assistance are from Spain, port tax. Although approved in 2003, the law has not been Taiwan, and Japan. implemented due to the lack of a regulatory decree, and the SIVIPAR fund was never set up. The government is in 86. Funding from external sources has typically been the process of reviewing the law with a view to either im- higher than from domestic sources, but this trend start- plementing or scrapping it. Road funds tend to reduce fiscal ed to change from 2013 onwards. There have been some flexibility and are often resisted by the Ministry of Finance major shifts in recent years away from dedicated line item even though they could, in certain circumstances, improve funds and toward bond issuance. Much of this funding has asset maintenance. Box 2 below offers some thoughts on been focused on works. Direct treasury funding has con- road funds as an interim step and the features of an effec- tinued, but is smaller in relation to the above-mentioned tive road fund. sources. Such funding is likely to be directed toward MOPC operational costs. Box 3: Road Funds as an Interim Step Gwilliam and Shalizi (1999) argue that insufficient or uncertain budgetary allocations to road maintenance result in road deterioration that significantly increases production and transport costs in many countries. To avoid this prob- lem, road professionals advocate the establishment of dedicated road funds managed by independent road boards consisting of user representatives. The road boards would have the power to determine both the level of charges for road use and the level of expenditure on road maintenance. By contrast, macroeconomists and public finance spe- cialists tend to resist the establishment of dedicated road funds, arguing that road funds reduce fiscal flexibility, do not adequately address problems associated with the provision of public goods or the internalization of externalities, and are often not well managed. In general, there are two long-term institutional options for reconciling fiscal prudence with asset maintenance: a road agency that is operated commercially (subject to the normal oversight accorded to privatized monopolies), or a reformed and well-functioning budget process. Gwilliam et al take the view that road funds must be viewed as a provisional, case-specific intermediate step toward one of these long-term solutions. If a country decides to proceed with an interim Road Fund, it should seek to address the following conditions. First, the Road Fund’s expenditure responsibilities should be limited to maintenance in order to correct a systematic bias against maintenance, despite the link between investment and maintenance. Second, the Fund’s revenues should come only from direct charges on road users, except in the case of fuel surcharges (which need to be separated from general taxes in agreement with the treasury). Where feasible, weight-distance charges should be introduced for trucks. Third, the Road Fund should have professional management under the direction of a user-representative board and be subject to strict oversight and auditing arrangements by third parties. Fourth, the Road Fund requires explicit transition arrangements as it moves toward a long-term solution. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 49 88. Paraguay has a history of tolling on some of its pri- A weight-distance toll on trucks would also avoid penalizing mary roads. The importance of tolling as a means of man- low-income users of tolled roads. Indeed, it is possible to de- aging traffic and generating revenue has expanded since the sign a toll system including, for example, discounted passes completion in the early 2000s of a major toll modernization to local traffic as a way of addressing issues of equity with program. This expanded the number of toll stations to 13 respect to the poor. and the length of road covered to 1,842.5 km. New technol- ogy was introduced through this program to automate and 90. Fuel taxes do not appear to be availed to their full improve secure revenue collection. Furthermore, a new tariff potential. The price of a liter of fuel in Paraguay is US$1.04 structure was introduced in June 2014. In June 2016 a policy compared to the world average of US$1.11, and lower than was introduced to charge vehicles traveling in both directions. in neighboring countries such as Brazil and Uruguay where However, toll charges remain on the low side for five differ- the unit price is US$ 1.09 and US$ 1.4722, respectively. It is ent vehicle classes ranging from US$2 equivalent to US$8 noted that in 2017 revenues from fuel taxes amounted to equivalent. In 2017, the revenue generated from 29.5 million a total of US$3.5 million, considerably lower than toll reve- journeys was equivalent to US$12.2 million. A high- level nues and suggesting room for a review and increase of such comparison of tolls, in US cents per km, shows Paraguay at taxes. While fuel taxes are a workable proxy for road usage the lower end of the range at 0.033, compared with Uruguay and externalities such as air pollution, they are not appro- at 0.061 or Brazil at 0.059 though this may reflect to some priate for other externalities like congestion (which varies extent Paraguay’s lower GDP per capita (less than half of by time and location) or safety. Given the multiple claims on Uruguay and sixty-fiver percent of Brazil). Revenue collection the surcharge, the allocation of revenues generated by the can be improved through modifications of the tariff structure surcharge should be calibrated rather than being directed to better align tariffs with the differential impacts of heavier entirely to road administration. freight transport versus lighter passenger vehicles on roads. Revenues are fairly distributed in this way and should help 91. In addition to these tolls, revenue is also gener- the country to progress towards achieving the levels of up- ated on concessioned roads managed by Rutas del Este keep required by the core road network. (Route No.2) and Tapa Pora (Route No. 7), covering 163.5 km and 502 km respectively. The Tapé Pora concession has 89. Aside from tolls, weight-distance charges have little its own tariff schedule, varying by location and number of conflict with fiscal objectives. Road use generates sev- lanes from a low tariff of US$2 equivalent to a high of US$20 eral externalities, of which road damage—requiring main- equivalent. Revenue generation from Tapé Pora appears tenance—is only one. A surcharge on fuel use provides a to be significantly higher than that of any other tolled sec- reasonable proxy for the road damage externality caused tion of the network, generating revenue in 2017 of US$50 by automobiles and implies a modest tax on gasoline. Fuel million equivalent (about double the 2013 take), from 6.6 use, however, is not as good a proxy for road damage by million journeys. It is worth noting that road concessions in trucks, which account for the bulk of road damage. The im- Paraguay are still at an early stage and expansion of such pact of a heavy truck is 10,000 times as great as that of a contracts depends on the extent to which confidence in the medium auto (Newbery 1987). It also varies nonlinearly with regulations governing private contracts increases. As reve- axle weight. MOPC could collect a weight-distance charge nue increases to meet core network maintenance and other from truckers and channel the revenue towards mainte- requirements, public resources should be freed up to focus nance activities, with or without a road fund, taking care to more on supporting the rural network, for which similar us- manage administration costs and vulnerabilities to evasion. er-fee-based concessions are not feasible. 22 https://www.globalpetrolprices.com/Paraguay/gasoline_prices/ 50 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS CONCLUSIONS 92. Despite considerable volatility, the road sector 95. In terms of budget allocation, the Government’s em- budget has expanded substantially since 2012. Budget phasis on the operations and maintenance of the paved volatility is seen in budget allocations across years and in network seems to have improved road conditions over the intra-year changes reflected in the approved, modified the past five years. Paved roads should continue to be the and executed budgets. The uncertainty of the annual final focus of the maintenance policy. However, any allocation of budget undermines effective planning, project selection and funding to the upgrade or reconstruction of secondary and implementation. The absence of multi-year budgeting, com- rural roads needs to take into account the underinvestment bined with insufficient technical and managerial capacity, experienced by some provinces, especially in the poorer has led to consistent underspending of the sector budget, north and Chaco regions with the lowest levels of rural ac- ranging between 25 and 35 percent. When budgets are not cess. Failure to address the discrepancies in rural access in executed in the way they were promised, there is generally these regions, combined with their remoteness and difficul- less confidence that the government is able to manage pub- ties in access to public services and markets, will inevitably lic finances effectively. Given that underspending could also perpetuate inequality. mean off-budget spending, there is a declining confidence in institutions and budget credibility. 96. Rural access is currently estimated at about 42 per- cent, far below the SDG goal of universal access. It is 93. Underspending, as well as overruns, is generally estimated that a move to universal access with gravel roads related to insufficiencies in budget preparation and would require about US$5 billion23, assuming all rural roads program preparation at MOPC and MOF. Development are improved. To reach the 60,000 people who currently do expenditures are difficult to plan accurately, but adequate not live within 2 kms of even a rural road would requires flexibility to reallocate funds from projects that are delayed another approximately US$ 2 billion. to projects that are proceeding well could allow satisfac- tory implementation of the overall expenditure program. 97. The MOPC can enhance the value of the Transport MOPC does not currently appear to have the flexibility to Master Plan of 2018, and its road sector strategy and make such changes. It is possible that cash-flow budgeting investment plan, by retaining the results of the asset is a means for the MOF to manage a development budget management modelling exercise. Asset management that it has not prepared and may explain the low execution models, using economic analysis, have been applied for rates which undermine budget credibility. When analyzing assessment of the paved road network and strategic plan- budget execution, it is important to consider both the risks ning of road investments, but the extent to which model of disruptive repetitive budgeting and the requirements for recommendations have been carried forward is not clear. cash control and compliance. The results of the modelling exercise have been modified with the application of socioeconomic criteria prior to final- 94. We find that the road sector budget is underspent izing the paving program. When the returns to investment to the tune of US$200-400 million annually. These inef- are so distorted, decisions to invest or not to invest, or what ficiencies may be attributed to the institutional “policy en- scale of investment to pursue, are also adversely affected. vironment” which determines what is allocated and what Changes to the modelling results are likely to have substan- is actually spent. Assuming that about half the amount tially reduced the allocative and spatial efficiency with which left on the table was applicable to maintenance activities, public resources are expended irrespective of any budget the cost of institutional inefficiency for the road sector in constraints. Disregarding the results of the asset manage- Paraguay is roughly between US$350-700 million every ment system has also had an important, adverse impact on year. Furthermore, the annual, modified budget of the the technical efficiency of expenditures. sector is more than sufficient to address the maintenance needs of rural roads that are currently underfunded by about 50 percent. 23 For simplicity, rehabilitation costs are assumed to be US$80,000 per km. Periodic and routine maintenance costs are also assumed to be US$10,000 for every 6 years and US$1,800 per km for every year. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 51 98. It is recommended that annual budget requests be aligned with the Medium-term Expenditure Framework (MTEF) and demand better preparation of technical and procurement documents. Currently, the link between the MTEF and annual budgets is tenuous at best and does not allow the linkages between expenditures and longer-term outcomes to be traced. Furthermore, project preparation processes and cycles are opaque and show considerable bunching over the calendar year. Increasing numbers of complaints from bidders and a large number of addenda in contracts signal weak document preparation linked to low technical and procurement capacity and efficiency. These issues can be addressed by requiring MOPC to start prepar- ing specifications and procurement documents as soon as a project enters the budget process. 52 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS CHAPTER 4. GOVERNMENT ROAD SECTOR PRACTICES AND MANAGEMENT (FUNDING, PROCUREMENT, BUDGET EXECUTION) MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 53 A. OVERVIEW OF PRACTICES 99. The Government recognizes the significance of an Uruguay 0.33 percent, and Brazil 0.12 percent in 2017, while adequate and efficient transport sector and has accord- Bolivia allocated 2.87 percent. Although the overall sector ed priority to infrastructure spending for the sector budget appears sufficient for road preservation needs, since 2012. Public expenditures on transport in Paraguay planning and utilization of the available resources leave increased from 1.04 percent of GDP in 2010 to about 2 per- room for improvements. There are potential inefficiencies cent in recent years, which is much higher than spending by in road maintenance expenditures in the paved road net- other countries in the region. For example, Argentina allo- work and financial requirements to achieve universal access cated an average of 0.58 percent of GDP to road transport, in rural and remote areas. B. MANAGEMENT SHORTCOMINGS: WEAK PLANNING AND LOW EXPENDITURE EFFICIENCY 100. The unpredictability of budget approvals contrib- allocated to a program and establish a link with longer- utes to significant spending inefficiencies. The substan- term, strategic objectives. Detailed data broken down by tial budget revisions from the initial allocations provide function are also in short supply. Poor expenditure report- incentives to divide projects into smaller entities and to ing practices signal uncertainty over total spending in the prioritize shovel-ready over higher-priority investments. For sector, and further complicates planning development of larger, multiyear projects, funding unpredictability can lead the sector. According to the MOPC, streamlining the budget to implementation delays and higher costs, both in terms of preparation and approval process, together with better capital costs (since delays increase the construction costs), record-keeping, could be important ways to improve the and in terms of recurrent costs, since the supervision and reliability and efficiency of spending in the roads sector. implementation support budgets that represent a significant part of total recurrent costs also increase. Budget unpre- 102. The Fiscal Responsibility Law No. 5098 mandates the dictability also weakens the project prioritization process by preparation of multiyear plans and budgets but it does reducing the time for implementing projects within a given not appear to be effectively utilized. It is unclear whether year, since budget modifications are only approved in mid- multiyear budgeting is applied at all, since both MOPC and year, thus creating an incentive to allocate money dispropor- the Ministry of Finance have confirmed that annual budgets tionally to shovel-ready, potentially low-impact projects. are based mostly, although not entirely, on historical data. 101. The same unpredictability highlights inefficiencies 103. The Transport Sector Master Plan, with a detailed in the central government’s budget processes and con- investment plan for the medium and long term, requires tributes to low budget execution ratios. The complexity significant strengthening to provide a strategic direc- of the budget approval process has a bearing on the predict- tion to the sector.24 For the road sector, the 2018 update of ability of the road sector budget. The budget classification the Plan could be improved by emphasizing the importance system does not facilitate the budgeting and monitoring of of a transparently derived, prioritized investment and main- expenditures by programs, apart from those funded by de- tenance program. It would also be enhanced by giving more velopment partners, and it is difficult to monitor spending attention to (i) ensuring adequate maintenance to minimize 24 A master plan provides a predictable and disciplined framework for multi-year budgeting which can ease the process of year- to-year budgeting, keeping a country’s overall development objectives in mind. Typically, such a plan sets out institutional priorities, and provides demand forecasts for the different modes under various economic growth scenarios, with alternative investment strategies. Typical sections in such a plan can include a situational analysis along with a gap analysis, a strategic framework, and an actual implementation plan to meet the objectives identified. 54 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS total road user costs and improve road safety; and (ii) en- 106. Experience with road administrations around the suring the sustainability of road programs by prioritizing in- world indicates the importance of separating policy and vestments in line with available resources, while expanding implementation functions. As we have seen in Chapter 2, the toll-road network and introducing user charges to in- MOPC has responsibility for both functions - policy formula- crease revenue sustainability. The lack of a well-defined and tion and its implementation, and there is no meaningful sep- articulated master plan means that there is no systematic aration between the two. There is also an almost complete framework for budget allocations for network development absence of mechanisms and tools to track performance and and maintenance. hold MOPC accountable for the same. 104. Budget execution can be improved by delegating 107. Accountability mechanisms are needed in order greater responsibilities to MOPC technical depart- to promote technical efficiency. An MTEF is the institu- ments. The MOPC Roads Department is responsible for tional mechanism that allows sector ministries to allocate most investment and maintenance projects. It comprises resources in line with their strategic priorities and within functional divisions (execution and conservation) and some the aggregate resource envelope presented by the Ministry units according to funding source – JICA, CAF, FONPLATA, of Finance. It helps to address information asymmetry be- GMANS, FOCEM. The implementation unit for World Bank tween central and sector ministries: sectors have better projects is separate from Vialidad and has considerably information for making disaggregated expenditure alloca- more autonomy than those within it. Technical staff have tions than Ministry of Finance. For this approach to produce very little autonomy, and ministerial and vice-ministerial the desired results, line ministries need to have incentives approvals are required at each step of the procurement and to use resources in a technically efficient manner. In other implementation processes. This creates serious bottlenecks words, they must be held accountable for the allocation and and impacts overall execution rates, exacerbating an already use of budgeted resources. Accountability mechanisms in- poor situation with respect to MOPC’s technical capacity. clude inter alia financial accountability and audits, value for money audits, ex post evaluations, and annual reports. 105. A further issue to be addressed is the strength or weakness of a country’s institutional system. In a best- case scenario, politicians and policy makers negotiate the prioritization of particular investments, while quasi-inde- pendent or independent regulatory institutions may con- tribute with checks and balances. However, the results of such a process can be distorted by rent-seeking behavior.25 It is important to understand the political, economic, and so- cial processes that promote or block pro-poor investments, and the role of oversight, rules, and regulations that permit more effective and politically feasible investments. 25 Rent-seeking behavior occurs in situations in which some interest groups manage to capture key state institutions or processes and pursue their particu- lar interests rather than those of society as a whole. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 55 C. PROCUREMENT AND VALUE FOR MONEY IN PUBLIC EXPENDITURES 108. The Dirección Nacional de Contrataciones Púbicas Figure 36: Number of Bids Invited, by month, 2015-18 (DNCP) has pioneered the adoption of open-contracting 25 data standards for procurement in all sectors, including the transport sector. But procurement still remains highly 20 inefficient. Law 2051/03 and its regulations establish the 15 functions and roles of all the units involved in procurement 10 and is implemented by DNCP. All the entities, including the MOPC, have a Unidad Operativa de Contratación (UOC) re- 5 sponsible for ensuring compliance with the UOC procure- 0 ment procedures and requirements. UOC also coordinates January February March April May June July August September October November December the finalization of bidding documents and their approval by the DNCP. Despite the country’s commitment to open con- tracting data standards, public procurement in Paraguay 2015 2016 2017 2018 experiences low levels of competition for government con- Source: Dirección Nacional de Contrataciones Públicas, 2019 tracts, procurement bunching, and poor contract prepara- tion, which generates successive complaints – all of which Figure 37: Number of Bid Invitations Closed, by month, 2015-18 point to the need to make value for money improvements in public contracting, including the need to boost low staff 25 capacities in the MOPC and sector ministries.26 20 109. Bunching of bidding processes (see Figures 36 and 15 37) and a high number of addendas per contract (Figure 10 38) indicate weak technical capacity within MOPC for bid 5 preparation and process management. There are no re- strictions on when a procurement process is launched, and 0 January February March April May June July August September October November December DNCP allows publication of bid documents prior to budget approval. A more even flow of new procurement starts would allow better preparation of technical documents, and 2015 2016 2017 2018 fewer addenda as a result. These enhancements could be achieved through improved work planning and monitoring Source: Dirección Nacional de Contrataciones Públicas, 2019 within MOPC. 110. Regression analysis was undertaken to test the Figure 38: Average number of Clarifications and Addendas, 2015-18 impact of seasonality on the contract award price. The 6 data tested did not show any significant impact: the award 5 price correlates with the estimated price and the number 4 of bidders. Competition measured by the number of bids was found to be higher for smaller contracts rather than 3 for higher-value ones. As final costs per kilometer of works 2 relate to the level of deterioration of project roads, a shorter 1 project cycle from identification to implementation is ex- 0 pected to yield lower costs per kilometer for maintenance 2015 2016 2017 2018 and development. Average number Average number of Clarifications of Addendas Source: Dirección Nacional de Contrataciones Públicas, 2019 26 World Bank. 2018 56 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 111. MOPC’s capacity for planning and management 114. It is recommended that the MOF simultaneously of the road sector is likely constrained by the 25 per- introduce ad hoc financial and physical audits of road cent staff reduction between 2013 and 2018, possibly sector expenditures to help improve discipline and in- implemented to counter overemployment at the Ministry. crease value for money in the sector. The execution of In 2017-2018, MOPC employed a total of 110 engineers of public investments in Paraguay is still highly complex and all disciplines (water, civil, etc.), and has recently hired and bureaucratic. While several of the process steps are impor- trained an additional 115 engineers. DCV currently employs tant to protect the quality of expenditures, a major overhaul 900 staff, DV has 330, UEP has 80, and DPV 14 staff. A staff- of the investment cycle would help to simplify and optimize ing and salary review, followed by the implementation of execution. Given that monitoring of public expenditures is recommendations for staffing and training, would go a long the MOF’s responsibility, it is imperative that the MOF de- way to resolving capacity issues at MOPC. velops both the tools and technical capacity to review the budget proposed by the MOPC, and subsequently confirm that expenditures conform to the agreed budget. CONCLUSIONS 115. A functional review of MOPC’s staffing aimed at 112. It is recommended that annual budget requests be identifying the number and profile of staff needed for aligned with the Medium-term Expenditure Framework optimal performance is strongly recommended. Many of (MTEF) and that technical and procurement documents the changes entailed by the recommendations above re- should be better prepared. Currently, the link between the quire an appropriate size and skills mix of the technical staff MTEF and annual budgets is at best tenuous and does not at MOPC. A functional review indicating gaps and a prior- allow the linkages between expenditures and longer-term itized staffing plan should be undertaken as a priority next outcomes to be traced. Furthermore, project preparation step to enhance sector efficiency. processes and cycles are opaque and show considerable bunching over the calendar year. Increasing numbers of complaints from bidders and a large number of addenda in contracts signal weak document preparation linked to low technical and procurement capacity and efficiency. These issues can be addressed by requiring MOPC to start prepar- ing specifications and procurement documents as soon as a project enters the budget process. 113. The Paraguayan context requires particular em- phasis on reforms that promote transparency and ac- countability since there is no separation between pol- icy development and road sector management – both functions being housed in MOPC. There are practically no mechanisms or tools in place to monitor and evaluate road sector performance and encourage efficient expenditure, either by the Ministry of Finance or another autonomous or semi-autonomous entity. An important first step towards accountability is for MOPC to commission and publish annu- al, audited reports on the income and expenditures of the road sector, to be followed by audits of physical progress and quality in subsequent years. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 57 CHAPTER 5. FINDINGS AND RECOMMENDATIONS 58 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 116. This PER presents several findings and recommendations to advance the management and administration of Paraguay’s roads sector in ways that tangibly improve the network while strengthening government capacity and sustaina- ble budgeting. These are organized under four main themes: • Better budget management and governance practices; • Striving for universal accessibility in rural areas; and • Addressing data and capacity issues; • Sustainable sources of revenue for the sector. A. BETTER BUDGET MANAGEMENT AND GOVERNANCE PRACTICES 117. While there is no blueprint for enhancing public 119. It is recommended that the MOF simultaneously sector efficiency, experience with road administrations introduces ad hoc financial and physical audits of road around the world reveals the importance of separating sector expenditures to help improve financial discipline policy and implementation functions of road admin- and to increase value for money in the sector. The execu- istration. Experience also shows that, for public sector tion of public investments in Paraguay is still highly complex agencies “transparency and clear disclosure policies, and and bureaucratic. While several of the process steps are im- a clear mechanism for evaluating performance, are asso- portant to protect expenditure quality, a major overhaul of ciated with high levels of performance, with performance the investment cycle would help simplify and optimize exe- orientation and professional management being the most cution. Given that public expenditure monitoring is an MOF important contributors.”27 The Ministry of Public Works, responsibility, it is crucial for this ministry to develop the however, is charged with both road administration policy tools and technical capacity to review the budget proposed and implementation functions. Furthermore, Paraguay has by the MOPC, and subsequently confirm that expenditures virtually no mechanisms or tools in place to monitor and correspond to the agreed budget. evaluate road sector performance, and to encourage effi- cient expenditure by the Ministry of Finance or any other 120. The MOPC can enhance the value of the Transport autonomous or semi-autonomous entity. Master Plan of 2018, and its road sector strategy and investment plan, by upholding the results of the asset 118. An important first step in this direction is for management modelling exercise. Asset management MOPC to publish annual, audited reports on the income models using economic analysis have been applied for as- and expenditures of the road sector, followed by phys- sessing the paved road network and for strategic planning ical progress and quality audits in subsequent years. of road investments, but it is not clear to what extent the The Paraguayan context requires particular emphasis to be model recommendations have been carried forward. The placed on reforms that promote transparency and account- results of the modelling exercise have been modified with ability, since policy development is not separate from road the application of socioeconomic criteria prior to finalizing sector management. Paraguay needs to consider long term the paving program. When the returns on investment are institutional options for reconciling fiscal prudence with so distorted, decisions to invest or not to invest, or what asset maintenance (an independent road entity with imple- scale of investment to pursue, are also adversely affected. mentation responsibility or a well- functioning budget pro- Changes to the modelling results are likely to have substan- cess). Meanwhile, sector accountability could be enhanced tially reduced the allocative and spatial efficiency with which by introducing independent annual audits of all capital and public resources are expended, irrespective of any budget recurrent expenditures and sharing these results in a trans- constraints. Setting aside results of the asset management parent manner. system also have an important, adverse impact on the tech- nical efficiency of expenditures. 26 World Bank. 2018 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 59 121. It is recommended that annual budget requests be 122. In terms of budget allocation, the Government’s aligned with the Medium-term Expenditure Framework emphasis on the operations and maintenance of the (MTEF) and that the preparation of technical and pro- paved network seems to have led to improvements in curement documents be improved. Currently, the link road conditions over the past five years, and should con- between the MTEF and annual budgets is at best tenuous tinue to be the focus. At the same time, funding allocations and does not allow the linkages between expenditures to the upgrading or reconstruction of secondary and rural and longer-term outcomes to be traced. Furthermore, roads need to consider the underinvestment experienced in project preparation processes and cycles are opaque and certain provinces, especially in the poorer rural north and show considerable bunching over the calendar year. Large Chaco regions which have the lowest levels of accessibili- amounts of addenda and increasing complaints from bid- ty. Failure to adequately address issues of rural access to ders signal weak document preparation and low technical public services and markets in these regions will inevitably and procurement capacity and efficiency. These issues can perpetuate inequality. be addressed by requiring MOPC to start preparing speci- fications and procurement documents as soon as a project enters the budget process. B. ADDRESSING DATA AND CAPACITY ISSUES 123. A systematic, up-to-date survey of the quality of 124. A functional review of road sector staff and pro- the paved network, as well as of the amount of traffic files is needed in order to develop a staffing and train- on it, is necessary for the meaningful application of as- ing plan commensurate with sector responsibilities and set management models and for determining network planned growth. Budget allocations for capital and recur- spending requirements. While maintenance spending for rent expenditures have increased substantially over the the paved network appears to be broadly adequate, there past seven years, challenging MOPC capacity to manage the is concern that the amount, spatial allocation, and type of annual budget and deliver high-quality works. Vialidad, the capital works funded may be inappropriate, especially if department for roads (primary and secondary), has seen expenditure requirements are based on guesswork as op- a substantial increase in its investment budget, while staff posed to hard data regarding traffic and road conditions. levels were reduced by about 25 percent in 2013–18. While Conducting a comprehensive survey of the paved road greater staff productivity can legitimately be expected, there network, developing mechanisms for regular updates of is a significant risk that the sector’s maximum absorptive ca- the roads database, and making this data publicly available pacity (given numbers of personnel, skill levels, responsibil- alongside budget data, are perhaps the best means of im- ities, and work plans) has been reached. A functional review proving value for money in the roads sector. could help bring staffing needs and skills into line with the future needs of the sector. 60 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS C. STRIVING FOR UNIVERSAL ACCESS IN RURAL AREAS 125. Develop a rural roads policy and strategy that roads, including multicriteria analysis and identification of address universal accessibility goals, appropriate tech- funding sources for the development and maintenance of nologies for all-weather access, and the cost and time the existing and upgraded network. frame required to achieve the strategic goals. The cur- rent lack of a roads subsector strategy results in fragment- 126. The rural roads strategy should guide the prioriti- ed efforts and uncoordinated construction are unlikely to zation and planning of rural roads investments for the enhance rural connectivity and could worsen regional ineq- next five to ten years. The rural road network remains a vi- uities. While about 70 percent of the general population of tal public resource for the country, despite the poorer rural the country live within 2 kilometers of a paved road, only 42 areas’ dispersed population and lack of economic weight. percent of the rural population has such access. A sustaina- A rural roads strategy would help to determine investment ble rural road development program is needed to meet the priorities based on transparent and objective socioeconom- huge demand for rural road connectivity. Developing a rural ic criteria that are appropriate in a rural context. This strat- roads policy in consultation with relevant stakeholders – the egy would help improve the efficiency of project selection MOPC and across key sectors (health, education, agricul- and public expenditures, as well as address current inequi- ture) – would also go a long way to support achievement ties in rural people’s access to economic, educational, and of the National Development Plan 2030. The policy should health facilities. include guidance on the methodology for prioritizing rural D. SUSTAINABLE SOURCES OF REVENUE FOR THE SECTOR 127. At present, Paraguay’s road funding mechanism is the two institutional options for the road sector. Paraguay’s to pool capital from various sources, relying heavily on paved network has been receiving adequate maintenance donor resources and more recently on domestic borrow- funds regardless of the uncertainty in the budget allocation ing. In the longer term, road user charges provide a more and transfer process. In this situation, a careful review of sustainable source of funding since these can be aligned the sector’s long-term institutional options and preferences, with the maintenance requirements of the network, espe- including the need for a provisional road fund in the form of cially if they are channeled through a road fund. Paraguay an active SIVIPAR, is recommended as a way forward. This has an infrastructure fund law, the SIVIPAR law adopted in review should also include an assessment of the sources of 2003, which theoretically functions as a road fund to collect revenues, the structures of user charges, the affordability of road user charges to defray the costs of upkeeping the road proposed user charges (specifically tolls and fuel taxes), and network. However, SIVIPAR has not yet been regulated or adequacy of the build-up of revenues toward a sustainable implemented. financial model. 128. Road funds, managed by independent road boards, 129. Expanding the use of tolls and concessions to fund are often recommended in situations of insufficient or maintenance of the core network is recommended. uncertain budgetary allocations to help avoid deteriora- Tolls offer a reliable source of funding to meet road upkeep tion of road assets. However dedicated funds also reduce costs. An increase in tolls to pay for road upkeep and dam- fiscal flexibility and may be just as poorly managed as budg- age needs to ensure a fair distribution of charges between etary allocations. In general, there are two long-term insti- heavier vehicles such as semi-trucks and lighter traffic, such tutional options for reconciling fiscal prudence with asset as cars and pickup trucks. Finally, while road concessions are maintenance: a road agency that is operated commercially, also an important source of private funding and manage- or a reformed and well-functioning budget process with ment for the road sector, and could be expanded, Paraguay adequate checks and balances. A well-designed road fund is first needs to boost the confidence of the private sector in an interim measure while the country moves towards one of the rules and regulations governing concession contracts. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 61 62 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS REFERENCES –– Archondo-Callao (2019). Rural Roads Maintenance Needs Evaluation. April 2019, First Draft, GIS Database –– Cook, C. C., T. Duncan, S. Jitsuchon, A. Sharma, and W. Guobao, (2005). Assessing the Impact of Transportand Energy Infrastructure on Poverty Reduction. Manila: The Asian Development Bank. –– DCV Budgets 2013-2018 by Type of Activity, Directorate of Rural Roads, Ministry of Public Works and Communications of Paraguay –– De Richecour, A. B., & Heggie, I. G. (1995). African Road Funds: What Works and Why? (No. SSATP Working Paper No. 14). Environmentally Sustainable Development Division, Africa Technical Department, the World Bank. –– Dirección Nacional de Contrataciones Públicas (2019). Datos Abiertos DNCP. Available at: https://www.contrataciones. gov.py/datos. Accessed 8 May 2019. –– Dirección Nacional de Contrataciones Públicas, 2019. Available at: https://www.contrataciones.gov.py/datos. Accessed on 8 May 2019 –– Directorate of Rural Roads, Ministry of Public Works and Communications of Paraguay, Databases. Allocation of resources to maintenance districts 2018 –– Edmonds, G. (1998). Wasted time: the price of poor access (Vol. 3). Geneva: ILO. –– Elaboracion de un Plan Inversion Vial, 2012. Ministerio de Obras Públicas y Comunicaciones –– Fan, S., Brzeska, J., & Shields, G. (2007). Investment priorities for economic growth and poverty reduction. Twenty- twenty (2020) focus brief on the world's poor and hungry people/International Food Policy Research Institute (IFPRI). –– Gwilliam, K., & Shalizi, Z. (1999). Road funds, user charges, and taxes. The World Bank Research Observer, 14(2), 159-186 –– Gyamfi, P., Gutierrez, L., & Boscán, G. Y. (1992). Infrastructure maintenance in LAC: the costs of neglect and options for improvement. –– Heggie, I. G., & Vickers, P. (1998). Commercial management and financing of roads. The World Bank. –– Infralatam, (2019). Datos de inversión en infraestructura económica. Available at http://www.infralatam.info/ –– Klaus Schwab, World Economic Forum. The Global Competitiveness Report 2017–2018. 2018. Accessed 8 May 2019. –– Ley Nro. 5.189 - 5.282 - Resumen Anual De Funcionarios. MOPC, 2019, https://www.mopc.gov.py/mopcweb.old/ resumen-anual-de-funcionarios-s334. Accessed 8 May 2019. –– Ministry of Finance of Paraguay. BOOST Database 2008-2017 –– MOPC. Budgets 2008-2018, Sistema Integrado de Administración Financiera (SIAF). Ministry of Public Works and Communications of Paraguay –– MOPC. Nómina Completa de los funcionarios del Ministerio de Obras Públicas y Comunicaciones. Available at: https:// www.mopc.gov.py/mopcweb.old/nomina-completa-de-los-funcionarios-permanentes-y-contratados-s67.Accessed 8 May 2019. –– OECD (2013). Spending on Transport Infrastructure 1995-2011: Trends, Policies, Data. Available at: https://www.itf- oecd.org/sites/default/files/docs/13spendingtrends.pdf –– Paraguay Comision de Equidad, Crecimiento y Desarrollo (2019). Towards the goal of becoming a high-income economy within a generation. Draft March 3, 2019. MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 63 –– PND (2014). Paraguay - Plan Nacional de Desarrollo 2030 –– Potter, M. B. H. (1997). Dedicated road funds: a preliminary view on a World Bank initiative (No. 97). International Monetary Fund. –– Raballand, G., & Macchi, P. (2010). Rural road investment efficiency: Lessons from Burkina Faso, Cameroon, and Uganda. Washington, D.C.: The World Bank. –– Rethinking Infrastructure in LAC, Fay et al, 2017 –– RONET. Road Network Evaluation Tools. Available at: http://www.ssatp.org/en/page/ road-network-evaluation-tools-ronet –– SEDAC (2019). Socioeconomic Data and Applications Center: Population Density in Paraguay in 2010. Available at: http://sedac.ciesin.columbia.edu/gpw/ –– World Bank (2008). Safe, Clean, and Affordable Transport for Development. The World Bank Group’s Transport Business Strategy for 2008–2012, Transport Sector Board. Washington, D.C.: The World Bank. –– World Bank (2016). “Measuring Rural Access: Using New Technologies. –– W orld Bank (2018) Paraguay. Invertir en capital humano: una revisión del gasto público y la gestión en los sectores sociales. –– World Bank (2018). Paraguay Systematic Country Diagnostic, June 27, 2018. –– World Bank (2019). Logistics Performance Index. Washington, D.C.: The World Bank –– World Bank (2019). World Development Indicators. Washington, D.C.: The World Bank. 64 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS APPENDIX A. RURAL ROAD NETWORK DISTRIBUTION BY CLIMATE ZONE Table 11: Network Distribution by Climate Zone Department Semi-Arid (30-70) Semi-Humid (70-130) Humid (130-220) Total Alto Paraguay 3,020 611 3,631 Alto Parana 17 303 8,893 9,213 Amambay 709 1,218 1,927 Asuncion Boqueron 4,013 152 4,166 Caaguazu 53 1,629 1,744 3,426 Caazapa 29 1,305 1,265 2,598 Canindeyu 45 1,300 4,015 5,360 Central 316 21 338 Concepcion 22 2,595 68 2,684 Cordillera 6 1,739 62 1,807 Guaira 883 1,240 2,123 Itapua 14 2,510 6,993 9,517 Misiones 1,104 329 1,433 Ñeembucu 1,221 65 1,286 Paraguari 2,337 45 2,382 Presidente Hayes 1,773 1,741 383 3,897 San Pedro 5,258 74 5,332 Total 8,992 25,715 26,415 61,122 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 65 APPENDIX B. TOTAL COSTS OF RURAL ROAD MAINTENANCE Table 12: Maintenance Total (US$ million per year) Department Asphalt Stone Gravel/ Earth Earth Earth(Low) Total Stone (High) (Average) Alto Paraguay 0.00 0.00 0.00 0.02 0.00 1.62 1.65 Alto Parana 0.01 0.61 0.75 0.88 1.91 2.86 7.03 Amambay 0.00 0.02 0.35 0.05 0.90 0.37 1.70 Asuncion 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Boqueron 0.00 0.00 0.00 0.00 0.34 1.56 1.91 Caaguazu 0.04 0.96 0.66 0.79 1.62 0.36 4.44 Caazapa 0.00 0.06 0.00 0.27 1.01 0.57 1.91 Canindeyu 0.22 0.00 0.62 0.17 1.40 1.60 4.00 Central 0.57 0.52 0.00 0.09 0.05 0.02 1.25 Concepcion 0.09 0.00 0.49 0.13 0.89 0.67 2.26 Cordillera 0.42 0.49 0.06 0.49 0.62 0.27 2.35 Guaira 0.00 0.39 0.09 0.43 0.77 0.41 2.08 Itapua 0.43 1.19 0.26 1.21 2.81 2.41 8.30 Misiones 0.00 0.05 0.00 0.14 0.57 0.31 1.07 Ñeembucu 0.01 0.00 0.00 0.03 0.33 0.40 0.78 Paraguari 0.00 0.51 0.00 0.37 0.93 0.44 2.26 Presidente Hayes 0.00 0.03 0.00 0.02 0.42 1.49 1.96 San Pedro 0.22 0.29 0.66 0.76 1.89 1.10 4.93 Total 2.02 5.12 3.94 5.85 16.48 16.47 49.87 Source: WB staff calculations. 66 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS APPENDIX C. MULTI-CRITERIA ANALYSIS TO IDENTIFY RURAL ROAD PRIORITY AREAS Table 12: Maintenance Total (US$ million per year) Weight (%) 20 20 20 20 20 100 Criteria Economic Development Poverty Road Access to Overall score Access health Department Region Land Number Total Paved Share of Overall Overall Under Cattle Poor Road Population Index Rank Cultivation (thousands) Population Access with <1 hr (Index) (no) (hectares) (%) (%) (%) Alto Paraguay Chaco 415 1,565 46.5 1.6 4.6 53 3 Alto Parana Oriental 7,314 210 21.4 71.9 75.7 45 8 Amambay Oriental 2,715 991 15.2 29.8 6.7 31 15 Asuncion Oriental 0 0 11.6 88.1 81.7 19 18 Boqueron Chaco 814 1,877 21.5 12.9 11.0 42 10 Caaguazu Oriental 7,419 553 43.7 59.1 35.7 54 2 Caazapa Oriental 2,646 343 47.0 30.0 30.6 49 5 Canindeyu Oriental 5,643 737 38.0 41.2 12.3 47 7 Central Oriental 3,450 64 16.2 96.1 93.5 31 14 Concepcion Oriental 3,959 1,210 44.0 45.0 6.3 48 6 Cordillera Oriental 3,003 255 26.9 77.2 32.1 28 17 Guaira Oriental 1,968 157 33.9 61.1 26.5 31 16 Itapua Oriental 8,714 453 33.2 58.3 61.0 57 1 Misiones Oriental 3,121 508 27.5 52.4 14.7 32 13 Ñeembucu Oriental 2,124 607 24.2 37.7 33.5 36 12 Paraguari Oriental 2,029 475 35.8 59.8 55.0 41 11 Presidente Chaco 1,599 2,497 20.0 28.1 10.7 44 9 Hayes San Pedro Oriental 3,113 1,355 43.6 45.1 19.8 50 4 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS 67 68 MORE, BETTER OR DIFFERENT? INVESTING IN PARAGUAY’S ROADS